Two-stage total hepatectomy and liver transplantation for acute deterioration of chronic liver disease: A new bridge to transplantation

Authors

  • Michael J. Guirl,

    Corresponding author
    1. Department of Internal Medicine, Division of Gastroenterology, Methodist Dallas Hospital, Dallas, TX
    • Department of Internal Medicine, Division of Gastroenterology, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, TX 75246
    Search for more papers by this author
    • Telephone: 214-820-2671; FAX: 214-820-3473

  • Jeffrey S. Weinstein,

    1. The Liver Institute, Methodist Dallas Hospital, Dallas, TX
    Search for more papers by this author
  • Robert M. Goldstein,

    1. Department of Transplantation Services, Division of Transplant Surgery, Baylor University Medical Center, Methodist Dallas Hospital, Dallas, TX
    Search for more papers by this author
  • Marlon F. Levy,

    1. Department of Transplantation Services, Division of Transplant Surgery, Baylor University Medical Center, Methodist Dallas Hospital, Dallas, TX
    Search for more papers by this author
  • Goran B. Klintmalm

    1. Department of Transplantation Services, Division of Transplant Surgery, Baylor University Medical Center, Methodist Dallas Hospital, Dallas, TX
    Search for more papers by this author

Abstract

Two-stage total hepatectomy and liver transplantation has been reported for acute liver disease such as fulminant hepatic failure, primary graft failure, severe hepatic trauma, and spontaneous hepatic rupture secondary to hemolysis, elevated liver function tests, low platelets syndrome, and preeclampsia. This is the first report of patients with cirrhosis to undergo a 2-stage total hepatectomy and liver transplantation. From 1984 to 2002, our institution performed 2008 orthotopic liver transplantations. We identified 4 patients with chronic liver disease who underwent a 2-stage hepatectomy and liver transplantation. This is a retrospective review of these 4 patients and a review of the literature on this procedure. All 4 patients were young men with an age range of 29–31 years and had underlying cirrhosis as well as a previous transjugular intrahepatic portosystemic shunt (TIPS)procedure. Acute decompensation fulfilling Ringes' criteria for toxic liver syndrome secondary to an upper gastrointestinal bleed occurred in all patients. The approximate average time between hepatectomy and liver transplantation was 20 hours (range: 8–42 hours). In all cases, the explanted liver showed histological changes of acute hepatic necrosis within the background of cirrhosis. After hepatectomy, vasopressor requirements were well documented in 2 patients. For 1 patient, there was a clear improvement in their hemodynamic status. The mean hospital stay of the 4 patients was 63 days. All patients were discharged from the hospital and are alive and well with adequate liver function at 6 to 37 months follow-up. Two-stage total hepatectomy and liver transplantation may be a life-saving procedure in highly selected cirrhotic patients with acute hepatic decompensation and multiorgan dysfunction. (Liver Transpl 2004;10:564–570.)

Immediate allocation of a donor organ is not always possible for the majority of patients who are awaiting liver transplantation. This is especially true for those who are either severely or acutely ill. In such situations, it would be helpful to have other “bridges” to liver transplantation until a suitable organ becomes available.

A 2-stage total hepatectomy and liver transplantation was first reported by Ringe et al. in 1988.1 The operation was performed in a patient with primary graft failure complicated by multiorgan dysfunction. The primary goal of hepatectomy is temporary hemodynamic and metabolic stabilization until an organ becomes available for liver transplantation. In the first stage of the procedure, the necrotic liver is removed after dissection of the hilar structures and hepatic veins. With the inferior vena cava retained, an end-to-side portocaval anastomosis, which allows systemic and portal venous drainage and decompression, is constructed. In the second stage, orthotopic liver transplantation using standard techniques is performed.1

Since the initial report, 2-stage hepatectomy and liver transplantation has been reported in patients with fulminant hepatic failure, primary graft failure, severe hepatic trauma, and spontaneous hepatic rupture secondary to hemolysis, elevated liver function tests, low platelets syndrome, and preeclampsia.1–12 All of the patients had acute liver failure complicated by toxic liver syndrome. The criteria for this syndrome was defined by Ringe et al.8 as complete liver necrosis associated with cardiovascular shock, renal, and perhaps respiratory failure requiring vasopressor support, hemodialysis, and mechanical ventilation. Although the toxic liver syndrome was first defined in patients with acute liver failure, similar findings may also occur in chronic liver disease patients suffering from an episode of acute decompensation. The following is the first report of 2-stage total hepatectomy and liver transplantation in 4 patients with underlying cirrhosis who had an acute decompensation after an episode of acute gastrointestinal bleeding.

