Hepatitis B prophylaxis in patients undergoing chemotherapy for lymphoma: A decision analysis model†
Article first published online: 6 AUG 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 46, Issue 4, pages 1049–1056, October 2007
How to Cite
Saab, S., Dong, M. H., Joseph, T. A. and Tong, M. J. (2007), Hepatitis B prophylaxis in patients undergoing chemotherapy for lymphoma: A decision analysis model. Hepatology, 46: 1049–1056. doi: 10.1002/hep.21783
Potential conflict of interest: Nothing to report.
- Issue published online: 25 SEP 2007
- Article first published online: 6 AUG 2007
- Manuscript Accepted: 18 APR 2007
- Manuscript Received: 19 JAN 2007
Hepatitis B reactivation is a major cause of morbidity and mortality in patients undergoing chemotherapy for lymphomas. These patients may experience direct liver-related complications or reduced cancer survival because of interruptions in chemotherapy. Our aim was to compare the costs and outcomes of 2 different chronic hepatitis B management strategies. In hepatitis B carriers undergoing chemotherapy, we pursued a decision analysis model to compare the costs and clinical outcomes of using lamivudine prophylaxis versus initiating lamivudine only when clinically overt hepatitis occurred. Our results indicate that the use of lamivudine prophylaxis is cost-effective. Even though the use of lamivudine prophylaxis was associated with an incremental cost of $1530 per patient ($18,707 versus $17,177), both the number and severity of hepatitis B reactivations were reduced. None of the patients in the prophylaxis group had liver-related deaths versus 20 who died in the no-prophylaxis group. Cancer deaths were also reduced from 47-39 with lamivudine prophylaxis, presumably because of the increased need for cessation or modification of chemotherapy in patients who had severe hepatitis B virus flares. The incremental cost-effectiveness ratio of using lamivudine prophylaxis was $33,514 per life year saved. Conclusion: Our results provide pharmacoeconomic support for the use of lamivudine prophylaxis in patients undergoing chemotherapy for lymphoma treatment. (HEPATOLOGY 2007.)
Hepatitis B virus (HBV) infection is an important medical and public health concern, and it is a leading cause of cirrhosis and hepatocellular carcinoma.1, 2 Patients with hepatitis B are at risk of viral reactivation when undergoing cytotoxic or immunosuppressive chemotherapy for malignancies, particularly lymphomas.3 The incidence of HBV reactivation may be as high as 50% and is associated with an overall liver-related mortality of 5%.4, 5 Even in patients who are able to recover from hepatitis B reactivation, their cancer survival may be impaired because of alterations or cessation of the initially intended chemotherapy course.6, 7
The pathogenesis of HBV reactivation during or after chemotherapy appears to be related to an increase in viral replication during immunosuppression with a resultant increase in hepatocyte destruction and inflammation upon immune reconstitution.3–5 Lamivudine, a nucleoside analogue, has been shown to be effective in inhibiting viral replication and decreasing viral load.8–10 Various strategies have been proposed in patients with hepatitis B undergoing chemotherapy to prevent viral replication. It has been suggested that the earlier use of lamivudine, before overt clinical hepatitis, is most efficacious.11 The use of prophylactic lamivudine can decrease the incidence of HBV reactivation from 30%-80% to 0%-17%.12–17
The aim of this study was to compare the efficacy and costs of 2 management strategies in patients with hepatitis B undergoing chemotherapy: (1) lamivudine prophylaxis and (2) lamivudine use only in patients with evidence of an HBV reactivation. Because lamivudine has been shown to be effective in preventing HBV reactivation, is inexpensive, and is associated with few adverse effects, we believed that the use of lamivudine prophylaxis would be most cost-effective.
