EXPERIMENTAL AND CLINICAL TRANSPLANTATION

Objectives: The effect of portal vein thrombosis on the progression of liver disease is controversial, with no consensus on optimal treatment. We aimed to assess how portal vein thrombosis affects wait list outcomes, identify risk factors associated with its development while on a wait list, and assess its effects on patient and graft survival.

Materials and Methods: This US-based retrospective cohort study analyzed 134 109 adult patients on wait lists for or undergoing primary orthotopic liver transplant between January 2002 and June 2014. Rate of portal vein thrombosis development, time from entry on wait list to transplant, comparisons of wait list drop-off rates between patients with versus those without portal vein thrombosis, risk factors associated with its development while on a wait list, and its effects on patient and graft survival were analyzed.

Results: We found that the rate of portal vein thrombosis at listing increased. Patients with the disease at listing were more likely to be removed from wait lists because of being too sick. Portal vein thrombosis at listing was an independent risk factor for being removed from a wait list. Of 63 265 patients who underwent primary orthotopic liver transplant, those with the disease were more likely to have higher Model for End-Stage Liver Disease scores and incidence of nonalcoholic steatohepatitis and diabetes mellitus. Portal vein thrombosis had a negative effect on patient and graft survival. Nonalcoholic steatohepatitis, body mass index, diabetes, and hepatocellular carcinoma were identified as risk factors for its development.

Conclusions: Portal vein thrombosis represents an increasing management and outcome burden in liver transplant. Having this disease at listing and/or at time of transplant is associated with worse patient and graft survival. Nonalcoholic steatohepatitis and hepato-cellular carcinoma are among the biggest risk factors for its development while on a wait list.

Key words : Diabetes mellitus, Hepatocellular carcinoma, Nonalcoholic steatohepatitis

Liver transplant (LT) in patients with portal vein thrombosis (PVT) remains an important challenge and historically has been considered to be an absolute contraindication for LT. The first successful LT in a patient with PVT was reported in 1985, and several surgical and medical strategies have been reported since then. The prevalence and incidence of PVT vary with different diagnostic methods and stages of cirrhosis from < 0.6% in those with compensated cirrhosis1 to 28% in patients listed for LT.^2-5 In patients with cirrhosis without PVT who are entering an LT wait list, the 1-year de novo PVT rate is 7.4% to 8.4%.^2-6 In the most recently published paper using the Organ Procurement and Transplant Network (OPTN) database, 4.9% of patients without PVT at the time of listing were found to have new PVT at the time of transplant after a median wait time of 3 months.⁷

The main known risk factors for the development of PVT are cirrhosis with associated hypercoagulable state and low portal venous flow in the setting of portal hypertension.^8,9 Acquired and inherited clotting abnormalities may also play a role at a lower scale.

Well known is the post-LT association between PVT and increased operative time, higher red blood cell transfusion requirements, higher rates of reoperation, longer intensive care unit and hospital stay, and worsening patient and graft survival.^7-10 However, the effects of PVT on the progression of liver disease is quite controversial, and there is no current consensus on the optimal treatment of patients with PVT and cirrhosis who are awaiting transplant.

The objectives of this study were as follows: (1) to assess the impact of PVT on wait-list outcomes, (2) to identify risk factors associated with the development of PVT while on a wait list, and (3) to assess how PVT affects patient and graft survival after LT.

Materials and Methods

We conducted a retrospective cohort analysis of adult patients in the United States on wait list for primary orthotopic liver transplant (OLT) or undergoing primary OLT between January 2002 and June 2014. Exclusion criteria for the analysis of wait list outcomes were age younger than 18 years, listing for multiorgan transplant, and listing for a liver retransplant. Exclusion criteria for post-OLT outcomes included multiorgan transplant, donations from living donors, and the use of split livers.

We analyzed the following: (1) rate of PVT during the study period, both at the time a patient entered a wait list (based on imaging data) and at the time of transplant (based on intraoperative findings); (2) wait time from entering a wait list until transplant, comparing patients with PVT at listing versus those without PVT; (3) drop-off rate from a wait list, comparing patients with PVT at listing versus those without PVT; (4) risk factors associated with the development of PVT while on a wait list; and (5) effects of PVT on patient and graft survival after LT.

Data source
The OPTN, under charter from the US Department of Health and Human Services, maintains a nationwide database of all patients who are waiting for matching for organ transplant and tracks the outcomes of those who receive a transplant. This database provides detailed clinical data on recipients, donors, and the recovery process.

