- Email: [email protected]
Severity and Duration of Metabolic Acidosis After Deep Hypothermic Circulatory Arrest for Thoracic Aortic Surgery
Severity and Duration of Metabolic Acidosis After Deep Hypothermic Circulatory Arrest for Thoracic Aortic Surgery Kamrouz Ghadimi, MD,* Jacob T. Gutsche, MD,* Samuel L. Setegne, BSc,* Kirk R. Jackson, MD,* John G.T. Augoustides, MD, FASE, FAHA,* E. Andrew Ochroch, MD, MSCE,* Joseph E. Bavaria, MD,† and Albert T. Cheung, MD* Objective: To determine the severity, duration, and contributing factors for metabolic acidosis after deep hypothermic circulatory arrest (DHCA). Design: Retrospective observational study. Setting: University hospital. Patients: Eighty-seven consecutive patients undergoing elective thoracic aortic surgery with DHCA. Interventions: Regression analysis was used to test for relationships between the severity of metabolic acidosis and clinical and laboratory variables. Measurements and Main Results: Minimum pH averaged 7.27 ⫾ 0.06, with 76 (87%) having a pH o 7.35; 55 (63%), a pH o 7.30; and 7 (8%), a pH o 7.20. The mean duration of metabolic acidosis was 7.9 ⫾ 5.0 hours (range: 0.0 - 26.8), and time to minimum pH after DHCA was 4.3 ⫾ 2.0 hours (1.0 - 10.0 hours). Hyperchloremia contributed to metabolic acidosis in 89% of patients. The severity of metabolic acidosis correlated with maximum lactate (p o 0.0001) and hospital length of stay (LOS) (r ¼ 0.22, p o 0.05), but not with DHCA time, DHCA temperature, duration of vasoactive
infusions, or ICU LOS. Patient BMI was the sole preoperative predictor of the severity of postoperative metabolic acidosis. Limitations: This retrospective analysis involved shortterm clinical outcomes related to pH severity and duration, which indirectly may have included the impact of sodium bicarbonate administration. Conclusions: Metabolic acidosis was common and severe after DHCA and was attributed to both lactic and hyperchloremic acidosis. DHCA duration and temperature had little impact on the severity of metabolic acidosis. The severity of metabolic acidosis was best predicted by the BMI and had minimal effects on short-term outcomes. Preventing hyperchloremic acidosis has the potential to decrease the severity of metabolic acidosis after DHCA. & 2015 Elsevier Inc. All rights reserved.
M
of DHCA,9 or factors unrelated to the conduct of CPB and DHCA.11 Understanding the factors that contribute to postoperative acidosis may provide information to improve and optimize the physiologic conditions of DHCA and CPB during surgery on the thoracic aorta. Identifying additional, preventable causes of postoperative metabolic acidosis may permit improved perioperative fluid and electrolyte management to mitigate the severity and duration of metabolic acidosis routinely encountered after thoracic aortic surgery. Deliberate hypothermia is utilized to prevent neurologic injury during circulatory arrest.12–14 Hypothermia also may attenuate the clinical and metabolic consequences of DHCA by reducing cellular metabolic demand.15 Although deliberate hypothermia and DHCA are used routinely in clinical practice, the metabolic profile of patients after DHCA has been characterized incompletely. The purpose of this study was to characterize the frequency, severity, and duration of metabolic acidosis after routine elective major thoracic aortic surgery performed using DHCA in a single institution with established CPB and DHCA protocols. Clinical variables, CPB, and DHCA parameters, as well as laboratory data, were examined with a hypothesis that modifiable factors existed that contributed to the severity and duration of postoperative metabolic acidosis. The identification of modifiable variables contributing to postoperative metabolic acidosis after DHCA would enable management strategies to decrease the magnitude of metabolic acidosis encountered in the routine care of patients undergoing thoracic aortic surgery requiring DHCA.
ETABOLIC ACIDOSIS is common in adult patients after cardiac surgery requiring cardiopulmonary bypass (CPB).1–4 Metabolic acidosis after CPB has been attributed to lactate production as a consequence of visceral hypoperfusion.5–7 The severity, duration, and pattern of metabolic acidosis in patients undergoing thoracic aortic surgery requiring deep hypothermic circulatory arrest (DHCA) have been described incompletely. Ischemia from temporary interruption of systemic blood flow would be expected to cause lactate production, but there may be other contributing factors to metabolic acidosis.8,9 The severity and duration of metabolic acidosis after DHCA may have clinical impact because severe metabolic acidosis may cause myocardial depression, attenuate the action of catecholamines on the circulation, prolong the need for mechanical ventilator support, or mask mesenteric ischemia.10 Understanding the expected pattern of arterial pH changes and the expected severity and duration of metabolic acidosis after DHCA is important for the routine perioperative care of patients after uncomplicated, elective thoracic aortic surgery. The severity and duration of the metabolic acidosis after DHCA may depend on the duration of DHCA, the temperature
From the Departments of *Anesthesiology and Critical Care; and †Surgery, Hospital of the University of Pennsylvania, Philadelphia, PA. Address reprint requests to Kamrouz Ghadimi, MD, Department of Anesthesiology, Division of Cardiothoracic Anesthesiology & Critical Care, Duke University Medical Center, 2301 Erwin Road, HAFS 5691G, Durham, NC 27710. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.07.025 1432
KEY WORDS: deep hypothermic circulatory arrest, thoracic aortic surgery, metabolic acidosis, lactate, hyperchloremia, critical care
METHODS
Patient Population Following institutional review board approval, 100 consecutive adult patients (age Z 18 years) undergoing elective
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 6 (December), 2015: pp 1432–1440
1433
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
thoracic aortic surgery requiring a period of DHCA in combination with retrograde cerebral perfusion (RCP) at the Hospital of the University of Pennsylvania from June 2008 to December 2009 were studied retrospectively. Patient-specific data from the electronic medical record and Sunrise Clinical Manager (Allscripts Healthcare Solutions Inc., Chicago, IL) were entered into a database by trained study personnel. Exclusion criteria involved patients with dialysis-dependent end-stage renal disease, preoperative hepatic dysfunction, emergent surgery, evidence of pre-existing malperfusion, mesenteric ischemia, use of selective antegrade cerebral perfusion, or hybrid endovascular surgery. Total arch cases were excluded from the authors’ analysis in order to minimize the impact of complex perfusion strategies on metabolic derangement. Thirteen patients met exclusion criteria and were removed from the authors’ analysis. Patient characteristics included age, body mass index (BMI), gender, preoperative laboratory tests, and medical comorbidities (Table 1). According to routine clinical protocol, arterial blood gas (ABG) analysis was performed from a time-stamped and datestamped arterial blood sample every 30 minutes (ABL 80 analyzer, Radiometer Inc., Copenhagen, Denmark) in the operating room and approximately every 1-to-3 hours postoperatively while the patient was cared for in a dedicated, cardiothoracic surgical ICU. Blood gas analyzers were inspected and maintained for accuracy according to a schedule recommended by the manufacturer and clinical engineering. All ABG values were reported and analyzed as alpha-stat and included measures of blood [Naþ], [Cl-], [HCO-3], [glucose], hemoglobin, and hematocrit. The ABG value in the period before CPB was considered the baseline (time 0) sample. All ABG samples during and after CPB until pH normalization were used in the analysis. Normalization of pH was defined as a pH Z 7.35. Anion gap (AG) was calculated from the arterial blood sample using AG ¼ [Naþ] - ([Cl-] þ [HCO3-]). Anion gap acidosis was defined as an AG Z 14 mEq/L. Blood lactate concentrations were obtained according to physician discretion and not measured routinely in every ABG sample, but all lactate samples that were measured among patients in the study were used in the statistical analysis. Hypernatremia was defined as [Naþ] 4 145 mEq/L. Hyperchloremia was defined as [Cl-] 4 110 mEq/L. The duration of metabolic acidosis was defined as the elapsed time between the end of DHCA to the time of pH normalization. Anesthesia Protocol All patients underwent general anesthesia with invasive arterial blood pressure, continuous cardiac output, and mixed venous oxygen saturation monitoring (Swan-Ganz CCOmbo Pulmonary Artery Catheter, Edwards Lifesciences, Inc., Irvine, CA). Nasopharyngeal, bladder, venous blood inflow and arterial blood outflow temperatures were measured continuously. Balanced anesthetic technique was provided using fentanyl, midazolam, isoflurane, and nondepolarizing muscle relaxant. Transesophageal echocardiography (TEE) (Philips Healthcare, Bothell, WA) was used to confirm surgical diagnosis and guide hemodynamic management. Unless contraindicated,
Table 1. Baseline Characteristics of the Study Population Characteristic
Age (years) BMI (kilograms/meter2) Weight (kilograms) Male Female a MDRD eGFR (mL/min) Preoperative Cr (mg/dL) Aortic Regurgitation Hypertension Bicuspid Aortic Valve Hyperlipidemia Congestive Heart Failure Aortic Stenosis Atrial fibrillation Coronary Artery Disease Tobacco Use Cardiomyopathy Mitral Regurgitation Diabetes Mellitus LV Dysfunction Stroke Marfan's Syndrome COPD Anemia Myocardial Infarction Peripheral Vascular Disease Cerebrovascular Disease Tricuspid Regurgitation
Study Population (n ¼ 87)
57.0 ⫾ 15.1 (24 – 83) 27.8 ⫾ 4.5 (19.3 – 41.1) 85.5 ⫾ 16.7 (50.4 – 130.2) 65 (75%) 22 (25%) 76.4 ⫾ 20.0 (20.3–144.0) 1.0 ⫾ 0.3 (0.5–2.6) 62% 55% 43% 36% 32% 26% 24% 17% 14% 12% 9% 9% 6% 6% 5% 5% 3% 3% 1% 1% 1%
NOTE. Values are listed as the mean ⫾ standard deviation with the range in parentheses. Abbreviations: BMI ¼ Body Mass Index; COPD ¼ Chronic obstructive pulmonary disease; Cr¼ serum creatinine concentration; LV¼left ventricular dysfunction (defined as LVEF o 50%). a MDRD eGFR ¼ Modification of Diet in Renal Disease (MDRD) Study Group equation.37