Patients and Methods
We compared 2 strategies for the management of chronic hepatitis B patients undergoing chemotherapy for lymphoma treatment. In the first strategy, all patients were given lamivudine prophylaxis before the initiation of chemotherapy, and this continued for 6 months after the cessation of chemotherapy. Adefovir was coadministered if patients developed a hepatitis B reactivation despite lamivudine use. In the second strategy, patients received lamivudine only if there was clinical evidence for HBV reactivation. Adefovir was used in this model if patients developed lamivudine resistance, which was estimated to be 20% for the year of the model.18 A decision analysis model was created for each strategy to simulate the costs and health outcomes for patients assigned to each of these strategies (Fig. 1). The model was constructed with TreeAge (Williamstown, MA) software.
In each strategy, patients had an opportunity to develop an HBV reactivation. For every patient that did develop a reactivation, they could have been continued on chemotherapy or have had their chemotherapy aborted or modified. From that point, they could then go on to recover from the HBV reactivation or die from fulminant hepatic failure. Patients who recovered from or did not have HBV reactivation either died or survived their lymphoma. The follow-up time was 1 year for each of the groups. The institutional perspective was adopted in all analyses.
Natural History of Hepatitis B Reactivation.
The natural history of HBV reactivation in patients undergoing chemotherapy was obtained from a systematic review of Medline literature. Data from 14 studies published from various parts of the world between 1990 and 2006 that examined HBV reactivation in lymphoma patients undergoing chemotherapy were collected (Table 1).12–17, 19–26 Seven of these studies compared the prophylactic use of lamivudine during chemotherapy to the empiric use of the drug,12–17, 26 6 studies restricted lamivudine to the treatment of reactivation,20–25 and 1 study examined only the prophylactic use of lamivudine.19
|Prophylaxis (N)||No Prophylaxis (N)||Lymphoma (N)||Chemotherapy Regimen (N)||Reactivation|
|Prophylaxis (N)||No Prophylaxis (N)|
|Li et al.12||40||116||NHL 147||Anthracyclines 146||7||60|
|HL 9||Steroids 144|
|Leaw et al.13||11||53||NHL 64||CEOP, MACOP-B, PACEBOM, BACOP, ACVB, m-BACOD||0||17|
|Idilman et al.14||4||3||NHL 6||CHOP 5||0||2|
|HL 1||ABVD 1, 2-CdA 1|
|Lau et al.15||15||15||NHL 24||CHOP 3, CEOP 7||0||8|
|HL 6||ABVD 4, COPP 3, Others 13|
|Shibolet et al.16||7||4||NHL 10||CHOP 7||0||2|
|HL 1||Others 4|
|Lee et al.17||11||20||NHL 31||CHOP 29, EDAP 1, proMACE-cytaBOM 1||1||17|
|Rossi et al.19||13||0||NHL 12||CHOP 6, FL 2||2||N/A|
|HL 1||ACVBP 2, others 3|
|Liao et al.20||0||11||NHL 11||CHOP 11||N/A||6|
|Faggioli et al.21||0||22||NHL 22||CNOP, CEOP, proMACE-CytaBOM||N/A||9|
|Yeo et al.22||0||15||Unspecified||Unspecified||N/A||6|
|Picardi et al.23||0||12||NHL 12||Fludarabine-based without steroids||N/A||5|
|Markovic et al.24||0||44||NHL 41||Unspecified||N/A||8|
|Lok et al.25||0||78||NHL 78||BACOP, m-BACOD, others||N/A||28|
|Persico et al.26||3||18||NHL 21||Unspecified||0||12|
Lymphoma and Chemotherapy.
Cancer survival was highest in patients receiving prophylaxis, regardless of the chemotherapy course.12–16, 19 Those who did not receive prophylaxis but were able to complete chemotherapy uneventfully12–16, 20, 22, 23 had higher survival rates than those who did not receive prophylaxis and had interruptions in their chemotherapy.12–16, 20, 22, 23 The cancer-related death rate for patients who did not have hepatitis B viral reactivation was not available from these studies and was assumed to be equal to the cancer death rate in the prophylaxis group of patients who did not have HBV reactivation and were able to complete their scheduled chemotherapy. If HBV reactivation did occur, the mean time to the episode was after the third cycle of chemotherapy.13–15, 19–25
Treatment and Follow-Up Strategies.
