Systemic Disorders

Chapter

36

Systemic Disorders Lynne G. Maxwell, Salvatore R. Goodwin, Thomas J. Mancuso, Victor C. Baum, Aaron L. Zuckerberg, Philip G. Morgan, Etsuro K. Motoyama, Peter J. Davis, and Kevin J. Sullivan

CONTENTS Endocrine Disorders  1099 l Diabetes Mellitus  1099 l Diabetes Insipidus  1102 l Syndrome of Inappropriate Antidiuretic Hormone Secretion  1104 l Adrenal Insufficiency  1104 l Hypothalamic-Pituitary-Adrenal Axis Suppression Caused by Exogenous Steroid Therapy  1105 l Thyroid Disorders  1107 l Pheochromocytoma  1110 Respiratory Disorders  1112 l Upper Respiratory Tract Infection  1112 l Reactive Airways Disease (Asthma)  1114 l Bronchopulmonary Dysplasia  1120 l Cystic Fibrosis  1122 Cardiovascular Disorders  1125 l Anesthetic Management  1125 l The Child with a Murmur and Possible Heart Disease  1127 l Noncardiac Manifestations of Congenital Heart Disease  1128 l Kawasaki Disease  1129 l Takayasu Arteritis  1130 Hematology and Oncology Issues: Hereditary and Congenital Disorders  1130 l Hemoglobinopathies  1130 l Oncology Issues  1138 Coagulation: Developmental Aspects, Disorders of Coagulation, and Perioperative Management of Hemostasis  1144 l Overview of Hemostasis  1144 l New Model of Cell-Based Coagulation  1145 l Modulators of Coagulation  1146 l Fibrinolysis  1147 l Antiplatelet and Anticoagulant Drugs  1148

1098

Laboratory Evaluation of Coagulation  1148 Developmental Hemostasis  1148 l Laboratory Evaluation of Coagulation in the Newborn  1149 l Developmental Changes Beyond the Neonatal Period  1150 l Developmental Aspects of Platelet Number and Function  1150 l Inherited Coagulopathies  1150 l Platelet Abnormalities  1153 l Acquired Coagulopathies  1154 l Acquired Thrombocytopathy  1155 l Intraoperative Coagulopathies  1156 l Intraoperative Evaluation of the Bleeding Patient  1158 Treatment of the Bleeding Patient  1160 l Safety of Transfusion and Factor Replacement  1160 l Safety of Hemostatic Agents  1160 l Anticoagulant-Induced Coagulopathy  1160 l Agents Used to Control Bleeding  1160 l Hemostatic Formulary  1163 l Complications of Blood Product Administration  1164 Miscellaneous Problems  1165 l Acquired Immunodeficiency Syndrome  1165 l Latex Allergy  1169 l Epidermolysis Bullosa  1170 l Down Syndrome  1172 Genetic Muscle Disorders  1174 l General Overview  1174 l Myasthenic Syndromes  1176 l Myotonias  1177 l Mitochondrial Myopathies  1178 l Muscular Dystrophy  1180 l Other Dystrophies  1181 l Metabolic Diseases  1181 l The Undiagnosed Myopathy  1182 Summary  1182 l l

C h a p t e r 36    Systemic Disorders   1099

patients who present special problems for A ­mong anesthesiologists are children whose underlying condi-

tions complicate anesthetic management and may be associated with an increased risk of morbidity. The number of rare diseases that may be encountered in infants and children is great, although only a few are mentioned here. Chosen for discussion are the diseases most commonly seen, those carrying an increased risk related to anesthetic management, and a few of unusual interest. Modifications to the understanding of mechanisms of coagulation are included, along with consideration of coagulopathic states, and there is a comprehensive review of the anesthetic implications of pediatric syndromes ­associated with genetic, metabolic, and dysmorphic features (Baum and O’Flaherty, 2006). A partial list of syndromes with possible anesthetic implications is included in Appendix D, which can be accessed online at www.expertconsult.com

ENDOCRINE DISORDERS

Diabetes Mellitus The endocrine condition most commonly dealt with in the perioperative period is the management of glucose homeostasis in children with diabetes mellitus. The prevalence of type 1 (­insulin-dependent) diabetes in the United States has remained stable for the past 15 years at 1 in 400 to 600 school-aged children, whereas the incidence of type 2 diabetes is increasing, especially among American Indian, Black, and Hispanic children and adolescents (CDC, 2007c). Diabetes mellitus is the result of an absolute or functional deficiency of insulin production by the pancreas. In type 1 diabetes, this deficiency is caused by an autoimmune pathophysiologic process. Insulin deficiency results in abnormalities of glucose transport and storage and of lipid and protein synthesis. Over time, these metabolic derangements result in the vascular pathology that leads to end-stage complications of renal, cardiac, and eye disease—diseases that typically do not occur before adulthood. The anesthetic implications of type 1 diabetes in children differ from those in adults with the same disease, for whom the primary concern is the type and severity of end-organ disease. Children with type 1 diabetes may be treated with various types of insulin on a daily basis to maintain tight glucose control with the aid of frequent blood-glucose monitoring. Since 1982, most newly approved insulin preparations have been produced using recombinant DNA technology with laboratory-cultivated bacteria or yeast. This process allows the bacteria or yeast cells to produce complete human insulin. Recombinant human insulin has mostly replaced animal-derived insulin (e.g., pork and beef insulin) in diabetes management (Plotnick and Henderson, 1998). Insulin products called insulin analogues are produced so that the structure differs slightly from human insulin (by one or two amino acids) to change onset and peak of action. An example of an analogue is human lispro, an ultra–short-acting insulin that is given only 15 minutes before a meal. Its peak and duration of action parallel the glucose rise that results from carbohydrate ingestion. Another new insulin is glargine, which almost mimics an insulin pump, providing a continuous, 24-hour, low background

TABLE 36-1. Kinetics of Commonly Used Insulins Insulin

Route

Onset (hr)

Peak (hr) Effective Duration (hr)

Human Lispro (Humalog)

SC

0.25

0.5–1.5

  3–4

  Regular

IV/SC

0.5–1.0

  2–3

  3–6

  NPH

SC

2–4

4–10

10–18

Glargine (Lantus)

SC

1

2–3

24

0.5–1

2–5

  4–6

Animal Regular

IV/SC

IV, Intravenous; SC, subcutaneous; NPH, neutral protamine Hagedorn.

level of insulin. The kinetics of some of the insulin preparations most commonly used in children are listed in Table 36-1. Some children’s diabetes may be managed with an external insulin pump, which provides a low, background, subcutaneous infusion of insulin and the ability to give small boluses before meals. Most children with diabetes administer insulin at least three times each day and check their blood sugar at least four times each day. Type 2 diabetes in children and adolescents may be controlled with diet and exercise, but these children also may be taking metformin. Metformin was recently shown to decrease gluconeogenesis by directly modulating copper-­binding protein (CBP) in the liver much as insulin itself does, rather than by overcoming the liver’s decreased sensitivity to insulin (He et al., 2009). Because of the effects of surgical stress on glucose homeostasis, children with type 1 diabetes are at risk for significant perioperative difficulties, even when their preoperative glucose control is good. Brittle or noncompliant patients with diabetes have additional problems, including an increased risk of perioperative hypoglycemia or hyperglycemia, osmotic diuresis with resultant hypovolemia, and altered mental status. The physician must document the child’s current insulin regimen, degree of compliance, preoperative glucose control, and risk of hypoglycemia from preoperative fast. Much of this information can be obtained from the patient’s endocrinologist or by examination of the child’s blood-glucose monitoring log. A recent growth history can indicate how well controlled the child’s diabetes may be. Coordination and cooperation among the patient, parents, pediatrician, endocrinologist, and anesthesiologist are essential if the goal of optimal perioperative glucose homeostasis is to be achieved. The anesthesiologist must particularly heed the recommendations of the child’s primary physician. Insulin is an anabolic hormone that promotes glycogen and triglyceride storage and protein synthesis. Present in small amounts even in the fasting state, it decreases glycogenolysis, gluconeogenesis, and lipolysis, with resultant ketogenesis and protein breakdown. Its complete absence at the time of surgery puts the patient in a state of starvation, in which caloric intake is greatly restricted and substrate demands (e.g., for healing) are at their highest. The risk of a catabolic state is increased by the release of stress hormones, including catecholamines, cortisol, and glucagon. Perioperative insulin administration is essential to control glucose and to promote an anabolic state, which

1100   P a r t  IV    Associated Problems in Pediatric Anesthesia

is most conducive to speedy healing and metabolic homeostasis. Preoperative anesthesia evaluation for elective procedures, informed by contemporaneous endocrine ­assessment of adequacy of glucose control, should be completed 7 to 10 days before the scheduled date of surgery to allow adjustment of treatment regimen or delay of procedure if control is not optimal. Rhodes and colleagues (2005) published a comprehensive review of concerns and perioperative management of pediatric patients with diabetes; it features an extremely useful clinical practice guideline that incorporates both preoperative assessment and choice of preoperative insulin regimen.

TABLE 36-2. Protocols for Perioperative Insulin Therapy Regimen

Morning of Surgery Procedure

Classic regimen

Start IV infusion of 5% dextrose in 0.45% saline or Ringer’s lactate solution at 1500 mL/m2/day. Administer half of usual morning insulin dose as regular insulin. Check blood glucose before induction, during and after anesthesia

Continuous insulin infusion

Start IV infusion of 5% dextrose in 0.45% saline or Ringer’s lactate solution at 1500 mL/m2/day. Add 1-2 units of insulin per 100 mL of 5% dextrose Starting insulin dose = 0.02 units/kg/hr Check blood glucose before induction, and during and after anesthesia

Insulin- and glucose-free regimen (for operative procedures of short duration)

Withhold morning insulin dose If indicated for procedure, give glucosefree solution (e.g., Ringer’s lactate) at maintenance rate. Check blood glucose before induction, and during and after anesthesia.

Preoperative Evaluation The preoperative evaluation should include measurements of the hematocrit, electrolyte levels, and glucose levels. A hemoglobin (Hb) A1c level (i.e., glycosylated Hb assay), although a useful index of long-term glucose control, is unlikely to affect the anesthetic plan and is not a necessary preoperative test (Nathan et al., 1984). If glycohemoglobin results are available, it is important to remember that different laboratories have different ranges for Hb A1c in normal subjects. Even in the same laboratory, the normal range may change from time to time. It is therefore important to know the laboratory’s normal range to interpret results in patients with diabetes. The normal range of Hb A1c is 4.5% to 6.1%, but the normal range also varies with age (Rhodes et al., 2005; Custer and Rau, 2009). Several systemic abnormalities may be present in the child with diabetes. Nineteen percent of children with diabetes have a vital capacity two standard deviations below the predicted mean value, suggesting the presence of restrictive lung disease (Buckingham et al., 1986). No apparent association exists between decreased vital capacity and duration of diabetes or presence of other diabetic complications. Abnormal lung elasticity and thickening of the alveolar basal laminae have been reported in children with diabetes (Schuyler et al., 1976; Vracko et al., 1980). Routine preoperative pulmonary function tests are not indicated in the asymptomatic child who has diabetes. Decreased atlantooccipital joint mobility, resulting in difficult intubation, may be present in a subset of adolescents with a syndrome of diabetes mellitus, short stature, and tightness of small joints of the fingers, wrists, ankles, and elbows (Salzarulo and Taylor, 1986). Abnormal cross-linking of collagen by nonenzymatic glycosylation is the postulated cause of this syndrome (Chang et al., 1980).

Perioperative Management Various regimens for managing insulin therapy perioperati­ vely have been proposed, three of which are discussed in the following sections and in Table 36-2: a classic regimen, the subcutaneous infusion insulin pump, and intravenous (IV) insulin infusion. Essential to optimal management, regardless of regimen, is the scheduling of elective surgery for the child with diabetes as early as possible in the day (first case) to minimize time that the patient must fast. The fasting interval should be the same as that recommended for patients who do not have diabetes: no solid food or milk for 8 hours, and clear liquids are permissible until 2 hours before the scheduled time of surgery (Schreiner et al., 1990). Children with diabetes should be encouraged to continue taking clear liquids until 2 hours before surgery. If this is not possible, an

IV fluid infusion should be started (described later). As recommended in adult patients with type 2 diabetes, metformin should be stopped 48 hours before surgery, based on reports of lactic ­acidosis in patients who remain on the drug and are in a fasting state perioperatively. Other orally administered medications (e.g., thiazolidinediones or sulfonylureas) may be continued through the day before surgery. Although some investigators have recommended the withholding of preoperative sedation from patients with diabetes to better monitor for signs of hypoglycemia, premedication is recommended in children. The use of agents such as benzodiazepines, opioids, or barbiturates does not alter glucose metabolism, and the failure to use such agents may elevate the blood sugar level as a result of anxiety, which causes a stress response with catecholamine release. Based on the blood-glucose level determined on arrival to the preoperative facility and before implementation of the regimens discussed in the following sections, glucose or insulin should be administered according to the scheme outlined in Table 36-3.

TABLE 36-3. Preoperative Glucose and Insulin Management for Diabetic Patients Blood Glucose Level

Management

<80 mg/dL

2 mL/kg D10W followed by glucose infusion

80-250 mg/dL

D5/0.45 NS or D10/0.45 NS solution at maintenance if insulin is to be administered; 0.9 NS if short case; no insulin

>250 mg/dL

Administer rapid-acting (lispro) or shortacting (regular) insulin SC to reduce blood sugar; use correction factor from patient’s endocrine provider or 0.2 unit/kg SC

>350 mg/dL

Consider canceling or postponing surgery, especially if ketonuria

NS, Normal saline; SC, subcutaneously.

C h a p t e r 36    Systemic Disorders   1101

On the morning of surgery, one half of the usual dose of longacting insulin (e.g., Neutral Protamine Hagedorn [NPH]) is administered subcutaneously after establishing an IV infusion of 5% glucose-containing solution at a rate of 100 mg/kg per hour of glucose (Table 36-2). Plasma-glucose concentrations should be maintained between 100 and 180 mg/dL. This target range is chosen because mild to moderate hyperglycemia (without ketosis) usually does not present a serious problem to the child, whereas hypoglycemia has devastating consequences. Hyperglycemia greater than 250 mg/dL should be avoided because of associated mental status changes, diuresis, and subsequent dehydration, which can occur because of the hyperosmolar state. Hyperglycemia has been associated with poorer outcomes in patients at risk for central nervous system (CNS) ischemia, including those undergoing cardiopulmonary bypass (Lanier et al., 1987; Lanier, 1991). Hyperglycemia has also been shown to impair wound healing and has adverse effects on neutrophil function in vitro. (Marhoffer et al., 1992; Delamaire et al., 1997). When the classic regimen is employed, supplemental subcutaneous doses of short-acting insulin can be given on a sliding scale postoperatively to maintain the desired plasma glucose level. This regimen should be restricted to patients who are scheduled for short surgical procedures, after which they are expected to resume eating promptly.

range of 100 to 180 mg/dL. This ­continuous-infusion regimen has been shown to yield better control of glucose concentrations than the regimen in which intermittent subcutaneous insulin is administered (Kaufman et al., 1996). The administration of intermittent large IV-insulin doses has no role, as it can result in large swings in glucose concentration (high and low) and a greater chance of lipolysis and ketogenesis. Patients with insulin pumps should have them turned off in the perioperative period, and pumps should be replaced by the continuous­infusion regimen, as most anesthesiologists are not familiar with the details of operation of such pumps. Fifty-percent dextrose solution should be available for administration in case of the development of hypoglycemia; 0.1  g/kg of dextrose raises the blood-glucose level by approximately 30 mg/dL. The glucose and insulin should be infused through a dedicated IV cannula to enable it to be well regulated apart from the non–glucose-containing crystalloid solutions that are administered to replace blood or fluid losses—especially important if the maintenance glucose solution contains potassium. Many institutions avoid potassium-containing solutions to prevent their inadvertent rapid administration in the setting of rapid administration of fluid for blood or fluid replacement. Most investigators believe that lactated Ringer’s solution should not be used for blood and fluid replacement, because lactate is a glycogenic precursor and may result in higher blood-glucose levels.

Subcutaneous Infusion Insulin Pump

Alternative Procedure (Insulin- and Glucose-free Regimen)

Increasing numbers of pediatric patients with type I diabetes are being managed with an external insulin pump that is capable of subcutaneous administration of both continuous and bolus doses of insulin. Such pumps afford excellent control, with changes in administration coordinated with eating, exercise, and stress. At this time, the proliferation of pumps from multiple manufacturers precludes the easy acquisition of knowledge and familiarity with their use in the perioperative setting. Some clinicians and institutions allow continued use of insulin pumps for short, uncomplicated procedures (e.g., less than 2 hours), whereas most recommend transition to a continuous insulin infusion, as described in the next section (Glister and Vigersky, 2003; Rhodes et al., 2005).

For extremely brief procedures, after which prompt resumption of oral intake is expected, an alternative protocol involves the administration of no insulin or glucose before or during surgery. When oral intake is established postoperatively, 40% to 60% of the usual daily insulin dose is given (Stevens and Roizen, 1987). Myringotomy with tube placement is an example of a procedure for which this regimen would be appropriate. The surgical procedure should be performed as the first case on the morning schedule to avoid prolonged fasting and excessive delay in insulin administration. The most serious perioperative complication that can occur in the diabetic child is hypoglycemia. Common signs of low blood-glucose levels include tachycardia, tearing, diaphoresis, and hypertension. In the anesthetized patient, these signs may be misinterpreted as the result of inadequate anesthesia. Because the clinical signs of hypoglycemia are masked by sedation or anesthesia, frequent (every hour) measurement of the serumglucose level is critical for the prevention of hypoglycemia, independent of the glucose-insulin regimen chosen. Glucose test strips, with or without the use of a reflectance photometer, provide quick, convenient, and reliable bedside blood-sugar measurements to guide therapy. Blood-glucose determinations performed with reflectance photometers provide results that are generally within 10% of clinical laboratory glucose determinations done on the same specimen (Chen et al., 2003). Visual evaluation of blood-glucose strips is less accurate (Arslanian et  al., 1994). Postoperative insulin administration is determined by the time of resumption of oral or enteral feeding and by the postoperative blood-glucose concentration. The endocrinologist and surgeon should be active partners in the choice of an appropriate insulin regimen, because they are responsible for monitoring glucose homeostasis after the patient leaves the recovery room. For day-surgery patients, contingency planning

Classic Regimen

Intravenous Insulin Infusion If a long procedure or a prolonged period of postoperative fasting is anticipated, the continuous IV infusion of glucose and insulin may provide the best control. On the morning of surgery, a glucose infusion is begun at a maintenance rate of 100  mg/kg per hour, with an insulin infusion of 0.02 to 0.05 unit/kg per hour “piggy-backed” into the glucose infusion. The glucose infusion can be D5 or D10 in half-normal saline with 10 to 20 mEq/L of potassium chloride. These infusions should be started 2 hours before surgery to minimize the duration of fasting and decrease the risk of the development of a catabolic state. Insulin is absorbed by IV bags and tubing. When the insulin solution is prepared, the first portion of the solution should be run through the tubing and discarded to saturate the sites in the tubing that bind insulin (Kaufman et al., 1996). Blood glucose levels should be checked hourly for the first few hours, and adjustments of 0.01 unit/kg per hour in the insulin rate should be made to keep the blood sugar in the acceptable

1102   P a r t  IV    Associated Problems in Pediatric Anesthesia

for insulin management and mechanisms for follow-up care and consultation should be clearly defined for members of the care team and family.

Anesthetic Management Regional or general anesthesia is appropriate for the child with diabetes mellitus. If tolerated with minimal sedation, regional anesthesia might be argued to offer the advantage of allowing for observation of the level of consciousness as a monitor of hypoglycemia. Practically speaking, most children require general anesthesia, even when regional techniques are employed. The ease and availability of point-of-care glucose determination from venous or finger-stick specimens obviate the need for monitoring of cerebral function.

Perioperative Management of Diabetic Ketoacidosis Occasionally, patients with diabetes require surgery for trauma or infection while they are in a state of ketoacidosis. Diabetic ketoacidosis includes hyperglycemia (plasma-glucose concen­ tration greater than 300 mg/dL) with glucosuria, ketonemia (ketones strongly positive at greater than 1:2 dilution of serum), ketonuria, and acidemia (pH lower than 7.30, serum bicarbonate lower than 15 mEq/L, or both). It is common for intraabdominal catastrophes with infection (e.g., appendicitis) to precipitate ketoacidosis. Foster and McGarry (1983) have succinctly summarized the pathophysiology of diabetic ketoacidosis. The initiating event is usually cessation of insulin therapy or onset of stress that renders the usual dose of insulin inadequate. Glucagon, catecholamines, cortisol, and growth hormone levels increase. A catabolic state is produced as substrates are mobilized, resulting in hepatic production of glucose and ketone bodies, which causes hyperglycemia and ketoacidosis. Subclinical brain swelling nearly always occurs during diabetic ketoacidosis therapy, although most patients remain asymptomatic (Krane et al., 1985). Fatalities from cerebral edema do occur, and some studies suggest that high rates of fluid administration early in treatment (more than 50 mL/kg in the first 4 hours) greatly increase the risk of herniation (Mahoney et al., 1999). Studies using 4 L/m2 for the first 24 hours followed by 1 to 1.5 times maintenance resulted in clearance of ketoacidosis equal to that in patients who were given more fluid, but a low but persistent incidence (0.35% to 0.5%) of symptomatic cerebral edema remained (Felner and White, 2001). The best methods to prevent the development of this devastating complication are administration of isotonic fluid only and frequent monitoring of serum osmolality (by direct measurement or calculation) to ensure that elevated osmolality is reduced gradually. Insulin therapy should be tailored to decrease the blood glucose concentration at a rate not greater than 100 mg/dL per hour. To prevent a more rapid decrease in blood glucose concentration, 5% dextrose and if necessary, 10% dextrose, should be added to the rehydration solution to slow the rate of fall, rather than decreasing the rate of insulin infusion (Arslanian et al., 1994). Fortunately, the anesthesiologist is rarely called on to administer anesthesia during this severe metabolic derangement. If an anesthetic is required during diabetic ketoacidosis, preoperative attention should be directed toward the correction of hypovolemia and hypokalemia, along with beginning an insulin infusion. Invasive hemodynamic monitoring may be indicated preoperatively to optimize the patient’s fluid and

electrolyte balance and to monitor the patient’s hemodynamic status accurately. Surgery should not be delayed inordinately because it may be impossible to correct the metabolic derangements before the underlying source of infection or organ dysfunction is corrected. For patients with signs of cerebral edema, monitoring of intracranial pressure may be necessary.

Diabetes Insipidus Diabetes insipidus (DI) is a clinical syndrome of hypotonic polyuria in the face of elevated plasma osmolality that results from inadequate production of, or inadequate response to, antidiuretic hormone (ADH). Central DI results from inadequate production or release of ADH from the posterior pituitary gland. ADH is synonymous with arginine vasopressin. Nephrogenic DI (also referred to as vasopressin-resistant DI) is characterized by partial or complete renal tubular unresponsiveness to endogenous ADH or exogenously administered arginine vasopressin. Congenital nephrogenic DI is caused by mutations in either the vasopressin receptor or aquaporin-2 gene. Inheritance is X-linked in the former and autosomal recessive or dominant in the latter (Sasaki, 2004). A combination of hydrochlorothiazide and a nonsteroidal antiinflammatory drug (NSAID) has been effective for the treatment of nephrogenic DI. Both cyclooxygenase-1 (COX-1; i.e., tolmetin and indomethacin) and cyclooxygenase-2 (COX-2; i.e., rofecoxib) drugs have been effective (Jakobsson and Berg, 1994; Pattaragarn and Alon, 2003). Toxic drug effects may also lead to acquired nephrogenic DI. Anesthetic implications of nephrogenic DI have been reviewed by Cramolini (1993) and Malhotra and Roizen (1987). The causes of DI are outlined in Box 36-1. This discussion focuses on central DI and its clinical manifestations, which are

Box 36-1 Causes of Diabetes Insipidus Vasopressin Deficiency (Neurogenic DI) Acquired Idiopathic Traumatic (accidental, surgical) Neoplastic (craniopharyngioma, metastasis, lymphoma) Granulomatous (sarcoid, histiocytosis) Infectious (meningitis, encephalitis) Vascular (Sheehan syndrome, aneurysm) Familial (autosomal dominant) Excessive Water Intake (Primary Polydipsia) Acquired Idiopathic (resetting of the osmostat) Psychogenic Vasopressin Insensitivity (Nephrogenic DI) Acquired Infectious (pyelonephritis) Postobstructive (urethral, ureteral) Vascular (sickle cell disease or trait) Infiltrative (amyloid) Cystic (polycystic disease) Metabolic (hypokalemia, hypercalcemia) Granulomatous (sarcoid) Toxic (lithium, demeclocycline) Solute overload (glucosuria, postobstructive) Familial (X-linked recessive) DI, Diabetes insipidus.

C h a p t e r 36    Systemic Disorders   1103

polyuria and polydipsia. The urine is hypotonic relative to the plasma. The urine osmolality is usually less than 200 mOsm/L, and urine specific gravity is less than 1.005 (Custer and Rau, 2009). When the patient has had inadequate access to water, severe dehydration and hypernatremia ensue, because a large volume of dilute urine is continually produced. Patients with preexisting DI may need incidental surgery. They are usually taking maintenance doses of vasopressin, which for relatively short, uncomplicated, elective procedures, should be continued through the perioperative period (Wise-Faberowski et al., 2004). Desmopressin (1-desamino8-d-arginine vasopressin, or DDAVP), a longer-acting (8 to 20 hours) vasopressin analogue, has a decreased vasopressor effect relative to its antidiuretic effect (Hays, 1990). DDAVP is usually given intranasally (2.5 to 10 mcg once or twice daily) or orally (25 to 200 mcg once or twice daily) to prevent diuresis (Lee et al., 1976). DDAVP also may be given subcutaneously or intravenously (1 to 2 mcg twice daily). An algorithm for the management of DI, whether preexisting or developing intraoperatively or postoperatively, is presented in Figure 36-1. The most common situation encountered by the anesthesiologist, however, is the new onset of DI intraoperatively or postoperatively in patients who are having surgery for pituitary or hypothalamic tumors, most commonly craniopharyngiomas; 70% to 90% of these patients develop DI (Lehrnbecher et al., 1998; Ghirardello et al., 2006;). Perioperative DI may present in one of four ways:

1. Transient polyuria probably is related to the onset and resolution of transient cerebral edema rather than to injury to the pituitary stalk. It usually resolves in 24 to 36 hours. 2. A triphasic pattern with an interval of normal urine output reflects the release of stored vasopressin from the

Pre-existing DI

Minor surgical procedure

Major surgical procedure

Usual AM dose of DDAVP

Hold AM dose of DDAVP

Intraoperative DI

No intraoperative DI

Vasopressin infusion

No postoperative DI

Postoperative DI

Vasopressin infusion n  FIGURE 36-1. Perioperative management of patients with preexisting DI.



posterior lobe or median eminence of the pituitary, followed by resumption of polyuria when the stored supply of vasopressin is exhausted. 3. Mild polyuria reflects partial DI, which is exaggerated by local edema and corticosteroid administration. 4. Permanent DI is caused by destruction or removal of all cells capable of producing and storing vasopressin.

If any degree of DI is going to occur, the onset is most commonly within 18 hours after the operation. Recommendations for therapy are reviewed by Wise-Faberowski and colleagues (2004). The goal of perioperative management of DI is to maintain normal fluid and electrolyte balance, urine output, and hemodynamic stability. Urine output may be prodigious (10 to 20 mL/kg per hour). Care must be taken to differentiate polyuria caused by DI (urine specific gravity of less than 1.005) from diuresis caused by mannitol administration, hyperglycemia (urine specific gravity that is usually greater than 1.015), or simple excessive administration of crystalloid (urine specific gravity of greater than 1.005). Patients with partial ADH deficiency usually do not require supplemental aqueous vasopressin perioperatively, because large quantities of ADH are produced in response to surgical stress (Malhotra and Roizen, 1987). However, serum osmolality should be measured often, and aqueous vasopressin should be given if the plasma osmolality exceeds 300 mOsm/L (Wise-Faberowski et al., 2004). If central DI is present preoperatively and the planned surgery is prolonged, an infusion of aqueous vasopressin is begun preoperatively and continued intraoperatively. The recommendations for adults include a bolus of 100 milliunits of aqueous vasopressin followed by a continuous infusion of 100 to 200 mU/ hr. This should be accompanied by the intraoperative administration of isotonic saline at two thirds of the maintenance rate, with additional fluid given for blood loss replacement and for maintaining hemodynamic stability (Malhotra and Roizen, 1987). Hypotonic fluids should be avoided, as hyponatremia may result. For the pediatric population, an infusion is begun at 0.5 mU/kg per hour and increased in 0.5 mU/kg per hour increments until a urine osmolality twice that of plasma and a urine output of less than 2 mL/kg per hour are achieved. It is rarely necessary to use more than 10 mU/kg per hour (Weigle, 1987). Side effects from vasopressin administration are minimal at doses used for antidiuresis; at larger doses, generalized vasoconstriction can occur and has resulted in tissue ischemia and myocardial infarction. DDAVP, rather than aqueous vasopressin, may also be used for treatment of perioperative DI because of its potent antidiuretic effect with minimal pressor activity or other side effects. In the perioperative period, it may be given intravenously until intranasal administration can be started or resumed. The suggested IV dose is 0.5 to 4 mcg, with a single dose having a duration of action of 8 to 12 hours (Muglia and Majzoub, 2008). It is important to note that this dose of DDAVP is one fortieth to one fourth of that used to prevent bleeding in patients with von Willebrand’s disease (vWD). The ease of intermittent dosing with DDAVP with low incidence of side effects must be balanced against the ability to titrate the continuous vasopressin infusion cited earlier. The long half-life of DDAVP (6 to 24 hours) in combination with intraoperative fluid administration may incur an increased chance of hyponatremia. In either case, careful monitoring of fluid balance is essential.

1104   P a r t  IV    Associated Problems in Pediatric Anesthesia

The anesthesiologist may occasionally encounter children who are receiving nightly nasal DDAVP for the treatment of enuresis. A review of its use reveals a negligible incidence of water intoxication (and no permanent effect on enuresis when treatment is stopped) (van Kerrebroeck, 2002). Given the known duration of action, DDAVP administered the night before outpatient surgery should not affect the urine output on the day of surgery.

Syndrome of Inappropriate Antidiuretic Hormone Secretion Just as central DI is caused by ADH deficiency, syndrome of inappropriate ADH secretion (SIADH) is caused by an excess production of ADH that is inappropriate with respect to the state of the intravascular volume. The most common causes of SIADH are listed in Box 36-2. The hallmark of SIADH is hyponatremia in the face of high urine osmolality and sodium levels. A comparison of the urine and serum electrolyte status seen in DI and SIADH is presented in Table 36-4. The treatment for mild cases of SIADH is fluid restriction (50% to 60% of maintenance fluid requirement) or insensible loss (400 mL/m2 per day), plus one half to three fourths of the urine output. If hyponatremia is severe enough to cause coma or seizures, treatment with hypertonic saline (3%) solution may be indicated, but caution should be employed because the administration of hypertonic saline may cause circulatory overload because the intravascular volume is already increased. A too-rapid rise of osmolarity (more than 20 mOsm/kg or more than 10 mmol/L of sodium in 24 hours) carries a risk of central pontine myelinolysis, a condition that can result in death (Laureno and Karp, 1997). This syndrome is thought to be caused by the sudden shrinkage of brain cells in response to rapidly increasing extracellular osmolality.

Box 36-2 Causes of Syndrome of Inappropriate Antidiuretic Hormone Secretion Central nervous system Infection Meningitis Encephalitis Abscess Guillain-Barré syndrome Neoplastic Tumor Trauma Subarachnoid hemorrhage

TABLE 36-4. Comparison of Diabetes Insipidus and SIADH Laboratory Test

Diabetes Insipidus

SIADH

Urine specific gravity

≤1.005

≥1.005

Urine osmolality

50-200 mOsm/L

>200 mOsm/L

Serum osmolality

>280 mOsm/L

<280 mOsm/L

Serum sodium

High (usually >148 mEq/L)

Low (usually <132 mEq/L)

Urine sodium

<20 mmol/L

>20 mmol/L

Adrenal Insufficiency Adrenal Insufficiency as a Result of Primary Abnormalities of the Hypothalamic-Pituitary-Adrenal Axis Adrenal insufficiency is an uncommon disease in children, but when it occurs it is associated with significant implications for the anesthesiologist. The causes of adrenal insufficiency are listed in Box 36-3. Adrenal insufficiency may include glucocorticoid deficiency with or without mineralocorticoid deficiency (Box 36-4). Isolated hypoaldosteronism is rare. In the perioperative period, children with congenital adrenal insufficiency require glucocorticoid and mineralocorticoid replacement. Chronic deficits in adrenal function result in the classic findings of Addison disease, including hyperpigmentation, weakness, and hyponatremia. The hyperpigmentation results from high levels of adrenocorticotropic hormone (ACTH) and unopposed melanophore-stimulating hormone caused by cortisol insufficiency. The additional presence of aldosterone insufficiency may produce hyponatremia, hyperkalemia, hypotension, and a small cardiac silhouette that results from hypovolemia (Keon and Templeton, 1993).

Box 36-3 Causes of Adrenal Insufficiency Primary adrenocortical insufficiency Congenital Enzyme deficiency Adrenal aplasia Adrenocortical unresponsiveness to ACTH Adrenoleukodystrophy or adrenomyeloneuropathy Trauma or septic Adrenal hemorrhage of newborn Adrenal hemorrhage of acute infection Chronic hypoadrenocorticism (Addison disease)

Positive-pressure ventilation

Secondary to deficient ACTH secretion Hypopituitarism Cessation of glucocorticoid therapy Resection of unilateral cortisol-producing tumor Infants born to steroid-treated mothers Respiratory distress syndrome Anencephaly Inanition, anorexia nervosa

Drugs Vincristine Vinblastine

Related to end-organ unresponsiveness Pseudohyopoaldosteronism Cortisol resistance

Infectious Pneumonia Tuberculosis Shigellosis Infant botulism

C h a p t e r 36    Systemic Disorders   1105

Box 36-4 Signs and Symptoms of Adrenal Insufficiency Glucocorticoid deficiency Fasting hypoglycemia Increasing insulin sensitivity Decreased gastric acidity Gastrointestinal symptoms (nausea, vomiting) Fatigue Mineralocorticoid deficiency Muscle weakness Weight loss Fatigue Nausea, vomiting, anorexia Salt craving Hypotension Electrolyte disturbance Hypokalemia Hyponatremia Acidosis Adrenal androgen deficiency Decreased pubic and axillary hair Decreased libido Increased b-Lipoprotein levels Hyperpigmentation

Perioperative Steroid Management The preoperative recognition of adrenal insufficiency and appropriate preoperative therapy minimize the likelihood of significant perioperative complications. Ninety percent of patients with congenital adrenal hyperplasia with adrenal insufficiency have 21-hydroxylase deficiency (Migeon and Donohoue, 1994). Virilization of the external genitalia occurs in female patients, and they often require surgical revision of their external genitalia. An abnormal genital pigmentation occurs in male patients, but this finding may be subtle. Infants with undiagnosed congenital adrenal hyperplasia may undergo exploratory laparotomy for acute abdomen because of nausea and vomiting. It is important to be attuned to the signs and symptoms in the history, physical examination, and laboratory evaluation that point to this diagnosis to prevent or treat shock, which may occur because of failure to administer steroid replacement. Mineralocorticoid deficiency can be managed by administering saline solution and avoiding potassium in IV fluids. Min­ eralocorticoid secretion rates in children are similar to those in adults, and the replacement dose is independent of age and weight. Desoxycorticosterone acetate may be administered intramuscularly in a dose of 1 mg/day. The intramuscular injection may be replaced by a single daily oral dose of 9-fluorocortisol (Florinef, 0.05 to 0.1 mg) when it is clear that an oral medication can be tolerated and absorbed. When cortisol is administered perioperatively, it has sufficient mineralocorticoid activity to obviate the need for any other replacement; 20 mg of hydrocortisone has mineralocorticoid activity equivalent to 0.1  mg 9a-fluorocortisol (Miller et al., 2008). Glucocorticoid deficiency is treated with cortisol (hydrocortisone) replacement. The importance of cortisol replacement for patients with known adrenal insufficiency should not

be underestimated, although vastly excessive doses are unwarranted. In the normal individual, the adrenal gland secretes 12 ± 2 mg of cortisol per square meter of body surface area every 24 hours (Kenny and Preeyasombat, 1966). The normal replacement dose prescribed for unstressed children is 25 mg/ m2 per day; the dose is double the normal production because of factors of bioavailability and half-life (Migeon and Donohoue, 1994). In response to stress (e.g., fever, acute illness, surgery, and anesthesia), the normal adrenal gland secretes 3 to 15 times this amount. Consequently, in the past, the recommendations for “stress” steroid coverage in the perioperative period ranged from 36 to 180 mg/m2 per day. More important than just the dose of steroid to be given, consideration should be devoted to the type of glucocorticoid administered, its half-life, the route of administration, and the timing of doses. The equivalencies for steroid preparations in terms of their relative glucocorticoid and mineralocorticoid effects are presented in Table 36-5. The most commonly cited recommendation for perioperative steroid coverage is hydrocortisone hemisuccinate (Solu-Cortef), given intravenously as 2  mg/kg immediately preoperatively and every 6 hours on the day of surgery, with reductions in the postoperative period depending on the degree of stress. Some practitioners feel that the half-life of hydrocortisone is so short that a 6-hour dosing interval may lead to periods of inadequate “coverage.” These practitioners recommend a preinduction dose of 25 mg/m2 of hydrocortisone given intravenously, followed by a continuous infusion of 50 mg/m2 administered during the estimated period of anesthesia. Postoperatively, 50 mg/m2 by continuous infusion is administered over the remainder of the first 24 hours. The total dose for the first 24 hours is 125 mg/m2, or 10 times normal physiologic production (Migeon and Donohoue, 1994). The first bolus dose must be administered before induction of anesthesia rather than waiting for an IV cannula to be placed after inhalational induction because of the stress associated with anesthetic induction itself. In the postoperative period, the steroid dose is tapered to a level commensurate with the residual stress. It is replaced with the child’s usual oral preparation when the child clearly can tolerate and absorb oral medication.

Hypothalamic-Pituitary-Adrenal Axis Suppression Caused by Exogenous Steroid Therapy In addition to the diseases discussed previously, suppression of the hypothalamic-pituitary-adrenal (HPA) axis can also occur after exogenous steroid usage, such as that administered for the treatment of inflammatory conditions (e.g., Crohn disease and asthma) or autoimmune disease (e.g., lupus and juvenile rheumatoid arthritis). Nearly 60 years ago, it was reported that two patients developed irreversible shock perioperatively after glucocorticoid administration was stopped preoperatively (Fraser et al., 1952; Lewis et al., 1953). Both patients were found to have adrenal atrophy and hemorrhage at autopsy. These two cases led to suggestions for “stress” steroid coverage in the perioperative period. HPA suppression places steroid-dependent children at increased risk for complications in the perioperative period, because they may be unable to respond to stress with an appropriate increase in the adrenal secretion of glucocorticoid. Dosages of cortisol or its equivalent that exceed 15 mg/m2 per day for more than 2 to 4 weeks invariably produce HPA suppression. A study in children with relatively short-term exposure to

1106   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-5. Potency of Commonly Used Steroid Preparations Sodium Retention Effect (= 0.1 mg Fludrocortisone [Florinef])

Duration of Action

100

20

S

Cortone

125

20

S

Generic Name

Trade Name

Hydrocortisone

Hydrocortisone Solu-Cortef

Cortisone

Glucocorticoid Effect (= 100 mg Cortisol)

Prednisolone

Delta-Cortef

  20

50

I

Prednisone

Deltasone Meticorten

  25

50

I

Methylprednisolone

Medrol Solu-Medrol

  15

No effect

I

Triamcinolone

Aristacort Kenacort

  10

No effect

I

Dexamethasone

Decadron Hexadrol

   1.5

No effect

L

Betamethasone

Celestone

   3

No effect or salt loss

L

Aldosterone

NCA

300

0.1-0.04

—

9-Fluorocortisol

Florinef

   6.5

0.1

I

Desoxycorticosterone acetate

NCA

   0

1 (IM)

I

Adapted from Migeon C, Donohoue PA: Adrenal disorders. In Kappy MS et al., editors: The diagnosis and treatment of endocrine disorders in childhood and adolescence, Springfield, Ill, 1994, Charles C. Thomas. S, Short (8-12 hr biological half-life); I, intermediate (12-36 hr biological half-life); L, long (36-72 hr biological half-life); IM, intramuscularly; NCA, not commercially available.

prednisolone or dexamethasone (5 and 3 weeks, respectively) for treatment of acute lymphoblastic leukemia showed that recovery of normal adrenal function (in response to ACTH stimulation) had a very wide range, occurring between 2 weeks and 8 months (Petersen et al., 2003). Although high dosages, prolonged therapy, and short duration between discontinuance of therapy and the surgical procedure increase the likelihood of HPA suppression, no practical test is available that unequivocally identifies patients who will need intraoperative steroids. Metyrapone depresses the production of cortisol by the adrenal glands and can be used to test the capacity of the pituitary gland to respond to decreased plasma cortisol concentrations by increasing ACTH secretion (Haynes, 1990). However, this test takes 3 days, is expensive, and has the risk of inducing adrenal insufficiency. Similarly, an ACTH-stimulation test can be performed at great expense to test adrenal responsiveness. However, even if cost and time were not issues, a study has shown a poor correlation between tests that indicate normal HPA function and dose or duration of glucocorticoid therapy or basal cortisol levels (Schlaghecke et al., 1992). Clinically significant events rarely occur during the perioperative period in unsupplemented patients who were receiving steroid medications for diseases other than adrenal insufficiency. Nevertheless, the potential for symptomatic adrenal insufficiency, although rare, coupled with the low risk of steroid-induced complications for short-term administration, suggests that steroids should be given in uncertain cases. If steroid therapy has been discontinued within the previous year, Donohoue (2005) makes the following recommendations: l If the dose of the glucocorticoid administered was less

than replacement levels, independent of the duration of

administration, there will be no major HPA suppression and therefore no need for supplementation. l If the dose of glucocorticoid administered was greater than replacement levels, HPA suppression will occur. If treatment lasted fewer than 2 weeks, suppression is transient, with prompt recovery (fewer than 2 weeks). If treatment lasted longer than 2 weeks, HPA suppression may persist for 1 week to 6 months, with 50% of patients recovering function within 6 weeks. This is the case even if the glucocorticoid was administered on an every-otherday basis. HPA suppression also can result from modes of steroid administration other than oral, including topical, nasal spray, and inhalers. Although adrenal suppression is rarely symptomatic with these modes of administration, some drugs, especially fluticasone propionate, in high doses have been associated with growth failure and adrenal suppression (Duplantier et al., 1998). With surgical stress, patients with adrenal suppression may become symptomatic, as has been reported for other kinds of stress. The patients reported by Drake et al. (2002) all had been taking fluticasone and had hypoglycemia at times of stress from intercurrent illness. Numerous cases of acute adrenal crisis have been reported in children receiving inhaled corticosteroids (ICS) for prolonged periods (Randell et al., 2003). Anesthesiologists should have a high index of suspicion of adrenal suppression if an asthmatic child on inhaled steroids develops hypotension or hypoglycemia in the perioperative period. For children who have been on prolonged courses of highdose steroids (e.g., for asthma or treatment of acute lymphocytic leukemia), the glucocorticoid regimen should follow that described earlier for patients with adrenal insufficiency. The dose

C h a p t e r 36    Systemic Disorders   1107

given should be commensurate with the normal physiologic corticosteroid production in response to stress (as described) and does not need to be a multiple of the pharmacologic dose being administered for the underlying medical illness. The dose administered should also be proportional to the perceived degree of surgical stress. For brief procedures, such as upper endoscopy, a single preoperative dose of steroids is suggested (50 mg/m2 of hydrocortisone); for more complicated cases, such as appendectomy or major intraabdominal operations, 100 mg/m2 is administered as a continuous infusion or divided into 4 doses per day. This dose is usually continued for 1 to 3 days after more complex surgical procedures (Krasner, 1999). The dose is tapered postoperatively and replaced with the patient’s usual oral steroid preparation and dose when the child is able to tolerate oral medications. The dose that these patients commonly take for the underlying disease often exceeds even maximum “stress” doses described for congenitally adrenal insufficient patients, and treatment required for the underlying disease may limit further tapering of the steroid dose. A small study of adults comparing “stress steroids” with saline showed no adverse effects in patients who continued their usual steroid dose for their underlying disease (Glowniak and Loriaux, 1997).

Thyroid Disorders

TABLE 36-6. Thyroid Function Tests Test

Age

Normal

T4 RIA (mcg/dL)

1-3 days 1-4 weeks 1-12 months 1-5 years 6-10 years 11-15 years 16-20 years

11-21.5 8.2-16.6 7.2-15.6 7.3-15 6.4-13.3 5.6-11.7 4.2-11.8

T3 RU

25%-35%*

T index

1.25-4.20†

Free T4 (ng/dL)

1-10 days >10 days

0.6-2 0.7-1.7

T3 RIA (ng/dL)

1-3 days 1-4 weeks 1-12 months 1-5 years 6-10 years 11-15 years 16-20 years

100-380 99-310 102-264 105-269 94-241 83-213 80-210

TSH RIA (mIU/mL)

1-3 days 1-4 weeks 1 month-15 years 16-20 years

<2.5-13.3 0.6-10 0.6-6.3 0.2-7.6

TBG (mg/dL)

1-3 days 1-4 weeks 1-12 months 1-5 years 6-20 years

— 0.5-4.5 1.6-3.6 1.3-2.8 1.4-2.6

Reverse T3‡ (mg/dL)

Newborns Adults

90-250 10-50

Hypothyroidism Hypothyroidism occurs because of abnormally low production of thyroid hormone. It may be caused by primary thyroid dysfunction or result from pituitary failure with decreased production of thyroid-stimulating hormone (TSH). Normal values for routinely performed thyroid function tests are presented in Table 36-6, and the interpretation of these test results with regard to diagnosis is presented in Table 36-7. Primary thyroid dysfunction may be congenital or acquired. Congenital hypothyroidism usually appears in infancy. Classic features in the infant include large fontanels, wide sutures, large tongue, umbilical hernia, and decreased deep tendon reflexes. In the older child, manifestations include slow heart rate, narrow pulse pressure, growth failure, hypothermia, and cold intolerance. Severe hypothyroidism is rare but may be associated with coma, cardiovascular collapse, hyponatremia, hypothermia, and respiratory failure. Keon and Templeton (1993) reviewed the anesthetic management of patients with hypothyroidism and stressed the importance of correcting hypothyroidism gradually over a 2-week period. Sudden death has been reported in children with myxedematous heart disease 2 to 3 weeks into therapy (LaFranchi, 1979). It is suggested that these children receive one fourth of the maintenance dose of thyroid hormone (6 to 8 mcg/ kg per day for an infant), with gradual incremental increases over 2 to 4 weeks until a maintenance dose is reached (Custer and Rau, 2009). Patients who are adequately treated will have normal thyroid hormone and TSH levels. Patients who do not respond to oral thyroid replacement may be given IV triiodothyronine (T3; loading dose of 0.7 mcg/kg), followed by an infusion titrated to T3 and TSH levels. Severe cardiac dysfunction in such a patient improved with T3 therapy (Mason et al., 2001). Patients with incompletely restored thyroid hormone levels may require hemodynamic monitoring and support to maintain hemodynamic stability. Patients with severe hypothyroidism may have

Modified from Johnson KB, editor: The Harriet Lane Handbook, St. Louis, 1993, Mosby. *Measures thyroid hormone binding, not T3. † T4 RIA × T3 RU. ‡ Reverse T3. RIA, Radioimmunoassay; RU, resin uptake; T3, triiodothyronine; T4, thyroxine; TBG, thyroid-binding globulin; TSH, thyroid-stimulating hormone.

TABLE 36-7. Interpretation of Thyroid Function Tests T4 RIA

T3 RU

T Index

Free T4

TSH

Primary hypothyroidism

L

L

L

L

H

Secondary hypothyroidism

L

L

L

L

L, N, or H

TBG deficiency

L

H

N

N

N

Hyperthyroidism

H

H

H

H

L

Modified from Siberry GK, Iannone R, editors: The Harriet Lane handbook, St. Louis, 2000, Mosby. L, Low; H, high; N, normal; RIA, radioimmunoassay; RU, resin uptake; T3, triiodothyronine; T4, thyroxine; TBG, thyroid-binding globulin; TSH, thyroidstimulating hormone.

associated adrenal insufficiency, and if so, they should receive stress steroid coverage as outlined earlier. The anesthetic care of the symptomatic patient with hypothyroidism can be problematic and requires caution when any depressant medications are given. Prolonged effects may result

1108   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-8. Anesthetic Implications of Hypothyroidism System

Anesthetic Considerations

Pharmacologic

Possible lower MAC value; prolonged recovery from opioid anesthesia

Cardiovascular

Decreased cardiac output, heart rate, and stroke volume; increased PVR and decreased intravascular volume; myocardial depression resulting from impaired cellular metabolism or myxedematous infiltration; baroreceptor dysfunction

Respiratory

Abnormal response to hypercapnia and hypoxia

Thermal regulation

Hypothermia resulting from reduced basal metabolic rate; reduced ability to increase core temperature

Endocrine

Increased incidence of adrenal insufficiency; consideration for stress steroid coverage

Metabolic

SIADH; hypoglycemia with prolonged fasting

Gastrointestinal

Delayed gastric emptying; consideration for full stomach precaution

MAC, Minimum alveolar concentration; PVR, peripheral vascular resistance; SIADH, syndrome of inappropriate antidiuretic hormone.

from decreased drug metabolism. Important considerations in the management of hypothyroidism as described by Keon and Templeton (1993) are outlined in Table 36-8. Invasive monitoring may be indicated when significant blood loss or fluid shifts occur. Care should be taken intraoperatively to minimize heat loss. Postoperative care should include monitoring of oxygen saturation, blood pressure, heart rate, and respiratory rate; postoperative ventilation may be necessary in the patient with delayed emergence from anesthesia.

Hyperthyroidism Hyperthyroidism is a syndrome produced by excess levels of ­circulating thyroid hormone. The most common causes are congenital hyperthyroidism and Graves disease (i.e., toxic goiter). Less commonly, acute suppurative thyroiditis, hyperfunctioning thyroid carcinoma, thyrotoxicosis factitia (i.e., exogenous administration of thyroid hormone), and toxic uninodular goiter (i.e., Plummer disease) may produce this syndrome. McCune-Albright syndrome (i.e., precocious puberty with polyostotic fibrous dysplasia) is also commonly associated with hyperthyroidism (Jones, 1988).

Congenital Hyperthyroidism Congenital hyperthyroidism is a transient phenomenon seen in newborns that results from the transplacental transfer of ­thyroid-stimulating antibody from mothers who commonly have a history of Graves disease. Most of these infants have a goiter and typically appear anxious and restless or irritable. Signs of hypermetabolism, including tachycardia, tachypnea, and elevated temperature, may be present. In the severely affected infant, symptoms may progress to weight loss, severe hypertension, and high-output cardiac failure with hepatomegaly (Smith et al., 2001). Appropriate medical therapy (methimazole) should be instituted early. Because maternal immunoglobulins have a short half-life in infants, the hyperthyroid state resolves in a few

weeks to a few months, and it sometimes may be followed by a period of hypothyroidism (Higuchi et al., 2001).

Graves Disease Diffuse toxic goiter (Graves disease) is the most common cause of hyperthyroidism in children. Its peak incidence occurs during adolescence, and it is five times more common in girls than in boys. The clinical course is generally gradual, with symptoms developing over 6 to 12 months. Early signs include motor hyperactivity, emotional disturbances, and nervousness. Affected children are progressively more irritable and restless and may have increased sweating, increased appetite, palpitations, and tremors of their fingers. Most children have obvious exophthalmos and an enlarged palpable thyroid. The cardiopulmonary symptoms of hyperthyroidism include systolic hypertension, tachycardia, palpitations, dyspnea, and cardiac enlargement, which may progress to frank cardiac decompensation. On rare occasions, atrial fibrillation or mitral regurgitation may also be present.

