- Email: [email protected]
Masking of Syndrome of Inappropriate Antidiuretic Hormone Secretion: The Isonatremic Syndrome
Masking of Syndrome of Inappropriate Antidiuretic Hormone Secretion: The Isonatremic Syndrome Irfan Khan, MD1, Bridget Zimmerman, PhD, MS2, Patrick Brophy, MD3, and Sameer Kamath, MD4 Objective To determine whether the administration of isotonic saline in patients undergoing spinal fusion surgery prevents the development of hyponatremia, thus masking the detection of syndrome of inappropriate antidiuretic hormone secretion (SIADH). Study design Prospective observational cohort study conducted in pediatric patients undergoing spinal fusion surgery. Using established criteria for diagnosing SIADH with the exception of serum sodium as a criterion, we separated patients into those with and without masked SIADH. Random cortisol levels were measured in the perioperative period to test for adrenal insufficiency to exclude it as a cause for natriuresis and hyponatremia. Results Of the 40 patients included in the study, 13 (32%; 95% CI, 19%-49%) met study criteria for masked SIADH. The serum sodium levels between the 2 groups were not different throughout the postoperative period. The antidiuretic hormone levels were increased at 24-48 hours after surgery (20.4 pg/mL in masked SIADH group vs 6.6 pg/mL in no masked SIADH group, P = .04). Subjects with masked SIADH demonstrated a tendency for weight gain (3.9 kg vs 2.5 kg, P = .058), which was maximal on postoperative day 2. Cortisol levels were similar between the groups. Conclusion Masked SIADH (SIADH-like state without hyponatremia) commonly occurs in the postoperative period in children and young adults undergoing spinal fusion surgery. Early postoperative evaluation and recognition may result in appropriate management of patient’s fluid balance. (J Pediatr 2014;165:722-6). See editorial, p 653
I
nappropriate antidiuretic hormone (ADH) secretion occurs in many clinical conditions and is designated as the syndrome of inappropriate ADH secretion (SIADH). It is a disorder of sodium (Na) and water balance characterized by impairment of urinary dilution and hypotonic hyponatremia in the absence of renal disease or other identifiable nonosmotic stimuli.1-3 The classic criteria used to diagnose SIADH are hyponatremia (serum Na <135 mEq/L), low serum osmolality (plasma osmolality <280 mOsm/kg), and high urine osmolality (urine osmolality >100 mOsm/kg).1 The incidence of SIADH after spinal fusion has ranged from 5% to 100%.1,2,4-6 The routine administration of isotonic saline at maintenance rates in the postoperative period after spinal fusion decreases the incidence of hyponatremia but does not result in hypernatremia.7-9 Most postoperative studies performed to evaluate the incidence of SIADH after spinal fusion surgery note the use of 0.45% normal saline (NS).2,4-6,10,11 All patients with SIADH in these studies had hyponatremia and fulfilled traditional criteria for diagnosis of this condition. We hypothesized that the routine use of isotonic saline in the postoperative period prevents the development of hyponatremia masking the detection of SIADH. We thus excluded hyponatremia as a diagnostic criterion for diagnosing a SIADH-like state.
Methods After approval by the institutional review board, we performed a prospective observational cohort study at the University of Iowa Children’s Hospital. Written informed consent was obtained from all parents, and assent was obtained from all children 12 years of age and older. Patients were excluded if they had preexisting renal, adrenal, pituitary, or thyroid disease. Patients on diuretics and/or with a history of congestive heart failure also were excluded.
ADH BNP FeNa Na NS SIADH UNa
Antidiuretic hormone Brain natriuretic peptide Fractional excretion of sodium Sodium Normal saline Syndrome of inappropriate antidiuretic hormone secretion Urine sodium
From the 1Division of Pediatric Critical Care, Presbyterian Hospital, Albuquerque, NM; 2Department of Biostatistics at the University of Iowa; 3Department of Pediatric Nephrology, and 4Division of Pediatric Critical Care, University of Iowa Children’s Hospital, Iowa City, IA Funded by Children’s Miracle Network Proposal #2231 and Pediatric Critical Care division at the University of Iowa Children’s Hospital. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2014.05.051
722
Vol. 165, No. 4 October 2014 Baseline weight (bed scale), vital signs, and laboratory testing were obtained (serum and urine electrolytes, osmolality, blood urea nitrogen, creatinine, serum albumin, urine-specific gravity, complete blood count, coagulation studies, venous blood gas, and random cortisol level). All patients received a combination of inhaled and intravenous anesthesia. All patients received NS, 5% albumin, or lactated Ringer solution in the operating room and 5% dextrose in NS at maintenance rates in the postoperative period. Pertinent data obtained from anesthesia records included medications, duration of hypotension, tachycardia, vital signs, ventilator data, vasopressor use, and antihypertensive use during surgery. Surgical data included the degree of spinal curvature, type of spinal fusion surgery, number of spinal segments fused, duration of surgery, and estimated blood loss. Body weight change, fluid intake and output, and the administration of blood products were noted. The laboratory tests in the postoperative period included serum and urine electrolytes, osmolality, blood urea nitrogen, creatinine, serum albumin, urine-specific gravity, complete blood count, coagulation studies, venous blood gas, random cortisol levels, and serum ADH levels. Patients were determined to have a SIADH-like state only if they met the following 7 criteria: serum osmolality <290 mOsm/kg, urine osmolality >600 mOsm/kg, fractional excretion of Na (FeNa) >0.5, urine output <0.5 mL/kg/h, urine Na (UNa) >30 mEq/L, serum bicarbonate >20 mEq/ L, serum potassium >3.5 mEq/L. We excluded serum Na <135 mEq/L as one of the criteria in making a diagnosis of “masked SIADH” to test our hypothesis. We separated the patients into 2 groups viz, masked SIADH and no masked SIADH based on the aforementioned criteria met any time during the first 72 hours of the postoperative period. Data were abstracted in the postoperative period at 6, 12, 24, 48, and 72 hours. Statistical Analyses The proportion (with 95% exact confidence limits) of patients that met the “masked SIADH” criteria within 72 hours after surgery was computed. The demographic and clinical
variables were compared between those with masked SIADH and those without masked SIADH with the use of Pearson c2 test or Fisher exact test for categorical variables, and 2-sample t test or Wilcoxon rank-sum test for the continuous variables. The comparison of fluid balance, fluid intake, and laboratory variables that were measured over time between subjects with masked SIADH and no masked SIADH was tested with the use of linear mixed model for repeated measures. The mixed model included masked SIADH status (with vs without), time, and masked SIADH*time interaction as the fixed effects. Specific comparisons of interest, such as between groups at each time, were tested using test of mean contrast with P values adjusted using Bonferroni method to account for the number of tests performed (ie, for test at each time point, adjusted P value = unadjusted P value 6 time points). P < .05 was considered statistically significant. All statistical analyses were performed using SAS (version 9.3; SAS Institute Inc, Cary, North Carolina).