The lamivudine prophylaxis strategy was made to simulate current recommendations for the management of patients with chronic hepatitis B undergoing chemotherapy for lymphoma.3 Patients were started on daily lamivudine 1 week before the initiation of chemotherapy, and this continued for 6 months after the cessation of chemotherapy. If a patient in either treatment strategy developed an HBV reactivation while on lamivudine, he was then started on daily adefovir in combination with lamivudine. The use of adefovir was based on studies that treated patients who developed lamivudine-escape mutants while undergoing chemotherapy.27–29 The rate of lamivudine resistance was estimated to be 20% at 1 year.18, 30, 31 Although there have been reports of HBV reactivation after lamivudine withdrawal, we did not include this phenomenon in our model because our study design used lamivudine for 6 months following chemotherapy cessation.32, 33 The intended chemotherapy course was, on average, 6 cycles, each 3 weeks apart. Patients were followed in an oncology clinic once per chemotherapy cycle (3 weeks) and in a hepatology clinic once every 3 months. If an HBV reactivation developed or if there was clinical suspicion of an HBV reactivation, the frequency of the hepatology clinic visits was increased to once every month. Laboratory tests were checked every 3 weeks, including a complete blood count, chemistry panel (including creatinine), and liver tests (to include bilirubin, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, total protein, and albumin). HBV DNA levels were checked only in those with suspected reactivation.
The institutional cost perspective was used in the decision analysis model. All costs for medical procedures and visits were obtained from Medicare reimbursement schedules or from current literature (Table 1), and all drug costs were derived from average wholesale prices.34 The cost of end-of-life care was derived from mean annual medical expenditures and adjusted for 2006 dollars.35 Adjustments were made by healthcare inflation rates according to the Bureau of Labor Statistics.36 All costs are denoted in US dollars for the year 2006 and rounded to the nearest dollar amount. One-way sensitivity analyses were carried out for corresponding costs of lymphoma and HBV treatments. The baseline costs increased or decreased by 50%.
Effectiveness was defined as increased survival in the number of years per patient. Adjustments for the quality of life were not performed. The incremental cost-effectiveness ratio (ICER) was calculated as the ratio of the change in the cost to the change in the health benefit (measured in life years).37 The ICER was compared to the generally accepted price ceiling of $50,000 per life year saved.37 All variables, including the costs of treatment, reactivation rates, and survival rates were evaluated in 1-way sensitivity analyses to determine their individual effects in the overall cost-effectiveness decision (shown later in Table 5). A 2-way sensitivity analysis was executed to establish the values at which the administration of prechemotherapy prophylaxis was most cost-effective. In this method, 2 parameters were varied simultaneously to model changing clinical environments. Pairs of values were elucidated that provided equal cost benefits for both clinical treatment options (Fig. 2).37, 38
|Baseline||Range||Prophylaxis Preferred||No Prophylaxis Preferred|
|Cancer mortality 1*||0.078||0–0.429||0–0.31||0.31–0.429|
|Cancer mortality 2†||0.078||0–0.429||0–0.23||0.23–0.429|
|Cancer mortality 3‡||0.089||0–0.75||0.016–0.75||0–0.016|
|Cancer mortality 4§||0.109||0–0.14||0.077–0.14||0–0.077|
|Cancer mortality without reactivation nor prophylaxis||0.078||0–0.429||0.0622–0.429||0–0.0622|
|Chance of reactivation with prophylaxis||0.096||0–0.175||0–0.175||No values|
|Chance of reactivation without prophylaxis||0.438||0.182–0.85||0.38–0.85||0–0.38|
|Chance of aborting/delaying chemotherapy without prophylaxis||0.711||0–1||0.4–1||0–0.4|
|HBV mortality without prophylaxis with aborted/delayed chemotherapy||0.109||0–0.14||0.073–0.14||0–0.073|
|HBV mortality on prophylaxis with aborted/delayed chemotherapy||0||0–1||0–0.2||0.2–1|
|HBV mortality on prophylaxis||0||0–1||0–0.3||0.3–1|
|Cancer mortality with prophylaxis and no reactivation||0.078||0–0.429||0–0.088||0.088–0.429|
In our model, we assumed the following:
- 1All issues not related to HBV should occur with equal frequency in all arms of the study, and they therefore were not included in the model.