Thyroid Storm An acute onset of hyperthermia, severe tachycardia, and restlessness comprises the syndrome of acute uncompensated thyrotoxicosis, or “thyroid storm.” Without appropriate and timely therapy, the patient’s condition may deteriorate to delirium, coma, and death. Therapy includes treatment of hyperthermia by cooling, maintenance of intravascular volume with balanced salt solutions, and β-adrenergic blockers such as propranolol titrated to ameliorate the cardiovascular response. Specific thyroid suppression therapy with propylthiouracil should be instituted. The clinical presentation of thyroid storm may occur intraoperatively, and this hypermetabolic state may be mistaken for malignant hyperthermia (Peters et al., 1981). The use of dantrolene mitigated the clinical signs in a patient who turned out to have thyroid storm (Bennett and Wainwright, 1989). It is well known that perioperative surgical stress can trigger the development of thyroid storm in a patient with previously unrecognized thyrotoxicosis (Stevens, 1983). For this reason, patients with signs and symptoms that may indicate the presence of hyperthyroidism should be carefully evaluated. Patients should be rendered euthyroid before any elective surgery, even if it is minor.

Laboratory Evaluation Serum levels of thyroxine (T4) and T3 are usually elevated in hyperthyroidism. TSH secretion is suppressed and may be unmeasurable. T3 toxicosis (elevated T3 level with normal amounts of T4) is more common in adult patients and rarely seen in children. For borderline cases, thyrotropin-releasing hormone (TRH) stimulation tests may be needed. Many patients with Graves disease of recent onset may have elevated levels of thyroid-stimulating immunoglobulin. Radionuclide scans can also be helpful in making a diagnosis. If a large goiter is present, neck radiographs, computed tomography (CT) scans, or magnetic resonance imaging (MRI) may be used to evaluate the degree of tracheal compression and deviation.

Treatment The management of hyperthyroidism is aimed at controlling the cardiovascular effects. β-adrenergic receptor blockade, ­usually

C h a p t e r 36    Systemic Disorders   1109

with propranolol (1 to 2 mg/kg per day), is titrated to effect. Antithyroid medications include propylthiouracil and methimazole, both of which inhibit the incorporation of inorganic iodide into organic compounds. Propylthiouracil inhibits the conversion of T4 to T3. Although early studies suggested that these agents might inhibit the formation of thyroid antibodies, later studies that included careful histopathologic analysis have shown this to be false (Paschke et al., 1995). The Food and Drug Administration (FDA) has recently recommended that propylthiouracil not be administered to pediatric patients because of reports of liver failure associated with its use. Methimazole is the antithyroid medication recommended for use in infants, children, and adolescents (FDA, 2009). Saturated solutions of potassium iodide may be administered orally (1 drop every 8  hours) to suppress thyroid hormone secretion. The clinical response to therapy is evident in 1 to 3 weeks, and the patient may require up to 3 months for adequate control to be achieved. Patients must have appropriate, regular surveillance to ensure that the T3 and T4 levels are in the normal range and that TSH concentrations are normal. Clinically, the patient demonstrates a euthyroid state by return of the heart rate, blood pressure, and reflexes to normal. Radioactive iodine is often used to treat hyperthyroidism in adults. However, such therapy is avoided in children because of side effects such as thyroid cancer, genetic damage to germ cells, and a higher incidence of hypothyroidism than occurs with pharmacologic therapy.

Anesthetic Management Preoperatively, patients should be pharmacologically euthyroid. Any residual cardiovascular signs and symptoms should be well controlled through the use of a β-adrenergic receptor blocker. Esmolol is an excellent choice for intraoperative use. A large goiter may produce tracheal deviation or compression, and the possibility of airway compromise should be evaluated preoperatively with radiographic studies. Any commonly used sedative may be given for premedication. Atropine and other anticholinergics should be avoided or used with extreme caution, because they decrease sweating and may interfere with thermoregulation. Medications, including antithyroid drugs and β-adrenergic blockers, should be administered through the morning of surgery.

Intraoperative Management of Patients for Thyroidectomy In children with large goiters who have a compromised airway, anesthesia should be managed with caution, as in any other child with upper airway obstruction. A sedated fiberoptic intubation may be chosen, or an inhalational induction performed, with maintenance of spontaneous ventilation until the airway is secured. If a large goiter has caused prolonged tracheal compression, a segment of tracheomalacia may exist, and an armored endotracheal tube (ETT) may be indicated. For patients without issues of tracheal compression, anesthesia may be induced with thiopental or propofol. However, if the patient’s laboratory evaluation indicates a continued hyperthyroid state, ketamine should be avoided because of its effect on catecholamine release. Mask inductions may be prolonged in these patients because of increased cardiac output, resulting in a slower rise in the alveolar concentration of the anesthetic if the ventilation is kept constant. Minute ventilation may be

reduced if significant airway obstruction resulting from goiter or tracheomalacia is present. Care should be taken to lubricate, pad, and appropriately protect the eyes, especially if they are protuberant because of Graves disease. With respect to choice of drugs, muscle relaxants with few cardiovascular side effects, such as cisatracurium, vecuronium, or rocuronium, provide a potential benefit by minimizing the occurrence of tachycardia. Similarly, for the maintenance of general anesthesia, anesthetics that have sympathomimetic effects should be avoided. If the child is in a hypermetabolic state, drug biotransformation may be accelerated; therefore, agents such as halothane, which has toxic metabolic products, are potentially more hazardous. Another reason to avoid halo­ thane is its sensitization of the myocardium to catecholamines. If the patient is in a hypermetabolic state, controlled ventilation should be employed during the surgical procedure to minimize the development of hypercapnia, which can contribute to further sympathetic stimulation. Anesthetic management at the conclusion of surgery may include deep extubation, with direct or fiberoptic laryngoscopy performed to evaluate the presence or absence of vocal cord paralysis, which may result from surgical trauma to the recurrent laryngeal nerve (traction or transection). Care should be taken after extubation to observe for signs of airway obstruction as a result of residual tracheomalacia. The decision to evaluate the airway prospectively at the time of extubation should be made jointly by the anesthesiologist and surgeon and should be based on the likelihood of residual tracheal obstruction caused by tracheomalacia or of vocal cord paralysis because the surgeon thinks that surgical trauma is likely.

Postoperative Care Children who have undergone thyroidectomy require close observation in the postoperative period (Fewins et al., 2003). They may develop postextubation croup or upper airway obstruction as a result of paralysis of the vocal cords, tetany (hypocalcemia), residual tracheomalacia, or tracheal compression resulting from a hematoma. Patients with postextubation croup may respond to supportive measures, including humidified supplemental oxygen, nebulized racemic epinephrine, and possibly, continuous positive airway pressure (CPAP) or bilevel positive airway pressure (BiPAP). Occasionally, these patients require brief reintubation before they can be successfully extubated. Unilateral vocal cord paralysis may go unnoticed or be associated with only mild stridor. Bilateral vocal cord paralysis, on the other hand, may manifest as severe stridor and upper airway obstruction. The child with bilateral vocal cord paralysis requires reintubation for airway support. A muscle relaxant such as rocuronium should be used to facilitate reintubation to avoid damage to the abducted cords. If the paralysis is prolonged, the child may subsequently require tracheostomy. Compression of the trachea by a hematoma may occur immediately after the operation or over the course of several hours. The child requires reintubation and surgical evacuation of the hematoma to relieve tracheal compression. Opening the wound in the recovery room may be necessary and lifesaving. After the extrinsic obstruction has been relieved and the incision closed again (if necessary), the child may be safely extubated. Inadvertent resection of the parathyroid glands during thyroidectomy may result in acute hypoparathyroidism after surgery. Clinical signs of hypocalcemia may become manifest

1110   P a r t  IV    Associated Problems in Pediatric Anesthesia

within the first postoperative day or take as long as 72 hours to develop. A low serum-ionized calcium level and a low concentration of parathyroid hormone are diagnostic. Clinical hypocalcemia, including tetany, is treated with IV calcium therapy. Surgical manipulation of the trachea and neck tissues can also lead to subcutaneous emphysema and the more serious possibility of pneumomediastinum or pneumothorax. Postoperative evaluation should include radiologic examination of the chest if respiratory distress occurs.

Pheochromocytoma Pheochromocytoma is a catecholamine-secreting tumor of chromaffin cells that most commonly arises in the adrenal medulla (DiGeorge, 1987). It may be found anywhere along the abdominal sympathetic chain, but it is most commonly near the aorta at the inferior mesenteric artery or the aortic bifurcation. Other sites include the neck, the mediastinum, and the walls of the bladder or ureters. Pheochromocytoma is a rare neoplasm in the pediatric population. Fewer than 5% of reported cases occur in children. The tumors may occur bilaterally or at multiple sites. This condition can be inherited as an autosomal-­dominant trait (most often in association with von Hippel–Lindau syndrome) or as part of a multiple endocrine neoplasia (MEN) type II or III (Table 36-9). The abnormally high plasma levels of epinephrine and norepinephrine produce a clinical syndrome with signs and symptoms related directly to the level of each hormone present in the patient. Hypertension is common and often leads to hypertensive encephalopathy and seizures. In particular, paroxysmal hypertension is most suggestive of pheochromocytoma. The patient may also complain of headaches and palpitations, pallor, sweating, and vomiting. In severe cases, patients develop chest pain that radiates to the arms, pulmonary edema, and cardiac decompensation. The catecholamine-induced hypermetabolism also can cause patients to have a voracious appetite but still lose weight and become cachectic. Polyuria, polydipsia, and abdominal pain may occur and be confused with diabetes mellitus.

Diagnosis It is extremely important to establish the diagnosis of pheochromocytoma before induction of anesthesia and start of surgery. The significant cardiovascular effects of excess catecholamines

TABLE 36-9. Multiple Endocrine Neoplasia Syndromes MEN Syndrome

Affected Organs

Disorder

Werner syndrome (type I, familial)

Parathyroid gland; pancreas; pituitary gland

Hypercalcemia; hypoglycemia; peptic ulcer

Sipple syndrome (type II, autosomal dominant)

Thyroid and parathyroid glands; adrenal medulla

Medullary carcinoma; hypercalcemia; pheochromocytoma

Type III

Nervous system; thyroid gland; adrenal medulla

Multiple neuromas; medullary carcinoma; pheochromocytoma

MEN, Multiple endocrine neoplasia.

can pose difficulties for the anesthesiologist and endanger the patient during the perioperative period if an appropriate diagnosis is not made preoperatively. These tumors can produce paroxysms of hypertension and other symptoms. Between paroxysms, the patient may be totally asymptomatic, making diagnosis extremely difficult. The demonstration of increased levels of catecholamines is the most specific diagnostic test. Although pheochromocytomas can produce norepinephrine and epinephrine, the predominant catecholamine produced in children is norepinephrine, which leads to chronic hypertension. Urine catecholamine concentrations are directly proportional to circulating levels, and determination of 24-hour urinary excretion of the primary catecholamines and their metabolites (i.e., 3-­methoxy 4-hydroxy vanillyl-mandelic acid [VMA] and metanephrine) used to be the primary means of establishing the diagnosis. The plasma free metanephrine determination has better sensitivity (100%) and specificity (94%). Because normal values differ with age, it is important to use age-specific norms when interpreting results (Weise et al., 2002). The differential diagnosis includes renal vascular disease, hyperthyroidism, Cushing syndrome, coarctation of the aorta, adrenal cortical tumors, and essential hypertension. Cerebral disorders, diabetes mellitus, and DI may produce similar symptoms. Neoplasms of neural origin (e.g., neuroblastoma or ganglioneuroma) may also secrete catecholamines. Before any contemplated anesthesia and surgery, the patient must undergo a complete evaluation to localize the tumor, including CT and MRI scans. Magnetic resonance angiography (MRA) and venography (MRV) may be used to identify vascular supply. A radionuclide (I-131 metaiodobenzylguanidine [MIBG]) scan may also help to localize the tumor. Differential venous catheterization may be needed to obtain blood from various sites for catecholamine levels if multiple tumors are suspected. Sedation or anesthesia may be required to perform these studies in infants and children. Before sedation for these diagnostic studies, hemodynamic abnormalities must be normalized. Drugs that induce catecholamine secretion or histamine release should be avoided. General anesthesia for the diagnostic procedures must be conducted with the same extreme caution one would exercise for resection of the tumor itself.

Preoperative Preparation and Evaluation Preoperative evaluation should include the measurement of serum electrolytes, determination of renal function, and fasting blood glucose. Excessive serum epinephrine levels may be associated with hyperglycemia and hypokalemia. An electrocardiogram and echocardiogram are important to evaluate cardiac rhythm, size, and function. Some patients with pheochromocytoma have a catecholamine-induced cardiomyopathy with decreased left ventricular contractility. The patient should be evaluated for associated endocrinopathies that may be present as part of MEN type II or III. Symptomatic treatment includes the administration of phenoxybenzamine over a period of several days to weeks before surgery. Phenoxybenzamine is a long-acting, orally administered α-adrenergic blocking agent that attenuates the effects of catecholamines on the peripheral circulation by blocking excessive vasoconstriction (Hoffman and Lefkowitz, 1990). The starting dose of phenoxybenzamine is 0.2 mg/kg once daily (maximum adult dose, 10 mg) (Taketomo et al., 2008). Then both dose and frequency are gradually increased (up to 3 times a day)

C h a p t e r 36    Systemic Disorders   1111

until a clinical effect is obtained; that is, the patient’s hematocrit decreases (because of vasodilation and increased blood volume), and the patient develops orthostatic changes in vital signs. Long-standing vasoconstriction produced by chronically high catecholamine levels causes decreased intravascular volume. Although the use of phenoxybenzamine restores vascular capacity to normal, increased oral fluid intake should accompany administration of phenoxybenzamine to avoid severe orthostatic changes. In some children, β-adrenergic blocking drugs such as propranolol may be needed to control heart rate and blood pressure. However, β-blocking agents should never be used without concurrent α-blockade therapy because of the deleterious effects of unopposed α-agonism, which may result in cardiac failure as a result of increased afterload. Labetalol may have a role in the management of pheochromocytoma, because it has α- and β-adrenergic blocking properties (Blom et al., 1987). It can be useful in minimizing the cardiovascular effects of excess catecholamines during the perioperative period, but it is not as potent an a blocker as is phenoxybenzamine for preoperative treatment. In addition to pharmacologic preparation, children with pheochromocytoma benefit from preoperative sedation to reduce the release of catecholamines caused by anxiety. Oral or IV midazolam alone or in combination with an opioid provides a good level of sedation.

Anesthetic Induction Induction of anesthesia is accomplished with IV anesthetics. Ketamine is specifically contraindicated, because it induces catecholamine release. Halothane should be avoided, because it may sensitize the myocardium to catecholamines and produce dysrhythmias. Mask induction with sevoflurane is well tolerated if hemodynamic parameters are well controlled, and IV cannula placement can be deferred until after induction. Intubation may proceed in the usual fashion, facilitated by a hemodynamically neutral nondepolarizing muscle relaxant such as vecuronium. Pancuronium, which causes muscarinic blockade and tachycardia, should be avoided. Despite the fact that atracurium causes histamine release and vasodilation, it has been used safely in adult patients with pheochromocytoma (Prys-Roberts, 2000). Before intubation, IV lidocaine (1 mg/kg), fentanyl (2 to 5 mcg/kg), or both are effective in minimizing the hemodynamic response to intubation.

Intraoperative Management After induction of anesthesia, an intraarterial catheter is placed to monitor the blood pressure continuously. Before induction, a reliable automated blood pressure device can provide frequent, accurate blood pressure measurements. A central venous catheter provides direct and reliable access for the assessment of intravascular volume and for infusions of fluids and emergency medications. Anesthesia may be maintained with isoflurane or sevoflurane, air, and oxygen. Both anesthetic agents have been used without exacerbation of hypertension, despite the fact that they do not blunt the production of norepinephrine in response to surgical stimulation (Suzukawa et al., 1983). Desflurane should be avoided because of its tendency to cause tachycardia and hypertension. The addition of moderately large doses of fentanyl (10 mcg/kg) or remifentanil (0.3 to 1 mcg/kg per

minute) minimizes the stress response and provides stable hemodynamics. If remifentanil is chosen, it is important to give a longer-acting opioid before the end of surgery to avoid hypertension as a result of pain on awakening. Adjunctive use of epidural anesthesia (i.e., local anesthetic with or without a small dose of fentanyl) is an excellent method of reducing the stress response and catecholamine release caused by usual surgical stimulation. However, none of these anesthetic strategies blocks catecholamine release that results from direct surgical manipulation of tumor tissue. Historically, controlling blood pressure during the induction and maintenance of anesthesia has been accomplished by an infusion of sodium nitroprusside or phentolamine. However, resection is increasingly being performed using a laparoscopic technique, and either nicardipine or magnesium sulfate infusions are used to maintain vasodilation and normotension during this type of surgery (Pretorius et al., 1998; Minami et al., 2002). In addition to propranolol use preoperatively, esmolol has been effective as a continuous infusion titrated to the level of surgical stimulation (Nicholas et al., 1988). Infusions of only short-acting vasodilators are recommended for control of hypertension before tumor resection, because with removal of the tumor, vasodilation caused by a persistent blockade and loss of excess catecholamines may lead to precipitous hypotension. Phenoxybenzamine has a long half-life; therefore, some physicians recommend discontinuing it 24 hours before surgery to decrease the likelihood that persistent vasodilation caused by a blockade will cause severe hypotension after the tumor is removed because of withdrawal of the catecholamines of tumor origin (PrysRoberts, 2000). Hypotension is best treated with discontinuation of vasodilator infusions, titration of anesthetic agents, and administration of crystalloid or colloid and blood products, if it is indicated by the magnitude of blood loss. If these measures are ineffective, vasopressors may be necessary, but the patient may be relatively resistant to α-agonists as a result of persistent α-blockade. If this occurs, judicious use of small doses of more potent, direct-acting vasoconstrictors (e.g., norepinephrine or epinephrine) may be necessary. Recently, vasopressin infusion has been effective in this situation (Deutsch and Tobias, 2006). During surgery, arterial blood gases, serum glucose, urine output, and body temperature should be closely monitored. Hyperglycemia may occur in response to high catecholamine levels. Hypoglycemia may occur when the tumor is removed and catecholamine levels decrease. Because of the potential for dysrhythmias, the electrocardiogram should also be vigilantly observed. At the end of surgery, the muscle paralysis is reversed. Extubation can be accomplished when the patient meets all normal extubation criteria.

Postoperative Care Postoperatively, the patient should be observed in an intensive care unit, with continuous monitoring of the arterial blood pressure and electrocardiogram. Hypertension usually resolves within 24 to 48 hours after surgery. If symptoms persist beyond this period, further investigation for residual pheochromocytoma is warranted. Good postoperative analgesia is provided with epidural infusion (if an open procedure was performed) or IV patient- (parent- or nurse-) controlled analgesia.

1112   P a r t  IV    Associated Problems in Pediatric Anesthesia

RESPIRATORY DISORDERS

Upper Respiratory Tract Infection Viral upper respiratory tract infections (URIs) are mild processes that do not preclude school attendance and other routine activities. However, URIs hold much greater significance for anesthesiologists. For many anesthesiologists, it is standard practice to avoid general anesthesia for elective surgery in children with URIs because of the respiratory complications during and after anesthesia reported in multiple small case series (McGill et al., 1979; Cohen and Cameron, 1991; Konarzewski et  al., 1992; Williams et al., 1992). Unfortunately, the vexing problem of runny noses in children is accentuated by the difficulty in differentiating URIs from other causes, such as allergic rhinitis, which does not increase the risk of complications.

Pathophysiology of Upper Tract Respiratory Infection Many investigators suggest that complications, including bronchospasm, intraoperative hypoxemia with an increased alveolar-arterial oxygen gradient, and postoperative hypoxemia, occur more often in children who undergo anesthesia while they have URIs (McGill et al., 1979; Olsson, 1987; DeSoto et al., 1988). The proclivity for these complications may be related to peripheral airway abnormalities, which have been demonstrated experimentally in adult humans and animals infected with viral respiratory pathogens (Johanson et al., 1969; Fridy et al., 1974; Dueck et al., 1991). These abnormalities include decreased diffusing capacity and increased closing volume— factors that can predispose patients to intrapulmonary shunting and hypoxemia, especially when they are combined with the effect of general anesthesia on lung volumes (decreased functional residual capacity) (see Chapter 3, Respiratory Physiology in Infants and Children) (Murat et al., 1985). These studies were done in adults who had infections involving their entire respiratory tracts rather than isolated URIs, and their results may support separation of treatment of patients with truly isolated URIs from those with any symptoms of more global airway or pulmonary parenchymal involvement (lower respiratory tract infection). Although the mechanisms by which viral respiratory infections lead to alterations in airway function are unclear, these experimental studies support the clinical impression that increased risk of perioperative hypoxemia occurs in patients with recent viral respiratory infection. Empey et al. (1976) demonstrated in adult patients that acute viral respiratory tract infection (influenza) produced marked bronchial reactivity to experimental bronchoconstrictor challenge that may persist for 6 weeks. Mechanisms by which viral infections lead to increased airway reactivity include the release of immunologic and inflammatory mediators such as leuko­ trienes, bradykinin, and histamine, which cause bronchoconstriction. Vagal-mediated mechanisms may be involved, because viral infections have been associated with changes in muscarinic receptors on airway smooth muscle (Fryer et al., 1990). Tissue concentrations of important enzymes such as neutral endopeptidase, which break down the neuropeptides that cause bronchoconstriction, are also decreased in viral infections (Jacoby et al., 1988; Dusser et al., 1989). However, these patients and animals cannot be said to have only URIs, because the airways below the larynx are also clearly affected. Patients whose infections are truly uncomplicated URIs or those with noninfectious

causes of runny nose should be differentiated from those who have evidence of lower respiratory involvement.

Perioperative Risk Many case reports in the literature document that in the perioperative period children with URIs have respiratory complications, including bronchospasm, stridor caused by subglottic edema, hypoxia, and atelectasis (McGill et al., 1979; Konarzewski et al., 1992; Williams et al., 1992). Three prospective studies have shown that patients with an active or recent URI had a 2- to 10-fold higher risk of bronchospasm or laryngospasm (Olsson and Hallen, 1984; Olsson, 1987; Cohen and Cameron, 1991). The incidence was higher among younger children, especially those younger than 2 years and those whose tracheas were intubated (Cohen and Cameron, 1991). Retrospective studies of much larger numbers of patients show a higher risk of respiratory complications may actually exist in asymptomatic children with history of URI within the 2 to 4 weeks preceding surgery than in those with acute URI (Tait and Knight, 1987b; Tait et al., 2001). Other studies found that the factors that increase the risk of adverse events were history of prematurity or reactive airways disease (RAD), parental smoking, copious secretions, nasal congestion intubation, and airway surgery (Tait et al., 2000, 2001). Despite this increased risk of adverse events, complications usually were easily treated and were not associated with any significant prolonged morbidity (Rolf and Cote, 1992; Tait et al., 2000, 2001). However, some patients in one study developed atelectasis severe enough to require bronchoscopy and prolonged postoperative mechanical ventilation. Although most of these studies evaluated children undergoing relatively minor elective surgery, another study reported that children with URI symptoms at the time of cardiac surgery also had increased risks for respiratory and other complications, including nonrespiratory infection (Malviya et al., 2003). Despite these findings, patients’ hospital stays were not prolonged, and the incidence of long-term sequelae was not increased. One study found that children with URIs who undergo mask halothane-nitrous oxide-oxygen anesthesia for myringotomy surgery had reduced severity and duration of URI symptoms in the postoperative period (Tait and Knight, 1987a). However, this reduction in symptoms may have resulted from the drainage and removal of infectious foci by the surgical procedure rather than from the beneficial effects of general anesthetics. Other investigators have reported no significant respiratory complications when children with URI were anesthetized (Hinkle, 1989; Jacoby and Hirshman, 1991). It is extraordinarily difficult to integrate the contradictory conclusions of these various series of patients to develop a logical algorithm for dealing with the child with a URI.

Anesthetic Decision Making It is apparent that no consensus has been reached in the literature or in the general anesthesia community with regard to the wisdom and safety of anesthetizing children with active or recent URIs. The bulk of the literature, clinical and experimental, suggests that recent viral infection increases the perioperative risk for respiratory complications, albeit mild and treatable, when the surgical and anesthetic plans require intubation. In children with underlying RAD, the risk for pulmonary complications immediately after acute URI is much greater than in

C h a p t e r 36    Systemic Disorders   1113

the normal patient population, making intraoperative bronchospasm much more likely. The threshold for postponing surgery in the asthmatic child with a recent URI who requires intubation is much lower. These risks must be weighed against the physiologic, psychological, and financial implications of delaying surgery. The most conservative approach to the child with a URI or recent URI is to postpone elective procedures for 1 to 2 weeks for uncomplicated rhinorrhea, congestion, and nonproductive cough and for 4 to 6 weeks for patients with lower-airway involvement (e.g., wheezing or productive cough). However, this may be an overly cautious and somewhat unrealistic recommendation. Children without other health conditions have an average of 3 to 8 colds per year, and children whose mothers smoke, who live in crowded conditions, and who attend daycare centers have a 61% incidence of URIs over a 2-week period (Fig. 36-2) (Fleming et al., 1987). It may be nearly impossible to find a time when the child does not have a URI or is not recovering from one. The needs of the family must be considered. Often, parents have traveled significant distances, taken time off from work, and made alternative childcare arrangements for their other children. Because the available data do not clearly indicate a single best approach to these patients, each anesthesiologist should develop a consistent approach appropriate to the individual practice. The following is a summary of an approach to the child with symptoms of URI. First, many children undergo operations directed at ameliorating their chronic upper respiratory tract symptoms. In cases such as myringotomy with tube placement, tonsillectomy, adenoidectomy, and cleft palate repair, the procedures are not automatically canceled or postponed unless the child’s signs and symptoms are clearly “different from baseline” or clearly involve more than the upper respiratory tract. Children who undergo the procedures just listed have a high incidence of upper respiratory tract symptoms and may always have manifestations consistent with URI. Most parents can advise whether their child is more congested than usual.

Box 36-5 Signs and Symptoms of Upper Respiratory Infection 1. Mild sore or scratchy throat 2. Mild malaise 3. Sneezing 4. Rhinorrhea 5. Nasal congestion or stuffiness 6. Nonproductive cough 7. Fever higher than 101° F (38° C) 8. Laryngitis To make a diagnosis of URI, two of any of the above signs or symptoms are required. If 1 and 2, 3 and 4, or 5 and 6 are combined, one additional sign or symptom is required. Modified from Tait AR, Knight PR: The effects of general anesthesia on upper respiratory tract infections in children, Anesthesiology 67:930, 1987a.

Elective surgeries other than those cited previously are postponed if any of the following are present: “croupy” cough; rectal temperature higher than 38.3° C associated with any URI sign or symptom; malaise or decreased appetite; and any evidence or recent history of lower respiratory tract involvement such as rales, wheezes, productive cough, or abnormal chest radiograph (Box 36-5). Laboratory and radiographic tests are usually not helpful in the decision-making process, although some investigators recommend obtaining a chest radiograph and a white blood cell count to evaluate the child with a URI. The white blood cell count is neither sensitive nor specific in identifying a URI, and chest radiography associated with a normal auscultative examination is unlikely to identify abnormalities (Brill et al., 1973). The presence of rales or wheezes should lead to postponement of elective surgery, regardless of findings on the chest radiograph. A suggested algorithm for making decisions about proceeding with surgery is presented in Figure 36-3.

Anesthetic Management

Age <36 mo.

Crowding

.35

.47

.49

.27

.28

.39

No crowding

.18

Age ≥36 mo.

.15

.22

.23

.33

None

Daycare only

Smoking only

Both

Probability of URI 2 weeks

.61

Risk factors n  FIGURE 36-2. Probability of upper respiratory tract infection according to age, crowding, maternal smoking, and daycare status. (From Fleming DW, et al.: Childhood upper respiratory tract infections: to what degree is incidence affected by daycare attendance? Pediatrics 79:55, 1987.)

Several major principles of anesthetic management can be suggested for dealing with children with acute or recent URIs (Tait and Malviya, 2005). These are especially important in anesthetizing children when surgery is urgent and cannot be delayed. In elective situations, it is best to avoid intubation if possible (if the surgical procedure allows), instead using regional anesthesia or general anesthesia by mask or laryngeal mask airway (LMA). A randomized, prospective study demonstrated a much lower incidence of bronchospasm in children with URI managed with LMA rather than ETT (0% vs. 12.2%) (Tait et  al., 1998). The incidence of all respiratory complications was reduced by 50% for the LMA group (19% vs. 35%). LMA may be an excellent alternative for airway management for patients with URI if the planned surgical procedure and fasting status are compatible with its use. If intubation is indicated, it should be accomplished when the patient is at a deep plane of anesthesia using an ETT at least 1 size (0.5 cm) smaller than age would determine. Any IV induction agent is acceptable, with the most important guiding principle being that enough should be given to achieve a deep level of anesthesia. An alternative is mask induction of inhalation anesthesia with sevoflurane, with or without nitrous oxide, and oxygen. If the procedure is expected to be prolonged,

1114   P a r t  IV    Associated Problems in Pediatric Anesthesia

informed fashion in the decision-making process; however, “in the final analysis, the name of the game is clinical judgment and a degree of good fortune” (Berry, 1990).

Emergency surgery Yes

No

Proceed

Systemic symptoms

Reactive Airways Disease (Asthma)

No

Yes Delay 4–6 weeks

General anesthetic required No

Yes

Proceed

Endotracheal intubation required No

Yes

Proceed

Other risk factors

Etiologic Factors and Pathophysiology

No Proceed

RAD, or asthma, the most common chronic disease of childhood in industrialized countries, has received wide public attention in recent years because of increases in morbidity and mortality. Asthma is the major cause of restricted activity, absence from school, and hospital admission in children, and it is responsible for significant health care costs in the United States (Newacheck and Halfon, 2000). The prevalence of asthma among children is greater than among adults, and it has increased by an average of 4.3% each year between 1980 and 1996 (Akinbami and Schoendorf, 2002). Beginning in 1997, the survey questions used to determine the prevalence of asthma were changed, which resulted in a slightly lower prevalence than in previous years, but the increasing trend has remained constant (MMWR, 2000).

Delay 4–6 weeks

n  FIGURE 36-3. Algorithm for clinical decision making for a patient with upper respiratory tract infection. (From Martin LD: Anesthetic implications of an upper respiratory tract infection in children, Pediatr Clin North Am 41:121, 1994.)

heated humidification should be used, because use of dry gas may lead to inspissation of secretions. Adjunctive agents, such as IV lidocaine (1 mg/kg), an opioid, or both, decrease airway reflexes (Hirshman, 1983). The preoperative use of anticholinergic agents (i.e., atropine and glycopyr­ rolate) theoretically may block muscarinic receptors, thereby interrupting the airway reflex arc (Jacoby and Hirshman, 1991). Their properties as antisialogogues may be helpful; however, these agents have not been shown to be beneficial in prospective studies (Tait et al., 2007). The use of glucocorticoids experimentally has decreased viral-associated, tachykinin-induced airway edema formation, although glucocorticoids are not routinely prescribed in this clinical situation—in contrast to their use in the patient with RAD (Piedimonte et al., 1990). Dexamethasone, however, has been used in an effort to prevent postextubation croup, but these studies have been performed in critically ill children who had undergone prolonged intubation and mechanical ventilation in the intensive care unit, rather than in children with respiratory infections who are undergoing surgery. In this high-risk population, dexamethasone may be effective (Markovitz and Randolph 2002; Lukkassen et  al., 2006). If intubation is necessary, tracheal suction of URI-associated secretions after intubation and before extubation may decrease the chance of atelectasis and mucus plugging, although this hypothesis has not been studied. Management of children with URI requires a logical approach. When this issue arises in the preoperative period, the patient, parents, surgeon, and anesthesiologist must participate in an

Asthma is a chronic inflammatory disorder of medium and small airways in which many cell types play a role, including mast cells and eosinophils. These cells release mediators of inflammation that, in susceptible individuals, cause symptoms associated with variable airway obstruction and airway hyperreactivity that is partially or completely reversible spontaneously or with appropriate treatment. Understanding the important role of inflammation in the immunopathogenesis of asthma in recent years has changed the focus to a newer therapeutic approach using antiinflammatory agents. Among the immune regulatory pathways involved in the pathogenesis of asthma, 2 cytokines—interleukin-4 and interferon-γ—appear to be important in controlling immunoglobin E (IgE) production, which is critical in the allergic inflammatory process. In individuals with asthma, mast cells and eosinophils are attracted to airways and release cytokines and lipid mediators that cause inflammation (Goldstein et al., 1994). The interplay of allergens and irritants, mast cells, eosinophils and their mediators, and the end effects on pulmonary vessels and airways is depicted in Figure 36-4. These mediators include histamines, leukotrienes, prostaglandins, kinins, and cytokines. Airway obstruction in asthma results from a combination of several factors, including airway smooth muscle spasm, airway mucosal edema, hypersecretion, and mucus plugging of small bronchi and bronchioles (Djukanovic et al., 1990). These changes result in airway obstruction, increased work of breathing, uneven distribution of ventilation, and in severe disease, air trapping, hyperinflation, and ventilation-perfusion imbalance, which leads to hypoxemia, diaphragmatic fatigue, hypercapnia, and respiratory failure. In longstanding RAD, mast cells infiltrate airway smooth muscle and in combination with chronic inflammation may result in airway remodeling, potentiating bronchoconstriction and airway hyperreactivity (Brightling et al., 2002). A strong association exists between asthma and allergy. Up to 90% of children with recurrent wheezing respond positively to bronchoconstrictor challenge, especially when associated with atopy (Clough et al., 1991). An increased prevalence of asthma is reported among first-degree relatives of asthmatic

C h a p t e r 36    Systemic Disorders   1115 ANTIGEN

lgE MAST CELL IL-4

Proinflammatory cytokines, IL-4

T CELL B CELL

IL-5 IL-8

Histamine Leukotrienes

Tryptase

IL-5

Proinflammatory cytokines

Chemokines LTB-4 BRONCHOSPASM

NEUTROPHILS

Epithelial thickening New blood vessels Muscle thickening

MACROPHAGE EOSINOPHIL

Activated

cytokines

INFLAMMATION Substance P Activation of afferent neurons

Muscle constriction

Increased mucus production

BRONCHOSPASM

n  FIGURE 36-4. Pathophysiology of asthma.

subjects; over two thirds of children with asthma appear to have a familial predisposition (Clifford et al., 1989). Various environmental factors precipitate airway hyperreactivity and trigger asthma. Recent evidence has shown that obesity is associated with a proinflammatory state and is an independent risk factor for the development of asthma (Visser et al., 2001; Guerra et al., 2004). The onset of asthmatic symptoms is often associated with viral infection of the lower respiratory tract, particularly respiratory syncytial virus (RSV) infection in infants and children (Rooney and Williams, 1971). Severe viral bronchiolitis in infancy is significantly associated with the subsequent development of airway hyperreactivity and asthma, although familial factors cannot be ruled out (Gurwitz et al., 1981; Gern et al., 2005; Lemanske et al., 2005). The development of IgE antibody to RSV may have an important role in inducing an allergic response to the virus (Welliver et al., 1989; Rakes et al., 1999). Children who experience respiratory failure and mechanical ventilation during infancy and early childhood, such as those with bronchopulmonary dysplasia (BPD), neonatal repair of congenital diaphragmatic hernia, or severe viral bronchiolitis, develop and sustain airway hyperreactivity even without a family history of asthma (Mallory et al., 1989; Nakayama et al., 1991). Prematurity alone may be associated with a higher incidence of asthma in preadolescent children (von Mutius et al.,

1993). The primary site of airway obstruction and hyperreactivity in children with a history of neonatal respiratory failure appears to be in relatively small airways—in contrast to relatively large central airway obstruction and hyperreactivity in those with typical allergic (i.e., IgE antibody-mediated) asthma (Mallory et al., 1991). In these patients, airway spasm is characterized by a rapid drop in oxygen saturation of Hb (SpO2) without audible wheezing by auscultation. In children with RAD, parental smoking (passive smoking) increases the severity of symptoms and exacerbates airway hyperreactivity (Soussan et al., 2003). Intrauterine exposure to maternal smoking also increases the incidence of airway hyperresponsiveness in infants (Singh et al., 2003). Infants with gastroesophageal reflux and chronic esophagitis often develop airway hyperresponsiveness with or without chronic aspiration and resultant tracheobronchial inflammation (Sheikh et al., 1999). Contributing factors responsible for the development of airway hyperreactivity and asthma are listed in Figure 36-5.

Precipitating Factors for Reactive Airways Disease Viral lower respiratory infections, particularly those caused by RSV and influenza, sensitize airways and provoke airway hyperreactivity even in individuals who are nonasthmatic and nonallergic for as long as 6 weeks (Empey et al., 1976). Exposure

1116   P a r t  IV    Associated Problems in Pediatric Anesthesia Allergen exposure in infancy

Genetic predisposition

Atopy

Parental smoking

Airway hyperreactivity

Asthma

ARF/ventilator in infancy

LRI (RSV) in infancy

ALGORITHM FOR PHARMACOLOGIC TREATMENT OF REACTIVE AIRWAYS DISEASE ICS (200 mcg BDP*)

Pharmacologic Agents for Asthma The pharmacologic management of asthma consists of bronchodilators and antiinflammatory drugs and includes six different classes of drugs: corticosteroids, leukotriene inhibitors, β-adrenergic agonists, theophylline, cromolyn (or nedocromil), and anticholinergics. Recently published guidelines for treatment of asthma in children include a useful algorithm (Fig.  36-6) (Bacharier et al., 2008). Initial treatment for asthma most commonly consists of inhaled corticosteroids (ICSs) and leukotriene-receptor antagonists, with ICSs usually being the first-line drugs. Montelukast has been used widely in children. Other first-line alternatives are cromolyn

Box 36-6 Precipitating Factors for Bronchospasm Lower respiratory infection (adenovirus, RSV) Irritants (cigarette smoke, inhaled anesthetics) Allergens (inhaled) Emotional stress; fear and excitement Exercise, hyperventilation Cold or dry gas (anesthetic gases without humidification) Manipulation/mechanical stimulation of pharynx and larynx Gastroesophageal reflux

LTRA† (dose is age dependent)

Insufficient control

Increase ICS dose (400 mcg BDP)

n  FIGURE 36-5. Contributing factors to the development of RAD. ARF, Acute renal failure; LRI, lower respiratory tract infection; RSV, respiratory syncytial virus.

to dry, cold air can precipitate tracheobronchial constriction in subjects with asthma, presumably in response to reduced tracheal temperature caused by evaporative heat loss (Gilbert et  al., 1988). The same mechanism appears to be responsible for exercise-induced bronchospasm and precipitation of asthma with excitement, anxiety, and hyperventilation (McFadden and Gilbert, 1994). The time before and during the induction of anesthesia is uniquely suited to trigger bronchospasm in susceptible individuals because of the patient’s emotional stress, fear, and excitement. Hyperventilation results, with mouth breathing of dry anesthetic gas mixtures, airway irritation by volatile anesthetics, and mechanical stimulation of the pharyngeal and laryngeal mucosa by laryngoscopy and endotracheal intubation (Box 36-6).

OR

Add ICS to LTRA

Insufficient control

Increase ICS dose (800 mcg BDP) OR Add LTRA to ICS OR Add LABA

Insufficient control

Consider other options Theophylline Oral corticosteroid n  FIGURE 36-6. Algorithm for treatment of RAD. Once control is achieved, therapy may be stepped down to previous level. ICS, Inhaled corticosteroid; BDP, beclomethasone equivalent; LTRA, leukotriene receptor antagonist, for dose, see Table 36-10. (Modified from Bacharier LB, et al: Diagnosis and treatment of asthma in childhood: a PRACTALL consensus report, Allergy 63:5, 2008.)

(or nedocromil). Because of safety concerns, long-acting β2agonists such as salmeterol are now reserved for patients with poor control despite use of an ICS and leukotriene-receptor antagonists (Martinez, 2005). Theophylline has reentered the treatment algorithm if patients have exacerbations despite ICS and β-adrenergic agonists. Oral corticosteroids in brief high-dose pulses are reserved for patients with moderate to severe asthma that is unresponsive to combinations of the other drugs (Stempel, 2003). Drugs commonly used in children are listed in Table 36-10.

Corticosteroids ICSs have become popular for the treatment of asthma because of their potent antiinflammatory effect on the airways with limited systemic effects (as compared with oral steroids) and are a first-line regimen recommended in treatment guidelines. Regular use of an ICS allows effective control of symptoms and improvement in lung function, reduces airway inflammation, and results in a gradual reduction in airway hyperreactivity (Konig, 1988; Juniper et al., 1991). Although corticosteroids inhibit the in vitro proliferation of airway smooth muscle

C h a p t e r 36    Systemic Disorders   1117

TABLE 36-10. Drugs Used for Asthma Drug

Formulation

Dosage

Antiinflammatory Drugs Inhaled Corticosteroids Beclomethasone dipropionate (QVAR)

MDI, 40, 80 mcg/puff

2-4 puffs bid (5-11 yr)

Budesonide (Pulmicort)

DPI, 90, 180 mcg/inhalation

1-2 inhalations bid (6-17 yr)

Flunisolide (AeroBid)

MDI, 250 mcg/puff

2 puffs bid (6-15 yr)

Fluticasone proprionate (Flovent)

MDI, 44, 110, or 220 mcg/puff

1-2 puffs bid (max, 440 mcg/d) (4-11yr)

Fluticasone proprionate (Flovent Diskus)

DPI, 50 mcg/inhalation

1-2 inhalations bid (4-11 yr)

Triamcinolone acetonide (Azmacort)

MDI, 75 mcg/inhalation

2-4 inhalations bid (6-12 yr)

Prednisone/prednisolone

Oral tablets (1, 2.5, 5, 10, 20 mg)

Acute: 1 mg/kg q d/bid × 5-14 d

Prednisone/prednisolone (Prelone, Pediapred)

Oral liquid (1 mg/mL)

Chronic: 0.25-2 mg/kg qod Preoperative: 1 mg/kg/d × 3 d Max, 60 mg/d

Oral granules, 4 mg; chewable tablets, 4.5 mg

12 mos-5 yrs: 4 mg qd; 6-14 yrs: 4.5 mg qd

Solution for nebulization (10 mg/mL)

1-2 inhalations tid-qid

  Albuterol (Proventil, Ventolin)

MDI, 90 mcg/inhalation

2 puffs q 4-6 hr prn

  Albuterol (generic)

Nebulized solution (0.63, 1.25, or 2.5 mg/3 mL; 2.5 mg/0.5 mL)

0.63 or 1.25 mg tid-qid prn

  Levalbuterol (Xopenex)

Nebulized solution (0.31, 0.63, or 1.25 mg/3 mL)

0.31-0.63 mg tid prn (6-11 yrs)

Oral Corticosteroids

Leukotriene Receptor Antagonist Montelukast (Singulair) Cromolyn Sodium (Intal) Cromolyn (generic) Inhaled b-2 Agonists (Bronchodilators) Short-acting (SABA)

Long-acting (LABA)*   Salmeterol (Serevent)

DPI, 50 mcg/blister

50 mcg bid (≥4 yrs)

Theophylline (generic)

Oral solution, 80 mg/15 mL

<1 yr: max, 0.2 × (age in weeks) + 5 = dose in mg/kg/d

Tablet (immediate release), 100 mg

>1 yr: 10 mg/kg/d; max, 16 mg/kg/d

Capsule (sustained release), 125, 200, 300 mg Anticholinergics Ipratropium bromide (Atrovent)

MDI, 17 mcg/puff

3-14 yr: 1-2 puffs qid (for acute exacerbation); >14 yr: 2 puffs qid

Ipratropium bromide (generic)

Nebulized solution (500 mcg/2.5 mL)

125-250 mcg q 6-8 hr mixed with albuterol (5-12 yr) 250-500 mcg q 6-8 hr (>12 yr)†

*Recommended only for use in combination with inhaled corticosteroid for children >4-5 years of age. Data from Bacharier LB et al.: Diagnosis and treatment of asthma in childhood: a PRACTALL consensus report, Allergy 63:5, 2008; modified from Treatment guidelines from the medical letter 6:86, 2008. † Zorc JJ et al.: Ipratroprium bromide added to asthma treatment in the pediatric emergency department, Pediatrics 103:748, 1999. DPI, Dry powder inhaler; MDI, metered-dose inhaler; prn, as needed

cells from subjects without asthma, they do not do so in airway smooth muscle cells from patients who have asthma (Roth et al., 2004). Recommended doses of ICS generally have minimal effects on the HPA axis; however, high doses, especially of fluticasone, have resulted in reduction of cortisol levels and symptomatic adrenal insufficiency in children (Barnes and Pedersen, 1993; Todd et al., 2002; Sim et al., 2003). Oral or parenteral corticosteroids are most effective for acute exacerbations of asthma unresponsive to maximal bronchodilator therapy (Chapman et al., 1991; Schuh et al., 2000).

Leukotriene-Receptor Antagonists Leukotriene-receptor antagonists are a class of drugs developed for the prevention and treatment of bronchial asthma. They are an alternative first-line treatment or may be added if ICSs do not provide adequate control. The formation of leukotrienes through the 5-lipoxygenase pathway depends on lipoxygenation of arachidonic acid, a major constituent of cell membrane phospholipids, detached by phospholipase A2 activity. Leukotrienes are potent bronchial smooth ­muscle

1118   P a r t  IV    Associated Problems in Pediatric Anesthesia

c­ onstrictors; on a molecular basis, leukotrienes C4 and D4 (LTC4 and LTD4) are approximately 1000 times more potent than histamine (Undem and Lichtenstein, 2001). Bronchial smooth muscle constriction by leukotrienes is considered a major cause of asthmatic symptoms. Leukotriene-receptor antagonists (e.g., zafirlukast, montelukast) are selective highaffinity LT1-receptor antagonists (Jones et al., 1995). Their use is associated with increased exhaled nitric oxide (NO) (Straub et al., 2005). Montelukast has been reported to be effective as maintenance therapy in children with moderate to severe asthma, with or without concomitant steroid therapy, with minimal side effects (Knorr et al., 2001; Phipatanakul et al., 2003).

higher levels. The great variability of drug metabolism and the necessity for monitoring of blood levels are two reasons the use of theophylline has declined (Drugs for Asthma, 2002).

β2-Adrenergic Agonists

Cromolyn sodium does not have a bronchodilator effect; therefore it is exclusively a prophylactic agent and has no bearing on anesthetic practice. It attenuates bronchoconstriction caused by allergen, exercise, and bronchial challenge (Stempel, 2003). Nedocromil sodium has similar chemical and biological properties to cromolyn, which became available in the early 1990s (van Bever and Stevens, 1992). Cromolyn and nedocromil are thought to act on pulmonary mast cells and stabilize cell membranes. They reduce IgE antibody-induced release of inflammatory mediators, including histamine and leukotrienes, from activated mast cells (Douglas, 1985). Maintenance therapy with cromolyn or nedocromil is recommended in children with moderate to severe asthma.

β2-adrenergic agonists initiate their action on the receptor sites of airway smooth muscle cells and increase adenylate cyclase activity, which produces cyclic adenosine monophosphate (cAMP) from adenosine triphosphate (ATP) and results in smooth muscle relaxation and bronchodilation. For maintenance therapy in more severe RAD and for the prevention of exercise-induced bronchoconstriction, long­acting β2-agonists (LABAs) are recommended (Bacharier et al., 2008). Studies suggest a small risk of increased asthma-related deaths in patients older than 12 years of age who take LABAs, especially if they are used regularly without an ICS; thus, the FDA requested that a “black box warning” be added to the labels of salmeterol and salmeterol/fluticasone in 2003 (Nelson et al., 2006). An FDA meeting was later convened to consider safety concerns regarding LABAs in adults and children. Although few children younger than 12 years of age have been studied, the risk of asthma exacerbations and hospitalizations was higher in children 4 to 11 years of age who received a combination drug with LABA and ICS when compared with patients who received ICS alone (Bisgaard and Szefler, 2006; Kramer, 2009). Until further studies are undertaken, it has been recommended that LABA not be given as monotherapy but only in combination with ICS (Drazen and O’Byrne, 2009). This concern does not apply to the use of short-acting β2-agonists (SABAs), which are the drugs of choice for the treatment of intermittent episodes, acute exacerbations, and prevention of exercise-induced asthma in patients with RAD. The recommended doses of commonly used β2-agonists are listed in Table 36-10.

Methylxanthines Methylxanthines (e.g., theophylline) produce bronchodilation by inhibiting adenosine-induced bronchoconstriction in asthmatic patients—not as was formerly thought, by competitively inhibiting phosphodiesterase, which metabolizes cAMP (Holgate, 1984). Although theophylline was the mainstay of pediatric asthma therapy in the 1970s and 1980s, it is no longer used for therapy in acute exacerbations of RAD (Rooklin, 1989). Theophylline still has a role in decreasing the severity of persistent bronchospasm, especially that which occurs at night. Theophylline has a narrow therapeutic index, with greatest efficacy at serum levels of 5 to 15 mcg/mL. Levels greater than 20  mcg/mL are associated with symptoms of toxicity, such as nausea, gastroesophageal reflux, irritability, learning difficulties in children, and headache (Creer and Gustafson, 1989; Ellis, 1985). Vomiting, tachyarrhythmias, and seizures can occur at

Anticholinergics Ipratropium bromide is an atropine derivative and is available as a metered-dose inhaler and as a nebulizable solution. Ipratropium has a slower onset of action than β2-agonists, but the duration of action is longer (up to 8 hours). Side effects are uncommon.

Cromolyn Sodium and Nedocromil Sodium

Preanesthetic Consideration The goal of the preoperative assessment of children with asthma is to ensure that each patient receives optimal treatment before reaching the operating room. The patient’s history, physical examination, and laboratory tests are all helpful to determine whether the patient’s condition is adequately managed. Children with RAD rarely require preoperative pulmonary function testing, but they are commonly monitored by pulmonology or allergy and immunology services with frequent assessment of pulmonary function testing (i.e., spirometry with flow-­volume curves). Some families use the peak expiratory flow rate for home assessment. If this is so, the family should be queried to ensure that the peak expiratory flow rate is maximized. Careful history taking is the single most important element of the preoperative evaluation of children with asthma. The profile of a typical acute episode, precipitating factors, and time of the most recent episode of asthma should be obtained. Previous and current drug therapy, dosages, effectiveness, and side effects, if any, should also be documented. Specific points of importance in the history include the following:



1. Determine whether the child has had episodes of bronchospasm and bronchodilator treatment in the previous 4 to 6 weeks. Ideally, elective surgery should be postponed for at least 4 to 6 weeks after an episode of symptomatic asthma, because airway hyperreactivity may be worsened after acute exacerbations, and pulmonary gas exchange may still be impaired because of bronchoconstriction, mucosal edema, and mucus plugs. 2. Determine whether the patient has a recent history of a URI or if the symptoms of URI still exist. URI in children with RAD is often associated with the exacerbation of bronchospasm and requires a more conservative approach

C h a p t e r 36    Systemic Disorders   1119

than in children who do not have asthma. Optimally, the child with a history of RAD should be free of URI symptoms for 4 to 6 weeks before an elective procedure, unless the URI symptoms recur so often that an asymptomatic period is difficult to attain. If the child has had a lower respiratory infection, such as influenza, within the previous 6 weeks, the postponement of scheduled surgery should be seriously considered, because airway hyperreactivity would be exaggerated as long as 6 weeks, even in nonasthmatic patients. 3. Ascertain the child’s steroid requirements over the previous year and the possible need for perioperative stress-dose steroid coverage (see preceding discussion). Children who often have bronchospasm that is poorly controlled with maximal therapy and require repeated courses of oral steroids may benefit from a short preoperative course of prednisone (1 mg/kg per day to a maximum of 60 mg once daily for 3 days, including the day of surgery), especially if endotracheal intubation is planned. Physical examination should be focused on careful auscultation of the chest for clinical evidence of bronchoconstriction (i.e., expiratory wheezing), use of the accessory muscles of respiration, and a prolonged expiratory phase. During severe episodes of bronchospasm, air movement may become so limited that wheezing may be barely audible. Patients with a history of BPD and asthma are most likely to have lower-airway obstruction and small airway hyperreactivity; wheezing and rhonchi may not be present (see Chapter 17, Neonatology for Anesthesiologists). The preanesthetic level of oxygen saturation should be obtained with a pulse oximeter while the child is breathing room air to determine the baseline oxygen saturation and assess for any preexisting hypoxemia. This information is exceptionally valuable for the postoperative assessment of lung function and gas exchange.

