Decision analysis model for hepatitis B prophylaxis one year after liver transplantation

Authors

  • Sammy Saab,

    Corresponding author
    1. Department of Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    2. Department of Surgery, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    • Pfleger Liver Institute, UCLA Medical Center, 200 Medical Plaza, Suite 214, Los Angeles, CA 90095
    Search for more papers by this author
    • Telephone: 310-206-6705; FAX: 310-206-4197

  • Maggie Y. Ham,

    1. Department of Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Michael A. Stone,

    1. Department of Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Curtis Holt,

    1. Department of Surgery, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    Search for more papers by this author
  • Myron Tong

    1. Department of Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    2. Department of Surgery, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA
    3. Huntington Medical Research Institutes, Pasadena, CA
    Search for more papers by this author

Abstract

In patients receiving orthotopic liver transplantation, hepatitis B recurrence rates have decreased significantly with the use of various methods for prophylaxis. At present, a combination of hepatitis B immunoglobulin (HBIG) and lamivudine is the standard of care, resulting in recurrence rates of 0% to 11%. Recent data suggest that the addition of adefovir to lamivudine is successful in treating patients with recurrent hepatitis B infection. A Markov model was used to compare costs and outcomes of 2 strategies for hepatitis B prophylaxis 1 year after transplantation. The first consisted of prophylaxis with lamivudine and adefovir (strategy 1), whereas the second consisted of intramuscular HBIG and lamivudine (strategy 2) with the addition of adefovir in patients who subsequently developed hepatitis B recurrence. Patients who failed with adefovir and lamivudine were then treated with tenofovir and entecavir. 16.8% of liver transplant recipients had hepatitis B recurrence after 10 years of treatment with lamivudine and HBIG. The medical costs for strategy 1 and strategy 2 after 10 years of therapy were $151,819 and $166,246, respectively, and this resulted in cost savings of $14,427. The decision analysis model began 1 year after liver transplantation. A 1-way sensitivity analysis demonstrated that the model was most sensitive to cost changes of adefovir and HBIG injections as well as variations in the hepatitis B virus recurrence rate. The model was robust to costs of lamivudine, laboratory costs, administrative fees, and office visit fees. Our decision analysis model resulted in marked savings in costs with strategy 1 (lamivudine and adefovir), providing pharmacoeconomic support for the use of this strategy as first-line therapy in hepatitis B prophylaxis in liver transplant recipients 1 year after liver transplantation. Liver Transpl 15:413–420, 2009. © 2009 AASLD.

The recurrence of hepatitis B virus (HBV) in liver transplant recipients was as high as 80%1 before the utilization of hepatitis B immunoglobulin (HBIG) and lamivudine. The use of HBIG and lamivudine for prophylaxis has virtually eliminated graft reinfection in liver transplant recipients, with current recurrence rates ranging from 0% to 11%.2–8 The decrease in reinfection has translated to improved outcome with survival. For instance, 2-year patient survival increased from 50% in 1980 to greater than 80% in 2000.9, 10 In fact, graft survival for hepatitis B is among the best for all indications for liver transplantation.11

However, treatment with HBIG and lamivudine has its limitations, which include high costs and the need for recurrent clinic visits for HBIG administration. Moreover, despite prophylaxis with the combination of lamivudine and HBIG, hepatitis B viral infection still may recur. Viral breakthrough is more commonly related to resistance to either HBIG or, less frequently, lamivudine.12 This is believed to be a result of either development of the hepatitis B surface antigen (HBsAg) escape variant or the emergence of the YMDD mutant, as demonstrated by polymerase chain reaction assays of sera collected from patients with HBV recurrence.12 Although the selection of resistant strains of HBV is a critical component in the development of recurrent disease, other factors such as noncompliance and patients' immune responses are involved. Additionally, a recent study found the recurrence of hepatocellular carcinoma post–liver transplant to be associated with HBV recurrence, possibly because of HBV replication in tumor cells.13

Because of HBV recurrence with lamivudine and HBIG, a number of studies have demonstrated the utility of adding adefovir to lamivudine for viral suppression.14–17 To date, the use of adefovir as salvage therapy in liver transplant recipients has been associated with complete viral suppression rates.2, 14, 15, 18, 19 No liver-associated deaths or viral breakthrough has been reported with the use of combination lamivudine and adefovir.15–17, 20 In addition, adefovir has been used in patients with known lamivudine-resistant YMDD mutants, resulting in complete suppression of viral replication.15

We compared the costs and efficacy of 2 strategies in patients with hepatitis B undergoing orthotopic liver transplantation: (1) lamivudine with adefovir and (2) lamivudine with HBIG. We also examined the costs and efficacy of salvage therapy with adefovir and lamivudine in patients who had HBV recurrence while on lamivudine and HBIG. Because of the low rates of HBV recurrence in patients receiving adefovir and its ease of administration in comparison with HBIG, we hypothesized that in patients who had received liver transplants, prophylaxis with adefovir and lamivudine would be more economically beneficial.

