The dynamic landscape of liver transplant in the era of effective hepatitis C virus therapy

Potential conflict of interest: Nothing to report. See Articles on Pages 777 and 804 Hepatitis C virus (HCV)‐related liver disease has remained the most common indication for liver transplantation (LT) in much of the world for decades. Unfortunately, universally recurrent disease with accelerated hepatic fibrosis and previously ineffective and intolerable therapies have led to diminished graft and patient survival among these transplant recipients. The advent of safe and effective directly acting antiviral (DAA) therapy to treat HCV infection has revolutionized the practice of LT. Despite the approval of DAA regimens, it has been estimated that the incidence of life‐threatening complications of HCV, including decompensated cirrhosis and hepatocellular carcinoma (HCC), will continue to rise into the next decade until widespread prevention and treatment efforts take hold.1 However, data continue to emerge that treatment with sustained virological response (SVR) decreases the risk of cirrhosis, hepatic decompensation, HCC, and all‐cause mortality.3 In addition, SVR in patients with decompensated cirrhosis (Childs‐Turcotte‐Pugh [CTP] class B and C) may lead to stabilization or improvement in liver function in up to 60% of patients.4 Thus, there has been reason to be optimistic that DAAs will lead to both improved posttransplant outcomes for patients with HCV and a significant decline in the need for LT in this large population. In this issue of Hepatology, Flemming et al.6 suggest that the revolution in HCV therapeutics may already be having a tangible impact on the LT waiting list. Based on data from the Scientific Registry of Transplant Recipients, annual age‐standardized incidence rates of wait listing for HCV, hepatitis B virus, and nonalcoholic fatty liver disease were compared across three historical HCV treatment periods (the 2003‐2010 “interferon” era, the 2010‐2013 “protease inhibitor” era, and the 2014‐2015 “DAA” era). Whereas the overall rate of listing for patients with HCV was similar between 2003 and 2015 (approximately 6.9 per 100,000 U.S. population), there were striking declines in listing for HCV‐related decompensated liver disease (defined as a Model for End‐Stage Liver Disease [MELD] score of at least 15) in the protease inhibitor (–4.8%) and DAA (–31.9%) eras, compared to the totality of the interferon era. Interestingly, this decline was somewhat offset by a significant overall rise in wait listing among patients with HCV with HCC as the primary indication. Although the researchers postulate that HCV treatment with DAAs has led to this drop in listing for HCV‐related decompensated cirrhosis, establishing causality is not possible in these observational studies and the short timeline for this magnitude of impact may indicate that there are additional important contributing factors. There appears to be a period of decline in listing for decompensated disease even in the late interferon period, and the aging of the “baby boomer” population is likely playing a prominent role. The significant rise in median age at listing from 53 to 57 years throughout the study period supports this aging cohort effect. Nonetheless, this analysis provides encouragement that the goal of avoiding life‐threatening indications for LT through effective treatment of patients with advanced fibrosis is feasible, and suggests that HCV may soon no longer be the leading indication for wait listing. Conversely, for those patients already on the transplant waiting list with decompensated cirrhosis, additional questions remain. Several DAA regimens have achieved remarkable safety and efficacy even in patients with decompensated cirrhosis or posttransplant, rendering SVR attainable in the vast majority of patients. Given the benefits of SVR on liver function and mortality in the pretransplant and posttransplant settings, the question is not whether all patients should be treated, but when an individual patient should be treated such that access to transplant is considered and overall survival is maximized. Also in this issue, Chhatwal et al. address this complex question in an elegant Markov‐based microsimulation (simulation of liver transplant candidates) to test which treatment strategy (either pretransplant and posttransplant) leads to the best survival, based on MELD score at the time of treatment.7 This model incorporates data from HCV treatment trials performed specifically among patients with decompensated cirrhosis or posttransplant (the SOLAR 1 and 2 trials utilizing sofosbuvir/ledipasvir and ribavirin),4 including SVR rates as well as probabilities of improvement or worsening in liver function with treatment of pretransplant patients. Importantly, United Network for Organ Sharing (UNOS) data were also incorporated to model the course of individual wait listed patients in terms of changes in MELD and probabilities of outcomes, including death, delisting, and transplant. This is an essential component of the simulation given the complexity of modeling patient outcomes throughout a broad range in MELD scores (10‐40) in the regional allocation system of the United States. The researchers conclude that the optimal cutoff below which patients on the wait list with decompensated cirrhosis should be treated ranges between 23 and 27, depending on the region and therefore median waiting time for transplant. This perhaps suggests that more pretransplant patients would benefit from treatment, even with MELD scores up to 27, despite concerns about the consequences of diminishing organ access due to a decline in MELD and elimination of the possibility of a HCV‐seropositive graft. The promise of improvement in liver function, prevention of extrahepatic manifestations of HCV including renal failure, and prevention of HCV recurrence in the allograft create an undeniable instinct to cure all patients with HCV on waiting list. However, some caution should be exercised in applying this analysis to individual patients given the uncertainty around several key model variables and the