Anesthetic Management The anesthesiologist must get to know the child with asthma and his or her parents and gain their confidence to minimize the child’s anxiety before anesthesia induction. The child should be well sedated to avoid struggle and hyperventilation, which can provoke exercise-induced asthma. Midazolam, which may be administered transmucosally (i.e., orally, nasally, or rectally) in infants and young children and orally or intravenously (if IV access is present) in older children, works well for sedation. A β2-adrenergic agonist may be given prophylactically using a metered-dose inhaler or nebulizer before induction (Table 36-10). Otherwise, the drug can be given after the induction of anesthesia through the ETT using the metered-dose inhaler and an aerosol chamber inserted in between the ETT adapter and the anesthesia circuit. The anesthetic approach is similar to that for children with URIs. After applying standard monitors (a minimum of a pulse oximeter and precordial stethoscope if the child is uncooperative), the inhalation induction should be smooth and progress swiftly with sevoflurane and nitrous oxide (see Chapter 11, Monitoring). For infants and young children, heated humidification should be used if available; the dry gas mixture from the anesthesia machine is a perfect environment for provocation of bronchospasm secondary to irritation and reduced tracheal

temperature from evaporative heat loss of the tracheal mucosa in a child with asthma (McFadden and Gilbert, 1994). For IV induction, propofol may be a better agent of choice than thiopental because it suppresses airway reflexes as compared with barbiturates, although thiopental, despite risk of histamine release, is not necessarily contraindicated in patients with asthma (Brown et al., 1992; Gal, 1994). Propofol may also produce bronchodilation in patients with other types of airway disease (Conti et al., 1993). Regardless of the drug chosen, it is important to give sufficiently large IV doses to blunt the reflex response, and sevoflurane should be added before the peak effect of the IV agent is lost. Whenever possible, endotracheal intubation should be avoided in patients with asthma, because the ETT stimulates large airway irritant receptors and can trigger bronchospasm (Hirshman, 1983). When no contraindications exist, an LMA is a good choice for patients with RAD, as its use avoids the laryngeal and tracheal stimulation of intubation (Groudine et al., 1995). It may also be prudent to avoid anesthetic agents that might release histamine (e.g., atracurium and morphine), although clinical evidence that such drugs actually cause intraoperative bronchospasm is limited. An anesthetic technique using a volatile anesthetic may be preferable to a balanced technique (i.e., nitrous oxide, opioid, and muscle relaxant) for patients who have asthma because of the salutary bronchodilating properties of volatile agents. Regional anesthesia can be combined with inhalation anesthesia with sevoflurane or isoflurane. Desflurane may cause more airway irritation in patients with RAD and lead to broncho­ spasm and associated coughing, especially during emergence (von Ungern-Sternberg et al., 2008).

Intraoperative Wheezing The differential diagnosis of intraoperative wheezing includes “light anesthesia,” kinked ETT, mainstem bronchial intubation, increased airway secretions, foreign body in the airway, pulmonary edema, embolus, and aspiration. In the child with RAD, wheezing can result from exacerbation of airway hyperreactivity and requires immediate attention. The treatment of intraoperative bronchospasm is detailed in Box 36-7. Treatment should begin after chest auscultation to confirm that there are bilateral breath sounds and therefore no mainstem intubation. The first step includes increasing the inhaled concentration of oxygen and deepening the level of anesthesia with volatile anesthetics, or administering IV propofol or

Box 36-7 Treatment of Intraoperative Bronchospasm Confirm diagnosis (exclude mainstem bronchus intubation, mucus plug, pneumothorax, anaphylaxis, and congestive heart failure). Deepen anesthesia with volatile agent. Administer inhaled β-agonists and ipratropium. Consider propofol or ketamine to further deepen anesthesia. Consider IV lidocaine, atropine, or both. Administer an IV corticosteroid. Modify ventilation to avoid stacking breaths, gas trapping, and barotrauma.

1120   P a r t  IV    Associated Problems in Pediatric Anesthesia

ketamine (0.5 to 2 mg/kg), a known bronchodilator (Corssen et al., 1972; Hirshman et al., 1979). Lidocaine (1 mg/kg) may also be given intravenously to reduce airway reactivity at the earliest sign of bronchospasm. Administration of muscle relaxant and suctioning of the ETT may be performed if the patient is intubated. The second step consists of the administration of β2-agonists given by a metered-dose inhaler and adapter through the ETT, followed by the squeezing of the anesthesia bag manually to provide a vital-capacity maneuver to distribute the bronchodilator mist to the tracheobronchial tree. Because only 5% to 10% of the administered dose may reach the end of the ETT and contact the airway, 4 to 8 puffs of the β2-agonist should be administered through the ETT. This maneuver may be repeated two or three times. Parasympatholytic agents (atropine, 0.02 to 0.03 mg/kg) or antihistamines (diphenhydramine, 0.5 mg/kg) are indicated when wheezing is associated with increased vagal tone or histamine release, respectively. The development of hypotension, urticaria, or flushing should lead to the consideration of anaphylaxis. Corticosteroids (e.g., 2 mg/kg of IV hydrocortisone) should be given, and circulation should be supported with appropriate vasoactive agents.

Techniques of Extubation At the conclusion of surgery and anesthesia, the patient with asthma can be extubated “deep” or “awake” to avoid laryngo­ spasm. Upper airway obstruction caused by soft-tissue collapse in the pharynx is the major disadvantage of deep extubation. Deep extubation can be accomplished safely provided the maintenance of upper airway patency was satisfactory during the induction of anesthesia before intubation and that no excessive secretions or blood are in the airway. If maintaining airway patency was difficult during induction, the patient’s airway may become obstructed during emergence. If this is the case, airway patency may be facilitated by prophylactic placement of an oropharyngeal or nasopharyngeal airway, well lubricated with lidocaine or lubricant jelly, when the patient is still deeply anesthetized. Successful deep extubation is facilitated by the attainment of spontaneous breathing before attempted extubation. For a successful “awake” extubation, prophylactic treatment with the inhalation of a β2-agonist must be given even if a dose was previously given during or after the induction of anesthesia. Tracheal suction of any secretions before emergence may decrease coughing caused by migration of mucus plugs. Lidocaine (1 mg/kg) given intravenously on emergence is helpful in minimizing tracheal stimulation as the patient awakens. The use of IV atropine (0.02 mg/kg), given for its vagolytic and bronchodilator effects, may be an additional safety precaution before extubation.

involved infants born at a mean gestational age of 34 weeks, all of whom had received excessive concentrations of oxygen during mechanical ventilation with a primitive ventilator by modern standards, over time, BPD has been seen in infants who had prolonged barotrauma (or volutrauma) in the absence of “excessive” oxygen. Early series were characterized by a high incidence of mortality with persistent respiratory symptoms and oxygen requirement beyond 4 weeks of age. Chest radiographs were abnormal and characterized by hyperinflation of the lungs with focal areas of increased density. Northway et al. called this condition BPD to “emphasize the involvement of all the tissues of the lungs in the pathologic process” (Northway et al., 1967; Northway, 2001). The number of infants with BPD has not decreased over the past two decades despite improved neonatal intensive care, probably because of the survival of more infants who are premature. However, the clinical picture has changed with the advent of antenatal steroids, the use of surfactant therapy, and advances in ventilatory strategies for reducing volutrauma and ventilatorinduced lung injury, including noninvasive techniques such as nasal continuous positive airway pressure (Geary et al., 2008, Greenough et al., 2008). Currently, most infants who develop BPD are born at 24 to 28 weeks’ gestation and are rarely born later than 32 weeks’ gestation, whereas the mean gestational age of Northway’s original series was 34 weeks (Hazinski, 1990). Because of changes in neonatal intensive care and affected patient population, many aspects of BPD have changed, including the definition, theories of pathogenesis, pathology, and clinical picture (Jobe and Ikegami, 2000; Jobe and Bancalari, 2001). Infants with BPD today are likely to have a minimal respiratory distress syndrome that does not progress after surfactant administration. The reason for prolonged ventilation in these premature infants is more commonly apnea or poor respiratory effort, which may be related to immaturity of central respiratory control mechanisms. These infants rarely require the high airway pressures and high oxygen concentration that led to the “old BPD.” This newer clinical picture had been referred to as chronic lung disease or “new BPD,” but it is now simply called BPD. The current definition of BPD is oxygen dependence for at least 28 postnatal days. Evaluation for the presence of BPD and grading of severity is assessed at 36 weeks’ postconceptual age in infants born at fewer than 32 weeks’ gestation and at 56 days after birth in infants born at or after 32 weeks’ gestation (Jobe and Bancalari, 2001). BPD is graded as mild, moderate, or severe, based on oxygen requirement and the need for ventilatory support (Table 36-11). Prevalence ranges from 67% in the smallest weight group to 1% in the largest, with an overall prevalence of 20% of infants born at less than 1500 g (Lemons et al., 2001; Bancalari et al., 2003).

Pathogenesis

Bronchopulmonary Dysplasia BPD is a chronic disease of lung parenchyma and small airways with chronic respiratory insufficiency that occurs in prematurely born infants (see Chapter 17, Neonatology for Anesthesiologists). As originally described by Northway’s group (1967), BPD develops after a period of acute and subacute ventilator-induced lung injury and oxygen toxicity, in prematurely born infants with severe respiratory distress syndrome (Hazinski, 1990). Although Northway’s original series

In the past, the development of BPD was associated with a condition that caused respiratory failure in the neonatal period (e.g., prematurity with respiratory distress syndrome, meconium aspiration syndrome, or congenital diaphragmatic hernia). Mechanical ventilation with high concentrations of oxygen (i.e., an acute insult to immature lungs) was employed, usually lasting longer than 1 week. Oxygen free radicals, which are not well handled by an immature antioxidant host-defense system in the neonatal lungs, can cause direct cellular injury (Ackerman, 1994). Although much lower concentrations of oxygen are now used

C h a p t e r 36    Systemic Disorders   1121

TABLE 36-11. Assessment of the Severity of Bronchopulmonary Dysplasia* Grade

Fio2 and Ventilatory Support GA at Birth <32 wks Assessed at 36 Weeks’ PCA

GA at Birth ≥32 wks Assessed at 56 Days Postnatally

Mild

0.21

0.21

Moderate

0.22-0.30

0.22-0.30

Severe

>0.30 and/or continuous positive airway pressure or mechanical ventilation

>0.30 and/or continuous positive airway pressure or mechanical ventilation

Modified from Baraldi E, Filippone M: Chronic lung disease after premature birth, N Engl J Med 357:1946, 2007. *Applies to patients who have been treated with >21% oxygen for at least 28 days. GA, Gestational age; PCA, postconceptual age.

than in the past, even room air (21% oxygen) is relatively hyperoxic for a premature infant whose in utero Po2 is less than 30 mm Hg (Hazinski, 1990). Excessive hydration and patent ductus arteriosus with increased pulmonary fluid have been recognized as additional important factors contributing to the development of BPD (Gerhardt and Bancalari, 1980; van Marter et al., 1992). The current theory of the mechanism of injury in BPD also emphasizes the role of infection and inflammation (Gonzalez et al., 1996; Sadeghi et al., 1998). Recurrent bacterial or viral infections in these infants may cause persistent alveolitis, which worsens alveolar and airway damage (Rojas et al., 1995; Hannaford et al., 1999). Multiple markers of inflammation (e.g., lipid mediators, proteases, oxygen free radicals, and cytokines) are elevated (Bose et al., 2008; Ryan et al., 2008). Nutritional deficiencies may also play a role (Sosenko et al., 2000; Geary et al., 2008). Immature, inflamed lungs with decreased compliance are most susceptible to high-volume (i.e., volutrauma) and lowvolume (i.e., shear stress trauma) trauma with marked distortion and distention of terminal bronchioles at high positive pressures (Hazinski, 1990). In earlier pathologic examination of lungs of infants dying with BPD, peribronchiolar fibrosis and smooth muscle thickening were seen. They were also found in animal models exposed to prolonged positive pressure ventilation and hyperdistention (Coalson, 1999). The pathology now seen in extremely premature infants reflects the immature state of their pulmonary parenchyma, with enlarged and simplified alveolar structure and reduced numbers of capillaries that are dysmorphic in appearance. Therefore, the new BPD is now regarded primarily as a disorder of lung development with superimposition of mechanical factors. Fibroproliferation may still occur but is more variable. Changes in larger blood vessels are less prominent with less indication of pulmonary hypertension than seen in old BPD. Airway smooth muscle hyperplasia may still occur, but it is also more variable (Coalson, 2006). After this damage has occurred to immature lungs, infants may require prolonged mechanical ventilation and high oxygen concentration for weeks or months, despite having not required high oxygen concentrations in the first few weeks of life. Although less common than with old BPD, progressive respiratory failure with associated pulmonary hypertension, with or without cor pulmonale, may follow.

Even after the perinatal period, RAD persists in infants with BPD. Mallory et al. (1991) studied lung function longitudinally in infants with moderate to severe BPD during the first 4 years of life with the forced deflation technique and found that airway hyperresponsiveness or hyperreactivity continued to be present in all children studied. They postulated that airway hyperreactivity is an important etiologic factor for the pathogenesis of lower airway obstruction in BPD. Attributing the cause of RAD to BPD is confounded by the recent evidence that antenatal steroid therapy to hasten lung maturation is a risk factor for the development of RAD between the ages of 3 and 6, as well as by the fact that the rates of RAD-like symptoms occur more often in prematurely born children than in those born at term (Doyle, 2006; Pole et al., 2009). Most long-term studies of lung function are reports of infants who had the old form of BPD. Little information is available regarding long-term outcomes of infants with new BPD.

Preanesthetic Considerations Most infants with moderate to severe BPD remain dependent on oxygen, with or without CPAP, or dependent on a ventilator beyond 4 weeks of age. They have persistent lower airway obstruction and airway hyperreactivity (Mallory et al., 1991). Tachypnea and dyspnea may be intermittently or chronically present. Growth failure because of chronic hypoxia despite oxygen therapy may occur, as well as cor pulmonale associated with pulmonary hypertension (Hazinski, 1990). Wheezing may or may not be present on auscultation, because the site of airway hyperreactivity is primarily in the periphery of the lungs as a result of increased thickness of the airway wall (Tiddens et al., 2008). In addition to lower airway obstruction primarily involving small airways, infants who have been intubated for prolonged periods sometimes develop large airway disease such as subglottic stenosis (that may or may not be recognized), tracheomalacia, and bronchomalacia (Miller et al., 1987; McCubbin et al., 1989). A later study also found a greater degree of upper airway obstruction in children with history of BPD, as compared with age-matched children with asthma (Sadeghi et al., 1998). Infants with mild forms of BPD improve with age and may become asymptomatic, but airway hyperreactivity may persist. Parents of such an infant may not be aware of the history of BPD even when their child received prolonged mechanical ventilation as a neonate. It is appropriate, therefore, to assume that a child has or had BPD and has RAD if he or she was born prematurely and was mechanically ventilated for more than 1 week during the neonatal period. Inguinal hernia is often present in infants with BPD, probably owing to prematurity and continually increased abdominal pressure that results from airway obstruction and increased inspiratory effort. Prematurely born infants may require postoperative admission for monitoring, because they have an increased risk of postoperative apnea as discussed in Chapters 3, Respiratory Physiology in Infants and Children; 17, Neonatology for Anesthesiologists; and 18, Anesthesia for General Surgery in the Neonate. As with asthmatic patients, careful history taking is of utmost importance before anesthetizing an infant with BPD or a history of BPD. These patients may have failure to thrive (a sign of chronic hypoxia). With lower respiratory tract infection, worsening of symptoms or even respiratory failure may occur. The patient may be taking SABAs or other treatments for asthma. Other medications may include diuretics. A family history of

1122   P a r t  IV    Associated Problems in Pediatric Anesthesia

allergy and asthma is significant, because premature birth may be linked to smooth muscle hyperresponsiveness and asthma (Bertland et al., 1985). Relatively common surgical conditions in infants and children with BPD or a history of BPD include inguinal hernia, direct laryngoscopy and bronchoscopy for subglottic stenosis, and surgical procedures of the larynx for the complications of prolonged intubation or tracheostomy (e.g., excision of granuloma or laryngotracheoplasty).

al., 2008). Of the other 1000 documented mutations, 20 account for most of the remaining 30% of cases, with the prevalence of mutations varying among ethnic groups (Tsui and Zielenski, 2007). Although the type of CFTR mutation does correlate with pancreatic function, poor correlation exists between the type of mutation and the severity of lung disease. Evidence suggests that other genetic polymorphisms influence the severity of lung disease (Drumm et al., 2005).

Anesthetic Management

Pathogenesis

Anesthetic management of infants and children with BPD or a history of BPD is similar to that for those with asthma. Before anesthetizing a child in this population, it is imperative to obtain a baseline oxygen saturation measurement with a pulse oximeter (SpO2), although a normal oxygen saturation level does not necessarily guarantee the absence of lung dysfunction. Many infants and young children with a history of BPD maintain remarkably good SpO2 values, presumably because of hypoxic pulmonary vasoconstriction (HPV). The infant with BPD with near-normal SpO2 in room air may develop marked desaturation after induction with sevoflurane, presumably because of a loss of HPV under general anesthesia, although HPV in healthy human volunteers may be insignificant (Benumof, 1994). In these patients, oxygen saturation may be maintained better with IV induction and maintenance techniques using opioids and propofol; however, no evidence supports this management protocol. Preoperative prophylactic treatment with a β2-adrenergic agonist by a metered-dose inhaler or handheld nebulizer may be beneficial for patients with potential airway hyperreactivity to prevent perioperative bronchoconstriction. In intubating a child with a history of mechanical ventilation, it is prudent to start with an ETT one size (0.5 mm inner diameter) smaller than the appropriate size for the age in anticipation of subglottic narrowing, which may be the result of prolonged intubation. If rapid sequence intubation is required because of fasting violation or intestinal obstruction, desaturation may be rapid when apnea occurs, and gentle ventilation by mask with maintenance of cricoid pressure may be necessary to maintain saturation if intubation is not rapidly accomplished.

CF is characterized by exocrine gland dysfunction that results in chronic pulmonary disease, pancreatic dysfunction, and abnormalities in electrolyte reabsorption in the sweat duct, with increases in sweat sodium and chloride concentrations and electrolyte imbalance. The fact that CF patients have sweat chloride levels in excess of 60 mEq/L (normal is less than 40  mEq/L), as measured by pilocarpine iontophoresis, is the basis of the sweat chloride test, which is still the gold standard for making the diagnosis of CF. Increasingly, the diagnosis of CF is being made earlier because of newborn screening techniques that are designed to detect elevated serum levels of immunoreactive trypsinogen and the most common CFTR mutations (Comeau et al., 2007). Such screening accounted for 21.6% of the diagnoses of CF in the United States in 2006 (Cystic Fibrosis Foundation, 2007). Even if these screenings are found to be positive, two sweat chloride test results of greater than 60 mEq/L are necessary to make the diagnosis of CF. Significant clinical manifestations of CF, other than pulmonary disease, include those listed in Table 36-12. Pulmonary disease is the most common cause of morbidity and death. Enhanced absorption of sodium across the airway

Cystic Fibrosis Cystic fibrosis (CF), an autosomal-recessive disorder, is the most common life-limiting inherited disorder among whites (Rosenstein and Cutting, 1998). In the United States, the gene frequency (heterozygotes) in whites is approximately 1 in 28; it is uncommon among Hispanics (1 in 46) and African Americans (1 in 65) and lowest among Asians and Native Americans (1 in 90) (Hamosh et al., 1998). The disease incidence among whites is approximately 1 in 2500 live births. Owing to early diagnosis and aggressive treatment over the past 40 years, in 2007 the median survival of a CF patient had increased to 37 years (Cystic Fibrosis Foundation, 2007). In 1985, Tsui et al. localized the gene responsible for the manifestation of CF to 250 kilobases on the long arm of chromosome 7. The deletion of three base pairs, removing a phenylalanine residue at position 508 (d508) from a 1480-amino acid protein called CF transmembrane conductance regulator (CFTR), a cAMP-dependent chloride ion channel, accounts for approximately 70% of CF chromosome abnormalities (Moskowitz et

TABLE 36-12. Organ System Involvement in Cystic Fibrosis Organ System

Incidence (%)

ENT Pansinusitis

90-100

Nasal polyps

20

Gastrointestinal Pancreatic Enzyme deficiency

85-90

Diabetes secondary to pancreatic failure

15

Intestinal Meconium ileus (newborn)

7-20

Distal intestinal obstruction syndrome (includes intussusception)

10-30

Rectal prolapse

20

Gastroesophageal reflux disease

50

Liver Hepatic failure

5-20

Coagulopathy caused by vitamin K deficiency

100 if untreated

Pulmonary Pneumothorax caused by bleb rupture

5-8

C h a p t e r 36    Systemic Disorders   1123

epithelium and failure to secrete chloride and fluid toward the airway lumen is thought to lead to dehydration and thickening of airway mucus and abnormal mucociliary clearance with subsequent bronchial inflammation and infection. Patients are initially colonized with Haemophilus influenzae and then by Staphylococcus aureus, and eventually by the mucoid variant of Pseudomonas aeruginosa. Colonization with Aspergillus and atypical mycobacteria may occur. The chronic infection in the periphery of the tracheobronchial tree results in bronchiolitis, which may lead to airway hyperresponsiveness, bronchiectasis, lobar or segmental atelectasis, and pneumothorax. Advanced disease is associated with destruction of airway architecture, fibrosis of lung parenchyma, and development of abscesses. Hemoptysis and eventually, cor pulmonale and respiratory failure ensue (Eckles and Anderson, 2003). Small airway obstruction, hyperinflation, and ventilationperfusion imbalance are the most common and important pulmonary changes in children with moderate-to-severe CF. The early signs of lung dysfunction include a reduction in maximum expiratory flow rates at low lung volumes (e.g., FEF25-75, FEF50, and FEF75) and an increase in the ratio of residual volume to total lung capacity (RV/TLC) (see Chapter 3, Respiratory Physiology in Infants and Children). Airway hyperreactivity is often present, probably in response to airway inflammation. Some patients have a good response to bronchodilators, but others have inconsistent or even paradoxical responses, sometimes worsening airway function because of the relaxation of airway smooth muscles and resultant increases in airway collapsibility (Brand, 2000).

Treatment Patients with CF take multiple medications. In 2007, a committee organized by the Cystic Fibrosis Foundation published guidelines for long-term medication used for optimal pulmonary symptom control in CF patients older than 6 years (Flume et al., 2007). Patients with a prominent bronchospastic component are on long acting β2-agonist therapy (Hordvik et al., 2002). They often take inhaled or oral antibiotics for prophylaxis or treatment of pulmonary infection. Patients infected with P.  aeruginosa often take aerosolized tobramycin, which when administered on a bimonthly basis has been shown to preserve pulmonary function and reduce hospitalization (Ramsey et al., 1999). Patients with infectious exacerbations are treated with IV antibiotics in a hospital or at home. Chest physiotherapy several times per day is a mainstay of CF treatment. Inhaled mucolytics (N-acetylcysteine) have long been used to decrease the viscosity of pulmonary secretions, but little is found in the literature documenting their efficacy (Duijvestijn and Brand, 1999). Inhaled hypertonic saline, which is thought to improve fluidity of pulmonary secretions, has been shown to decrease the frequency of pulmonary exacerbations (Flume et al., 2007). Dornase alfa (i.e., human recombinant DNase), which dissolves deoxyribonucleic acid (DNA) released from neutrophils, has improved pulmonary function and reduced the frequency of infection (Jones et al., 2003; Flume et al., 2007). High-dose ibuprofen therapy has been found to slow the decline of forced expiratory volume during the first second of measurement (FEV1) in children and adolescents if started when the FEV1 is greater than 60% of that predicted (Lands and Stanojevic, 2007). This effect is related to reduction in inflammation (Nichols et al., 2008). Doses between 20 and 30 mg/kg every 6 hours are administered to achieve plasma concentrations of 50 to 100 mcg/mL

(Arranz et al., 2003). Ibuprofen treatment has been associa­ ted with a significant increase in gastrointestinal (GI) bleeding com­pared with patients who are not taking ibuprofen, although the absolute number of patients affected is small (Konstan et al., 2007). Patients with pancreatic impairment require pancreatic enzyme replacement.

Preanesthetic Considerations Common surgical indications in infants and children with CF are listed in Table 36-13. Although some patients with CF require surgery in the neonatal period for meconium ileus, no special anesthetic considerations are necessary. Management of children with CF is a challenge to the anesthesiologist. These patients are often frail and malnourished. Decreased plasma albumin levels may affect anesthetic potency. Intravascular volume may be diminished because of chronic diarrhea, poor oral intake, and diuretic therapy. Electrolyte imbalance may result from excessive chloride and sodium losses. Pulmonary function ranges from near normal without airway obstruction to severe obstruction, air trapping, hypoxemia, and hypercapnia. Copious thick secretions and resultant ventilation-perfusion imbalance may prolong mask induction with volatile anesthetics. Secretions may irritate the larynx and precipitate laryngospasm. Nasal polyps may block the nasal airway completely during mask induction. Pathophysiologic considerations in patients with CF that may affect anesthetic management are listed in Table 36-14. The preoperative evaluation should include the assessment of pulmonary function by history, physical examination, and pulmonary-function testing. Pulmonary-function testing should include assessments of maximum expiratory flow­volume curves, lung volume measurements, and response to bronchodilators. An increase in TLC and the RV/TLC ratio with decreased vital capacity indicates the presence of hyperinflation and air trapping. Lower airway obstruction with small airway involvement is demonstrated when FEF25-75, FEF50, and especially FEF75 are markedly decreased from predicted values. A recent preoperative chest radiograph is needed in patients with moderate to severe pulmonary disease. Preoperative oxygen saturation should be obtained by means of pulse oximetry in room air for postoperative comparison. Recent tracheal

TABLE 36-13. Surgical Indications for Patients with Cystic Fibrosis Conditions

Typical Age Range

Meconium ileus or equivalent

1 d-3 yr

Nasal polyp/sinusitis

10-18 yr

Other procedures

10-18 yr

  Bronchoscopy Feeding gastrostomy; central venous access (port or PICC)   Lobectomy or thoracoplasty/thoracoscopy Organ transplantation (double lungs; heart-lungs) PICC, Peripherally inserted central catheter.

1124   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-14. Pathophysiology of Cystic Fibrosis: Effect on Anesthetic Management Pathophysiology

Possible Outcome

Pulmonary Dysfunction Airway obstruction V/Q imbalance

Prolonged mask induction

Copious secretions Airway hyperreactivity

Laryngospasm, bronchospasm

Nasal polyp

Upper airway obstruction

Gastrointestinal and Hepatobiliary Disorders Decreased serum albumin levels

Increased drug potency

Coagulopathies

Increased bleeding

Diabetes or glucose intolerance

Hyperglycemia, acidosis

Abnormal sweat gland function

Electrolyte imbalance

Cor pulmonale

Hemodynamic instability; dysrhythmia

­culture results should be reviewed as a guide to choice of perioperative antibiotic therapy. In patients with significant lower airway obstruction and air trapping, preoperative arterial blood gas measurement is recommended to assess the degree of hypoxemia more accurately and to evaluate the presence of hypercapnia and acid-base status. In those with long-­standing hypoxemia, pulmonary hypertension and cor pulmonale should be suspected. These patients should have preoperative electrocardiography and echocardiography to evaluate myocardial function and reserve. Blood sugar, liver-function tests, and coagulation studies may be indicated. Eight percent of CF patients between the ages of 11 and 17 develop CF-related diabetes mellitus as a result of fibrosis and fatty infiltration of the pancreas (Cystic Fibrosis Foundation, 2007). Some, but not all, may require insulin, because some insulin continues to be produced. Malabsorption and liver disease may be associated with abnormal coagulation. Additional vitamin K supplementation may be necessary to correct coagulopathy. The patient’s pulmonary physician should be consulted to ensure that the patient’s condition is optimized as much as is possible before surgery. Ibuprofen should be stopped 2 days before surgery to allow its inhibitory effects on platelet aggregation to dissipate. The child with CF and his or her family are exceedingly knowledgeable regarding the pathogenesis and treatment of the disease. A lack of knowledge of CF in general, and of the patient’s past history and present conditions in particular, at the time of the preoperative visit could quickly undermine the confidence of the family in the anesthesiologist. More importantly, the patient with CF is often petrified by the thought of death under anesthesia, in part because of a longheld belief in the pulmonary medicine community that general anesthesia leads to prolonged deterioration of pulmonary function (Price, 1986). It is therefore prudent for the anesthesiologist to gain the patient’s and the parents’ confidence and administer preoperative sedation such as oral midazolam (Della Rocca, 2002). Opioid premedication should be avoided in severe cases because of possible respiratory depression and hypoxemia.

Anesthetic Management Because of copious secretions in affected patients, it is preferable to schedule surgery later in the day to allow enough time for ambulation and chest physiotherapy in the morning to ­facilitate expectoration of secretions retained overnight. The baseline oxygen saturation in room air is measured with a pulse oximeter before administering oxygen and anesthetics. In patients with significant pulmonary involvement, IV access should be established before the induction of anesthesia because of prolonged mask induction and possible nasal obstruction from nasal polyps. An anticholinergic may be given during induction. Concern about excessive drying of secretions is unfounded, because atropine decreases secretions without changes in viscosity and has not been a significant problem in clinical practice (Lamberty and Rubin, 1985). IV propofol may be preferred to thiopental because it is less irritating to the upper airways and actually causes bronchodilation. Ketamine, despite its bronchodilating properties, is relatively contraindicated because it tends to increase secretions and may cause laryngospasm. Fifty percent of children with CF have gastro­ esophageal reflux disease and may require rapid-sequence intubation, although chronic treatment with H2-blockers and or a proton-pump inhibitor along with an adequate fasting interval mitigates the risk of aspiration. Inhalation induction is usually satisfactory in young children with mild lung disease. Anesthetic gases should be heated and humidified to prevent irritation of the upper airways and laryngospasm and to avoid drying and inspissation of secretions. When trapped gas volume is suspected or proven by pulmonary function test, nitrous oxide should be avoided to prevent expansion of the emphysematous area and the potential danger of bleb rupture. Endotracheal intubation with muscle relaxation is mandatory in patients with severe respiratory involvement, although the anesthesiologist should be exceedingly careful not to overdistend already air-trapped lungs. When a nondepolarizing muscle relaxant is chosen, the effect of aminoglycoside antibiotics to prolong the duration of action of such drugs must be kept in mind, and monitoring of train-of-four should be used to guide relaxant administration. It is also mandatory to carefully monitor end-tidal carbon dioxide to prevent hyperventilation and to maintain preoperative arterial partial pressure of carbon dioxide (Pco2) levels, which may be elevated. Sudden hypocapnia in a chronically hypercapnic patient can disrupt the patient’s ventilatory control mechanisms, increasing the chance that the patient might require postoperative ventilation. After intubation, tracheobronchial suction should be performed and repeated at intervals throughout surgery and before extubation to improve pulmonary gas exchange. Although the use of an LMA might be an option for short cases, disadvantages include the inability to suction secretions, obstruction of the LMA by thick secretions, risk of laryngospasm, and risk of aspiration in patients with gastroesophageal reflux disease. For bronchoscopy for the purpose of removal of secretions for culture and bronchial lavage, the LMA has been safely used in patients with CF (Nussbaum and Zagnoev, 2001). Intraoperatively, glucose should be monitored in patients with glucose intolerance. Care should be taken to conserve heat in these patients with reduced body fat. Regional anesthesia should be considered whenever applicable. Although regional anesthetic techniques without general anesthesia might be useful in some situations, these techniques

C h a p t e r 36    Systemic Disorders   1125

should be carefully considered in children with severe pulmonary disease. Depression of abdominal and intercostal muscle function by thoracic levels of spinal or epidural anesthesia may not be tolerated. Pediatricians and pulmonologists often request regional anesthesia instead of general anesthesia because of fear that severely affected patients with CF will not tolerate general anesthesia or may become dependent on a ventilator after endotracheal intubation. However, most of these sick patients with CF, dyspneic or orthopneic with hypercapnia and oxygen dependence, may not tolerate even a short surgical intervention, such as insertion of a central venous catheter or MediPort, with local anesthesia and sedation. Instead, general endotracheal anesthesia with an inhaled agent, supplemented by caudal, lumbar, or thoracic epidural anesthesia for abdominal or thoracic procedures, is much better tolerated, safer, and provides good operative conditions, rapid emergence, and a pain-free postoperative state (Dalens et al., 1986). If epidural anesthesia is to be used, coagulopathy should be ruled out, and the appropriate concentration of local anesthetic drugs should be chosen to minimize motor block. Continuous caudal or epidural anesthesia with local anesthetic with or without carefully chosen doses of an opioid provides prolonged postoperative pain relief and facilitates coughing and deep breathing after upper abdominal or thoracic procedures. If regional anesthesia is not appropriate, judicious use of inhalation agents and short-acting opioids and wound infiltration with local anesthetic by the surgeon should facilitate early extubation, which is desirable in most cases. After surgeries without a high risk of postoperative bleeding, the use of NSAIDs may be effective in reducing the amount of opioid needed for analgesia.

Cystic Fibrosis and Hemoptysis Hemoptysis may occur in older children with more severe lung disease. These children may come to the hospital for angiography and bronchial artery embolization. Preanesthetic considerations as described above should be evaluated, and the patient’s pulmonary status should be optimized to the greatest extent. Pulmonary hemorrhage has been reported after induction of general anesthesia, intubation, and institution of ­positive-pressure ventilation (McDougall and Sherrington, 1999). It is possible that airway and parenchymal distention associated with positive pressure ventilation results in the breach of relatively thinwalled vessels. These authors have recommended that sedation and local anesthesia may be safer in this situation (McDougall and Sherrington, 1999; Barben et al., 2002)

Cystic Fibrosis and Lung Transplantation For patients with end-stage pulmonary disease, lung transplantation may be the final surgical option. Fifty percent of patients with CF and FEV1 of less than 30%, partial pressure of oxygen (Pao2) less than 55 mm Hg, or partial pressure of carbon dioxide (Paco2) greater than 50 mm Hg will survive for 2 years and may be candidates for lung transplantation (i.e., double-lung or heart-lung procedure) (Belkin et al., 2006). Approximately 175 patients with CF per year undergo bilateral lung transplantation in the United States, of whom the minority are children (Cystic Fibrosis Foundation, 2007). The 3-year survival rate is 60%, which is similar to that seen in patients who do not have CF (Sweet et al., 1997). Some (27%) lungtransplant patients develop bronchiolitis obliterans, which is

responsible for 40% of deaths that occur more than 1 year after transplantation (Boucek et al., 2004). Among the survivors of lung transplantation, no recurrence of CF in the transplanted lungs has been reported, as measured by the transepithelial potential differences (Alton et al., 1991). The management of end-stage CF patients for lung transplantation is described in Chapter 3, Respiratory Physiology, and Chapter 28, Anesthesia for Organ Transplantation.

CARDIOVASCULAR DISORDERS Cardiovascular disorders are commonly encountered in the pediatric population. The baseline incidence of congenital heart disease in the population is approximately 0.8 in 100 births, on which is superimposed an incidence of acquired heart disease. However, more and more congenital cardiac defects have been appreciated to have either a genetic or an environmental basis, either alone or in combination, such that incidence and risks in offspring may be higher in selected populations (Jenkins et al., 2007). Both congenital and acquired diseases have the ability to affect myocardial function, valve function, and conduction tissue, all of which can also be affected by anesthetics. In addition, anesthetic effects on vascular tone can have a positive or negative impact on myocardial function and shunting of blood through intracardiac defects. Patients with cardiac disease should be identified preoperatively. Although children with congenital heart disease who undergo noncardiac surgery should generally do well with appropriate anesthetic and perioperative care, preliminary information suggests that in the aggregate, congenital cardiac disease of even a moderate degree can increase mortality and adverse events during and after noncardiac surgery (Baum et al., 2000). Even hemodynamically insignificant lesions may necessitate perioperative endocarditis prophylaxis (Table 36-15). However, not all surgical procedures or all children with cardiac disease require endocarditis prophylaxis. Recommendations for endocarditis prophylaxis were significantly changed in 2007, with a general decrement in the use of prophylaxis, because previous recommendations were recognized to be lacking in validity or efficacy or were overly complicated (Wilson et al., 2007). Current recommendations are outlined in Boxes 36-8 and 36-9.

Anesthetic Management Although the specifics of the anesthetic management of individual cardiac problems are discussed in Chapters 20, Anesthesia for Congenital Heart Surgery, and 21, Anesthesia for Children with Congenital Heart Disease Undergoing Non-Cardiac Surgery, the following comments are generally applicable.

Preoperative Period Prolonged preoperative fasting should be avoided in children who have cyanotic heart disease with significant erythrocytosis to avoid dehydration and further increase of elevated hematocrit and blood viscosity. Small infants with significant heart failure and failure to thrive can have inadequate glycogen reserves and are at risk for hypoglycemia if they fast for many hours. Otherwise conventional age-appropriate preoperative sedation is

1126   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-15. Endocarditis Prophylaxis Regimens for Dental Procedures Reason

Agent

Regimen

Standard general prophylaxis

Amoxicillin

50 mg/kg orally (adults: 2 g)

Unable to take orally

Ampicillin

50 mg/kg IM or IV* (adults: 2 g)

OR

Allergic to penicillin

Cefazolin or ceftriaxone

50 mg/kg IM or IV* (adults: 1 g)

Clindamycin

20 mg/kg orally (adults: 600 mg)

OR Cephalexin

50 mg/kg orally (adults: 2 g)

OR

Allergic to penicillin and unable to take orally

Azithromycin or clarithromycin

15 mg/kg orally (adults: 500 mg)

Clindamycin

20 mg/kg IV* (adults: 600 mg)

OR Cefazolin or ceftriaxone

50 mg/kg IM or IV* (adults: 1 g)

Modified from Wilson W et al: Prevention of infective endocarditis: guidelines from the American Heart Association, Circulation 116:1736, 2007. *It is appreciated that many children do not have IV access prior to surgery. IV antibiotics should be given as soon as possible and before surgical incision. Children’s dose should not exceed adult dose. IM, Intramuscularly; IV, intravenously.

in no way contraindicated in children with cyanotic or acyanotic heart disease unless the child is in profound heart failure.

Intraoperative Period Although much discussion is appropriately devoted to the specifics of cardiac pathophysiology, most children with congenital heart disease who develop problems during anesthesia do so for primarily noncardiac reasons, particularly airway compromise (Strafford and Henderson, 1991). Infants who are cyanotic, in particular, begin with decreased oxygen saturation and can rapidly desaturate with transient interruption in ventilation, whether as a result of apnea or airway obstruction. Children with severe congestive failure or cyanosis have a decreased margin of safety and tolerate failures of respiratory or hemodynamic management poorly. Much time is spent discussing the effects of left-to-right and right-to-left shunts on the onset time of IV and volatile anesthetics. Although differences exist, with currently available anesthetic gases of relatively low solubility, these differences are almost always so small as to be clinically irrelevant. In the absence of a complication such as loss of the airway or the development of a hypercyanotic “tet” spell in children with tetralogy of Fallot or variants, oxygen saturation in children with cyanotic disease almost invariably increases with induction of anesthesia (Laishley et al., 1986; Greeley et al., 1986). Several reasons are possible, one of the most likely being a decrease in oxygen

Box 36-8 Changes in Most Recent Endocarditis Prophylaxis Recommendations The number of cardiac conditions that require prophylaxis has been reduced. Moderate- and high-risk groups are not differentiated. l Dental prophylaxis is reserved solely for conditions in Box 36-9 and solely for manipulation of gingival tissues or the periapical region of teeth, or for perforation of oral mucosa. l Prophylaxis is reasonable for patients with conditions in Box 36-9 for invasive procedures of the respiratory tract that involve incision or biopsy, such as tonsillectomy and adenoidectomy. For patients who have an established infection of the respiratory tract, treatment with an agent active against Viridens streptococci is recommended, unless the infection is known or suspected to be caused by staphylococcus, in which case the regimen should include an appropriate agent such as an antistaphylococcal penicillin, cephalosporin, or vancomycin. l Prophylaxis for procedures on infected skin or musculo­ skeletal tissue is indicated only for patients with conditions in Box 36-9. As tissue is already infected, it is reasonable to use an antibiotic such as an antistaphylococcal penicillin or a cephalosporin that is active against those organisms (staphylococcus and β-hemolytic streptococci) that cause these infections and may lead to endocarditis. Vancomycin or clindamycin can be substituted in case of allergy or antibiotic resistance. l Antibiotic prophylaxis is not recommended for GI or genitourinary procedures, including vaginal delivery and hysterectomy. For patients in a high-risk group with established infections of the GI or genitourinary tracts who require surgery or cystoscopy, an appropriate antibiotic effective against enterococci, such as penicillin, ampicillin, piperacillin, or vancomycin, can be used, without empirical evidence of clinical efficacy in preventing endocarditis (Box 36-9). l Prophylaxis is not required for endotracheal intubation. l Prophylaxis is no longer required for patients with mitral valve prolapse. l Timing is simplified to 30 to 60 minutes before incision for all regimens. l

Data from Wilson W, et al: Prevention of infective endocarditis: guidelines from the American Heart Association, Circulation 116:1736, 2007.

Box 36-9 Cardiac Conditions that Necessitate Antibiotic Endocarditis Prophylaxis Prosthetic valves (bioprosthetic and homograft) Previous bacterial endocarditis Complex cyanotic heart disease Systemic pulmonary shunts (e.g., Blalock-Taussig shunt) Unrepaired cyanotic heart disease Completely repaired congenital heart defects with prosthetic material or device, whether by surgery or by catheter intervention, during the first 6 months after procedure Cardiac transplant recipients who develop cardiac valvulopathy Data from Wilson W, et al: Prevention of infective endocarditis: guidelines from the American Heart Association, Circulation 116:1736, 2007.

C h a p t e r 36    Systemic Disorders   1127

consumption that causes an increase in mixed venous oxygen saturation and subsequently higher arterial oxygen saturation when some of this blood is shunted right to left. Minimization of right-to-left shunting at the atrial level is addressed primarily by increasing intravascular volume. Minimization of shunting at the ventricular and great vessel levels is primarily modulated by changes in pulmonary vascular resistance (PVR) and systemic vascular resistance. Increasing systemic resistance or decreasing PVR increases left-to-right shunting (or decreases right-to-left shunting) and vice versa. Although α-adrenergic receptors are found in the pulmonary circulation (stimulation of which can increase PVR), they are denser in the systemic circulation, and α-agonists increase systemic vascular resistance significantly more than they increase PVR, with a decrease in right-to-left shunting. Nitrous oxide is a mild myocardial depressant. In adult patients, it can increase PVR, particularly in patients in whom PVR is already elevated. However, in children, no significant increase in PVR has been observed with 50% nitrous oxide, regardless of the preexisting PVR (Hickey et al., 1986). Patients who are cyanotic, and in particular patients with elevated central venous pressure, are at risk for increased perioperative blood loss and require adequate IV access. Not only do all patients with cyanotic disease need IV catheters to be kept clear of air bubbles to avoid systemic air emboli, but small amounts of transient right-to-left shunting can occur during the cardiac cycle, even with lesions thought of as left-to-right shunting lesions. Therefore, IV catheters must be cleared of air for all patients with shunt lesions, regardless of predominant direction of shunt flow. Stopcocks are common sites where air can be introduced inadvertently. End-tidal Pco2 correlates with arterial Pco2 in acyanotic patients. However, in children and adults with cyanotic congenital heart disease, end-tidal Pco2 tends to underestimate arterial Pco2 (Burrows, 1989).

in response to supplemental oxygen. Similarly, oxygen saturation is not markedly decreased by removing supplemental oxygen (other causes for postoperative hypoxemia being absent). Knowledge of the patient’s normal preoperative range of oxygen saturation helps avoid unnecessary prolongation of the PACU stay for fear of removing supplemental oxygen. Hypovolemia from continued surgical blood or fluid loss postoperatively can worsen right-to-left shunting in patients with cyanosis and should be rapidly corrected. The onset of hypovolemia can be insidious if it is as a result of gradual oozing from surgical drains. Patients with cyanotic disease should have repeated measurement of Hb after surgery, especially after significant blood loss. They may require a higher than normal hematocrit level to ensure adequate oxygen delivery. In general, a level similar to the preoperative hematocrit should be maintained. Patients with labile pulmonary arterial hypertension particularly benefit from good postoperative analgesia, because pain increases pulmonary vascular tone. Even patients with cyanosis have a normal ventilatory response to hypercarbia and respond in a normal fashion to appropriate doses of parenteral, intrathecal, or epidural opiates; therefore, age- and weight-appropriate doses of analgesic drugs should be given. Patients who have had a Glenn or Fontan procedure are dependent on low PVR for maintenance of adequate pulmonary blood flow. If these patients should require postoperative ventilation, PVR should be minimized by limiting positive inspiratory pressure and using low levels of positive end­expiratory pressure to optimize functional residual capacity, which in turn, minimizes PVR.

Postoperative Period

Cardiac murmurs are exceedingly common in healthy children, with an overall lifetime incidence of at least 50%. Most of these are the somewhat inappropriately named functional murmurs (also called innocent). The incidence of functional murmurs is highest at 3 to 4 years of age. Functional murmurs represent the sound of blood flowing through a structurally normal heart (Fig. 36-7). No anesthetic concern is associated with these murmurs, and the family should be reassured. Several functional murmurs are commonly recognized, almost all of which are short and soft but become louder when the patient lies supine.

The Child with a Murmur and Possible Heart Disease

The specific length of observation in a postanesthesia care unit is dependent on the patient and the surgical procedure and cannot be generalized. Patients with good hemodynamic function may undergo relatively minor noncardiac surgery on an ambulatory basis and are not automatically excluded because of their cardiac disease. When not under anesthesia, patients with cyanotic disease have relatively little increase in systemic oxygen saturation

S1

S1

S1

Carotid bruit

Venous hum S1

A2

P2

S1 Pulmonary ejection murmur of children and newborns

Vibratory systolic murmur

n  FIGURE 36-7. Innocent murmurs. Not shown is the mammary souffle over the breast of a pregnant or lactating woman. The pulmonary ejection murmur is similar to the murmur of peripheral pulmonic stenosis in neonates. Vibratory systolic murmur is also called Still murmur. (From Park MK: Pediatric cardiology for practitioners, ed 5, Philadelphia, 2008, Mosby.)

1128   P a r t  IV    Associated Problems in Pediatric Anesthesia

Most functional murmurs become louder with increased ­cardiac output, as would occur with anemia, fever, exercise, or anxiety. The most common functional murmur is the Still murmur, which has a typical musical or vibratory quality and is a midsystolic murmur heard between the midleft sternal border and the apex. Soft pulmonary flow murmurs at the upper left sternal border are commonly heard in thin-chested older children and adolescents. The murmur is softer than true pulmonic stenosis and is unaccompanied by a systolic ejection click. Peripheral pulmonic stenosis generates an ejection murmur from the left upper sternal border to the axillae and back and is common in neonates. It is generated by turbulent flow when blood passes from the main to the branch pulmonary arteries. In the neonate, the branch pulmonary arteries, unaccustomed to accommodating large amounts of pulmonary blood flow in utero, form an acute angle with the main pulmonary artery. By approximately 6 months of age the vessels remodel and the murmur disappears. Less common innocent murmurs are the venous hum and the mammary souffle, both of which are continuous murmurs and are thus exceptions to the rule that diastolic murmurs are always pathologic. The venous hum represents blood draining down the jugular into the subclavian veins. It is heard over the left or right upper chest with the patient upright and disappears when the patient lies down, with gentle compression of the jugular vein, or with a Valsalva maneuver. The mammary souffle can be heard over the breasts of pregnant or lactating women. Unlike functional murmurs, pathologic murmurs are generated by a normal amount of blood across an abnormal valve or opening or by an abnormal amount of blood passing through normal valves. Occasionally, a murmur is appreciated for the first time when children arrive for a preanesthetic evaluation. The exact method of evaluation remains somewhat controversial (Yu et al., 2002). Isolated chest radiographs and electrocardiograms are generally

a poor investment (Yu et al., 2002). In addition, electrocardiograms that are interpreted by computer or an adult cardiologist may need to be reinterpreted using age-appropriate normal values. Location of the murmur can also aid in the diagnosis (see Chapter 9, Preoperative Preparation, Fig. 9-3). In general, children who are acyanotic and growing well, with a soft systolic murmur and good exercise tolerance, tolerate anesthesia well. Signs of heart disease in infants differ somewhat from those in adults and older children. Perioral cyanosis can be a normal finding in neonates, especially with crying, and must be differentiated from central cyanosis (confirmed by pulse oximetry). Heart failure is often manifested in young infants by tachypnea, diaphoresis with eating (in excess of the normal sweating of the head that many infants have), and hepatomegaly. Increased pulmonary blood flow can impinge upon small bronchioles, causing airway obstruction and expiratory wheezing (“cardiac asthma”). Peripheral edema as a result of congestive failure is distinctly uncommon in children. Blood pressure measurements in both arms and a leg help detect or exclude coarctation of the aorta. When caring for children with known heart disease or a history of cardiac surgery, the child’s pediatrician or cardiologist should be contacted, and a copy of the most recent evaluation should be obtained.

Noncardiac Manifestations of Congenital Heart Disease Longstanding cyanotic and acyanotic congenital disease can have effects on the function of a variety of other organ systems. Some of these effects may not become clinically apparent until years after surgical correction of the underlying cardiac defect (Table 36-16).

TABLE 36-16. Potential Noncardiac Manifestations of Congenital Heart Disease System

Causes/Implications for Anesthetic Management

Pulmonary/Thoracic Decreased lung compliance

Occurs in lesions with increased pulmonary blood flow (i.e., left-to-right shunting) or pulmonary venous obstruction; can require higher airway pressure for ventilation; can impinge on small airways, resulting in air trapping, wheezing

Scoliosis

More common with cyanotic lesions; can manifest in adolescence, years after corrective cardiac surgery

Hemoptysis

Can occur in end-stage Eisenmenger syndrome (pulmonary hypertension caused by prolonged excessive pulmonary blood flow)

Phrenic nerve injury

From prior surgery; more common after surgery at the apices of the thorax (e.g., patent ductus arteriosus ligation, coarctation, pulmonary artery banding, or Blalock-Taussig shunt)

Recurrent laryngeal nerve injury

From prior surgery or from an enlarged hypertensive pulmonary artery; see phrenic nerve injury (above)

Blunted ventilatory response to hypoxemia

In cyanotic patients; normalizes after surgical repair; normal ventilatory response to hypercarbia

Hematologic Symptomatic hyperviscosity

Due to chronic hypoxemia; occurs with hematocrit above approximately 65% (or lower if iron deficient); may cause neurologic symptoms

Bleeding diathesis

Abnormalities of a variety of factors have been described in patients who have cyanosis with no consistent pattern; elevated central venous pressure can cause increased operative bleeding, as can increased tissue vascularity with cyanotic disease (collateral blood vessel formation); increased risk of bleeding with prior thoracic surgery during subsequent thoracic procedures

Gallstones

Calcium bilirubinate stones from increased heme turnover in cyanotic disease; may not become symptomatic until years after corrective cardiac surgery

C h a p t e r 36    Systemic Disorders   1129

TABLE 36-16. Potential Noncardiac Manifestations of Congenital Heart Disease—cont’d System

Causes/Implications for Anesthetic Management

Neurologic Paradoxical emboli to central nervous system

Can manifest even with a predominantly left-to-right shunt lesion

Brain abscess in patients with right-to-left shunts

Can manifest with seizure focus years later

Cerebral thrombosis

In children with erythrocytosis but not in adults

Compression of nerve by vascular structure

Recurrent laryngeal nerve by enlarged, hypertensive pulmonary artery

Nerve injury during prior surgery

Recurrent laryngeal, phrenic, or sympathetic chain; see phrenic nerve injury (above) for high-risk surgeries

Vascular Femoral vein complications

Thrombosis or ligation from prior cardiac catheterization

Anatomic discontinuity of inferior vena cava and right atrium

Congenital, associated with polysplenia syndrome

Anatomic discontinuity of right internal jugular and innominate vein with right atrium

As a result of Glenn shunt

Reduced lower extremity blood pressure

Coarctation of the aorta; left arm involvement is variable

Discontinuity of subclavian artery

With classic Blalock-Taussig anastomosis; stenosis of the subclavian artery after the modified Blalock-Taussig anastomosis; currently, classic Blalock-Taussig shunt is rarely performed

Artifactually elevated right arm blood pressure

Supravalvar aortic stenosis (the Coanda effect)

Kawasaki Disease Originally named mucocutaneous lymph node syndrome after its major manifestations, Kawasaki disease is the most common cause of acquired heart disease in children in the United States. The etiology for this disease, with characteristics of an infectious disease without an identifiable agent and a vasculitis that is not easily treated with steroids, has yet to be determined. In the United States, the peak incidence occurs between 13 and 24 months of age. Diagnosis and therapy have been reviewed in detail, and a general overview has been presented by Burns (2007) (Newburger et al., 2004). No diagnostic test has been developed. The acute illness is associated with fever; intense conjunctival injection; red, cracked lips and oral mucosa; erythema of the palms and soles; and less commonly, cervical lymphadenitis of the neck followed weeks later by desquamation of the skin of the fingers and toes. Arthritis and arthralgia can also develop. Other uncommon systemic manifestations include transient high-­frequency hearing loss, gastrointestinal complaints, hepatomegaly, and hydrops of the gallbladder. Erythema and induration at the site of a previous bacillus Calmette-Guérin (BCG) vaccination can also occur. The most concerning feature of the disease is that it causes an infantile periarteritis nodosa-like vasculitis of medium and large arteries in 10% to 15% of children. Of particular concern is involvement of coronary arteries, with the risk of subsequent thrombosis or, less commonly, rupture (Fig. 36-8). The risk of coronary artery aneurysms is higher in infants. In addition, the acute phase of the illness can be associated with ­myocarditis—usually mild but sometimes associated with heart failure. Myocarditis is usually transient, lasting several weeks. Laboratory findings during the acute phase reveal an intense acute-phase response that includes neutrophilia, elevated sedimentation rate and C-reactive protein, and thrombocytosis to over 800,000/mcL3.

n  FIGURE 36-8. Angiogram of the left main coronary artery in a young child with Kawasaki disease showing multiple fusiform aneurysms of both the right and left coronary arteries. (Courtesy Jonathan Rome, MD).