Results In a period of 15 months, between July 2010 and October 2011, 47 patients met study criteria and 40 consented. Thirty-five patients underwent posterior spinal fusion and 5 underwent anteroposterior spinal fusion. Of all the study subjects, 13 (32%; 95% CI, 19%-49%) fit the definition of masked SIADH. The demographic and clinical variables for the 2 groups are listed in Table I. We found no statistically significant difference in the 2 groups (masked SIADH vs no masked SIADH). Patients in the masked SIADH group demonstrated weight gain, which was maximal on postoperative day 2 (3.9 kg vs 2.5 kg, P = .058). There was a trend towards increased pressor use in the operating room in the no masked SIADH group. Descriptive data of patients in each group are listed in Table II. Hemoglobin levels were similar between groups (P > .99). Mean postoperative hemoglobin at 0-6 hours was 9.54 0.39 g/dL for masked SIADH and 9.84 0.27 g/dL for no masked SIADH. At 48-72 hours, mean hemoglobin was 8.73 0.39 g/dL for masked SIADH and 9.37 0.27 g/ dL for no masked SIADH, with a mean difference of
Table I. Comparison of demographic and clinical variables between masked SIADH and no masked SIADH group Variables Male sex, n (%) Age, y, mean (SD) Preoperative weight in kilograms, mean (SD) Weight change (kg), mean (SD) Postoperative day 1 Postoperative day 2 Postoperative day 3 Idiopathic scoliosis, n (%) Antihypertensive use in the operating room, n (%) Pressor use in the operating room, n (%) Segments fused, median (25th-75th percentile) Surgical time in minutes, median (25th-75th percentile) Estimated blood loss, mL, median (25th-75th percentile) Volume of packed red blood cells transfused, mL, median (25th-75th percentile) Duration of hypotension in operating room, min, median (25th-75th percentile)
Masked SIADH (n = 13) 2 (15) 14.6 (2.5) 59.3 (18.2) 2.5 (1.7) 3.9 (2.0) 4.2 (3.0) 12 (92) 9 (69) 5 (38) 13 (12-13) 300 (270-300) 325 (300-400) 600 (0-675) 25 (10-95)
No masked SIADH (n = 27) 8 (30) 13.9 (2.3) 52.2 (17.0) n = 24; 2.0 (1.9) n = 23; 2.5 (2.0) n = 16; 2.7 (2.5) 17 (63) 11 (41) 19 (70) 13 (12-15) 280 (260-300) 300 (300-400) 300 (0-800) 55 (25-140)
P value .451 .396 .235 .529 .058 .217 .068 .091 .054 .353 .423 .956 .674 .136 723
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 165, No. 4
Table II. Descriptive data for patients in the masked SIADH and no masked SIADH group Masked SIADH (n = 13)
No masked SIADH (n = 27)
Variables
Mean
SD
Min
Max
Mean
SD
Min
Max
Urine osmolality, mOsm/kg Serum osmolality, mOsm/kg Serum potassium, mEq/L Urine output, mL/kg/h UNa, mEq/L Serum bicarbonate, mEq/L FeNa
828.15 286.15 4.14 0.38 201.69 24.46 0.94
139.83 4.51 0.34 0.10 57.83 1.76 0.42
609 276 3.60 0.14 117 22 0.60
1066 296 4.60 0.50 292 28 2
768.26 287.78 4.04 0.99 126.44 23.78 0.67
172.80 5.57 0.48 0.78 63.40 2.14 0.70
436 272 3.10 0.11 10 20 0.10
1024 300 5.40 3.00 246 28 2.90
postoperative hemoglobin of 0.39 between the groups. ADH levels were greater in the masked SIADH group and attained statistical significance at the 24- to 48-hour time point in the postoperative period (20.4 6.9 pg/mL vs 6.6 1.6 pg/mL, P = .049). There was no statistically significant difference in the serum cortisol levels between groups (P > .99). The mean postoperative serum cortisol at 6-12 hours was 10.6 3.4 mg/dL in the masked SIADH group and 16.6 3.1 mg/dL in the no masked SIADH group. At 4872 hours, mean serum cortisol was 7.6 1.1 mg/dL for masked SIADH and 10.0 1.3 mg/dL for no masked SIADH. Serum Na levels were similar between groups as shown in Figure 1. Fluid intake was greater in the masked SIADH group (P = .013) along with a significant positive fluid balance (P = .008; Figure 2). A total of 12 (92%) patients in the masked SIADH group met the required criteria within the first 24 hours of surgery with 50% meeting the same within 6 hours. Only one patient met the criteria after 24 hours, with none meeting the same after 48 hours.