Statistical analyses
The cohorts of patients were compared descriptively in terms of demographic characteristics, diagnoses, and clinical status using t test for continuous variables and chi-square test for categorical variables. In addition, the rate of PVT at the time of listing was evaluated chronologically over the entire study period to investigate trends in overall prevalence of PVT.

We compared patient cohorts in terms of patient survival using the Kaplan-Meier estimator and the log-rank chi-square test for equality. To evaluate sources for confounding factors, we constructed multivariable logistic regression models to evaluate the risk associated with falling off the wait list and to study the risks factors associated with development of PVT while on the wait list. The University of Washington Human Subjects Division deems the OPTN database to be deidentified and publicly available and thus not human subject data. Therefore, this study was exempt from human subject review. All analyses were performed in JMP Pro 11.2.1 software (SAS Institute, Cary, NC, USA).

The OPTN dataset reported 134 109 adult patients listed for primary OLT between January 2002 and June 2014. Patients with missing information on PVT at listing (5949, 4.8%) were excluded from the analyses (Table 1). Rate of PVT at listing was significantly increased over the time course (Figure 1). Patients with PVT at the time of listing versus those without PVT were more likely to be White (73.2% vs 71.3%) or Hispanic (16.2% vs 14.1%), have nonalcoholic steatohepatitis (NASH) as the cause of end-stage liver disease (19.7% vs 13%; 20.5%), and have shorter waiting times (350 ± 558 vs 439 ± 668 days; P < .0001). The rate of transplant was the same among patients with PVT (66 681; 54.2%) versus those without PVT (2761; 55.1%) at listing (P = not significant). Patients with PVT at listing were more likely to be removed from a wait list due to being too sick for transplant (11 071; 9%) than those without PVT (541; 11%) (P <.0001); however, the chances of death while on a wait list was higher among patients without PVT at listing (16 343; 13.7%) than those with PVT (591; 11.8%) (P < .01). On multivariable analyses, the presence of PVT at listing was an independent risk factor for being removed from a wait list for being too sick. The other risk factors identified were age, hepatocellular carcinoma, wait time, Model for End-Stage Liver Disease (MELD) score, and being on life support at listing (Table 2).

A total of 63 265 adult patients underwent a primary OLT during the same time frame. Of these, most patients (61 557) did not have PVT at listing, 57 945 remained without PVT at the time of transplant, 1708 patients were listed with PVT and had PVT at the time of transplant, and 3612 patients were listed without PVT but PVT was identified at the time of transplant (Table 3). Patients with PVT versus those without PVT at the time of transplant were more likely to be male (70% vs 66%), be Hispanic or White race, have history of previous abdominal surgery, have higher MELD score, have NASH as the cause of end-stage liver disease, have higher incidence of diabetes mellitus, and have higher rates of transjugular intrahepatic porto-systemic shunt (TIPS) procedures. The use of deceased donors after cardiac death and the deceased donor risk index were similar among both groups. The presence of PVT either at listing or at the time of transplant had a negative effect on patient and graft survival after OLT (Figure 2). On multivariable logistic regression analyses, several independent predictors of PVT development while on a wait list were identified (Table 4), including number of days on a wait list, age, prior abdominal surgery, hepatocellular carcinoma, ascites, history of variceal bleed, NASH, higher body mass index (BMI) at listing, and diabetes mellitus.

Although PVT is no longer considered a contra-indication for OLT, the presence of PVT in LT candidates remains technically challenging and is associated with a poor prognosis, mostly due to increased operative time, higher transfusion require-ments, increased rate of reoperation, and longer intensive care unit and hospital stay duration.^7-10

This study has several interesting findings. Despite the inferior outcomes, the rate of PVT in patients listed for OLT has increased over the years, despite the likelihood that the prevalence and incidence of PVT has not changed over these years. The reported annual incidence of PVT in patients with cirrhosis is variable, ranging from 4.6% to 17.9%.^2-5 The possible explanation for this finding is a more aggressive approach for transplant of patients who were likely turned down in the past. We believe this is related to better intra- and perioperative patient care, including improvements in surgical management of PVT both before and at the time of transplant. Another explanation for the higher incidence of PVT in the listed patients is an increased rate of end-stage liver disease due to NASH and therefore higher BMI in the listed patients, with both variables considered risk factors for the development of PVT.

