Intraoperative Renal Support During Liver Transplantation

Transcription

1 LIVER TRANSPLANTATION 15:73-78, 2009 ORIGINAL ARTICLE Intraoperative Renal Support During Liver Transplantation Derek R. Townsend, 1,2 Sean M. Bagshaw, 1 Michael J. Jacka, 1,2 David Bigam, 3 Dominic Cave, 2 and R. T. Noel Gibney 1 1 Division of Critical Care Medicine, 2 Department of Anesthesiology and Pain Medicine, and 3 Department of Surgery, University of Alberta Hospital, University of Alberta, Edmonton, Alberta, Canada Acute kidney injury (AKI) is common in liver failure prior to orthotopic liver transplantation (OLT) and may complicate the intraoperative course. We describe the logistics of intraoperative continuous renal replacement therapy (CRRT) during OLT and the associated clinical outcomes. We performed a retrospective review of adult patients (age 18 years) receiving intraoperative CRRT during OLT at the University of Alberta Hospital between January 1, 1996 and December 31, Demographic, detailed clinical, and perioperative data, physiologic and laboratory measures, details of renal replacement therapy (RRT) provided, and data on renal recovery and survival were ascertained. Of 636 OLTs, 41 (6.4%) received intraoperative CRRT. The most common indications for OLT in these patients were hepatitis C (34.2%) and alcoholic (29.3%) cirrhosis. The median [interquartile range Model for End-Stage Liver Disease score was 38 (31-43), and 90.2% were classified as Child-Pugh class C. Preoperatively, 70% were in the intensive care unit, 58.5% were mechanically ventilated, and 48.7% required vasopressor support. The median (IQR) duration of intraoperative CRRT was 258 ( ) minutes, representing 57% of the total operative time. Filter circuit clotting occurred in 40% but was not associated with a shorter CRRT duration (P 0.41). No other complications were described. CRRT allowed an even or negative intraoperative fluid balance in 92.7%. CRRT was continued in 78% after OLT for a median (IQR) of 5 (3-11) days. Of these, 24 (75%) were transitioned to intermittent hemodialysis for a median (IQR) of 15 (4-39) days. Survival was 97.6% at 1 month and 75.6% at 1 year. Renal recovery to RRT independence occurred in 100% of survivors by 1 year; however, the mean (standard deviation) estimated glomerular filtration rate (egfr) was 54.7 (19.1) ml/minute/m 2, with 62.1% having an egfr 60 ml/minute/m 2. In conclusion, our data suggest that intraoperative CRRT during OLT is achievable and safe. Intraoperative CRRT may be a valuable adjuvant therapy for those with preoperative AKI. Additional investigations are warranted. Liver Transpl 15:73-78, AASLD. Received April 29, 2008; accepted August 13, Orthotopic liver transplantation (OLT) can be a technically complex and lengthy procedure. 1 The intraoperative course can be associated with major alterations to systemic hemodynamics, metabolic control, and coagulation status that require the administration of large volumes of blood products and crystalloid solutions. 2 Continuous renal replacement therapy (CRRT) is now routinely used for supportive management of critically ill patients with acute kidney injury (AKI). 3 This modality of renal support is ideally suited for critically ill patients with liver failure due to enhanced hemodynamic tolerance and reduced risk for exacerbation of cerebral edema, along with superior metabolic, acidbase, and azotemic control. 4,5 Although there is a biological rationale for the use of intraoperative CRRT during OLT for patients with concomitant preoperative AKI, there is a paucity of available data, which are generally restricted to a few case reports or small series. 5-7 Abbreviations: AKI, acute kidney injury; CKD, chronic kidney disease; CRRT, continuous renal replacement therapy; egfr, estimated glomerular filtration rate; ICU, intensive care unit; IHD, intermittent hemodialysis; INR, international normalized ratio; IQR, interquartile range; MELD, Model for End-Stage Liver Disease; OLT, orthotopic liver transplantation; PaCO 2, partial pressure of carbon dioxide in arterial blood; RRT, renal replacement therapy; SD, standard deviation. Potential conflict of interest: Noel Gibney received a research grant and consulting fees from Gambro, Inc. Address reprint requests to Sean M. Bagshaw, Division of Critical Care Medicine, University of Alberta Hospital, 3C1.12 Walter C. Mackenzie Centre, Street, Edmonton, Alberta T6G2B7, Canada. Telephone: ; FAX: ; bagshaw@ualberta.ca DOI /lt Published online in Wiley InterScience ( American Association for the Study of Liver Diseases.