patients received epsilon-aminocaproic acid, 75 mg/kg IV load, and 1-to-2 g/h IV infusion until admission to the ICU. Heparin anticoagulation for CPB was titrated to maintain an activated clotting time (ACT) greater than 400 seconds. Fifteen-hundred milliliters of normal saline solution were used to prime the CPB circuit prior to cannulation. Autologous blood priming of the CPB circuit was performed whenever possible in retrograde fashion through the venous cannula just prior to initiation of CPB. Normal saline within the circuit was displaced into a container and utilized as needed for intravascular volume expansion during CPB. This served to minimize crystalloid dilution of hematocrit. Normal saline was used sparingly in the operating room, outside of CPB, mainly as the carrier fluid for vasoactive infusions (25-100 mL/h) and as the crystalloid solution within blood tubing. Albumin or plasma protein fractions were not used as volume expanders during the authors’ study period. CPB and DHCA Protocol RCP was provided via the superior vena cava (SVC) venous cannula. Perfusion management and cannulation for DHCA with RCP were according to a standardized institutional
1434
GHADIMI ET AL
protocol using alpha-stat pH management strategy.16 Arterial cannulation was performed via the ascending aorta. Arterial pressure during CPB was maintained in the range of 50 mmHg to 70 mmHg. The partial pressure of oxygen (PaO2) from the arterial blood was maintained Z100 mmHg, PaCO2 between 35 mmHg to 45 mmHg, and hematocrit 421% during active cooling on CPB (Sarns 8,000 Heart-Lung Machine, model 11160 heater/cooler unit, Terumo Cardiovascular Systems, Ann Arbor, MI). The rate of cooling on CPB gradually was increased until the onset of ventricular fibrillation or until a venous blood inflow temperature of 271C was achieved. Cooling was continued until the target DHCA temperature was achieved or electrocortical silence appeared by electroencephalography (EEG). The SVC was snared between the right atrium and the azygos vein for RCP. Retrograde cerebral perfusion consisted of oxygenated blood at 121C delivered through the SVC cannula at a pressure not exceeding 25 mmHg measured in the SVC. After aortic arch repair was completed, systemic CPB was resumed via cannulation of the aortic graft. Rewarming was initiated after a period of at least 5 minutes of hypothermic reperfusion. The rate of rewarming on CPB was controlled to ensure a temperature gradient between the venous inflow and the heat exchanger of no greater than 101C and a nasopharyngeal temperature less than 36.51C. Separation from CPB was accomplished with external cardiac pacing if necessary, positive-pressure mechanical ventilator support, intravenous epinephrine infusion to maintain a cardiac index Z2.3 L/min/M2, and intravenous phenylephrine to maintain a mean arterial pressure Z60 mmHg. Phenylephrine, vasopressin, or norepinephrine was added for low systemic vascular resistance. The duration of inotropic and vasopressor infusions were recorded from the time of initiation of an infusion in the operating room after DHCA to the time of discontinuation of all vasopressor and inotropic infusions in the surgical ICU. Patients who did not receive vasoactive infusions (n ¼ 9) during the study period were assigned an infusion duration of 0 hours. Blood products were transfused to maintain post-CPB hematocrit Z25% and to correct coagulopathy after normalizing the ACT with protamine. All patients were managed postoperatively in the ICU by a dedicated intensivist according to a fast-track protocol that included active warming to normothermia (371C) and tracheal extubation as soon as metabolic, circulatory, and respiratory parameters were acceptable. Intravenous inotropic and vasopressor medications were actively and continuously titrated to achieve a target cardiac index Z2.3 L/min/M2 and MAP between 60 mmHg and 70
mmHg, then discontinued when no longer necessary. Sodium bicarbonate was administered for the management of metabolic acidosis at the discretion of the cardiac anesthesiologist in the operating room or the attending intensivist, attending aortic surgeon, nurse practitioners, and resident physicians while in the CTICU. The administration of sodium bicarbonate was limited to the treatment of metabolic acidosis. There was no established institutional protocol or algorithm to dictate sodium bicarbonate administration in the CTICU at the time the study was performed. Statistical Methods All data were analyzed using STATA 11 (StataCorp LP, College Station, TX). Maximum, minimum, and average ABG values were determined from all ABG results available from the end of DHCA until the end of POD 2. This time course was chosen since pH normalization was expected to have occurred by 48 hours after cessation of DHCA and initiation of reperfusion. The end of DHCA was chosen as the starting point of metabolic data collection due to the onset of reperfusion and redistribution of lactate into the systemic circulation.5–7 The relationships between pH values and DHCA or clinical parameters were tested using the Pearson productmoment correlation coefficients (r). A relationship was considered significant for p o 0.05. All values were presented as mean ⫾ SD (range). To establish the relationship among variables, linear regression models were developed first to determine a priori univariate predictors followed by building parsimonious models to determine which univariate predictor (s) determined the minimum pH and duration of metabolic acidosis after DHCA. Trend lines were fitted to key laboratory data using the mean values for each unit of time (hourly) where measurements were available. RESULTS
A total of 87 patients were analyzed after undergoing elective, routine thoracic aortic surgery involving the ascending aorta with graft replacement utilizing DHCA (Table 2). One patient underwent a left carotid endarterectomy prior to thoracic aortic surgery. Although this may have prolonged anesthetic exposure, undergoing this operation did not impact CPB and DHCA variables. The average DHCA time was 23.3 ⫾ 6.4 minutes (range: 10.0-46.0 minutes), with a mean DHCA temperature of 15.3 ⫾ 2.0ºC (range: 10.6–18.8ºC). There were no operative deaths, no patients required reexploration for operative bleeding in the postoperative period, and the 30-day survival rate was 100% in the authors’ study population.
Table 2. Surgery of the Thoracic Aorta Performed With DHCA Operation
Composite AVR and root graft, ascending aorta þ hemiarch graft replacement AoV repair and root graft, ascending aorta þ hemiarch graft replacement (AoV-sparing root replacement) Ascending aorta þ hemiarch graft replacement Composite AVR and root graft, ascending aorta þ hemiarch graft replacement þ mitral valve surgery Composite AVR and root graft, ascending aorta þ hemiarch graft replacement þ left carotid endarterectomy Abbreviations: AoV ¼ Aortic valve; AVR ¼ aortic valve replacement.
Number of Patients (n ¼ 87)
65 12 7 2 1
1435
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
A total of 1,173 ABG samples were obtained and analyzed among the 87 patients in the study. Seventy-six patients (87%) had a minimum pH o7.35, 55 (63%) had a minimum pH o7.30, 28 (32%) had a pH o7.25, and 7 (8%) had a pH o7.20 after DHCA. The mean minimum pH was 7.27 ⫾ 0.06, and the mean PaCO2 at time of minimum pH was 42.6 ⫾ 6.4 mmHg. The time to minimum pH was approximately 4.3 ⫾ 2.0 hours after DHCA and returned to normal at an average of 12 to 14 hours after DHCA (Fig 1, top
Table 3. Physiologic Parameters Related to the Conduct of DHCA Sample Parameter
CPB Time (minutes) DHCA Time (minutes) Minimum DHCA Temp (ºC) Minimum pH Duration of Metabolic Acidosis (hours) Time to Minimum pH Vasoactive Infusion (hours) Maximum Serum Chloride (mEq/L) Maximum Serum Sodium (mEq/L) Maximum Serum Bicarbonate (mEq/L) Maximum Anion Gap (mEq/L)a Maximum Serum Lactate (mmol/L) ICU LOS (days) Hospital LOS (days)
Size
Mean ⫾ SD
Range
87 87 87 87 87
225.8 ⫾ 55.7 151.0 – 535.0 23.7 ⫾ 5.6 14.0 – 46.0 15.3 ⫾ 2.1 10.6 – 20.7 7.27 ⫾ 0.06 7.13 – 7.41 7.9 ⫾ 5.0 0.0 – 26.8
87 87 87
4.3 ⫾ 2.0 1.0 – 10.0 28.1 ⫾ 36.9 0.0 – 342.5 114.6 ⫾ 3.6 106.0 – 124.0
87
146.6 ⫾ 3.2
138.0 – 155.0
87
27.2 ⫾ 2.4
21.8 – 33.0
87 34
10.4 ⫾ 3.1 7.8 ⫾ 4.1
2.6 – 19.8 1.0 – 16.9
87 87
1.7 ⫾ 1.2 9.6 ⫾ 3.3
1.0 – 8.0 6.0 – 21.0
NOTE. Values are listed as the mean ⫾ standard deviation. Abbreviations: CPB ¼ cardiopulmonary bypass; DHCA ¼ deep hypothermic circulatory arrest; ICU ¼ intensive care unit; LOS ¼ length of stay a Anion Gap ¼ Sodium – (Chloride þ Bicarbonate).
Fig 1. Scatter plots displaying the time course of arterial pH (top panel), lactate (middle panel), and chloride (bottom panel) after the end (time 0) of DHCA among 87 patients undergoing thoracic aortic surgery. Values obtained prior to initiation of CPB and prior to DHCA are displayed left of time ¼ 0 on the abscissa. The pH and chloride values displayed were obtained through t ¼ 48 hours (n ¼ 87). Lactate values were available up through t ¼ 38 hours (postoperative day 2) after the discontinuation of DHCA in 34 patients.
panel). The duration of metabolic acidosis averaged 7.9 ⫾ 5.0 hours but lasted up to 26.8 hours (Table 3). Arterial lactate concentration was measured in 111 blood samples among 34 patients in the sample population. The average maximum lactate concentration among these patients was 7.8 ⫾ 4.1 mmol/L (range: 1.0-16.9 mmol/L). Lactate concentrations peaked at approximately 6 hours and returned to normal (lactate o2.2 mmol/L) between 18 to 20 hours after DHCA (Fig 1, middle panel). Seven (8%) patients had AG acidosis and 77 (89%) had hyperchloremia. Preoperative serum chloride measured 107.0 ⫾ 3.3 mEq/L (101 mEq/L-115 mEq/L). Chloride concentrations increased immediately after initiation of measurements during reperfusion and peaked at 14 hours to 18 hours after DHCA (Fig 1, bottom panel). Minimum pH correlated significantly with maximum lactate (Table 4) (Fig 2) but not with the maximum AG, duration of CPB, duration of DHCA, DHCA temperature, or ICU LOS (Table 4). There was a weak but significant correlation between the severity of metabolic acidosis with hospital LOS (r ¼ 0.22, p ¼ 0.04). The duration of metabolic acidosis correlated significantly with maximum AG, but did not correlate significantly with maximum lactate concentration (p ¼ 0.059), duration of CPB, duration of DHCA, DHCA temperature, ICU LOS, or hospital LOS (Table 4). Patient BMI was the sole preoperative predictor of the severity of postoperative metabolic acidosis (r ¼ 0.005, 95% C.I. 0.007 0.002, p o 0.001). Preoperative creatinine and eGFR were predictors of minimum pH in the univariate analysis but did not contribute towards prediction of minimum pH in the multivariate model. The BMI displayed correlation with severity (r ¼ 0.36, p o 0.001) (Fig 3) and duration of metabolic acidosis (r ¼ 0.26, p ¼ 0.009).