Abbreviations

ADV, adefovir; CMP, comprehensive metabolic panel; HBIG, hepatitis B immune globulin; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; IM, intramuscular; IV, intravenous; LAM, lamivudine; PCR, polymerase chain reaction.

PATIENTS AND METHODS

Decision Model

We compared 2 strategies for hepatitis B prophylaxis in orthotopic liver transplantation recipients. Strategies were based on initiation of the strategies 1 year following liver transplantation and in recipients with recurrent hepatitis B infection. In strategy 1, patients were given adefovir and lamivudine as hepatitis B prophylaxis after their first year of orthotopic liver transplantation. HBIG was discontinued after the initiation of adefovir. In strategy 2, patients received both HBIG [intramuscular (IM)] and lamivudine as prophylaxis. In the latter instance, adefovir was initiated if patients developed HBV recurrence, and HBIG was discontinued. Tenofovir and entecavir were initiated for patients who developed HBV recurrence while on lamivudine and adefovir.9, 21 A decision analysis model was constructed with a commercial spreadsheet (Excel, Microsoft, Seattle, WA) to simulate 10-year costs for each strategy.

Possible outcomes included the following: (1) no recurrence of HBV with HBIG and lamivudine, (2) recurrence of HBV in patients receiving HBIG and lamivudine requiring the addition of adefovir, (3) no recurrence of HBV on adefovir and lamivudine, and (4) recurrence of HBV on adefovir and lamivudine requiring tenofovir and entecavir (Fig. 1). In strategy 1, liver transplant recipients were provided oral lamivudine and adefovir. The recipients were seen in a clinic every 3 months and were tested for HBsAg and HBV DNA every 3 months. In strategy 2, patients received daily oral lamivudine and monthly IM HBIG injections. These patients were seen in a clinic every 3 months with a physician and every month for HBIG administration during a nurse visit. Laboratory tests for antibody to HBsAg titers, HBsAg, and HBV DNA were performed every 3 months. Because we could not identify cases in the literature of viral breakthrough on oral combination therapy, we assumed a HBV recurrence rate of 0.25% per year in patients initially treated with adefovir and lamivudine (strategy 1). We also assumed a higher HBV recurrence rate of 0.5% per year with adefovir and lamivudine if they had previously failed lamivudine and HBIG (ie, originally in strategy 2).

Figure 1.

Decision analysis model: a simulation of costs and health outcomes for patients assigned to each strategy starting 1 year after liver transplant. Abbreviations: HBIG, hepatitis B immune globulin; HBV, hepatitis B virus.

Hepatitis B Recurrence

Hepatitis B recurrence was defined as the reappearance of serum HBsAg after its initial disappearance after liver transplantation5, 6, 8 with detectable serum HBV DNA levels with or without abnormal liver function tests.18 Recurrence rates were obtained from a critical review of the literature (Table 1).2–8, 14, 16, 18, 22 A linear regression analysis was then performed with HBV recurrence rates obtained from the literature to estimate the recurrence rate as a function of time from liver transplant (Fig. 2). A multiple regression analysis was performed with the sample size and time to reinfection as covariates to establish the linear equation. We then used the obtained regression equation to produce estimated rates of HBV recurrence at 6-month intervals during the 10-year cycle of the postoperative simulation. The rate of recurrence was cumulative at the end of the 10-year cycle.

Table 1. Publications of Hepatitis B Prophylaxis and Rescue Strategies in Post–Orthotopic Liver Transplantation Patients
StudynTreatmentFollow-Up DurationRecurrenceSurvival
  1. Abbreviations: ADV, adefovir; HBIG, hepatitis B immune globulin; HBV, hepatitis B virus; IM, intramuscular; IV, intravenous; LAM, lamivudine.

Dan et al.31207HBIG (IV) + LAM1 year0%98%
   3 years7%93%
   5 years11%90%
Marzano et al.877HBIG (IV) + LAM39 months5% 
Dumortier et al.417HBIG (IV) + LAM30 months0%100%
Marzano et al.2133HBIG (IV) + LAM1 year4% 
Markowitz et al.713HBIG (IV) + LAM1.1 years0%100%
Han et al.559HBIG (IV) + LAM459 days0%98%
Buti et al.29HBIG (IM) + LAM83 months11%100%
Han et al.659HBIG (IM) + LAM1051 days0%95%
Buti et al.23ADV + LAM rescue6 months0%100%
Lo et al.1516ADV + LAM10 months0% 
   15 months0%94%
Schiff et al.1257ADV + LAM528 days0% 
Neff et al.1610ADV + LAM31 months0% 
Gane et al.27147HBIG (IM) + LAM1860 days4%88%
Figure 2.

Hepatitis B rate of recurrence in patients receiving hepatitis B immune globulin and lamivudine prophylaxis.

Costs

Costs for medications were obtained from the literature and Medicare reimbursement schedules (Table 2). Drug costs were derived from average wholesale prices.23 The HBIG dose was 2000 units per month. Costs are expressed in US dollars for 2008 on a monthly basis per patient. Costs are adjusted for inflation at an approximate average yearly rate of 3% for hospitals, as determined by the Bureau of Labor Statistics of the US Department of Labor.24

Table 2. Costs Entered into the Markov Model
Variable Baseline Cost (Range)References
  • NOTE: All costs are in US dollars (2008 dollars).

  • Abbreviations: HBIG, hepatitis B immune globulin; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; PCR, polymerase chain reaction.

  • *

    Per month.

  • Sodium, potassium, chloride, carbon dioxide, blood urea nitrogen, creatinine, glucose, albumin, alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, total bilirubin, calcium, and protein.

Medications*Lamivudine$340 ($170–$680)5
 Intramuscular HBIG (2000 units, 2.5–cc vial)$805 ($402–$1610)5, 10
 Adefovir$772 ($386–$1544)5
 Tenofovir$614 ($307–$1228)19
 Entecavir$682 ($341–$1364)19
Laboratory feesHBsAg$60 ($30–$120)Medicare
 HBV DNA quantitative PCR$120 ($60–$240)Medicare
 Comprehensive metabolic panel$84 ($42–$168)Medicare
 Hepatic panel$60 ($30–$120)Medicare
 Hepatitis B surface antibody$130 ($65–$260)Medicare
OtherInfusion$21 ($11–$42)Medicare
 Nurse visit$20 ($10–$40)27
 Physician visit$89 ($45–$178)27

Sensitivity Analysis

Costs of treatment, laboratory tests, and visits were varied at 50% and 150% of the baseline costs. Reactivation rates were also varied at 50% and 150%. A threshold analysis was then conducted on variables to which the model was sensitive.

Assumptions

We assumed (1) stable immunosuppression, (2) no prednisone use, (3) similar rejection and nonspecific causes of elevated liver enzymes, (4) similar compliance, (5) similar causes of nonhepatic deaths, and (6) stable renal function.

RESULTS

Baseline Assumptions

We used a decision analysis model to simulate clinical outcomes and costs associated with HBV prophylaxis in patients who had received orthotopic liver transplants. In order to increase the generalizability of the model, we varied event probabilities and costs (Table 3). We then extrapolated clinical costs and outcomes to a cohort of 100 patients within each treatment strategy group on the basis of data and assumptions from the literature (Fig. 1). In strategy 2, patients had a rate of hepatitis B recurrence of 16.8% after 10 years while on HBIG and lamivudine. This equated to 16.8 patients in our cohort of 100 patients.

Table 3. Sensitivity Analysis Varying the Costs and Hepatitis B Recurrence Rate
Variable Varied Costs
50%150%
  1. NOTE: The cost difference was calculated as strategy 2 − strategy 1. Costs refer to percentages of baseline values.

  2. Abbreviations: CMP, comprehensive metabolic panel; HBIG, hepatitis B immune globulin; HBsAg, hepatitis B surface antigen; HBV DNA, hepatitis B DNA quantitative polymerase chain reaction.

MedicationsLamivudine$12,887$15,967
 Adefovir$41,544−$12,691
 Tenofovir$17,211$11,642
 Entecavir$17,518$11,336
 HBIG−$17,461$46,315
 HBIG administration$13,594$15,259
VisitsPhysician$14,427$14,427
 Nurse$13,634$15,220
LaboratoryLabs (HBIG)$9,273$19,580
 Labs (HBsAg, HBV DNA, and CMP)$14,427$14,427

In our model, the cost per patient of prophylaxis for 10 years was $151,819 with adefovir and lamivudine and $166,246 with HBIG and lamivudine. The costs related to strategy 2 included $8717 for HBIG and lamivudine, $7499 per patient for salvage therapy using adefovir and lamivudine, and $8603 per patient for salvage therapy with tenofovir and entecavir. Strategy 1 resulted in cost savings per patient of $14,427 for 10 years.

Sensitivity Analyses

Prophylaxis with adefovir and lamivudine was the less costly strategy in the 1-way sensitivity analyses for most varied variables. The model was robust to costs of lamivudine, clinical visits, laboratory tests, and HBIG administration. However, the model was most sensitive to the costs of adefovir and HBIG and to hepatitis B recurrence. When the monthly cost of adefovir was greater than $978, strategy 2 became less costly. Strategy 2 was also less costly when the monthly cost of HBIG was less than $623 (Fig. 3). The rate of hepatitis B recurrence was varied by 50% and 150% for both strategies. When the rate of recurrence was varied by 50%, strategy 2 was less costly than strategy 1 by $17,631. Costs between strategies were not equal until the rate of viral recurrence was increased by 695%. Variance in the rate of viral recurrence was not able to shift the model to favor strategy 2. The rate of recurrence varied linearly with time (Fig. 2).

Figure 3.

Threshold analysis varying the costs of adefovir and hepatitis B immune globulin. Values above the threshold line result in strategy 1 being less costly than strategy 2; values below the threshold line result in strategy 2 being less costly than strategy 1. Abbreviation: HBIG, hepatitis B immune globulin.

DISCUSSION

Lamivudine is an oral nucleoside analogue with efficacy against hepatitis B. It is safe and well tolerated. However, its long-term use is associated with viral resistance in both liver transplant and nontransplant patients. In nontransplant patients, this resistance has been shown to be prevented with combination antiviral therapy. In a large randomized study, Peters et al.25 demonstrated no viral resistance in patients treated with the combination of adefovir and lamivudine. In contrast, 10% to 20% of patients developed viral resistance when treated with lamivudine monotherapy. Rapti et al.17 confirmed the efficacy of adefovir as salvage therapy in nontransplant patients. The authors found that the earlier treatment is instituted, the more quickly viral suppression is achieved. In transplant recipients, adefovir has also been used for viral suppression in patients developing viral breakthrough.14, 16, 17

We used IM HBIG in our model, as it has previously been shown to be as effective as intravenous (IV) HBIG, in addition to being more cost-effective.6 Yearly costs of IM HBIG or IV HBIG in combination with lamivudine were $13,500 and $181,130, respectively, resulting in $52,600 per recurrence prevented with IM HBIG versus $371,000 per recurrence prevented with IV HBIG. We standardized our model to reflect 1 HBIG injection per month per patient ($805); however, if a higher dose is required to maintain titers within the therapeutic range (eg, bimonthly) the cost of IM HBIG would be closer to $1600 per month, and so the actual cost would be underestimated. Therefore, with increases in the HBIG injection frequency, there is a bias against our model. Our model suggests that in recipients of orthotopic liver transplants, it is less costly to treat with adefovir, rather than HBIG, in combination with lamivudine. Baseline costs for strategy 1 and strategy 2 were $151,819 and $166,246, respectively, with a difference of $14,427. On the basis of these results, we infer that initiating prophylaxis with adefovir and lamivudine is advantageous after the first year post-transplantation.

Our model was based on treatment strategies after the first year of transplantation. This may explain the lower expected savings with an oral antiviral therapy. Most of the data used in this study were derived when adefovir was added as salvage therapy in recipients who had received HBIG and lamivudine for variable periods of time. Few studies used combination lamivudine and adefovir immediately after liver transplantation. In fact, a recent study demonstrated the effectiveness of converting to an oral combination therapy consisting of lamivudine and adefovir after a median of 32 months of HBIG and lamivudine.17 Additional studies are needed to understand if an oral therapy regiment would be safe and effective. The closest data that we have demonstrating the possibility were reported by Schiff et al.,14 who found that patients who were wait-listed for liver transplant received adefovir for a median of 15 weeks prior to receiving a liver transplant, and it was then continued for a median of 36 weeks after transplantation.

Not only is there a pharmacoeconomic advantage to using an oral regiment, there are also quality of life issues that have not been assessed in published studies. Administration of HBIG requires frequent blood tests to assess antibody titers as well as monthly office visits. On the other hand, adefovir is an oral medication with rare adverse effects, and there is less of a need for frequent monitoring. Indeed, compliance with HBIG administration is a concern, with 1 study reporting 15% noncompliance,2 whereas noncompliance with adefovir has been reported to be as low as 1%.14 In our study, we assumed 100% compliance in our model, which would bias the model against an oral treatment regimen. Reasons for overall noncompliance include medication costs and their adverse effects.26, 27 Thus, the use of a less expensive and cumbersome strategy will likely lead to improved compliance. Furthermore, the incremental increase in oral drugs taken with combination oral therapy is minimal. Patients were assumed to have equal rates of non–liver-associated deaths in each strategy.