- 2All patients who did not have HBV reactivation should be able to complete chemotherapy without interruption.
- 3All lymphoma patients were treated with cyclophosphamide, hydroxydaunomycin, vincristine, and prednisone, as this was the most frequently used regimen in the referenced publications (Table 1).
- 5There was no significant morbidity or mortality related to the use of either lamivudine or adefovir.
- 6Cancer survival was similar between patients who had a reactivation but were able to complete chemotherapy and those who did not have an HBV reactivation.
- 7Liver transplantation was not an option for patients with severe HBV reactivation because of lymphoma.
We also assumed 100% compliance in all patients for both chemotherapy and HBV-related medications.
Using the decision analysis model, we were able to simulate the clinical outcomes, effectiveness, and costs associated with prophylaxis for hepatitis B reactivation in patients receiving chemotherapy. To increase the generalizability of our model, we varied all event probabilities and costs (Tables 2 and 3).
|End-of-life care||$35,000 (17,500-52,500)||35, 36|
|CHOP cycle||$446 (184-669)||34|
|Clinical visit*||$242 (121-363)||Medicare|
|Comprehensive metabolic panel||$70 (35-105)||Medicare|
|Hepatitis B DNA||$120 (60-180)||Medicare|
|CBC with differential and platelets||$110 (55-165)||Medicare|
|Reactivations among HBV carriers|
|Prophylaxis||9.6 (0–17.5)||12–17, 19, 26|
|No prophylaxis||43.8 (18.2–85.0)||12–17, 20–26|
|Disruption of chemotherapy due to HBV reactivation|
|Prophylaxis||60.0 (50–100)||12–14, 17, 19|
|No prophylaxis||71.1 (0–100)||12, 14, 17, 20, 23|
|HBV-associated liver death in those who continued chemotherapy|
|Prophylaxis||0||12, 14, 17, 19|
|No prophylaxis||3.8 (0–25)||12, 14, 17, 20, 23|
|HBV-associated liver death in those who had disruption in chemotherapy|
|Prophylaxis||0||12, 14, 17, 19|
|No prophylaxis||10.9 (0–14.0)||12, 14, 17, 20, 23|
|Cancer death in those who had HBV reactivation and continued chemotherapy|
|Prophylaxis||7.8 (0–42.9)||12–16, 19|
|No prophylaxis||8.9 (0–75)||12–16, 20, 22, 23|
|Cancer death in those who had HBV reactivation with disruption in chemotherapy|
|Prophylaxis||7.8 (0–42.9)||12–16, 19|
|No prophylaxis||10.9 (0–75)||12, 14–16, 19, 20, 22, 23|
|Cancer death in those who did not have HBV reactivation|
|No prophylaxis||7.8 (0–42.9)||—†|
Using the baseline assumptions derived from published literature, we extrapolated the clinical costs and outcomes to a cohort of 1000 patients, with 500 in each treatment strategy group (Table 4). We found that prophylaxis increased the cost to $18,707 from $17,177 per patient. However, prophylaxis was effective in reducing the number of HBV reactivations (48 versus 219), liver-associated deaths (0 versus 20), and cancer-associated deaths (39 versus 47). Prophylaxis was also associated with prolonging life years from 0.876 in the no-prophylaxis group to 0.922 in the prophylaxis group. The ICER was $33,514 per life year. With a ceiling of $50,000 per patient per year, prophylaxis was the preferred strategy.
