Objectives: Hepatitis C virus infection is the most common underlying reason for hepatocellular carcinoma and indication for liver transplant. The increased availability of non-interferon–based therapy has expanded the number of treatment-eligible patients.
Materials and Methods: We used a decision analysis model to compare 2 strategies of treating hepatitis C virus. Included patients were followed for 1 year after liver transplant. The probabilities and costs were obtained from a literature review, an expert panel, and our institution’s experience. Sensitivity analyses were performed on all variables.
Results: Our model demonstrated that it would be less costly to treat patients after liver transplant than to treat patients while they wait for transplant. When we compared baseline values, the cost difference between the 2 strategies was $25,011 per patient and $41,535 per sustained viral response. Overall survival was 60.1% for both strategies. Our model was robust across most of the variables tested in the sensitivity analysis.
Conclusions: Our results indicated that there is no substantial pharmacoeconomic or survival advantage of treating hepatitis C virus in patients with compensated cirrhosis and hepatocellular carcinoma before liver transplant versus after transplant.
Key words : Antiviral therapy, Health care use, Cirrhosis
Hepatitis C virus (HCV) is the leading indication for liver transplant in the United States.1 After transplant, recurrent HCV infection is universal, resulting in the risk of cirrhosis development in less than a decade.2,3 Five-year survival rates for transplant recipients for HCV are among the lowest of all other indications.4 Previous studies have assessed the use of interferon-based agents to treat HCV infection in patients while they wait for liver transplant and after transplant.5 However, widespread use of interferon has been limited by safety, tolerability, and efficacy concerns. Moreover, interferon is contraindicated in patients with advanced liver disease, particularly when used with first-generation protease inhibitors.6 Interferon-based therapy after transplant also has been associated with substantial complications.7-9
There is great enthusiasm for non-interferon–based therapy in patients waiting for liver transplant. Recently, sofosbuvir and ribavirin have been shown to be effective in suppressing viral replication in patients on wait lists for liver transplant with HCV with a diagnosis of hepatocellular carcinoma (HCC).10 In that study, when viral suppression was achieved for at least 2 months before transplant, the likelihood of achieving a sustained viral response (SVR) after liver transplant was over 95%.10 Other antiviral treatment combinations also have been effectively administered in patients with cirrhosis,11-13 with recent studies using these direct-acting agents without interferon in liver transplant recipients.14-16 Non-interferon–based therapies appear safe, tolerable, and effective in the posttransplant setting despite potential drug-drug interactions.
There is insufficient evidence regarding whether suppression of HCV in patients with HCC changes either the natural history of the disease or the likelihood of HCC recurrence. Moreover, dropout rates from transplant wait lists for patients with liver cancer and HCV vary because of disease progression.17-20 In this study, we investigated whether it would be less costly to treat HCV after rather than before liver transplant in patients with HCV and HCC.
Materials and Methods
We compared 2 strategies for treating patients with HCV genotype 1 on wait lists for liver transplant with compensated liver disease and HCC. In the first strategy (strategy 1), patients received oral antiviral treatment when they were placed on the liver transplant wait list. In the second strategy (strategy 2), patients received treatment after liver transplant. A decision analysis model was constructed using Excel (Seattle, WA, USA). Follow-up continued for 1 year after liver transplant. Outcomes of interest included overall SVR, total cost per patient, antiviral drug costs, and overall survival.
The natural history of patients on liver transplant wait lists with HCV and HCC was derived from the literature (Table 1). The dropout rate in these patients was estimated from a validated model by Toso and associates.18,21 Because there is no consistent consensus that viral eradication affects the rate of HCC recurrence, health care use and dropout rates were modeled to be equal between patients who were still infected versus those who were no longer infected with HCV. After liver transplant, we assumed that the natural history and health care use for patients who achieved an SVR were the same as for patients who received liver transplant for hepatitis B and HCC. We obtained the time to discharge after liver transplant, rate of readmission, and mean duration of hospitalization during readmission from a random sample of 25 transplant recipients for HCC who had either hepatitis B virus or HCC in 2013 at our institution. The incidence of fibrosing cholestatic hepatitis was obtained from a literature review.22,23 In our model, we calculated that fibrosing cholestatic hepatitis would occur within 3 months of liver transplant.
