Potassium (K+) disturbances, hypokalemia or hyperkalemia, are common intraoperative complications during orthotopic liver transplantation (OLT) and can cause serious perioperative morbility and mortality (1–4). Current understanding of this important topic is largely acquired from early studies that included small cohorts of adult patients with a focus on postreperfusion hyperkalemia (1,5–8). Pediatric patients present a variety of unique surgical and anesthetic challenges during OLT (2,9). Although some evidence suggests that children undergoing OLT had different patterns of K+ disturbances compared with adults (10,11), few studies have specifically examined the intraoperative K+ disturbances in pediatric OLT.
Advances in surgical and anesthetic techniques in the last two decades have dramatically improved the safety of patients during OLT (12). Many strategies have been developed to reduce the incidence of severe intraoperative hyperkalemia. Although K+-reducing strategies are effective and hyperkalemia-related cardiac complications are now rare, the incidence of intraoperative hypokalemia, especially in pediatric patients, may be under-appreciated. Risk factor analysis using a multivariate model is an important tool to identify patients at risk before the event occurs, and such analyses may assist clinicians in decision making. Although multivariate analyses were used for hyperkalemia during OLT (8), studies on risk factors associated with hypokalemia in children undergoing OLT have not been reported.
The aims of this study were 1) to investigate the incidence of intraoperative K+disturbances with an emphasis on hypokalemia in children undergoing OLT; and 2) to identify risk factors for hypokalemia in the preperfusion and postreperfusion periods during OLT.
After approval by the IRB, a retrospective study was performed. All pediatric OLT patients (age 18 yr or less) who underwent OLT surgery at the Dumont– University of California Los Angeles Transplant Center between January 1, 1998, and December 31, 2004, were included.
Anesthetic management followed the standard care at our institution. After anesthetic induction, patients were endotracheally intubated, and their anesthesia was maintained with isoflurane, fentanyl, and neuromuscular blockade. Transfusion of packed red blood cells (RBCs) was typically administered to keep the hematocrit in the mid-20s. Hyperkalemia prophylaxis (furosemide or insulin), antifibrinolytics or veno-venous bypass were not routinely used in pediatric patients during OLT at our center. Transfused RBCs were not routinely washed before administration. Plasma-reduced RBCs (RBCs centrifuged to remove plasma) were prepared by the blood bank staff on request for cases in which a large amount of blood loss was anticipated. K+ disturbances were treated at the discretion of anesthesiologists. Blood samples for K+ measurement were drawn from the arterial catheter intraoperatively by anesthesiologists.
For the purposes of this study, the transplant procedure was divided into 2 periods: prereperfusion and postreperfusion. Postreperfusion began when the portal circulation of the liver graft was reestablished. The intraoperative K+ values were recorded at hourly intervals except for the period immediately after reperfusion. Data from 4 prereperfusion intervals (3 to 4 h, 2 to 3 h, 1 to 2 h, and 0 to 1 h before reperfusion) and 4 postreperfusion intervals (0 to 15 min, 16 min to 1 h, 1 to 2 h, and 2 to 3 h after reperfusion) were collected. Hypokalemia was defined as a K+ value below 3.5 mmol/L and hyperkalemia as a K+ value exceeding 5.0 mmol/L.
For hypokalemic risk factor analysis, patients were divided into two groups according to their K+ levels in either the prereperfusion or postreperfusion period. Patients having one or more episodes of hypokalemia in the prereperfusion period were characterized as the prereperfusion hypokalemic group, and patients having no episodes of hypokalemia in the prereperfusion period were characterized as the prereperfusion nonhypokalemia group. The postreperfusion hypokalemic and nonhypokalemic groups were classified according to the same guidelines.
The following 23 patient-related (n = 9), baseline laboratory (n = 5), and intraoperative (n = 9) variables were evaluated as potential risk factors for hypokalemia in either the prereperfusion or postreperfusion period.
- Patient-related variables: age, weight, gender, diagnosis of congenital biliary atresia or acute hepatic failure (defined as massive liver necrosis with encephalopathy developing within 8 wk from the first sign of illness in a patient without underlying chronic liver disease), retransplantation, combined transplantation with kidney or small bowl, prior surgery, and presence of ascites at surgery.
- Baseline laboratory variables: international normalized ratio (INR) of prothrombin time, serum creatinine (Cr), blood urea nitrogen (BUN), K+ and base excess (BE).
- Intraoperative variables: numbers of units of transfused RBCs and fresh-frozen plasma (FFP), urine output, surgical time, cold and warm graft ischemic times, use of piggyback technique, living-related graft, and whole graft. The last five variables were used only for postreperfusion hypokalemia analysis.