2 74 TOWNSEND ET AL. We have performed intraoperative CRRT during OLT on critically ill patients with liver failure and AKI since Accordingly, we describe our institutional experience with intraoperative CRRT during OLT. Our objectives were (1) to describe the logistics of intraoperative CRRT and (2) to describe the clinical outcomes for patients receiving OLT supported with intraoperative CRRT. PATIENTS AND METHODS Study Population We reviewed all adult patients (age 18 years) who received intraoperative CRRT during isolated OLT (no concomitant kidney transplant) at the University of Alberta Hospital between January 1, 1996 and December 31, The University of Alberta Hospital Liver Transplant Program was started in 1989 and now performs approximately 80 liver transplants annually. This study was approved by the University of Alberta Health Research Ethics Board. Study Protocol This was a retrospective cohort study. For each patient, demographic, detailed clinical, and perioperative data, physiologic and laboratory measures, details of renal replacement therapy (RRT) provided, and data on renal recovery and survival were ascertained. Intraoperative CRRT The decision to initiate intraoperative CRRT was determined preoperatively by the transplant team (ie, anesthesiologist, intensivist, and surgeon) in all cases. Although no specific clinical or biochemical criteria were used, patients were selected for intraoperative CRRT on the basis of preoperative kidney function (ie, azotemia, oliguria, acidemia, or RRT already being received), the severity of illness, concern for poor tolerance to intraoperative management (ie, transfusions, fluid administration, and donor graft reperfusion), or an anticipated need for initiation or continuation of RRT after surgery. Central venous access was via an 11.5F dual lumen hemodialysis catheter (Mahurkar catheter, Tyco Healthcare, Mansfield, MA) inserted preoperatively into the jugular, subclavian, or femoral vein. Initiation and monitoring of CRRT in the operating theatre were performed by the anesthesiologists and by clinical equipment specialist technicians. All CRRT was performed with commercially available CRRT machines. The Prisma CFM (Gambro Canada, St. Leonard, Canada) was used with an M100 or ST100 hemofilter with an AN69 membrane. All CRRT was performed in the continuous venovenous hemodiafiltration mode with a blood warming circuit. The blood flow rate was set in the range of 125 to 150 ml/minute. Dialysate and replacement fluids were commercially available solutions (Hemosol B0 or Prismocal, Gambro Canada) and contained bicarbonate (32 mmol/l). As these solutions are calcium-free, additional calcium was added to each bag. The addition of extra bicarbonate or potassium to either the dialysate or replacement solution was determined individually for each patient. Dialysate and replacement solution rates were set in the range of 1 to 1.5 L/hour each. No venovenous bypass was performed during the procedure. No anticoagulation was used in the majority of patients out of concern for risk of bleeding with heparin or toxicity with regional citrate. However, in 5 patients, regional citrate anticoagulation was used with no evidence of toxicity. 8,9 Three patients also received anticoagulation with prostacyclin at a fixed dose infusion of 4 ng/kg/minute. The primary objective was to commence CRRT in the operating theater after induction of anesthesia and prior to the start of the procedure. CRRT was continued for the duration of the procedure and electively discontinued to facilitate transfer to the intensive care unit (ICU). Study Definitions The presence of AKI was defined by an acute rise in serum creatinine to 133 mol/l or a reduction in urine output to 200 ml/12 hours. We defined renal recovery as independence from RRT. Preoperative and postoperative estimated glomerular filtration rates (egfrs) were estimated by the Modification of Diet in Renal Disease equation. 