groups, and no patient in the masked SIADH group developed hyponatremia. The incidence of classic SIADH after spinal fusion surgery ranges from 5% to 100%.2,4-6,10,11 In our patients we found the incidence of a SIADH-like state to be approximately 32%, which is within the range described in literature.2 The presence of low serum osmolality, high urine osmolality, high UNa, and greater FeNa helped us exclude subjects who had hypovolemia. Hato and Ng12 in their study of patients with SIADH showed that patients with UNa >50 mEq/L were unlikely to be hypovolemic. In our patients with masked SIADH, the median UNa was 206 mEq/L. This argues against hypovolemia as the etiology for oliguria. The serum cortisol levels were no different between groups which helped us exclude adrenal dysfunction or insufficiency as the etiology for inappropriate renal Na handling. The hallmark of SIADH is excess free water retention. Use of a greater Na load only prevents hyponatremia. It does
Discussion In this prospective observational study, we have shown that a SIADH-like state occurs in the absence of hyponatremia with the use of isotonic saline, thus supporting the hypothesis of “masked SIADH.” Serum Na levels were similar between
Figure 1. Serum Na levels in the masked SIADH and no masked SIADH groups. 724
Figure 2. A, Fluid intake and B, fluid balance between masked SIADH and no masked SIADH groups. Khan et al
ORIGINAL ARTICLES
October 2014 not however affect the excess ADH secretion and resulting free water retention.9,13 Coulthard et al8 showed that none of the 82 children receiving Hartmann solution after spinal instrumentation and craniotomy developed hypo- or hypernatremia in the postoperative period. They did not evaluate for SIADH in their study population. Although it is possible that none of the patients in this study developed SIADH, the fact that no one developed hyponatremia is worth noting. Burrows et al6 in a study of 41 patients undergoing spinal fusion noted an 80% incidence of SIADH. All patients diagnosed with SIADH received hypotonic saline in the postoperative period. Twenty percent of their patients received isotonic saline and did not develop hyponatremia. These patients had oliguria, weight gain, and high UNa levels (UNa >50 mEq/L) but absence of hyponatremia likely precluded establishing a diagnosis of SIADH using traditional criteria.1 Our study patients were similar to those described by Burrows et al. We used isotonic saline in the perioperative period in all our patients which likely prevented the development of hyponatremia in the sub group with masked SIADH. None of our patients developed hypernatremia similar to Coulthard et al.8 Lieh-Lai et al2 looked at the ADH and brain natriuretic peptide (BNP) levels in 30 children undergoing spinal fusion surgery and found a 33% incidence of SIADH based on clinical criteria. They monitored ADH and BNP levels in the perioperative period and found that ADH levels peaked around the first 6-12 hours of surgery. Highest peaks were seen in patients with SIADH. No significant difference was noted in BNP levels between patients with and without SIADH however any difference may have been hidden due to the small number of subjects. We did not evaluate BNP levels, but our results for ADH elevation in the masked SIADH group are similar to that described by Lieh-Lai et al2 in the SIADH group. A total of 92% of our patients met criteria for masked SIADH within 24 hours of surgery with 50% of them meeting the same within 6 hours. No patient developed masked SIADH after 48 hours. These results are comparable with the study by Burrows et al.6 We were able to demonstrate a trend towards statistical significance in post-operative weight gain and positive fluid balance in patients with masked SIADH. We used weight as a surrogate for fluid overload and marker of fluid retention as described in literature.14 None of the other studies that reported on occurrence of SIADH in spinal fusion patients reported weight changes. Perioperative weight gain and fluid overload has been linked to complications like increased duration of mechanical ventilation, altered wound healing, and increased risk of infections in the postoperative period.14,15 We did not study any of these outcomes and hence are unable to comment on the impact of the mild fluid overload on our patients with masked SIADH. Our study has several strengths. We used strict inclusion and exclusion criteria. A single surgeon performed all the procedures thus eliminating variation in surgical technique as a potential confounder. Uniform anesthetic techniques
were used, however anesthesia was not by any protocol. Order sets were used to eliminate variation in timing of laboratory draws and electronic medical record was used for data collection. Extensive nursing and ancillary staff education was completed before the study launch to ensure adherence to the study lab protocol. There are a few limitations to our study. We noticed a Hawthorne effect due to the increased awareness among caregivers about the possibility of SIADH-like state in our patient population, leading to more restrictive fluid management practices. Although this may have helped the study subjects, it could have diluted the results of our study. To better characterize this we evaluated the difference in the mean fluid balance between the first and last 20 patients enrolled in our study. We observed that the first 20 patients had positive fluid balance of 626 mL in the first 6 hours from surgery, as compared with 27 mL in the second 20 patients, with P = .019. Because many of our patients were managed on the pediatric floor in the postoperative period, we could not monitor their central venous pressures. Although the use of central venous pressures helps distinguish hypovolemic hyponatremia from SIADH, we alternatively used FeNa and UNa levels as one of the criteria to exclude hypovolemic hyponatremia, as patients with a FeNa >0.5, and UNa >50 mEq/L are unlikely to have hypovolemic hyponatremia.12 Having a control group would have made the study stronger. Multiple studies have shown the implications of using hypotonic fluids such as 5% dextrose in one-half NS in the postoperative period causing hyponatremia. Isotonic saline use is the standard practice in our institution and we did not feel it would be safe to use hypotonic fluids in a patient population that is known to develop SIADH in the postoperative period. Our primary objective with this pilot study was to demonstrate that “masked SIADH” is an SIADH-like state that exists even in the absence of hyponatremia but goes largely undetected using traditional criteria. Having BNP levels would have been helpful and strengthened our data. In our review we found that BNP levels were not significantly different in patients with SIADH as described by Lieh-Lai et al.2 Although any difference in BNP levels in this study may have been hidden due to the small number of patients, we chose not to measure BNP levels in our patients due to financial constraints. We would recommend that future studies looking at SIADH in the postoperative period have BNP levels measured to better understand body Na handling. Meaningful adverse outcomes were not assessed in our study. We looked at weight gain as a surrogate for fluid overload and also investigated post operative anemia. Patients with masked SIADH had more weight gain but there was no difference in hemoglobin levels between groups. Given the number of patients enrolled, we did not feel we would be able to demonstrate statistically significant differences in clinically relevant outcomes. A large multicenter study would be needed to test this hypothesis and detect statistically significant difference in clinically relevant outcomes.