Coronary artery aneurysms become apparent within the first 2 weeks of disease in 3% to 5% of children who have been treated with IV gamma globulin (IVIG) and in 20% to 25% of children who have not. Early aneurysms can resolve spontaneously or progress.

1130   P a r t  IV    Associated Problems in Pediatric Anesthesia

Treatment in the acute phase includes IVIG and aspirin. Aspirin is begun at a high dosage (80 to 100 mg/kg per day) and continued either until the child is afebrile or until day 14 of illness, after which the high antiinflammatory dose is decreased to antiplatelet levels (3 to 5 mg/kg per day). The lower dose is continued until the child shows no evidence of aneurysm formation (6 to 8 weeks after onset) or is continued indefinitely if coronary aneurysms develop. Although the mechanism is undetermined, IVIG (2 g/kg) is an effective acute therapy that is successful in approximately 80% of cases. If IVIG therapy fails, children are at high risk for the development of coronary artery complications, which may be treatable with steroids, although a recent study suggests this may not be effective (Newburger et al., 2007). If coronary artery aneurysms do develop, approximately one half to two thirds regress within 2 years, and approximately one fifth develop coronary stenoses. Smaller aneurysms (less than 8 mm in diameter) and fusiform aneurysms are more likely to regress than larger or saccular aneurysms. Larger aneurysms can develop thromboses or stenoses with subsequent ischemia, or they can rupture. Rupture is rare and usually occurs within the first month or two of disease. Ischemia can develop years after the acute illness. Even if aneurysms regress, intimal proliferation can result in endothelial dysfunction (Furuyama et al., 2003). Warfarin has been used at some centers to treat children with giant aneurysms. Angioplasty has been attempted at several centers with mixed results, and surgical bypass grafting has been performed on occasion for high-grade obstruction of the left main coronary artery or at least two of the major coronary arteries. Because of the young age of the patients, grafting is completed with arterial rather than venous grafts (Kitamura et al., 1994).

Takayasu Arteritis Takayasu arteritis, a vasculitis of the aorta and its major branches that is sometimes known by its catchy synonym “pulseless disease,” is an uncommon disease in children. However, 75% of patients will have begun to develop symptoms during adolescence. An important cause of hypertension in teenagers in Asia, where it is more common, it occurs 8 times more often in females. Narrowing of major arteries results in limb claudication or end-organ disease. Blood pressure in limbs can be artifactually low or unobtainable. Early vessel inflammation is followed by fibrosis. Headaches are a common symptom. The subclavian artery is involved in 90% of cases, and two thirds of all cases involve the aorta, both supradiaphragmatic and infradiaphragmatic. The carotid artery, almost always the left, is involved in half of cases. Stenoses are more common than occlusion, and occlusion is more common than aneurysm formation. Mural inflammation and thickening occur, and coronary and pulmonary arteries are uncommonly affected. Aortic root dilation can result in aortic valve insufficiency. Initial treatment is with corticosteroids. Long-term therapy is often required, and cytotoxic drugs are sometimes added. Once fibrosis has occurred, treatment is by stenting or surgery (Rigby et al., 2002; Kalangos et al., 2006). Unfortunately, there is often recurrence with tapering of steroid therapy or with time after surgical treatment (Maksimowicz-McKinnon et al., 2007).

HEMATOLOGY AND ONCOLOGY ISSUES: HEREDITARY AND CONGENITAL DISORDERS

Hemoglobinopathies Hemoglobin Structure, Development, and Function A Hb molecule is composed of two pairs of subunits comprised of protoheme and globin. The globin imparts the spatial structure responsible for characteristics of Hb, including oxygen affinity. Globin chains differ in the number and sequence of amino acids and are designated by α, β, γ, δ, ε, ζ, and θ. Human adult red blood cells (RBCs) contain three types of Hb: 95% HbA (α2, ß2), 2% to 3% HbA2 (α2, δ2), and less than 2% HbF (α2, γ2). The physical properties and spatial relationships of the four chains determine oxygen affinity and Hb solubility. At birth, erythrocytes contain 70% to 90% HbF, and HbF predominates until 2 to 4 months of age in normal patients. β-chain production begins and γ-chain production decreases before birth, resulting in a normal adult Hb profile by the age of 4 months. Therefore, disorders of β-chain production do not manifest clinically, nor can they be reliably detected by Hb precipitation tests, before 4 months of age. However, neonatal Hb electrophoresis screening is included as part of testing for inborn errors of metabolism and detects most newborn patients with sickle cell disease (SCD) in the nursery. Persistence of HbF may occur naturally; it also results from hydroxyurea administration. An elevated HbF concentration is protective against complications of SCD. Hemoglobinopathies result from either production of abnormal Hb or decreased production of a chain. The most relevant example of the former mechanism is amino acid substitution. Decreased chain production results in thalassemia. Hb profiles for clinically relevant hemoglobinopathies are detailed in Table 36-17 (Lane, 1996).

Genetics and Pathophysiology of Sickle Cell Disease SCD refers to a clinical phenotype marked by erythrocyte deformation, hemolysis, anemia, microvascular occlusion, and recurrent ischemic injury in all organ systems. In the United States, 8% of the African American population carries one of the recessive genes that result in the sickle cell trait (HbAS, SCT), and one in 625 African Americans is homozygous for mutant alleles that result in sickle cell anemia (HbSS, SCA). SCD is a leading cause of morbidity and mortality among African Americans and is associated with important considerations for perioperative management. The expression of the SCD phenotype is not limited to patients with SCA. Its expression is caused by the inclusion of a mutant β-globin chain in the Hb tetramer, resulting in clinical courses of variable severity. SCD phenotypes of clinical relevance include the following genotypes: sickle cell anemia (HbSS, SCA), HbSCD (HbSC), sickle-β0 thalas­semia (S-β0), sickle-β+ thalassemia (S-β+), and HbS-OArab (HbS-OArab). By definition, patients with the sickle cell trait (HbAS) have at least 50% HbA. Under physiologic conditions, clinical problems are rare. Hb polymerization begins below oxygen saturations of 85% in patients with HbSS, but this does not occur in patients with HbAS until saturations are less than 40%. Thus, it is accepted that patients with HbAS do not require preoperative

C h a p t e r 36    Systemic Disorders   1131

TABLE 36-17. Clinical Severity and Diagnostic Testing for the Common Sickle Cell Syndromes Relative Clinical Severity Syndrome

Genotype

Hemolysis

Vasoocclusion

Hemoglobin Electrophoresis in Older Children Neonatal Screening*

HbA (%)

HbS (%)

HbF (%)

HbA2 (%)

HbC (%)

Solubility Test†

Sickle cell anemia (HbSS)

S-S

++++

++++

FS

0

80-95

2-20

<3.5

0

Pos

Sickle β0 thalassemia‡

S-β0

+++

+++

FS

0

80-92

2-15

3.5-7

0

Pos

Sickle HbC disease (HbSC)

S-C

+

++

FSC

0

45-50

1-5

NA§

45-50

Pos

Sickle β thalassemia

S-beta

+

+

FSA or FS

5-30

65-90

2-10

35-6

0

Pos

+

‡

+

¶

Sickle cell trait

A-S

0

0

FAS

50-60

35-45

<2

<3.5

0

Pos

Normal

A-A

0

0

FA

95-98

0

<2

<3.5

0

Neg

From Lane PA: Sickle cell disease, Pediatr Clin North Am 43:639, 1996. *Hemoglobin reported in order of quantity (e.g., FSA = F > S > A). † False-negative results occur during infancy in all sickle syndromes. ‡ 0 β indicates thalassemia mutation with absent production of β-globin; β+ indicates thalassemia mutation with reduced (but not absent) production of β-globin. § Quantity of HbA2 cannot be measured in presence of HbC. ¶ Quantity of HbA at birth sometimes insufficient for detection. F, Fetal Hb; S, sickle Hb; C, HbC: A, HbA.

100

Oxygen saturation %

transfusion and are managed with standard attention to hydration, oxygenation, and temperature control. The fundamental defect in SCD genotypes is amino acid substitution of valine for glutamine at position 6 of the β chain. The resulting HbS is prone to polymerization under conditions of oxygen desaturation caused by formation of bonds between the β-6 valine and the β chains of adjacent tetramers, resulting in the formation of erythrocyte-deforming Hb polymers, erythrocyte cell membrane dysfunction, intracellular oxidant injury, cellular dehydration, and irreversible erythrocyte membrane deformity (Eaton and Hofrichter, 1987; Hebbel et al., 1988; Repka and Hebbel, 1991; Aslan et al., 2000). Deformed erythrocytes obstruct the microvasculature, resulting in tissue ischemia and organ injury. Polymerization is not initially irreversible, however, and prompt return of damaged erythrocytes to the oxygenated state can result in restoration of Hb and cell membrane properties. Survival of sickle cells in vivo is 5 to 15 days, compared with 120 days for RBCs that contain HbA. The oxygen-dissociation curve in SCA is shifted to the right (i.e., the Hb molecules’ affinity for oxygen is less), and theoretically the cells are predisposed to sickling. The cause of this rightward shift is not known, but it is probably related to increased 2,3-diphosphoglycerate levels and to increased mean corpuscular Hb concentration. Of interest is the heterogeneous nature of the P50 (Pao2 at 50% Hb saturation) values among the erythrocyte population of individuals with sickle cell disease (Fig. 36-9) (Seakins et al., 1973). Although it is apparent that erythrocyte and Hb factors are critical in the pathogenesis of the complications of SCD, recent investigations have identified other important contributors that include the coagulation cascade, platelets, leukocytes, endothelial cells, systemic inflammation, oxidant-mediated injury, and abnormalities of NO metabolism (Hammerman et  al., 1999; Belcher et al., 2000; Morris et al., 2000a; Aslan et al., 2001; Pawloski et al., 2001; Cosby et al., 2003; Assis et al., 2005; Morris et al., 2005). The endothelium interacts with sickle erythrocytes, platelets, leukocytes, and the coagulation cascade through the expression of vascular cell adhesion molecule-1 (VCAM-1) and elaboration of potent vasoconstrictor compounds responsible for ­exacerbation

75

50 AA, normal SS, top SS, bottom

25

0 0

25

50

75

100

125

PO2 mm Hg

n  FIGURE 36-9. Oxygen-Hb dissociation curves from top and bottom layers of HbS RBCs. Notice the heterogeneous nature of the P50 values. (From Seakins M, et al: Erythrocyte HbS concentration: an important factor in the low oxygen affinity of blood in sickle cell anemia, J Clin Invest 52:422, 1973.)

of tissue ischemia and pulmonary artery hypertension (PAH) (Mehta and Mehta, 1980; Brittain et al., 1993; Kurantsin-Mills et al., 1994; Peters et al., 1994; Gee and Platt, 1995; Hagger et al., 1995; Hammerman et al., 1997, 1999; Hebbel, 1997; GraidoGonzalez et al., 1998; Liesner et al., 1998; Belcher et al., 2000; Assis et al., 2005). Systemic inflammation and oxidant-mediated injury also appear to propagate endothelial injury and tissue is­chemia (Hebbel et al., 1982; Schacter et al., 1988; Dias-Da-Motta et al., 1996; Xia et al., 1996; Solovey et al., 1997; Demiryurek et al., 1998; Belcher et al., 2000; Aslan et al., 2001; Assis et al., 2005). SCD is a systemic vasculopathy adversely affecting every organ system in the body.

1132   P a r t  IV    Associated Problems in Pediatric Anesthesia

Abnormalities in the arginine-NO pathway are well described in SCD. Arginine, the precursor for the production of NO, is deficient in SCD and is depleted further during acute chest syndrome (ACS) and vasoocclusive crisis (VOC) (Enwonwu et al., 1990; McDonald et al., 1997; Morris et al., 2000a, 2000b, 2005). The pathophysiology of vasoocclusion and its biochemical contributors are detailed in Figure 36-10 (Gladwin and Vichinsky, 2008). NO is bound by Hb and is exported to the peripheral vasculature in the forms of nitrite and S-nitrosohemoglobin (Pawloski et al., 2001; Cosby et al., 2003). Deficient NO bioactivity occurs secondary to excessive NO and arginine consumption, arginase-mediated destruction of arginine, and abnormal conversion of arginine and NO to injurious oxidative metabolites (Boucher et al., 1999; Hammerman et al., 1999; Mori and Gotoh, 2000; Morris et al., 2000a). Deficient NO bioactivity contributes to the pathogenesis of SCD through diminished inhibitory effects of NO on VCAM-1 expression, leukocyte adhesion, coagulation cascade, and platelet activation, and through diminished NO-mediated vasodilation, which promotes vasoconstriction, ischemia, and resultant erythrocyte sickling. The interactions of deoxygenated sickle Hb, deformed erythrocyte cell membranes, and activated endothelial cells can be summarized by the “four S’s”; unstable Hb that precipitates and becomes insoluble, leading to sickling deformation of erythrocyte membranes, which stick to activated endothelial cells.

Acute Complications of Sickle Cell Disease Relevant to Pediatric Anesthesia Pediatric anesthesiologists encounter patients with SCD when the patients are experiencing acute complications of their disease. Acute splenic sequestration usually occurs in children between the ages of 5 months and 2 years and may appear as late as teenage years in patients with sickle-thalassemia (S-β+). This condition results from the pooling of large quantities of blood in the spleen and leads to shock with profound anemia. Aplastic crisis results when the normal brisk reticulocytosis associated with SCD is suppressed, usually as a result of an intercurrent viral infection with parvovirus B19. Hemolytic crisis occurs when a patient with SCD is exposed to a precipitant that causes an abrupt increase in hemolytic stress (infection or medication). Many patients with hemolytic crisis are also deficient in the enzyme glucose-6-phosphate dehydrogenase (G6PD). These disorders are treated with intravascular volume expansion and transfusion of packed RBCs. Treatment of infection and discontinuation of offending medications are also required for hemolytic crisis. Sepsis and septic shock are serious complications that occur in patients with SCD, who generally experience autoinfarction of the spleen in early childhood, rendering them susceptible to infection with encapsulated organisms. The importance of aseptic techniques and wound infection prophylaxis cannot be overstated.

Hemolysis, endothelial dysfunction

Viscosity, vaso-occlusion

Precapillary arteriole Smooth-muscle cells ET-1

Capillary

Hb

α4β1

– Arg NO O2

NOS

Decreased NO bioactivity

Monocyte

Erythrocyte

X NO X

Endothelial cells

Postcapillary venule

VCAM-1

Platelets

XO

Pulmonary hypertension Leg ulceration Priapism Stroke

Pain crisis Acute chest syndrome Osteonecrosis

Increased vaso-occlusion

n  FIGURE 36-10. Hypothetical mechanisms of clinical subphenotypes of SCD. It is hypothesized that many of the complications of SCD can be divided into two overlapping subtypes, each driven by distinct mechanisms. Cutaneous leg ulceration, priapism, pulmonary hypertension, sudden death, and stroke are associated with low steady-state Hb levels and an increased rate of intravascular hemolysis, shown on the left side of the figure. These vasculopathic complications probably result from endothelial dysfunction, mediated by inactivation of NO by plasmafree Hb and vascular reactive oxygen species, and by arginine catabolism by plasma arginase. The process of hemolysis-associated endothelial dysfunction may also cause hemostatic activation and intimal and smooth muscle proliferation. Such clinical complications (e.g., VOC, ACS, avascular necrosis of bones, and retinal vasculopathy) are associated with high steady-state leukocyte counts and high Hb levels and are likely to result from obstruction of capillaries and postcapillary venules by erythrocytes that contain polymerized HbS and by leukocytes (a monocyte is shown), as shown on the right side of the figure. ET-1, Endothelin 1; NOS, nitric oxide synthase; O2-, superoxide, VCAM-1, vascular cell adhesion molecule 1; XO, xanthine oxidase. (From Gladwin MT, Vichinsky E: Pulmonary complications of sickle cell disease, N Engl J Med 359:2254, 2008.)

C h a p t e r 36    Systemic Disorders   1133

VOC manifests as episodes of painful ischemia and tissue infarction that result from small-vessel occlusion by sickle cells. The most common types of VOC include dactylitis (handfoot crisis) in infancy, painful crisis in children and adolescents, and priapism in male patients. Stroke and ACS are the most serious forms of VOC that warrant further discussion. ACS is a common and potentially lethal complication of SCD with complex pathophysiology that can occur in the perioperative period (Fig. 36-11). ACS is the second most common cause of hospital admission and is a leading cause of premature death (mortality rates of 2% to 12%) (Castro et al., 1994; Vichinsky et al., 1997). The etiology of ACS is multifactorial and includes regional hypoxemia secondary to atelectasis or pneumonia, pulmonary vasoocclusion, abnormal NO metabolism, fat emboli, and systemic liberation of inflammatory mediators (Garza, 1990; Vichinsky et al., 1994; Castro, 1996; Styles et al., 1996). ACS occurs in the setting of pulmonary infection or painful crisis, and the symptoms of pneumonia (i.e., fever, pleuritic chest pain, cough, dyspnea, hypoxemia, and progressive infiltrates on chest radiograph) may be difficult to distinguish

from ACS. Respiratory failure may develop quickly, and patients with ACS are treated with hydration, oxygen, antibiotics, simple transfusion, and in severe cases, exchange transfusion. Case reports have described treatment with inhaled NO and extracorporeal membrane oxygenation (Atz and Wessel, 1997; Pelidis et al., 1997; Sullivan et al., 1999). The best treatment for ACS is prevention. Stroke is the other serious complication of SCD that may occur in the perioperative period. Stroke in children with SCD is most commonly ischemic, but one third of strokes in adult patients with SCD are hemorrhagic (Pavlakis et al., 1989). Many more patients with SCD suffer silent ischemic infarcts that are evident only on neuroimaging, and their presence predicts future is­chemic neurologic injury. Indeed, 20% of asymptomatic adolescents have evidence of silent cerebral infarction on neuroimaging studies. Because of the prevalence of silent infarcts in the SCD population, hematologists monitor middle cerebral artery flow velocity with transcranial Doppler (TCD) ultrasonography. Patients who suffer stroke, as well as those with abnormally elevated TCD flow rates, are treated with chronic HbS erythrocyte

Vaso-occlusive crisis

Erythrocyte

Increased polymerization and erythrocyte rigidity

Regional hypoxia

α4β1 VCAM-1 Decreased oxygen delivery Desaturated hemoglobin

NO

NO

Increased endothelial VCAM-1 expression and adhesion

Increased erythrocyte adhesion in lung causing pulmonary infarction

Fat

Microvascular occlusion and bone marrow infarction

Shunt

Secretory phospholipase

Hypoventilation and atelectasis resulting from rib and vertebral infarction

Pulmonary infection Acute chest syndrome

n  FIGURE 36-11. The vicious cycle of the ACS. ACS is a lung-injury syndrome initiated by three major triggers, all related to vasoocclusion by sickle cells: infection, embolization of bone marrow fat, and intravascular sequestration of red cells. All of these cause lung injury and infarction. Lung injury results in ventilation-perfusion mismatch and hypoxemia, which leads to increased deoxygenation of HbS, followed by Hb polymerization and erythrocyte vasoocclusion, which in turn promotes bone marrow infarction and pulmonary vasoocclusion. NO, Nitric oxide; VCAM-1, vascular cell adhesion molecule 1. (From Gladwin MT, Vichinsky E: Pulmonary complications of sickle cell disease, N Engl J Med 359:2254, 2008.)

1134   P a r t  IV    Associated Problems in Pediatric Anesthesia

transfusion to keep HbA at less than 30%. Patients are treated with chronic transfusion for 10 years or longer, because cessation of therapy is associated with increased risk of stroke and recurrence of abnormal TCD flow rates (Adams and Brambilla, 2005). Alloimmunization to erythrocyte antigens is associated with chronic transfusion. Iron overload may also result and may require parenteral deferoxamine or oral deferasirox chelation therapy. Extended phenotypical matching of patient and donor red-cell antigens may result in decreased alloimmunization. Angiographic and pathologic studies of patients with SCD who have experienced stroke have demonstrated proximal intracranial arterial stenosis of the internal carotid artery, which can be associated with segmental thickening of the arterial wall, and intimal hyperplasia, which tends to occur at sites of arterial bifurcation. A subset of patients with SCD may also demonstrate predominantly small-vessel disease of the CNS vasculature. The potential for presence of flow-limiting lesions in the CNS should be considered when modulating mean arterial blood pressure in the anesthetic management of patients with SCD.

Chronic Complications of Sickle Cell Disease Relevant to Pediatric Anesthesia Chronic complications of SCD reflect the accumulation of a lifetime of ischemic insults and subsequent decrements in function in all organ systems. Complications include decreased growth and maturation, increased nutritional requirements, retinop­athy, stroke, cognitive dysfunction, cardiac dysfunction, elevated PVR, chronic lung injury, diminished renal tubular function, icterus, bone and joint destruction, leg ulceration, and splenic infarction with consequent susceptibility to infectious risk. Of particular importance to the perioperative physician is an appreciation for chronic changes that occur with time in the cardiovascular, respiratory, and renal systems. The effects of SCD on the cardiovascular system may not be clinically evident to the anesthesiologist. Patients with chronic anemia maintain systemic oxygen delivery through increases in stroke volume and, to a lesser degree, heart rate (in older children and adolescents). Stroke volume increases secondary to increased end-diastolic volume with minimal change in ejection fraction. This condition manifests clinically as cardiomegaly with ventricular dilation. More ominous than the common changes evident in the left heart is the development of PAH and right ventricular dysfunction. Children with SCD may demonstrate elevated pulmonary vascular resistance during adolescence and early adulthood, and the development of PAH is a robust predictor of premature death (Castro et al., 2003). Most concerning is that in several series of echocardiographic evaluations of patients with SCD, the incidence of PAH was found to be between 20% and 32%, and many of the patients were asymptomatic (Sutton et al., 1994; Gladwin et al., 2004). The mechanisms for the development of PAH in patients who have SCD have been postulated to include diffuse arteriolar thrombosis, scavenging of NO by free Hb in the circulation, arginase-mediated destruction of arginine (limiting NO bioavailability), systemic oxidant-mediated vascular injury, and increased expression of cellular adhesion molecules. SCD patients with or without acute lung injury may have significant PAH and right ventricular dysfunction. Older patients with SCD may also have significant chronic deterioration of pulmonary function. Sickle cell chronic lung

disease is marked by recurrent episodes of chest pain associated with progressive pulmonary fibrosis and dyspnea (Powars et al., 1988). The etiology of this disease is not known but is associated with chronic hemolysis and increased plasma arginase activity. Anesthesiologists should be aware that restrictive lung disease, obstructive lung disease, airway smooth muscle hyperreactivity, and significant derangements of ventilationperfusion matching have all been described in SCD. Repeated ischemic insults to the kidneys may result in significant deterioration in glomerular and tubular function in patients who have SCD. Although this condition is rare in children, older patients may suffer impairment severe enough to require dialysis or transplantation. Patients with SCD demonstrate isosthenuria—an inability to concentrate urine—making the patient prone to dehydration. The potential for presence of renal dysfunction should be considered in the management of perioperative fluids, acid-base balance, electrolyte levels, and medications.

Anesthetic Management of Patients with Sickle Cell Disease Patients with SCD are surviving later into adulthood as a consequence of availability of more effective immunization, antimicrobial prophylaxis and treatment of sepsis, surveillance measures to detect PAH and impending stroke, and diseasealtering therapies such as hydroxyurea, chronic transfusion, and stem cell transplantation. Optimal perioperative management of this population requires an understanding of the pathophysiology of the serious complications of SCD and emphasizes basic strategies to minimize their occurrence. Perioperative care of these patients can be considered in terms of preoperative care (including the role for prophylactic preoperative RBC transfusion), intraoperative care, and postoperative care.

Preoperative Preparation of Patients with Sickle Cell Disease The goal of preoperative preparation of the patient with SCD is to optimize medical conditions that impact the patient’s surgery. Identification of patients with SCD is accomplished through newborn Hb electrophoresis screening. Most patients with a SCD phenotype will have been referred to a pediatric hematology practice for longitudinal follow-up care. Rarely, a child may elude detection, with potentially catastrophic results caused by lack of daily antibiotic prophylaxis. Patients older than 6 months can be screened for sickle hemoglobinopathies with “sickle prep” tests in which Hb polymerization is provoked in vitro. In the presence of significant HbF concentrations, these sickle prep examinations lack sensitivity, and Hb electrophoresis is required. Children younger than 6 months and older children with positive sickle prep results require Hb electrophoresis to delineate their hemoglobinopathy. By their tenth birthday, 90% of patients with SCD have clinical manifestations of their disease. Cardiovascular, respiratory, and renal systems, as well as the CNS should be screened for dysfunction that may impact perioperative management. Patients should be treated for infection when appropriate, because infection and systemic inflammation contribute to development of VOC and ACS. In that dehydration may predispose to VOC and ACS, patients with SCD should be well hydrated before surgery. Administration of IV hydration at 1.5 times the maintenance

C h a p t e r 36    Systemic Disorders   1135

rate overnight before surgery has been recommended in the past; however, this is not a practice supported by randomized trials. A recent small retrospective study demonstrated the safety and effectiveness of outpatient preoperative management (transfusion and liberal oral hydration) and goaldirected postoperative management, which obviated the need for preoperative admission and extended postoperative observation in this cohort of patients with SCD who were undergoing adenotonsillectomy (Duke et al., 2006). In short, patients with SCD should not be permitted to become dehydrated or volume overloaded during any phase of perioperative management.

Role of Routine Preoperative Blood Transfusion for Patients with Sickle Cell Disease The most debated topic relevant to the perioperative management of patients with SCD concerns the role of prophylactic preoperative blood transfusion. It has long been appreciated that surgical morbidity and mortality are increased in this population. Early reviews reported mortality rates as high as 10% and complication rates of 50% (Holzmann et al., 1969). Data reflective of more recent advances in anesthetic monitoring and pharmacology demonstrate a 1.1% perioperative mortality rate among adult and pediatric patients with SCD—a level still clearly greater than that of the general population (Koshy et al., 1995). No perioperative deaths were reported in children younger than 14 years of age, nor were perioperative deaths noted in 54 children who underwent 66 elective surgical procedures without preoperative transfusion (Griffin and Buchanan, 1993; Koshy et  al., 1995). Although perioperative mortality among patients with SCD is recognized to be more common in older children and adults, serious complications and mortality are well described in all age groups and warrant meticulous attention to perioperative management. Beginning in the late l980s, anesthesiologists, hematologists, and surgeons became interested in preoperative transfusion as a means of decreasing the perceived risk associated with surgery and anesthesia for patients with SCD. Simple transfusion strategies refer to the transfusion of small aliquots of blood with the goal of increasing the Hb level to 10 g/dL without regard to the final HbS percentage. Aggressive transfusion strategies refer to either repetitive, simple transfusions over time or acute exchange transfusion with the goals of Hb concentration of 10  g/dL and HbS less than 30%. The rationale for preoperative transfusion in both cases is to increase the ­oxygen-carrying capacity of the blood and in the case of aggressive transfusion to decrease the percentage of HbS in the circulation to prevent red cell sickling. Potential detrimental effects of preoperative prophylactic transfusion include increased risk for acquisition of blood-borne infection, hemolytic transfusion reactions, febrile reactions to leukocyte antigens, transfusion-associated acute lung injury, iron overload, and immunosuppression. Alloimmunization to foreign RBC antigens may develop, which results in difficulty in cross-matching RBCs for transfusion. The past two decades have witnessed the publication of numerous trials of variable quality and size that have attempted to define the utility of routine preoperative transfusion in the population with SCD. A succinct answer to whether preoperative transfusion is beneficial remains elusive because of confounding variables that include the following:

l Different surgical procedures appear to be associated with

different degrees of risk for morbidity and mortality;

l Perioperative morbidity and mortality risk from the same

surgical procedures may not be equivalent in infants, children, adolescents, and young adults; l Changes in surgical and anesthetic techniques, as well as advances in disease-altering medical therapies for SCD, may result in changes in perioperative morbidity and mortality; l Differing thresholds among studies for exclusion and inclusion of perioperative complications to be considered as clinically relevant endpoints to warrant transfusion in an effort to minimize its occurrence; and l Lack of randomized prospective data that compare outcomes in patients with SCD who are managed with simple transfusion or exchange transfusion with those who are managed without preoperative transfusion. Despite the difficulties discussed above, approximately 14 studies published during the past 20 years addressed, in some fashion, the utility of preoperative transfusion (Bischoff et al., 1988; Griffin and Buchanan, 1993; Koshy et al., 1995; Vichinsky et al., 1995, 1999; Haberkern et al., 1997; Neumayr et al., 1998; Waldron et al., 1999; Hirst and Williamson, 2001; Wali et al., 2003; Buck et al., 2005; Fu et al., 2005; Leff et al., 2007; Al-Samak et al., 2008). Considering these studies together, the practitioner can glean the following conclusions: l Patients with SCD are at much higher risk for periopera-

tive morbidity and mortality, and meticulous attention to preoperative, intraoperative, and postoperative care is warranted. l Different surgical procedures are associated with different degrees of perioperative risk, and it may not be reasonable to require the same preoperative therapy and preparation for all surgical procedures. l Among surgical procedures, adenotonsillectomy, laparotomy, and thoracotomy are common procedures associated with the greatest perioperative risk. l Preoperative transfusion may be associated with improved perioperative outcomes (but not as low as in the population without SCD). l When transfusion is used for preoperative management, simple transfusion is as effective as exchange transfusion in decreasing SCD-related morbidity for most patients. l Simple transfusion is more effective than exchange transfusion in reducing transfusion related morbidity (Table 36-18) (Vichinsky et al., 1995). Data from large, randomized, prospective studies that stratify patients according to age, hemoglobinopathy, surgical procedure, and prior morbidity have not been published. Therefore, the approach to preoperative transfusion should be appropriately individualized after considering the patient’s comorbidities, history of SCD-related complications, proposed surgical procedure, and the opinions of the surgeon, anesthesiologist, and hematologist involved in the patient’s care. Typically, patients with SCD who are undergoing procedures associated with moderate and increased risk (i.e., laparotomy, thoracotomy, and adenotonsillectomy) are managed with simple transfusion to correct anemia. High-risk patients, such as those with stroke and recurrent ACS, and patients who

1136   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-18. Aggressive vs. Simple Transfusion Protocols for Sickle Cell Patients Characteristic

Aggressive Arm

Simple Arm

Hemoglobin S (HbS)

31%

59%

Hemoglobin (Hb)

11.1 g/dL

10.6 g/dL

Units transfused

5

2.5

Hospital days (for transfusion)

4

2.5

Complications

31%

35%

Transfusion-related

14%

7%

Acute chest syndrome

10%

10%

Death

1%

0

Data from Vichinsky EP, et al: A comparison of conservative and aggressive transfusion regimens in the perioperative management of sickle cell disease: the preoperative transfusion in sickle cell disease study group, N Engl J Med 333:206, 1995.

are undergoing very high-risk procedures (i.e., cardiovascular and cerebrovascular surgery) may require an aggressive exchange-transfusion approach that targets a particular HbS percentage (less than 30%). Many patients with a history of stroke are treated with a chronic-transfusion protocol designed to maintain the Hb concentration at 10 to 11 g/dL and the HbS percentage at less than 30% to 40%. For such patients, further preoperative transfusion therapy is probably not necessary provided the patients have been compliant with the chronic-transfusion regimen. Verification of appropriate Hb concentration and HbS percentage through consultation with the referring hematology staff is appropriate in these circumstances. In summary, in most patients undergoing most procedures, some form of simple transfusion to correct anemia is usually provided. Close consultation with the hematology service should be sought to help to determine which low-risk patients and low-risk procedures may be managed without transfusion. To maximize the benefit of transfusion therapy while preventing acute increase in viscosity and risk of stroke, the targeted Hb concentration after transfusion should be 10 to 11 g/dL.

Operative Anesthetic Management of Patients with Sickle Cell Disease Monitoring and maintenance of arterial oxygenation are recommended in the anesthetic management of all patients. Although it is intuitive that maintenance of oxygenation is particularly critical in patients who have SCD, the medical literature contains many descriptions of patients with SCD who were exposed to hypoxemia (i.e., cyanotic congenital heart disease, occlusive tourniquet use, experimental inhalation of hypoxic gas mixtures, and chronic lung disease) and did not suffer immediate SCD-related complications. Similarly, postoperative SCD-related complications continue to occur despite administration of supplemental oxygen and avoidance of intraoperative hypoxemia. Nevertheless, close monitoring of arterial oxygenation and administration of supplemental oxygen are recommended to maintain arterial oxygen saturations in the normal range.

The consequences of Hb desaturation may be mitigated if Hb is promptly reoxygenated after short transit time through the circulation. It is imperative to avoid physiologic perturbations that result in both Hb desaturation and vascular stasis. Preservation of cardiac output, oxygen-carrying capacity, and oxygen saturation, and avoidance of dehydration, hypotension, and vasoconstriction are recommended to minimize conditions conducive to Hb desaturation and prolonged vascular transit time. Maintenance of normal body temperature is also probably advisable during the anesthetic management of patients with SCD. Whereas hypothermia inhibits oxygen unloading at the tissue level and in this sense might be protective against Hb polymerization, it also results in vasoconstriction in vascular beds and prolongs vascular transit time, thus promoting Hb polymerization. Hyperthermia promotes oxygen unloading at the tissue level and may be associated with vasoconstriction or vasodilation. Close monitoring and maintenance of normal body thermal homeostasis are the basic tenets of critical care and anesthetic management. At the tissue level, systemic acidosis promotes—and alkalosis inhibits—oxygen unloading. Clinical evidence linking acidosis to precipitation of SCD events is lacking, and sodium bicarbonate administration has not been shown to prevent SCD complications. Further, it is usually difficult to separate the clinical effects of acidosis from its underlying cause. It is recommended to keep systemic acid-base balance close to the normal range. No particular anesthetic management technique has proven to be more or less effective than others. Arguments can be made for and against regional techniques for patients with SCD. In The Cooperative Study of Sickle Cell Disease, non–SCD-related complications of fever and infection were noted more often in the patients who received regional anesthesia than in those who received general anesthesia (Koshy et al., 1995). However, the group that received regional anesthesia contained more obstetric patients, a subpopulation known to have higher complication rates than other surgical groups. The theoretical rationale for the development of more complications in the regional group includes the presence of compensatory vasoconstriction in areas above the block, lack of controlled ventilation, and the potential for vascular stasis during regional anesthesia (ScottConner and Brunson, 1994). Other authors feel that regional techniques do not increase SCD complication risk and, in fact, have used them to treat complications of SCD, including painful crisis and priapism (Yaster et al., 1994). The anesthetic technique chosen is probably not as important as proper preoperative preparation and meticulous intraoperative attention to control of precipitating physiologic factors. The safety of tourniquet use in orthopedic surgery for patients with SCD has been debated. Although prospective studies that examine the safety of tourniquet use in this population are lacking, retrospective studies supporting its safety have been published (Adu-Gyamfi et al., 1993). If a tourniquet is to be used, it is recommended that the extremity be meticulously exsanguinated before inflating the tourniquet. Otolaryngology procedures are among the most common surgical procedures performed on children. In a 1999 study that compared patients who underwent adenotonsillectomy and myringotomy after simple and aggressive transfusion, 118 patients were enrolled and randomized to one of the two preoperative transfusion regimens, and an additional 47 patients were enrolled and not randomized; 20 of the 47 nonrandomized

C h a p t e r 36    Systemic Disorders   1137

patients were not transfused (Waldron et al., 1999). No differences between transfusion strategies were reported with respect to complications. History of pulmonary disease predicted perioperative SCD-related complications in adenotonsillectomy patients. The authors speculated that because myringotomy is associated with short duration of surgery, minimal blood loss, and surgery remote from the airway, preoperative transfusion might not improve the safety of myringotomy. However, this hypothesis was not specifically studied, and the importance of attention to all other aspects of perioperative care (hydration, oxygenation, and temperature control) was emphasized. Patients with SCD who have sleep-disordered breathing require meticulous perioperative observation, because they have particularly severe nocturnal oxygen desaturation and hypercarbia when compared with age- and gender-matched patients in a control group (who do not have SCD) with sleep-disordered breathing (Kaleyias et al., 2008). Cholecystectomy is another common procedure performed in patients with SCD (Haberkern et al., 1997). Increased use of laparoscopic techniques for gall bladder removal results in less postoperative pain, splinting, and atelectasis. Wales and colleagues (2001) examined the effect of laparoscopic vs. open cholecystectomy in this patient population. Increased risk for ACS in all types of abdominal surgery was noted, irrespective of technique (20% overall, 22% in laparoscopic cases, 15% in open cases), but most of these patients were not transfused. The authors concluded that the postoperative benefits of laparoscopic surgery do not result in decreased SCD complications, but they questioned whether more liberal preoperative transfusion would have resulted in improved outcomes.

Postoperative Management of the Patient with Sickle Cell Disease Most serious SCD-related complications occur in the postoperative period, and therefore the advisability of outpatient surgery for patients with SCD is questionable. VOC is the most common postoperative SCD-related complication. For this reason, it is prudent to comply with the paradigm used by the Preoperative Transfusion in Sickle Cell Disease Study Group, which included postoperative hospitalization, oxygen supplementation, hydration, and pulse-oximeter monitoring. The importance of adequate postoperative analgesia (e.g., oral or IV opioids, NSAIDs, and regional anesthesia techniques) cannot be overstated, because inadequate pain management results in pulmonary splinting, hypoventilation, and ACS. On the other hand, excessive administration of pain medication may result in hypoventilation, atelectasis, and ACS. Postoperative length of stay in the hospital is dictated to a large degree by the type of surgery performed, but patients can be considered for discharge when they are able to sustain oral hydration, are free from fever, can demonstrate good oxygen saturations and pulmonary toilet, and when pain is well controlled on oral analgesics. Other groups have advocated an individualized approach to the perioperative management of surgical patients with SCD. For children undergoing adenotonsillectomy, Duke and colleagues (2006) recommended preoperative outpatient transfusion, liberal oral clear liquid hydration at home, postoperative monitoring with pulse oximeter, and oxygen administration only to those patients with saturations of less than 94%. With this paradigm, which mandated controlled responses to hypoxemia and fever in the postoperative period, complications were noted

that included 20% with hypoxemia, 10% with fever, and 8% with ACS. Most of their patients were discharged from the hospital after fewer than 24 hours without adverse sequelae.

Thalassemia Thalassemia refers to any of several genetic defects in the production of globin chains of Hb. Patients may have deficient production of the α-globin chain (α thalassemia) or the β-globin chain (β thalassemia). The clinical thalassemic syndromes can be understood in terms of the corresponding genotypes. The α-thalassemia syndromes, which contain four allele loci, have more genotypical possibilities than the β thalassemias. The severity of the syndrome is dependent on how many of the four α-globin genes are absent. The absence of functional α-globin genes (–/–) is not compatible with life and leads to death in utero (hydrops fetalis). One functional α-globin gene (–/α-) leads to a severe microcytic, hemolytic anemia called HbH disease. Two nonfunctional genes (αα/–) gives rise to a mild anemia (α thalassemia minor syndrome). A single nonfunctional α-globin gene is clinically silent. For the β thalassemias (with two β-globin gene loci), mutations on both of the β-globin gene loci result in homozygous disease; however, each locus has more than 60 possible mutations that result in heterogeneous clinical syndromes for homozygous patients, depending on which mutant alleles are present at the two loci. For some patients with very dysfunctional mutant genes present at the two loci, severe anemia results with a lifelong transfusion requirement—referred to as β-thalassemia major. Other patients with more functional mutant alleles at these loci also have severe anemia but do not require lifelong ­transfusions—this condition is referred to as β-thalassemia intermedia. Patients who are heterozygous (one normal β-globin gene and one mutant β-globin) have more moderate anemia and produce enough β-globin chains to obviate the need for ongoing transfusions—referred to as β-thalassemia trait, or β-thalassemia minor. In patients with thalassemia, the issues encountered by the anesthesiologist are anemia, the problems associated with chronic anemia, and numerous transfusions. These patients do not recover from surgical blood loss with RBC production, as is expected in other patients. It should be remembered that thalassemia patients can be heterozygous for two different disorders of β-globin production. The most clinically relevant examples are patients who have both a sickle gene and a mutant β-thalassemia gene at the β-globin chain loci. The modifiers β+ and β0 are included as descriptive terminology for these patients. β0 indicates that the patient cannot produce any normal β-globin chains; therefore, patients with S-β0 thalassemia have no HbA. Their clinical course is similar to those of patients with HbSS disease. β+ does not refer to a specific genotype but indicates that the patients can produce variable quantities of normal β-globin chains (depending upon the mutant gene present). Patients with S-β+ thalassemia produce variable amounts of HbA and generally have a more benign clinical course than patients with HbSS. Both groups of heterozygotes warrant the same meticulous perioperative management that is provided to patients with sickle cell anemia.

1138   P a r t  IV    Associated Problems in Pediatric Anesthesia

Oncology Issues Relationship to the Pediatric Anesthesiologist As anesthesiologists play greater roles in the management of pediatric patients with cancer, their interactions with these patients and their families have become increasingly important. From the time children are diagnosed with malignancy until the end of a successful treatment, the anesthesiologist has assumed an increasingly important role in their care. Historically, repetitive painful procedures such as lumbar puncture, bone marrow aspiration, and others were performed with physical restraint under local anesthesia and without sedation. This protocol was minimally effective in relieving pain and anxiety, and families were confronted not only with the anxiety and uncertainty of the diagnosis and prognosis of their children, but also with the dread of repetitive painful procedures. In many pediatric hospitals, an anesthesiologist is present when diagnostic procedures are performed, tumors are removed, central venous access devices are implanted and removed, and radiation treatments are administered. Increasingly, anesthesiologists are also being asked to assist with pain management during end-of-life care, in response to painful complications of chemotherapy, and during recovery from hematopoietic stem cell transplantation (HSCT). Childhood cancer remains one of the leading causes of ­disease-related mortality among pediatric patients, but progressive advances in its management have led to increased numbers of long-term survivors. Indeed, 79% of children diagnosed with cancer before the age of 15 survive for at least 5 years (Jemal et al., 2006). Acute lymphoblastic leukemia and CNS tumors are the two most common malignancies in children, and advances in their management have exposed the anesthesiologist to increased numbers of surviving children with delayed organ toxicity related to treatment. The following section is a review of the anesthetic implications of chemotherapy, radiation therapy, and HSCT.

Myelosuppression Acute and delayed toxicities result from the administration of chemotherapy. Acute toxicities common to most chemotherapy agents include myelosuppression, alopecia, nausea, vomiting, mucositis, and liver dysfunction. Myelosuppression results in profound pancytopenia and is of concern in perioperative management. Neutropenia renders patients susceptible to bacterial and viral infections. Recommended precautions to prevent introduction of bacterial infections include protective and reverse isolation, avoidance of rectal temperature measurement and medication administration, strict hand washing, and aseptic technique during procedures, including before and during medication administration through central venous catheters. Thrombocytopenia is common in oncology patients, and the need for perioperative platelet transfusion is dependent on the type of procedure to be performed, the potential for bleeding, and the function of the existing platelets. Neuraxial anesthesia may be contraindicated in these situations. These decisions should be made in consultation with the patient, the family, and the hematologist directing the patient’s care. The decision as to when to transfuse packed RBCs should also be discussed with the hematologist. In the absence of significant cardiac or pulmonary disease or the need for increased

oxygen-carrying capacity, hematocrit values in the low to mid 20s are common and well tolerated. RBCs may require radiation to prevent graft-versus-host disease (GVHD) in the immunocompromised oncology patient, as well as leukocyte depletion to prevent transmission of cytomegalovirus. Individual chemotherapeutic drugs and radiation produce unique toxicities that are of concern in the anesthetic management of oncology patients (Table 36-19) (Bleyer, 2007). Particularly important among them are medications that, alone or in combination with anesthetic drugs, result in cardiac and pulmonary toxicity, and these are discussed in detail.

Cardiac Toxicity Anthracycline Antibiotics Several chemotherapeutic agents cause cardiac toxicity that may be acute or chronic. The agents most commonly associated with cardiac toxicity are the anthracycline antibiotics doxorubicin, daunorubicin, epirubicin, and idarubicin (Giantris et al., 1998; Singal and Iliskovic, 1998; Balis et al., 2002). Acute cardiac changes associated with anthracycline administration include decreased QRS amplitude, nonspecific ST-T wave changes, supraventricular and ventricular rhythm disturbances, and reduction in ventricular ejection fraction, reaching a nadir at 24 hours after administration. A severe form of this constellation of effects, called myocarditis-pericarditis syndrome, results in congestive heart failure with cardiogenic shock. Chronic toxicity may also occur weeks, months (early form), and years (late form) after anthracycline administration. The early form is associated with cytoplasmic vacuolization, myofibrillary lysis, and degeneration of nuclei and mitochondria. Oxidative damage that results from anthracycline metabolism is believed to be responsible for these changes. Myocardial dysfunction with congestive heart failure that is poorly responsive to cardiotonic medications occurs with a cumulative dosedependent risk. Acute myocardial injury occurs in fewer than 1% of pediatric patients; the incidence of congestive heart failure increases with doxorubicin doses higher than 450 mg/m2 but has been observed with doses as low as 220 mg/m2 (Giantris et al., 1998). The toxic threshold for idarubicin is not known, but doses up to 150 mg/m2 appear to be tolerated (Allen, 1992; Swafford and Gibbs, 1998). Mediastinal irradiation increases the risk of anthracycline-induced myocardial dysfunction. Late-form cardiotoxicity, which occurs 7 to 14 years after therapy, is more common in children and may be related to the inability of the heart to grow with the child. Late toxicity is associated with cumulative doses of doxorubicin more than 300 mg/m2 but has been described to occur in children after lower doses. Children treated with these medications receive follow-up care with baseline and serial echocardiograms for many years after treatment. Risk for death from cardiac-related events is eight times higher among survivors of pediatric malignancies than it is in the general population (Green et al., 2001; Mertens et al., 2001). Risk factors for anthracycline-mediated cardiac toxicity are listed in Table 36-20 (Swafford and Gibbs, 1998). Methods to prevent or minimize anthracycline-induced cardiac disease have been and continue to be investigated. These include limiting cumulative dose and altering administration methods of anthracycline, coupling anthracyclines with liposo­mal carriers and polyethylene glycol, using anthracycline analogues, and adding cardioprotective medications

Folic acid antagonist; inhibits dihydrofolate reductase

Purine analogue; inhibits purine synthesis

Pyrimidine analogue; inhibits DNA polymerase

Alkylates guanine; inhibits DNA synthesis

Alkylates guanine; inhibits DNA synthesis

Binds to DNA, intercalation

Binds to DNA, inhibits transcription

Binds to DNA cleaves DNA strands

6-Mercaptopurine (Purinethol)

Cytarabine (Ara-C)

Cyclophosphamide (Cytoxan)

Ifosfamide (Ifex)

Doxorubicin (Adriamycin) and daunorubicin (Cerubidine)

Dactinomycin

Bleomycin (Blenoxane)

Mechanism of Action or Classification

Methotrexate

Drug

Hodgkin disease, nonHodgkin lymphoma, germ cell tumors

Wilms tumor, rhabdomyosarcoma, Ewing sarcoma

ALL, AML, osteosarcoma, Ewing sarcoma, Hodgkin lymphoma, non-Hodgkin lymphoma, neuroblastoma

Non-Hodgkin lymphoma, Wilms tumor, sarcoma, germ cell and testicular tumors

ALL, non-Hodgkin lymphoma, Hodgkin lymphoma, soft tissue sarcoma, Ewing sarcoma

ALL, AML, non-Hodgkin lymphoma, Hodgkin lymphoma

ALL

ALL non-Hodgkin lymphoma, osteosarcoma, Hodgkin lymphoma, medulloblastoma

Indication(s)

TABLE 36-19. Common Chemotherapeutic Agents Used In Children

Nausea, vomiting, pneumonitis, stomatitis, Raynaud phenomenon, pulmonary fibrosis, dermatitis

Nausea, vomiting, tissue necrosis on extravasation, myelosuppression, radiosensitizer, mucosal ulceration

Nausea, vomiting, cardiomyopathy, red urine, tissue necrosis on extravasation, myelosuppression, conjunctivitis, radiation dermatitis, arrhythmia

Nausea, vomiting myelosuppression, hemorrhagic cystitis, pulmonary fibrosis, inappropriate ADH secretion, bladder cancer, CNS dysfunction, cardiac toxicity, anaphylaxis

Nausea, vomiting, myelosuppression, hemorrhagic cystitis, pulmonary fibrosis, inappropriate ADH secretion, bladder cancer, anaphylaxis

Nausea, vomiting, myelosuppression, conjunctivitis, mucositis, CNS dysfunction; With intrathecal administration, arachnoiditis, leukoencephalopathy, leukomyelopathy

Myelosuppression, hepatic necrosis, mucositis; allopurinol increases toxicity

Myelosuppression, mucositis, stomatitis, dermatitis, hepatitis; With long-term administration, osteopenia and bone fractures; With high-dose administration, renal and CNS toxicity; With intrathecal administration, arachnoiditis, leukoencephalopathy, leukomyelopathy

Adverse Reactions

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Plasma levels must be monitored with high-dose therapy and when low doses are administered to patients with renal dysfunction and leucovorin rescue applied accordingly

Monitory Drug Level

Continued

Requires hepatic activation and is thus less effective in presence of liver dysfunction

Systemic administration may be PO, IM, or IV; may also be administered intrathecally

Allopurinol inhibits metabolism

Systemic administration may be PO, IM, or IV; also may be administered intrathecally

Comments

C h a p t e r 36    Systemic Disorders   1139

Inhibits microtubule formation

Depletion of L-asparagine

Polyethylene glycol conjugate of L-asparagine

Lymphatic cell lysis

Carbamylation of DNA; inhibits DNA synthesis

Inhibits DNA synthesis

Topoisomerase inhibitor

Enhances normal differentiation

Vinblastine (Velban)

L-Asparaginase

Pegaspargase (Pegaspar)

Prednisone and dexamethasone (Decadron)

Carmustine (nitrosourea)

Carboplatin and cisplatin (Platinol)

Etoposide (VePesid)

Etretinate (Tegison) (vitamin A analogue) and tretinoin

Acute progranulocytic leukemia, neuroblastoma

ALL, non-Hodgkin lymphoma, germ cell tumor

Gonadal tumors; osteosarcoma, neuroblastoma, CNS, tumors, germ cell tumors

CNS tumors, non-Hodgkin lymphoma, Hodgkin disease

ALL; Hodgkin disease, nonHodgkin lymphoma

ALL

ALL; AML, when used in combination with asparaginase

Hodgkin disease; Langerhans cell histiocytosis

ALL, non-Hodgkin lymphoma, Hodgkin disease, Wilms tumor, Ewing sarcoma, neuroblastoma rhabdomyosarcoma

Indication(s)

Dry mouth, hair loss, pseudotumor cerebri, premature epiphyseal closure

Nausea, vomiting, myelosuppression, secondary eukemia

Nausea, vomiting, renal dysfunction, myelosuppression, ototoxicity, tetany, neurotoxicity, hemolytic-uremic syndrome, anaphylaxis

Nausea, vomiting, delayed myelosuppression (4-6 wk); pulmonary fibrosis, carcinogenic stomatitis

Cushing’s is syndrome, cataracts, diabetes, hypertension, myopathy, osteoporosis, infection, peptic ulceration, psychosis

Indicated for prolonged asparagine depletion and for patients with allergy to L-asparaginase

Allergic reaction pancreatitis, hyperglycemia, platelet dysfunction and coagulopathy, encephalopathy

Local cellulitis, leukopenia

Local cellulitis, peripheral neuropathy, constipation, ileus, jaw pain, inappropriate ADH secretion, seizures, ptosis, minimal myelosuppression

Adverse Reactions

From Bleyer A: Principles of treatment. In Kliegman RM, et al., editors: Nelson textbook of Pediatrics, ed 18, Philadelphia, 2007, Saunders. IM, Intramuscularly; IV, intravenously, PO, orally.