10 We characterized selected comorbid diseases. The presence of cardiovascular disease was defined as any of the following: documented coronary artery or ischemic heart disease, previous myocardial infarction, valvular abnormality, congestive heart failure, hypertension, or peripheral vascular disease. The presence of respiratory disease was defined as any of the following: documented smoking history, asthma, chronic obstructive pulmonary disease, or interstitial lung disease. Severity of cirrhosis was determined by the Child-Turcotte-Pugh score. 11 Disease severity and mortality risk were also determined by calculation of the Model for End-Stage Liver Disease (MELD) score. 12 Data Collection All data were collected on standardized case report forms. Demographic data included age, sex, comorbid disease, and date of hospital/icu admission and discharge. A regional database maintained by the University of Alberta Hospital Liver Transplant Program was interrogated for perioperative data, which included the etiology of the preoperative liver disease and details of surgery (ie, date, type, operative and anhepatic time, and transfusions/fluid therapy). Individual patient medical records were reviewed for preoperative data (ie, location, acute physiology, laboratory values, mechanical ventilation, vasopressors, and receiving RRT) and postoperative data (ie, acute physiology, laboratory data, duration of mechanical ventilation, and duration and modality of RRT). Medical records were also reviewed for detailed data on liver disease severity (ie, MELD score and Child-Turcotte-Pugh score) and on intraoperative CRRT (ie, duration, blood flow, dialysate

3 RENAL SUPPORT FOR LIVER TRANSPLANTATION 75 TABLE 1. Preoperative Demographics and Baseline Characteristics for Liver Failure Patients Receiving Continuous Renal Replacement Therapy During Liver Transplantation (n 41) Baseline Characteristic Age [mean 50.4 (10.6) Male sex (%) 63 Etiology of primary liver disease (%) Alcohol 29.3 Hepatitis C virus 34.2 Hepatitis B virus 7.3 Non A-B-C hepatitis virus 4.9 Cryptogenic cirrhosis 7.3 Primary biliary cirrhosis 9.8 Primary sclerosing cholangitis Antitrypsin disease 2.4 Hemachromatosis 7.3 Drug toxicity* 4.9 Autoimmune hepatitis 2.4 Hepatocellular carcinoma 22 Repeat transplant (%) 17 Comorbidities Cardiovascular disease (%) 41.5 Respiratory disease (%) 41.5 Abbreviation: SD, standard deviation. *Drug toxicity included acetaminophen-induced and minocycline-induced liver failure. rate, replacement fluid rate, filter circuit changes, anticoagulation, and documented complications). Survival, need for RRT, and kidney function at 1, 3, and 6 months and 1 year were ascertained by a review of individual medical records, hospital administrative data, or the Liver Transplant Program database. Statistical Analyses Analysis was performed with Intercooled Stata (Stata Corp., College Station, TX). In the event of missing data values, data were not replaced. Normally or near normally distributed variables are reported as means with standard deviations (SDs) and compared by the Student t test. Nonnormally distributed continuous data are reported as medians with interquartile ranges (IQRs) and compared by the Mann-Whitney U test. Categorical data are reported as proportions and compared with Fisher s exact test. The standardized mortality ratio was calculated as the ratio of observed deaths to expected deaths. The probability of death was predicted with the preoperative MELD score. A P value of 0.05 was considered statistically significant for all comparisons. RESULTS A total of 636 patients received OLT during the 10-year study period, 41 (6.4%) of which were supported with intraoperative CRRT and composed the primary study TABLE 2. Summary of Preoperative Characteristics, Physiology, and Laboratory Parameters for Liver Failure Patients Receiving Continuous Renal Replacement Therapy During Liver Transplantation (n 41) Preoperative Feature MELD score [median 38 (31 43) Child-Pugh score [mean 11.4 (1.5) Child-Pugh