|Total costs per patient||$18,707||$17,177|
|Reactivations (number per arm)||48||219|
|Effectiveness (LY per patient)||0.922||0.876|
|Deaths (number per arm)|
Lamivudine prophylaxis was the most cost-effective strategy in a 1-way sensitivity analysis for most varied variables (Tables 4 and 5). The incremental cost-effectiveness of prophylaxis exceeded $50,000 per life year saved when the costs of dying were less than $18,500 and the monthly lamivudine cost was greater than $333. Also, as the incidence of reactivations in the no-prophylaxis group decreased below 0.38, no prophylaxis became cost-effective. As cancer mortality increases for patients receiving prophylaxis or as cancer mortality decreases for patients not receiving prophylaxis, no prophylaxis becomes cost-effective. In a 2-way sensitivity analysis (Fig. 2), the changing cost of lamivudine was compared to a changing rate of HBV reactivation without prophylaxis. Rising costs coupled with proportionally increased reactivations did not modify the most cost-effective decision.
Hepatitis B reactivation poses a significant risk to patients undergoing chemotherapy. Recent studies have shown that prophylaxis with lamivudine before the initiation of chemotherapy reduces both the incidence and severity of HBV reactivations.12–17, 19–26 However, no studies have investigated the use of prophylactic lamivudine in an economic analysis. In this study, we compared the costs and clinical outcomes of 2 different strategies for chronic hepatitis B carriers undergoing chemotherapy for lymphoma treatment. Our results indicate that the prophylactic use of lamivudine is a cost-effective strategy in comparison with the initiation of lamivudine only when overt hepatitis becomes evident.
Under baseline assumptions, the lamivudine prophylaxis strategy was cost-effective, with an ICER of $33,514 per life year. No liver-related deaths occurred in the prophylaxis group versus 20 deaths in the no-prophylaxis group. Cancer deaths were also reduced from 47 in the no-prophylaxis group to 39 in the prophylaxis group. To increase the generalizability of our study, 1-way sensitivity analyses were performed on all variables. Lamivudine prophylaxis was the preferred strategy when the cost of death was high, the cost of lamivudine was low, and, for most clinical event ranges, all baseline assumptions were included.
Our study has several limitations. First, our tested parameters were from studies including few patients with nonrandomized groups. The studies also differed in their lymphoma subtypes, inclusion criteria, and management strategies. Nevertheless, we were able to calculate a weighted mean and varied rates like those described in the literature to increase generalizability. Another limitation was that the use of adefovir in lamivudine-resistant patients undergoing chemotherapy has been described only in the context of case reports.27–29 In nonchemotherapy patients, the sequential use or combination use of adefovir has been established and is believed to be effective in lamivudine resistance.42, 43 Other medications against HBV such as entecavir and telbivudine were not included in our model because of the lack of data. The duration of time for which lamivudine was continued after the cessation of chemotherapy was another potential limitation. This varied significantly between the 14 studies, from 4 weeks to 2 years. There is no general consensus for the optimal duration of lamivudine treatment after the cessation of chemotherapy, but on the basis of the discussions in several publications, we proposed continuing lamivudine for 6 months after the cessation of chemotherapy.16, 33 The wide range in the results may have been due in part to the variability of lamivudine usage.