Hepatitis C antiviral treatment
We reviewed the efficacy rates of US Food Drug and Administration-approved non-interferon–based therapies or as recommended by the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America guidelines in HCV treatment.11-13,24 We assumed a 12-week regimen for treatment-naïve HCV patients. The costs associated with a 24-week regimen were captured in the sensitivity analysis. We did not include patients with decompensated liver disease in our model. For patients treated after transplant, we used response rates of recently presented studies for our analyses.14-16 Responses to antiviral treatment that were classified as an SVR were those that met the definition of an undetectable hepatitis C viral load at 12 weeks after completion of treatment. During antiviral therapy, patients underwent viral testing 1 month after starting HCV treatment and 3 months after completing therapy according to the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America guidelines.24,25 Routine blood work results were not incorporated into the model because no differences would be shown between treatment strategies. There were also no differences between the 2 groups regarding number of office visits during and after treatment. Patients who died of fibrosing cholestatic hepatitis were assumed to have had 2 months of office visits. Few studies have assessed antiviral retreatment of such patients who did not respond to an initial course of therapy. We assumed antiviral therapy resulted in a 90% SVR.15,26 In our model, we further assumed that patients would start antiviral therapy soon after a diagnosis of fibrosing cholestatic hepatitis since time to diagnosis can be a predictor of mortality.27,28
In our model, we had a number of important assumptions: (1) repeat liver transplant is not available; (2) pretransplant SVR is durable after transplant; (3) HCV infection does not affect the likelihood of acute or chronic rejection; (4) viral suppression does not affect HCC recurrence before or after transplant; (5) posttransplant survival after an SVR is similar to that shown with hepatitis B virus; (6) the dropout rate of patients waiting for liver transplant is similar between treatment responders and nonresponders, with dropout rate increasing linearly from 3 to 12 months; (7) all patients receive a liver transplant or die 12 months after being placed on a liver transplant wait list; and (8) the number of clinic visits is the same in responders and nonresponders to treatment.
All patients in the model with HCC were believed to have met the Milan criteria for wait list inclusion.29,30 The dropout rate was estimated and varied in a sensitivity analysis to generalize patients enrolled with different risks of cancer recurrence.18,19 Because our follow-up was 1 year after transplant, rates of HCC recurrence were not incorporated in the model.
Estimations of costs were obtained from published studies, Medicare reimbursement rates from the year 2014, and from the Red Book 2014 (Table 2).31-36 In the posttransplant setting, the costs of conducting annual protocol liver biopsies and of managing subsequent complications were included, as well as the costs of managing rejection during antiviral treatment. Cost estimates were converted to 2014 US dollars using the Bureau of Labor Statistics Consumer Price Index Inflation Calculator.37 Costs were not discounted. The study perspective was of a third-party payer.
We conducted a univariate sensitivity analysis to assess whether the results of our model were robust. All variables used in the model for clinical assumptions, annual probabilities, and costs (Table 1) were tested over a wide range of values to determine their effects on total costs per patient and the costs per SVR. Ranges used in the sensitivity model were varied up to 20% of the baseline value. Total costs and outcomes were compared between the 2 strategies.
When we compared baseline values, our model showed that a pretransplant HCV treatment strategy (strategy 1) was associated with increased total costs, increased costs per patients, increased costs per SVR, and similar overall survival versus a posttransplant HCV treatment strategy (strategy 2). Pretransplant HCV treatment was associated with a mean increased cost of $25,011 per patient versus posttransplant HCV treatment. In addition, with a pretransplant treatment strategy, the costs per SVR increased by $41,535.
The results of our model consistently showed that pretransplant antiviral treatment (strategy 1) was more costly than posttransplant treatment across all variables tested except for length of stay related to immediate posttransplant care. When the length of stay was increased in the sensitivity model to 22 days in patients who still had HCV infection after transplant, the total cost per patient was $11,987 less costly with the pretransplant HCV strategy than with the posttransplant HCV strategy, with costs per SVR being $20,710 less costly for strategy 1 versus strategy 2. Likewise, our model was sensitive regarding length of stay in patients who achieved a SVR before transplant. When the length of stay was decreased to 13 days in patients who no longer showed HCV infection, the total costs per patient was $4,370 less costly with the pretransplant HCV strategy, with costs per SVR also $7,845 less costly with the pretransplant HCV strategy.
Our model was robust regarding other variables in the sensitivity analysis. For example, despite the decreased cost of treating antiviral agent-naïve patients of $75,200 in the sensitivity analysis, the pretransplant strategy was still associated with an increase in total costs of $17,979 per patient and $29,736 costs per SVR versus the posttransplant antiviral strategy. The increase of SVR to 100% also did not favor a pretransplant strategy. The total costs per patient and costs per SVR were $11,494 and $13,072 over the posttransplant antiviral strategy.