class C (%) 90.2 ICU admission (%) 70.7 Mechanical ventilation (%) 58.5 Vasoactive agents (%) 48.7 Bilirubin [mmol/l; median 323 ( ) Platelets [10 9 cells/ml; 45 (27 51) median INR [median 2.0 ( ) PaCO 2 [mm Hg; mean 23 (14) Hemoglobin [g/l; mean 98 (16) Sodium [mmol/l; median 138 ( ) Potassium [mmol/l; 4.1 ( ) median Lactate [mmol/l; median 2.8 ( ) Serum creatinine [ mol/l; 172 ( ) median * egfr [ml/minute/m 2 ; 39.2 (22.7) mean Serum urea [mmol/l; 15 (9.4 22) median * Abbreviations: egfr, estimated glomerular filtration rate; ICU, intensive care unit; INR, international normalized ratio; IQR, interquartile range; MELD, Model for End-Stage Liver Disease; PaCO 2, partial pressure of carbon dioxide in arterial blood; SD, standard deviation. *The SI unit conversion for serum creatinine is 1 mg/dl 88.4 mol/l and for serum urea is 1 mg/dl mmol/l. Preoperative egfr for those not already receiving renal replacement therapy. cohort (Table 1). Details of the preoperative status, including liver and kidney function, are shown in Table 2. The majority were already receiving RRT preoperatively (63.4%, n 26), and nearly all were receiving CRRT (96%, n 25). A summary of contributing factors for the initiation of intraoperative CRRT is shown in Table 3. Intraoperative details, including the duration of surgery, anhepatic time, and fluid therapy, are shown in Table 4. All patients received an intraoperative transfusion of packed red blood cells and fresh frozen plasma, whereas 85.4% received platelets and only 43.1% received cryoprecipitate. A summary of details of intraoperative CRRT is shown in Table 5. CRRT was operational for 50% of the total operative time in 68% of

4 76 TOWNSEND ET AL. TABLE 3. A Summary of Preoperative Factors Contributing to the Decision To Initiate Intraoperative Continuous Renal Replacement Therapy (n 41) Preoperative Factor n (%) Illness severity MELD score (78) Vasopressor dependent 20 (48.7) Probable need for transfusion Hemoglobin 80 g/l 5 (12.1) Platelets cells/ml 29 (71.4) INR (56.1) Acute kidney injury 41 (100) Receiving RRT 26 (63.4) Hyperkalemia (K 5.0 mmol/ 5 (12.1) L) Dysnatremia (Na 150 or 9 (22) 130 mmol/l) Lactic acidosis ( 4.0 mmol/l) 7 (17) Abbreviations: INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; RRT, renal replacement therapy. TABLE 5. Summary of Continuous Renal Replacement Therapy During Liver Transplantation (n 41) Details Blood flow [ml/minute; median 125 ( ) RRT time [minutes; median 258 ( ) Number of filters [median (range)] 1 (1 4) Replacement fluid rate [ml/hour; 1000 ( ) median Bicarbonate added (%) 51.5 Dialysate rate [ml/hour; median 1000 ( ) Bicarbonate added (%) 35.5 Net delivered dose [ml/hour; 2000 ( ) median Intraoperative cumulative fluid balance (%) Neutral 8 (19.5) Negative 30 (73.2) Positive 3 (7.3) Abbreviations: IQR, interquartile range; RRT, renal replacement therapy. TABLE 4. Summary of Intraoperative Details During Liver Transplantation (n 41) Characteristic Surgical time [hours; mean 8.8 (2.8) Anhepatic time [minutes; 60 (55 76) median Transfusions [median Packed red blood cells 7 (5 10) (units) Fresh frozen plasma 10 (7 16) (units) Platelets (units) 8 (5 15) Cryoprecipitate (units)* 0 (0 8) Crystalloid [plasmolyte; ml; 3000 ( ) median Albumin [ml; median 25% (n 16) 275 ( ) 5% (n 4) 1250 ( ) Abbreviations: IQR, interquartile range; SD, standard deviation. *Only 18 received cryoprecipitate (median: 8 units; range: 5-24). patients (n 28), with the average duration representing 57% of the total operative time. In all, 92.7% of patients achieved an even or negative intraoperative fluid balance. Additional bicarbonate was added to the replacement fluid and dialysate in 35% of patients. Filter circuit clotting occurred in 40% of patients, and clotted filters and circuits were rapidly replaced. The reasons for filter failure were not described. Filter clotting was not associated with a significantly shorter duration of intraoperative CRRT (P 0.41). In those treated with no