Inhibits microtubule formation

Mechanism of Action or Classification

Vincristine (Oncovin)

Drug

TABLE 36-19. Common Chemotherapeutic Agents Used In Children—cont’d

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Therapeutic drug monitoring not available or indicated

Monitory Drug Level

Aminoglycosides may increase nephrotoxicity

Phenobarbital increases metabolism, decreases activity

PEG-asparaginase now preferred to L-asparaginase

IV administration only; must not be allowed to extravasate

IV administration only; must not be allowed to extravasate

Comments

1140   P a r t  IV    Associated Problems in Pediatric Anesthesia

C h a p t e r 36    Systemic Disorders   1141

TABLE 36-20. Risk Factors for Anthracycline Cardiac Toxicity Risk Factor

Description

Cumulative dose

Risk <1% for doses <300 mg/m2 5%-10% for doses 350-450 mg/m2 30% for doses >550 mg/m2

Schedule of administration

Risk greatest with bolus administration Less risk with continuous infusion Less risk with dexrazoxane

Mediastinal irradiation

Strong association with increasing risk

Cardiac disease

Preexisting coronary artery disease, valvular heart disease, hypertension

Age

Young children Adults aged >70 years

From Swafford J, Gibbs HR: Cardiac complications of cancer treatment, Anesth Clin N Am 16:598, 1998.

such as dexrazoxane (that chelate iron molecules and prevent anthracycline-induced oxidative injury) into the chemotherapy regimen (Legha et al., 1982; Shapira et al., 1990; Dorr et al., 1991; Herman et al., 1997; Cottin et al., 1998; Creutzig et al., 2001; Lipshultz et al., 2002a; O’Byrne et al., 2002; Safra, 2003; Levitt and Dorup, 2004; O’Brien et al., 2004; Barry et al., 2007). Patients who suffer from anthracycline-induced cardiotoxicity are treated with afterload-reducing medications to minimize ventricular wall stress. The most common medications used for this purpose are the angiotensin-converting enzyme (ACE) inhibitors. ACE inhibitors administered to survivors of pediatric cancers have been beneficial in improving left-ventricle dimensions, fractional shortening, left-ventricle mass, and decreasing afterload (Lipshultz et al., 2002b). However, this benefit appears to be transient, because many patients regress to baseline and deteriorate further after 6 to 10 years of treatment (Lipshultz et al., 2002b, 2004). Other medications used for cardiac dysfunction and congestive heart failure include β-blockers such as carvedilol. The beneficial effects of β-blockers in the setting of idiopathic dilated cardiomyopathy and ischemic heart disease with congestive heart failure in adult patients include reduction in mortality, increase in left-ventricular ejection fraction, and prevention of adrenergically mediated intrinsic myocardial dysfunction and remodeling (Bristow, 1997; Lechat et al., 1998).

Cyclophosphamide Cyclophosphamide, especially in high doses (exceeding 100 to 200 mg/kg), can cause severe congestive heart failure and hemorrhagic myocarditis. Pericardial tamponade has also been described with cyclophosphamide and has resulted in cardiac tamponade. Toxicity occurs with lower doses in children who have also received anthracyclines.

Radiation Radiation, when used for the treatment of thoracic tumors, has also resulted in cardiac toxicity, with early toxicity associated with pericarditis, pericardial effusions, and tamponade.

Radiation dose is expressed in terms of joules deposited per kilogram of body weight (J/kg), which is denoted as Gray (Gy) units. One Gy is equal to 100 rads in the old system; thus, 1 rad is equivalent to 1 cGy. Most long-term effects occur with cumulative doses that exceed 40 Gy and may not become manifest for up to 10 years after treatment (Applefield et al., 1982). The anesthetic management of patients with known or suspected cardiomyopathy necessarily includes cardiology consultation and echocardiographic assessment of myocardial and valvular function. Oncology patients may have clinically silent deterioration in cardiovascular function for many years after completion of therapy. Selection of induction agents, maintenance agents, and invasive hemodynamic monitors is made on the basis of cardiovascular status. Intraoperative fatalities have occurred in children with chemotherapy-induced cardiomyopathy.

Pulmonary Toxicity Many chemotherapeutic agents have some element of pulmonary toxicity. The etiology of pulmonary dysfunction in the immunosuppressed, and especially in stem cell-transplant recipients, may include infectious etiologies, nonspecific lung injury related to dysregulated inflammation and immune function, and toxic injury from medications and radiation. Because of the multifactorial causes for lung injury in oncology patients, it is often difficult to pinpoint the offending agent responsible for the injury.

Alkylating agents Drugs such as carmustine, lomustine, busulfan, melphalan, chlorambucil, and cyclophosphamide are all associated with cytotoxic lung injury. Busulfan may cause lung injury when given as a single agent, whereas the other medications usually do so only in high doses or when part of multidrug therapy. Busulfan lung injury occurs 6 weeks to 10 years after therapy, with an average time interval from treatment to symptoms of 3 years. Dyspnea, fatigue, nonproductive cough, weight loss, unexplained fever, and bi-basilar pulmonary infiltrates are hallmarks of this complication, and it carries a poor prognosis. Bleomycin (a cytotoxic antibiotic) and mitomycin (an alkylating agent) have pulmonary toxicities. Bleomycin lung injury is the prototype for interstitial pneumonitis and pulmonary fibrosis. It has been clearly recognized that a toxic synergistic relationship exists between high inspired-oxygen concentration and bleomycin. Although the exact mechanism is unclear, oxygen concentrations above 30% can rapidly precipitate acute lung injury and acute respiratory distress syndrome in patients who have previously received bleomycin (Maher and Daley, 1993; Mathes, 1995). Mortality from this injury ranges between 13% and 83% in various studies. High-dose corticosteroids have had a beneficial effect in the treatment of this injury (Maher and Daley, 1993). Antimetabolites such as cytosine arabinoside, fludarabine, methotrexate, and 6-mercaptopurine also cause varying degrees of lung injury in dose-related fashion but generally carry better long-term prognoses than does bleomycin­mediated pulmonary toxicity.

Radiation Thoracic radiation causes clinically significant lung injury in 5% to 15% of patients. Several phases of lung injury are described.

1142   P a r t  IV    Associated Problems in Pediatric Anesthesia

The latent or early phase occurs within 1 to 2 months of exposure; the exudative phase occurs 4 to 6 months afterward, and symptoms of pneumonitis develop. The late phase is heralded by the development of pulmonary fibrosis from 6 to 12 months after exposure. Factors that influence the development of pulmonary toxicity include the total dose of irradiation, volume of lung treated, fraction size (radiation given per treatment), and patient age at the time of treatment (younger patients are more prone to toxicity). Pulmonary toxicity has decreased significantly over the past decade because of refined techniques of radiotherapy (Hassink et al., 1993). Toxicity usually does not occur until more than 30 Gy are delivered to more than 50% of the lung when radiation is used alone in adults. The mechanism of injury appears to be different in children younger than 3 years of age, in whom interference with lung and chest wall growth may occur (Miller et al., 1986). In these children, restrictive lung disease has developed with radiation doses as low as 11 to 14 Gy. Anesthetic management of patients who have—or who are at risk for the development of—radiation or chemotherapyinduced pulmonary toxicity requires careful preoperative assessment and intraoperative management. Preoperative assessment may include chest x-ray or CT, spirometric assessment, oxygen saturation or arterial blood gas measurement, and pulmonology consultation. Children at risk for pulmonary toxicity should be treated with the lowest inspired oxygen concentration that provides acceptable oxygen saturation values, and the lowest airway pressures (peak inspiratory pressure and positive end-expiratory pressure) that provide adequate ventilation and preservation of functional residual capacity should be selected.

Hematopoietic Stem Cell Transplantation HSCT has become an accepted therapy for many pediatric disorders, including hematologic malignancies, aplastic anemia, immunodeficiency disorders, congenital hematologic defects, inborn errors of metabolism, and some solid tumors. Pediatric patients who have received HSCT present several unique considerations for anesthetic management, including pretransplant conditioning, acute and chronic GVHD, adverse organ system effects of GVHD, non–GVHD-related morbidity, and unique pharmacologic considerations.

Pretransplant Conditioning All patients who have undergone HSCT first undergo pretransplant conditioning to ablate or impair the native bone marrow. Chemotherapeutic agents commonly employed for this conditioning include cyclophosphamide, busulfan, and fludarabine with or without total body irradiation. In addition to the adverse effects of the chemotherapy agents already discussed, adverse effects of cyclophosphamide include hemorrhagic cystitis and pulmonary fibrosis, and busulfan causes hepatic venoocclusive disease and seizures (Stein et al., 1990; Culshaw et al., 2003). Patients with hemorrhagic cystitis may require blood transfusion or may need anesthesia for cystoscopy and possible suprapubic tube placement. Total body irradiation is associated with restrictive cardiomyopathy, interstitial pneumonitis, and hypothyroidism. Appropriate preoperative evaluation, induction and maintenance drug selection, placement of invasive hemodynamic monitors, and

provision for ­postoperative ­analgesia and monitoring are dictated by cardiovascular status and pulmonary reserve.

Graft-versus-Host Disease Acute GVHD is a clinical syndrome that usually develops in 40% to 60% of patients receiving HSCT by 60 to 100 days after allogeneic HSCT. It is characterized by the development of erythematous skin rash; hepatic involvement with cholestatic jaundice; GI disease marked by the presence of abdominal pain, excessive vomiting, ileus, bleeding, and diarrhea; fever, thrombocytopenia, and anemia; and occasional pulmonary involvement with vascular leak. Chronic GVHD occurs in 20% to 40% of patients who have undergone HSCT and may occur as an extension of acute GVHD, or it may occur in the absence of preceding GVHD (Table 36-21) (Venkatesan and Jacob, 2007). The diagnosis is occasionally made after day 100 of HSCT and is rarely made after more than 500 days after HSCT (Arai and Vogelsang, 2000). Chronic GVHD is a distinctive syndrome that resembles autoimmune collagen vascular disease, with manifestations evident in every organ in the body.

Organ System Considerations in Patients with HSCT The respiratory system is a major source of morbidity among patients receiving HSCT. Diffuse lung injury is a major TABLE 36-21. Clinicopathologic Features of Chronic Graft-Versus-Host Disease System

Features

Systemic

Recurrent infections with immunodeficiency, weight loss, sicca syndrome, debility

Skin

Lichen planus, scleroderma, hyperpigmentation or hypopigmentation, dry scale, ulcerated, freckling, flexion contractures

Hair

Alopecia

Mouth

Sicca syndrome, depapillation of tongue with variegations scalloping of lateral margins, lichen planus and ulcer, angular tightness, salivary gland inflammation, fibrosis

Joints

Decreased range of motion, diffuse myositis/ tendonitis

Eyes

Decreased tearing, injected sclerae, conjunctivae

Liver

Cholestasis, cirrhosis

Gastrointestinal

Failure to thrive, esophageal strictures, malabsorption, chronic diarrhea

Lung

Bronchiolitis obliterans can manifest as dyspnea, cough, wheezing with normal CT scan and marked obstructive ventilatory defects, pneumothorax, chronic sinopulmonary symptoms and/or infections

Heart

Bradycardia, chest pain

Hematopoietic

Refractory thrombocytopenia, eosinophilia

Immune system

Profound immunodeficiency, functional asplenia, risk of pneumococcal sepsis

From Venkatesan T, Jacob R: Anesthesia and graft-versus-host disease after hematopoietic stem cell transplantation, Pediatr Anesth 17:7, 2007.

C h a p t e r 36    Systemic Disorders   1143

c­ omplication of HSCT that accounts for 50% of transplantrelated mortality and may be classified as either infectious or noninfectious. Acute noninfectious lung injury is called idiopathic pneumonia syndrome and refers to widespread alveolar injury that occurs in the absence of left-ventricular dysfunction or respiratory tract infection. Current treatment strategies for idiopathic pneumonia syndrome include oxygen administration, mechanical ventilation with small tidal volumes and low peak-inspiratory pressures, broad-spectrum antibiotic administration, and increased immune suppression. Chronic GVHD may involve the respiratory system, resulting in recurrent sinopulmonary infections and restrictive or obstructive lung disease. A severe form of obstructive lung disease associated with chronic GVHD is bronchiolitis obliterans, which is characterized by progressive dyspnea, cough, wheezing, reduced FEV1, reduced expiratory flow rates, and increased residual lung volume. Small-airway plugging with inflammatory cells, emphysematous distal airways, hyperinflation, interstitial pneumatosis, and pneumothorax may be noted. Patients with bronchiolitis obliterans have a very poor prognosis, and no therapies have been demonstrated to favorably influence the outcome of this disease. Therapy for pulmonary disease with expiratory obstruction associated with chronic GVHD includes administration of supplemental oxygen, bronchodilators, enhanced immune suppression, antimicrobial prophylaxis, and IV immune globulin (Arai and Vogelsang, 2000; Yanik and Cooke, 2006). Patients with pulmonary involvement may need anesthesia for bronchoscopy with bronchoalveolar lavage or biopsy to discern the etiology of pulmonary disease (e.g., infectious vs. bronchiolitis obliterans). Pulmonary function may temporarily worsen after bronchoscopy, especially if bronchoalveolar lavage is performed. The parent/guardian should be informed of this possibility during the consent process. Preoperative evaluation of patients with advanced respiratory disease may include chest radiographs, spirometry, arterial blood gas measurement, and pulse oximetry. Intraoperative management emphasizes monitoring of airway pressure, humidification and warming of airway gases, and continuation of perioperative antibiotics. Patients with severe lung disease may require postoperative mechanical ventilation or placement of an arterial catheter for blood gas analysis. The skin and its appendages are targets in chronic GVHD, resulting in dermal sclerosis with loss of hair and sweat glands. Patients with sclerotic disease have fragile skin prone to tissue trauma and have impaired ability to sweat, resulting in predisposition to the development of hyperthermia. The oral manifestations of acute and chronic GVHD can result in severe mucositis, restricted mouth opening, and difficult orotracheal intubation (Schubert et al., 1984; Arai and Vogelsang, 2000). Ocular manifestations of chronic GVHD include decreased lacrimation, which predisposes patients to corneal ulceration (Grover et al., 1998). The musculoskeletal system may be adversely affected by HSCT, resulting in polymyositis, polyserositis, and myop­ athy (Reyes et al., 1983; Beredjiklian et al., 1998). Joint involvement in the form of flexion contractures and ulcerations may be present. Decreased range of motion between cervical vertebrae and diminished mouth opening should be sought in sclerodermoid chronic GVHD, as these may complicate airway management. Polyserositis may be associated with ascites as well as pericardial and pleural effusions. Careful attention to assessment of potential for oral airway difficulties, positioning,

padding, monitoring of temperature, application of warming devices, eye lubrication, and eye protection is required. The immune system can be assumed to be significantly impaired in recipients of HSCT (Atkinson et al., 1982; Lapp et al., 1985). Continuation of antibacterial, antiviral, and antifungal therapies during the perioperative period is imperative, and many clinicians insert bacterial and viral filters into the breathing circuit. Strict aseptic technique is required during invasive procedures. The rectal route of medication administration and temperature monitoring is discouraged. The GI tract is a major source of morbidity in recipients of HSCT. Acute GVHD manifests with diarrhea, abdominal pain, GI bleeding, ileus, and loss of electrolytes, fluid, and blood. GI tract signs seen in chronic GVHD include esophageal, gastric, and duodenal ulcerations, esophageal web, diarrhea, vomiting, chronic malabsorption, and gastroesophageal reflux (Arai and Vogelsang, 2000). Because of mucosal abnormalities, it is prudent to avoid the use of esophageal stethoscopes and rectal temperature probes. Rapid-sequence induction and intubation should be considered if GI dysfunction appears to put the patient at risk for aspiration pneumonia. Mucositis and swelling of oral structures may impede visualization of airway structures. Hepatic dysfunction is common in GVHD. Acute GVHD results in cholestatic jaundice, and chronic GVHD is associated with elevation of alkaline phosphatase, bilirubin, and transam­ inases. Many patients with GVHD are also dependent on parenteral nutrition, which may exacerbate hepatic injury. Selection of anesthetic agents with minimal hepatic metabolism and toxicity and minimal effects on hepatic blood flow may be indicated. Close attention to glucose homeostasis is necessary to prevent hypoglycemia under anesthesia. Monitoring and correction of coagulation parameters are essential for operative procedures associated with significant bleeding and when regional anesthesia is contemplated. Hematologic dysfunction after HSCT may manifest as refractory thrombocytopenia, isolated neutropenia or anemia, or pancytopenia. Adequate blood and blood products should be available for cases in which blood loss is expected to be high and when hemostasis is critically important at the surgical site (CNS and ocular cases).

Non–Graft-versus-Host Disease HSCT Complications Complications of HSCT not related to GVHD include renal dysfunction, hepatic venoocclusive disease, and impaired thyroid and adrenal function. Renal dysfunction occurs commonly as a result of nephrotoxic medications and immunosuppressive therapy. Preoperative assessment of renal function, careful attention to intravascular volume status, preservation of cardiac output, and avoidance of nephrotoxic medications (when possible) are the cornerstones of preservation of renal function. Hepatic venoocclusive disease is a nonspecific hepatic injury that occurs after the administration of some drugs used for pretransplant conditioning (e.g., busulfan or actinomycin D). The mechanism is unclear, but pathology involves vasculitis of the sinusoids of the hepatic venules. Hepatic venoocclusive disease initially presents as jaundice, right upper-quadrant pain, and ascites, but it may progress to include coagulopathy and encephalopathy. Thyroid and adrenal function are critical for cardiovascular stability and drug clearance. Endocrino­ logy consultation and appropriate hormone replacement are recommended.

1144   P a r t  IV    Associated Problems in Pediatric Anesthesia

Pharmacologic Considerations Immune-modulating medications used in HSCT include cyclosporine A, azathioprine, tacrolimus, mycophenolate mofetil, and corticosteroids. Relevant side effects of azathioprine include pancytopenia, and those of steroids include hypertension, diabetes, and neurotoxicity. Hypertension, seizures, diabetes, and renal insufficiency are toxicities associated with cyclosporine and tacrolimus (Kostopanagiotou et al., 2003). Adverse effects of immunosuppressive medications particularly relevant to children include growth retardation, hirsutism, obesity, and osteoporosis (Ellis et al., 2000). In the presence of adequate renal and hepatic function, no contraindications exist to the use of anesthetic agents. Sevoflurane can be used for the induction of anesthesia, and sevoflurane, desflurane, and isoflurane, can be used for the maintenance of anesthesia. IV agents and local anesthetics can also be used. Cyclosporine increases the effects of nondepolarizing muscle relaxants, resulting in decreased dosage requirement. Neuromuscular blockade should be monitored closely to prevent inadvertent overdose. The use of cisatracurium is preferred with renal or hepatic impairment. Nitrous oxide inhibits bone marrow function after prolonged exposure (longer than 6 hours) at high concentrations (Weimann, 2003). Methotrexate inhibits folate metabolism as well as marrow function. Nitrous oxide may amplify the side effects of methotrexate therapy and therefore should probably be avoided for long procedures in patients who are treated with methotrexate and in patients with marginal marrow function. NSAIDs may potentiate nephrotoxicity of cyclosporine and tacrolimus. Potent anesthetic opioids (e.g., fentanyl, sufentanil, and alfentanil) and some of the benzodiazepines (e.g., midazolam and diazepam) are metabolized by the cytochrome P450 enzymes (Leather, 2004). Serum concentrations of these drugs are elevated if drugs that inhibit the cytochrome P450 system (e.g., azole antifungals, calcium-channel blockers, and macrolide antibiotics) are coadministered. Calciumchannel blockers, azole antifungals, and macrolide antibiotics increase levels of tacrolimus and cyclosporine through enzyme inhibition. Children with extensive systemic involvement after HSCT as well as patients undergoing extensive surgical procedures may require postoperative admission to an intensive care unit as a result of limited physiologic reserve. Neuraxial blocks, local anesthesia infiltration, and parenteral opioids are appropriate methods for the provision of postoperative analgesia under close monitoring.

COAGULATION: DEVELOPMENTAL ASPECTS, DISORDERS OF COAGULATION, AND PERIOPERATIVE MANAGEMENT OF HEMOSTASIS Coagulation abnormalities provide many challenges for the pediatric anesthesiologist. Children with known coagulation disorders require disease-specific perioperative management and must often be treated for complications of their bleeding diathesis. More challenging is the intraoperative investigation and management of children who develop a coagulopathy in the OR, either from preexisting, albeit undiagnosed, diseases or from acquired disorders. The following sections focus on

endogenous control of hemostasis, developmental changes in coagulation, and commonly inherited coagulopathies and their management.

Overview of Hemostasis The hemostatic system is designed to maintain blood in a fluid state until vessel injury occurs, at which point an explosive cascade of events terminates blood loss by sealing off the vascular defect. Hemorrhage occurs if the response is inadequate; thrombosis occurs if the response is dysregulated. The vascular endothelial cell is at the fulcrum of this delicate balance. The normal endothelial cell maintains blood in its fluid state by inhibiting platelet aggregation and blood coagulation through the production of prostacyclin, NO, and the ectonucleotidase CD39, as well as by promoting fibrinolysis through the antithrombin III-mediated conversion of plasminogen to plasmin (Ignarro et al., 1987; Marcus et al., 2005). Physically, the endothelial cell is a barrier between the platelets and procoagulant proteins derived from reactive components present in the deeper layers of the vessel wall. These components include collagen, fibronectin, von Willebrand factor (vWF), and tissue factor (TF), all of which stimulate platelet adhesion and aggregation and trigger the coagulation cascade (Fig. 36-12).

Primary Phase of Hemostasis The platelet is central to the primary phase of hemostasis (Fig. 36-13). The normal circulating platelet count ranges from 150,000 to 400,000/mL. An additional 33% of all platelets are

Collagen Tissue factor Thrombus

Endothelium

Smooth muscle

Subendothelial matrix n  FIGURE 36-12. Response to vascular injury. Collagen and tissue factor (TF) associated with the vessel wall provide a hemostatic barrier to maintain the high-pressure circulatory system. Collagen (white arrows), located in the subendothelial matrix beneath the endothelium, is not exposed to flowing blood under normal conditions. TF (black arrows), located in the medial (smooth muscle) and adventitial layers of the vessel wall, comes in contact with flowing blood when the vessel is disrupted or punctured. Both collagen and thrombin initiate thrombus formation. Collagen is a first line of defense, and TF is a second line of defense. (From Furie B, Furie BC: Mechanisms of thrombus formation, N Engl J Med 359:938, 2008.)

C h a p t e r 36    Systemic Disorders   1145 Cell surface

Tissue factor

Factor XIa

Factor VIIa

Factor IXa Tissue-factorpathway inhibitor

Factor VIIIa

Factor Xa Factor Va

Antithrombin III, heparin

Factor IIa

Factor IIa

Fibrinogen Plasminogen

Fibrin

u-PA t-PA

Plasmin

Thrombomodulin Protein S Protein C n  FIGURE 36-13. The clotting cascade. Coagulation is initiated by the exposure of blood to TF bound to cell membranes. TF interacts with factor VIIa to convert factor IX to factor IXa and factor X to factor Xa (only the activated forms are shown). Factor IXa converts factor X to factor Xa. Factor Xa generates factor IIa (thrombin) from factor II (prothrombin). Each of these reactions takes place on an activated cell surface. Once factor IIa is generated, it cleaves plasma fibrinogen to generate fibrin. The TF-pathway inhibitor forms a quaternary structure with TF, factor VIIa, and factor Xa (shown in blue). The thrombomodulin-protein C-protein S pathway (shown in yellow) inactivates factors Va and VIIIa. Antithrombin III inactivates factors XIa, IXa, Xa, and IIa (shown in orange) in a reaction that is accelerated by the presence of heparin sulfate. In the fibrinolytic pathway, tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA) convert plasminogen to plasmin. Once generated, plasmin proteolytically degrades fibrin (shown in purple). (From Rosenberg RD, Aird WC: Vascular-bed-specific hemostasis and hypercoagulable states, N Engl J Med 340:1555, 1999.)

sequestered within the spleen. After vascular injury, the affected vessel constricts proximally, diverting blood flow away from the site of endothelial disruption. Extravasated blood is exposed to subendothelial structures, and the platelets are stimulated by their contact with collagen. The platelets become adherent to the subendothelial infrastructure, anchored by the binding of vWF to the platelet surface glycoprotein Ib, as well as by the interaction of platelet glycoprotein VI with collagen (Ruggeri, 2000). Once platelet adhesion occurs, platelet activation results in the glycoprotein VI-mediated release of platelet agonists such as epinephrine and serotonin from the dense granules (Furie

and Furie, 2008). The synthesis of thromboxane-A2 (TXA2) from arachidonic acid by COX and thromboxane is then released from the platelet lipid membrane, and there is a 50% increase in the number of platelets and a conformational change in the platelet surface glycoprotein receptor IIb-IIIa (GPIIb-IIIa), which binds to fibrinogen, vWF, and fibronectin (Shattil et al., 1998). GPIIbIIIa is now referred to as αIIbβ3 (Bennett, 2005). TF can also trigger platelet activation independent of endothelial disruption, vWF contact, or glycoprotein VI (Dubois et al., 2006). The TF VIIa (factor VIIa) complex activates factor IX, which ultimately generates thrombin. Thrombin then cleaves the platelet surface receptor Par1, activating the platelet (Dubois et al., 2007). A developing thrombus recruits unstimulated platelets. Platelet activation increases integrin αIIbβ3’s affinity for fibrinogen and vWF, clustering this receptor on the platelet surface (Bennett, 2005). Platelet aggregation, through the linkage of αIIbβ3 to the αIIbβ3 on other platelets bridged by fibrinogen or vWF, increases the size of the initial platelet plug, creating a mass of aggregated platelets at the injury site. At low shear rates, fibrinogen is the principle ligand, whereas at higher shear rates vWF plays a predominant role (Goto et al., 1998). Synthesized TXA2 induces aggregation of other platelets and promotes vascular smooth muscle constriction, producing local vasoconstriction, which limits blood loss and increases the effectiveness of the platelet plug by decreasing the effective surface area that the platelet plug needs to cover. TXA2 amplifies the platelet’s responses to weak agonists such as ADP and epinephrine (Funk, 2001). Not all of the recruited platelets undergo activation. Many remain loosely associated but inactive and ultimately fall away from the initial platelet plug (Dubois et al., 2007). TXA2 binds to a G protein-coupled receptor on the platelet surface, leading to an increase in intracellular calcium and activation of protein kinase C. The initial platelet plug is friable. It is stabilized by the increase in platelet cytosolic calcium, which initiates actin filament turnover that in turn modulates the cytoskeletal changes. These changes enable integrin aIIbß3 clustering and the binding of both fibrinogen and vWF. The conformational change of aIIbß3, induced through ligand binding, exposes new ligand binding sites and further merging of platelet surface receptors, resulting in clot retraction (Warltier et al., 2002).

Secondary Phase of Hemostasis The exposure of subendothelial structures to circulating blood simultaneously activates the coagulation cascade (the secondary phase of hemostasis) to produce a cross-linked fibrin clot. Of the coagulation proteins, prothrombin (II), protein C, protein S, and factors VII, IX, and X are synthesized as prozymogens and activated to serine proteases through a vitamin K-dependent hepatic enzyme (Furie and Furie, 1990). This modification is required for calcium binding, serving as a bridge for binding the factors to the phospholipid surface.

New Model of Cell-Based Coagulation In earlier schemes, the coagulation cascade had been divided into intrinsic (e.g., factors XII, XI, IX, and VIII) and extrinsic (e.g., TF and factor VII) pathways (Fig. 36-14). This model is primarily useful for the interpretation of the in vitro laboratory

1146   P a r t  IV    Associated Problems in Pediatric Anesthesia n  FIGURE 36-14. Mechanism of hemostasis. HMWK, High-molecularweight kininogen; vWF, von Willebrand factor.

Kallikrein F XII

XIIa

F XI

XIa

HMWK Prekallikrein

Xa

VII

VIIa

Ca++ F XIX

IXa IIa or Xa

vWf

F VIII

PI Ca++

Tissue factor

VIIIa FX

Xa Ca++

Prothrombin (F II) Fibrinogen (I)

FV

IIa

PI Va

Thrombin (IIa) Fibrin XIIIa

IIa

FXIII

Clot

tests, the activated partial thromboplastin time (aPTT) and the prothrombin time (PT). The common pathway of the clotting cascade is the production of factor Xa, which in concert with factor Va, cleaves prothrombin to thrombin, resulting in fibrin production. It became apparent that deficiencies in the intrinsic pathway did not produce bleeding conditions and that the model was not clinically relevant. In vivo, the critical component in coagulation initiation is TF, a membrane protein receptor for factor VII. TF is expressed constitutively on cells that are not in direct contact with blood, such as vascular smooth muscle, fibroblasts, and macrophages, forming a hemostatic envelope around the vascular endothelium (Mann et al., 1998). The endothelium acts as a barrier that separates the cellular sources of TF from factor VIIa in flowing blood, preventing inadvertent coagulation initiation. Once endothelial injury occurs, activated endothelial cells and platelets activate TF through the release of protein disulfide isomerases (Cho et al., 2008; Reinhardt et al., 2008). Activated TF forms a complex with factor VII, producing factor VIIa. The TF/factor VIIa complex then activates factors X and IX (Orfeo et al., 2005). Factor IXa binds to factor VIII, forming the “tenase complex” (factor IXa/VIII). This complex inefficiently activates factor X. The TF/factor VIIa complex also converts factor X to factor Xa directly. Factor Xa combines with its cofactor (factor Va) on the TF-bearing cell and generates small amounts of thrombin. Although this limited amount of thrombin is sufficient to activate platelets in the local area and to activate factors V, VIII, and IX, it is insufficient to cleave fibrinogen. Factor VIII is cleaved off vWF and is then activated to factor VIIIa, which complexes with factor IXa to form the “activated tenase complex” (factor VIIIa/IXa). The activated tenase complex binds to the negatively charged phospholipids that coat the surface of activated platelets and produce factor Xa. This platelet-localized factor Xa binds to factor Va, also bound to the platelet surface, and catalyzes the conversion of prothrombin to thrombin at the rapid rate necessary for adequate hemostasis. Thrombin formation is accelerated, fibrinogen is cleaved, and the coagulation complex is

further activated, augmenting thrombin production. The fibrin monomers undergo spontaneous polymerization to form the fibrin clot, which is then stabilized by crosslinking, mediated by factor XIIIa. A firm platelet-fibrin clot results, which over the course of time, mediated by platelets, decreases in size. The graphic representation of this new model of cell-based coagulation, which replaces the former intrinsic pathway, is shown in Figure 36-15 (Gailani and Broze, 1991). Although factor XII is not involved in physiologic hemostasis, accruing evidence suggests that it plays a key role in abnormal thrombosis (Renne and Gailani, 2007). Extracellular ribonucleic acid (RNA) released from damaged cells induces arterial thrombosis through factor XII activation (Kannemeier et al., 2007). Inorganic polyphosphates released from activated platelets trigger factor XII activation and provide an additional pathway for fibrin formation by stimulated platelets (Johne et al., 2006).

Modulators of Coagulation Coagulation is modulated by a number of plasma proteins, the most important of which are antithrombin III, thrombomodulin, TF pathway inhibitor, protein C, protein S, and factor V. Antithrombin III is a potent inhibitor of thrombin and factors IXa, Xa, and XIIIa. Heparin potentiates the inhibition of thrombin by antithrombin III. Thrombomodulin, expressed on the endothelial cell surface, binds and neutralizes thrombin. TF pathway inhibitor limits factor Xa production by the activated tenase complex (factor IXa/VIIIa). Activated protein C (APC), activated by the presence of thrombin and accelerated by thrombin-thrombomodulin, proteolytically inactivates factor VIII/ VIIIa and factor V/Va, thereby down-regulating the production of factor Xa and thrombin. Protein S has two anticoagulant roles; it acts as a cofactor for APC in the inactivation of factors Va and VIIIa and as a cofactor for TF pathway inhibitor in the inactivation of factors Xa and VIIa (Castoldi and Hackeng, 2008). Factor V uniquely has procoagulant and anticoagulant

C h a p t e r 36    Systemic Disorders   1147

EC EC

Endothelial cell injury results in exposure of subendothelial cell collagen

Collagen

Multimers of vWF have specific binding sites for both collagen and platelet GP Ib activated by high shear flow.

GP Ib Platelet

Fibrinogen

GP IIb IIIa

vWF Fibronectin

GP IIb IIIa

Thrombospondin

Thromboxane A2

Tissue factor pathway

XIa

PK K

II Common pathway

IXa·VIII

Fibrinogen

Fibrin XIIIa monomer X-linked FSP fibrin

Contact activation pathway TPA TPA PAI-1 X-linked fibrin TPA Plasminogen Plasmin

EC

TPA

Exposure of subendothelial cell TF activates fibrin formation through the dominant tissue factor pathway

XI K

BK

EC

X

IIa IX

XIIa HMWK

Exposed TF complexes with and activates circulating VII to activate tissue factor pathway leading to crosslinked-fibrin clot Xa·V

VIIa

VII

Stimulates vessel contraction Stimulates aggregation of passing platelets

ADP

TF

Platelet adherence to collagen via vWF stimulates intracellular signal transduction resulting in platelet shape change, granule release and formation of the heterodimer platelet receptor GP IIb IIIa. Platelets aggregate via intercellular links; contractile proteins pull together the platelet plug.

FSP D-dimer

Contact activation pathway causes generation of bradykinin and kinin resulting in activation of inflammation and fibrinolysis. Endothelial cell release of TPA mediates generation of plasmin that lysis clot

Plasmin Antiplasmin

n  FIGURE 36-15. Hemostatic mechanism: cell-based coagulation model. BK, Bradykinin; EC, endothelial cell; FSP, fibrin split products; GP, glycoprotein; K, kallikrein; PAI, plasminogen activator inhibitor; PK, prekallikrein; TF, tissue factor; TPA, tissue plasminogen activator; vWF, von Willebrand factor. (Redrawn from Manco-Johnson M, Nuss R: Hemostasis in the neonate, Neoreviews 1:191, 2000.)

activities. Genetic alterations in factor V (i.e., factor V Leiden) or deficiencies of protein C may result in excessive procoagulant activity, which may lead to venous thromboembolism (Tormene et al., 2002). When factor V is cleaved by thrombin, a number of intermediates are formed in addition to factor Va, the essential cofactor to factor Xa. These intermediates are cofactors for APC and act as anticoagulants. APC can cleave factor V directly, producing an anticoagulant and precluding factor V’s transformation to factor Va (Thorelli, 1999).

Fibrinolysis Fibrinolysis occurs simultaneously with the initiation of clot formation, limits thrombosis to the local area of injury, and begins the processes of clot revision, vascular damage repair,

and ultimately, vessel recanalization. During the initial phase of hemostasis, endothelial cells and platelets release plasminogen activator inhibitors (PAIs), which facilitate fibrin formation. In response to thrombin, endothelial cells begin to release tissue plasminogen activator (TPA) that along with prourokinase, converts plasminogen to plasmin. The plasminogen, which is bound to fibrin in the hemostatic plug, is much more reactive to TPA than circulating plasminogen. After plasmin is produced locally at the site of the hemostatic plug, fibrinolysis or fibrin degradation can occur. Fibrinolysis at the hemostatic plug is opposed by ongoing coagulation and by antifibrinolysis, mediated by α2-plasmin inhibitor, which is also bound to fibrin. The spectrum of endothelial cell and platelet interactions in the setting of clotting factors, adhesive proteins, fibrinolytic proteins, and the myriad inhibitors promotes an equilibrium

1148   P a r t  IV    Associated Problems in Pediatric Anesthesia

that promotes fluidity of circulating blood, and localization of hemostasis and injury repair. Derangements in any portion of this precariously balanced mechanism can lead to a hemorrhagic or thrombotic complication. A defect in clot formation in the setting of physiologic fibrinolysis leads to bleeding, as does normal clot formation in the setting of premature fibrinolysis. Thrombosis can occur in the setting of endothelial cells expression of TF, reduction in antithrombin function, and when platelet aggregation and activation are excessive.

Antiplatelet and Anticoagulant Drugs The increased understanding of physiologic hemostasis has led to the development of new antithrombotic agents to modulate the consequences of thrombosis in critical illness and many systemic diseases. Antiplatelet drugs target key steps in platelet secretion, adhesion, and activation. COX-1 inhibitors such as acetylsalicylic acid (ASA) inactivate platelet COX-1, a pivotal enzyme in TXA2 and prostacyclin synthesis, important for platelet secretion and aggregation. Some COX-1 drugs inhibit the enzyme transiently, whereas others (e.g., ASA) permanently inhibit TXA2 production for the lifespan of the platelet. ADP antagonists such as ticlopidine irreversibly block platelet receptor P2Y12, inhibiting platelet aggregation by the platelet agonists, thrombin, epinephrine, ADP, and collagen (Savi and Herbert, 2005). αIIbβ3-receptor antagonists block platelet integrin αIIbβ3 from binding to fibrinogen and vWF on activated platelets, thus reducing platelet aggregation by 80% (Warltier et al., 2002; Patrono et al., 2004). Anticoagulant drugs are designed to perturb the precariously balanced factor-based physiologic coagulation system and its modulators. Vitamin K antagonists, exemplified by warfarin, inhibit the vitamin K-dependent γ-carboxylation of factors II, VII, IX, and X, and proteins C and S (Levy et al., 2008). The antithrombin (AT) III agonist unfractionated heparin binds to AT, resulting in a 1000-fold increase in both thrombin and factor Xa inactivation. Factor Xa inhibitors such as low­molecular-weight heparin (LMWH) and fondaparinux also bind to AT, augmenting factor Xa inhibition; however, these drugs are too small to simultaneously bind both AT and thrombin, which is necessary for thrombin inactivation. Increasingly, LMWH is replacing unfractionated heparin in the treatment of pediatric thrombotic disease (Dix et al., 2000; Merkel et al., 2006). Direct thrombin inhibitors lepirudin, bivalirudin, and argatroban, bind to thrombin’s active site, preventing the conversion of fibrinogen to fibrin, both in the circulation as well as on the evolving thrombus. Argatroban has been successfully used as an anticoagulant for cardiopulmonary bypass in an infant with heparin-induced thrombocytopenia (Malherbe et al., 2004).

Laboratory Evaluation of Coagulation The panel of tests routinely used to evaluate coagulation includes the platelet count, PT, aPTT, and bleeding time. The normal platelet count is between 150,000 and 500,000/mcL, but increased bleeding as a result of thrombocytopenia rarely occurs at counts above 50,000/mcL. Bleeding may also occur when platelets are relatively normal in number but dysfunctional with regard to their role in coagulation.

The normal PT, which ranges from 11.5 to 14 seconds, reflects normal amounts of factors II, V, VII, and X, which are the vitamin K–dependent factors. Defects in vitamin K–dependent clotting factors may be caused by a deficiency of the vitamin, poor responsiveness to the vitamin because of liver disease, or exposure to warfarin agents that impair vitamin K’s transition to the reduced form. The International Normalized Ratio (INR) can be used to estimate the degree of factor deficiency. An INR in the range of 2 to 3 correlates with factor concentrations of 10%; between 3 and 4 correlates with concentrations of 5%; and an INR higher than 4 correlates with concentrations of 1% (Boulis et al., 1999). The normal aPTT is between 25 and 40 seconds and requires normal levels of vWF and factors XII, XI, IX, and VIII. The bleeding time test assesses the integrity of the vascular and platelet aspects of coagulation and is prolonged in patients with reduced numbers of platelets (fewer than 100,000/mcL), in patients with defective platelet function (e.g., after aspirin, NSAID, or valproic acid administration), and in patients with vWD. The test is performed using a standardized template device to guide the incision with a tourniquet placed on the arm at 40 mm Hg for children, 30 mm Hg for newborns at term, and 20 mm Hg for premature infants. The bleeding-time evaluation has suffered from a reputation of being difficult to perform and poorly reproducible, especially in newborns and infants. Results may be affected by skin thickness and the device used to puncture the skin (e.g., surgical blade or manufactured bleeding time device), location of incision, and body temperature. Recently, a bleeding-time device (Surgicutt; ITC, Edison, New Jersey) has been developed that comes in three sizes for adults, children (5 months to 15 years), and newborns (up to 5 months). The blades are different lengths (and depths for newborns) and afford a standardized incision rather than a puncture. This standardization should decrease the variability in bleeding times that result from varying incision techniques and skin thicknesses related to age, but this hypothesis has not yet been studied in a controlled fashion. The blood is blotted with filter paper every 30 seconds. Normal bleeding time is between 4 and 8 minutes in adults and may be as low as 2 minutes in newborns. The bleeding time is prolonged in very–low-birth-weight infants with hematocrit of less than 28%; bleeding time may be reduced in these infants after transfusion to hematocrit of more than 28% (Sola et al., 2001). It is unknown whether the longer bleeding times at lower hematocrits are associated with an increased risk of clinical bleeding.

Developmental Hemostasis The hemostatic system of the newborn and young child is significantly different than that of the adult. Although considered immature, the hemostatic system is functional; the young child is successfully protected from hemorrhagic and thrombotic complications. These differences are most exaggerated in the hemostatic mechanism of the newborn. As the coagulation and fibrinolytic factors do not cross the placenta, the proposed causes for the differences in the newborn’s hemostatic system include decreased factor synthesis, enhanced clearance, general activation of the coagulation system with resulting factor consumption, and the synthesis of less active fetal forms of some proteins.

C h a p t e r 36    Systemic Disorders   1149

TABLE 36-22. Concentration of Coagulation Factors in the Term and Preterm Newborn Relative to Adult Values

1.2

U/mL

1.0

Level at Preterm (% of Adult Value)

Factor

0.6

Thrombin

  50

  40

Factor VII

  66

  66

Factor IX

  50

  35

II VII IX X

0.4 0.2 0 Fetal

Preterm

Full term

0.5

1–5 6–10 11–18 Adult

Years 1.2 1.0 0.8 U/mL

Level at Term (% of Adult Value)

0.8

0.6 V VIII XI XIIIa

0.4 0.2 0 Fetal

Preterm

Full term

0.5

1–5 6–10 11–18 Adult

Years n  FIGURE 36-16. Developmental hemostasis: changes in plasma concentration of coagulation proteins over the course of development. (Redrawn from Andrew M: Developmental hemostasis: relevance to thromboembolic complications in pediatric patients, Thromb Haemost 74[Suppl]:415, 1995.)

One of the most notable features of the coagulation system of the infant is that plasma clotting factors are inconsistently different from adult levels (Fig. 36-16). The most well-known ontogenetic differences in the hemostatic system involve the vitamin K–dependent factors. These proteins are present in low levels at birth, and coagulation is severely impaired in the absence of vitamin K supplementation; this is called hemorrhagic disease of the newborn (discussed more fully in the next section). The levels of the four vitamin K–dependent coagulation factors (factors II, VII, IX, and X) and the contact factors (factors XI and XII, prekallikrein, and kininogen) are all less than 50% of adult values and slowly rise to within 20% of adult levels by 6 months of age (Andrew et al., 1990a). Factor VII levels increase rapidly after birth in premature and full-term infants. Factors II and VII remain less than adult values for most of childhood (Andrew et al., 1992). Factor IX activity can be as low as 15% of adult levels and reach adult levels at 9 months of age. On the other hand, the levels of fibrinogen and factors V, VIII, and XIII are normal at birth; fibrinogen and factor VIII levels are at the high end of the normal range (Andrew, 1997). vWF levels are increased during the first weeks of life and return to adult levels

Factor X

  40

  40

Fibrinogen

100

100

Factor V

  75

  88

Factor VIII

100

110

Factor X

  40

  40

Factor XI

  40

  40

Factor XII

  50

  30

Factor XIII

  75

  80

Heparin cofactor II

  25

  33

Antithrombin

  50

  40

α2-Macroglobulin

150-200

150-200

Protein C

  25

  10

Protein S

  40

  25

Plasminogen

  50

  55

α2-Antiplasmin

100

130

Tissue plasminogen activator

200

200

Plasminogen activator inhibitor

180

180

between 2 and 6 months of life (Thomas et al., 1995). The net effect of these differences in the newborn’s hemostatic system is delayed thrombin generation, similar to that seen in adults on receiving anticoagulant medications such as warfarin or subcutaneous heparin (Andrew et al., 1990b). The concentration of coagulation factors in term and preterm infants relative to adult values is summarized in Table 36-22.

Laboratory Evaluation of Coagulation in the Newborn Newborns with suspected coagulopathies are evaluated with PT, aPTT, platelet count, and levels of fibrinogen and fibrin degradation products. In newborns, as a result of the developmental differences in the coagulation protein levels, both PT and aPTT are prolonged compared with adult values. Because of the low plasma concentrations of many of the clotting factors in the newborn, the aPTT is markedly elevated. The developmental progression of the PT and aPTT in term and preterm infants is summarized in Table 36-23. Infants with vitamin-K deficiency have a prolonged PT when compared with age-appropriate norms. Fibrinogen concentrations less than 100 mg/dL and platelet counts of less than 100,000/mcL are indicative of a pathologic process (Manco-Johnson, 2008).

1150   P a r t  IV    Associated Problems in Pediatric Anesthesia

TABLE 36-23. Developmental Changes in the Prothrombin Time and Activated Partial Thromboplastin Time Age of Infant

Day 1

Day 5

Day 30

Day 90

Adult

PT

  13 ± 1.4

12.4 ± 2.5

11.8 ± 1.3

11.9 ± 1.2

12.4 ± 0.8

aPTT

42.9 ± 5.8

42.6 ± 8.6

40.4 ± 7.4

37.1 ± 6.5

33.5 ± 3.4

PT

  13 ± 1.5

12.5 ± 1.3

11.8 ± 0.9

12.3 ± 1.2

12.4 ± 0.8

aPTT

53.6 ± 13

50.5 ± 12

44.7 ± 9

50.5 ± 12

33.5 ± 3.4

Term

Preterm (30-36 weeks)

Developmental Changes Beyond the Neonatal Period Many of the proteins that regulate coagulation and thrombin generation are also decreased in early infancy. Antithrombin III and heparin cofactor II are markedly decreased to levels that might predispose to spontaneous thromboembolic events. The α2 macroglobulin levels at birth are greater than those of adults and remain so until the third decade of life. By 6 months of life, AT III levels exceed levels seen in adults. Protein C and protein S concentrations at birth are also substantially less than those seen in adults and remain low throughout childhood (Andrew et al., 1992). Fibrinolysis is suppressed throughout childhood. During childhood, plasminogen levels increase to adult levels, but the TPA/PAI-1 ratio is significantly lower than that in adults, which explains the decrease in fibrinolysis in children (Siegbahn and Ruusuvaara, 1988).

Developmental Aspects of Platelet Number and Function Although normal in number, neonatal platelets are hyporeactive. Neonatal platelet aggregation is diminished in response to certain physiologic agonists such as ADP, thromboxane, and epinephrine (Levy-Shraga et al., 2006). Intracellular calcium transport is decreased as well, resulting in diminished granule release and slowed conformational change (Kuhne and Imbach, 1998). Increased vWF levels in the newborn period contribute to decreased bleeding times (Andrew, 1997). Platelet function in the neonate is aided by a relatively high hematocrit, causing an increased concentration of platelets directed to the wall by the dynamics of laminar flow. Platelet function improves over the first 2 weeks of life (Israels, 2009). The most serious manifestation of thrombocytopenia in the newborn period is intraparenchymal brain hemorrhage. Of neonates with spontaneous intraparenchymal hemorrhage, one third have an associated coagulopathy such as vitamin-K deficiency, hemophilia, or thrombocytopenia. Neonates with a platelet count of less than 50,000/mcL are at significant risk for intraparenchymal hemorrhage (Sandberg et al., 2001; Jhawar et al., 2003).

Inherited Coagulopathies Hemostasis usually requires activity levels of coagulation factors at least 30% of normal. The aPTT may be normal with a factor level as low as 15% to 18% of normal. Past medical

­ istory and family history are invaluable tools in the evaluation h of a bleeding patient. Substantial hemorrhage after oral cavity manipulation (whether by dentist or toothbrush) is often a sign of an underlying bleeding disorder, reflecting an imbalance between abnormal clot formation and normal salivary fibrinolysis. Patients with mild bleeding disorders who have never had trauma or surgery may experience symptoms rather late in life and may have normal screening tests. Prenatal diagnosis of most congenital factor deficiencies can now be made from fetal DNA.

Hemophilia Of the inherited deficiencies of coagulation factors, the most common are the X-linked recessive hemophilias; hemophilia A is factor VIII deficiency, and hemophilia B (Christmas disease) is factor IX deficiency. Thirty percent of cases arise from spontaneous mutation. Approximately 50% of mutations of factor VIII result from inversions of the DNA sequence within intron 22. The incidence of hemophilia A is 1 in 5000 male live births, and that of hemophilia B is 1 in 30,000. Hemophilia B can result from spontaneous mutations, which cause decreased rates of activation of factor IX, altered binding of factor IX to phospholipid membranes, or reduced circulation times. Patients with hemophilia B, Leyden, a single nucleotide substitution in the transcriptional promoter, have severe hemophilia until puberty, at which point, factor IX spontaneously increases to 50%, suggesting that transcription of the factor IX gene is in part hormonally mediated. Factor IX is a smaller molecule than factor VIII and has a greater volume of distribution. The clinical severity of hemophilia is usually dictated by the degree of clotting factor deficiency. Patients with severe hemophilia, less than 1% of normal plasma levels, have an annual average of 20 to 30 bleeds; these are bleeding events that may be spontaneous or marked by excessive bleeding after minor trauma, characteristically into joints or muscle. These patients are usually diagnosed within the first 2 years of life. Bleeding is less common in the newborn period than in later months, but when it occurs, it is most common after circumcision. In a study of newborns with hemophilia, 30% bled from circumcision sites, 27% had intracranial hemorrhage, 16% had persistent bleeding from puncture sites, and 1% had subgaleal hematomas or cephalohematomas (Kulkarni and Lusher, 2001). Infants with severe hemophilia A or B have a 2% to 8% risk of spontaneous intracranial hemorrhage (Rodriguez et al., 2005).

C h a p t e r 36    Systemic Disorders   1151

Hemophilia A and hemophilia B are characterized by a prolonged aPTT with a normal PT. Because the neonate’s aPTT is physiologically prolonged, a factor VIII level must be directly measured to establish the diagnosis. Factor IX levels are physiologically low at birth and do not reach adult levels until 6 months of age, making confirmation of the diagnosis of hemophilia B uncertain until later infancy, except in severe cases. The treatment of hemophilia has changed dramatically over the past three decades from the availability of plasma-derived replacement factors in the 1970s to the engineering of recombinant factors in the 1990s to the recently begun trials of gene replacement therapy (Mannucci and Tuddenham, 2001). Patients with mild and moderate disease, corresponding to 6% to 30% and 1% to 5% of normal factor levels, respectively, usually bleed excessively only after trauma or surgery and are managed with on-demand factor replacement. These patients are often diagnosed later in life.

Preoperative Preparation of the Child with Hemophilia Preparation of the child with hemophilia for surgery depends on the severity of the patient’s disease and the proposed procedure. Patients with mild hemophilia A who have demonstrated an adequate response to DDAVP in the past can undergo minor procedures after IV DDAVP administration (0.3 mcg/kg) 30 minutes before surgery (Mannucci, 1997; Rodriguez and Hoots, 2008).