Masking of Syndrome of Inappropriate Antidiuretic Hormone Secretion: The Isonatremic Syndrome
725
THE JOURNAL OF PEDIATRICS
www.jpeds.com
The etiology of the development of SIADH in patients undergoing spinal fusion is speculative. The most common hypotheses include hemodynamic instability, stretch on spinal axis, blood loss, and effect of anesthetics. Although the mechanism is most likely multifactorial, it is clear that this group of patients is at risk for development of SIADH in the early postoperative period. As it stands, clinicians are unlikely to consider SIADH as a possibility in the absence of hyponatremia thus delaying the diagnosis of the same. In the era of standard use of isotonic solutions to prevent iatrogenic hyponatremia, practitioners may fail to detect SIADH-like state resulting in inappropriate fluid management strategies. The downstream effects of this may include dilutional anemia with inappropriate transfusions, fluid overload with more diuretic usage causing fluid electrolyte imbalance, longer length of stay, and possibly wound infections.14,15 Our study highlights the utility of a set of seven clinical and laboratory criteria (serum osmolality, urine osmolality, FeNa, UNa, urine output, serum potassium, and serum bicarbonate) for making the diagnosis of “masked SIADH.” We have shown that SIADH like state occurs in the absence of hyponatremia when using isotonic salt solutions as maintenance fluids in spinal fusion patients. We propose that our study be considered as a “proof of concept” which needs to be investigated in the future. n We thank Barbara Freyenberger, Kristen Brown, Jodi Becker, and Tricia Michna (Division of Pediatric Critical Care); Elizabeth Rossiter and Stuart Weinstein, MD (Department of Orthopedics; funded by the National Institutes of Health [NIH; 2R01AR052113-06], receives royalties for editing textbooks by Lippincott, Williams, and Wilkins, and serves on the Board of Trustees for the Journal of Bone and Joint Surgery); and Jack Widness, MD (funded by NIH [P01 HL046925] and research supported by the Thrasher Research Fund 0285-3). Submitted for publication Sep 23, 2013; last revision received Apr 4, 2014; accepted May 7, 2014. Reprint requests: Sameer Kamath, MD, University of Iowa Children’s Hospital, 200 Hawkins Drive, Iowa City, IA 52242. E-mail: [email protected]
726
Vol. 165, No. 4
References 1. Schwartz WB, Bennett W, Curelop S, Bartter FC. A syndrome of renal sodium loss and hyponatremia probably resulting from inappropriate secretion of antidiuretic hormone. Am J Med 1957;23:529-42. 2. Lieh-Lai MW, Stanitski DF, Sarnaik AP, Uy HG, Rossi NF, Simpson PM, et al. Syndrome of inappropriate antidiuretic hormone secretion in children following spinal fusion. Crit Care Med 1999;27:622-7. 3. Esposito P, Piotti G, Bianzina S, Malul Y, Dal Canton A. The syndrome of inappropriate antidiuresis: pathophysiology, clinical management and new therapeutic options. Nephron Clin Pract 2011;119:c62-73. 4. Elster AD. Hyponatremia after spinal fusion caused by inappropriate secretion of antidiuretic hormone (SIADH). Clin Orthop Relat Res 1985;136-41. 5. Bell GR, Gurd AR, Orlowski JP, Andrish JT. The syndrome of inappropriate antidiuretic-hormone secretion following spinal fusion. J Bone Joint Surg Am 1986;68:720-4. 6. Burrows FA, Shutack JG, Crone RK. Inappropriate secretion of antidiuretic hormone in a postsurgical pediatric population. Crit Care Med 1983;11:527-31. 7. Holliday MA, Segar WE. The maintenance need for water in parenteral fluid therapy. Pediatrics 1957;19:823-32. 8. Coulthard MG, Long DA, Ullman AJ, Ware RS. A randomised controlled trial of Hartmann’s solution versus half normal saline in postoperative paediatric spinal instrumentation and craniotomy patients. Arch Dis Child 2012;97:491-6. 9. Choong K, Arora S, Cheng J, Farrokhyar F, Reddy D, Thabane L, et al. Hypotonic versus isotonic maintenance fluids after surgery for children: a randomized controlled trial. Pediatrics 2011;128:857-66. 10. Callewart CC, Minchew JT, Kanim LE, Tsai YC, Salehmoghaddam S, Dawson EG, et al. Hyponatremia and syndrome of inappropriate antidiuretic hormone secretion in adult spinal surgery. Spine (Phila Pa 1976) 1994;19:1674-9. 11. Brazel PW, McPhee IB. Inappropriate secretion of antidiuretic hormone in postoperative scoliosis patients: the role of fluid management. Spine (Phila Pa 1976) 1996;21:724-7. 12. Hato T, Ng R. Diagnostic value of urine sodium concentration in hyponatremia due to syndrome of inappropriate antidiuretic hormone secretion versus hypovolemia. Hawaii Med J 2010;69:264-7. 13. Moritz ML, Ayus JC. Prevention of hospital-acquired hyponatremia: a case for using isotonic saline. Pediatrics 2003;111:227-30. 14. Lowell JA, Schifferdecker C, Driscoll DF, Benotti PN, Bistrian BR. Postoperative fluid overload: not a benign problem. Crit Care Med 1990;18:728-33. 15. Misteli H, Kalbermatten D, Settelen C. Simple and complicated surgical wounds [in German]. Ther Umsch 2012;69:23-7.