Abbreviations:

TIPS, transjugular intrahepatic portosystemic shunt; HD, hospital day; CVVHD, continuous venovenous hemodialysis; UGIB, upper gastrointestinal bleed.

Case Reports

Demographics, periprocedural data, and outcomes for the 4 patients are listed in Tables 1 and 2. Laboratory data and vasopressor requirements for which data is available is illustrated in Figs. 1 and 2.

Table 1. Patient Demographics
Patient No.Age (yr)/GenderLiver Disease EtiologyCTP ClassUGIB EtiologyOpen TIPSMinnesota Tube Placed
  1. Abbreviations: CTP, Child-Turcotte-Pugh; UGIB, upper gastrointestinal bleed; TIPS, transjugular intrahepatic portosystemic shunt.

130/MBudd-Chiari syndromeBGastric varicesNoYes
231/MHepatitis CCProbable gastric varicesNoYes
329/MPrimary sclerosing cholangitisCEsophageal varicesYesYes
429/MCryptogenicCPortal hypertensive gastropathyYesNo
Table 2. Periprocedural Details and Patient Outcomes
Patient No.CVVHD Prior to HepatectomyCVVHD Total Time (d)Anhepatic PeriodABO MatchAcute RejectionHospital Stay (d)Follow-up (months)Retransplantation
  1. Abbreviations: CVVHD, continuous venovenous hemodialysis.

1Yes6Approx 42 hYesNo226No
2No2517 h and 20 minYesYes15017No
3Yes1Approx 11 hYesYes4137No
4Yes177 h and 56 minNoYes4025Yes
Figure 1.

Laboratory data. A, 48 hours before UGIB; B, after UGIB; C, immediately before hepatectomy; D, immediately after hepatectomy; E, 24 hours after liver transplantation; F, at the time of discharge. Data was not available at all time points.

Figure 2.

Vasopressor requirements for Patients 1 and 2. Dopamine (μg/kg/min, solid line), phenylephrine (μg/min, dashed line), and norepinephrine (μg/min, dotted line); A, Initiation of vasopressor support; B, CVVHD started; C, immediately before hepatectomy; D, immediately after hepatectomy; E, after liver transplantation. Complete vasopressor data is not available for Patients 3 and 4.

Patient 1

A 30-year-old man with a history of Budd-Chiari syndrome and cirrhosis secondary to factor V leiden mutation had a history of nonbleeding gastroesophageal varices, refractory ascites, and a recurrent hepatic hydrothorax. In May 2002, a month before his admission, a transjugular intrahepatic portosystemic shunt (TIPS) was placed. Initially, he had some improvement in his fluid retention but his TIPS occluded and his symptoms recurred. For technical reasons, his TIPS was not amenable to revision. He was subsequently hospitalized with dyspnea secondary to a hepatic hydrothorax. He was treated with thoracentesis, paracentesis, diuretics, and salt poor albumin. On hospital day (HD) #6, he developed melena and hematemesis. Esophagogastroduodenoscopy revealed bleeding gastric varices. He was treated successfully with octreotide and supportive care. On HD #9, he had recurrent variceal bleeding. A Minnesota tube was placed and he was intubated and placed on mechanical ventilation for airway protection. His bleeding stopped; however, he had a marked increase in his liver function tests along with worsening coagulopathy consistent with ischemic hepatitis. He also developed hypotension associated with a refractory metabolic acidosis and anuria. He was treated with vitamin K, fresh frozen plasma, cryoprecipitate, sodium bicarbonate-buffered replacement fluids, continuous venovenous hemodialysis (CVVHD), and multiple vasopressor agents. His condition failed to stabilize and he underwent a total hepatectomy with an end-to-side portocaval shunt. After hepatectomy, he was taken back to the intensive care unit where he continued to receive supportive care including CVVHD. His vasopressor requirements decreased and his metabolic acidosis improved. He was anhepatic for approximately 42 hours before an ABO compatible donor organ became available and he then underwent orthotopic liver transplantation. The explanted liver histology showed submassive hepatocellular necrosis with underlying cirrhosis. After liver transplantation, he continued on CVVHD for 4 days. He had no complications related to the 2-stage procedure. He was hospitalized for 22 days. At 6 months follow-up, he is alive and well with adequate liver and renal function.

Patient 2

A 31-year-old man with a history of hepatitis C and cirrhosis complicated by ascites and nonbleeding gastroesophageal varices had a TIPS placed in March 2001 for refractory ascites. The TIPS occluded and for technical reasons was not amenable to any further revision. He presented 1 month later with fatigue and an elevated bilirubin felt secondary to worsening hepatic function. On HD #4, he developed renal insufficiency with worsening ascites. On HD #9, he developed massive hematemesis. Esophagogastroduodenoscopy revealed blood in his lower esophagus, stomach, and duodenum. A bleeding source was not identified, although his gastric fundus and cardia were never completely visualized. A Minnesota tube was placed and an octreotide infusion was initiated for gastrointestinal bleeding presumably related to gastric varices. Despite control of his bleeding, he developed a marked increase in his liver function tests and worsening coagulopathy consistent with ischemic hepatitis. He also developed hypotension associated with a refractory metabolic acidosis and anuria. He then had a cardiac arrest secondary to pulseless electrical alternans and was resuscitated with cardiopulmonary resuscitation, mechanical ventilation, atropine, and epinephrine. He was treated with vitamin K, fresh frozen plasma, cryoprecipitate, sodium bicarbonate-buffered replacement fluids, and multiple vasopressor agents. His condition failed to stabilize, and he then underwent a total hepatectomy with an end-to-side portocaval shunt. After hepatectomy, he was taken back to the intensive care unit where he received supportive care including CVVHD. His vasopressor requirements did not change significantly but his metabolic acidosis improved. He was anhepatic for 17 hours and 20 minutes before an ABO compatible donor organ became available, and he then underwent orthotopic liver transplantation. The explanted liver histology showed submassive hepatocellular necrosis with underlying cirrhosis. After liver transplantation, he required CVVHD for 25 days. He had numerous complications including acute rejection; multiple hepatic and splenic infarcts with necrosis, peritonitis, and sepsis; respiratory failure requiring reintubation and mechanical ventilation; nosocomial pneumonia; pleural and pericardial effusions requiring thoracentesis and pericardiocentesis; and papillary stenosis requiring endoscopic retrograde cholangiopancreatography and sphincterotomy. He was hospitalized for 150 days. After hospital discharge, he required several months of physical rehabilitation. At 17 months follow-up, he is alive and well with adequate liver function and chronic renal insufficiency with a glomerular filtration rate of 28 mL/hr.

Patient 3

A 29-year-old man with a history of ulcerative colitis, primary sclerosing cholangitis, and cirrhosis complicated by ascites was in his usual state of health until September 1999 when he developed progressive jaundice. He underwent an endoscopic retrograde cholangiopancreatography for a suspected dominant biliary stricture. During the procedure, acute esophageal variceal bleeding was noted. He was treated successfully with sclerotherapy but had recurrent bleeding and underwent a successful TIPS. On HD #4, he had recurrent bleeding with evidence of an occluded TIPS that was revised. His bleeding persisted and a Minnesota tube was placed and an octreotide infusion was initiated. Despite control of his bleeding, he developed an ischemic hepatitis with a marked increase in his liver function tests and worsening coagulopathy. He then developed refractory hypotension associated with metabolic acidosis, anuria, and, ultimately, a cardiac arrest secondary to ventricular tachycardia. He was resuscitated with cardiopulmonary resuscitation, mechanical ventilation, cardioversion, and a lidocaine drip. He was treated with vitamin K, fresh frozen plasma, cryoprecipitate, sodium bicarbonate-buffered replacement fluids, CVVHD, and multiple vasopressor agents. His condition failed to stabilize and he then underwent a total hepatectomy with an end-to-side portocaval shunt. He did not leave the operating room and was anhepatic for approximately 11 hours before an ABO compatible donor organ became available. The explanted liver histology showed massive hepatocellular necrosis with underlying cirrhosis. After liver transplantation, he continued on CVVHD for 1 day. His postoperative course was complicated by acute rejection, a bleeding gastric ulcer requiring endoscopic therapy, and a pericardial effusion requiring pericardiocentesis. He was hospitalized for 41 days. At 37 months follow-up, he is alive and well with adequate liver and renal function.

Patient 4

A 29-year-old man had a history of cryptogenic cirrhosis complicated by hepatic encephalopathy and refractory ascites. To manage his ascites, he had an elective TIPS in October 2000 and subsequent revision. He was admitted with recurrent hepatic encephalopathy. On HD #2, he developed hematemesis. An esophagogastroduodenoscopy revealed diffuse bleeding from the gastric body secondary to portal hypertensive gastropathy. Ultrasound with Doppler showed a patent TIPS. He was treated with octreotide and supportive care. There was no significant change in his liver function tests; however, he developed worsening coagulopathy, anuria, pulmonary failure, and hypotension. He was treated with vitamin K, fresh frozen plasma, cryoprecipitate, sodium bicarbonate-buffered replacement fluids, CVVHD, mechanical ventilation, and multiple vasopressor agents. His condition failed to stabilize, and he then underwent a total hepatectomy with an end-to-side portocaval shunt. He did not leave the operating room and was anhepatic for 7 hours and 56 minutes before an ABO incompatible donor organ became available and he then underwent orthotopic liver transplantation. The explanted liver histology showed submassive hepatocellular necrosis with underlying cirrhosis. After liver transplantation, he continued on CVVHD for 10 days. His immediate postoperative course was complicated by acute rejection and papillary stenosis requiring endoscopic retrograde cholangiopancreatography and sphincterotomy. He was hospitalized for 40 days. His long-term postoperative course was complicated by multiple biliary strictures likely secondary to ABO incompatibility. Twenty-four months after his initial liver transplantation, he underwent successful retransplantation. At 25 months follow-up, he is alive and well with adequate liver and renal function.

Discussion

This is the first report on the use of a 2-stage total hepatectomy and liver transplantation in a group of highly selected cirrhotic patients. We describe 4 patients with cirrhosis complicated by acute hepatic and multiorgan failure manifested by ischemic hepatitis, renal dysfunction, hypotension, and metabolic acidosis who had some degree of hemodynamic and metabolic stabilization after total hepatectomy allowing for a successful second stage liver transplantation. None of the patients were treated with any other supportive interventions previously reported in patients with acute liver failure including N-acetylcysteine,13 steroids, extracorporeal liver assist devices,14 or xenotransplantation.15

In 1988, Ringe et al.1 reported the first case of a patient with primary graft and multiorgan failure undergoing a 2-stage total hepatectomy and liver transplantation. Since then, this procedure has been reported in nearly 50 patients with acute liver failure. The mortality of acute hepatic necrosis with multiorgan failure is nearly 100%.16–17 The etiology of multiorgan failure in acute liver failure is unknown but is likely multifactorial. It has been hypothesized that toxic metabolites released from the necrotic liver or vasoactive effects play a major pathophysiological role.18 The mediators have not been identified, but cardiosuppressive factors released by the liver in the hepatic ischemia-anoxia rat model have been described.19 Purposed cytokines include Kupffer cell derived tumor necrosis factor α, which is known to cause cardiovascular and pulmonary instability.20–21 In addition, at the time of liver transplantation, surgeons have long observed an improvement in hemodynamic stability and a diminished degree of metabolic acidosis after interrupting the blood supply to the native liver.1, 22

There are several important issues that this case series highlights. First, our patients were all young and had no comorbidities except cirrhosis and its complications. Most of the published cases involving a 2-stage total hepatectomy and liver transplantation have included patients younger than 40 years old.1–12 This may be due to the bias of the transplant team as well as the fact that most patients with fulminant hepatic failure, spontaneous hepatic rupture, or severe hepatic trauma are relatively young.1–12 Second, all of our patients had acute decompensation and multiorgan dysfunction that was precipitated by a UGIB related to portal hypertension. Each of our patients underwent a TIPS procedure before their UGIB and the TIPS may have played a role in the pathogenesis of acute hepatic decompensation. One could hypothesize that suboptimal hepatic blood flow due to a patent or dysfunctional TIPS placed our patients at risk for hepatic ischemia in the setting of hemorrhagic shock. Third, 3 patients required CVVHD before hepatectomy. CVVHD has been reported to result in greater circulatory stability, less change in intracranial pressure, and less cerebral edema when compared with standard hemodialysis.23 In addition, CVVHD has been shown to facilitate fluid removal and improve metabolic imbalance in patients with fulminant hepatic failure and following hepatectomy with portocaval shunting.24 Thus, we believe the use of CVVHD was essential to our success in this series of patients.

All patients met Ringes' criteria for toxic liver syndrome defined as complete liver necrosis associated with cardiovascular shock, renal, and perhaps respiratory failure requiring vasopressor support, hemodialysis, and mechanical ventilation.8 Moreover, histological evidence for cirrhosis along with either massive or submassive necrosis was confirmed in all 4 patients. These findings support the diagnosis of toxic liver syndrome and demonstrate that acute liver failure can occur in those who are healthy as well as in those with chronic liver disorders.25

In the case series by Oldhafer et al.,6 all long-term survivors had anhepatic times of less than 24 hours. Interestingly, our patient who was anhepatic for more than 24 hours not only survived but had the shortest hospital stay and did not have acute rejection. The longest period of a patient having a 2-stage hepatectomy and liver transplant with long-term survival is 48 hours.5 In addition, 6 of the 7 long-term survivors in the Oldhafer et al.6 series had catecholamine treatment discontinued after hepatectomy. Only 1 patient who continued to require vasopressor support had a long-term survival.6 In our series, all patients had improvement in acidosis and 2 patients (for whom data is available) continued to require vasopressor support albeit at a lower dose. Therefore, it may be suggested that all the short-term benefits of total hepatectomy can be better gauged by improvement in acidosis and the absence of increased vasopressor needs as opposed to decreasing or discontinuing vasopressors.

In the largest series to date, consisting of 32 patients with acute liver failure who underwent 2-stage total hepatectomy and liver transplantation, the long-term survival rate was approximately 30%.8 None of the patients who died before liver transplantation showed signs of metabolic or hemodynamic stabilization after hepatectomy. Nineteen patients went on to liver transplantation with 9 dying 1 to 46 days after transplantation, most secondary to sepsis. Of the 10 patients with long-term survival, 3 patients died 3 to 22 months after liver transplantation from causes not related to the procedure. In contrast, all patients in our series are alive and well with normal liver function at 6 to 37 months follow-up. Patient 4 did require retransplantation 24 months after his first transplantation secondary to biliary strictures presumably because he received an ABO incompatible liver allograft. The prolonged hospitalization seen in our patient group likely reflects the severity of their illness at the time of liver transplantation. The better survival in our series may be related to patient selection and to improvements in supportive care, such as CVVHD, which have evolved over the last decade.

The aim of this report was to demonstrate that total hepatectomy may be a potential “bridge” to liver transplantation in highly selected patients with cirrhosis who acutely decompensate and develop complications related to a toxic liver syndrome. The retrospective nature of this report, the small number of patients, the incomplete data, and the use of multiple therapeutic interventions all make it impossible to establish that there is a definite role for total hepatectomy in this situation. Each patient in this case series had a rapid deterioration in their clinical course resulting in multiple complications that led to multiple therapeutic interventions over a short period of time. Intensive critical care, CVVHD, and total hepatectomy were all factors that may have contributed to the successful outcomes in these 4 patients. However, all of the patients were receiving intensive critical care, and 3 patients were on CVVHD before hepatectomy. Therefore, it is unlikely that the observed improvement in inotropic use and acidosis after hepatectomy could have resulted from these interventions alone. It is our belief that these successful outcomes were primarily related to the decision to perform a total hepatectomy before the onset of a terminal state of irreversible multiorgan failure.

Many questions pertaining to this approach in severely ill cirrhotic patients remain unanswered. This procedure should be considered when there is no hope for immediate liver transplantation or for recovery in the presence of a failing liver.7 The precise indications for the procedure and criteria before hepatectomy need to be established. However, based on our limited experience, this procedure appears to be successful in stable young cirrhotic patients who develop a toxic liver syndrome after an acute UGIB from portal hypertension. Further clinical and laboratory work should be concentrated on other supportive measures such as bioartificial livers and xenotransplantation. Until then, 2-stage total hepatectomy and liver transplantation may offer a reasonable chance to rescue highly selected patients with end-stage liver disease.

Acknowledgements

The authors thank Carol A. Santa Ana for preparing the figures.

Ancillary