Data are expressed as mean ± sd for continuous variables and as proportions for binary variables. For continuous variables, the median and range are also reported. Before univariate analysis, each continuous variable was dichotomized at its median or at a meaningful value indicated by a scatterplot. The variables were then analyzed univariately by comparing the proportions using the χ2 test with the corresponding odds ratio (OR) and its 95% confidence interval reported. Variables with P values <0.1 in univariate analyses were further analyzed in a multivariate forward-stepwise logistic regression model. Interactions were tested among significant variables from multivariate logistic regression models. The ability of the model to discriminate between patients with and without hypokalemia was evaluated by using the area under the receiver operating characteristic (ROC) curve. A P value <0.05 was considered statistically significant. Statistical analyses were performed using the Statistical Package for the Social Sciences 13.0 for Windows (SPSS, Inc., Chicago, IL) and SAS software (version 9.1, SAS Institute, Cary, NC).
During the 7-yr study period, 275 OLTs were performed in 232 children. Seven operations were excluded from the study because of either an inadequate number (<3) of intraoperative K+ specimens or an indeterminate reperfusion time, leaving 268 OLTs for analysis. The patients ranged in age from 1 mo to18 yr. Of 268 patients, 168 (62.7%) were 3 yr old or less. The median age of the study population was 1 yr. Congenital biliary atresia (31.0%) and acute liver failure (17.2%) were the two most common indications for OLT. Other continuous variables are summarized in Table 1.
A total of 1510 intraoperative K+ specimens were recorded, with an average of 5.6 per operation. Hypokalemia was noted in 38.5% of all specimens (581/1510). Of 268 patients, 193 (72.0%) developed one or more episodes of hypokalemia: 156 (58.2%) in the prereperfusion period, 149 (55.6%) in the postreperfusion period, and 115 (42.9%) in both the prereperfuion and postreperfusion periods. Severe hypokalemia (K+ ≤2.5 mmol/L) occurred in 37 patients (13.8%). The lowest K+ value recorded was 1.6 mmol/L. Hyperkalemia occurred only in 4.6% of all specimens (70/1510). Of 268 patients, 43 (16.1%) had one or more episodes of hyperkalemia during OLT.
Figure 1 illustrates the frequency of hypokalemia and hyperkalemia before and after reperfusion. Hyperkalemia occurred in only 5.1% of specimens in the period 3 to 4 h before reperfusion. Not surprisingly, the most frequent incidence of hyperkalemia (9.1%) was in the period immediately after reperfusion. However, hypokalemia occurred in a significant 44.4% of specimens at baseline and 40.3% in the period 3 to 4 h before reperfusion. After a slight decrease, the incidence of hypokalemia increased significantly in the postreperfusion period and reached 50% in the period 2 to 3 h after reperfusion.
Results of univariate analyses of 18 variables for prereperfusion hypokalemia are shown in Table 2. There were no differences between the two groups with respect to all 9 patient-related variables, 4 intraoperative variables, baseline INR, or BUN. Patients with baseline K+ ≤3.5, BE >5, or Cr ≤0.5 had a significantly more frequent incidence of intraoperative hypokalemia in the prereperfusion period. Forward-stepwise multivariate logistic regression analysis (Table 4) confirmed these three variables as independent risk factors for prereperfusion hypokalemia. A significant interaction between baseline K+ and Cr was also noticed. Patients with baseline K+ ≤3.5 and Cr >0.5 had a more than 16-fold increased odds to develop prereperfusion hypokalemia compared with patients with baseline K+ >3.5 and Cr >0.5. Patients with baseline Cr ≤0.5 and K+ >3.5 had more than fivefold increased risk for this complication than did those with Cr >0.5 and K+ >3.5. For patients who had both risk factors, the joint effect of baseline K+ ≤3.5 and Cr ≤0.5 led to a more than 20-fold increase in the odds of prereperfusion hypokalemia compared with patients without those two risk factors. Baseline BE >5 was independently (without interaction with other variables) associated with significantly increased risks (OR, 6.62) for hypokalemia in the prereperfusion period. A model including the 3 predictors, baseline K+ ≤3.5, BE > 5, Cr ≤0.5 and the interaction between baseline K+ and Cr yielded an ROC area of 0.76 (95% confidence interval 0.70-0.82), indicating a clinically useful accuracy.
As shown in Table 3, univariate analysis of 23 variables revealed that the following 7 variables were associated with a significantly more frequent incidence of hypokalemia in the postreperfusion period: age ≤3 yr, body weight ≤15 kg, baseline K+ ≤3.5, Cr ≤0.5, FFP transfusion >90 mL/kg, urine output ≥21 mL/kg, and absence of ascites at surgery. Multivariate analysis including these 7 variables (Table 4) confirmed that 4 variables—body weight ≤15 kg, baseline K+ ≤3.5 mmol/L, FFP >90 mL/kg, and absence of ascites at surgery—were independent risk factors for postreperfusion hypokalemia. Patients with each of these 4 factors had about 3 times the risk of developing postreperfusion hypokalemia than patients without these characteristics. There were no interactions among significant variables for postreperfusion hypokalemia. An ROC area derived from the model containing 4 predictors for postreperfusion hypokalemia was 0.72 (95% confidence interval, 0.65–0.78), again indicating the model as clinically useful.