Khan et al
I
nappropriate antidiuretic hormone (ADH) secretion occurs in many clinical conditions and is designated as the syndrome of inappropriate ADH secretion (SIADH). It is a disorder of sodium (Na) and water balance characterized by impairment of urinary dilution and hypotonic hyponatremia in the absence of renal disease or other identifiable nonosmotic stimuli.1-3 The classic criteria used to diagnose SIADH are hyponatremia (serum Na <135 mEq/L), low serum osmolality (plasma osmolality <280 mOsm/kg), and high urine osmolality (urine osmolality >100 mOsm/kg).1 The incidence of SIADH after spinal fusion has ranged from 5% to 100%.1,2,4-6 The routine administration of isotonic saline at maintenance rates in the postoperative period after spinal fusion decreases the incidence of hyponatremia but does not result in hypernatremia.7-9 Most postoperative studies performed to evaluate the incidence of SIADH after spinal fusion surgery note the use of 0.45% normal saline (NS).2,4-6,10,11 All patients with SIADH in these studies had hyponatremia and fulfilled traditional criteria for diagnosis of this condition. We hypothesized that the routine use of isotonic saline in the postoperative period prevents the development of hyponatremia masking the detection of SIADH. We thus excluded hyponatremia as a diagnostic criterion for diagnosing a SIADH-like state.
Methods After approval by the institutional review board, we performed a prospective observational cohort study at the University of Iowa Children’s Hospital. Written informed consent was obtained from all parents, and assent was obtained from all children 12 years of age and older. Patients were excluded if they had preexisting renal, adrenal, pituitary, or thyroid disease. Patients on diuretics and/or with a history of congestive heart failure also were excluded.
ADH BNP FeNa Na NS SIADH UNa
Antidiuretic hormone Brain natriuretic peptide Fractional excretion of sodium Sodium Normal saline Syndrome of inappropriate antidiuretic hormone secretion Urine sodium
From the 1Division of Pediatric Critical Care, Presbyterian Hospital, Albuquerque, NM; 2Department of Biostatistics at the University of Iowa; 3Department of Pediatric Nephrology, and 4Division of Pediatric Critical Care, University of Iowa Children’s Hospital, Iowa City, IA Funded by Children’s Miracle Network Proposal #2231 and Pediatric Critical Care division at the University of Iowa Children’s Hospital. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2014.05.051
722
Vol. 165, No. 4 October 2014 Baseline weight (bed scale), vital signs, and laboratory testing were obtained (serum and urine electrolytes, osmolality, blood urea nitrogen, creatinine, serum albumin, urine-specific gravity, complete blood count, coagulation studies, venous blood gas, and random cortisol level). All patients received a combination of inhaled and intravenous anesthesia. All patients received NS, 5% albumin, or lactated Ringer solution in the operating room and 5% dextrose in NS at maintenance rates in the postoperative period. Pertinent data obtained from anesthesia records included medications, duration of hypotension, tachycardia, vital signs, ventilator data, vasopressor use, and antihypertensive use during surgery. Surgical data included the degree of spinal curvature, type of spinal fusion surgery, number of spinal segments fused, duration of surgery, and estimated blood loss. Body weight change, fluid intake and output, and the administration of blood products were noted. The laboratory tests in the postoperative period included serum and urine electrolytes, osmolality, blood urea nitrogen, creatinine, serum albumin, urine-specific gravity, complete blood count, coagulation studies, venous blood gas, random cortisol levels, and serum ADH levels. Patients were determined to have a SIADH-like state only if they met the following 7 criteria: serum osmolality <290 mOsm/kg, urine osmolality >600 mOsm/kg, fractional excretion of Na (FeNa) >0.5, urine output <0.5 mL/kg/h, urine Na (UNa) >30 mEq/L, serum bicarbonate >20 mEq/ L, serum potassium >3.5 mEq/L. We excluded serum Na <135 mEq/L as one of the criteria in making a diagnosis of “masked SIADH” to test our hypothesis. We separated the patients into 2 groups viz, masked SIADH and no masked SIADH based on the aforementioned criteria met any time during the first 72 hours of the postoperative period. Data were abstracted in the postoperative period at 6, 12, 24, 48, and 72 hours. Statistical Analyses The proportion (with 95% exact confidence limits) of patients that met the “masked SIADH” criteria within 72 hours after surgery was computed. The demographic and clinical
variables were compared between those with masked SIADH and those without masked SIADH with the use of Pearson c2 test or Fisher exact test for categorical variables, and 2-sample t test or Wilcoxon rank-sum test for the continuous variables. The comparison of fluid balance, fluid intake, and laboratory variables that were measured over time between subjects with masked SIADH and no masked SIADH was tested with the use of linear mixed model for repeated measures. The mixed model included masked SIADH status (with vs without), time, and masked SIADH*time interaction as the fixed effects. Specific comparisons of interest, such as between groups at each time, were tested using test of mean contrast with P values adjusted using Bonferroni method to account for the number of tests performed (ie, for test at each time point, adjusted P value = unadjusted P value 6 time points). P < .05 was considered statistically significant. All statistical analyses were performed using SAS (version 9.3; SAS Institute Inc, Cary, North Carolina).