Several clinical problems arise with the use of lamivudine as a monotherapy. For instance, the rate of lamivudine resistance is estimated to be approximately 20% at 1 year.18, 30, 31 Our model accounted for the risk of lamivudine resistance by the addition of adefovir in cases of escape mutants. This strategy has been shown to be effective in a number of studies.27–29 The selection of lamivudine-resistant strains may also lead to cross-resistance to future therapies.44, 45 Our analysis was performed with lamivudine, which, to the best of our knowledge, has been the only drug studied for prophylactic purposes in patients receiving chemotherapy. The lack of clinical experience with alternative therapies such as adefovir, entecavir, and telbivudine in this setting limits their incorporation into the current cost-effectiveness study. Another potential limitation with the use of lamivudine is the phenomenon of HBV reactivation after drug withdrawal.32, 33 However, the phenomenon has been described in only 2 studies in which patients were treated for a short time after chemotherapy cessation (4 weeks to 3 months). In our model, we did not include the phenomenon of HBV reactivation because lamivudine was continued for a prolonged period of time (i.e., 6 months) after chemotherapy cessation.
Lymphomas are a heterogeneous group of diseases. Although certain subtypes may not require any treatment, others are almost universally fatal. Our results are not likely to be generalizable to other disease states because of potentially unique rates and timing of HBV reactivation. Future studies should explore the costs and effectiveness of HBV prophylaxis in different disease states such as hepatocellular carcinoma,46 breast cancer,47 nasopharyngeal carcinoma,48 rheumatologic diseases,49 renal transplants,50 lung transplants,51 and inflammatory bowel diseases.52
In our model, the prophylactic use of lamivudine was associated with 171 fewer HBV reactivations and 28 fewer overall deaths. It follows logically then that chronic HBV carriers should be started on lamivudine prophylaxis before chemotherapy and those whose HBV status is unknown should be screened for the presence of HBV infection. In parts of the world where HBV is endemic, such as Southeast Asia, sub-Saharan Africa, and parts of the Pacific and Amazon basins, the prevalence of HBV is as high as 8%-20%, and screening is certainly warranted. Even in nonendemic parts, such as North America and Western/Northern Europe, the prevalence is estimated to be 0.2%-0.5%.53 Given the relative ease of screening with the hepatitis B surface antigen, we believe that this is a cost-effective and life-saving strategy.
To date, there is no consensus on whether patients with positive DNA levels should be treated before the initiation of chemotherapy or whether there is an HBV DNA level below which the initiation of chemotherapy is considered safe. A recent case series by Yeo et al.11 suggested that viral replication precedes clinical hepatitis by approximately 2 weeks in patients undergoing chemotherapy. However, because the level of HBV DNA that may be associated with hepatitis B reactivation has not been determined, a safer strategy may be to treat all patients with a positive hepatitis B surface antigen test at the initiation of chemotherapy, regardless of the HBV DNA status.
In summary, this study provides strong support for the use of lamivudine for prophylaxis against hepatitis B reactivation in patients undergoing chemotherapy for lymphomas. The results of our model indicate that it is cost-effective to prescribe lamivudine in patients with known hepatitis B 1 week before chemotherapy and for 6 months after the cessation of chemotherapy.
We thank Martin L. Lee, Ph.D., of the Department of Biostatistics at the University of California at Los Angeles for his expert opinion regarding the statistical analysis. We also thank Mark Roach for his assistance in obtaining cost data.
- 10Lamivudine therapy for chemotherapy-induced reactivation of hepatitis B virus infection. Am J Gastroenterol 1999; 94: 249–251., .Direct Link:
- 29Adefovir dipoxil to prevent exacerbation of lamivudine-resistant hepatitis B infection during chemotherapy for non-Hodgkin's lymphoma. Am J Gastroenterol 2004; 99: 1619–1620., , , , .Direct Link:
- 34PDR Staff. Red Book. 110th ed. Thomson Healthcare (Montvale, NJ); 2006.
- 36U.S. Department of Labor Bureau of Labor Statistics. Consumer price index calculator. Available at: http://www.bls.gov. Accessed May 2007.
- 38Meta-Analysis, Decision Analysis, and Cost-Effectiveness Analysis. New York, NY: Oxford University Press; 2000..
- 53World Health Organization. Hepatitis B. Available at: http://www.who.int/csr/disease/hepatitis/whocdscsrlyo20022/en/index1.html. Accessed May 2007.