In this study, we demonstrated that a significant number of children (72.0%) developed intraoperative hypokalemia at some point during OLT. Overall, 30% to 50% of blood samples were hypokalemic during OLT, except for the time interval immediately after reperfusion. These findings suggest that children are prone to hypokalemia and resistant to hyperkalemia during OLT. This is in contrast to previous studies suggesting that hyperkalemia was the predominant disturbance and that K+ disturbances were the same in children compared to those in adults during OLT (7). Several factors likely contribute to the differences between our findings and those in previous studies. First, much of the literature examining the K+ patterns during OLT were from studies done decades ago, and our data collected from current practice may reflect a fact that the overall incidence of hyperkalemia in patients undergoing OLT has decreased with advances in surgical and anesthetic management. Although the incidence of hyperkalemia decreases, the incidence of hypokalemia during OLT may increase. Second, compared with adults, pediatric patients may be indeed be more prone to develop hypokalemia during OLT.
The reasons children are more prone to intraoperative hypokalemia are not entirely understood and need further study. However, in our multivariate study of 18 variables in the prereperfusion period, baseline K+ ≤3.5 was an important independent predictor for hypokalemia during pediatric OLT. Compared with patients with baseline K+ >3.5, patients with baseline ≤3.5 had a 16- to 20-fold increased risk of developing hypokalemia in the prereperfusion period regardless of baseline Cr levels. Because half of our patients had baseline K+ ≤3.5, pretransplant factors, such as inadequate K+ intake, diarrhea, or use of diuretics, may have played important roles in the development of prereperfusion hypokalemia. Acid-base balance is a major force in determining the transmembrane shift of K+ and patients with metabolic alkalosis demonstrated by BE >5 were at significantly increased risk for prereperfusion hypokalemia. Children with good baseline kidney function (Cr ≤0.5) were also at increased risk for hypokalemia in the prereperfusion period. We speculate that excretion of K+ by kidneys does occur and that such excretion leads to the development of intraoperative hypokalemia in those patients.
The risk factors for hypokalemia in the postreperfusion period were significantly different from those in the prereperfusion period. Baseline K+ ≤3.5 remained a risk factor for postreperfusion hypokalemia, but its importance decreased noticeably (OR 2.72 versus 16.89 in the postreperfusion versus prereperfusion period, respectively). Two important risk factors for hypokalemia in the prereperfusion period (Cr ≤0.5 and BE >5) were either significant in only univariate analysis or not significant at all in the postreperfusion period. Low body weight (≤15 kg) was a risk factor only for postreperfusion hypokalemia. This is probably because, in general, the grafts for low body weight recipients are of good quality and high-quality grafts can uptake a significant amount of K+ and therefore cause hypokalemia in the postreperfusion period (13). It was not surprising to find that transfusion of a large quantity of K+-poor solutions, such as FFP, was associated with hypokalemia. Why there was such an association only in the postreperfusion period and not in the prereperfusion period is not understood and needs to be studied.
K+ is a major intracellular cation and plays an important role in determining the membrane potentials of the cells. The hazards of hypokalemia, particularly in a relationship with cardiac complications, including electrical conduction and contractile abnormalities, are well recognized (14,15). Hypokalemia-induced life-threatening arrhythmia, such as torsades de pointes, has been reported in a child during OLT (16). Hypokalemia-associated metabolic alkalosis has also been linked to a longer hospital stay in OLT patients (17,18). In addition, hypokalemia can cause generalized muscle weakness, muscle necrosis, rhabdomyolysis, paralytic ileus, metabolic alkalosis (19), enhancement of neuromuscular blockers (20), and nephrogenic diabetes insipidus (21). Although most hypokalemia-related complications were reported outside of the transplant setting, such a high frequency of (sometimes severe) hypokalemia in pediatric patients during OLT warrants further study.
Prophylaxis using furosemide, insulin, washed RBCs, or other measures to reduce severe hyperkalemia is widely used in adults during OLT and also in pediatric patients in some centers (11). The findings from this study do not support universal prophylaxis for hyperkalemia in pediatric patients during OLT, as a significant number of pediatric patients develop hypokalemia even without prophylaxis, and prophylaxis can only exacerbate the frequency or severity of hypokalemia. Furthermore, the finding of infrequent hyperkalemia in pediatric patients during OLT suggests that more generous K+ replacement should be considered to maintain normokalemia and to avoid the potential complications caused by hypokalemia, especially in children with the risk factors for hypokalemia.