Factor Replacement Hemophilia A. The type, timing, and dose of factor to be administered should be decided in advance in consultation with the patient’s hematologist. In general, children with hemophilia A who require major procedures should have their factor VIII level maintained close to 100% of normal from 30 minutes before surgery through the first 2 to 7 days of the postoperative period. Factor VIII levels can then be reduced to 30% to 50% of normal for the next 3 to 7 days. Children who undergo minor procedures can be adequately covered with factor VIII levels of 50% after the second postoperative day (Martlew, 2000). In determining factor replacement, it should be remembered that the plasma volume is 45 to 50 mL/kg. Because 1 mL of plasma contains 1 unit of factor VIII, 50 units/kg of factor VIII will increase the patient’s level to 100% (rise of 2% per unit of factor VIII/kg). In the absence of inhibitors, the half-life of factor VIII in vivo is 8 to 12 hours; subsequent doses are timed to maintain the desired level of activity. The first dose has a somewhat shorter half-life than subsequent doses. Therefore, the second dose should be given after a somewhat shorter interval (6 hours). Factor VIII may be administered as cryoprecipitate (0.2 bags/kg should raise factor VIII level to 50%), but the factor VIII level in cryoprecipitate is variable, and plasma levels should be followed if bleeding is not well controlled. The use of heat- or detergent-treated factor concentrates (e.g., Monoclate-P and Hemofil-M) has been replaced with recombinant factor VIII (rFVIII) preparations (e.g., Humate-P). Although the treated factor concentrates had a lower risk of transmission of viruses (e.g., human immunodeficiency virus [HIV] and hepatitis A, B, or C) compared with cryoprecipitate, rFVIII carries no risk of viral transmission; however, it is unknown whether the albumin in the recombinant preparation may have a risk of prion

t­ ransmission. Albumin-free rFVIII concentrate has become available, which may eliminate concerns about infection (Josephson and Abshire, 2004).

Hemophilia B. As in the case of hemophilia A, rFIX has started to replace plasma-derived factor IX concentrate for prophylaxis and treatment of bleeding in children with hemophilia B (Shapiro et al., 2005). In adults, 1 IU/kg of rFIX increased circulating factor IX activity by 0.8 IU/dL. Children have been found to have a smaller increase—0.68 IU/dL per 1 IU/kg infused in children 1 month to 12 years, 0.46 IU/dL in neonates, and 0.93 IU/dL in adolescents, but sample sizes were small. A mean dose of 90 IU/kg (range 29 to 260 IU/kg) was given to 23 patients before surgery (Shapiro et al., 2005). Because of the low yield of circulating factor IX in neonates after rFIX administration, as noted, higher doses are required in this population with regular measurement of serum factor IX levels. A recent report described the administration of rFIX by continuous infusion at a rate of 30 to 35 IU/kg per hour for successful treatment of intracranial and extracranial hemorrhages in the neonate (Guilcher et al., 2005). Because the half-life of factor IX is 12 to 24 hours, it requires less frequent dosing than does factor VIII to maintain adequate levels. As with factor VIII, the second dose should be given at a somewhat shorter interval than subsequent doses (6 to 8 hours). If rFIX is unavailable, the factor IX level is raised 1% for each unit of factor IX concentrate/kg. Hemophilia C. Hemophilia C is factor XI deficiency (Rosenthal syndrome), an autosomally recessive disease that is most commonly reported in Askhenazic Jews. The incidence in the Ashkenazic population is 3 in 1000, compared with a rate of 1 in 1,000,000 in the general population (Gomez and BoltonMaggs, 2008). Incidence of hemophilia C is increased among people with Noonan syndrome (Bertola et al., 2003). Factor IX deficiency presents with a prolongation of the aPTT, with a normal PT. There is an incomplete correlation between the severity of factor deficiency and hemorrhagic symptoms, in that some patients with very low factor levels have no history of bleeding. Bleeding typically occurs after trauma or surgery and is commonly seen in sites that have a high fibrinolytic rate, such as the genitourinary tract, and after circumcision (Andrew, 1997). Because factor XI levels are physiologically low in the neonatal period, the diagnosis is confirmed through levels obtained in later infancy. The perioperative management of these patients is dictated by their bleeding risk. Patients with factor XI levels of more than 15% without a previous history of bleeding, or patients with levels of 5% to 14% who have had previous surgery without significant bleeding in the absence of fresh frozen plasma (FFP) administration can be considered to be at low risk. Patients with factor levels of less than 15%, a history of spontaneous bleeding, bleeding during previous surgeries, or those with a family history of such bleeding complications can be considered to be at high risk. Depending on the surgical procedure, patients who are considered to be at low risk for bleeding can be managed with FFP immediately available. High-risk patients should receive FFP 2 hours before surgery.

Factor Inhibitors in Patients with Hemophilia A and B Factor replacement results in the development of neutralizing antibodies called inhibitors, which complicates the treatment of affected hemophiliacs. The incidence of inhibitors in hemophilia

1152   P a r t  IV    Associated Problems in Pediatric Anesthesia

A is 30%; in hemophilia B, it is 5% (Astermark, 2006). These antibodies are classified as low- or high-titer inhibitors. Patients with low-titer inhibitors may be treated with high doses of factor replacement. Patients with high titers of these inhibitors often develop severe bleeding complications that are recalcitrant to standard factor replacement algorithms and require bypassing agents such as rFVIIa or activated prothrombin complex concentrates (Matthew, 2006; Young, 2006; Rodriguez and Hoots, 2008). For patients with inhibitory antibodies, immunomodulation therapy should be instituted concomitantly with hemostatic therapy. No consensus has been reached on the optimal regimen. IV immune globulin, corticosteroids, and alkylating agents all have been shown to reduce inhibitor levels.

Adjuvant Treatment Options for Hemophilia Patients Antifibrinolytics. These agents (ε-aminocaproic acid [EACA] and tranexamic acid) are very useful adjuvants for patients with mild to severe hemophilia who are experiencing mucosal bleeding, primarily of oral, nasal, and menstrual origins. They are discussed in greater detail below. Recombinant Factor VIIa (rFVIIa). rFVIIa was initially licensed for use in hemophilia patients with inhibitors (Roberts, 2001). rFVIIa binds to TF and to the surface of activated platelets, both of which result in factor X activation and thrombin production. The standard dose is 90 to 120 mcg/kg every 2 to 3 hours until hemostasis is achieved. More than 90% of patients with hemophilia who have a low risk for thrombosis have an effective response to rFVIIa (O’Connell et al., 2002). Anecdotally, a child with hemophilia A and inhibitors was successfully managed intraoperatively with rFVIIa in combination with EACA (Simic and Milojevic, 2007).

von Willebrand Disease Characteristics vWD, the most common congenital bleeding disorder, is a deficiency or dysfunction of the adhesive glycoprotein, which is produced in endothelial cells and megakaryocytes and is stored in Weibel-Palade bodies in endothelial cells and platelets. vWF is fundamental in the binding of platelets to damaged endothelial surfaces, promotes the secretion of factor VIII, and binds, carries, and protects factor VIII in plasma. vWD affects 1 in 100 individuals and is inherited in an autosomal fashion (usually dominant), with males and females affected equally. It is characterized by impaired platelet adhesion to exposed subendothelium in high shear vessels (Federici, 2006). Because of the reduced or defective vWF, factor VIII is reduced to a mild and variable degree because of decreased secretion and enhanced clearance. Because vWD is a disorder of the protein responsible for the adherence of platelets to damaged endothelial surfaces, the clinical manifestations in affected individuals resemble those seen in patients with platelet disorders (i.e., mucocutaneous bleeding [e.g., nose or gingiva], menorrhagia, and increased bleeding with trauma or surgery). vWD has three variants. Patients with the type 1 variant, the most common (80%), have a heterozygous quantitative deficiency of vWF (20% to 40% of normal) associated with diminished factor VIII levels. Patients with type 1 vWD often have menorrhagia or mild to moderate bleeding from mucocutaneous sites. Medications that affect platelet function (e.g., NSAIDS)

can cause hemorrhage in a previously asymptomatic patient with type 1 vWD. Type 2 vWD (17% of patients with vWD) is characterized by the production of qualitatively abnormal vWF. Some of these patients have an associated thrombocytopenia, whereas others have a factor VIII deficiency that is out of proportion to the level of vWF. Type 3 vWD (3%) is marked by profound deficiencies of vWF and factor VIII. Homozygous patients may experience severe bleeding and spontaneous hemarthrosis. Because vWF levels are higher at birth and the proportion of the most functional high-molecular-weight multimeric units is increased, the incidence of bleeding in newborns is very low. When newborns with vWD bleed, it is the result of concomitantly low factor VIII levels. Acquired vWD has been associated with Wilms tumor, systemic lupus erythematosus (SLE), congenital heart defects, and HbE thalassemia. The laboratory diagnosis of vWD is based on a prolonged bleeding time and on decreased levels of vWF antigen and factor VIII in the face of normal PT, fibrinogen, and platelet count. The aPTT may be normal or mildly prolonged. Ristocetin cofactor assay, which measures vWF-induced platelet agglutination, is used to identify type 2 vWD. It is recommended that screening be performed on three separate occasions before ruling out vWD, because functional and antigenic vWF levels may overlap those of normal patients and vWF levels can fluctuate in unpredictable ways. The vWF levels rise during pregnancy. There is a significant linkage between ABO locus and the vWF antigen, such that patients with A and B blood types have markedly higher levels (100% to 115%) of factor than those with type O (75%) (Souto et al., 2000).

Preoperative Preparation Most patients with type 1 disease respond to the IV administration of 0.3 mcg/kg of DDAVP 30 minutes before surgery. DDAVP induces the release of vWF from endothelial storage granules (Weibel-Palade bodies) into the circulation and results in a 2- to 3-fold increase in plasma von Willebrand antigen levels within 30 to 60 minutes, with the effect lasting more than 6 hours (Mannucci, 1997; Robertson et al., 2008). DDAVP may be administered two to three times a day, although tachyphylaxis may develop (Federici, 2008). Because 10% of patients with type 1 vWD fail to respond to DDAVP, the response (increased factor VIII levels and normalization of bleeding time) should be documented before surgery to make sure it is adequate to prevent excessive perioperative bleeding (Nolan et al., 2000). If it is effective, it may be used as the sole agent for the treatment of minor bleeding (e.g., epistaxis) or perioperatively for minor surgery, such as dental extraction. Patients with type 2 or type 3 diseases usually require replacement factor VIII and vWF to control bleeding. DDAVP is contraindicated in type 2b disease, because it may exacerbate thrombocytopenia. Those who do respond to DDAVP may also need factor replacement before major surgical procedures or for major trauma. Plasma-derived human factor VIII concentrate, which has a high concentration of vWF (e.g., Humate-P), is effective and is approved for replacement therapy in vWD (Federici et al., 2002). If the patient’s response to DDAVP is adequate and bleeding is not a problem, liberal DDAVP administration may be substituted for factor concentrate in the postoperative period. Close monitoring of bleeding and communication with the hematologist should guide postoperative management of these patients.

C h a p t e r 36    Systemic Disorders   1153

Factor XIII Deficiency

Thrombocytopenia of Critical Illness

The role of factor XIII in hemostasis is to stabilize newly formed clots by crosslinking fibrin monomers. Plasma levels as low as 1% to 2% are usually adequate for hemostasis. Patients with factor XIII deficiency have bleeding despite a normal PT, aPTT, and platelet count. Factor XIII deficiency is a rare bleeding disorder that is inherited in an autosomalrecessive manner and has an estimated incidence of 1 in 2 million. Typical symptoms are delayed hemorrhages after mild trauma. The most common manifestation is prolonged bleeding from the newborn’s umbilical stump, which is virtually pathognomonic for this deficiency, or after circumcision. The major morbidity in children with factor XIII deficiency is a marked propensity for intracranial hemorrhages (Gordon et al., 2008). Seriously affected patients are treated with cryoprecipitate or purified factor concentrate. All traumatic brain or closed-head injuries are treated prophylactically.

Thrombocytopenia is a common problem in the neonatal and pediatric intensive care units, with a prevalence of 25%. A manifestation of endothelial activation in critically ill children, thrombocytopenia is an independent predictor of mortality and prolonged length of hospital stay (Krishnan et al., 2008).

Factor VII Deficiency Factor VII has the shortest half-life of all of the clotting factors, estimated to be 6 hours. Factor VII deficiency is a rare autosomal-recessive disorder. The severity of the hemorrhagic diathesis does not correlate with factor VII levels. Many individuals have mutations of factor VII but are asymptomatic, and they come to medical attention as a result of an isolated PT prolongation. Much more common than an inherited factor VII deficiency is an acquired factor VII deficiency. Because of the exquisitely short half-life of factor VII, liver failure, vitamin-K deprivation, or oral anticoagulant toxicity first manifests as factor VII deficiency with an isolated increased PT value.

Thrombocytopenia of Immune Origin Immune-mediated thrombocytopenias occur in the setting of isoimmunization, with transplacental transfer of maternal alloantibodies directed against paternally inherited antigens present on fetal platelets or with transfer of maternal autoimmune diseases such as idiopathic thrombocytopenic purpura (ITP) and SLE. Isoimmune thrombocytopenia occurs in 1 of 1000 deliveries. The distinguishing characteristic is the maternal platelet count, which is normal in isoimmune disease and decreased in autoimmune disease. Infants are usually asymptomatic unless the platelet count is less than 10,000. Mucocutaneous, spinal cord, and intracranial hemorrhages are seen prenatally and postnatally in isoimmune disease (Abel et  al., 2003). The bleeding in autoimmune thrombocytopenia is usually not as severe, but the risk of intracranial hemorrhage increases when the platelet count is less than 40,000 in the newborn. Both diseases are treated with platelet transfusions, IVIG, and corticosteroids. The established treatment of alloimmune thrombocytopenia is the administration of washed, irradiated maternal platelets (10 mL/kg), but donor platelets screened for the absence of human platelet antigen 1a (HPA-1a) have been shown to be effective (Rothenberger, 2002; Rayment et al., 2003).

Thrombocytopathies

Platelet Abnormalities Congenital Coagulopathies as a Result of Platelet Abnormalities Inherited coagulopathies also include diseases associated with quantitative and qualitative platelet dysfunction. Wiskott-Aldrich syndrome is an X-linked syndrome characterized by thrombocytopenia, immunodeficiency, and eczema. Newborns typically have thrombocytopenia caused by underproduction. These platelets are abnormally small. The disease results from the mutation of the Wiskott-Aldrich syndrome protein (WASP), a cytoskeletal regulatory protein found in megakaryocytes and lymphocytes (Caron, 2002). Symptomatic bleeding is treated with platelet transfusions. Congenital bone-marrow–failure syndromes that result in congenital thrombocytopenia include Diamond-Blackfan syndrome anemia, Schwachman-Diamond syndrome, and Fanconi anemia. These infants have severe mucocutaneous bleeding or intracranial hemorrhage as a result of profound thrombocytopenia, and they depend on platelet transfusions. Thrombocytopenia with absent radii is another cause of neonatal thrombocytopenia that is associated with skeletal anomalies. The thrombocytopenia is most pronounced in the first year of life, when mucocutaneous bleeding commonly occurs, and platelet transfusions are required.

Inherited qualitative platelet defects are uncommon conditions that also present with bleeding in the newborn period. Glanzmann thrombasthenia is an autosomal-recessive deficiency of αIIbβ3 that impairs fibrinogen binding to platelets. Patients experience mucocutaneous bleeding in the neonatal period and have a lifelong risk of bleeding. Platelet count and morphology are normal; however, bleeding time, clot retraction, and platelet aggregation tests are all abnormal, and flow cytometry is required to confirm the αIIbβ3 deficiency. Bleeding is managed with platelet transfusions. BernardSoulier syndrome is an autosomal-recessive deficiency of the platelet vWF receptor. These patients have mild to moderate bleeding and have unusually large platelets. The diagnosis is confirmed by failure of agglutination in the presence of ristocetin.

Bleeding Diathesis Associated with Blood Vessel Abnormalities Hereditary blood vessel disorders associated with bleeding diathesis include uncommon connective tissue diseases such as Ehlers-Danlos and Marfan syndromes. Osteogenesis imperfecta, a heterogeneous group of disorders of collagen synthesis, may have increased perioperative bleeding as a result of increased capillary fragility and abnormalities in collagen-induced platelet aggregation (Edge et al., 1997).

1154   P a r t  IV    Associated Problems in Pediatric Anesthesia

Acquired Coagulopathies Vitamin K Deficiency Hemorrhagic disease of the newborn is a bleeding disorder that is caused by a deficiency of vitamin K. Clinical bleeding occurs in 1 in 1000 to 1 in 10,000 babies who do not receive vitamin K supplementation. Vitamin K is poorly transferred across the placenta and is present only in low concentration in breast milk. Hemorrhagic disease of the newborn can be temporally divided into three types: early, classic, and late onset. Bleeding within the first 24 hours of life is defined as early disease and is generally seen in infants born to mothers who receive oral anticoagulants or antiepileptic drugs. These infants often have serious bleeding, including intracranial hemorrhage. Bleeding within the first week of life is classic disease and usually involves cutaneous, GI, or circumcision bleeding in infants who did not receive vitamin K supplementation at birth and who are usually breastfed. Bleeding in the first 3 months of life is referred to as late-onset disease, and it is seen in exclusively infants who were breastfed and in those with disorders of fat absorption such as CF, biliary atresia, and celiac disease. The diagnosis is confirmed with a prolonged PT, increased levels of proteins produced in the absence of vitamin K, and a low vitamin K level. Administration of vitamin K subcutaneously or intravenously increases coagulation factors within 2 hours, with complete correction within 24 hours. Serious bleeding may be treated with FFP (10 to 20 mL/kg) or with purified factor IX product.

Hepatic Dysfunction-Related Coagulopathy Liver disease that results in synthetic dysfunction has a major impact on hemostasis because many of the coagulation factors are synthesized in the liver. Levels of these proteins are the first to decline with worsening liver disease, especially the very short-lived factor VII. Measurement of factor V, a hepatically synthesized, non–vitamin K-dependent protein, which is present in similar amounts in the newborn and the adult, is useful in differentiating vitamin K deficiency from hepatic dysfunction. Fibrin-degradation products are increased as a result of their decreased clearance in the setting of hepatic dysfunction. The development of ascites results in further loss of coagulation proteins. Perioperative management of these patients includes determination of their exact deficiencies and correcting them with targeted management. Prolongation of the PT, resulting from depletion of vitamin K-dependent factors, can be treated with vitamin K or FFP. Vitamin K should normalize the PT in as little as 6 to 8 hours. Hypofibrinogenemia should be treated with cryoprecipitate. rFVIIa has been used successfully to correct the coagulopathy associated with liver failure (Atkison et al., 2005). Factor VIIa has been used in small numbers of patients to decrease transfusion requirements during liver transplantation, despite the fact that bleeding in that setting is multifactorial in nature (Busani et al., 2008).

Anticoagulant-Related Coagulopathy Pediatric thrombotic complications are being recognized, diagnosed, and treated in great numbers; neonates are at higher risk for thromboembolic complications than are older ­children

(Kenet and Nowak-Gottl, 2006). Patients who receive ­therapeutic anticoagulation are at risk for devastating hemorrhagic complications. They have a 0.25% to 1% annual risk of intracranial hemorrhage, either intracerebral or subdural, both associated with high morbidity and mortality rates (Leissinger et al., 2008). If surgery is contemplated or when procedural heparinization must be reversed, protamine can be administered IV over 10 minutes. The dosage of protamine is based on the time since the last dose of heparin and can be calculated using the formula shown in Table 36-24 (Monagle, et al., 2008). LMWH anticoagulation is also reversed with protamine, in a dose of 1 mg of protamine per 1 mg (100 units) of LMWH (Monagle et al., 2001). Protamine is given slowly, because rapid administration may cause profound hypotension. It only partially (60%) reverses the anti-Xa effects of LMWH administered within the previous 3 to 4 hours (Monagle et al., 2001). Although fondaparinux and the direct thrombin inhibitors have no specific reversal agents, some experience with using rFVIIa has been reported (Bijsterveld et al., 2002). Children receiving oral anticoagulation may be difficult to maintain in a therapeutic range because of variations in diet, concurrent medications, and underlying disease processes. Breastfed infants are very sensitive to oral anticoagulants because of low concentrations of vitamin K in breast milk. Many of the common medications that are prescribed for children, including prednisone, amoxicillin, trimethoprim-sulfamethoxazole, and ranitidine, increase the INR of children taking oral anticoagulants. In preparation for elective surgery, vitamin K antagonists can be discontinued 4 days before the procedure, with the goal to restore the patient’s INR to the range of 0.8 to 1.2 (Ansell et al., 2008). In this setting, vitamin K supplementation normalizes the INR within 18 to 24 hours after IV administration (Douketis et al., 2008). In situations of acute bleeding or when more rapid reversal is necessary, both IV vitamin K and either FFP or prothrombin complex concentrate should be used. In adults, the INR is corrected at a rate of 0.18 INR/hour after FFP and IV vitamin K administration (Boulis et al., 1999). Prothrombin complex concentrate rapidly normalizes the INR more effectively than does FFP, at a much smaller volume, with a low risk of thrombotic events (Leissinger et al., 2008; Levy et al., 2008). rFVIIa has been used as an alternative agent; it quickly corrects elevated INR in the setting of intracranial hemorrhage, which has been reported to allow for rapid neurosurgical intervention or, in some cases, to actually successfully halt the progression of hemorrhage, obviating the need for a craniotomy (Bartal et al., 2007).

TABLE 36-24. Protamine Reversal of Heparin Therapy Time Since Last Heparin Dose (minutes)

Protamine Dose* (mg/100 units Heparin)

<30

1

30-60

0.5-0.75

60-120

0.375-0.5

>120

0.25-0.375

Data from Monagle P, et al.: Antithrombotic therapy in neonates and children. In American College of Chest Physicians evidence-based clinical practice guidelines, ed 8, Chest 133:887S, 2008. *Maximum protamine dose = 50 mg; Infusion rate should not exceed 5 mg/min.

C h a p t e r 36    Systemic Disorders   1155

Acquired Thrombocytopathy Aspirin and NSAIDs are the most commonly used medications that affect the coagulation system. These medications inhibit platelet COX, blocking thromboxane synthesis and leading to a partial impairment of platelet function. Two COX isoenzymes have been characterized: COX-1, which is always present on platelets and the gastric mucosa, and COX-2, which is up-­regulated during inflammation. Nonspecific COX inhibitors have been shown to increase perioperative bleeding complications after adenoidectomy. Aspirin ingestion prolongs the bleeding time by 2 to 3 minutes. Perioperative administration of NSAIDs (e.g., ketorolac) increased the number of children with bleeding after tonsillectomy (Marret et al., 2003; Dsida and Cote, 2004). Many anesthetic agents have been implicated in platelet dysfunction, as measured by platelet aggregometry; among them are inhaled anesthetics such as sevoflurane and the IV anesthetics propofol and ketamine (Nakagawa et al., 2002). However, no data demonstrate that these agents increase perioperative bleeding or transfusion requirements in the clinical setting. Another class of drugs that may interfere with platelet function is the anticonvulsants such as sodium valproate. Valproate has caused mild thrombocytopenia, neutropenia, and even red cell aplasia, and patients taking valproate should be evaluated before surgery with a complete blood count with platelet count. Bone marrow suppression usually occurs with levels higher than 100 mcg/mL and usually responds to a decrease in dose (Acharya and Bussel, 2000). Bleeding time may be prolonged in patients taking valproic acid but usually not to a clinically significant extent. In a small series of patients taking valproic acid, 20% were shown to have acquired vWD, with low ristocetin cofactor activity, although only 2 of 6 affected children were symptomatic (epistaxis) (Serdaroglu et al., 2002; Koenig et al., 2008). One case report documented profound factor XIII deficiency that resulted in severe intracranial bleeding after craniotomy for epilepsy surgery (Pohlmann-Eden et al., 2003). This deficiency was resolved after discontinuation of valproic acid therapy. In a small study of children with cerebral palsy who underwent femoral osteotomy, those children who took valproic acid had greater blood loss and need for transfusion than did those with cerebral palsy who were either taking no anticonvulsants or anticonvulsants other than valproic acid (Carney and Minter, 2005). Therefore, the anesthesiologist should have a heightened awareness of the possibility of excessive surgical bleeding, especially during and after craniotomy, in patients taking valproic acid. Some hematologists recommend a bleeding time for children who are taking valproic acid and who are scheduled for craniotomy, with further investigation including vWF and ristocetin cofactor, as well as possible preoperative administration of DDAVP if indicated (Acharya and Bussel, 2000; Koenig et al., 2008)

Disseminated Intravascular Coagulation Disseminated intravascular coagulation (DIC) is the unregulated activation of the hemostatic system characterized by the generation of activated clotting factors, fibrin, and accelerated fibrinolysis. Conditions associated with DIC are listed in Box 36-10. Patients can have bleeding or thrombosis or may have only laboratory evidence of DIC. DIC is the result of significant

Box 36-10 Conditions Associated with Disseminated Intravascular Coagulation Sepsis Shock Heat stroke Acidosis Hypoxia* Trauma Head injury Fat embolism Crush injury Burn injury Toxin exposure† Severe allergic reaction Intravascular hemolysis Liver disease Cancer Myeloproliferative disease Vascular anomalies Kasabach-Merritt syndrome Extracorporeal circulation Obstetric complications Amniotic fluid embolism Placental abruption Preeclampsia *Antenatal hypoxia, from Hannam S et al: Neonatal coagulopathy in preterm, small-for-gestational-age infants, Biol Neonate 83:177, 2003. † Snake bite, from Gold BS et al.: Bites of venomous snakes, N Engl J Med 83:177, 2003.

exposure of circulating blood to TF, commonly from endothelial disruption or from hypoxia, acidosis, and sepsis. No single laboratory test can establish or exclude the diagnosis of DIC. The most common laboratory abnormalities include thrombocytopenia or a rapidly falling platelet count and elevated D-dimer levels. Less commonly, microangiopathic hemolytic anemia, hypofibrinogenemia, and PT and aPTT prolongation are seen. D-dimers may be present in premature infants without DIC. Antithrombin III levels can be markedly depressed as well. The treatment of DIC is principally focused on eradicating the precipitating process; treatment of the consequences without treating the underlying cause is certain to fail. No evidence exists that prophylactic administration of platelets or plasma improves the outcome of a patient who is not bleeding. FFP, cryoprecipitate, and platelet transfusions are used only to treat active bleeding in the older child. However, in the neonatal period, because of the risk of intracerebral hemorrhage, many aim to correct platelet counts of less than 50,000, fibrinogen levels of less than 100 mg/dL, and INR of higher than 1.5. Sequential thromboelastograms (TEGs) are useful in monitoring the correction of DIC in the perioperative period (see “Thromboelastograph”).

Nutritional Coagulopathies In addition to vitamin K deficiency, vitamin C deficiency (i.e., scurvy) can result in a bleeding diathesis. Because vitamin C plays a fundamental role in collagen formation, patients with vitamin C deficiency have impaired collagen synthesis and capillary fragility, which leads to gingival bleeding and petechiae. Children with scurvy are at risk for significant ­oropharyngeal bleeding during airway manipulation (Disma et al., 2008).

1156   P a r t  IV    Associated Problems in Pediatric Anesthesia

Acquired Hemophilia Antibodies to coagulation factors develop in patients with hemophilia who are treated with factor replacement and in patients who do not have hemophilia who have no prior exposure to hemostatic therapy. These acquired inhibitors to coagulation factors occur most commonly against factor VIII. Although many patients with acquired hemophilia are elderly, children can be affected in a devastating manner. Patients who develop inhibitors commonly have coexisting diseases such as lupus or rheumatoid arthritis, or they have been recently treated for rheumatic fever or nephrotic syndrome (Sakai et  al., 2005). Transplacental transfer of acquired inhibitors has been reported, resulting in hemorrhagic complications in the newborn period. Patients with acquired inhibitors most commonly have bleeding into fascial planes and mucous membranes, rather than into joints. The diagnosis is elusive because of inconsistent test results. No correlation exists between inhibitor titers and the severity or pattern of bleeding (Huth-Kuhne et al., 2009). In patients with inhibitors, even large amounts of cryoprecipitate or FFP may not promote satisfactory hemostasis. FFP administration leads to an anamnestic response, increasing inhibitor levels. Factor VIII autoantibodies in acquired hemophilia are usually incompletely inhibitory, so that factor VIII levels are usually detectable and may be as high as 10% to 20% of normal values. First-line therapy for inhibitor-associated hemorrhage is either rFVIIa or one of the commercially available activated prothrombin complexes. If one of these agents is not immediately ­available, DDAVP can be employed (Huth-Kuhne et al., 2009).

Recombinant Factor V  IIa (rFVIIa) The success of rFVIIa, a synthetic clotting factor, originally developed for use in patients with hemophilia and inhibitors to

Resting platelets

factors VIII or IX, dramatizes the new paradigm of cell-based hemostasis presented earlier and offers an additional therapeutic option for other bleeding situations in the perioperative period (Roberts et al., 2004). Because of deficiencies in either factors VIII or IX, patients with hemophilia are unable to generate the platelet-localized factor Xa that is necessary for explosive thrombin production. rFVIIa forms a complex with TF exposed at areas of vascular injury, acting as a local catalyst for coagulation. The resultant rFVIIa-TF-Xa complex can overcome a deficiency of factor VIII or IX, the foundation for rFVIIa’s primary indication. rFVIIa also binds to activated platelets and enhances thrombin generation. rFVIIa’s mechanism of action is illustrated in Figure 36-17 (Mannucci and Levi, 2007).

Intraoperative Coagulopathies Patients with no preoperative disorders of coagulation who have surgery may develop coagulopathy because of a combination of blood loss, fluid replacement, and other intraoperative circumstances. Conditions associated with the development of intraoperative coagulopathy are listed in Box 36-11.

Colloid-Induced von Willebrand Syndrome The administration of some synthetic colloids as volume expanders may be associated with the development of acquired vWD (Chappell et al., 2008; Bailey et al., 2010). Large amounts of dextran decrease vWF and factor VIII levels and enhance fibrinolysis. The increase in bleeding times after dextran infusion can be completely normalized by the administration of DDAVP.

Activated platelet

Factor V

Factor X

Factor VIII

Factor Va

Factor Va Tissue factor Factor VIIa

Factor Xa Factor VIIIa

Factor X

Factor II

Factor IX

Factor IXa

Factor VIIIa

Factor IIa (thrombin) Fibrinogen

Factor Xa

Factor II

A

Factor VIIa

Factor IIa (thrombin)

B

Fibrin

n  FIGURE 36-17. rFVIIa’s mechanism of action. When the vessel wall is disrupted, subendothelial TF becomes exposed to circulating blood and may bind factor VIIa (A). This binding activates factor X, and activated factor X (factor Xa) generates small amounts of thrombin. The thrombin (factor IIa) in turn activates platelets and factors V and VIII. Activated platelets bind circulating factor VIIa (B), resulting in further factor Xa generation, as well as activation of factor IX. Activated factor IX (factor IXa) (with its cofactor VIIIa) yields additional factor Xa. The complex of factor Xa and its cofactor Va then converts prothrombin (factor II) into thrombin (factor IIa) in amounts that are sufficient to induce the conversion of fibrinogen to fibrin. (From Mannucci PM, Levi M: Prevention and treatment of major blood loss, N Engl J Med 356:2301, 2007.)

C h a p t e r 36    Systemic Disorders   1157

Box 36-11 Conditions Associated with the Development of Intraoperative Coagulopathy Neurologic conditions Intracranial surgery Traumatic brain injury Cardiovascular conditions Congenital heart disease Shock Kasabach-Merritt syndrome Trauma Orthopedic conditions Fat embolism Scoliosis surgery Osteogenesis imperfecta Intramedullary nailing of long-bone fractures Miscellaneous conditions Citrate-induced hypocalcemia Factor V inhibition from exposure to bovine topical thrombin Data from Iberti TJ, Miller M, Abalos A, et al: Abnormal coagulation profile in brain tumor patients during surgery, Neurosurgery 34:389–394, 1994; Vavilala MS, Dunbar PJ, Rivara FP, et al: Coagulopathy predicts poor outcome following head injury in children less than 16 years of age, J Neurosurg Anesthesiol 13:13–18, 2001; Murshid WR, Gader AG. The coagulopathy in acute head injury: comparison of cerebral versus peripheral measurements of haemostatic activation markers, Br J Neurosurg 16:362– 369, 2002; Hymel KP, Absire TC, Luckey DW, et al: Coagulopathy in pediatric abusive head trauma, Pediatrics 99:371–375, 1997; Robinson CM, Ludlam CA, Ray DC, et al: The coagulative and cardiorespiratory responses to reamed intramedullary nailing of isolated fractures, J Bone Joint Surg Br 83:963–973, 2001; Keegan MT, Whatcott BD, Harrison BA: Osteogenesis imperfecta, perioperative bleeding, and desmopressin, Anesthesiology 97:1011–1013, 2002; Neschis DG, Heyman MR, Cheanvechai V, et al: Coagulopathy as a result of factory V inhibitor after exposure to bovine topical thrombin, JVasc Surg 35:400–402, 2002.

Of the available colloids, children commonly receive human serum albumin and only rarely receive hydroxyethyl starch. Human serum albumin has minimal effects on coagulation (Schramko et al., 2009). However, albumin does prolong the bleeding time based on impairment of platelet aggregation. In adult studies, when albumin is compared with dextran and high-molecular-weight hydroxyethyl starch, albumin is associated with less postoperative blood loss. Preliminary studies comparing middle-molecular-weight hydroxyethyl starch and albumin suggest that a minimal difference exists between them in postoperative blood loss. In patients with even mild forms of vWD, the administration of artificial colloid in patients can be associated with significant hemorrhagic complications; albumin or crystalloid should be used preferentially (De Jonge and Levi, 2001).

Hypothermia Substantial data suggest that hypothermia is an independent and dramatic contributor to coagulopathy. Mild hypothermia of 35° C significantly prolongs the PT, aPTT, and bleeding time. At a core temperature of 34° C, coagulation and platelet function are severely altered, despite normal fibrinolytic function

(Wolberg et al., 2004). The coagulopathy of hypothermia is exacerbated by acidosis (Dirkman et al., 2008). Transfusion of platelets and clotting factors does not correct the hypothermic coagulopathy completely in the absence of rewarming. Both components of the hemostatic mechanism appear to be deleteriously affected by hypothermia. Platelet function is seriously impaired by mild hypothermia as a result of a reduction in TXA2 release. The clotting cascade involves a series of enzymatic reactions, all of which are slowed by hypothermia (Hoffman and Monroe, 2007). The laboratory detection of hypothermic coagulopathy is often missed because most laboratories perform clotting tests at 37° C (Rossaint and Spahn, 2006). The results of the PT and aPTT performed at the patient’s actual core temperature are prolonged. TEG data suggest that hypothermia impairs clot formation rather than enhancing fibrinolysis (Dirkman et  al., 2008). Thromboelastography can be adjusted to a patient’s core body temperature to adequately evaluate the role played by hypothermia in hypothermic coagulopathy (Ramaker et al., 2009).

Hemodilution Acute normovolemic hemodilution to minimize red cell transfusion requirements often results in alterations of hemostasis, especially when colloid is used as the diluent. Quantitative modeling of acute normovolemic hemodilution demonstrates that patients often develop inadequate fibrinogen levels (less than 100 mg/dL) before they reach the hematocrit threshold for red cell transfusion or the threshold for platelet transfusions (Singbartl et al., 2003). Fibrinogen administration has been shown to reverse the hemostatic consequences of colloid hemodilution (Mittermayr et al., 2007).

Massive Transfusion Massive blood loss is defined as the loss of more than 1 blood volume in a 24-hour period, the normal blood volume being 7% of ideal body weight in an adult and 8% to 9% in an infant. In the operating room, early recognition of major blood loss can be appreciated using the definitions of massive blood loss as occurring at the rate of 2 to 3 mL/kg per minute or 50% of blood volume in over 3 hours. The progression from dilutional coagulopathy to dilutional thrombocytopenia is seen in massive transfusion and in extreme hemodilution. During the era of whole-blood administration, thrombocytopenia was the initial coagulopathic event. Currently, blood loss is most often replaced with plasma-poor RBC products that are devoid of most coagulation factors. Under these conditions, the initial coagulopathic event is the dilution of coagulation factors (marked by prolongation of the PT); hypofibrinogenemia is the first factor deficiency that occurs, at a time when the platelet count remains more than 150,000/mcL (Perkins et al., 2008). The PT becomes prolonged when less than 1 blood volume is lost, but a clinical coagulopathy does not occur until the PT and PTT exceed 1.5 to 1.8 times the control values (Hirshberg et al., 2003). In the clinical setting, fibrinogen concentrations fall below the hemostatically critical level of 100 mg/dL when blood loss is in excess of 150% of the patient’s blood volume, and the remaining coagulation factors fall below 25% after 200% blood loss. A platelet count of less than 50,000 should be anticipated when more than 2 blood volumes have been lost (Stainsby et al., 2000).

1158   P a r t  IV    Associated Problems in Pediatric Anesthesia

FFP is the first choice in treating a coagulopathy that results from massive transfusion (Gonzalez et al., 2007). Although the optimal FFP/RBC ratio is unknown, aggressive adherence to a ratio of 1:1 to 1:2 helps prevent the onset of a dilutional coagulopathy and minimize the need for cryoprecipitate (Hirshberg et al., 2003; Ho et al., 2005). These authors have also suggested an ideal platelet-to-RBC ratio of 0.8:1 for the massively transfused patient (Hirshberg et al., 2003). Survival in both civilian and combat victims requiring massive transfusions was dependent on the FFP/RBC ratio that they received; mortality was lowest when this ratio was approximately 1:1.5 (Borgman et al., 2007; Sperry et al., 2008). In this setting, if rapid laboratory corroboration is possible, a complete coagulation profile should be obtained to determine the patient’s specific replacement needs. In the absence of timely data, or with continuing hemorrhage, empirical therapy with FFP is justified. FFP at a dose of 20 mL/kg increases fibrinogen by approximately 60 mg/ dL and increases clotting factors by 20%.

Traumatic Coagulopathy Independent of Blood Loss or Replacement Laboratory manifestations of trauma-related coagulopathy include abnormal PT and aPTT, decreased protein C levels, and an increase in plasma thrombomodulin. These patients have increased transfusion requirements. Abnormal aPTT and low protein C are independent predictors of mortality (Macleod et al., 2003; Brohi et al., 2007a). Traumatic coagulopathy manifests as a hypocoagulable state associated with hypothermia, acidosis, clotting factor dilution, and tissue destruction, all of which are directly proportional to the severity of the traumatic injury (Hess et al., 2008). Nearly 25% of trauma patients arrive at the emergency department with an established clinically significant coagulopathy, and these patients are four times more likely to die than are trauma victims without a coagulopathy (Brohi et al., 2007b). The variables responsible for this coagulopathy include tissue trauma, shock, dilution, hypothermia, acidemia, and inflammation (Armand and Hess, 2003). Mild to moderate hypothermia affects coagulation and platelet function and is associated with clinical bleeding (Martini et al., 2005). Animal investigation confirms that acidemia slows the kinetics of coagulation reactions and enhances fibrinolysis (Martini et al., 2007). Shock and the hemodilution that result from fluid resuscitation are the early instigators of the traumatic coagulopathy process. The aggressive fluid management required by these trauma patients produces a concomitant dilutional coagulopathy as well. Vigorous fluid resuscitation can result in a confluence of circumstances that in and of itself contributes to the traumatic coagulopathy. By increasing intravascular volume, blood pressure, and vasodilation, hemorrhage is further potentiated. These hemodynamic and rheologic perturbations increase the likelihood that hemostatic plug formation will not occur. Specific injuries modulate the course of the coagulopathy. Extensive tissue destruction releases tissue thromboplastins, which activate the clotting process and result in a consumptive coagulopathy. tPA and PAI-1 are released from the extensively damaged tissue bed. In the first few hours after traumatic injury, tPA increases out of proportion to PAI-1 and produces a systemic activation of fibrinolysis (Brohi et al., 2008). This coagulopathy resembles that of DIC, but the diffuse microthrombi that are classically seen in DIC are not seen in traumatic coagulopathy.

Traumatic brain injury produces a consumptive coagulopathy as the result of the release of TF and other tissue thromboplastins into the circulation and fibrinolysis activation, which is associated with a dramatic increase in mortality (Stein and Smith, 2004; Talving et al., 2009). Long-bone fractures are also associated with coagulopathy, perhaps through the mechanism of subclincial fat embolism.

Intraoperative Evaluation of the Bleeding Patient Activated clotting time (ACT) is a modification of a whole-blood clotting test that uses kaolin or Celite to accelerate coagulation by activating the contact pathway (Box 36-12). A fixed volume of blood is placed into a tube with activator at 37° C for 60 seconds, after which the contents are stirred until a clot is formed. The normal ACT range is 80 to 120 seconds. This test can be performed easily at the bedside using commercially available equipment. ACT levels correlate well with antifactor Xa heparin levels in the precardiopulmonary bypass period and are commonly used to monitor the adequacy of heparin anticoagulation

Box 36-12 Evaluation of the Bleeding Child Platelet Assessments Abnormal number: thrombocytopenias, hemangiomas Abnormal morphology: inherited platelet defects Abnormal function: inherited or acquired PT and aPTT Assessments Abnormal PT, normal aPTT Factor VII deficiency Vitamin K deficiency, liver disease Factor deficiencies: II, V, VII, and X Drug related: Coumadin Normal PT, abnormal aPTT Factor deficiencies: VIII, IX, XI, XII, kallikrein, prekallikrein, high-molecular-weight kininogen, and vWF Drug related: heparin Abnormal PT, abnormal aPTT Vitamin K deficiency, liver disease Factor deficiencies: II, V, and X Dysfibrinogenemia Drug related: heparin and coumadin Abnormal PT, abnormal aPTT, thrombocytopenia l DIC l Liver disease l Dysfibrinogenemia Normal PT, normal aPTT, normal platelets Factor XIII deficiency α2-antiplasmin deficiency Thrombin time* Normal Liver disease, vitamin K deficiency Factor deficiencies: II, V, X Drug related: Coumadin Abnormal Liver disease DIC Dysfibrinogenemia Drug related: heparin *Useful in cases of abnormal PT and abnormal aPTT.

C h a p t e r 36    Systemic Disorders   1159

levels during extracorporeal circulation. The relationship of ACT with heparin dosing is linear in the setting of normal antithrombin III concentrations and factor XII activity, ­normothermia, a platelet count of more than 50,000, intact platelet function, and a fibrinogen level of more than 100 mg/mL. However, in the absence of extracorporeal circulation, ACT has a much poorer correlation with plasma heparin concentrations than does the aPTT. ACT is insensitive to many coagulation abnormalities; clotting deficiencies and platelet abnormalities can be present with a normal ACT (Girardi et al., 2000). Nevertheless, in the setting of major trauma, intraoperative ACT measurements were able to discriminate between patients who became coagulopathic and those who did not. However, because the ACT test is performed at 37° C, the contribution of hypothermia to the development of this coagulopathy is underappreciated (Aucar et al., 2003). ACT levels do not correlate well with LMWH antiXa levels.

Thromboelastograph The TEG is a point-of-care evaluation of a patient’s hemostatic balance, from initial clot formation to clot retraction or dissolution. Coagulation of blood has been compared with the building of a house: TEG profiling does not end when the foundation stone is laid, as do the other clinically employed clotting studies. The TEG also reflects the speed of the building process, whether the building will be sturdy, and whether it is likely to be damaged soon after it is built. The TEG examines the viscoelastic properties of blood as it clots. Placed into a rotating cup with an immersed pin, liquid blood begins to clot by forming fibers between the cup and the pin, transmitting motion to the pin. The TEG measures the elastic shear modulus of the clot, providing information about the rate of clot formation, clot strength, platelet function, and fibrinolytic activity, reflected in the characteristics of the tracing produced (Venkatesh et al., 2001). The maximal elastic sheer modulus depends on platelet count and function and the amount of fibrin deposited on the pin. A typical TEG tracing and the parameters of the TEG are illustrated in Figure 36-18 and described in the following list: R (reaction time): Latency to initial clot formation; from onset of tracing until a 2-mm amplitude on tracing. This

Coagulation

Clotting time (R)

is similar to whole-blood clotting time; it depends on an intact intrinsic pathway and an adequate generation of thrombin. Prolonged R is associated with clotting deficiencies, heparin administration, and thrombocytopenia. The coagulation test that correlates with R is aPTT. K (coagulation rate): Rate of fibrin buildup and crosslinking, occurring from 2 to 20 mm. Prolonged K is associated with clotting deficiencies, platelet dysfunction, thrombocytopenia, and hypofibrinogenemia. a angle (rate of increase in elastic shear modulus): The rate of fibrin buildup and crosslinking; slope of divergence of tracing from R. MA (maximal elastic shear modulus): Measured at maximal divergence of the graph, after the clot is entirely formed, this is the maximal clot strength. A typical clot has an MA of 50 mm, which is equivalent to 5000 dynes/cm2. MA is the best description of the competency of the clot. It depends on platelet count and function and on the fibrinogen level (Chandler, 1995): Decreased MA is associated with thrombocytopenia, platelet dysfunction hypofibrinogenemia, and deficiencies of factors VIII and XIII. Increased MA is associated with a prothrombotic state. The tracings seen on the TEG for common disorders of hemostasis are shown in Figure 36-19. The TEG is a global measure of hemostasis, useful when multiple hemostatic defects are present. The weakness of TEG is its inability to identify specific clotting abnormalities. The TEG is a sensitive indicator of hypocoagulable and hypercoagulable perturbations. The typical hypercoagulable TEG profile has the appearance of a cognac glass (Traverso et al., 1995). The R of patients receiving coumadin increases with increased INR, but it may remain within a normal range. Patients with decreased clotting factors have decreased R and K values and prolonged PT, or they have decreased angle and prolonged aPTT. Hypofibrinogenemia is associated with decreased R and K times, decreased angle, and decreased MA. Hyperfibrinolysis is characterized by reduced amplitude at 30 and 60 minutes (decreased A30, A60). In addition to its point-of-care availability, the conditions under which a TEG is performed can be optimized to best define the clinical scenario. TEG offers a number of other advantages

Fibrinolysis

Clot kinetics (K, α)

Clot strength (MA)

Lysis time

n  FIGURE 36-18. Parameters of the thromboelastograph.

1

2

3

4

5

n  FIGURE 36-19. Thromboelastographic tracings of common hemostatic abnormalities. 1, Normal; 2, hemophilia; 3, thrombocytopenia; 4, fibrinolysis; 5, hypercoagulation.

1160   P a r t  IV    Associated Problems in Pediatric Anesthesia

in the assessment of hemostasis that conventional coagulation tests do not. It is performed at the patient’s temperature to reflect its effects on hemostasis; conventional parameters are measured at 37° C. The anticoagulation effects of LMWH can be rapidly assessed with TEG tests, whereas factor Xa measurements are not readily available (Coppell et al., 2006). Hypercoagulability can be diagnosed by TEG but usually cannot be determined through conventional means (Ganter and Hofer, 2008). The patient’s state of coagulability can be quantified with the coagulation index (CI) (Chan et al., 2007): CI = (–0.6516)R – (0.3772)K + (0.1224 )MA + (0.0759)α – 7.7922 CI > 3, hypercoagulable stage CI < 3, hypocoagulable stage TEG has been employed in the pediatric population. The parameters of children younger than 1 year of age undergoing noncardiac surgery were different from those of an adult control group. The rates of clot initiation and clot buildup, as well as clot strength were all greater in the infant groups. Despite their observations of developmental differences in clotting factor concentrations, the authors concluded that the hemostatic mechanism of the infant is balanced (Chan et al., 2007). TEG has been used in a variety of clinical pediatric perioperative settings, including cardiac surgery, extracorporeal membrane oxygenation, liver transplantation, and neurosurgery (Miller et al., 2000; Goobie et al., 2001; Davis et al., 2006;). Because TEG has been a registered trademark since 1996 and because of the availability of other similar devices (rotation thromboelastometry [ROTEM] and Sonoclot), it has been suggested that the term viscoelastic point-of-care coagulation analysis should replace TEG as the generic term for this methodology. As of this writing, these newer devices have not been extensively used in children.

been developed, including screening potential blood, plasma, and platelet donors and leukoreducing whole-blood and red cell products (Ludlam and Turner, 2006).

Safety of Hemostatic Agents Recently, the medications that have been most extensively evaluated as alternative or adjunctive hemostatic agents include the antifibrinolytic lysine analogues, EACA and tranexamic acid (a bovine-derived protease inhibitor), DDAVP, and rFVIIa. Designed to assess therapeutic efficacy, most trials were not optimally designed to assess potential adverse effects; therefore, definitive safety data are lacking for all agents. Although these drugs are often associated with dramatic success—“the liquid intracranial hemorrhage became a clot pancake before my eyes”—they are not panaceas. Adverse events should be anticipated when tightly regulated systems such as the hemostatic and fibrinolytic systems are pharmacologically perturbed. Of particular concern are the thrombotic adverse events associated with these hemostatic agents (Mannucci and Levi, 2007). Most safety data regarding the complications of these hemostatic agents are based on results of studies in adult patients. Is it reasonable for pediatric providers to extrapolate safety concerns from the adult experience? Until sufficiently large pediatric studies are conducted to evaluate the safety of these agents, they must certainly pay close attention to the concerns raised by the adult studies. Given that these drugs have life-saving potential, the dilemma, as Twite and Hammer (2008) point out, is that in the absence of data, providers must balance “do what is best” with “do no harm.” Informed consent, when deemed necessary, and heightened vigilance for the occurrence of adverse events must accompany the use of such agents in children.

Anticoagulant-Induced Coagulopathy TREATMENT OF THE BLEEDING PATIENT

Safety of Transfusion and Factor Replacement Complications of life-threatening bloodborne virus transmission have been markedly reduced by the institution of multiple screening steps in the procurement of donor-derived plasma products. However, the risks of transfusion-associated transmission of thermostable viruses such as hepatitis A and parvovirus B19 remain. Reports of transfusion-associated transmission of West Nile virus highlight the continued risks of using the blood supply as the principal source for hemostatic agents. Nucleic acid amplification testing for West Nile virus has now been shown to be effective in detecting virus in asymptomatic donors (Busch et al., 2005). The list of potential bloodborne pathogens is long and continues to grow, although not all bloodborne viruses have been demonstrated to be pathogenic to humans. Unfortunately, these concerns are not completely alleviated by the use of recombinant factor concentrates. With the outbreak of variant Creutzfeldt-Jakob disease in the United Kingdom, concern has been raised that the human albumin used in the manufacture and formulation of some recombinant factors may contain and transmit prion proteins. In countries where bovine spongiform encephalopathy has occurred, strategies to minimize the risk of transmission of Creutzfeldt-Jakob disease have

Patients who are experiencing severe bleeding should receive FFP in a dose of 20 mL/kg until vitamin K administration increases endogenous synthesis. Infusion of factor IX complex concentrate, which contains high concentrations of the activated vitamin K–dependent factors (II, VII, IX, and X) can correct the effects of anticoagulants rapidly without the excessive volume exposure of FFP. Factor IX complex (40 IU/kg) in addition to FFP infusion (as compared with FFP alone) successfully corrected the INR in patients with intracranial hemorrhage in one third of the time with 9% of total fluid volume administered (Boulis et al., 1999). rFVIIa can also correct oral anticoagulant-induced coagulopathy.

Agents Used to Control Bleeding DDAVP DDAVP is an analogue of vasopressin that recruits factor VIII from storage sites within endothelial cells via activation of the endothelial vasopressin V2 receptor and cAMP-mediated signaling; it may raise baseline factor VIII levels by 2- to 20-fold in normal patients and in patients with mild hemophilia (Mannucci, 1997; Kaufmann and Vischer, 2003). This increase in factor level is often sufficient to prevent or limit minor ­bleeding.

C h a p t e r 36    Systemic Disorders   1161

DDAVP also increases the release of ultra large vWF multimers from the endothelium. DDAVP is the treatment of choice for children with type 1 vWD, mild hemophilia, or platelet dysfunction, including uremia and drug-induced bleeding diatheses. DDAVP is also effective in treating hemorrhage in patients with acquired inhibitors to factors VIII or IX. Anecdotally, DDAVP has controlled perioperative bleeding in patients with Ehlers-Danlos syndrome, Marfan syndrome, and osteogenesis imperfecta, in which bleeding is thought to be the result of dysfunctional platelet aggregation caused by abnormal collagen (Keegan et al., 2002; Franchini, 2007). DDAVP has been employed in a variety of pediatric procedures associated with risk of large blood loss in an attempt to decrease perioperative blood loss and transfusion requirements. Despite early studies to suggest that DDAVP reduced blood loss in cardiac surgery patients, most studies suggest that DDAVP is not efficacious in uncomplicated cardiac surgery. Children undergoing complex congenital heart repair did not have any reduction in bleeding or transfusion requirements with the prophylactic use of DDAVP (Oliver et al., 2000). Similarly, DDAVP does not appear to reduce blood loss in spinal fusion in patients with idiopathic scoliosis or in those with cerebral palsy-­associated neuromuscular scoliosis (Theroux et al., 1997; Alanay et al., 1999). A dose of 0.3 mcg/kg administered IV over 20 to 30 minutes or intranasally (150 mcg for children weighing less than 50 kg, and 300 mcg for those weighing more than 50 kg) increases factor VIII levels by 62%, and this dose may be repeated every 8 to 12 hours to control bleeding. Peak effects occur within 30 to

Plasminogen

60 minutes after IV infusion and 60 to 90 minutes after intranasal administration. Tachyphylaxis may occur after three or four doses. Because severe hyponatremia-associated seizures have been reported with the use of DDAVP, close observation of fluid status and electrolyte balance is mandated. Although thrombotic complications have been reported, in two meta-analyses of DDAVP use in the adult perioperative period, no significant difference in the incidence of these complications between DDAVP and placebo was reported (Mannucci, 1997; Crescenzi et al., 2008).