1436
GHADIMI ET AL
Table 4. Relationship Between Minimum pH and Duration of Metabolic Acidosis With Study Parameters Correlation Coefficient (r)
Minimum pH CPB Time DHCA Time Minimum DHCA Temperature Maximum Anion Gap Maximum Chloride Maximum Lactateb Duration of Vasoactive Infusions ICU Length of Stay Hospital Length of Stay Duration of Metabolic Acidosis CPB Time DHCA Time Minimum DHCA Temperature Maximum Anion Gap Maximum Chloride Maximum Lactateb Duration of Vasoactive Infusions ICU Length of Stay Hospital Length of Stay
a
p-value
-0.013 0.081 -0.092 -0.061 -0.093 -0.583 0.106 0.162 0.218
0.895 0.410 0.350 0.535 0.344 o0.0001 0.282 0.101 0.028
0.162 0.003 0.102 0.222 0.157 0.295 -0.050 -0.073 0.081
0.101 0.976 0.299 0.026 0.112 0.059 0.611 0.630 0.591
Abbreviations: CPB ¼ cardiopulmonary bypass; DHCA ¼ deep hypothermic circulatory arrest; ICU ¼ intensive care unit. a Pearson rank correlation coefficient. b Only includes patients who had serum lactate measurements (n ¼ 34).
The impact of vasoactive infusions on the pattern of metabolic acidosis and lactate production was evaluated by determining the duration of infusions from initiation in the operating room to termination in the CTICU. Neither severity nor duration of metabolic acidosis correlated with the duration of vasoactive infusions (Table 4). In addition, lactate values were not impacted by the duration of vasoactive infusions (eg, epinephrine, phenylephrine) (r ¼ 0.14, p ¼ 0.61). DISCUSSION
Metabolic acidosis was common and often severe after routine, elective thoracic aortic surgery requiring DHCA, with
87% of patients having a minimum pH less than 7.35, 63% with a minimum pH o7.30, 32% with pH o7.25, and 8% with pH o7.20. The average minimum pH was 7.27 ⫾ 0.06 units (range: 7.13-7.41), and the average duration of metabolic acidosis was 7.9 ⫾ 5.0 hours but lasted up to 27 hours after DHCA. The end of DHCA was chosen as the starting point of metabolic data collection because it marked the onset of reperfusion and redistribution of lactate generated during the period of ischemia into the systemic circulation.5–7,17 Metabolic acidosis was most severe at approximately 4.3 ⫾ 2.0 hours after DHCA and differed from the typical patterns described after cardiac surgery performed with deliberate hypothermia on CPB without DHCA.6,18 The delayed and gradual onset of metabolic acidosis during reperfusion after DHCA may reflect the time course required for recovery from ischemia experienced during the period of circulatory arrest. This includes lactate washout from the viscera (mainly) and other tissue. The time course of pH changes in the authors’ study showed that recovery from the metabolic effects of DHCA also was gradual and did not begin until an average of 4.3 ⫾ 2.0 hours after the initiation of reperfusion. Recovery from metabolic acidosis and the normalization of pH took an average of 12 hours to 14 hours after DHCA. This time course may have been influenced by impaired lactate metabolism caused by the delayed return of hepatorenal function as a consequence of ischemia or hypothermia.6 The significant relationship observed between the minimum pH and the maximum serum lactate suggested that ischemia and reperfusion were important contributors to postoperative metabolic acidosis after thoracic aortic surgery requiring DHCA (Figs 1 and 4). Tissue and end-organ ischemia with lactic acidosis was an expected consequence of deliberate hypothermia and CPB, mainly thought to be related to the washout of lactate into the systemic circulation and its subsequent metabolism during reperfusion.5 Another potential contributor to lactic acidosis may have been peripheral vasoconstriction as a consequence of vasopressors causing tissue hypoxia by reducing blood flow to the splanchnic, skeletal muscle, or peripheral circulation.19 In addition, epinephrine, as a consequence of its metabolic actions, also has been shown to
Fig 2. Minimum arterial pH after DHCA correlated with maximum lactate after DHCA among patients (n ¼ 34) undergoing thoracic aortic surgery (r ¼ -0.58, p o 0.0001).
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
1437
Fig 3. Minimum arterial pH after DHCA correlated inversely with BMI among patients (n ¼ 87) undergoing thoracic aortic surgery (r ¼ -0.36, p o 0.001).
be an independent cause of hyperlactatemia after CPB in patients undergoing CABG and valve surgery.19 However, the pattern of hyperlactatemia observed after DHCA was unlikely to be explained by vasopressor therapy. The duration
of vasopressor use in the study population averaged 28.1 ⫾ 36.9 hours (0.0-254.7 hours), whereas the average duration of metabolic acidosis was 7.9 ⫾ 5.0 hours (0.0-26.8 hours). In addition, no correlation was found between the
Fig 4. Characteristic time course of arterial pH, lactate, and chloride in a single patient who had undergone ascending aorta and hemiarch graft replacement using DHCA and RCP. The gray zone represents the time period prior to the initiation of DHCA, the blue zone represents the time period of DHCA, the red zone indicats the time period during reperfusion and rewarming that was associated with the onset of metabolic acidosis, and the green zone represents the period during which the patient recovered from metabolic acidosis. CPB, cardiopulmonary bypass; CTICU, cardiothoracic surgical intensive care unit; DHCA, deep hypothermic circulatory arrest.
1438
duration of vasoactive infusions and the severity or duration of metabolic acidosis after DHCA among the study population. Lactic acidosis has been shown to be a predictor of poor outcomes among cardiac surgical patients and has been associated with increased 30-day mortality, the incidence of neurologic injury, hospital LOS, and ICU LOS.20–24 In contrast, only a weak but significant association between the severity of postoperative metabolic acidosis and hospital length of stay was detected in this study of patients undergoing elective, routine surgery with DHCA. Despite common observation of severe metabolic acidosis and lactatemia in the study population, there were no significant relationships between the severity and duration of postoperative metabolic acidosis with any of the other short-term clinical outcomes. It was possible that the magnitude and duration of postoperative lactic acidosis observed in the study population was an expected and obligate consequence of CPB with DHCA and had little effect on clinical outcomes. Precise characterization of the severity, duration, and timing of metabolic changes that normally occur after surgery requiring DHCA will be important for distinguishing normal patterns of recovery from pathologic patterns caused by new or evolving complications producing tissue ischemia such as bowel or limb ischemia in the postoperative period. Precise description of the severity, duration, and timing of metabolic changes that normally occur after surgery employing DHCA with a largely standardized, alpha-stat management protocol also will permit studies to compare the metabolic consequences of alternative perfusion paradigms used for thoracic aortic surgery (eg, pH-stat management, antegrade cerebral perfusion, etc.).25,26 A patient characteristic that predicted the severity of metabolic acidosis was BMI. The observation that patients with greater BMI had more severe metabolic acidosis may be explained by the greater mass of metabolically active tissue subjected to ischemia during DHCA and the subsequently greater amount of lactic acid released into the circulation during reperfusion. Another contributing factor to explain this relationship was that body mass may have affected the ability to achieve uniform systemic metabolic suppression by deliberate hypothermia delivered using CPB. It was unexpected that the severity or duration of metabolic acidosis was unrelated to the duration or the temperature of DHCA. The absence of a relationship between severity of metabolic acidosis and DHCA parameters may have indicated that physiologic conditions for DHCA had been optimized for the patients in this study sample. An alternative explanation could be that the range of DHCA durations and temperatures were too limited to generate a significant correlation among the sample population. Further studies are warranted to investigate these possibilities and to examine the effects of temperature gradients between central and peripheral body compartments on the severity and duration of postoperative metabolic acidosis. The study finding that BMI predicted the severity of metabolic acidosis suggested that the standard conditions for DHCA may need to be modified or adjusted according to BMI in patients with larger BMI undergoing CPB with deliberate hypothermia. The absence of a relation between the severity of acidosis and the AG suggested that acute non-AG acidosis was an important clinical contributor to postoperative metabolic
GHADIMI ET AL
acidosis in patients undergoing thoracic aortic surgery with DHCA. This was an inherent issue with AG calculation in the setting of hyperchloremia. The almost immediate onset of hyperchloremia from the start of metabolic data collection and the persistence of hyperchloremia in the postoperative period indicated that hyperchloremic acidosis was an important and potentially preventable contributor to postoperative acidosis in this patient population. Clinical studies have demonstrated that intravascular volume expansion with colloids, normal saline, or other crystalloid solutions containing chloride as the primary anion could cause hyperchloremic acidosis in both cardiac and noncardiac surgical patients.3,27–30 In the time period that the operations were performed, 0.9% sodium chloride was the primary crystalloid solution used to prime the CPB circuit for intravascular volume expansion on CPB and in the ICU and for compounding medications administered for intravenous administration. Other medications commonly administered to cardiac surgical patients that also may have contributed to hyperchloremia included calcium chloride and potassium chloride. The routine administration of chloride-rich solutions likely contributed to postoperative hyperchloremia that persisted for 14 hours to 16 hours after DHCA. Preventing hyperchloremic acidosis potentially could reduce the severity of metabolic acidosis in this patient population and may even have the potential to decrease hospital length of stay. Preventing hyperchloremic acidosis could be accomplished by limiting patient exposure to non-buffered chloride-containing intravenous solutions and the judicious use of balanced crystalloid solutions containing acetate or gluconate as anions for intravascular volume expansion.29 The findings of this study supported an institutional decision to replace normal saline with a balanced salt mixture (Plasmalyte As, Baxter Inc., Deerfield, IL) as the primary crystalloid solution used for the CPB prime and perioperative volume expansion for cardiac surgical patients. LIMITATIONS