anticoagulation, 38% had filter circuit clotting; however, there was no significant difference in filter circuit clotting by anticoagulation type (P 1.0). For those with filter circuit clotting, there was no significant difference in the median number of units of packed red blood cells (P 0.29), units of fresh frozen plasma (P 0.20), or units of cryoprecipitate transfused (P 0.66); however, these patients did receive more units of platelets (P 0.03). There were no other significant complications associated with intraoperative CRRT described. Postoperatively, 32 patients (78%) were continued on CRRT in ICU for a median (IQR) duration of 5 (3-11) days. Of these patients, 24 (75%) were eventually transitioned to intermittent hemodialysis (IHD) and had a median (IQR) duration of treatment of 15 (4-39) days. Of those dead at 1 year, 7 patients (70%) had received IHD for a median (IQR) duration of 37 (4-44) days. Clinical outcomes are summarized in Tables 6 and 7. Despite the overall high preoperative risk for death in this cohort, crude survival was 97.6% at 1 month and 75.6% at 1 year. Renal recovery occurred in 100% of survivors by 1 year; however, the mean (SD) egfr was 54.7 (19.1) ml/minute/m 2, with 62.1% having an egfr 60 ml/minute/m 2. There were no statistical differences in survival or renal recovery to RRT independence when patients were stratified by preoperative RRT status. Of the 10 patients who had died at 1 year (24.4%), the median (IQR) survival was 116 (45-151) days. Based on a predicted mortality by MELD score of 75% (n 31) and the observed mortality of 24.3%, the estimated standardized mortality rate was At the time of

5 RENAL SUPPORT FOR LIVER TRANSPLANTATION 77 TABLE 6. Summary of Secondary Outcomes for Liver Failure Patients Receiving CRRT During Liver Transplantation Clinical Outcome Total (n 41) Survivor (n 31) Nonsurvivor (n 10) P Value Mechanical ventilation [days; median * 5 (2 10) 5 (2 9) 8 (3 24) 0.30 Duration of postoperative RRT [days; median CRRT (n 41) 3 (1 7) 3 (1 26) 6 (1 26) 0.33 IHD (n 24) 15 (4 39) 14 (7 35) 37 (4 44) 0.31 ICU length of stay [days; median 8 (5 20) 8 (5 20) 14 (4 28) 0.41 Hospital length of stay [days; median 43 (22 72) 34 (22 72) 55 (37 77) 0.28 Abbreviations: CRRT, continuous renal replacement therapy; ICU, intensive care unit; IHD, intermittent hemodialysis; IQR, interquartile range; RRT, renal replacement therapy. *Duration of postoperative mechanical ventilation. Stratified by survival status at 1 year. TABLE 7. Summary of Survival and Renal Recovery for Liver Failure Patients Receiving Continuous Renal Replacement Therapy During Liver Transplantation Outcome 1 Month 3 Months 6 Months 1 Year Survival (%) 40 (97.6) 37 (90.2) 34 (82.93) 31 (75.6) Survival and need for chronic RRT (%) 14 (35) 2 (5.4) 2 (6.6) 0 (0) Serum creatinine [ mol/l; median * 122 (78 138) 115 (91 137) 124 (95 157) 126 (95 170) egfr [ml/minute/m 2 ; mean * 49.9 (32.0) 57.9 (21.1) 54.6 (20.7) 54.7 (19.1) egfr 60 ml/minute/m 2 (%) Abbreviations: egfr, estimated glomerular filtration rate (Modification of Diet in Renal Disease equation); IQR, interquartile range; RRT, renal replacement therapy; SD, standard deviation. *Serum creatinine and egfr for those patients with renal recovery post liver transplantation. The SI unit conversion for serum creatinine is 1 mg/dl 88.4 mol/l. death, 60% had recovered sufficient kidney function to be independent of RRT. Of those independent of RRT, the median (IQR) egfr was 46 (36-65) ml/minute/m 2, with 4 patients having an egfr 60 ml/minute/m 2. DISCUSSION We performed a retrospective analysis of all liver failure patients who received OLT and were supported with intraoperative CRRT. Our objectives were to evaluate the logistics of intraoperative CRRT during OLT and to describe the associated postoperative clinical outcomes for this cohort with a high probability of death. We found that intraoperative CRRT during liver transplantation is logistically achievable. Second, we showed that