Antifibrinolytics Two classes of antifibrinolytics have been employed to optimize hemostasis in the setting of a bleeding diathesis and to reduce blood loss and transfusion requirements during major procedures: the synthetic lysine analogues EACA and tranexamic acid, and the serine protease inhibitor aprotinin. Their mechanism of action is illustrated in Figure 36-20. Antifibrinolytic drugs are commonly employed when bleeding occurs in sites that are rich in plasminogen activator and other fibrinolytic enzymes, such as the endometrium, the GI tract, and the urinary tract (Mannucci, 1998). These drugs are contraindicated in the setting of hematuria, DIC, or thromboembolic disease. EACA has been successfully used in managing bleeding crises in patients with hemophilia and inhibitors (Ghosh et al., 2004). EACA binds to the lysine site on plasminogen and plasmin, preventing plasmin from binding to fibrin; fibrinolysis is inhibited and clot stabilization continues. EACA has been shown to be

Plasminogen activator

Plasminogen

Lysine-binding site

Plasminogen activator

Lysine analogue

Fibrin Lysine analogue

A

Fibrin-degradation products

B

n  FIGURE 36-20. Antifibrinolytics mechanism of action. Activation of plasminogen by endogenous plasminogen activators results in plasmin, which causes degradation of fibrin. Binding of plasminogen to fibrin makes this process more efficient and occurs through lysine residues in fibrin that bind to lysine-binding sites on plasminogen (A). In the presence of lysine analogues, these lysine-binding sites are occupied, resulting in an inhibition of fibrin binding to plasminogen and impairment of endogenous fibrinolysis (B). (From Mannucci PM, Levi M: Prevention and treatment of major blood loss, N Engl J Med 356:2301, 2007.)

1162   P a r t  IV    Associated Problems in Pediatric Anesthesia

­effective in decreasing blood loss in children undergoing cardiac surgery (Eaton, 2008). Children undergoing posterior lumbar fusion who received 100 mg/kg of EACA followed by an infusion of 10 mg/kg per hour had less blood loss and required less RBC transfusions than did a randomized control group (Florentino-Pineda et al., 2001). Aprotinin is a naturally occurring serine protease inhibitor that affects hemostasis through several mechanisms; it is an antifibrinolytic and inhibits kallikrein, plasmin, trypsin, and APC. Aprotinin inhibits the initiation of fibrinolysis and the contact phase of coagulation, but it has no effects on platelet function. It is inactive when given orally. Aprotinin was first shown to be effective in reducing blood loss in coronary-artery–bypass graft operations and subsequently was shown to be effective in reducing perioperative bleeding and transfusion requirements in pediatric cardiac surgery, liver transplantation, hip replacements, and posterior spine fusion (Lemmer et al., 1996; Urban et al., 2001; Samama et al., 2002; Eaton, 2008). Mouth bleeding is common in hemophilia patients, resulting in part from the potent fibrinolytic activity of saliva. For oral or GI bleeding in patients with hemophilia, EACA is often very effective and may dramatically reduce the need for additional coagulation factor infusions when given at a dose of 100 mg/kg orally every 6 hours for 5 to 10 days after a hemorrhagic episode. Antifibrinolytic mouthwashes allow for the performance of dental extractions on patients receiving long-term oral anticoagulant treatment without lowering the degree of anticoagulation (Patatanian and Fugate, 2006).

Safety of Antifibrinolytic Agents Clinical trials demonstrate that all three drugs (EACA, aprotinin, and tranexamic acid) are effective at reducing transfusion requirements as compared with placebo (Henry et al., 2007). Recent studies of aprotinin’s use in adult patients have documented an increased incidence of renal failure and cardiovascular complications; a study called Blood Conservation Using Antifibrinolytics in a Randomized Trial (BART) reported an increase in perioperative mortality (Mangano et al., 2006; Fergusson et al., 2008). Safety data on pediatric aprotinin use are limited. The largest pediatric review of aprotinin usage (865 children) was designed to evaluate the incidence and impact of aprotinin hypersensitivity reactions; this study was not designed to identify thrombotic events or renal impairment, and none was reported (Jaquiss et  al., 2002). In a retrospective review of aprotinin’s use in pediatric cardiac surgery, using historical controls, no differences in the complication rates were reported; however, thrombotic events were not one of the primary end points (Backer et al., 2007). Recent retrospective studies found no increase in renal dysfunction in neonates who received aprotinin during cardiopulmonary bypass (Guzzetta et  al., 2009a; Manrique et al., 2009). At the current time, the issue of whether aprotinin is safe and effective in children is moot, as the manufacturer has removed the drug from the market because of concerns raised by the BART trial.

Recombinant Activated Factor VIIa (rFVIIa) rFVIIa is a synthetic clotting factor that was originally developed for use in patients with hemophilia who developed inhibitors to factors VIII or IX. The rationale behind the development and clinical success of rFVIIa underscores the new paradigm

of cell-based hemostasis presented earlier. Those with hemophilia are unable to generate the platelet-localized factor Xa necessary for explosive thrombin production because of deficiencies in factors VIII or IX. Administered factor VIIa is effective in improving hemostasis in patients who have hemophilia by complexing with TF exposed at areas of vascular injury and by binding to activated platelets and restoring platelet surfacethrombin generation. This augmented thrombin production further increases platelet activation and thrombin-activatable fibrinolysis inhibitor (TAFI), which decreases fibrinolysis. In this way, the resultant factor VIIa-TF-Xa complex thus overcomes a deficiency of factor VIII or IX. For patients with hemophilia who have high inhibitor titers, randomized, controlled trials in adults and children have shown that rFVIIa minimizes spontaneous bleeding and decreases intraoperative blood loss (O’Connell et al., 2002). After rFVIIa’s dramatic use in an exsanguinating 19-year-old soldier with a vascular injury, an increasing number of anecdotal and small case series have described rFVIIa’s successful use in mitigating uncontrollable bleeding in patients without hemophilia in a variety of clinical situations—hepatic dysfunction, thrombocytopenia, platelet dysfunction, dilutional coagulop­ athies, DIC, and after cardiopulmonary bypass (Kenet et al., 1999; Poon et al., 1999; Martinowitz et al., 2001; Blajchman, 2003; Goodnough, 2003; Tobias et al., 2003a, 2003b; Guzzetta et al., 2009b). Because factor VII is the first of the clotting factors to become deficient in liver failure, rFVIIa has been used in the setting of hepatic dysfunction and vitamin K antagonism. The enhanced platelet activation that results from increased thrombin production has led to the use of rFVIIa in treating quantitative and qualitative platelet dysfunction (Patel et al., 2001). rFVIIa resulted in a reduction in bleeding time in 50% of patients with thrombocytopenia. The clot formed with rFVIIa use has a denser mesh of fibrin fibers that is more resistant to plasmin degradation (Hedner, 1998). rFVIIa has been used in the patient receiving massive transfusions whose coagulopathy has been unresponsive to conventional plasma-component therapy. The recommended dose of rFVIIa is 90 mcg/kg administered over 2 to 5 minutes. Because of its short half-life, additional doses must be given every 2 to 4 hours until hemostasis is achieved. The dose for patients with factor VII-deficiency is 25 mcg/kg. rFVIIa has many advantages. It is completely synthetic, decreasing the risks of infectious complications; it can be quickly reconstituted from powder, eliminating the time needed for thawing and procurement of products from the blood bank; and it is dissolved in a small volume, minimizing excessive volume load, electrolyte perturbations, and the latency between institution of treatment and correction of the hemostatic defect. rFVIIa has limitations as well. Its efficacy depends on the presence of the other clotting factors. In the setting of a massive transfusion, factor VIIa is not the only clotting factor that is deficient. Most anecdotal reports concerning the use of rFVIIa in massive transfusion-associated bleeding are in the context of earlier administration of FFP. rFVIIa has a short half-life, necessitating frequent dosing. In the context of liver failure or dilutional coagulopathy, dosing is required every 12 hours. In the presence of inhibitors to factors VIII or IX, doses must be given every 2 to 4 hours until hemostasis is maintained. The risk of rFVIIa’s thrombotic complications in hemophiliacs appears to be low. In contrast, most of the rFVIIa’s thrombotic complications are in the context of its “off-label usage” and

C h a p t e r 36    Systemic Disorders   1163

are associated with serious morbidity and mortality (O’Connell et al., 2006; Abshire, 2008; Abshire and Kenet, 2008). Although potentially promising, at the present time the role of rFVIIa in the treatment of perioperative bleeding should be restricted to that of rescue therapy for the intractably bleeding patient unresponsive to conventional transfusion treatment. Specifically, rFVIIa should be used cautiously in patients at risk for catastrophic thrombotic complications (e.g., as children with small-vessel anastomoses or palliative cardiac shunts) until further experience is garnered. After randomized, controlled trials have been completed to better define the thrombotic and pulmonary risks, rFVIIa could become a first-line agent for many of the coagulopathies that plague patients in the perioperative period (Alten et al., 2009).

Hemostatic Formulary The blood-based components available for treatment of the bleeding patient, their constituents, and indications for their use are summarized in Table 36-25. The American Society of Anesthesiologists Task Force on Blood Component Therapy (2006) recommended that the triggers for treatment of the patient who is at risk for bleeding be multiple and not depend on a single factor. Often, the critically ill child may require an emergent invasive procedure before a complete hemostatic profile has been determined. In these circumstances, bleeding from puncture sites or general oozing in the surgical field may necessitate empirical treatment with FFP. Generally accepted treatment triggers for component therapy are listed in Box 36-13.

Platelets Each unit of platelets contains 5.5 to 7.5 × 1010 platelets diluted in 50 mL of plasma. An apheresis platelet unit contains more than 3 × 1011 platelets in 250 to 300 mL of plasma. Normally, one third of all transfused platelets undergo splenic ­sequestration.

Box 36-13 Recommendations for Blood-Component Therapy Platelets Less than 50 k for acute bleeding Less than 100 k for intracranial, subarachnoid, or extracorporeal circulation procedures FFP aPTT more than 1.5 times normal PT more than 1.5 times normal Fibrinogen Less than 100 mg/dL

In nonsurgical patients with platelet counts of more than 20,000/mcL, spontaneous bleeding is uncommon. ASA guidelines suggest that patients receive transfusions for platelet counts less than 50,000/mcL and that platelet transfusions be considered for platelet counts between 50,000 and 100,000/ mcL, taking into consideration the risks and consequences of postoperative bleeding from the surgical site. A platelet count of 50,000/mcL is considered sufficient for spinal anesthesia, 80,000/mcL for epidural anesthesia, and 100,000/mcL is recommended for neurosurgical or posterior ophthalmologic procedures (Rochon and Shore-Lesserson, 2006). Because of the risk of intracranial hemorrhage, sick premature infants are transfused when their platelet counts fall below 50,000/mcL (Roseff et al., 2006). If concurrent platelet dysfunction exists, the threshold for platelet transfusion should be lowered. The ideal platelet dose is between 0.07 and 0.15 × 1011 platelets/kg, which is approximately 10 mL/kg. A dose of 5 to 10 mL/kg or 1 unit/10 kg increases the platelet count by 50,000/mcL. Platelets are much more likely than RBCs to cause bacterial sepsis, because they are stored at room temperature for up to 5 days and potentially have a higher bacterial load. The reported incidence of bacterial contamination of ­platelet ­products is 1  case per 2000 units (Blajchman et al., 2005). Storing the

TABLE 36-25. Hemostatic Formulary Component

Contents

Indications

Dose

Outcome

Whole blood

Hematocrit: 30%-40% Most clotting factors ↓ FV, FVIII No platelets

Neonatal surgery Massive transfusion

Fresh frozen plasma, 225 mL

All clotting factors 2 mg/mL fibrinogen

Hemodilution Liver failure DIC Warfarin toxicity FXI deficiency

10-15 mL/kg (max, 2 units) 5-8 mL/kg

↑15% in factors ↑40 mg/dL fibrinogen

Platelets, 50 mL

>5.5 × 1010 50 mL plasma

Thrombocytopenia Platelet dysfunction

1 unit/10 kg (max, 10 units)

↑50,000

Cryoprecipitate, 25 mL*

Fibrinogen >150 mg Factor VIII >80 units von Willebrand factor >80 units FXIII >80 units

↓ fibrinogen Hemophilia A vWD

1-2 units/10 kg (max, 10 units)

↑ 60-100 mg/dL fibrinogen

*Cryo units = [(desired fibrinogen - initial fibrinogen) × plasma volume (dL)]/150 mg fibrinogen/unit. DIC, Disseminated intravascular coagulation; vWD, von Willebrand disease.

1164   P a r t  IV    Associated Problems in Pediatric Anesthesia

platelets in galactose-containing solution has been found to preserve platelet function despite chilling, and this approach may reduce the risk of bacterial contamination (Jhang and Spitalnik, 2005; Hornsey et al., 2008). Platelet transfusions can also be associated with the development of pulmonary microvascular injury, called transfusionrelated acute lung injury (TRALI). TRALI is clinically similar to acute respiratory distress syndrome (discussed in a later section). Within 6 hours of receiving a plasma-containing product, fever, tachypnea, dyspnea, progressive hypoxemia, radiographic evidence of pulmonary edema, and hypotension occur. The prevalence of TRALI in patients who have received platelet transfusions is estimated to be 3 per 1000 units of concentrate (Silliman, 1999).

Fresh Frozen Plasma FFP contains 250 mL of plasma and 500 mg of fibrinogen in a citrate anticoagulant. One unit of FFP has a concentration of coagulation factors similar to that of 4 to 5 units of platelet concentrates, 1 apheresis unit of platelets, and 1 unit of fresh whole blood. FFP, 1 mL/kg, raises most factor levels by approximately 1%. After a dose of 10 to 15 mL/kg of FFP, plasma clotting factors rise approximately 15%, and the fibrinogen rises by 40 mg/dL. However, FFP contains only 0.6% of factor VIII. FFP use is indicated for treatment of microvascular bleeding in patients who have had massive transfusions, for documented coagulopathy (PT more than 1.5 times normal) in patients who have had massive transfusions, for urgent reversal of anticoagulant therapy, and for active bleeding in patients with history or course suggestive of an inherited or acquired coagulopathy for which specific factor concentrates are not available.

Cryoprecipitate Cryoprecipitate is the most practical source for fibrinogen rep­lacement. Each unit contains approximately 200 mg of fibrinogen, as well as more than 80 units of factor VIII, vWF, fibronectin, and factor XIII. Achieving fibrinogen plasma levels of 80 to 100 mg/dL and maintaining this level above 50 to 60 mg/dL usually controls hemorrhagic symptoms. To raise the fibrinogen level 100 mg/dL, 0.17 unit of cryoprecipitate per 1 kg of body weight should be infused. Fibrinogen has a long half-life; therefore, replacement therapy can be given at intervals of 3 to 4 days. Cryoprecipitate is indicated for several uses: l For prophylactic use in patients with congenital fibrinogen

deficiencies, vWD unresponsive to DDAVP, and factor VIII deficiency when factor VIII concentrate is not available; l In bleeding patients with vWD or factor VIII deficiency when factor VIII concentrate is not available; l For consumptive coagulopathies when the fibrinogen level is less than 80 to 100 mg/dL; and l For microvascular bleeding in the patient who has had massive transfusion when hypofibrinogenemia cannot be immediately documented.

Factors V, X, XI, and XIII The only current source of factor V is FFP. Normal hemostasis is achieved with levels higher than 25 units/dL. These levels can be achieved with a loading dose of 20 mL/kg of FFP followed by infusions of 6 mL/kg every 12 hours. Factor

V is very labile in FFP, and recently donated FFP should be used. FFP (1 mL/kg) increases the plasma level of factor X by 1 unit/dL. FFP (1 mL/kg) increases circulating factor XI by 1.5 unit/dL. Hemostatic levels of factor XIII are 2 to 3 units/ dL. FFP (5 to 10 mL/kg) is adequate to achieve therapeutic levels. Cryoprecipitate may also be used. One bag of cryoprecipitate contains 75 units of factor XIII.

Complications of Blood Product Administration Citrate Intoxication Citrate, when infused rapidly as the storage solution of blood products, can cause a temporary reduction in ionized calcium levels. FFP has considerably more citrate than do RBCs in citrate-phosphate-dextrose-adenine (CPDA-1). The signs of citrate intoxication include hypotension, narrow pulse pressure, flattening of the instantaneous slope of the arterial catheter tracing, elevated end-diastolic pressures, prolongation of the QT interval, widening of the QRS complexes, and flattening of the T waves. Hypocalcemia is directly related to the rate of citrate administration and is unlikely to occur unless transfusions exceed 1 mL/kg per minute. Impaired perfusion or liver dysfunction lowers this threshold for potential hypocalcemia. Slow calcium administration during rapid or large-volume blood product administration can avert this induced hypocalcemia.

Transfusion-Related Acute Lung Injury (TRALI) TRALI usually manifests as bilateral pulmonary infiltrates within 4 hours of transfusion. The clinical picture consists of acute respiratory distress syndrome, with hypoxia, hypotension, and bilateral pulmonary edema, leading to radiographic opacification and fever within 2 to 6 hours of transfusion (Roseff et al., 2006). Although in most patients, TRALI symptoms usually resolve within 48 to 72 hours, a 5% to 10% mortality rate is associated. TRALI has two proposed mechanisms. In the first mechanism, leukocytes from transfused blood products interact with antibodies in the pulmonary microvasculature, leading to endothelial injury and alveolar exudation. Antileukocyte antibodies are found in this group of patients (Roseff et al., 2006). The other proposed mechanism involves clinical settings such as trauma, sepsis, or massive transfusion, in which cytokine production partially activates endogenous neutrophils. These neutrophils become adherent to the pulmonary microvascular endothelium. On exposure to the lipids contained in transfused blood products, neutrophil activation becomes complete and endothelial damage results in the clinical picture of TRALI (Silliman, 1999). The care of the patient with TRALI is supportive; the role of diuretics and steroids is unproven.

Transfusion-Associated Graft-Versus-Host Disease Transfusion-associated GHVD (TA-GVHD) results when immunocompetent donor lymphocytes are transfused into an immun­ odeficient host patient who is unable to destroy them. These transfused lymphocytes react with host antigens, producing fever, skin rash, pancytopenia, diarrhea, and abnormal liver function test results. TA-GVHD can occur 4 to 30 days after transfusion (Alter and Klein, 2008). The pancytopenia is profound, and

C h a p t e r 36    Systemic Disorders   1165

mortality approaches 100% (Slichter, 2007). Median survival is only 21 days after transfusion. Patients who are at high risk for developing TA-GVHD include neonates; patients with congenital immunodeficiency, leukemia, or lymphoma; and those who have received intensive chemotherapy and bone marrow or solid-organ transplants. Infants and children with severe combined immunodeficiency syndrome (SCIDS) are of most concern, in part because SCIDS is often unrecognized at birth or in early infancy. Although many children with SCIDS are diagnosed by 6 months of age, children have been diagnosed with SCIDS up until 2 years of age. Children with the DiGeorge syndrome (22q11 deletion) and those with conotruncal defects or tetralogy of Fallot in whom immunodeficiency has not been excluded should be considered at risk for TA-GVHD as well. Patients with HIV do not appear to be at increased risk for TA-GVHD. The absence of TA-GVHD in HIV patients may underscore the key role of the recipient’s CD4 cells (depleted early in the course of HIV infection) in the pathogenesis of TA-GVHD (Alter and Klein, 2008). TA-GVHD can occur in patients who do not have an immunodeficiency. Patients who receive products from a donor who is homozygous for a shared haplotype are also at risk for TA-GVHD. The chance of receiving haplotype-homozygous blood from an unrelated donor varies among countries, ranging from 1 in 874 in Japan to 1 in 7147 whites in the United States, and 1 in 16,835 in France. Trauma patients with no known risk factors for TA-GVHD had evidence of microchimerism for as long as 1.5 years after transfusion (Lee et al., 1999). The spectrum of patients who are at potential risk for TA-GVHD is likely to increase with the use of immunosuppressive regimens in the treatment of autoimmune and inflammatory bowel diseases. The prevention of TA-GVHD lies in attenuation of donor lymphocyte reactivity. The only method approved by the FDA is irradiation (Alter and Klein, 2008). Irradiation damages the DNA in donated T cells, which precludes their proliferation and prevents the development of TA-GVHD (Shlomchik et al., 1999). Most blood centers rely on a nominal dose of 25 Gy. Patients at risk for TA-GVHD should receive only irradiated RBCs, platelets, and granulocytes. A summary of situations in which irradiated blood products should be administered is presented in Box 36-14.

Box 36-14 Indications for Irradiated Blood Products Intrauterine transfusions Patients younger than 2 years old Transplantation Bone marrow transplantation Organ transplantation Oncologic conditions Lymphoma Neuroblastoma Glioblastoma Rhabdomyosarcoma Immunodeficiency Congenital immunodeficiency l Wiskott-Aldrich syndrome l Conotruncal abnormalities Acquired immunodeficiency l HIV infection l Patients receiving immunosuppressive drugs Recipients of blood products from first-degree relatives

To avoid the risk of TA-GVHD in patients with late-presenting SCIDS, many blood banks routinely irradiate all cellular blood products given to young children. At present, no consensus exists as to the cut-off age for this practice (Roseff et al., 2006). The Children’s Hospital of Philadelphia irradiates products for children who are younger than 3 months of age, whereas Johns Hopkins Hospital irradiates all products for recipients younger than 6 years of age. Blood components donated from first- or second-degree relatives should be irradiated because of the possibility of HLA haplotype homozygosity. FFP and cryoprecipitate need not be irradiated, because most authorities feel that the freezing and thawing process destroys any donor T cells (Luban, 2002).

Hyperkalemia Acute hemodynamic decompensation after red cell transfusions should be considered hyperkalemia, until proved otherwise. Large volumes (more than 25 mL/kg) of stored red cells given rapidly to infants have been associated with hyperkalemic cardiac arrest (Smith et al., 2008). The combination of perturbations of calcium and potassium concentrations in the context of central venous blood administration may be sufficient to produce hyperkalemic dysrhythmia (Eder, 2002).

MISCELLANEOUS PROBLEMS

Acquired Immunodeficiency Syndrome Epidemiology The epidemic of HIV and acquired immunodeficiency syndrome (AIDS) has had an enormous deleterious impact on worldwide health, and children have not been spared. In 2007, the World Health Organization (WHO) estimated that 33 million people worldwide were infected with HIV. Of this number, 2 million were younger than 15 years of age (WHO, 2007). Most of those infected with HIV live outside of the United States. The Centers for Disease Control and Prevention (CDC) estimated that 1 to 1.2 million people in the United States were infected with HIV in 2003; during 2006, the most recent year with complete statistics, approximately 50,000 new HIV infections were reported (Glynn, 2005; CDC, 2007a; Hall et al., 2008). The incidence of HIV infection varies by both ethnicity and geography. In the United States, AIDS is seen more often in Hispanic and African American children than in Caucasians. In 2008, the HIV/AIDS surveillance system of the CDC reported that 49% of children with AIDS were African American, 18% were Hispanic, and 30% were Caucasian. Most cases in the United States have been reported in the Northeast (23%) and South (25%), primarily in the large metropolitan areas (CDC, 2007a). Transmission of HIV can occur through parenteral exposure to blood, sexual contact, or through vertical transmission from mother to child. In the United States, nearly all HIV infections in children younger than 13 years are the result of vertical transmission. Approximately 6000 to 7000 children are born to HIV-positive mothers annually in the United States. Women of childbearing age comprise one of the fastest growing groups with HIV infection in the United States, accounting for more than 27% of adult HIV cases reported (CDC, 2007b). Although transmission of HIV from mother to child can occur before, during, or after delivery, the highest percentage of HIV-infected children acquire the virus during delivery, most likely through exposure

1166   P a r t  IV    Associated Problems in Pediatric Anesthesia

to infected blood and secretions during delivery. Chance of transmission is increased with preterm birth, low birth weight, low maternal CD4 counts, and IV drug use during pregnancy. With cesarean section and prenatal, intrapartum, and neonatal antiretroviral treatment, vertical transmission in the United States has been decreased dramatically (AAP Committee on Pediatric AIDS, 2000; CDC, 2007b). In the United States, postpartum transmission via breastfeeding is the least common mode of perinatal transmission; however, it is quite common in developing countries. The number of perinatally acquired HIV cases in the United States peaked in 1992 at nearly 1000 and decreased to fewer than 100 in 2006 (Mofenson et al., 2006). This decline is thought to be related to more widespread use of the Public Health Service guidelines for universal counseling and voluntary HIV testing of pregnant women and more effective use of retroviral therapy in these patients (AAP Committee on Pediatric AIDS, 2000). Although adolescents account for a small percentage of AIDS cases, the number of adolescents with AIDS is growing rapidly. In 2007, nearly 2000 young people in the United States were diagnosed with HIV infection or AIDS (CDC, 2007a). Transmission of HIV through contaminated blood or blood products accounts for approximately 3% of cases of HIV infection in the United States. Screening of blood products for HIV began in the United States in 1985, and since then the risk of transmission has decreased dramatically. Sexual transmission of HIV, although relatively rare in pediatrics, is a growing problem (Romero et al., 2007). The age at which HIV infection is diagnosed in children varies with the mode of transmission. Children who were infected by transfusion of contaminated blood or blood products, primarily between the years of 1978 and 1985, are now adults (Grubman et al., 1995). Children who

acquired the disease during birth can be diagnosed as infants, with detection of the virus as early as 1 week of age. The viral load increases with time, and the virus is detectable in the peripheral blood of almost all infected infants by the age of 4  months. HIV has three patterns of progression. Newborns who were infected during gestation develop symptoms within the first few months of life and have detectable virus early in life. Without treatment, they often die before their first birthday. Most newborns with HIV have acquired the infection during birth, and these infants have a much slower progression of disease. The viral load increases in the first few months of life and then declines over the subsequent 2 years. A small percentage of infants with perinatal infection survive for an extended time with minimal progression of the disease. Children infected with HIV have generally the same immunologic manifestations as adults. Because infants and children have a lymphocytosis at baseline, true lymphopenia is relatively rare. However, relentless depletion of CD4 lymphocytes occurs, resulting in the development of opportunistic infections. Infants and children with HIV/AIDS have many opportunistic infections. The most common serious infections are pneumonia, bacteremia, and sepsis, and nearly every organ system may be affected. The CDC has classified HIV in young children based on the presence of signs and symptoms, the state of immunodeficiency, and the immunologic category of the disease (Caldwell et al., 1994). The current recommendations for initiation of antiretroviral therapy in infants, children, and adolescents were summarized in 2008 by the Working Group on Antiretroviral Therapy and Medical Management of HIV Infected Children, and these guidelines can be found on the CDC website and in Box 36-15 (Guidelines for Use of Antiretroviral Agents, 2009).

Box 36-15 Guidelines for the Use of Antiretroviral Agents in Pediatric HIV Infection Preferred Regimen Children 3 years and older: two NRTIs plus efavirenz* Children younger than 3 years old or who cannot swallow ­capsules: two NRTIs plus nevirapine Alternative Two NRTIs plus nevirapine1 (children 3 years and older) PROTEASE INHIBITOR-BASED REGIMENS Preferred Regimen Two NRTIs plus lopinavir/ritonavir Alternative Two NRTIs plus atazanavir plus low-dose ritonavir (children older than 6 years) Two NRTIs plus fosamprenavir plus low-dose ritonavir (children older than 6 years) Two NRTIs plus nelfinavir (children older than 2 years) USE IN SPECIAL CIRCUMSTANCES Two NRTIs plus atazanavir unboosted (for treatment-naïve adolescents older than 13 years and weighing more than 39 kg)

Two NRTIs plus fosamprenavir unboosted (children older than 2 years) Two NRTIs plus saquinavir plus low-dose ritonavir only in postpubertal adolescents who weigh enough to receive adult doses Zidovudine plus lamivudine plus abacavir TWO NRTI BACKBONE OPTIONS Preferred Abacavir plus (lamivudine or emtricitabine) Didanosine plus emtricitabine Tenofovir plus (lamivudine or emtricitabine) (for Tanner stage 4 or postpubertal adolescents only) Zidovudine plus (lamivudine or emtricitabine) Alternative Abacavir plus zidovudine Zidovudine plus didanosine Use In Special Circumstances Stavudine plus (lamivudine or emtricitabine)

From Working Group on Antiretroviral Therapy and Medical Management of HIV-Infected Children, 2009: Guidelines for the use of antiretroviral agents in pediatric HIV infection. Available at http://aidsinfo.nih.gov/ContentFiles/PediatricGuidelines.pdf; accessed May 8, 2009. *Efavirenz (a nonnucleoside reverse transcriptase inhibitor [NNRTI]) is currently available only in capsule form and should only be used in children aged 3 years and older who weigh 10 kg or more; nevirapine would be the preferred NNRTI for children aged younger than 3 years or who require a liquid formulation. Unless adequate contraception can be assured, efavirenz-based therapy is not recommended for adolescent females who are sexually active and may become pregnant. PI, Protease inhibitor; NRTI: nucleoside/nucleotide reverse transcriptase inhibitors.

C h a p t e r 36    Systemic Disorders   1167

Clinical Presentation In the perioperative environment, caring for children with HIV/ AIDS is focused on two major considerations: effects of this systemic infection and its treatment on a child’s readiness for anesthesia and surgery, and protection of health care workers from acquiring infection as a result of exposure to the blood or secretions of these children. As many as 80% of children with HIV develop lung disease (McSherry, 1996). Pulmonary function is compromised by bacterial and viral infections, but in addition many children with AIDS also develop lipoid interstitial pneumonia (LIP) (Rubinstein et al., 1986). The incidence of LIP in children infected by HIV is 20% to 30%. Clinical characteristics of this condition include tachypnea, wheezing and cough, hypoxemia, and even clubbing. This condition presents with bilateral infiltrates. Severely affected children may have bronchiectasis and lung cysts. The most common opportunistic infection responsible for pneumonia in children with HIV infection is Pneumocystis carinii pneumonia. Most people are infected with this common organism in childhood, but disease is caused only in immunocompromised individuals. The peak incidence of this infection occurs in children who are younger than 1 year of age. P. carinii is often characterized by the acute onset of fever, hypoxemia, and respiratory distress, but a disease with a more desultory onset is also seen (Simonds et al., 1998). Infection with respiratory viruses in children with HIV is common and often more severe in these patients. In children with AIDS, infections with RSV, parainfluenza, and influenza viruses are more likely to be symptomatic, and infections with adenovirus or measles may lead to serious morbidity (Englund et al., 1998). Children with LIP or infectious pneumonia may come to the operating room for bronchoscopy and bronchoalveolar lavage for diagnosis (Birriel et al., 1991). They are often hypoxemic and in respiratory distress before the procedure. These cases are often quite challenging, because the diagnosis must be established to prescribe the correct therapy; the procedure cannot be delayed until the child’s condition improves. Children who are infected in the perinatal period with HIV often have involvement of the CNS, although opportunistic infections of the CNS are uncommon. In children infected during birth, encephalopathy is seen in 10% of those infected with HIV and in 23% of those diagnosed with AIDS (Navarro and Hanson, 1996). In toddlers, the presentation is that of a progressive encephalopathy with loss of developmental milestones or arrest of development. CNS pathology includes low brain weight, acquired microcephaly, inflammatory infiltrates, and calcific vasculopathy of the basal ganglia vessels. As the encephalopathy progresses, loss of fine and gross motor skills and language skills may occur and behavioral problems may develop. In older children, the clinical picture often becomes one of a static encephalopathy (Bowers et  al., 1998). Seizures and focal neurologic signs are unusual and their presence should prompt a search for other causes, such as infection, stroke, or a tumor. It is important to document signs and symptoms of encephalopathy or other CNS involvement in the preoperative evaluation. Approximately 10% to 20% of children with HIV infections have clinically significant cardiovascular involvement. Careful echocardiographic and electrocardiographic evaluations of children with HIV infections may uncover subtle abnormalities in a much higher percentage of patients (Lipshultz et al., 1989). Common abnormalities include resting sinus tachycardia, sinus

dysrhythmias, and ventricular hypertrophy. Echocardiography studies of children with HIV infection have demonstrated both left ventricular diastolic and systolic dysfunction. One center reported that 10% of children with HIV required temporary treatment for congestive heart failure, generally during an intercurrent illness (Luginbuhl et al., 1993; Keesler et al., 2001). Hepatosplenomegaly and a gallop are indications of congestive heart failure in these patients, and medical therapy has been generally effective in reversing the symptoms. In children with advanced AIDS, hemodynamic instability may occur (Evenhouse et al., 1987; Stewart et al., 1989). Children with AIDS often have lowered counts of all the formed elements of the blood. Poor bone marrow function in these patients may be caused by the disease itself, by poor nutrition, or as a side effect of the medications used to treat the disease. As with many chronic conditions, the anemia seen in AIDS patients is often normochromic and normocytic, with low reticulocyte counts. In the preoperative evaluation of these children, other causes for anemia, such as occult bleeding, should be ruled out (Russell and Nedeljkovic, 2001). Treatment of anemia in HIV-infected children should address the cause and should include administration of erythropoietin (rh-EPO). Perioperative transfusion of RBCs should be undertaken only to provide a minimum level of oxygen delivery and after careful consideration of the possible deleterious effects in patients with HIV/AIDS (Hillyer et al., 1999). Only cytomegalovirus-­negative, leukocyte-depleted RBCs should be used. Patients with HIV have a high prevalence of thrombocytopenia. The cause is sometimes difficult to determine, but both impaired production and increased destruction have been found in patients with AIDS. In addition, a lupus-type anticoagulant was found in 20% of AIDS patients undergoing routine coagulation testing (Cohen et al., 1986). GI and nutritional problems can be common and difficult to treat in children with AIDS. Infections of the GI tract cause major morbidity and can be quite severe (Doyle and Pickering, 1990). Oral infections with Candida or ulcerative gingivitis are seen in children infected with HIV. Bacterial, viral, or fungal infections may cause diarrhea. With recurrent or chronic diarrhea, children develop malnutrition and failure to thrive. Growth failure in these children can result from malabsorption, poor nutrient intake, and possibly altered energy use. Hepatosplenomegaly is seen in up to 80% to 90% of children infected with HIV and is associated elevated serum aminotransferases. Coinfection with hepatitis B or C virus is more prevalent in children with HIV than in the general population. Pancreatitis can occur, usually as a complication of the drug therapy used to treat HIV or one of the opportunistic infections. Other clinical signs of HIV infection include various skin rashes such as eczema or seborrhea and manifestations of renal dysfunction such as proteinuria, hematuria, hypoalbuminemia, and edema.

Anesthesia and Procedures Children with HIV/AIDS often undergo diagnostic bronchoscopies with bronchoalveolar lavage or biopsy, diagnostic upper and lower endoscopies, placement of gastrostomy tubes for nutritional support, and placement of central venous ­catheters (Birriel et al., 1991). In addition to procedures specific to their underlying systemic viral infection, these children can require anesthesia for any other routine or emergent surgical or diagnostic procedure such as myringotomy or herniorraphy. In the

1168   P a r t  IV    Associated Problems in Pediatric Anesthesia

evaluation of these children before anesthesia, the effects of the infection on organ systems as outlined above should be evaluated. The anesthesiologist should be aware of medications taken, their effects, and their side effects. In addition to medications specific for the treatment of AIDS, these children are often being treated with antibiotics and corticosteroids. Pulmonary insufficiency is often seen in patients with AIDS and despite meticulous management during the procedure, clinical deterioration commonly occurs after the procedure. During the preanesthetic visit, the anesthesiologist should, when indicated, discuss the possibility of postoperative intubation and ventilation. In an advanced case in which a do not resuscitate (DNR) order or other advance directive is in place, the anesthesiologist should discuss these thoroughly with the child and family (Truog et al., 1999). It may be that the person caring for the child is not a biological parent, necessitating additional administrative steps in the informed consent process. Although the specific agents and techniques chosen for a case depend on the particular child and situation, the physician should consider several points when making those choices. CNS involvement is relatively common; therefore, CNS depressants such as barbiturates, benzodiazepines, and opioids should be carefully titrated. If liver or renal dysfunction is present, drug metabolism and elimination are impaired. Attention to sterile technique, often not a priority among anesthesiologists, is paramount in these children who are severely immunocompromised.

Pain Management Children with HIV/AIDS may have many causes for pain and suffering independent of postoperative pain (Box 36-16) (Nedeljkovic, 2001). The anesthesiologist or pediatric pain specialist should be prepared to participate in the management of both the acute and chronic pain that afflicts these children. In one report, 59% of children with HIV described pain as having an important impact on their lives (Hirschfeld et al., 1996; Yaster and Schecter, 1996). The clinical presentations of pain and suffering in children with AIDS are varied, and the pharmacologic and nonpharmacologic treatments that may be employed are broad (Box 36-16) (Gaughan et al., 2002). Assessment is often difficult because of the nature of the discomfort and the difficulty in communicating with children who are afflicted with encephalopathy.

Box 36-16 Common Pain Syndromes in HIV/AIDS Patients Gastrointestinal Tract Visceral pain Esophagitis Pancreatitis Sclerosing cholangitis Nervous System Headache Peripheral neuropathies Musculoskeletal System Arthralgia Myositis

HIV/AIDS and Health Care Providers Individuals who care for HIV-infected children must take prudent steps to prevent transmission (CDC, 2002). Although HIV has been isolated from saliva, the titer is generally low. Studies of hundreds of household contacts have confirmed that the risk of transmission from casual contact is nearly zero. It is therefore unlikely that operating-room personnel would contract HIV from passive contact with an child who is HIV positive. Exposures that place these health care personnel at risk for contracting HIV include needlesticks, cuts with sharp objects, and contact of mucous membranes or nonintact skin with blood, tissue, or other body fluids. The greatest risk of contracting HIV for health care workers is via needlestick with a contaminated needle. Hollow-bore needles (those used to administer medications) give a much larger inoculum of blood and infectious agent than do solid needles. Estimates of the average risk for various types of exposure are 0.3% to 0.4% after percutaneous exposure (needlestick or laceration), 0.09% after mucous membrane exposure, and less than 0.09% after exposure to nonintact skin. After parenteral exposure to HIV, specific steps should be taken; within existing state and local laws, evaluation of the HIV status of the source of the exposure, postexposure prophylaxis, postexposure treatment, and follow-up care should occur. A health care worker who has a parenteral exposure to blood or body fluids from a child known or suspected to be infected with HIV should have the wound thoroughly washed and then irrigated with saline. Exposed mucous membranes should be thoroughly irrigated with saline. The exposure should be immediately reported to the institutional employee health service or to the “stick” team if one exists. In addition, the patient should be tested for hepatitis, which may coexist with HIV. If the source is known to be HIV positive, details of the infection (such as CD4+ count, viral-load tests, and the source’s current treatment) should be gathered quickly, so that the medications for postexposure prophylaxis can be chosen. Postexposure prophylaxis should be undertaken as close to immediately after parenteral exposure to the blood or body fluid from a child suspected of having or known to have HIV (Panlilio et al., 2005). In situations in which the HIV ­status is not known, prophylaxis should be given on a caseby-case basis. In most cases, administration of two antiretroviral drugs is indicated. The addition of a third medication should be considered in cases with increased risk for transmission. Before prophylaxis is begun, the employee should be tested to document HIV status, with subsequent testing at 6 and 12 weeks. A retrospective study undertaken by the CDC to determine the rate of seroconversion of health care workers in England, France, and the United States reported three conclusions: exposure to a large quantity of blood was associated with a higher rate of seroconversion; seroconversion was more likely when the exposure was from a patient in the terminal stages of AIDS; and a 79% decrease in seroconversion occurred when zidovudine was begun after exposure. Anesthesiologists should be familiar with the needlestick policies of their institutions and be prepared to follow them and advise others of the policies in the event of an exposure. Universal precautions, as recommended by the CDC, should be followed by all health care workers who have direct patient contact or exposure to a patient’s body fluids (CDC, 2009).

C h a p t e r 36    Systemic Disorders   1169

Latex Allergy The first published report in an American medical journal about an allergy to rubber gloves appeared in 1933 (Downing, 1933). Sporadic reports followed until the late 1980s and early 1990s, when reporting of allergic reactions to latex rose sharply. This increase is thought to have been the result of increased exposure of health care workers and patients after the publication of the Universal Precautions Guidelines by the CDC in 1987 (CDC, 1987). After 1987 the use of surgical gloves in the United States increased by a factor of 25, from 800 million to 20 billion annually. Allergic reactions to latex were first reported by pediatric anesthesiologists in 1991, before circulation of the Medical Alert (in  1991) by the FDA that warned health care workers of this emerging problem (Holzman, 1993). Latex allergy is a significant problem in health care. As of 1997, through its mandatory reporting mechanism, the FDA had received more than 2300 reports of allergic reactions involving medical products that contain latex, with 225 cases of anaphylaxis, 53 cardiac arrests, and 17 deaths. Individuals at high risk for latex allergy have certain common characteristics, which are listed in Box 36-17 (Hochleitner et al., 2001; Randolph, 2001; Hourihane et al., 2002). Many children who have had latex reactions in the operating room have had spina bifida or urinary tract anomalies (Cremer et al., 2007). These two groups, not surprisingly, undergo multiple surgical procedures, making it difficult to ascertain whether the high prevalence of latex allergy is simply the result of repeated exposure or to an immunologic response associated with specific conditions. Reactions to latex have been divided into three types: irritant contact dermatitis, type IV hypersensitivity (skin reactions similar to poison ivy), and type I, or IgEmediated, hypersensitivity. Type I hypersensitivity is by far the more severe reaction. All 17 of the above-mentioned deaths reported to the FDA were the result of type I hypersensitivity. This type of response to latex has been reported in many clinical settings, including intraabdominal surgery, genitourinary surgery, and dental procedures. Some have reported reactions associated with airborne exposure as a result of being in the vicinity of someone donning latex-containing gloves. Manifestations of type I hypersensitivity are listed in Box 36-18.

Box 36-17 Individuals at High Risk for Latex Allergy Patients who have undergone multiple surgical procedures Patients with spina bifida (meningomyelocele) Health care personnel Individuals with a history of atopy Individuals with a history of allergy to tropical fruits

Generally, intraoperative type I hypersensitivity does not occur immediately at the beginning of a surgical procedure, but rather after exposure of the peritoneum or other mucous membranes to latex. The presentation includes bronchospasm, hypoxemia, hypotension, and tachycardia. Skin manifestations, such as urticaria or flushing, may also occur. In a series of patients reported by Holzman, the mean SpO2 fell from 100% to 92% (Holzman, 1993). Bronchospasm and hypotension are difficult to treat, even with IV epinephrine, and the manifestations may persist until the exposure is stopped.

Treatment of Intraoperative Anaphylaxis As increasingly more latex-free medical equipment is manufactured, it is important to remain vigilant for the possibility of inadvertent latex exposure with an at-risk patient. Delay in diagnosis of an episode of latex anaphylaxis only makes treatment more difficult and likely continues exposure to the offending allergen. The mainstays of treatment are stopping the latex exposure and resuscitation. All latex must be removed from the surgical field, as well as those materials for which the latex content is unknown, and the procedure must be ended as rapidly as possible. If blood or antibiotics are being administered, this administration should be stopped. Consideration should be given to evaluating the patient for a transfusion reaction if the symptoms and signs of anaphylaxis began during blood administration. Resuscitation efforts are directed toward stabilization of vital signs and reversal of the pathophysiology of anaphylaxis. Because these reactions often occur during intraabdominal surgery, patients are often already intubated when the reaction occurs. If the patient is not intubated, strong consideration should be given to intubation. If possible, based on the progress of the surgery and the patient’s vital signs, administration of anesthetic agents should be stopped. IV fluid and epinephrine doses (starting at 0.001 mg/kg) should be given to maintain blood pressure in the normal range for the patient, a Foley catheter should be placed, invasive hemodynamic monitoring should be instituted, and if bronchospasm is present, inhaled bronchodilators should be given through the ETT. If repeated doses of epinephrine are needed, as often occurs, an infusion of 0.05 to 0.1 mcg/kg per minute should be started. The patient may require treatment as outlined above for several hours, and admission to the intensive care should be arranged. A Latex Alert sign should be placed outside of the patient’s operating room and in the intensive care unit, and the condition should be noted prominently in and on the medical record. Once the vital signs are stable, secondary treatments can be instituted, including administration of diphenhydramine, ranitidine, and hydrocortisone. Further therapy depends on the patient’s condition as the resuscitation progresses.

Diagnosis of Latex Anaphylaxis Box 36-18 Manifestations of Type I (IgE-Mediated) Hypersensitivity Hives, urticaria, red eyes, angioedema of the eyelids Nasal congestion GI cramping, nausea, diarrhea Headache, anxiety Shortness of breath, bronchospasm, tachycardia, hypotension, anaphylaxis

After the patient is stabilized, tests to document the diagnosis of latex allergy can be performed. Although many tests are available to confirm diagnosis, a universally accepted serum test for the diagnosis of a type I hypersensitivity reaction is not available. An elevated level of serum tryptase occurs within the first 4 hours in patients who have experienced anaphylaxis with mast-cell degranulation, regardless of the cause. The radioallergosorbent (RAST) or enzymeallergosorbent (EAST) tests are available for specific proteins. Blood should be sent for testing,

1170   P a r t  IV    Associated Problems in Pediatric Anesthesia

with the realization that a specific individual may or may not react to that particular protein. To determine type I hypersensitivity to latex, current practice is to use a skin prick test that employs antigen extracted from latex similar to that in medical products. Testing materials and methods have been standardized for some time (Hamilton et al., 2002). The patient should be referred to a specialist in allergy and immunology for complete evaluation. Skin testing should be delayed for 4 to 6 weeks after the episode of anaphylaxis to allow time for cellular inflammatory mediators released during the reaction to be reconstituted (Dakin and Yentis, 1998). Testing before allowing time for these proteins to be reconstituted increases the risk for false-negative results. This testing must be performed carefully and in the proper setting; severe reactions have been seen in sensitive individuals even with the minimal exposure that occurs with this test (Kelly et al., 1993). Once the diagnosis is confirmed, patients should be encouraged to wear a medical alert bracelet.

Recommendations Avoidance of latex exposure is by far the best safeguard against the risks associated with latex sensitivity (Holzman, 1997). Although operating rooms nationwide are working toward becoming completely free of latex, 100% compliance has not yet been achieved. Avoidance, therefore, depends on recognizing those at risk for latex sensitivity (Box 36-17). Most pediatric anesthesiologists in the United States avoid latex exposure from birth in patients with meningomyelocele because of the high prevalence of sensitivity. In some hospitals, children with bladder exstrophy are treated similarly. The preanesthetic assessment should include questions about atopy and allergies to foods, especially tropical fruits. Children thought to be at risk based on history simply should not be exposed to latex. If a child who is thought to be at risk for latex sensitivity by virtue of a medical or surgical condition denies (or the parents deny) facial redness after touching balloons or after dental care and has not undergone latex-sensitivity testing, it still seems prudent to avoid latex-containing products, especially as more and more products are manufactured to be free of latex. All equipment used by anesthesiologists can be obtained in forms that are free of latex, including IV sets and tubing, breathing circuits and breathing bags, and sterile and nonsterile gloves. Most multidose vials are made with latex-free stoppers. Latex Alert warning signs should be placed outside the operating room door to avoid inadvertent latex exposure in these patients. Clinicians must keep themselves informed about the progress in this area. It is important to note that not all medical equipment and products are free of latex; therefore, each new product should be checked for possible latex content. It is true that products that were formerly unsafe become safe with a change in manufacturing. However, given the high morbidity of an anaphylactic reaction, it is essential that caregivers be certain of the safety of medical products used in children at risk for latex sensitivity. The FDA approved the marketing of a patient-­examination glove that is produced from a new form of natural rubber latex made from the guayule bush, a desert plant found in the American Southwest. Available data on this glove show that individuals who are allergic to latex made from the sap of the rubber tree (heva braziliensis) do not, on first exposure, react to the guayule latex (FDA, 2008).

Prophylaxis and Desensitization Although it is difficult to create a completely latex-free environment, the current consensus seems to be that prophylaxis of patients with known or suspected latex sensitivity need not be undertaken. Those who endorse prophylaxis propose administration of diphenhydramine, ranitidine, and corticosteroids from the preoperative through the postoperative period. The literature contains case reports of patients who developed anaphylaxis after exposure to latex despite preoperative administration of the recommended prophylactic medication (Kwittken et al., 1992; Setlock et al., 1993). Desensitization has been successful in a limited number of reported cases (Patriarca et al., 2002a, 2002b). Many of the participants in these efforts were actually health care workers with documented IgE-mediated latex allergy, but none of them had previously suffered anaphylaxis after exposure to latex. The clinical manifestations of latex allergy included asthma, angioedema, and urticaria. Although most patients were adults, one report included subjects as young as 8 years of age. Techniques used were cutaneous exposure over a 12-month period and 4-day desensitization via sublingual exposure. At this point, however, desensitization does not appear to be an option for children with type I hypersensitivity to latex. As more details of the immunologic basis for latex reactions become known, new and safer desensitization programs may be developed (Rolland and O’Hehir, 2008).

Occupational Latex Allergy In 1998, a report of latex sensitivity among the staff of the Department of Anesthesiology at the Johns Hopkins University School of Medicine documented a 24% incidence of irritant or contact dermatitis and nearly 13% incidence of latex-specific IgE positivity, although pediatric anesthesiologists may have a somewhat lower incidence (Brown et al., 1998; Greenberg et al., 1999). A large meta-analysis of studies in health care workers showed a 0% to 30% prevalence of type I latex allergy in that group. The authors did not have data that elucidated the reasons for the large variation in prevalence (Garabrant and Schweitzer, 2002). Other reports suggest that avoidance of latex reverses the sensitivity, at least in health care workers (Zeldin et al., 1996). The creation of a latex-free operating room environment will benefit both the patients and those who care for them.

Epidermolysis Bullosa Epidermolysis bullosa (EB) encompasses a heterogeneous group of congenital, hereditary, blistering disorders and is subdivided into three major subtypes: EB simplex, junctional EB, and dystrophic EB. These types differ in histology, clinical severity, and mode of inheritance, but all are characterized by the easy development of blisters after minor trauma or friction. The literature contains several reviews of the perioperative and anesthetic care of children with one of the variants of EB (Tetzlaff and Fleisher, 2005; Lin and Golianu, 2006; Lindemeyer et al., 2009). Junctional EB is often clinically apparent early in life and heals with scarring. A discriminating feature of this variant is the relative sparing of the hands and feet. However, involvement of the mucous membranes may be severe, and ulceration of the respiratory epithelium has been documented. The recessive variant of dystrophic EB may be the most severe

C h a p t e r 36    Systemic Disorders   1171

form of the condition, in which mucous membrane lesions are common. Treatment of children with this condition is supportive. Infections are common and should be promptly treated. Adequate nutrition is paramount but often difficult to provide in cases with esophageal blisters with subsequent stricture formation. Children with the more severe forms may come to the operating room for a variety of procedures such as scar revisions, corrections of digital fusions, placement of gastrostomy tubes, or even colonic interpositions.