In a recent study, a liver transplant recipient treated with an oral combination had recurrent HBV defined as the serological finding of HBsAg.28 However, HBV DNA was undetectable. The significance of HBsAg with undetectable HBV DNA is unclear. Indeed, in other studies, antiviral therapy resulted in a decrease in viral load that was associated with histologic improvement,29–38 an increase in the survival of patients with decompensated liver disease from hepatitis B,39–47 and improved clinical outcomes in patients with HBV reactivation.48, 49

Our study has several important limitations. First, the data used in these studies came from retrospective case series in which the number of patients varied from 3 to 1207, and the follow-up ranged from 6 months to 5 years.2–8, 14, 16, 18, 22, 50 We controlled for sample size by multiple regression analysis and varied the recurrence rates in a sensitivity analysis. Nevertheless, the results all consistently showed the efficacy of combination therapy in viral suppression. Moreover, combination therapy has been demonstrated to be effective in nontransplant mono-infected and co-infected patients.25, 51 Another limitation of our current model is that we did not consider the effect of pretransplant drug resistance on posttransplant outcome. However, we do not believe that this is a major limitation because 1 study has shown that patients who develop the YMDD mutation prior to transplant respond well to adefovir add-on therapy with undetectable serum HBV DNA after a median of 35.2 months.16 Similarly, in the study by Schiff et al.,14 40% of viremic posttransplant patients with lamivudine resistance who received adefovir had negative serum HBV DNA levels at 48 weeks, and prolonged treatment led to undetectable HBV DNA levels in 78% of patients by week 144. Additionally, Rapti et al.17 found that adding adefovir to lamivudine in patients with lamivudine resistance successfully suppressed HBV replication, which was maintained for at least 3 years. However, there are case reports in which patients with the hepatitis B YMDD mutant have suppression of viral replication with continued HBIG and lamivudine.52, 53

A third limitation is that there are few data concerning the treatment of resistance with lamivudine and adefovir combination therapy, likely because of the rarity of this scenario. In our model, we assumed a low rate of resistance to lamivudine and adefovir combination therapy and employed the use of entecavir and tenofovir. Entecavir and tenofovir are likely the most potent oral drugs in their respective classes.30, 54–56 In adefovir-resistant hepatitis B nontransplant recipients, the combination of tenofovir and lamivudine suppressed HBV DNA levels in all patients at a median duration of 16.5 months.57 Similarly, van Bömmel et al.58 demonstrated that the addition of tenofovir decreased HBV DNA in 95% of patients with adefovir- and lamivudine-resistant HBV infection within a median of 3.5 months. Further clinical trials are required to clarify this issue.

Another potential confounding variable that we did not take into account in our decision analysis model is renal impairment. Rates of renal failure in patients receiving lamivudine and adefovir, defined as elevations of creatinine of at least 0.5 mg/dL from baseline, have been reported to range from 0% to 12%.14, 16, 59 Although renal failure is a rare complication of adefovir, it is difficult to attribute renal impairment to adefovir, as most patients were also taking immunosuppressants. Using the Scientific Registry of Transplant Recipients, Ojo et al.60 found that slightly more than 25% of liver transplant recipients developed renal failure after 10 years. A study of adefovir for the treatment of nontransplant patients with chronic hepatitis B demonstrated no renal dysfunction in patients receiving the standard oral daily dose of 10 mg.29 However, 8% of patients receiving a high dose of 30 mg developed renal failure, with a maximum creatinine of 1.8 mg/dL.29 In all cases, the renal impairment resolved with dose reduction or interruption of treatment. Thus, patients should be monitored closely for evidence of renal failure, and the dosage of adefovir should be adjusted immedately.29, 61

The ease of administration, lower cost of adefovir and its administration, and decreased rates of HBV recurrence in strategy 1 promote its use. Thus, in institutions in which HBIG is more costly than adefovir, our decision analysis model demonstrates the pharmacoeconomic support for the use of lamivudine and adefovir as the first-line therapy for hepatitis B prophylaxis in liver transplant recipients after control of HBV disease post-transplantation.

Acknowledgements

The authors thank Professor Richard S. Marken, Ph.D., for statistical advice, as well as Michael D. Stone, D.O.

Ancillary