no anticoagulation was generally needed to maintain filter circuit patency. Third, we found that intraoperative CRRT was associated with most patients achieving a neutral or negative fluid balance for the procedure. Fourth, we found that those alive at 1 year had all recovered sufficient kidney function to be independent of RRT. However, we also found that most had incomplete recovery with evidence of chronic kidney disease (CKD). Finally, we showed trends suggesting that nonsurvivors incurred longer durations of postoperative RRT support, mechanical ventilation, ICU stay, and hospital stay. Our findings suggest that intraoperative CRRT during OLT is feasible and safe. In our cohort, CRRT was applied for more than 50% of the total operative time (4-5 hours). Importantly, this was generally achieved without the use of systemic anticoagulation of the extracorporeal circuit (ie, unfractionated heparin) because of the concern for intraoperative bleeding complications. This practice is supported by observational data showing that CRRT can be successfully performed with acceptable filter circuit lifespan without anticoagulation. 3,13 Although a small number of patients in our study received anticoagulation with regional citrate or prostaglandin, there are theoretical concerns about the use of either of these methods for anticoagulation, including metabolic toxicity (ie, citrate lock) and hypotension. In 1 patient, the filter clotted on several occasions. Although all patients in our cohort were coagulopathic (by conventional measures such as INR and activated partial thromboplastin time) prior to surgery, 7 required a filter change intraoperatively because of circuit clotting. Although we did not have additional measures of global hemostatic function (ie, thromboelastogram), it is possible that these patients were in fact prothrombotic. 14,15 Overall, no adverse effects were described in these patients, and our experience would suggest that CRRT can be applied without anticoagulation for the duration of surgery. There were no other significant complications attributable to CRRT described. Our

6 78 TOWNSEND ET AL. data have also shown that, despite the administration of considerable intraoperative fluid therapy, the majority of patients completed surgery with a cumulative intraoperative neutral or negative fluid balance. We found 75.6% 1-year survival in our cohort. This would appear consistent with the survival described by Gonwa et al. 16 In an observational study of 2 distinct OLT cohorts, survival at 1 year for those with preoperative AKI requiring RRT was in the range of 74% to 90%, whereas for those whose course was complicated by AKI requiring RRT after OLT, survival at 1 year was only 42% to 55%. Moreover, in our cohort, only 78% required continuation of CRRT in the immediate postoperative period. The postoperative course for nonsurvivors in our cohort was more complicated than that for survivors and characterized by several key trends, including a longer period of mechanical ventilation and longer stays in both the ICU and hospital. In addition, for those who had died at 1 year, the duration of RRT was longer, 70% were temporarily transitioned to IHD, and 40% died with end-stage renal disease. In contrast, all survivors at 1 year post-olt in our cohort had recovered sufficient function to gain RRT independence. Despite this, survivors exhibited a high incidence of CKD, with approximately 62% having an egfr 60 ml/minute/m 2. This finding is consistent with previous investigations. In a small retrospective analysis of kidney function following OLT complicated by AKI, McCauley et al. 17 reported that 83% of patients had CKD after a median of 30 months. Our data suggest that the incidence of CKD at 1 year in this cohort is high and would indicate further investigation of the long-term kidney prognosis of peri-olt AKI (together with the potential nephrotoxic effects of immunosuppressive agents) is warranted. Our study has notable