Results In a period of 15 months, between July 2010 and October 2011, 47 patients met study criteria and 40 consented. Thirty-five patients underwent posterior spinal fusion and 5 underwent anteroposterior spinal fusion. Of all the study subjects, 13 (32%; 95% CI, 19%-49%) fit the definition of masked SIADH. The demographic and clinical variables for the 2 groups are listed in Table I. We found no statistically significant difference in the 2 groups (masked SIADH vs no masked SIADH). Patients in the masked SIADH group demonstrated weight gain, which was maximal on postoperative day 2 (3.9 kg vs 2.5 kg, P = .058). There was a trend towards increased pressor use in the operating room in the no masked SIADH group. Descriptive data of patients in each group are listed in Table II. Hemoglobin levels were similar between groups (P > .99). Mean postoperative hemoglobin at 0-6 hours was 9.54 0.39 g/dL for masked SIADH and 9.84 0.27 g/dL for no masked SIADH. At 48-72 hours, mean hemoglobin was 8.73 0.39 g/dL for masked SIADH and 9.37 0.27 g/ dL for no masked SIADH, with a mean difference of
Table I. Comparison of demographic and clinical variables between masked SIADH and no masked SIADH group Variables Male sex, n (%) Age, y, mean (SD) Preoperative weight in kilograms, mean (SD) Weight change (kg), mean (SD) Postoperative day 1 Postoperative day 2 Postoperative day 3 Idiopathic scoliosis, n (%) Antihypertensive use in the operating room, n (%) Pressor use in the operating room, n (%) Segments fused, median (25th-75th percentile) Surgical time in minutes, median (25th-75th percentile) Estimated blood loss, mL, median (25th-75th percentile) Volume of packed red blood cells transfused, mL, median (25th-75th percentile) Duration of hypotension in operating room, min, median (25th-75th percentile)
Masked SIADH (n = 13) 2 (15) 14.6 (2.5) 59.3 (18.2) 2.5 (1.7) 3.9 (2.0) 4.2 (3.0) 12 (92) 9 (69) 5 (38) 13 (12-13) 300 (270-300) 325 (300-400) 600 (0-675) 25 (10-95)
No masked SIADH (n = 27) 8 (30) 13.9 (2.3) 52.2 (17.0) n = 24; 2.0 (1.9) n = 23; 2.5 (2.0) n = 16; 2.7 (2.5) 17 (63) 11 (41) 19 (70) 13 (12-15) 280 (260-300) 300 (300-400) 300 (0-800) 55 (25-140)
P value .451 .396 .235 .529 .058 .217 .068 .091 .054 .353 .423 .956 .674 .136 723
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 165, No. 4
Table II. Descriptive data for patients in the masked SIADH and no masked SIADH group Masked SIADH (n = 13)
No masked SIADH (n = 27)
Variables
Mean
SD
Min
Max
Mean
SD
Min
Max
Urine osmolality, mOsm/kg Serum osmolality, mOsm/kg Serum potassium, mEq/L Urine output, mL/kg/h UNa, mEq/L Serum bicarbonate, mEq/L FeNa
828.15 286.15 4.14 0.38 201.69 24.46 0.94
139.83 4.51 0.34 0.10 57.83 1.76 0.42
609 276 3.60 0.14 117 22 0.60
1066 296 4.60 0.50 292 28 2
768.26 287.78 4.04 0.99 126.44 23.78 0.67
172.80 5.57 0.48 0.78 63.40 2.14 0.70
436 272 3.10 0.11 10 20 0.10
1024 300 5.40 3.00 246 28 2.90
postoperative hemoglobin of 0.39 between the groups. ADH levels were greater in the masked SIADH group and attained statistical significance at the 24- to 48-hour time point in the postoperative period (20.4 6.9 pg/mL vs 6.6 1.6 pg/mL, P = .049). There was no statistically significant difference in the serum cortisol levels between groups (P > .99). The mean postoperative serum cortisol at 6-12 hours was 10.6 3.4 mg/dL in the masked SIADH group and 16.6 3.1 mg/dL in the no masked SIADH group. At 4872 hours, mean serum cortisol was 7.6 1.1 mg/dL for masked SIADH and 10.0 1.3 mg/dL for no masked SIADH. Serum Na levels were similar between groups as shown in Figure 1. Fluid intake was greater in the masked SIADH group (P = .013) along with a significant positive fluid balance (P = .008; Figure 2). A total of 12 (92%) patients in the masked SIADH group met the required criteria within the first 24 hours of surgery with 50% meeting the same within 6 hours. Only one patient met the criteria after 24 hours, with none meeting the same after 48 hours.
groups, and no patient in the masked SIADH group developed hyponatremia. The incidence of classic SIADH after spinal fusion surgery ranges from 5% to 100%.2,4-6,10,11 In our patients we found the incidence of a SIADH-like state to be approximately 32%, which is within the range described in literature.2 The presence of low serum osmolality, high urine osmolality, high UNa, and greater FeNa helped us exclude subjects who had hypovolemia. Hato and Ng12 in their study of patients with SIADH showed that patients with UNa >50 mEq/L were unlikely to be hypovolemic. In our patients with masked SIADH, the median UNa was 206 mEq/L. This argues against hypovolemia as the etiology for oliguria. The serum cortisol levels were no different between groups which helped us exclude adrenal dysfunction or insufficiency as the etiology for inappropriate renal Na handling. The hallmark of SIADH is excess free water retention. Use of a greater Na load only prevents hyponatremia. It does
Discussion In this prospective observational study, we have shown that a SIADH-like state occurs in the absence of hyponatremia with the use of isotonic saline, thus supporting the hypothesis of “masked SIADH.” Serum Na levels were similar between
Figure 1. Serum Na levels in the masked SIADH and no masked SIADH groups. 724
Figure 2. A, Fluid intake and B, fluid balance between masked SIADH and no masked SIADH groups. Khan et al
ORIGINAL ARTICLES
October 2014 not however affect the excess ADH secretion and resulting free water retention.9,13 Coulthard et al8 showed that none of the 82 children receiving Hartmann solution after spinal instrumentation and craniotomy developed hypo- or hypernatremia in the postoperative period. They did not evaluate for SIADH in their study population. Although it is possible that none of the patients in this study developed SIADH, the fact that no one developed hyponatremia is worth noting. Burrows et al6 in a study of 41 patients undergoing spinal fusion noted an 80% incidence of SIADH. All patients diagnosed with SIADH received hypotonic saline in the postoperative period. Twenty percent of their patients received isotonic saline and did not develop hyponatremia. These patients had oliguria, weight gain, and high UNa levels (UNa >50 mEq/L) but absence of hyponatremia likely precluded establishing a diagnosis of SIADH using traditional criteria.1 Our study patients were similar to those described by Burrows et al. We used isotonic saline in the perioperative period in all our patients which likely prevented the development of hyponatremia in the sub group with masked SIADH. None of our patients developed hypernatremia similar to Coulthard et al.8 Lieh-Lai et al2 looked at the ADH and brain natriuretic peptide (BNP) levels in 30 children undergoing spinal fusion surgery and found a 33% incidence of SIADH based on clinical criteria. They monitored ADH and BNP levels in the perioperative period and found that ADH levels peaked around the first 6-12 hours of surgery. Highest peaks were seen in patients with SIADH. No significant difference was noted in BNP levels between patients with and without SIADH however any difference may have been hidden due to the small number of subjects. We did not evaluate BNP levels, but our results for ADH elevation in the masked SIADH group are similar to that described by Lieh-Lai et al2 in the SIADH group. A total of 92% of our patients met criteria for masked SIADH within 24 hours of surgery with 50% of them meeting the same within 6 hours. No patient developed masked SIADH after 48 hours. These results are comparable with the study by Burrows et al.6 We were able to demonstrate a trend towards statistical significance in post-operative weight gain and positive fluid balance in patients with masked SIADH. We used weight as a surrogate for fluid overload and marker of fluid retention as described in literature.14 None of the other studies that reported on occurrence of SIADH in spinal fusion patients reported weight changes. Perioperative weight gain and fluid overload has been linked to complications like increased duration of mechanical ventilation, altered wound healing, and increased risk of infections in the postoperative period.14,15 We did not study any of these outcomes and hence are unable to comment on the impact of the mild fluid overload on our patients with masked SIADH. Our study has several strengths. We used strict inclusion and exclusion criteria. A single surgeon performed all the procedures thus eliminating variation in surgical technique as a potential confounder. Uniform anesthetic techniques
were used, however anesthesia was not by any protocol. Order sets were used to eliminate variation in timing of laboratory draws and electronic medical record was used for data collection. Extensive nursing and ancillary staff education was completed before the study launch to ensure adherence to the study lab protocol. There are a few limitations to our study. We noticed a Hawthorne effect due to the increased awareness among caregivers about the possibility of SIADH-like state in our patient population, leading to more restrictive fluid management practices. Although this may have helped the study subjects, it could have diluted the results of our study. To better characterize this we evaluated the difference in the mean fluid balance between the first and last 20 patients enrolled in our study. We observed that the first 20 patients had positive fluid balance of 626 mL in the first 6 hours from surgery, as compared with 27 mL in the second 20 patients, with P = .019. Because many of our patients were managed on the pediatric floor in the postoperative period, we could not monitor their central venous pressures. Although the use of central venous pressures helps distinguish hypovolemic hyponatremia from SIADH, we alternatively used FeNa and UNa levels as one of the criteria to exclude hypovolemic hyponatremia, as patients with a FeNa >0.5, and UNa >50 mEq/L are unlikely to have hypovolemic hyponatremia.12 Having a control group would have made the study stronger. Multiple studies have shown the implications of using hypotonic fluids such as 5% dextrose in one-half NS in the postoperative period causing hyponatremia. Isotonic saline use is the standard practice in our institution and we did not feel it would be safe to use hypotonic fluids in a patient population that is known to develop SIADH in the postoperative period. Our primary objective with this pilot study was to demonstrate that “masked SIADH” is an SIADH-like state that exists even in the absence of hyponatremia but goes largely undetected using traditional criteria. Having BNP levels would have been helpful and strengthened our data. In our review we found that BNP levels were not significantly different in patients with SIADH as described by Lieh-Lai et al.2 Although any difference in BNP levels in this study may have been hidden due to the small number of patients, we chose not to measure BNP levels in our patients due to financial constraints. We would recommend that future studies looking at SIADH in the postoperative period have BNP levels measured to better understand body Na handling. Meaningful adverse outcomes were not assessed in our study. We looked at weight gain as a surrogate for fluid overload and also investigated post operative anemia. Patients with masked SIADH had more weight gain but there was no difference in hemoglobin levels between groups. Given the number of patients enrolled, we did not feel we would be able to demonstrate statistically significant differences in clinically relevant outcomes. A large multicenter study would be needed to test this hypothesis and detect statistically significant difference in clinically relevant outcomes.
Masking of Syndrome of Inappropriate Antidiuretic Hormone Secretion: The Isonatremic Syndrome
725
THE JOURNAL OF PEDIATRICS
www.jpeds.com
The etiology of the development of SIADH in patients undergoing spinal fusion is speculative. The most common hypotheses include hemodynamic instability, stretch on spinal axis, blood loss, and effect of anesthetics. Although the mechanism is most likely multifactorial, it is clear that this group of patients is at risk for development of SIADH in the early postoperative period. As it stands, clinicians are unlikely to consider SIADH as a possibility in the absence of hyponatremia thus delaying the diagnosis of the same. In the era of standard use of isotonic solutions to prevent iatrogenic hyponatremia, practitioners may fail to detect SIADH-like state resulting in inappropriate fluid management strategies. The downstream effects of this may include dilutional anemia with inappropriate transfusions, fluid overload with more diuretic usage causing fluid electrolyte imbalance, longer length of stay, and possibly wound infections.14,15 Our study highlights the utility of a set of seven clinical and laboratory criteria (serum osmolality, urine osmolality, FeNa, UNa, urine output, serum potassium, and serum bicarbonate) for making the diagnosis of “masked SIADH.” We have shown that SIADH like state occurs in the absence of hyponatremia when using isotonic salt solutions as maintenance fluids in spinal fusion patients. We propose that our study be considered as a “proof of concept” which needs to be investigated in the future. n We thank Barbara Freyenberger, Kristen Brown, Jodi Becker, and Tricia Michna (Division of Pediatric Critical Care); Elizabeth Rossiter and Stuart Weinstein, MD (Department of Orthopedics; funded by the National Institutes of Health [NIH; 2R01AR052113-06], receives royalties for editing textbooks by Lippincott, Williams, and Wilkins, and serves on the Board of Trustees for the Journal of Bone and Joint Surgery); and Jack Widness, MD (funded by NIH [P01 HL046925] and research supported by the Thrasher Research Fund 0285-3). Submitted for publication Sep 23, 2013; last revision received Apr 4, 2014; accepted May 7, 2014. Reprint requests: Sameer Kamath, MD, University of Iowa Children’s Hospital, 200 Hawkins Drive, Iowa City, IA 52242. E-mail: [email protected]
726
Vol. 165, No. 4
References 1. Schwartz WB, Bennett W, Curelop S, Bartter FC. A syndrome of renal sodium loss and hyponatremia probably resulting from inappropriate secretion of antidiuretic hormone. Am J Med 1957;23:529-42. 2. Lieh-Lai MW, Stanitski DF, Sarnaik AP, Uy HG, Rossi NF, Simpson PM, et al. Syndrome of inappropriate antidiuretic hormone secretion in children following spinal fusion. Crit Care Med 1999;27:622-7. 3. Esposito P, Piotti G, Bianzina S, Malul Y, Dal Canton A. The syndrome of inappropriate antidiuresis: pathophysiology, clinical management and new therapeutic options. Nephron Clin Pract 2011;119:c62-73. 4. Elster AD. Hyponatremia after spinal fusion caused by inappropriate secretion of antidiuretic hormone (SIADH). Clin Orthop Relat Res 1985;136-41. 5. Bell GR, Gurd AR, Orlowski JP, Andrish JT. The syndrome of inappropriate antidiuretic-hormone secretion following spinal fusion. J Bone Joint Surg Am 1986;68:720-4. 6. Burrows FA, Shutack JG, Crone RK. Inappropriate secretion of antidiuretic hormone in a postsurgical pediatric population. Crit Care Med 1983;11:527-31. 7. Holliday MA, Segar WE. The maintenance need for water in parenteral fluid therapy. Pediatrics 1957;19:823-32. 8. Coulthard MG, Long DA, Ullman AJ, Ware RS. A randomised controlled trial of Hartmann’s solution versus half normal saline in postoperative paediatric spinal instrumentation and craniotomy patients. Arch Dis Child 2012;97:491-6. 9. Choong K, Arora S, Cheng J, Farrokhyar F, Reddy D, Thabane L, et al. Hypotonic versus isotonic maintenance fluids after surgery for children: a randomized controlled trial. Pediatrics 2011;128:857-66. 10. Callewart CC, Minchew JT, Kanim LE, Tsai YC, Salehmoghaddam S, Dawson EG, et al. Hyponatremia and syndrome of inappropriate antidiuretic hormone secretion in adult spinal surgery. Spine (Phila Pa 1976) 1994;19:1674-9. 11. Brazel PW, McPhee IB. Inappropriate secretion of antidiuretic hormone in postoperative scoliosis patients: the role of fluid management. Spine (Phila Pa 1976) 1996;21:724-7. 12. Hato T, Ng R. Diagnostic value of urine sodium concentration in hyponatremia due to syndrome of inappropriate antidiuretic hormone secretion versus hypovolemia. Hawaii Med J 2010;69:264-7. 13. Moritz ML, Ayus JC. Prevention of hospital-acquired hyponatremia: a case for using isotonic saline. Pediatrics 2003;111:227-30. 14. Lowell JA, Schifferdecker C, Driscoll DF, Benotti PN, Bistrian BR. Postoperative fluid overload: not a benign problem. Crit Care Med 1990;18:728-33. 15. Misteli H, Kalbermatten D, Settelen C. Simple and complicated surgical wounds [in German]. Ther Umsch 2012;69:23-7.
Khan et al