Interestingly, despite having a similar rate of transplant, patients with PVT at listing receive transplants quicker than patients without PVT. This is explained by a higher Model for End-Stage Liver Disease score at the time of listing in patients with PVT. On the basis of our data, we are unable to draw conclusions regarding why PVT patients have higher MELD scores; however, we assume that part of it may be secondary to the use of anticoagulation, which inherently causes an elevation of the MELD score. Several studies have shown that PVT is more frequently found in patients with more severe liver disease, with a prevalence ranging from < 1% in well-compensated cirrhosis to > 25% in patients on transplant wait list.^2-9 Whether PVT is the cause or the consequence of liver deterioration is still uncertain.

The effect of PVT on entrance to wait list is also a controversial issue. We have demonstrated that the presence of PVT is associated with a significantly increased risk of being removed from a wait list due to being too sick for transplant. Consequently, the wait list mortality rate is increased for patients without PVT. These findings echo some previous data from the SRTR/OPTN dataset and data from a single-center study that showed that the presence of PVT does not increase the risk of wait list mortality.^11,12 Fouzas and associates¹³ found that patients with PVT on wait list have lower mortality than patients without PVT. Unfortunately, all of the studies, including ours, do not differentiate partial versus complete PVT. We do not have data to explain this finding, but we suspect that patients with PVT might be removed from the wait list before they die as to not affect the center’s wait list mortality rate.

One of the main findings of our study was that the presence of PVT was associated with worse posttransplant patient and graft survival. Of note, this association persisted over time and was independent of the quality of the donor, which is reflected by the same deceased donor risk index between groups. The impact of PVT on survival after OLT is controversial.¹¹ Survival may depend on the grade of PVT.¹⁴ According to a recent meta-analysis,¹¹ partial PVT seems to have no effect on survival rates, whereas patients who have complete PVT have inferior 1-year survival rates. On further analyses, we found that survival of patients with PVT who are listed is even worse than those patients who are listed without PVT but are found to have PVT at the time of transplant. In a recent paper, Ghabril and associates⁷ noticed the same finding as ours. The explanation of this cannot be surmised from our data; however, we assume that the development of PVT while on a wait list produces a clot more easily amenable to thrombectomy at the time of transplant compared with patients with long-standing PVT, which is sometimes associated with calcification of the portal vein thrombus.

The identification of risk factors for the development of PVT is paramount to working to minimize the risk of developing PVT while on a wait list. In patients without PVT at the time of listing, several risk factors were identified for PVT at the time of transplant. Among these, NASH, increased BMI, and presence of diabetes mellitus all had significant roles as independent risk factors for the development of PVT. This supports the suggested association of fatty liver disease and obesity with increased thrombotic risk.^15,16 Nonalcoholic steato-hepatitis was also found to be an independent risk factor for being removed from a wait list due to being too sick. In an era where NASH (and its inherent association with metabolic syndrome) as a cause of end-stage liver disease is dramatically increasing, the transplant community must be prepared to deal with the increased rate of PVT in wait-list candidates. Cancer is a well-known risk factor for a prothrom-botic state,¹⁷ and we found that a diagnosis of hepatocellular carcinoma has one of the highest odds ratios of PVT (1.26). Ascites and a history of variceal bleeding were also shown to be associated with the development of PVT while on a wait list, which concurs with the association between PVT and lower portal flow in the setting of severe portal hypertension.⁹

Another important finding was the protective effect of TIPS on the development of PVT, which supports the current increasing trend of using TIPS for the treatment of PVT for patients on wait lists. It is widely believed that, by increasing blood flow, TIPS may improve portal venous hemodynamics and induce flow-enabled clot dissolution. It has been shown that TIPS produces and sustains an open portal venous system in LT candidates, which would allow transplant in patients previously denied.^18,19

This study has certainly several limitations, including those limitations inherent to a retrospective observational study with selection bias, our inability to assess the effects of PVT on the decision to list the patient, and our inability to rule out errors in the report regarding the extent of PVT. The strengths of this study are its large sample size of listed patients and those who underwent LT and our contem-poraneous assessment of patients with PVT on wait lists and at the time of transplant.

In summary, PVT represents an increasing management and outcome burden in LT. The presence of PVT at listing and/or at the time of transplant is associated with worse patient and graft survival. Nonalcoholic steatohepatitis and hepato-cellular carcinoma are among the biggest risk factors for the development of PVT while on a wait list. The use of TIPS while on a wait list has a protective effect in the development of PVT. Larger studies are needed to better develop optimal therapeutic and preventive strategies for PVT in LT candidates.

Volume: 16 Issue: 2 April 2018

FULL TEXT