dynamic nature of the organ allocation system. Although the SVR rates were extrapolated from the largest published trials in patients with decompensated liver disease and the model was relatively stable with variation in SVR rate in sensitivity analyses, the true SVR rate in the population modeled may be significantly lower. When pooled across all clinical trials with all regimens used in CTP class B and C patients of all HCV genotypes (n = 801), the overall SVR rate was recently found to be only 83.5%,5 and SVR declines in patients with more‐severe hepatic dysfunction (CTP class C and/or high MELD score). Although patients with MELD scores up to 40 were included in this model, and those with MELD scores up to 27 appeared to benefit from pretransplant treatment, there were almost no patients with MELD scores over 20 included in any clinical trials to date. These trials have defined hepatic decompensation upon entry by CTP class (B or C) rather than MELD. Almost all were performed with sofosbuvir‐based therapy and therefore excluded patients with advanced renal impairment, which is common in patients with MELD scores in this high range. Thus, modeling the impact of treatment in patients with this degree of baseline hepatic impairment in terms of SVR rate (which has been reported to be as low as 25% in patients with MELD scores over 20,9) safety, and the impact on liver function may not currently be possible. Furthermore, the mid‐ and long‐term outcomes for patients who achieve SVR and improvement in liver function and are thus not transplanted are not known. In pooled analysis, up to 60% will experience some improvement in liver function as measured by MELD and/or CTP score, with many experiencing a decline in MELD to less than 15.4 However, some with MELD scores that decline to under 15 will not experience full restoration of health and may have ongoing indications for transplant. In addition, 23% may experience clinical worsening initially, and it is clear that there is ongoing risk of HCC despite SVR, especially in patients with cirrhosis.10 In fact, as noted above, the impact of broader HCV treatment on HCC incidence may lag behind that of decompensation of liver function in terms of trends in transplant listing.6 Finally, the dynamic nature of organ allocation adds additional significant complexity to the simulation of LT candidates model. The utilization of UNOS data to attempt to approximate individual patient scenarios is a major strength of this analysis, and the researchers are commended for taking this approach. However, the model may therefore be susceptible to changes in allocation practices and policy in the near future. This will include broader sharing and redistricting to address regional disparities in MELD score at allocation. It may also include a better understanding of how best to utilize HCV‐seropositive donors, perhaps even in HCV‐negative recipients, removing the incentive to wait for posttransplant treatment in selected patients. It will be of interest to see updated reports from this simulation of LT candidates model as better data on treatment of high MELD patients are available and allocation policies shift. This is a remarkable time for the field of HCV treatment. Thousands of patients have been treated and cured in just the few short years since the first interferon‐free regimen was approved. On an individual patient level, the benefits of safe and effective therapies are perhaps nowhere more profound than among the LT candidates and recipients who face the immediate risks of ongoing infection and who previously had almost no hope for cure. Due to successful treatment of patients with advanced fibrosis, we may already be realizing the benefits of SVR on the population‐based need for LT. However, for those who are already suffering the complications of end‐stage liver disease, some essential questions remain unanswered. Delaying treatment that could stabilize or improve health on the wait list seems a difficult paradigm, and if transplant were accessible to all in need and costs related to retreatment were not a factor, immediate treatment would be universal. However, there may be patients who are ill‐served by treatment on the wait list in the current system, and treatment on the wait list must currently remain individualized (a proposed algorithm is presented in Fig. 1). It is clear that treatment is indicated to prevent recurrent disease in patients with compensated cirrhosis and/or low MELD who have access to transplant through MELD exception points or living donation. Patients with low MELD may also benefit from treatment to improve liver function and avoid transplant, but whether MELD or CTP score are the best measures and the optimal cutoffs are not known. There are also patients with MELD scores above which treatment is either detrimental or not possible, such as in patients with both severe hepatic and renal impairment that represent contraindications for protease inhibitors and sofosbuvir, respectively. The patients with moderate MELD scores in between these two extremes are those who remain underserved by our lack of treatment data and inability to predict clinical outcomes, and are an important focus of ongoing study in this dynamic landscape of HCV and LT.Figure 1: Proposed algorithm for HCV treatment in LT candidates with chronic HCV. *MELD < 20 is based upon the limited data on safety and efficacy of treatment in patients with MELD >20; however, treatment above this MELD may be beneficial in some patients. **GFR > 30 mL/min/1.73 m2 is based upon the need for use of sofobuvir in all currently recommended treatment regimens for patients with decompensated cirrhosis and the contraindication to use of sofosbuvir in patients with severe renal impairment. ***MELD > 27 is based upon the simulation of LT model presented in this issue,7 which found that even if treatment is assumed to be safe and effective above this MELD, there may be a survival benefit by delaying therapy to the posttransplant setting. Abbreviations: GFR, glomerular filtration rate, expressed as mL/min/1.73 m2; LDLT, living donor liver transplant.