The results of our study indicate that treating HCV in patients with compensated cirrhosis while waiting for liver transplant with the main indication of HCC is substantially more costly than treating the HCV infection after transplant. We included a number of factors in our model that could affect the cost differential and thus support the generalizability of our study. Our study suggested that there is a pharmacoeconomic advantage of treating liver transplant recipients versus treating patients while on the liver transplant wait list.
Antiviral therapy has led to substantial gains in the treatment of patients with compensated cirrhosis. With currently available and soon to be US Food and Drug Administration-approved regimens, the SVR in patients with cirrhosis is over 90%. Sustained viral responses in patients with cirrhosis treated with interferon-based therapy are associated with decreased morbidity and mortality.38-41 As antiviral therapies are increasingly used for patients with greater severity of liver disease, it is unclear whether the same magnitude of improved outcomes will be realized. For example, the platelet and albumin threshold levels are likely to be substantially lower with non-interferon–based therapy.
In our model, we did not include a benefit of achieving a SVR in pretransplant patients. This is in direct contrast to studies that assessed viral suppression in patients with hepatitis B. With hepatitis B, there is ample evidence to suggest viral suppression leads to improved liver function and decreased HCC rates.42-45 However, to the best of our knowledge, there is no evidence that suppression of HCV reduces the likelihood of recurrent HCC. Because we assumed that antiviral therapy had no effect on pretransplant care, we did not include the associated costs of cirrhosis management or liver transplant surgery. We did, however, incorporate differences in hospitalization stay after transplant between patients who had or no longer had HCV. In our model, we assumed that an SVR pretransplant was durable and was maintained posttransplant. Furthermore, we assumed that only HCV patients who were still infected could develop fibrosing cholestatic hepatitis. Posttransplant length of stay, monitoring, and readmission rates were greater in patients who were still infected with HCV.
Our model was also sensitive to the length of stay after liver transplant. Pretransplant antiviral therapy was favored if the length of stay after transplant was increased in patients with HCV infection or decreased in patients without HCV infection. We obtained baseline length of stay values from transplant recipients for liver cancer with underlying hepatitis B or C. However, Henry and associates recently suggested that the average national length of stay may be lower than our own experience.46
There are several important limitations to our study. We did not incorporate differences in organ accessibility between the treatment strategies. Many centers may limit the use of HCV-positive grafts in patients who have achieved an SVR. Thus, the wait time for liver transplant may be longer in patients who are cured of their viral infection. Between 2010 and 2014, the percentage of patients who have received a donor graft infected with hepatitis C ranges from 1.9% in Organ Procurement and Transplantation Network region 6 to 6.62% in Organ Procurement and Transplantation Network region 2 (personal communication, Scientific Registry of Transplant Recipients). Specifically, in our Organ Procurement and Transplantation Network region 5, the percentage is 3.65%. Another limitation of our model is the 1-year follow-up after transplant. We did not believe this would introduce significant bias since all patients with HCV infection are modeled to start antiviral therapy 3 months after transplant. Patients who develop fibrosing cholestatic hepatitis start antiviral therapy sooner in our model. A third limitation is lack of confirmatory studies regarding treating patients who failed direct acting antiviral agents. In our model, we assumed that centers would offer to retreat patients, which would double the duration of therapy. This was based on recent findings of high efficacy with 24-week sofosbuvir and ledipasvir treatment. Treatment failure was more common in the posttransplant treatment strategy.47,48
The results of our study indicate that treating HCV is substantially less costly when initiated after transplant instead of before. Additional studies are needed to understand the effects of viral suppression after transplant on HCC recurrence rates in candidates who had been on liver transplant wait lists with compensated cirrhosis and HCC.
Volume : 14
Issue : 1
Pages : 66 - 71
DOI : 10.6002/ect.2015.0069
From the Departments of 1Medicine and 2Surgery,
University of California at Los Angeles, Los Angeles, and the 3Liver
Center, Huntington Medical Research Institutes, Pasadena, California, USA
Acknowledgements: The authors declare that they have no sources of funding for this study, and they have no conflicts of interest to declare. Authors had the following study roles: study concept and design (SS), acquisition of data (TF, CW, ME, MJ), analysis and interpretation of data (TF, CW, MJ), drafting of the manuscript (SS, MJ), critical revision of the manuscript for important intellectual content (SS, ME, MT), statistical analysis (TF, CW, MJ), and study supervision (SS). The authors thank Naadir Jamal for administrative analysis.
Corresponding author: Sammy Saab, Pfleger Liver Institute, UCLA Medical Center, 200 Medical Plaza, Suite 214, Los Angeles, CA 90095, USA
Phone: +1 310 206 6705
Fax: +1 310 206 4197
Table 1. Probability Analysis
Table 2. Estimation of Costs