Anesthetic Management The preanesthetic evaluation of a child with EB should involve the patient’s dermatologist and pediatrician, who can advise the anesthesiologist of the child’s general course with regard to blistering, skin infections, nutritional status, and the possible usefulness of additional steroid administration. In addition to assessing the child’s general condition and health, the physical examination should focus on the airway (that may be compromised by scarring around the mouth), IV access sites, and the location and condition of existing and recent blisters. Friction and, secondarily, pressure must be avoided in caring for children with this condition in the perioperative period. Monitoring must adhere to the ASA standards, but the application of the monitors should be modified. Pulse oximetry may be accomplished with an adult clip-on probe (that may be impossible if the patient has complete pseudosyndactyly of the fingers and toes), or the adhesive strip oximeter probe may be placed over a clear plastic bag covering the hand or foot that is then wrapped with Webril or Coban (3M; Saint Paul, Minnesota) (Fig. 36-21). The precordial stethoscopes should simply be placed (without adhesive) onto the chest; temperature should be monitored with an axillary probe; soft padding should be placed between the skin and noninvasive blood pressure cuffs; electrocardiograph leads should be nonadhesive (e.g., needle electrodes); and IV and arterial catheters should be sutured and lightly covered with a gauze bandage (e.g., Webril or Coban). A nonadhesive, silicone-based dressing (Mepilex; Mölnlycke Health Care AB, Göteborg, Sweden) that is used for protecting blistered areas in these patients has recently become available. It has been found to be useful for securing venous and arterial catheters, holding precordial stethoscopes in place, and facilitating contact

n  FIGURE 36-21. A thin plastic bag is placed over the digit; the adhesive pulse oximeter probe is then placed over the plastic bag. The extremity is then wrapped with Webril to keep the pulse oximeter bag assembly in place.

of electrocardiograph gel pads in which the adhesive has been removed (Figs. 36-22, 36-23, and 36-24). It can also be used to cover the contact surfaces of the anesthesia mask to substitute for the previous practice of coating the mask surface with steroid ointment, which creates a slippery environment for subsequent handling of the patient. The eyes should be lubricated but not taped closed with adhesive tape. (Mepilex may be used.) A variety of techniques for anesthetic induction and maintenance have been reported for procedures performed on these challenging patients (Holzman et al., 1987; Farber et al., 1995; Herod et al., 2002; Lindemeyer et al., 2009). An oral premedication may be useful in children with EB who are taken to the operating room, because a struggling child who is restrained may develop blisters where held by the operating team. Alternatively, intramuscular ketamine has been used to both induce and maintain anesthesia in these patients (Ames et al., 1999). The

n  FIGURE 36-22. The IV cannula and tubing are held in place with a strip of Mepilex transfer.

n  FIGURE 36-23. An arterial cannula and tubing are held in place with Mepilex transfer strips on the skin and adhesive tape applied to the Mepilex transfer.

1172   P a r t  IV    Associated Problems in Pediatric Anesthesia

n  FIGURE 36-25. A nasal ETT is secured to a head wrap. No adhesive is applied to the face.

n  FIGURE 36-24. Adhesive is trimmed from the contact area of infant electrocardiograph pads. Contacts consisting only of gel and wire are secured to patient with strips of Mepilex transfer.

­induction technique chosen depends on the preoperative assessment and the planned procedure. The induction of general anesthesia can be via an inhalational technique, but contact with the child’s face must be gentle. Oropharyngeal airways should be avoided. Laryngoscopy with a straight blade without contact with the epiglottis is preferred, and intubation with a smallerthan-predicted, softened (in warm, sterile saline), lubricated ETT after administration of muscle relaxant will minimize trauma to the oral, supraglottic, and tracheal mucosa. Most physicians recommend securing the tube with umbilical tape tied around the back of the head. If there are no contraindications, a deep extubation decreases tracheal trauma caused by coughing. Cases of postoperative bullae in the pharynx, some of which caused airway obstruction, have been reported, but predisposing factors are difficult to identify (James and Wark, 1982). Most series report no airway complications despite the formation of new oral bullae. Limited mouth opening in many patients with EB necessitates fiberoptic nasotracheal intubation. This route is a good choice for all patients with EB—even those with adequate mouth opening—because oral intubation causes more trauma and bulla formation to the tongue and oral mucosa because of pressure of the laryngoscope on the supraglottic area (Baum and O’Flaherty, 2006). The nasal mucosa is composed of pseudostratified cylindrical ciliated epithelium with goblet cells (i.e., respiratory epithelium), stratified cuboidal epithelium, and stratified squamous nonkeratinized epithelium. The first two types of epithelium are less vulnerable to bulla formation than the latter two, which comprise the epithelium of the oral mucosa. In addition, the nasotracheal tube may be secured more easily without tape than an oral tube (Fig. 36-25). Care should be taken to use a tube small enough to avoid pressure on the skin surface at the entrance to the naris. As is customary for

nasal intubation, the nare should be prepared with lubrication and vasoconstriction (e.g., oxymetazoline 0.05% or phenylephrine 0.25%), and the tube should be softened as described. Oral rehabilitation procedures, commonly required by EB patients because of the inability to brush their teeth, are more easily accomplished with a nasotracheal tube (Griffin and Mayou, 1993). Some clinicians have recommended nasal intubation (with fiberoptic guidance) to eliminate the trauma and friction of direct laryngoscopy. It is also thought that the passage of the tube is not as traumatic and the nasopharyngeal mucosa is not as vulnerable to bulla formation as is the oral, gingival, and supraglottic mucosa, especially with the pressure of a laryngoscope blade. In addition, a nasal tube can be secured to a head-wrap dressing, eliminating the need for any ties that may abrade the face. The decision to admit these children after the procedure must be made on an individual basis. However, they warrant careful observation in a well-monitored environment after surgery and anesthesia. Even though general anesthesia is used often in these patients, reports of successful regional techniques have been published (Kaplan and Strauch, 1987; Cakmakkaya et al., 2008; Nasr et al., 2008).

Down Syndrome The incidence of Trisomy 21 is 1 in 600 to 800 live births. More than half of Trisomy conceptions spontaneously abort early in pregnancy. The syndrome has many clinical manifestations, some of which are of particular note to the anesthesiologist. Approximately 40% of children with Trisomy 21 have anomalies of the cardiovascular system. The three most common anomalies seen in these children are complete atrioventricular canal (CAVC) (comprising approximately 40% of the total), ventricular septal defect (25%), and atrial septal defect (10% to 15%). Children with Down syndrome who undergo repair of CAVC have significantly higher perioperative mortality than those without Trisomy 21. However, the outcome after surgery

C h a p t e r 36    Systemic Disorders   1173

is not different for other cardiac anomalies. These defects have in common the propensity for increased pulmonary blood flow, which the anesthetic management should plan to minimize. Repair or palliation of a cardiac defect does not eliminate the need for particular attention to the cardiovascular system in the evaluation of these patients preoperatively. Children with Down syndrome have varying degrees of mental retardation, and it is important to be aware of the degree of intellectual impairment when meeting and talking with them. Hypotonia is one of the most common clinical features seen in these children, and it may affect the patency of their upper airways. The relatively large tongue, short neck, and crowded midface and laryngomalacia contribute to upper airway obstruction (Kanamori et al., 2000; Mitchell et al., 2003). Partial airway obstruction, while awake and during sleep, is often seen in children with Trisomy 21. This situation is exacerbated by the administration of sedatives and during inhalation induction (Luscri and Tobias, 2006; Ng et al., 2006). Children with Trisomy 21 have an increased incidence of subglottic stenosis, and often the proper size of ETT for a given child is smaller than would have been predicted. The incidence of tracheal stenosis is also significant because of complete tracheal rings. An orthopedic anomaly of great concern in these children is ligamentous laxity of the atlantoaxial joint, which may predispose affected individuals to C1-C2 subluxation and possible spinal cord damage. The incidence of this anomaly is 12% to 32%, depending on the ages of the children studied and the exact definition of laxity used (Hata and Todd, 2005; Pizzutillo and Herman, 2005). The incidence of hearing loss and of hypothyroidism is increased in these children (Tuysuz and Beker, 2001). Other associated findings in patients with Down syndrome are included in Box 36-19.

Perioperative Management The preoperative evaluation of a child with Down syndrome should give particular attention to the organ systems commonly affected by this condition (Mitchell et al., 1995). The history of prior surgeries should be reviewed. These children may have undergone cardiac procedures, removal of the tonsils and adenoids, myringotomy tube placement, and other common pediatric procedures. Records from other doctors may contain helpful information about associated conditions such as obstructive sleep apnea syndrome, atlantoaxial laxity, or subluxation. Management of these children regarding possible C1-C2 subluxation is a difficult matter. The American Academy of Pediatrics (AAP) has published statements by the AAP Committee on Genetics and AAP Committee on Sports Medicine and Fitness that include a discussion of this clinical problem. The Committee on Genetics policy statement on health care supervision of children with Down syndrome recommends that radiographs looking for evidence of atlantoaxial instability or subluxation be obtained at between 3 and 5 years of age (AAP Committee on Genetics, 2001). The Committee on Sports Medicine and Fitness reviewed the topic of atlantoaxial instability in Down syndrome in a 1995 publication and tentatively concluded that lateral plain films are of potential but unproved value in detecting patients at risk for developing spinal cord injury during participation in sports (AAP Committee on Sports Medicine and Fitness, 1995). The Special Olympics does not plan to remove its requirement that all athletes with Down syndrome receive lateral spine radiographs (Special Olympics,

Box 36-19 Associated Findings in Patients with Down Syndrome General findings Low birth weight Short stature Cardiovascular findings Congenital heart disease Increased susceptibility to pulmonary hypertension Atropine sensitivity Respiratory findings High-arched narrow palate Macroglossia Micrognathia Increased susceptibility to respiratory infections Subglottic stenosis Postextubation stridor Upper airway obstruction, sleep apnea GI findings Dental abnormalities Duodenal obstruction Gastroesophageal reflux Hirschprung disease CNS findings Mental retardation Epilepsy Strabismus Musculoskeletal findings Hypotonia Hyperextensibility or flexibility Dysplastic pelvis Atlantoaxial subluxation Immune system findings Immunosuppression Leukemia (acute lymphoblastic, acute myeloid forms) Hematologic findings Neonatal polycythemia Endocrine findings Low circulating level of catecholamine Hypothyroidism

1983). Some ­conclusions can be drawn from the published case reports summarized in the AAP Committee on Sports Medicine Subject Review and the recommendations cited above. The preoperative history and physical examination should include a careful search for evidence of cervical instability. Of course, the parent should be questioned about past cervical x-rays, if they were taken. The family and patient should be questioned about the occurrence of any type of neck pain, limitation of neck mobility, tortocollis, head tilt, abnormalities of gait, or other signs of upper motor neuron dysfunction. The examination should also look for spasticity, hyperreflexia, extensor­plantar reflex, or clonus. If the history and physical examination reveal problems or the cervical radiographs show an atlantodens interval of greater than 5 mm, the child’s elective surgery should be delayed and neurosurgical consultation sought (Hata and Todd, 2005). If the child had previous negative radiographs and the history and physical do not suggest a problem, it is not clear whether the films should be repeated (Pueschel,  1998).

1174   P a r t  IV    Associated Problems in Pediatric Anesthesia

Although the incidence of worsening of the atlantodens interval over time is low, some patients do show progression. Even in asymptomatic patients, the possibility exists that a postoperative neurologic disability may occur (Williams et al., 1987; Litman and Perkins, 1994). Whatever the result of the evaluations, if surgery and anesthesia are undertaken, the general consensus is that the heads and necks of these patients should be kept in the neutral position throughout the perioperative period. A reassuring study of children with Down syndrome with normal cervical radiographs who underwent tonsillectomy and adenoidectomy in the usual position showed no changes in the latency or amplitude of the somatosensory potentials (Abramson et al., 1995). Atlantoaxial instability occurs in conditions other than Down syndrome and at a higher rate (Box 36-20).

GENETIC MUSCLE DISORDERS The last decade has seen a tremendous increase in genetic and genomics research, which has allowed researchers to reach a clearer picture of the molecular causes of genetic-based muscle diseases. In most cases, identification of the proteins altered in the disease states has added to the understanding of the development of muscle and the neuromuscular junction, and the anesthetic management suggested for these diseases and syndromes is beginning to be modified as a result of this knowledge. This section contains a review of what is presently known concerning the molecular nature of diseases of the neuromuscular system and their clinical presentations. A general overview of the generation of muscle contraction in a normal cell (Fig. 36-26) is followed by a discussion of recent molecular and genetic studies that involve the major muscle diseases, because knowledge of their results may be of use to the clinician in the near future.

General Overview The genetic muscle diseases can be divided into 4 broad categories—muscular dystrophies, myotonic syndromes, mitochondrial myopathies, and myasthenic syndromes. To these,

Box 36-20 Conditions Associated with Atlantoaxial Dislocation Congenital abnormalities Trisomy 21 Klippel-Feil syndrome Larsen syndrome Mucopolysaccharidoses Spondyloepiphyseal dysplasia Metatropic dwarfism Kniest syndrome Chondodysplasia punctata 22q syndrome Infection Pharyngeal Tumors Trauma Postoperative complications (especially after airway surgery) Arthritis Rheumatoid Ankylosing spondylitis

Axon

MUSCLE CELL Transverse tubules

Action potential

Synapse Actin-myosin filaments Ca++

Sarcoplasmic reticulum

Ca++

Mitochondria

n  FIGURE 36-26. Schematic diagram of the muscle cell and motor neuron. An action potential (AP), generated in the membrane of the muscle by sodium influx through voltagegated sodium channels, travels down the motor neuron, causing the release of acetylcholine at the synapse. It travels along the membrane and down the transverse tubules. There it causes the influx of calcium at the base of the tubules (through the dihydropyridine receptor). The small calcium currents trigger release of larger amounts of calcium (through the ryanodine receptor) from the sarcoplasmic reticulum into the matrix of the muscle cell, activating the actin-myosin filaments to contract. The filaments attach (not shown) to the surface of the cell and the extracellular matrix to cause effective mechanical contraction (see Fig. 36-27, D).

malignant hyperthermia (MH) must be added as a separate category, though some overlap between MH and the other diseases does exist (see Chapter 37, Malignant Hyperthermia). Of course, a few muscle diseases do not easily fit into these groupings. Each of these categories is discussed separately and related to the MH syndrome where possible. The general locations in the muscle cell of the molecular changes that lead to these classes of disease are shown in Figure 36-27. The syndromes are caused by changes listed below: l The myasthenic syndromes affect transmission of the

action potential (AP) from the motor neuron to the muscle cell; this involves a disruption of the signal carried by the neurotransmitter, acetylcholine, at the ­synapse (see Fig. 36-27, A). l Myotonic syndromes affect transmission of the AP along the muscle membrane and are caused by abnormalities in sodium, chloride, potassium, or calcium channels (see Fig. 36-27, B). These changes cause a prolonged depolarization of the muscle membrane, which leads to prolonged contraction of the muscle. l Mitochondrial myopathies are (as their name implies) caused by abnormalities in mitochondrial function. Because mitochondria are important for supplying ATP in most tissues (most importantly nerve and muscle), the symptoms often involve the nervous system and muscle (see Fig. 36-27, C). The lack of ATP in muscle leads primarily to weakness and wasting of muscle. Mitochondria are also important in triggering cell death, or apoptosis; therefore, mitochondrial diseases may also lead to muscle wasting via this mechanism.

C h a p t e r 36    Systemic Disorders   1175

Axon

Action potential

MUSCLE CELL

Axon MUSCLE CELL Transverse tubules

Ca++

Sarcoplasmic reticulum Ca++

A

Ca++

Sarcoplasmic reticulum

Ca++

B Axon

Axon MUSCLE CELL

MUSCLE CELL

Actin-myosin filaments

Ca++

C

Sarcoplasmic reticulum

Ca++

Mitochondria

Ca++

Sarcoplasmic reticulum Ca++ Dystrophin

D

n  FIGURE 36-27. Detail of Figure 36-26 that shows the main areas of defects that lead to muscle disease. A, Disruption of the signal across the synapse leads to myasthenic syndromes. B, Defects in the calcium, potassium, and sodium channels in the muscle cell membrane give rise to myotonias. The membranes remain depolarized too long, causing the inability to relax. C, Mitochondrial dysfunction leads to decreased intracellular levels of ATP and to cell death. The decreased ATP concentration is responsible for the inability of the muscles to contract strongly and for insufficient reuptake of calcium into the sarcoplasmic reticulum. The latter effect may lead to inability to relax under certain circumstances. D, Defects in the attachment of the actin-myosin filaments to the cell surface and extracellular matrix produce muscular dystrophies. These attachments are important for mechanical force and for organizing and stabilizing the membrane.

l Muscular dystrophies result from the dissociation of con-

tractile force from the muscle to the surrounding connective tissue. The actin-myosin filaments contract, but they are no longer connected well to the cell membrane or the surrounding tissue. As a result, the equivalent of electromechanical dissociation occurs; i.e., the electrical signal is not translated into effective mechanical force (see Fig. 36-27, D). In addition, the membrane defects cause instability in the cell membrane integrity that may result in deleterious responses to anesthetic agents. The dotted circles in each panel of Figure 36-27 indicate the locations of the molecular changes. It is important to note that this figure is an oversimplification, but it gives the general pattern of the molecular causes of the syndromes. Skeletal muscle contraction is accomplished by the generation of a neuronal AP that terminates at the neuromuscular synapse (see Fig. 36-26). The neuronal AP stimulates sodium channels in the neuronal axon that propagate the signal along the axon. As the AP reaches the end of the axon, voltage-gated calcium channels are activated that allow the influx of calcium

into the neuron. This influx of calcium, in turn, stimulates the release of a neurotransmitter, acetylcholine, from the nerve terminal into the synapse. The acetylcholine binds to receptors on the cell surface of the postsynaptic cell—the muscle, in this case. Binding of the acetylcholine to its receptors allows influx of sodium into the muscle and generates a new AP, which propagates a transmembrane signal that spreads along the membrane of the cell. The AP is carried from the cell surface into the interior of the cell by a series of invaginations of the cell membrane known as T tubules. These structures allow for transmembrane electrical depolarizations to be carried deeply within the cell, where they would otherwise not be generated. At the ends of the T tubules, the sodium currents are again replaced by calcium currents, resulting from the activation of a voltage-gated calcium channel known as the dihydropyridine receptor (Fig. 36-28). These calcium currents, in their turn, stimulate larger calcium release from the sarcoplasmic reticulum through a calcium-sensitive calcium channel called the ryanodine receptor. These larger fluxes of calcium stimulate movement of the

1176   P a r t  IV    Associated Problems in Pediatric Anesthesia Na+, Ca++ channels Cell membrane Ca++

Sarcoplasmic reticulum Ca++

Ryanodine receptor

Dihydropyridine receptor

Actin-myosin contraction n  FIGURE 36-28. Further detail of the region of the T tubules. Sodium, potassium, and calcium channels on the cell surface are responsible for propagation of the AP along the cell membrane and into the T tubules. When the AP reaches the terminus of the T tubule, the sodium currents are replaced by calcium currents through the voltage-gated dihydropyridine receptor. These calcium currents trigger release of calcium through the ryanodine receptor from the large calcium stores in the sarcoplasmic reticulum.

actin-myosin filaments, an ATP-requiring step (and therefore dependent on functioning mitochondria). The filaments are attached to the surface of the muscle and the surrounding matrix through a variety of proteins, most notably, dystrophin. Movement of the filaments is transduced into shortening of the cell (muscle contraction) by the connection to the cell surface and surrounding matrix. Relaxation is accomplished by reuptake of the intracellular calcium primarily back into the sarcoplasmic reticulum. This reuptake is energy requiring and dependent on mitochondrial function, ATP generation, and ATP-dependent calcium pumps. Loss of this energy source is the cause of rigor mortis. The normal flow of electrical signal transduced to mechanical force can be disrupted at many places. Anesthesiologists often inhibit the transmission of the signal across the neuromuscular junction with the use of neuromuscular blockers (NMBs) such as vecuronium. Such an effect is conceptually similar to a myasthenic syndrome. Local anesthetics applied to the motor neuron inhibit the propagation of an AP along the neuron, resulting in decreased release of neurotransmitters. In this manner, they also mimic the effect of myasthenia. The use of local anesthetics directly on muscle blocks voltage-gated sodium channels in the muscle membrane, with the resulting inhibition of AP propagation, acting in the opposite manner to the changes seen in myotonic syndromes. Volatile anesthetics are also inhibitors of the voltage-gated membrane channels (e.g., sodium, potassium, and calcium) and thus also act in a manner opposite to that seen in myotonia. However, these drugs also inhibit mitochondria and are capable of causing a relaxation effect in a manner similar to a mitochondrial myopathy. These examples are given only to further acquaint the anesthetist with the underlying causes of the myopathies; drugs certainly do not cause these diseases. However, these similar effects help explain the interaction of the drugs with the disease states.

Myasthenic Syndromes Myasthenic syndromes are the result of the failure of transmission of the signal from the terminal of a motor neuron to the muscle innervated by the neuron. Most myasthenic syndromes are the result of immune responses against components of the neuromuscular junction (primarily the postsynaptic acetylcholine receptors) and are not classic genetic diseases. The symptoms result from decreased neurotransmission across the neuromuscular junction, and task-specific fatigue is the hallmark of these diseases (see Fig. 36-27, A). The well-known disease myasthenia gravis (MG) is an example of such a disorder, although it is primarily a disease of adulthood. MG can occur in the neonatal period because of placental transfer of maternal antibodies. In addition, juvenile-onset MG is seen in association with thymoma (Kiran et al., 2000). Congenital myasthenic syndromes (CMSs) are genetic in origin, and analyses have identified more than ten different genes that are important in development of the neuromuscular synapse (Palace and Beeson, 2008). Rarely, inherited disorders of neuromuscular transmission (CMSs), can result from acetylcholine receptor mutations or other mutations that involve the release of acetylcholine. Included in this group are familial infantile myasthenia, familial limb-girdle myasthenia, endplate acetylcholinesterase deficiency, and syndromes with altered or deficient acetylcholine receptors (Menold et al., 1998; Maselli et al., 2001). In addition, CMS may result from defects in development of the synapse. One such form of congenital myasthenia is caused by a defect in the gene DOK7, which is important in the formation or maintenance of synaptic structure (Müller et al., 2007). These genetic diseases mimic MG in their presentation and implications for anesthesia, and they manifest during infancy or childhood (Dalal et al., 1972).

Anesthetic Considerations for Myasthenic Syndromes The primary concern during the perioperative period for patients with myasthenic syndromes is to avoid respiratory compromise from weakened respiratory muscles or upper airway muscles (Brown et al., 1990; Abel and Eisenkraft, 2002). For this reason, nondepolarizing muscle relaxants are used sparingly, if at all, in these patients. As a result of the blockade and destruction of acetylcholine receptors, patients with MG or myasthenic syndromes are often resistant to succinylcholine (Baraka, 2001). It is important to remember that patients can appear strong on awakening only to become fatigued later in the recovery period. Itoh and others (2002) showed that seronegativity for the antiacetylcholine receptor antibody did not predict a normal response to muscle relaxants. In particular, patients with MG who were successfully treated with thymectomy may also retain a high sensitivity to muscle relaxants. The conclusion of these reports is that the anesthesiologist must presume a high sensitivity to muscle relaxants in all patients with myasthenic syndromes, even if they are functioning well after medical or surgical treatment. Techniques that employ a variety of short-acting anesthetics without the addition of muscle relaxants have been very successful (Della Rocca et al., 2003; Bouaggad et al., 2005). Others have reported encouraging results after use of regional anesthesia in these patients (Caliskan et al., 2008). However, one report cautions that even with a stable anesthetic, tourniquet release may trigger an exacerbation of symptoms (Brodsky and Smith, 2007).

C h a p t e r 36    Systemic Disorders   1177

Myotonias Myotonia is a temporary involuntary contraction of muscle fibers caused by transient hyperexcitability of the surface membrane (Miller, 1989; Bernard and Shevell, 2008). The persistent contracture of the skeletal muscle generally occurs after muscle stimulation but may be triggered by other stimuli such as cold, pain, or stress. A classic finding in patients with myotonia is the inability to easily relax after a firm handshake. Myotonias can be subdivided into 2 general groups — dystrophic and nondystrophic. The dystrophic group (represented by myotonic dystrophy [DM]) shows a progressive wasting of muscle mass and strength; the nondystrophic group does not show such progressive changes. In general, the myotonias may be thought of as a family of channelopathies that mostly affect muscle (Jurkat-Rott et al., 2002; Rosenbaum and Miller, 2002; Bernard and Shevell, 2008). The abnormalities in the channels lead to prolonged depolarization in the membrane once an AP is generated (Fig. 36-28). This in turn leads to prolonged or increased release of calcium into the cell, resulting in prolonged contraction. Two forms of myotonia (myotonia congenita and Becker disease) result from defects in the same skeletal muscle chloride channel (termed CLC1) (Jurkat-Rott et al., 2002; Pusch, 2002; Renner and Ptacek, 2002). Myotonia congenita (Thomsen disease) is an autosomaldominant disease that manifests in childhood and is associated with a normal life expectancy and minimal symptoms (Grunnet et al., 2003). Becker disease, not to be confused with Becker muscular dystrophy (BMD), is an autosomal recessive form of this channelopathy that also appears in childhood (Pusch, 2002). In addition, some mutations in this chloride channel cause a variant of dominant myotonia with a milder phenotype, myotonia levior (Ryan et al., 2002; Farbu et al., 2003). These myotonic diseases are nonprogressive and do not have a dystrophic component (i.e., the muscle does not deteriorate over time). Other less severe myotonias result from abnormalities in sodium or potassium channels on the muscle cell membrane. These include paramytonia congenita (sodium channel), hyperkalemic periodic paralysis (sodium channel), and hypokalemic periodic paralysis (calcium, sodium, or potassium channel) (Jurkat-Rott et al., 2002). As noted above, myotonic contractions may be precipitated by stress, cold, and pain. Thus, these triggering factors must be aggressively avoided in this population during the perioperative period. Regional anesthesia and NMBs do not reverse the contractions, because they act upstream from the molecular causes of the syndrome (compare Fig. 36-27, A and B). However, succinylcholine has been noted to precipitate contractions, as have anticholinesterases when used for NMB reversal. These contractions have been most notable in the occurrence of masseter spasm after the use of succinylcholine but can also involve other muscles and lead to extreme difficulty with positive pressure ventilation and intubation (Farbu et al., 2003). For these reasons, the use of succinylcholine is discouraged in patients with myotonia. If an episode of myotonia occurs during anesthesia, volatile anesthetics, quinine, or procainamide can be used for relaxation. If a myotonic episode occurs in patients with periodic paralysis, use of the carbonic anhydrase inhibitor dichlor­ phenamide has been shown to be useful (Cleland and Griggs, 2008). In that the myotonias occur as the result of abnormal ion channels, great care must be taken to keep electrolytes normal at all times. Whereas myotonic syndromes may have symptoms

in common with MH, they are not associated with true MH (see Chapter 37, Malignant Hyperthermia). Steinert muscular dystrophy (DM) is the most common form of myotonia (Anderson and Brown, 1989). This disease is a form of muscular dystrophy and includes congenital DM. It is discussed here rather than with other muscular dystrophies, because its presentation is different—it more closely resembles the myotonias. DM was shown to actually include two different molecular diseases (Ranum and Day, 2002). DM type 1 (DM1) results from alterations in the human dystrophia myotonica-protein kinase gene (DMPK), leading to an increase in unstable CTG repeats in the 3′ untranslated region of the DMPK gene (Amack and Mahadevan, 2004; Botta et al., 2008; Orengo et al., 2008). DMPK codes for a serine-threonine protein kinase. The changes in DMPK lead to abnormal splicing of the message for the calcium channel protein CLCN1, with resulting defects in function of the channel (Lueck et al., 2007). The myotonia may then result from defects in calcium channels, resulting in altered transmembrane potentials in muscle (Kaliman and Llagostera, 2008). However, others have suggested that the myotonia is probably the result of abnormal phosphorylation of sodium channels, resulting in delayed inactivation after channel opening (Lee et al., 2003). The precise mechanisms by which this mutation causes the changes in muscle function are not yet clear, but they appear to involve several pathways (Dansithong et al., 2008). As noted above, the changes in the protein kinase gene are in the promoter or starting region of the gene and are the result of duplications in short repetitive sequences (CTG triplets). The number of repetitive sequences is often increased in the offspring compared with an affected parent. As a result, each successive generation tends to exhibit a more severe form of the disease. DM type 2, or DM2, has a clinically diverse presentation, including myotonia, proximal muscle wasting, and endocrine, cardiac, and cerebral abnormalities. It results from expansion of a similar sequence in an intron of a gene; however, the gene is separate from the one that causes DM1 and codes for a probable transcription factor, ZFN9 (Liquori et al., 2001; Finsterer, 2002; Liquori et al., 2003). The precise physiologic changes that lead to myotonic or dystrophic changes are not known. In both disease states, the abnormalities result from abnormal RNA species that disrupt normal development of the cells (Mankodi and Thornton, 2002).

Anesthetic Considerations for Myotonias Attainment of muscle relaxation can be difficult in these patients. As with the other forms of myotonia discussed previously, cold, stress, pain, and succinylcholine can precipitate myotonia. Additionally, because this is a dystrophy with muscle wasting, succinylcholine can elicit a hyperkalemic response and should be avoided. Unlike the other myotonias, DM leads to deterioration of the muscle fibers and is associated with weakness and hypotonia in the infant and child. Paradoxically, however, these patients can still experience a myotonic episode as well. They can have profound respiratory depression, severe cardiac conduction abnormalities, cardiomyopathy, developmental delay, dysphagia, and decreased gastric motility. Muscle relaxants must be used with great care, if at all, in these patients. Smaller doses are probably indicated, and an NMB monitor is advised. As with nondystrophic myotonias, reversal agents may induce myotonic episodes.

1178   P a r t  IV    Associated Problems in Pediatric Anesthesia

Because respiratory muscle weakness is notable in these patients, the respiratory status is potentially fragile when any narcotic or general anesthetic is used. Thus, their care presents challenges involving several physiologic systems. White and Bass (2003) presented a thoughtful review of the anesthetic care of patients with DM. DM is also commonly thought to be associated with MH. However, as with the myotonias discussed above, although this syndrome shares features with MH, it is not associated with true MH (see Chapter 37, Malignant Hyperthermia).

mitochondrial inner membrane and electrons are donated to oxygen to generate water. The proton gradient is then used to drive an ATP synthase (complex V). The coupling of electron transfer to phosphorylation is known as oxidative phosphorylation and is overwhelmingly the major source of ATP and other high-energy phosphate bonds that supply energy to the cell. Mitochondrial complexes are composed of groups of proteins, ranging from just a few (complex II) to over 40 (complex  I). In addition, the dehydrogenases, membrane transporters, and structural proteins raise the number of functional proteins in the mitochondria into the hundreds. Genes in the cell’s chromosomes encode for most of the proteins of the ETC, whereas mitochondrial DNA encodes a minority of the ETC proteins. The other enzymes in the mitochondria are entirely encoded by the nuclear genome. The genetics of mitochondrial disease are complicated by the fact that mitochondria are inherited from the mother. However, different populations of the maternal mitochondria may be passed to different offspring, so the inheritance pattern can be quite varied. Finally, mitochondrial dysfunction has effects other than energy depletion. Increased free radical damage to other cellular components and alterations in protein phosphorylation may be seen with mitochondrial disease. Each of these effects can give rise to mixed but wide-ranging functional changes in affected individuals. It is a common mistake to group all mitochondrial diseases together as similar entities; however, mutations in any of the mitochondrial proteins may result in dramatically different functional changes. In addition, mitochondria in different tissues can be quite varied in their activity. The differences between tissues in sensitivity to abnormalities of mitochondrial function (and the varied inheritance pattern discussed previously) give rise to different symptoms even within members of a family who carry the identical mutation. Because of this variability, it is dangerous to imply that because an anesthetic

Mitochondrial Myopathies Mitochondrial dysfunction is being recognized as the cause of an increasingly large list of disease syndromes. The more commonly seen mitochondrial syndromes are Leigh disease, Kearns-Sayre syndrome, and Leber hereditary optic neurop­ athy. However, mitochondrial dysfunction is also associated with unnamed myopathies and encephalopathies and with symptoms of failure to thrive. Mitochondrial abnormalities have been shown to be involved with some forms of autism and Parkinson’s disease (Shoffner et al., 1991; Schaefer et al., 2004). It is clear that the presentation of mitochondrial disease may be quite varied. Mitochondria are the principal source of energy metabolism within cells, especially those of nerve and muscle (see Fig. 36-27, C). Within mitochondria reside the enzymes responsible for the Krebs cycle, fatty acid β oxidation, and most importantly, oxidative phosphorylation (Fig. 36-29). Mitochondria contain the enzymes that metabolize glucose, fatty acids, and amino acids to generate NADH and succinate, which in turn, are used as electron donors for the electron transport chain (ETC). By passing electrons down the ETC (complexes I to IV), a proton gradient is generated across the Oxaloacetate α ketoglutarate Matrix

Acetyl-CoA H+

NAD+

ATP

NADH Fumarate ATP synthetase

I Inner membrane Malate Glutamate Succinate Pyruvate

Intermembrane space

II

CoQ

III

IV H2O O2

Cytochrome

H+

H+

DHQ

H+

H+

Pi

ADP ATP

TMPD + ascorbate

n  FIGURE 36-29. ETC of mitochondria. Substrates for the ETC are transported into the mitochondria (light blue arrow) and NADH, or succinate, is generated via the Krebs cycle. NADH donates electrons to the ETC at complex I; they then follow the path indicated by the dark blue arrows. Succinate donates electrons to complex II; they follow the path indicated by the red arrows. Once the electrons reach complex III, their paths are functionally the same (gray arrow). Protons are pumped into the intermembrane space (green arrows) to generate a transmembrane gradient. The protons then leak back into the matrix through complex V to generate the energy to drive the ATPase (red oval). DHQ and TMPD/ascorbate are indicated and are artificial electron donors that can be used experimentally to drive complexes III and IV, respectively. DHQ, Dihydroquinidine; TMPD, tetramethyl-p phenylenediamine.

C h a p t e r 36    Systemic Disorders   1179

technique was successful in a few patients with mitochondrial disease, the technique is safe for all patients with mitochondrial dysfunction. Muscle and nerve cells are uniquely dependent on the energy delivered by the mitochondria. Mutations in mitochondrial proteins are responsible for striking clinical features in those two tissues, including myopathy, cardiomyopathy, encephalop­athy, seizures, and cerebellar ataxia. Of course, cardiac muscle and the CNS are also the two main targets of general anesthetics. Thus, particular care must be taken when exposing such a patient to these agents. As motor neurons may be affected, a hyperkalemic response to succinylcholine may be seen. Lastly, MH is thought to be associated with some forms of mitochondrial myopathies, but the nature of this relationship is unclear (Keyes et al., 1996; Fricker et al., 2002).

Anesthetic Considerations for Mitochondrial Myopathies The perioperative period is a time during which a patient may be exposed to periods of stress. Under conditions of stress, ATP levels may be inadequate to meet demand. Shivering caused by hypothermia probably represents the greatest threat to these patients. However, hyperthermia and stress from untreated pain also represent serious risks. The failure of ATP production to meet metabolic demands inevitably leads to lactic acidosis, often of profound significance. To avoid such problems, great care must be taken to keep patients normothermic during surgical cases and to provide adequate treatment for postoperative pain. Postoperative pain represents a particularly troublesome problem, because narcotics can further compromise respiratory status. Mitochondrial patients may become acidotic, caused by high levels of lactate as a result of hypovolemia. Prolonged preoperative fasting should be avoided in these patients. If fasting is necessary, IV fluids should be started with glucose added to avoid anaerobic metabolism. As cyanide inhibits the respiratory chain, sodium nitroprusside should probably be avoided. For similar reasons, tourniquets should be avoided if possible. It is not clear what hematocrit value is adequate in these patients, but it is probably wise to maintain hematocrit at closer to normal levels than at the reduced levels allowed in other patients. Lastly, whereas mild levels of hypotension are commonly used in many patients to avoid blood loss, such an approach is less desirable in patients with mitochondrial disease. These patients are probably less able to compensate for decreased oxygen delivery. Essentially, every general anesthetic studied has been shown to depress mitochondrial function. The most notable of these are the volatile anesthetics and propofol. It is often said that these agents only depress mitochondria at doses higher than their clinical concentrations. However, Miro and colleagues (1999) showed that even at doses commonly used in the operating room, anesthetics cause a significant depression of mitochondria in normal patients. Studies in model organisms have shown that when complex I is abnormal, sensitivity to volatile anesthetics is markedly increased (Kayser et al., 1999). Case reports have also indicated that some children exhibit an increased sensitivity to sevoflurane (Morgan et al., 2002). In addition, a strong clinical impression exists that children with mitochondrial myopathies have an increased risk during surgery (Morgan et al., 2002; Bolton et al., 2003; Farag et al., 2005). Because metabolism is altered in patients with mitochondrial

disease, the abilities of the cell to generate ATP and to effectively use oxygen are diminished, and exposure to anesthetics may represent an increased risk compared with other patients (Bolton et al., 2003). In contrast, others have found in retrospective studies that patients with mitochondrial defects did not appear to have an increased rate of perioperative complications (Driessen et al., 2007). Clearly, a prospective study is warranted to resolve these concerns. The use of regional anesthetics should be considered if appropriate for the case. However, it has also been noted that mitochondria are the probable target for the cardiac complications of bupivacaine; thus, patients with mitochondrial myopathies may be at increased risk with this drug as well (Weinberg et al., 2000). From an anesthesiologist’s point of view, the primary complications of mitochondrial myopathies include respiratory failure, impairment of myocardial function, conduction defects, and dysphagia. Each of the volatile anesthetics depresses respiration, though to varying degrees. Isoflurane and desflurane depress the ventilatory response to CO2 more than does sevoflurane. In addition, isoflurane and desflurane cause more direct muscle relaxation. Thus, from this standpoint, sevoflurane would seem to be advantageous. However, isoflurane and desflurane are noted for their ability to maintain cardiac output to a greater degree than does sevoflurane (Weiskopf, 1995; Lowe et al., 1996). In short, each of the volatile anesthetics presently in use is capable of interacting negatively with a mitochondrial myopathy. Fortunately, patients may be supported during the perioperative period to avoid the theoretic or real side effects of these drugs. The volatile anesthetics do not require metabolism for excretion; rather, they are removed by ventilation, thus representing an advantage over IV anesthetics, which are dependent on energy-requiring metabolism. Lastly, the respiratory status of these patients (as with all patients with myopathies) must not be stressed, even when not exposed to anesthetics. Great care must be exercised while weaning them from ventilatory support to ensure adequate spontaneous ventilation. During the past decade, the IV anesthetic propofol has become increasingly popular as a maintenance anesthetic. However, it has many of the same side effects as volatile anesthetics and is a profound inhibitor of several aspects of mitochondrial function. One notable exception is that it is not known to cause much muscle relaxation. However, it is quite capable of decreasing ventilatory drive as well as cardiac output and contractility. Although it is viewed as a short-acting drug, its ultimate elimination is metabolism dependent. Theoretic concern exists that patients with mitochondrial myopathy may have an increased risk of developing propofol-infusion syndrome during prolonged exposure. However, propofol has been used successfully both as an induction agent and as an infusion and seems to have little effect in this setting (Driessen, 2008). All of the general anesthetic agents are known to directly inhibit mitochondrial function and may add to preoperative problems. However, each of the anesthetics already discussed has been used successfully when caring for patients with mitochondrial disease. It may be that, as the various types of mitochondrial disease are better defined, anesthetic preferences in certain cases may become clear. However, such a recommendation cannot be made at the present time. What is clear is that these patients must be monitored more closely than other patients when a general anesthetic is used and that great care must be exercised to document that the effects of the anesthetics are largely gone before assuming that the patient can ventilate adequately.

1180   P a r t  IV    Associated Problems in Pediatric Anesthesia

Patients with mitochondrial myopathies respond well to the nondepolarizing relaxants, although they probably have an increased sensitivity. These drugs should be titrated to the desired effect and monitored closely by nerve stimulation. As noted for general anesthetics, great care must be exercised to ensure that these patients are fully reversed before removing them from ventilatory support. As motor neurons may be affected, a hyperkalemic response to succinylcholine may be seen. This drug should only be used when absolutely necessary in these patients. Succinylcholine and volatile anesthetics are also known as triggers for MH. MH is thought to be associated with some forms of mitochondrial myopathies, but the nature of this relationship is unclear. It may be that a group of mitochondrial myopathies mimics MH by having inadequate ATP for reuptake of calcium from the cytoplasm of muscle cells into the sarcoplasmic reticulum. Such a failure could cause prolonged muscle contraction and lead to increased metabolism. Clearly, this represents a complicated problem in these patients. Whether they increase their temperature or become acidemic depends on the nature and severity of their disease. In general, regional anesthesia is well tolerated by patients with mitochondrial myopathies. This tolerance is in spite of data that indicate that local anesthetics, especially bupivacaine, are capable of potently depressing mitochondrial function. Although most neural blockade uses doses of anesthetic that are well below the doses necessary for mitochondrial effects, it is important to be aware that these drugs are also capable of such effects, especially on the heart. Despite this risk, however, because of the low doses commonly used, regional anesthesia represents a valuable mode of treatment. Ideally, such blocks avoid exposure of the CNS and cardiac muscle to potentially toxic side effects. Not all cases can be completed solely with regional anesthesia, especially in children. However, consideration should be given to this approach when possible, owing to the benefits of effective postoperative analgesia without sedation or respiratory compromise. It is important to remember that mitochondria are inhibited by local anesthetics. If a muscle biopsy is being performed for the diagnosis of mitochondrial disease, the muscle itself must not be exposed to the local anesthetic. Such exposure would likely lead to a false defect in mitochondrial function. A similar concern can be raised for the use of propofol in muscle biopsy cases, although it would be expected that the concentration of propofol in the muscle should be quite low. At the present time, there is no perfect anesthetic for these patients. When possible, consideration should be given to the use of local anesthetics in small amounts for a regional anesthetic. When a general anesthetic is necessary, probably each of the general anesthetics in use has its place. At present it is not possible to entirely eliminate one group as less safe than others. What is clear is that these patients must be monitored more closely than other patients. The use CNS monitors to gauge their depth of anesthesia more closely may allow anesthesiologists to expose these patients only to the minimum amount of drug necessary to carry out the surgical procedure.

however, many of their implications for anesthesiologists are similar (Kerr et al., 2001; Schmidt et al., 2003). Duchenne’s muscular dystrophy (DMD) is an X-linked disorder that results from deletion mutations in the dystrophin gene and leads to a complete lack of dystrophin in skeletal muscles (see Fig. 36-27, D). The defect is present in approximately 1 in 3500 live births, with the onset of disease often before school age and progressing to wheelchair dependence by the second decade of life. Dystrophin is a large protein that helps to anchor the contractile components (the actin myosin filaments) to the cell membrane and indirectly to the surrounding extracellular matrix (Fig. 36-30). Loss of this protein leads to profound muscle weakness and then respiratory failure, cardiomyopathy, cardiac conduction defects, and occasionally, mild mental retardation (Finsterer and Stollberger, 2003; Muntoni et al., 2003). Other, less global changes in this same gene cause Becker Muscular Dystrophy (BMD) and the related disease, X-linked dilated cardiomyopathy. Cardiomyopathy is occasionally seen in female carriers of the mutation. The clinical presentation of patients with various forms of muscular dystrophy are known to most anesthesiologists and are not reviewed in detail here other than as affects their anesthetic implications.

Muscular Dystrophy

n  FIGURE 36-30. Dystrophy-related proteins; attachments of the actin filaments to the cell membrane and extracellular matrix. Disruption of these attachments alters protein distribution in the membrane and membrane stability. Some of the specific proteins (e.g., laminin, sarcoglycan, and dystrophin) that cause dystrophies are shown. ACh, Acetylcholine.

At least five forms of muscular dystrophy are clinically relevant for anesthesiologists: Becker, Duchenne, facioscapulohumeral, Emery-Dreifuss, and limb-girdle muscular dystrophies (Farrell, 1994). These entities vary greatly in severity of presentation;

Anesthetic Considerations for Muscular Dystrophy The approach to patients with DMD or BMD has changed in recent years. The main anesthetic implications of DMD and BMD are related to the severity of these profound myopathies. As would be expected in patients with muscle weakness, significant respiratory insufficiency can occur postoperatively with either disease. Cardiac muscle and conduction are also involved, and drugs that further depress cardiac function or that increase the likelihood of dysrhythmias should be avoided. All patients with DMD or BMD should receive a full cardiology evaluation with echocardiography within 3 months of surgery and pulmonary function testing before any surgery. Lastly, dysphagia is common, and gastric motility may be decreased, requiring expeditious control of the airway. The specific history of each patient

DYSTROPHY-RELATED PROTEINS

Extra cellular matrix Laminincongenital dystrophy

Sarcoglycan– limb girdle dystrophy Cell membrane

Ach receptor Dystrophin – Duchenne’s/Becker’s Actin filaments

C h a p t e r 36    Systemic Disorders   1181

with DMD or BMD must be evaluated closely to determine the best technique to secure the airway. In addition to playing a key role in anchoring contractile components to the cell membrane and the surrounding extracellular matrix, dystrophin is also important in organizing the postsynaptic acetylcholine receptors (Muntoni et al., 2003). In its absence, abnormalities occur both in the types of receptors and in their number and location, which may lead to abnormal responses to depolarizing muscle relaxants. Patients with BMD or DMD can have rhabdomyolysis and hyperkalemia in response to succinylcholine; thus, succinylcholine is contraindicated for them. It should also be noted that congenital forms of muscular dystrophy exist that probably involve other proteins necessary for attaching the contractile machinery to the extracellular matrix (see Fig. 36-30). Although this is a heterogeneous group of mutations, it is probably best to treat these patients as if they had DMD. In the past few years, appreciation has increased concerning the risk of rhabdomyolysis and hyperkalemia in patients with DMD (and BMD) from exposure to volatile anesthetics alone. Hayes and colleagues (2008) present an important review of patients who had apparent hyperkalemic arrest after exposure to volatile anesthetics. These patients were felt to not have MH but to be still at risk for an untoward response to a volatile anesthetic. Although these events may be confused with MH, the association of DMD and BMD with true MH appears to be only coincidental. As a result, the authors strongly suggest using a “trigger-free” anesthetic in patients with known muscular dystrophy. This opinion is reinforced in a Consensus Statement from the American College of Chest Physicians (Birnkrant et al., 2007). However, at present no such recommendation has come from the ASA. The Association of Unusual Enzymopathies, Channelopathies, and Muscular Disorders with MH has been reviewed in a series of articles (Benca and Hogan, 2009; Davis and Brandom, 2009; Gurnaney et al., 2009; Hogan and Vladutiu, 2009; Klingler et al., 2009; Litman and Rosenberg, 2009; Parness et al., 2009).

Other Dystrophies The remaining dystrophies—facioscapulohumeral (FSH), Emery-Dreifuss, and limb-girdle muscular dystrophy (LG)— are much milder in their presentations (Emery, 2002). FSH is one of the most common dystrophies; it is the most benign and usually is associated with little respiratory involvement (Fitzsimons, 1999). However, the neck, face, and scapular stabilizing muscles are often weak, and the ability to raise the head may be of little use in determining respiratory muscle strength (Dresner and Ali, 1989). Defects in the gene FRG1 are associated with FSH and with disrupted muscle growth in vertebrates; however, the function of FRG1 is unknown at present (Hanel et al., 2009). Emery-Dreifuss muscular dystrophy usually has its onset in the teenage years and results from mutations in lamin (part of the nuclear matrix) or in emerin, an inner nuclear membrane protein that interacts with lamin and transcription regulators (Bione et al., 1994; Méjat et al., 2009). These patients have cardiac conduction defects, dilated cardiomyopathy, contractures (positioning problems), and often, fusion of C3 to C5, resulting in a less mobile neck (associated with difficult intubation) (Aldwinckle and Carr, 2002; Shende and Agarwal, 2002).

LG results from mutations in several proteins (at least 11 known), such as α-sarcoglycan, which associate with dystrophin. Some of these mutations are in proteins that overlap with those causing congenital muscular dystrophy. As in EmeryDreifuss muscular dystrophy, LG is associated with some respiratory muscle weakness and cardiac conduction abnormalities, and at least some cases result from a defect in a protein that interacts with muscle cell membrane and is implicated in membrane repair (Capanni et al., 2003). The literature contains little regarding experience with the interaction of anesthetics with LG dystrophy (Pash et al., 1996). In all three forms of muscular dystrophy, succinylcholine should be avoided, because hyperkalemia can result. MH is not reported in these three milder forms of muscular dystrophy. Thus, although not universal, the recurring themes with muscular dystrophy are to avoid succinylcholine, to watch for respiratory depression, and to avoid cardiac depressants and arrhythmogenic drugs. Volatile agents must be used with extreme caution because they are associated with rhabdomyolysis. It is important to note that anesthetic complications have been reported in most of these types of muscular dystrophy (Farrell, 1994). These events most commonly involve a hyperkalemic episode, and sudden cardiac arrest may occur. Such events can occur in patients who are still in a subclinical stage of their disease and in whom the crisis may be the first manifestation. For this reason, many clinicians reserve their use of succinylcholine in children to only those cases in which there exists a specific indication for its use.

Metabolic Diseases Several other genetic diseases of muscle exist that are of interest to anesthesiologists, including primary diseases of metabolism that lead to chronically weak muscles or muscles that are prone to damage when exposed to high metabolic stress. McArdle disease, an inherited disorder of muscle phosphorylase activity that down-regulates a Na+K+ membrane pump, is a prototypical disorder of this type (Clausen, 2003). Muscle from patients with McArdle disease can function normally until stressed by exercise or ischemia, at which time severely painful, electrically silent contracture develops, which can be followed by rhabdomyolysis, myoglobinuria, renal failure, and death. This sequence shares many elements with anesthesia-induced MH. Several MH-susceptible patients exhibited altered energy balance in response to caffeine, raising the possibility that disorders of intracellular energy production may contribute to the MH phenotype (Textor et al., 2003). Patients with other less common diseases, such as SchwartzJampel syndrome, King-Denborough syndrome, and Brody disease, are also prone to anesthetic complications. In each case, MH-like responses have been reported, but their exact relationship to MH is not clear. King-Denborough syndrome is a progressive myopathy in which patients have short stature, severe scoliosis, pectus deformities, ptosis, low-set ears, and cryptorchism (Heiman-Patterson et al., 1986). Anesthesia-related deaths, fulminant episodes of MH, and positive in vitro contracture test results have been commonly associated with this disorder (Isaacs and Badenhorst, 1992). Because this disease is so rare, information about its genetic etiology is still lacking. Patients with Brody disease (also known as Lambert-Brody disease) elicit a decreased ability to relax on repeated ­activation

1182   P a r t  IV    Associated Problems in Pediatric Anesthesia

of their muscle, which results from an identified defect of the sarcoplasmic reticulum Ca2+-ATPase (Odermatt et al., 2000). An in vitro contracture test performed on muscle from such a patient was abnormal, which may indicate that such patients may be at risk for MH-like syndrome (Froemming and Ohlendieck, 2001). Severe myotonia or neuromyotonia is a common feature of Schwartz-Jampel syndrome, in which hyperexcitability can be exacerbated by anesthetic agents and muscle relaxants are considered to be triggering agents for MH (Seay and Ziter, 1978; Ray and Rubin, 1994). It is unclear whether true MH is associated with this syndrome. In addition, difficult intubation because of neck and laryngeal anatomic abnormalities has been reported (Stephen and Beighton, 2002). Clearly, these studies present as many questions as they do answers. Precise recommendations for the anesthetic care of these patients are not available. However, in general, the anesthesiologist probably should avoid the use of succinylcholine and volatile anesthetics while otherwise delivering normal meticulous anesthetic care. At present, the use of nontriggering anesthetics seems safe for this population, although any muscle relaxant must be used with the utmost caution. These patients should have prolonged observation for prolonged or recurrent weakness in the postoperative period.

The Undiagnosed Myopathy In reviewing the above suggestions, an unpleasant scenario becomes apparent. What is the best anesthetic approach in the child with an undiagnosed myopathy? Depending on the molecular basis for the myopathy, the patient may turn out to have a diagnosis of mitochondrial defect or muscular dystrophy and may possibly be at risk for MH. The first possibility carries the recommendation to avoid propofol, and the second and third scenarios carry the recommendation to avoid volatile anesthetics (absolutely, if MH, and if possible, if ­muscular dystrophy). An insightful discussion of this conundrum was presented by Ross (2007) and Davis and Brandom (2009).

It  is clear that volatile agents should be avoided if MH is a serious possibility. In those cases in which MH is not a concern, the choice of anesthetics depends on the likelihood of a mitochondrial defect vs. muscular dystrophy. Two reports of anesthetic management of children with muscle weakness who underwent muscle biopsy found that the likelihood of untoward effects using either class of drugs was low, in the range of 1% or lower (Driessen et al., 2007; Flick et al., 2007). To improve this number, it is incumbent on the anesthesiologist consult, the referring neurologist, and the geneticist/ metabolism specialist to be aware of all aspects of the patient’s history to determine which disease group is more likely. Until further information is available to guide the anesthesiologist, it remains crucial to closely monitor these patients during any anesthetic exposure and to maintain the close observation for several hours postoperatively.

SUMMARY The anesthetic care of the pediatric patient with a systemic disorder provides myriad challenges for the anesthesiologist. A thorough understanding of the patient’s disease and the effect that the anesthetic will have on the disease process are the two most important issues the anesthesiologist must address. For some of these patients, appropriate consultation with the surgeon, pediatrician, and pediatric subspecialist is essential to the proper management of these potentially difficult and challenging anesthetics. For questions and answers on topics in this chapter, go to “Chapter Questions” at www.expertconsult.com.

R eferences Complete references used in this text can be found online at www.expertconsult.com.