In conclusion, hypokalemia is the predominant K+ disturbance among children undergoing OLT. Baseline serum K+ ≤3.5, BE >5, and Cr ≤0.5 are the predictors for prereperfusion hypokalemia, and body weight ≤15 kg, baseline serum K+ ≤3.5, absence of ascites at surgery, and FFP >90 mL/kg are the independent predictors for postreperfusion hypokalemia. These findings support the use of K+ replacement to maintain normokalemia and avoid the potential complications related to hypokalemia in pediatric OLT, especially in children with the risk factors for hypokalemia.
We thank Dr. Jeff Gornbein, the UCLA Department of Biomathematics, for statistical support and helpful discussion and Dr. Marie Csete, Emory University Department of Anesthesiology, Atlanta, GA, for critical review of the manuscript.
1. Carmichael, FJ, Lindop MJ, Farman JV. Anesthesia for hepatic transplantation: cardiovascular and metabolic alterations and their management. Anesth Analg 1985;64:108–16.
2. Borland LM, Roule M, Cook DR. Anesthesia for pediatric orthotopic liver transplantation. Anesth Analg 1985;64:117–24.
3. Kang YG, Freeman JA, Aggarwal S, DeWolf AM. Hemodynamic instability during liver transplantation. Transplant Proc 1989;21:3489–92.
4. Aggarwal S, Kang Y, Freeman JA, et al. Postreperfusion syndrome: cardiovascular collapse following hepatic reperfusion during liver transplantation. Transplant Proc 1987;19:54–5.
5. De Wolf A, Frenette L, Kang Y, Tang C. Insulin decreases the serum potassium concentration during the anhepatic stage of liver transplantation. Anesthesiology 1993;78:677–82.
6. Farman JV, Lines JG, Williams RS, et al. Liver transplantation in man: anaesthetic and biochemical management. Anaesthesia 1974;29:17–32.
7. Veyckemans F, Carlier M, Scholtes JL, et al. Anesthetic experience in adult and pediatric orthotopic liver transplantation. Acta Anaesthesiol Belg 1986;37:77–87.
8. Nakasuji M, Bookallil MJ. Pathophysiological mechanisms of postrevascularization hyperkalemia in orthotopic liver transplantation. Anesth Analg 2000;91:1351–5.
9. Belani KG, Batra YK, Naasz M, et al. Infants are a higher intraoperative risk group for orthotopic liver transplantation. Transplant Proc 1994;26:196–7.
10. Xia V, Liu L, Tran A, et al. Intraoperatrive potassium disturbances are common in patients undergoing orthotopic liver retransplantation. Liver Transpl 2004;10:C37.
11. Belani KG, Reardon RF, Payne WD, et al. Acute hypokalemia is well tolerated in infants during liver transplantation. Transplant Proc 1994;26:147–50.
12. Steadman RH. Anesthesia for liver transplant surgery. Anesthesiol Clin North America 2004;22:687–711.
13. Abouna GM, Aldrete JA, Starzl TE. Changes in serum potassium and pH during clinical and experimental liver transplantation. Surgery 1971;69:419–26.
14. Rastegar A, Soleimani M, Rastergar A. Hypokalaemia and hyperkalaemia. Postgrad Med J 2001;77:759–64.
15. Cohn JN, Kowey PR, Whelton PK, Prisant LM. New guidelines for potassium replacement in clinical practice: a contemporary review by the National Council on Potassium in Clinical Practice. Arch Intern Med 2000;160:2429–36.
16. Chvilicek JP, Hurlbert BJ, Hill GE. Diuretic-induced hypokalaemia inducing torsades de pointes. Can J Anaesth 1995;42:1137–9.
17. Raj D, Abreo K, Zibari G. Metabolic alkalosis after orthotopic liver transplantation. Am J Transplant 2003;3:1566–9.
18. Contreras G, Garces G, Reich J, et al Predictors of alkalosis after liver transplantation. Am J Kidney Dis 2002;40:517–24.
19. Soleimani M, Bergman JA, Hosford MA, McKinney TD. Potassium depletion increases luminal Na+/H+ exchange and basolateral Na+:CO3=. J Clin Invest 1990;86:1076–83.
20. Miller RD, Roderick LL. Diuretic-induced hypokalaemia, pancuronium neuromuscular blockade and its antagonism by neostigmine. Br J Anaesth 1978;50:541–4.
© 2006 International Anesthesia Research Society
21. Amlal H, Krane CM, Chen Q, Soleimani M. Early polyuria and urinary concentrating defect in potassium deprivation. Am J Physiol Renal Physiol 2000;279:F655–63.