limitations. Importantly, our study is single-center, small, and retrospective. In addition, we do not have a control group of OLT patients with perioperative AKI who did not receive intraoperative CRRT for comparison. As such, we recognize that our study is largely descriptive and is prone to bias. However, 1 objective of our study was to demonstrate that intraoperative CRRT during OLT is logistically possible. Our study is the largest to date to describe intraoperative CRRT and, we believe, shows that it is, when indicated, both feasible and safe. We now believe that additional prospective controlled studies from multiple centers are needed to support the development of consensus criteria and guidelines for the use of intraoperative CRRT during OLT. In summary, we believe that this study establishes that intraoperative CRRT during OLT in critically ill patients with preoperative AKI can be safely achieved. We contend that this supportive therapy may be an important adjuvant measure, in particular for patients with preoperative AKI, for preserving intraoperative electrolyte, acid-base, and volume homeostasis; however, we recognize that further confirmatory investigations are needed. We also contend that if intraoperative CRRT during OLT is to be performed, a multidisciplinary and protocol-driven approach is required. REFERENCES 1. Wiklund RA. Preoperative preparation of patients with advanced liver disease. Crit Care Med 2004;32:S106 S Rettke SR, Janossy TA, Chantigian RC, Burritt MF, Van Dyke RA, Harper JV, et al. Hemodynamic and metabolic changes in hepatic transplantation. Mayo Clin Proc 1989; 64: Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, et al. Continuous renal replacement therapy: a worldwide practice survey. The Beginning and Ending Supportive Therapy for the Kidney (B.E.S.T. Kidney) Investigators. Intensive Care Med 2007;33: Davenport A, Will EJ, Davison AM, Swindells S, Cohen AT, Miloszewski KJ, Losowsky MS. Changes in intracranial pressure during haemofiltration in oliguric patients with grade IV hepatic encephalopathy. Nephron 1989;53: Bellomo R, Harris C, Kang Y, Daniel E, Fung JJ, Bronsther O. Combined veno-venous bypass and high volume hemofiltration during orthotopic liver transplantation. ASAIO J 1993;39: Korbet SM, Casey C, Rodby RA, Williams W. The use of continuous arteriovenous hemofiltration in orthotopic liver transplantation. Clin Nephrol 1995;43: Salord F, Bailly MP, Gaussorgues P, Workineh S, Pouyet M, Robert D. Continuous arteriovenous haemodialysis during emergency hepatic retransplantation: two case reports. Intensive Care Med 1990;16: Kutsogiannis DJ, Mayers I, Chin WD, Gibney RT. Regional citrate anticoagulation in continuous venovenous hemodiafiltration. Am J Kidney Dis 2000;35: Kutsogiannis DJ, Gibney RT, Stollery D, Gao J. Regional citrate versus systemic heparin anticoagulation for continuous renal replacement in critically ill patients. Kidney Int 2005;67: Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999;130: Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding esophageal varices. Br J Surg 1973;60: Kamath PS, Wiesner RH, Malinchoc M, Kremers W, Therneau TM, Kosberg CL, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001;33: Uchino S, Fealy N, Baldwin I, Morimatsu H, Bellomo R. Continuous venovenous hemofiltration without anticoagulation. ASAIO J 2004;50: Wanless IR, Wong F, Blendis LM, Greig P, Heathcote EJ, Levy G. Hepatic and portal vein thrombosis in cirrhosis: possible role in development of parenchymal extinction and portal hypertension. Hepatology 1995;21: Kang Y. Thromboelastography in liver transplantation. Semin Thromb Hemost 1995;21(suppl 4): Gonwa TA, Mai ML, Melton LB, Hays SR, Goldstein RM, Levy MF, Klintmalm GB. Renal replacement therapy and orthotopic liver transplantation: the role of continuous veno-venous hemodialysis. Transplantation 2001;71: McCauley J, Van Thiel DH, Starzl TE, Puschett JB. Acute and chronic renal failure in liver transplantation. Nephron 1990;55: