Presentation, diagnosis, and management of early hepatic venous outflow complications in whole cadaveric liver transplant

Patient Group: Identifying, Diagnosing, and Definition of Early HVOO

A retrospective analysis of 777 consecutive LTs over a 5-year period (August 2007 to August 2012) at a single institution was undertaken. Pediatric recipients (<16 years), living-related, auxiliary, and reduced/split livers were excluded from the analysis. Clinical data were extracted from prospectively maintained transplant and theatre databases, then correlated with electronically documented clinical outcome (iSOFT Clinical Manager Client, version 1.4.579 Computer Sciences Corporation, USA) and supplemented where need be with review of traditional article clinical notes. The study was performed with institutional ethical approval and within the remit of the ethical guidelines of the 1975 Declaration of Helsinki.

Recipients with potential early outflow problems were identified as recipients with problematic ascites, requiring diuretics (spironolactone and/or frusemide) in the first 3 months after LT, as recorded on discharge summaries, clinic follow-up letters, or identified from their listing diagnosis for retransplant. Investigations for recipients with problematic ascites typically included ascitic fluid analysis for albumin, amylase, triglycerides, bacterial culture, cell count, and cytology. Screening ultrasound was performed for vascular anatomy patency and flow direction, supplemented with liver biopsy if required. A computed tomography (CT) scan was then used for visualization of the vascular implanting anatomy and graft appearance. If preceding investigations remained negative and outflow remained a diagnostic concern, then venography and pressure studies were performed.14-17 At the time of venography, transjugular liver biopsy was often performed if there was no recent biopsy. Definitive diagnosis of HVOO was made if the wedged HV pressure was greater than the sinusoidal pressure, ie, >12 mm Hg.18

Surgical Techniques

Donation after brain death (DBD) and donation after cardiac death (DCD) livers were procured using standard techniques. Only controlled DCD livers were used (category 3 and 4) according to Maastricht classification19 and a modified Casavilla's super-rapid technique was employed, with emphasis on rapid donor hepatectomy.20, 21 Institutionally, the maximum accepted warm ischemia time (from oxygen saturations of 70% or systolic pressure of 50 mm Hg to aortic cannulation) was 30 minutes. The preservation solution was University of Wisconsin in all cases.

Implanting techniques used during the period of study included both caval replacement and PB. In PB implantation, 3 HV, 2 HV with extension, side-to-side, and cavostomy/triangulation variants were used. In PB implantation, a TPCS was typically fashioned.3 Institutionally, veno-veno (VV) bypass is no longer routine but was used on a selective basis where the recipient was not tolerant of a trial cross clamp as reflected by mean arterial pressures of less than 60 mm Hg or if there was excessive bleeding during hepatectomy. In cases of domino transplant, institutional practice was to put the amyloid donor on VV bypass to facilitate hepatectomy en bloc with cava. The amyloid liver would then be implanted using the PB technique, and the amyloid donor would have caval implantation.

Operative cavoplasty for HVOO involved Pringle of the hilus, clamping of the suprahepatic and infrahepatic cava. The caval anastomosis was then taken down, and the common orifice of the recipient HVs and donor cava were refashioned to facilitate triangulation of the anastomosis that was sutured with 4/0 Prolene. VV bypass facilities were available if needed.

Statistical Analysis

Results are reported as mean ± standard deviations or percentages. Continuous variables were compared using the Wilcoxon rank sum test and categorical variables were entered into contingency tables to be evaluated by the Pearson chi-square and Fisher's exact test. Statistical significance was set at P < 0.05. Analysis was performed using SPSS, version 17 (IBM, Armonk, NY).

Transplant Activity

Over the 5-year period of study 777 adult transplants were done, of which 25 were living related, 49 right lobes, 2 left lobes, and 6 auxiliary. The remainder (n = 695) received whole cadaveric livers of which DCD made up 24%, DBD made up 75%, and domino made up 1% of grafts used (see Table 1). Considering only recipients of whole livers, the average age was 55 ± 2 years with an average Model for End-Stage Liver Disease (MELD) score of 14 ± 2. Retransplantation made up 10% (n = 69/695) of transplant activity. Chronic liver disease (CLD; 85%, n = 592/695) was the main indication for transplant and acute liver failure (ALF; 15%, n = 103/695) accounted for the remainder (see Table 1). In CLD, the leading reasons for transplant were alcohol-related liver disease (23%, n = 136/592), hepatitis C virus (HCV; 17%, n = 101/592), primary sclerosing cholangitis (PSC; 8%, n = 50/592), primary biliary cirrhosis (8%, n = 47/592), and hepatocellular carcinoma (HCC; 22%, n = 131/592).

Implantation Techniques

Table 1 summarizes the main implanting techniques used in the time period of study. PB was the dominant approach for implantation in 88% of cases (n = 613/695) over caval replacement (12%, n = 82/695), and of the PB variant with 2 HVs with extensions was favored (69%, n = 423/613), followed by 3 HVs (29.7%, n = 182/613) with triangulation (0.5%, n = 3/613), and side to side (0.8%, n = 5/613). VV bypass was used in 8% (n = 55/695) of transplants over the time period studied, typically being used in the retransplants (31%, n = 17/55), acute Budd-Chiari (15%, n = 8/55), and familial amyloid polyneuropathy (18%, n = 10/55). In ALF (n = 103), a PB approach was preferred (91%, 94/103) with caval replacement being used only occasionally (9%, 9/103). ALF recipients, where caval replacement was the favored implanting technique, was in the setting of acute Budd-Chiari (n = 8) or an early retransplant for HVOO (n = 1).

Early Ascites After Transplant

Of the 695 recipients who received a whole liver, 3% (n = 20/695) were identified with symptomatic ascites needing medical management and investigation in the first 3 months after transplant. The majority were still inpatients, 75% (n = 15/20), presenting with either ongoing high drain losses (n = 8) and/or abdominal distension (n = 11), and 1 patient was associated with a pleural effusion. The readmissions with symptomatic ascites, 25% (n = 5/20), typically occurred at a month after transplant with the presenting complaint of abdominal distension. After further investigation (CT, pressure studies, and biopsy), just more than one-third (35%) of these early cases of ascites were confirmed to have HVOO (n = 7). The remaining diagnoses of early ascites after transplant were initial poor graft function (n = 3), chylous fluid loss (n = 2), no clear diagnosis (n = 4), veno-occlusive disease (n = 2), pancreatitis (n = 1), and portal vein stenosis (n = 1). These non-HVOO early cases of ascites were managed medically, and the portal vein stenosis was stented by interventional radiology.

Characteristics and Management of Early HVOO

Seven cases of early HVOO were identified; these cases were dispersed throughout the 5-year period of study (Table 2 summarizes recipient and donor clinical characteristics). Overall, the incidence of early HVOO was 1% (7/695) with diagnosis being confirmed at 35 ± 23 days after transplant based on CT, pressure studies, and biopsy. Of these, graft function was normal in 1 with the remaining cases having mild cholestasis at the time of diagnosis. All had preserved renal function before commencement of diuretics (estimated glomerular filtration rate > 60). Typical CT findings for the diagnosis of HVOO were a heterogeneous liver with nonvisualization of HVs and normal inflow vasculature (see Fig. 1). The average wedge pressure was 22 ± 10.5 mm Hg (see Fig. 2 for illustrative case) and characteristic histology findings were perivenular cell drop out and hemorrhagic congestion (see Fig. 3).

Considering implantation technique, the cases of early HVOO that were identified occurred when 2 HVs with extension PB had been used, giving an incidence of 1.6% (7/423) for that technique. Statistically, with the present study design and numbers for each implanting technique, it cannot be demonstrated that 2 HVs with extension gives a higher risk of early HVOO (P = 0.114). In just under one-third of HVOO cases (29%, n = 2/7) medical management with diuretics was successful with the remainder requiring surgical management (71%, n = 5/7). Operative cavoplasty was successful in 80% of cases (n = 4/5). The HVOO cases that were successfully managed by cavoplasty were all inpatients and revision was within a month of transplant (see Table 3 for summary of HVOO clinical, diagnostic findings, and management strategy).

The 1 patient for whom cavoplasty was unsuccessful was readmitted 6 weeks after transplant presenting with tense ascites and diuretic-related renal dysfunction. The indication for transplant was HCC on a background of HCV. There was no underlying procoagulant disorder identified in the recipient and the cause of death in the donor was an intracerebral bleed. Figures 1 and 2 show the CT and pressure studies that supported the diagnosis of HVOO. A right lobe transjugular biopsy at the time of the pressure studies also corroborated the radiological and clinical diagnosis of HVOO. Histologically the findings were of centrilobular patchy acute hemorrhage with centrilobular necrosis. There was no associated thrombosis, rejection, or steatosis (see Fig. 3). Exploratory laparotomy for cavoplasty was decided upon, but at the time of the operation, the finding of severe portal hypertension combined with a large, congested liver led to the procedure being abandoned. Interventional radiology then deployed stents in the cava and HV in an attempt to relieve HVOO (see Fig. 4). After 3 days, the stents thrombosed despite anticoagulation (see Fig. 5). The patient was relisted and successfully retransplanted a week later. The implanting technique was caval replacement on VV bypass.