The findings of this retrospective observational study were subject to the limitations of bias and data gathering inherent in this methodology. The full consequences of metabolic acidosis on mortality may have been assessed incompletely because only patients undergoing elective surgery were studied, and postoperative 30-day survival was 100% in this sample population. The selective performance of lactate measurements in 34 out of 87 patients also was considered a limitation of this study. It was likely that lactate measurements were prompted by clinical reasons such as an elevated base deficit so that the patients with lactate measurements represented a subset of patients with a greater severity of metabolic acidosis. This explanation may explain the observed correlation between maximum lactate concentrations and minimum pH (Fig 2), but the overall pattern of lactate increase after DHCA corresponded closely with the overall pattern of pH changes in the entire study population (Fig 2), clearly supported by the peak trends of lactate values (n ¼ 34) corresponding in time with minimum arterial pH measurements (n ¼ 87) (Fig 1, top and middle panels). One shortcoming in the authors’ study that deserves discussion was the administration of sodium bicarbonate at the discretion of clinical providers for the management of metabolic acidosis. The
1439
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
administration of sodium bicarbonate may have attenuated pH changes but would not have been expected to affect the course of lactate production or metabolism.31,32 However, the administration of sodium bicarbonate may have been effective for the treatment of non-AG metabolic acidosis such as hyperchloremia through the ability to reestablish pH neutrality vis-à-vis normalizing electrolyte differences between strong cations and anions.33–36 Although the exact volume of normal saline administration could not be quantified accurately during the retrospective chart review, it was appropriately presumed that the large majority of hyperchloremia resulted from saline administration. The selective administration of sodium bicarbonate may explain the absence of any observed correlations between serum chloride levels and the severity or duration of metabolic acidosis despite the presence of worsening hyperchloremia after DHCA in 89% of patients (Fig 1, bottom panel). The authors’ analysis involved short-term clinical outcomes related to pH severity and duration, which indirectly may have captured the impact of sodium bicarbonate administration. Future prospective studies will be necessary to specifically define the effects of administering sodium bicarbonate or other buffered solutions for the prevention or treatment of hyperchloremic metabolic acidosis after DHCA. CONCLUSIONS
In conclusion, mixed lactic and hyperchloremic metabolic acidosis was a common finding after elective, routine hemiarch
surgery of the thoracic aorta using DHCA. The severity or duration of metabolic acidosis was not determined by DHCA or CPB parameters. Metabolic acidosis after these operations was often severe and accompanied by elevated lactate concentrations in the postoperative period. Time to minimum pH was approximately 4.3 ⫾ 2.0 hours and lasted for an average of 7.9 ⫾ 5.0 hours after the onset of reperfusion. The severity of metabolic acidosis correlated with patient BMI and had a weak association with hospital length of stay but did not correlate with other clinical outcomes. Hyperchloremic acidosis was identified as a potentially preventable contributor to postoperative acidosis and could be prevented by limiting the administration of non-buffered, chloride-rich solutions. ACKNOWLEDGMENTS
The authors would like to acknowledge the support and assistance of Stuart J. Weiss MD, PhD, and the Division of Cardiothoracic Anesthesiology and Critical Care, Michael A. Acker, MD and the Division of Cardiothoracic Surgery, the Clinical Perfusion team, the Surgical ICU Pharmacy Staff, and the Nursing staff of the Heart and Vascular Surgical Intensive Care Unit at the Hospital of the University of Pennsylvania for their assistance in carrying out this project. The work for this project was funded entirely by the department of Anesthesiology and Critical Care at the Hospital of the University of Pennsylvania.
REFERENCES 1. Ballinger WF, Vollenweider H, Templeton JY, et al: Acidosis of hypothermia. Ann Surg 154:517-523, 1961 2. Carson SA, Morris LE, Edmark KW, et al: Acid-base management for open-heart surgery. Circulation 29:456-466, 1964 3. Ernest D, Herkes RG, Raper RF: Alterations in anion gap following cardiopulmonary bypass. Crit Care Med 20:52-66, 1992 4. Harris EA, Seelye ER, Barratt-Boyes BG: Respiratory and metabolic acid-base changes during cardiopulmonary bypass in man. Br J Anaesth 42:912-922, 1970 5. Landow L: Splanchnic lactate production in cardiac surgery patients. Crit Care Med 21:S84-S91, 1993 6. Landow L, Phillips DA, Heard SO, et al: Gastric tonometry and venous oximetry in cardiac surgery patients. Crit Care Med 19: 1226-1233, 1991 7. Takala J, Uusaro A, Parviainen I, et al: Lactate metabolism and regional lactate exchange after cardiac surgery. New Horiz 4:483-492, 1996 8. Vincent JL, De Backer D, et al: Circulatory shock. N Engl J Med 369:1726-1734, 2013 9. Walpoth BH, Walpoth-Aslan BN, Mattle HP, et al: Outcome of survivors of accidental deep hypothermia and circulatory arrest treated with extracorporeal blood warming. N Engl J Med 337:15001505, 1997 10. Wildenthal K, Mierzwiak DS, Myers RW, et al: Effects of acute lactic acidosis on left ventricular performance. Am J Physiol 214: 1352-1359, 1968 11. Makino J, Uchino S, Morimatsu H, et al: A quantitative analysis of the acidosis of cardiac arrest: A prospective observational study. Crit Care 9:R357-R362, 2005 12. Griepp RB, Stinson EB, Hollingsworth JF, et al: Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 70:1051-1063, 1975
13. James ML, Andersen ND, Swaminathan M, et al: Predictors of electrocerebral inactivity with deep hypothermia. J Thorac Cardiovasc Surg 147:1002-1007, 2014 14. Stecker MM, Cheung AT, Pochettino A, et al: Deep hypothermic circulatory arrest: I. Effects of cooling on electroencephalogram and evoked potentials. Ann Thorac Surg 71:14-21, 2001 15. Bigelow WG, Callaghan JC, Hopps JA: General hypothermia for experimental intracardiac surgery; the use of electrophrenic respirations, an artificial pacemaker for cardiac standstill and radio-frequency rewarming in general hypothermia. Ann Surg 132:531-539, 1950 16. Appoo JJ, Augoustides JG, Pochettino A, et al: Perioperative outcome in adults undergoing elective deep hypothermic circulatory arrest with retrograde cerebral perfusion in proximal aortic arch repair: Evaluation of protocol-based care. J Cardiothorac Vasc Anesth 20: 3-7, 2006 17. Mustafa I, Roth H, Hanafiah A, et al: Effect of cardiopulmonary bypass on lactate metabolism. Intensive Care Med 29:1279-1285, 2003 18. Maillet JM, Le Besnerais P, Cantoni M, et al: Frequency, risk factors, and outcome of hyperlactatemia after cardiac surgery. Chest 123:1361-1366, 2003 19. Phillipson EA, Sproule BJ: The clinical significance of elevated blood lactate. Can Med Assoc J 92:1334-1338, 1965 20. Demers P, Elkouri S, Martineau R, et al: Outcome with high blood lactate levels during cardiopulmonary bypass in adult cardiac operation. Ann Thorac Surg 70:2082-2086, 2000 21. Hugot P, Sicsic JC, Schaffuser A, et al: Base deficit in immediate postoperative period of coronary surgery with cardiopulmonary bypass and length of stay in intensive care unit. Intensive Care Med 29: 257-261, 2003
1440
22. Lindsay AJ, Xu M, Sessler DI, et al: Lactate clearance time and concentration linked to morbidity and death in cardiac surgical patients. Ann Thorac Surg 95:486-492, 2013 23. Ranucci M, De Toffol B, Isgro G, et al: Hyperlactatemia during cardiopulmonary bypass: Determinants and impact on postoperative outcome. Crit Care 10:R167, 2006 24. Toraman F, Evrenkaya S, Yuce M, et al: Lactic acidosis after cardiac surgery is associated with adverse outcome. Heart Surg Forum 7:E155-E159, 2004 25. Ganapathi AM, Hanna JM, Schechter MA, et al: Antegrade versus retrograde cerebral perfusion for hemiarch replacement with deep hypothermic circulatory arrest: Does it matter? A propensitymatched analysis. J Thorac Cardiovasc Surg 148:2896-2902, 2014 26. Urbanski PP, Lenos A, Bougioukakis P, et al: Mild-to-moderate hypothermia in aortic arch surgery using circulatory arrest: A change of paradigm? Eur J Cardiothorac Surg 41:185-191, 2012 27. Krajewski ML, Raghunathan K, Paluszkiewicz SM, et al: Meta-analysis of high- versus low-chloride content in perioperative and critical care fluid resuscitation. Br J Surg 102:24-36, 2015 28. Liskaser F, Story DA, Hayhoe M, et al: Effect of pump prime on acidosis, strong-ion-difference and unmeasured ions
GHADIMI ET AL
during cardiopulmonary bypass. Anaesth Intensive Care 37:767772, 2009 29. Liskaser FJ, Bellomo R, Hayhoe M, et al: Role of pump prime in the etiology and pathogenesis of cardiopulmonary bypass-associated acidosis. Anesthesiology 93:1170-1173, 2000 30. Scheingraber S, Rehm M, Sehmisch C, et al: Rapid saline infusion produces hyperchloremic acidosis in patients undergoing gynecologic surgery. Anesthesiology 90:1265-1270, 1999 31. Boyd JH, Walley KR: Is there a role for sodium bicarbonate in treating lactic acidosis from shock? Curr Opin Crit Care 14:379-383, 2008 32. Forsythe SM, Schmidt GA: Sodium bicarbonate for the treatment of lactic acidosis. Chest 117:260-267, 2000 33. Alston RP, Cormack L, Collinson C: Metabolic acidosis developing during cardiopulmonary bypass is related to a decrease in strong ion difference. Perfusion 19:145-152, 2004 34. Fencl V, Jabor A, Kazda A, et al: Diagnosis of metabolic acidbase disturbances in critically ill patients. Am J Respir Crit Care Med 162:2246-2251, 2000 35. Stewart PA: Independent and dependent variables of acid-base control. Respir Physiol 33:9-26, 1978 36. Story DA, Kellum JA: New aspects of acid-base balance in intensive care. Curr Opin Anaesthesiol 17:119-123, 2004
infusions, or ICU LOS. Patient BMI was the sole preoperative predictor of the severity of postoperative metabolic acidosis. Limitations: This retrospective analysis involved shortterm clinical outcomes related to pH severity and duration, which indirectly may have included the impact of sodium bicarbonate administration. Conclusions: Metabolic acidosis was common and severe after DHCA and was attributed to both lactic and hyperchloremic acidosis. DHCA duration and temperature had little impact on the severity of metabolic acidosis. The severity of metabolic acidosis was best predicted by the BMI and had minimal effects on short-term outcomes. Preventing hyperchloremic acidosis has the potential to decrease the severity of metabolic acidosis after DHCA. & 2015 Elsevier Inc. All rights reserved.
M
of DHCA,9 or factors unrelated to the conduct of CPB and DHCA.11 Understanding the factors that contribute to postoperative acidosis may provide information to improve and optimize the physiologic conditions of DHCA and CPB during surgery on the thoracic aorta. Identifying additional, preventable causes of postoperative metabolic acidosis may permit improved perioperative fluid and electrolyte management to mitigate the severity and duration of metabolic acidosis routinely encountered after thoracic aortic surgery. Deliberate hypothermia is utilized to prevent neurologic injury during circulatory arrest.12–14 Hypothermia also may attenuate the clinical and metabolic consequences of DHCA by reducing cellular metabolic demand.15 Although deliberate hypothermia and DHCA are used routinely in clinical practice, the metabolic profile of patients after DHCA has been characterized incompletely. The purpose of this study was to characterize the frequency, severity, and duration of metabolic acidosis after routine elective major thoracic aortic surgery performed using DHCA in a single institution with established CPB and DHCA protocols. Clinical variables, CPB, and DHCA parameters, as well as laboratory data, were examined with a hypothesis that modifiable factors existed that contributed to the severity and duration of postoperative metabolic acidosis. The identification of modifiable variables contributing to postoperative metabolic acidosis after DHCA would enable management strategies to decrease the magnitude of metabolic acidosis encountered in the routine care of patients undergoing thoracic aortic surgery requiring DHCA.
ETABOLIC ACIDOSIS is common in adult patients after cardiac surgery requiring cardiopulmonary bypass (CPB).1–4 Metabolic acidosis after CPB has been attributed to lactate production as a consequence of visceral hypoperfusion.5–7 The severity, duration, and pattern of metabolic acidosis in patients undergoing thoracic aortic surgery requiring deep hypothermic circulatory arrest (DHCA) have been described incompletely. Ischemia from temporary interruption of systemic blood flow would be expected to cause lactate production, but there may be other contributing factors to metabolic acidosis.8,9 The severity and duration of metabolic acidosis after DHCA may have clinical impact because severe metabolic acidosis may cause myocardial depression, attenuate the action of catecholamines on the circulation, prolong the need for mechanical ventilator support, or mask mesenteric ischemia.10 Understanding the expected pattern of arterial pH changes and the expected severity and duration of metabolic acidosis after DHCA is important for the routine perioperative care of patients after uncomplicated, elective thoracic aortic surgery. The severity and duration of the metabolic acidosis after DHCA may depend on the duration of DHCA, the temperature
From the Departments of *Anesthesiology and Critical Care; and †Surgery, Hospital of the University of Pennsylvania, Philadelphia, PA. Address reprint requests to Kamrouz Ghadimi, MD, Department of Anesthesiology, Division of Cardiothoracic Anesthesiology & Critical Care, Duke University Medical Center, 2301 Erwin Road, HAFS 5691G, Durham, NC 27710. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.07.025 1432
KEY WORDS: deep hypothermic circulatory arrest, thoracic aortic surgery, metabolic acidosis, lactate, hyperchloremia, critical care
METHODS
Patient Population Following institutional review board approval, 100 consecutive adult patients (age Z 18 years) undergoing elective
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 6 (December), 2015: pp 1432–1440
1433
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
thoracic aortic surgery requiring a period of DHCA in combination with retrograde cerebral perfusion (RCP) at the Hospital of the University of Pennsylvania from June 2008 to December 2009 were studied retrospectively. Patient-specific data from the electronic medical record and Sunrise Clinical Manager (Allscripts Healthcare Solutions Inc., Chicago, IL) were entered into a database by trained study personnel. Exclusion criteria involved patients with dialysis-dependent end-stage renal disease, preoperative hepatic dysfunction, emergent surgery, evidence of pre-existing malperfusion, mesenteric ischemia, use of selective antegrade cerebral perfusion, or hybrid endovascular surgery. Total arch cases were excluded from the authors’ analysis in order to minimize the impact of complex perfusion strategies on metabolic derangement. Thirteen patients met exclusion criteria and were removed from the authors’ analysis. Patient characteristics included age, body mass index (BMI), gender, preoperative laboratory tests, and medical comorbidities (Table 1). According to routine clinical protocol, arterial blood gas (ABG) analysis was performed from a time-stamped and datestamped arterial blood sample every 30 minutes (ABL 80 analyzer, Radiometer Inc., Copenhagen, Denmark) in the operating room and approximately every 1-to-3 hours postoperatively while the patient was cared for in a dedicated, cardiothoracic surgical ICU. Blood gas analyzers were inspected and maintained for accuracy according to a schedule recommended by the manufacturer and clinical engineering. All ABG values were reported and analyzed as alpha-stat and included measures of blood [Naþ], [Cl-], [HCO-3], [glucose], hemoglobin, and hematocrit. The ABG value in the period before CPB was considered the baseline (time 0) sample. All ABG samples during and after CPB until pH normalization were used in the analysis. Normalization of pH was defined as a pH Z 7.35. Anion gap (AG) was calculated from the arterial blood sample using AG ¼ [Naþ] - ([Cl-] þ [HCO3-]). Anion gap acidosis was defined as an AG Z 14 mEq/L. Blood lactate concentrations were obtained according to physician discretion and not measured routinely in every ABG sample, but all lactate samples that were measured among patients in the study were used in the statistical analysis. Hypernatremia was defined as [Naþ] 4 145 mEq/L. Hyperchloremia was defined as [Cl-] 4 110 mEq/L. The duration of metabolic acidosis was defined as the elapsed time between the end of DHCA to the time of pH normalization. Anesthesia Protocol All patients underwent general anesthesia with invasive arterial blood pressure, continuous cardiac output, and mixed venous oxygen saturation monitoring (Swan-Ganz CCOmbo Pulmonary Artery Catheter, Edwards Lifesciences, Inc., Irvine, CA). Nasopharyngeal, bladder, venous blood inflow and arterial blood outflow temperatures were measured continuously. Balanced anesthetic technique was provided using fentanyl, midazolam, isoflurane, and nondepolarizing muscle relaxant. Transesophageal echocardiography (TEE) (Philips Healthcare, Bothell, WA) was used to confirm surgical diagnosis and guide hemodynamic management. Unless contraindicated,
Table 1. Baseline Characteristics of the Study Population Characteristic
Age (years) BMI (kilograms/meter2) Weight (kilograms) Male Female a MDRD eGFR (mL/min) Preoperative Cr (mg/dL) Aortic Regurgitation Hypertension Bicuspid Aortic Valve Hyperlipidemia Congestive Heart Failure Aortic Stenosis Atrial fibrillation Coronary Artery Disease Tobacco Use Cardiomyopathy Mitral Regurgitation Diabetes Mellitus LV Dysfunction Stroke Marfan's Syndrome COPD Anemia Myocardial Infarction Peripheral Vascular Disease Cerebrovascular Disease Tricuspid Regurgitation
Study Population (n ¼ 87)
57.0 ⫾ 15.1 (24 – 83) 27.8 ⫾ 4.5 (19.3 – 41.1) 85.5 ⫾ 16.7 (50.4 – 130.2) 65 (75%) 22 (25%) 76.4 ⫾ 20.0 (20.3–144.0) 1.0 ⫾ 0.3 (0.5–2.6) 62% 55% 43% 36% 32% 26% 24% 17% 14% 12% 9% 9% 6% 6% 5% 5% 3% 3% 1% 1% 1%
NOTE. Values are listed as the mean ⫾ standard deviation with the range in parentheses. Abbreviations: BMI ¼ Body Mass Index; COPD ¼ Chronic obstructive pulmonary disease; Cr¼ serum creatinine concentration; LV¼left ventricular dysfunction (defined as LVEF o 50%). a MDRD eGFR ¼ Modification of Diet in Renal Disease (MDRD) Study Group equation.37
patients received epsilon-aminocaproic acid, 75 mg/kg IV load, and 1-to-2 g/h IV infusion until admission to the ICU. Heparin anticoagulation for CPB was titrated to maintain an activated clotting time (ACT) greater than 400 seconds. Fifteen-hundred milliliters of normal saline solution were used to prime the CPB circuit prior to cannulation. Autologous blood priming of the CPB circuit was performed whenever possible in retrograde fashion through the venous cannula just prior to initiation of CPB. Normal saline within the circuit was displaced into a container and utilized as needed for intravascular volume expansion during CPB. This served to minimize crystalloid dilution of hematocrit. Normal saline was used sparingly in the operating room, outside of CPB, mainly as the carrier fluid for vasoactive infusions (25-100 mL/h) and as the crystalloid solution within blood tubing. Albumin or plasma protein fractions were not used as volume expanders during the authors’ study period. CPB and DHCA Protocol RCP was provided via the superior vena cava (SVC) venous cannula. Perfusion management and cannulation for DHCA with RCP were according to a standardized institutional
1434
GHADIMI ET AL
protocol using alpha-stat pH management strategy.16 Arterial cannulation was performed via the ascending aorta. Arterial pressure during CPB was maintained in the range of 50 mmHg to 70 mmHg. The partial pressure of oxygen (PaO2) from the arterial blood was maintained Z100 mmHg, PaCO2 between 35 mmHg to 45 mmHg, and hematocrit 421% during active cooling on CPB (Sarns 8,000 Heart-Lung Machine, model 11160 heater/cooler unit, Terumo Cardiovascular Systems, Ann Arbor, MI). The rate of cooling on CPB gradually was increased until the onset of ventricular fibrillation or until a venous blood inflow temperature of 271C was achieved. Cooling was continued until the target DHCA temperature was achieved or electrocortical silence appeared by electroencephalography (EEG). The SVC was snared between the right atrium and the azygos vein for RCP. Retrograde cerebral perfusion consisted of oxygenated blood at 121C delivered through the SVC cannula at a pressure not exceeding 25 mmHg measured in the SVC. After aortic arch repair was completed, systemic CPB was resumed via cannulation of the aortic graft. Rewarming was initiated after a period of at least 5 minutes of hypothermic reperfusion. The rate of rewarming on CPB was controlled to ensure a temperature gradient between the venous inflow and the heat exchanger of no greater than 101C and a nasopharyngeal temperature less than 36.51C. Separation from CPB was accomplished with external cardiac pacing if necessary, positive-pressure mechanical ventilator support, intravenous epinephrine infusion to maintain a cardiac index Z2.3 L/min/M2, and intravenous phenylephrine to maintain a mean arterial pressure Z60 mmHg. Phenylephrine, vasopressin, or norepinephrine was added for low systemic vascular resistance. The duration of inotropic and vasopressor infusions were recorded from the time of initiation of an infusion in the operating room after DHCA to the time of discontinuation of all vasopressor and inotropic infusions in the surgical ICU. Patients who did not receive vasoactive infusions (n ¼ 9) during the study period were assigned an infusion duration of 0 hours. Blood products were transfused to maintain post-CPB hematocrit Z25% and to correct coagulopathy after normalizing the ACT with protamine. All patients were managed postoperatively in the ICU by a dedicated intensivist according to a fast-track protocol that included active warming to normothermia (371C) and tracheal extubation as soon as metabolic, circulatory, and respiratory parameters were acceptable. Intravenous inotropic and vasopressor medications were actively and continuously titrated to achieve a target cardiac index Z2.3 L/min/M2 and MAP between 60 mmHg and 70
mmHg, then discontinued when no longer necessary. Sodium bicarbonate was administered for the management of metabolic acidosis at the discretion of the cardiac anesthesiologist in the operating room or the attending intensivist, attending aortic surgeon, nurse practitioners, and resident physicians while in the CTICU. The administration of sodium bicarbonate was limited to the treatment of metabolic acidosis. There was no established institutional protocol or algorithm to dictate sodium bicarbonate administration in the CTICU at the time the study was performed. Statistical Methods All data were analyzed using STATA 11 (StataCorp LP, College Station, TX). Maximum, minimum, and average ABG values were determined from all ABG results available from the end of DHCA until the end of POD 2. This time course was chosen since pH normalization was expected to have occurred by 48 hours after cessation of DHCA and initiation of reperfusion. The end of DHCA was chosen as the starting point of metabolic data collection due to the onset of reperfusion and redistribution of lactate into the systemic circulation.5–7 The relationships between pH values and DHCA or clinical parameters were tested using the Pearson productmoment correlation coefficients (r). A relationship was considered significant for p o 0.05. All values were presented as mean ⫾ SD (range). To establish the relationship among variables, linear regression models were developed first to determine a priori univariate predictors followed by building parsimonious models to determine which univariate predictor (s) determined the minimum pH and duration of metabolic acidosis after DHCA. Trend lines were fitted to key laboratory data using the mean values for each unit of time (hourly) where measurements were available. RESULTS
A total of 87 patients were analyzed after undergoing elective, routine thoracic aortic surgery involving the ascending aorta with graft replacement utilizing DHCA (Table 2). One patient underwent a left carotid endarterectomy prior to thoracic aortic surgery. Although this may have prolonged anesthetic exposure, undergoing this operation did not impact CPB and DHCA variables. The average DHCA time was 23.3 ⫾ 6.4 minutes (range: 10.0-46.0 minutes), with a mean DHCA temperature of 15.3 ⫾ 2.0ºC (range: 10.6–18.8ºC). There were no operative deaths, no patients required reexploration for operative bleeding in the postoperative period, and the 30-day survival rate was 100% in the authors’ study population.
Table 2. Surgery of the Thoracic Aorta Performed With DHCA Operation
Composite AVR and root graft, ascending aorta þ hemiarch graft replacement AoV repair and root graft, ascending aorta þ hemiarch graft replacement (AoV-sparing root replacement) Ascending aorta þ hemiarch graft replacement Composite AVR and root graft, ascending aorta þ hemiarch graft replacement þ mitral valve surgery Composite AVR and root graft, ascending aorta þ hemiarch graft replacement þ left carotid endarterectomy Abbreviations: AoV ¼ Aortic valve; AVR ¼ aortic valve replacement.
Number of Patients (n ¼ 87)
65 12 7 2 1
1435
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
A total of 1,173 ABG samples were obtained and analyzed among the 87 patients in the study. Seventy-six patients (87%) had a minimum pH o7.35, 55 (63%) had a minimum pH o7.30, 28 (32%) had a pH o7.25, and 7 (8%) had a pH o7.20 after DHCA. The mean minimum pH was 7.27 ⫾ 0.06, and the mean PaCO2 at time of minimum pH was 42.6 ⫾ 6.4 mmHg. The time to minimum pH was approximately 4.3 ⫾ 2.0 hours after DHCA and returned to normal at an average of 12 to 14 hours after DHCA (Fig 1, top
Table 3. Physiologic Parameters Related to the Conduct of DHCA Sample Parameter
CPB Time (minutes) DHCA Time (minutes) Minimum DHCA Temp (ºC) Minimum pH Duration of Metabolic Acidosis (hours) Time to Minimum pH Vasoactive Infusion (hours) Maximum Serum Chloride (mEq/L) Maximum Serum Sodium (mEq/L) Maximum Serum Bicarbonate (mEq/L) Maximum Anion Gap (mEq/L)a Maximum Serum Lactate (mmol/L) ICU LOS (days) Hospital LOS (days)
Size
Mean ⫾ SD
Range
87 87 87 87 87
225.8 ⫾ 55.7 151.0 – 535.0 23.7 ⫾ 5.6 14.0 – 46.0 15.3 ⫾ 2.1 10.6 – 20.7 7.27 ⫾ 0.06 7.13 – 7.41 7.9 ⫾ 5.0 0.0 – 26.8
87 87 87
4.3 ⫾ 2.0 1.0 – 10.0 28.1 ⫾ 36.9 0.0 – 342.5 114.6 ⫾ 3.6 106.0 – 124.0
87
146.6 ⫾ 3.2
138.0 – 155.0
87
27.2 ⫾ 2.4
21.8 – 33.0
87 34
10.4 ⫾ 3.1 7.8 ⫾ 4.1
2.6 – 19.8 1.0 – 16.9
87 87
1.7 ⫾ 1.2 9.6 ⫾ 3.3
1.0 – 8.0 6.0 – 21.0
NOTE. Values are listed as the mean ⫾ standard deviation. Abbreviations: CPB ¼ cardiopulmonary bypass; DHCA ¼ deep hypothermic circulatory arrest; ICU ¼ intensive care unit; LOS ¼ length of stay a Anion Gap ¼ Sodium – (Chloride þ Bicarbonate).
Fig 1. Scatter plots displaying the time course of arterial pH (top panel), lactate (middle panel), and chloride (bottom panel) after the end (time 0) of DHCA among 87 patients undergoing thoracic aortic surgery. Values obtained prior to initiation of CPB and prior to DHCA are displayed left of time ¼ 0 on the abscissa. The pH and chloride values displayed were obtained through t ¼ 48 hours (n ¼ 87). Lactate values were available up through t ¼ 38 hours (postoperative day 2) after the discontinuation of DHCA in 34 patients.
panel). The duration of metabolic acidosis averaged 7.9 ⫾ 5.0 hours but lasted up to 26.8 hours (Table 3). Arterial lactate concentration was measured in 111 blood samples among 34 patients in the sample population. The average maximum lactate concentration among these patients was 7.8 ⫾ 4.1 mmol/L (range: 1.0-16.9 mmol/L). Lactate concentrations peaked at approximately 6 hours and returned to normal (lactate o2.2 mmol/L) between 18 to 20 hours after DHCA (Fig 1, middle panel). Seven (8%) patients had AG acidosis and 77 (89%) had hyperchloremia. Preoperative serum chloride measured 107.0 ⫾ 3.3 mEq/L (101 mEq/L-115 mEq/L). Chloride concentrations increased immediately after initiation of measurements during reperfusion and peaked at 14 hours to 18 hours after DHCA (Fig 1, bottom panel). Minimum pH correlated significantly with maximum lactate (Table 4) (Fig 2) but not with the maximum AG, duration of CPB, duration of DHCA, DHCA temperature, or ICU LOS (Table 4). There was a weak but significant correlation between the severity of metabolic acidosis with hospital LOS (r ¼ 0.22, p ¼ 0.04). The duration of metabolic acidosis correlated significantly with maximum AG, but did not correlate significantly with maximum lactate concentration (p ¼ 0.059), duration of CPB, duration of DHCA, DHCA temperature, ICU LOS, or hospital LOS (Table 4). Patient BMI was the sole preoperative predictor of the severity of postoperative metabolic acidosis (r ¼ 0.005, 95% C.I. 0.007 0.002, p o 0.001). Preoperative creatinine and eGFR were predictors of minimum pH in the univariate analysis but did not contribute towards prediction of minimum pH in the multivariate model. The BMI displayed correlation with severity (r ¼ 0.36, p o 0.001) (Fig 3) and duration of metabolic acidosis (r ¼ 0.26, p ¼ 0.009).
1436
GHADIMI ET AL
Table 4. Relationship Between Minimum pH and Duration of Metabolic Acidosis With Study Parameters Correlation Coefficient (r)
Minimum pH CPB Time DHCA Time Minimum DHCA Temperature Maximum Anion Gap Maximum Chloride Maximum Lactateb Duration of Vasoactive Infusions ICU Length of Stay Hospital Length of Stay Duration of Metabolic Acidosis CPB Time DHCA Time Minimum DHCA Temperature Maximum Anion Gap Maximum Chloride Maximum Lactateb Duration of Vasoactive Infusions ICU Length of Stay Hospital Length of Stay
a
p-value
-0.013 0.081 -0.092 -0.061 -0.093 -0.583 0.106 0.162 0.218
0.895 0.410 0.350 0.535 0.344 o0.0001 0.282 0.101 0.028
0.162 0.003 0.102 0.222 0.157 0.295 -0.050 -0.073 0.081
0.101 0.976 0.299 0.026 0.112 0.059 0.611 0.630 0.591
Abbreviations: CPB ¼ cardiopulmonary bypass; DHCA ¼ deep hypothermic circulatory arrest; ICU ¼ intensive care unit. a Pearson rank correlation coefficient. b Only includes patients who had serum lactate measurements (n ¼ 34).
The impact of vasoactive infusions on the pattern of metabolic acidosis and lactate production was evaluated by determining the duration of infusions from initiation in the operating room to termination in the CTICU. Neither severity nor duration of metabolic acidosis correlated with the duration of vasoactive infusions (Table 4). In addition, lactate values were not impacted by the duration of vasoactive infusions (eg, epinephrine, phenylephrine) (r ¼ 0.14, p ¼ 0.61). DISCUSSION
Metabolic acidosis was common and often severe after routine, elective thoracic aortic surgery requiring DHCA, with
87% of patients having a minimum pH less than 7.35, 63% with a minimum pH o7.30, 32% with pH o7.25, and 8% with pH o7.20. The average minimum pH was 7.27 ⫾ 0.06 units (range: 7.13-7.41), and the average duration of metabolic acidosis was 7.9 ⫾ 5.0 hours but lasted up to 27 hours after DHCA. The end of DHCA was chosen as the starting point of metabolic data collection because it marked the onset of reperfusion and redistribution of lactate generated during the period of ischemia into the systemic circulation.5–7,17 Metabolic acidosis was most severe at approximately 4.3 ⫾ 2.0 hours after DHCA and differed from the typical patterns described after cardiac surgery performed with deliberate hypothermia on CPB without DHCA.6,18 The delayed and gradual onset of metabolic acidosis during reperfusion after DHCA may reflect the time course required for recovery from ischemia experienced during the period of circulatory arrest. This includes lactate washout from the viscera (mainly) and other tissue. The time course of pH changes in the authors’ study showed that recovery from the metabolic effects of DHCA also was gradual and did not begin until an average of 4.3 ⫾ 2.0 hours after the initiation of reperfusion. Recovery from metabolic acidosis and the normalization of pH took an average of 12 hours to 14 hours after DHCA. This time course may have been influenced by impaired lactate metabolism caused by the delayed return of hepatorenal function as a consequence of ischemia or hypothermia.6 The significant relationship observed between the minimum pH and the maximum serum lactate suggested that ischemia and reperfusion were important contributors to postoperative metabolic acidosis after thoracic aortic surgery requiring DHCA (Figs 1 and 4). Tissue and end-organ ischemia with lactic acidosis was an expected consequence of deliberate hypothermia and CPB, mainly thought to be related to the washout of lactate into the systemic circulation and its subsequent metabolism during reperfusion.5 Another potential contributor to lactic acidosis may have been peripheral vasoconstriction as a consequence of vasopressors causing tissue hypoxia by reducing blood flow to the splanchnic, skeletal muscle, or peripheral circulation.19 In addition, epinephrine, as a consequence of its metabolic actions, also has been shown to
Fig 2. Minimum arterial pH after DHCA correlated with maximum lactate after DHCA among patients (n ¼ 34) undergoing thoracic aortic surgery (r ¼ -0.58, p o 0.0001).
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
1437
Fig 3. Minimum arterial pH after DHCA correlated inversely with BMI among patients (n ¼ 87) undergoing thoracic aortic surgery (r ¼ -0.36, p o 0.001).
be an independent cause of hyperlactatemia after CPB in patients undergoing CABG and valve surgery.19 However, the pattern of hyperlactatemia observed after DHCA was unlikely to be explained by vasopressor therapy. The duration
of vasopressor use in the study population averaged 28.1 ⫾ 36.9 hours (0.0-254.7 hours), whereas the average duration of metabolic acidosis was 7.9 ⫾ 5.0 hours (0.0-26.8 hours). In addition, no correlation was found between the
Fig 4. Characteristic time course of arterial pH, lactate, and chloride in a single patient who had undergone ascending aorta and hemiarch graft replacement using DHCA and RCP. The gray zone represents the time period prior to the initiation of DHCA, the blue zone represents the time period of DHCA, the red zone indicats the time period during reperfusion and rewarming that was associated with the onset of metabolic acidosis, and the green zone represents the period during which the patient recovered from metabolic acidosis. CPB, cardiopulmonary bypass; CTICU, cardiothoracic surgical intensive care unit; DHCA, deep hypothermic circulatory arrest.
1438
duration of vasoactive infusions and the severity or duration of metabolic acidosis after DHCA among the study population. Lactic acidosis has been shown to be a predictor of poor outcomes among cardiac surgical patients and has been associated with increased 30-day mortality, the incidence of neurologic injury, hospital LOS, and ICU LOS.20–24 In contrast, only a weak but significant association between the severity of postoperative metabolic acidosis and hospital length of stay was detected in this study of patients undergoing elective, routine surgery with DHCA. Despite common observation of severe metabolic acidosis and lactatemia in the study population, there were no significant relationships between the severity and duration of postoperative metabolic acidosis with any of the other short-term clinical outcomes. It was possible that the magnitude and duration of postoperative lactic acidosis observed in the study population was an expected and obligate consequence of CPB with DHCA and had little effect on clinical outcomes. Precise characterization of the severity, duration, and timing of metabolic changes that normally occur after surgery requiring DHCA will be important for distinguishing normal patterns of recovery from pathologic patterns caused by new or evolving complications producing tissue ischemia such as bowel or limb ischemia in the postoperative period. Precise description of the severity, duration, and timing of metabolic changes that normally occur after surgery employing DHCA with a largely standardized, alpha-stat management protocol also will permit studies to compare the metabolic consequences of alternative perfusion paradigms used for thoracic aortic surgery (eg, pH-stat management, antegrade cerebral perfusion, etc.).25,26 A patient characteristic that predicted the severity of metabolic acidosis was BMI. The observation that patients with greater BMI had more severe metabolic acidosis may be explained by the greater mass of metabolically active tissue subjected to ischemia during DHCA and the subsequently greater amount of lactic acid released into the circulation during reperfusion. Another contributing factor to explain this relationship was that body mass may have affected the ability to achieve uniform systemic metabolic suppression by deliberate hypothermia delivered using CPB. It was unexpected that the severity or duration of metabolic acidosis was unrelated to the duration or the temperature of DHCA. The absence of a relationship between severity of metabolic acidosis and DHCA parameters may have indicated that physiologic conditions for DHCA had been optimized for the patients in this study sample. An alternative explanation could be that the range of DHCA durations and temperatures were too limited to generate a significant correlation among the sample population. Further studies are warranted to investigate these possibilities and to examine the effects of temperature gradients between central and peripheral body compartments on the severity and duration of postoperative metabolic acidosis. The study finding that BMI predicted the severity of metabolic acidosis suggested that the standard conditions for DHCA may need to be modified or adjusted according to BMI in patients with larger BMI undergoing CPB with deliberate hypothermia. The absence of a relation between the severity of acidosis and the AG suggested that acute non-AG acidosis was an important clinical contributor to postoperative metabolic
GHADIMI ET AL
acidosis in patients undergoing thoracic aortic surgery with DHCA. This was an inherent issue with AG calculation in the setting of hyperchloremia. The almost immediate onset of hyperchloremia from the start of metabolic data collection and the persistence of hyperchloremia in the postoperative period indicated that hyperchloremic acidosis was an important and potentially preventable contributor to postoperative acidosis in this patient population. Clinical studies have demonstrated that intravascular volume expansion with colloids, normal saline, or other crystalloid solutions containing chloride as the primary anion could cause hyperchloremic acidosis in both cardiac and noncardiac surgical patients.3,27–30 In the time period that the operations were performed, 0.9% sodium chloride was the primary crystalloid solution used to prime the CPB circuit for intravascular volume expansion on CPB and in the ICU and for compounding medications administered for intravenous administration. Other medications commonly administered to cardiac surgical patients that also may have contributed to hyperchloremia included calcium chloride and potassium chloride. The routine administration of chloride-rich solutions likely contributed to postoperative hyperchloremia that persisted for 14 hours to 16 hours after DHCA. Preventing hyperchloremic acidosis potentially could reduce the severity of metabolic acidosis in this patient population and may even have the potential to decrease hospital length of stay. Preventing hyperchloremic acidosis could be accomplished by limiting patient exposure to non-buffered chloride-containing intravenous solutions and the judicious use of balanced crystalloid solutions containing acetate or gluconate as anions for intravascular volume expansion.29 The findings of this study supported an institutional decision to replace normal saline with a balanced salt mixture (Plasmalyte As, Baxter Inc., Deerfield, IL) as the primary crystalloid solution used for the CPB prime and perioperative volume expansion for cardiac surgical patients. LIMITATIONS
The findings of this retrospective observational study were subject to the limitations of bias and data gathering inherent in this methodology. The full consequences of metabolic acidosis on mortality may have been assessed incompletely because only patients undergoing elective surgery were studied, and postoperative 30-day survival was 100% in this sample population. The selective performance of lactate measurements in 34 out of 87 patients also was considered a limitation of this study. It was likely that lactate measurements were prompted by clinical reasons such as an elevated base deficit so that the patients with lactate measurements represented a subset of patients with a greater severity of metabolic acidosis. This explanation may explain the observed correlation between maximum lactate concentrations and minimum pH (Fig 2), but the overall pattern of lactate increase after DHCA corresponded closely with the overall pattern of pH changes in the entire study population (Fig 2), clearly supported by the peak trends of lactate values (n ¼ 34) corresponding in time with minimum arterial pH measurements (n ¼ 87) (Fig 1, top and middle panels). One shortcoming in the authors’ study that deserves discussion was the administration of sodium bicarbonate at the discretion of clinical providers for the management of metabolic acidosis. The
1439
SEVERITY AND DURATION OF METABOLIC ACIDOSIS
administration of sodium bicarbonate may have attenuated pH changes but would not have been expected to affect the course of lactate production or metabolism.31,32 However, the administration of sodium bicarbonate may have been effective for the treatment of non-AG metabolic acidosis such as hyperchloremia through the ability to reestablish pH neutrality vis-à-vis normalizing electrolyte differences between strong cations and anions.33–36 Although the exact volume of normal saline administration could not be quantified accurately during the retrospective chart review, it was appropriately presumed that the large majority of hyperchloremia resulted from saline administration. The selective administration of sodium bicarbonate may explain the absence of any observed correlations between serum chloride levels and the severity or duration of metabolic acidosis despite the presence of worsening hyperchloremia after DHCA in 89% of patients (Fig 1, bottom panel). The authors’ analysis involved short-term clinical outcomes related to pH severity and duration, which indirectly may have captured the impact of sodium bicarbonate administration. Future prospective studies will be necessary to specifically define the effects of administering sodium bicarbonate or other buffered solutions for the prevention or treatment of hyperchloremic metabolic acidosis after DHCA. CONCLUSIONS
In conclusion, mixed lactic and hyperchloremic metabolic acidosis was a common finding after elective, routine hemiarch
surgery of the thoracic aorta using DHCA. The severity or duration of metabolic acidosis was not determined by DHCA or CPB parameters. Metabolic acidosis after these operations was often severe and accompanied by elevated lactate concentrations in the postoperative period. Time to minimum pH was approximately 4.3 ⫾ 2.0 hours and lasted for an average of 7.9 ⫾ 5.0 hours after the onset of reperfusion. The severity of metabolic acidosis correlated with patient BMI and had a weak association with hospital length of stay but did not correlate with other clinical outcomes. Hyperchloremic acidosis was identified as a potentially preventable contributor to postoperative acidosis and could be prevented by limiting the administration of non-buffered, chloride-rich solutions. ACKNOWLEDGMENTS
The authors would like to acknowledge the support and assistance of Stuart J. Weiss MD, PhD, and the Division of Cardiothoracic Anesthesiology and Critical Care, Michael A. Acker, MD and the Division of Cardiothoracic Surgery, the Clinical Perfusion team, the Surgical ICU Pharmacy Staff, and the Nursing staff of the Heart and Vascular Surgical Intensive Care Unit at the Hospital of the University of Pennsylvania for their assistance in carrying out this project. The work for this project was funded entirely by the department of Anesthesiology and Critical Care at the Hospital of the University of Pennsylvania.
REFERENCES 1. Ballinger WF, Vollenweider H, Templeton JY, et al: Acidosis of hypothermia. Ann Surg 154:517-523, 1961 2. Carson SA, Morris LE, Edmark KW, et al: Acid-base management for open-heart surgery. Circulation 29:456-466, 1964 3. Ernest D, Herkes RG, Raper RF: Alterations in anion gap following cardiopulmonary bypass. Crit Care Med 20:52-66, 1992 4. Harris EA, Seelye ER, Barratt-Boyes BG: Respiratory and metabolic acid-base changes during cardiopulmonary bypass in man. Br J Anaesth 42:912-922, 1970 5. Landow L: Splanchnic lactate production in cardiac surgery patients. Crit Care Med 21:S84-S91, 1993 6. Landow L, Phillips DA, Heard SO, et al: Gastric tonometry and venous oximetry in cardiac surgery patients. Crit Care Med 19: 1226-1233, 1991 7. Takala J, Uusaro A, Parviainen I, et al: Lactate metabolism and regional lactate exchange after cardiac surgery. New Horiz 4:483-492, 1996 8. Vincent JL, De Backer D, et al: Circulatory shock. N Engl J Med 369:1726-1734, 2013 9. Walpoth BH, Walpoth-Aslan BN, Mattle HP, et al: Outcome of survivors of accidental deep hypothermia and circulatory arrest treated with extracorporeal blood warming. N Engl J Med 337:15001505, 1997 10. Wildenthal K, Mierzwiak DS, Myers RW, et al: Effects of acute lactic acidosis on left ventricular performance. Am J Physiol 214: 1352-1359, 1968 11. Makino J, Uchino S, Morimatsu H, et al: A quantitative analysis of the acidosis of cardiac arrest: A prospective observational study. Crit Care 9:R357-R362, 2005 12. Griepp RB, Stinson EB, Hollingsworth JF, et al: Prosthetic replacement of the aortic arch. J Thorac Cardiovasc Surg 70:1051-1063, 1975
13. James ML, Andersen ND, Swaminathan M, et al: Predictors of electrocerebral inactivity with deep hypothermia. J Thorac Cardiovasc Surg 147:1002-1007, 2014 14. Stecker MM, Cheung AT, Pochettino A, et al: Deep hypothermic circulatory arrest: I. Effects of cooling on electroencephalogram and evoked potentials. Ann Thorac Surg 71:14-21, 2001 15. Bigelow WG, Callaghan JC, Hopps JA: General hypothermia for experimental intracardiac surgery; the use of electrophrenic respirations, an artificial pacemaker for cardiac standstill and radio-frequency rewarming in general hypothermia. Ann Surg 132:531-539, 1950 16. Appoo JJ, Augoustides JG, Pochettino A, et al: Perioperative outcome in adults undergoing elective deep hypothermic circulatory arrest with retrograde cerebral perfusion in proximal aortic arch repair: Evaluation of protocol-based care. J Cardiothorac Vasc Anesth 20: 3-7, 2006 17. Mustafa I, Roth H, Hanafiah A, et al: Effect of cardiopulmonary bypass on lactate metabolism. Intensive Care Med 29:1279-1285, 2003 18. Maillet JM, Le Besnerais P, Cantoni M, et al: Frequency, risk factors, and outcome of hyperlactatemia after cardiac surgery. Chest 123:1361-1366, 2003 19. Phillipson EA, Sproule BJ: The clinical significance of elevated blood lactate. Can Med Assoc J 92:1334-1338, 1965 20. Demers P, Elkouri S, Martineau R, et al: Outcome with high blood lactate levels during cardiopulmonary bypass in adult cardiac operation. Ann Thorac Surg 70:2082-2086, 2000 21. Hugot P, Sicsic JC, Schaffuser A, et al: Base deficit in immediate postoperative period of coronary surgery with cardiopulmonary bypass and length of stay in intensive care unit. Intensive Care Med 29: 257-261, 2003
1440
22. Lindsay AJ, Xu M, Sessler DI, et al: Lactate clearance time and concentration linked to morbidity and death in cardiac surgical patients. Ann Thorac Surg 95:486-492, 2013 23. Ranucci M, De Toffol B, Isgro G, et al: Hyperlactatemia during cardiopulmonary bypass: Determinants and impact on postoperative outcome. Crit Care 10:R167, 2006 24. Toraman F, Evrenkaya S, Yuce M, et al: Lactic acidosis after cardiac surgery is associated with adverse outcome. Heart Surg Forum 7:E155-E159, 2004 25. Ganapathi AM, Hanna JM, Schechter MA, et al: Antegrade versus retrograde cerebral perfusion for hemiarch replacement with deep hypothermic circulatory arrest: Does it matter? A propensitymatched analysis. J Thorac Cardiovasc Surg 148:2896-2902, 2014 26. Urbanski PP, Lenos A, Bougioukakis P, et al: Mild-to-moderate hypothermia in aortic arch surgery using circulatory arrest: A change of paradigm? Eur J Cardiothorac Surg 41:185-191, 2012 27. Krajewski ML, Raghunathan K, Paluszkiewicz SM, et al: Meta-analysis of high- versus low-chloride content in perioperative and critical care fluid resuscitation. Br J Surg 102:24-36, 2015 28. Liskaser F, Story DA, Hayhoe M, et al: Effect of pump prime on acidosis, strong-ion-difference and unmeasured ions
GHADIMI ET AL
during cardiopulmonary bypass. Anaesth Intensive Care 37:767772, 2009 29. Liskaser FJ, Bellomo R, Hayhoe M, et al: Role of pump prime in the etiology and pathogenesis of cardiopulmonary bypass-associated acidosis. Anesthesiology 93:1170-1173, 2000 30. Scheingraber S, Rehm M, Sehmisch C, et al: Rapid saline infusion produces hyperchloremic acidosis in patients undergoing gynecologic surgery. Anesthesiology 90:1265-1270, 1999 31. Boyd JH, Walley KR: Is there a role for sodium bicarbonate in treating lactic acidosis from shock? Curr Opin Crit Care 14:379-383, 2008 32. Forsythe SM, Schmidt GA: Sodium bicarbonate for the treatment of lactic acidosis. Chest 117:260-267, 2000 33. Alston RP, Cormack L, Collinson C: Metabolic acidosis developing during cardiopulmonary bypass is related to a decrease in strong ion difference. Perfusion 19:145-152, 2004 34. Fencl V, Jabor A, Kazda A, et al: Diagnosis of metabolic acidbase disturbances in critically ill patients. Am J Respir Crit Care Med 162:2246-2251, 2000 35. Stewart PA: Independent and dependent variables of acid-base control. Respir Physiol 33:9-26, 1978 36. Story DA, Kellum JA: New aspects of acid-base balance in intensive care. Curr Opin Anaesthesiol 17:119-123, 2004