- © 2012 by American Society of Clinical Oncology
Hepatitis B, Rituximab, Screening, and Prophylaxis: Effectiveness and Cost Effectiveness
- Corresponding author: Dr. Ralph M. Meyer, Director, NCIC CTG, Cancer Research Institute, Queen's University, 10 Stuart St., Kingston, Ontario, Canada, K7L3N6; e-mail: RMeyer{at}ctg.queensu.ca.
An important role of medical journals is to communicate information between stakeholder groups.1,2 These communications may be between researchers about findings that inform future research; results of phase I-II trials in Journal of Clinical Oncology (JCO) exemplify this purpose. Alternately, communications between practitioners can advise about implementing clinical practices, such as with narrative reviews and case-based manuscripts including JCO's Oncology Grand Rounds, which provide guidance to practitioners.3 Communications from investigators to practitioners and policy makers include results of randomized controlled trials (RCTs) and systematic reviews. These communications inform decisions about managing individual patients and health care delivery policies. For policy determination, economic evaluations also have an important role. Implicit in conducting an economic analysis is prior demonstration that the intervention is effective. With this knowledge in hand, understanding economic ramifications of adopting an intervention may be very helpful: a central premise is that resources are scarce and decisions about alternatives are associated with an opportunity cost because resources used for one purpose are unavailable for another use. JCO has provided guidance to researchers intending to submit a report of an economic analysis4; high priority is given to analyses that affect decisions about adoption.
In the article that accompanies this editorial, Zurawska et al5 provide a cost-effectiveness analysis assessing hepatitis B virus (HBV) screening before administering rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP) chemotherapy to patients with diffuse large B-cell lymphoma (DLBCL). Their analysis links data demonstrating that when treated with R-CHOP, patients with DLBCL who are chronically infected with HBV can experience reactivation of HBV infection leading to acute hepatitis with consequences including morbidity associated with infection, compromise of chemotherapy delivery that might result in morbidity and mortality associated with suboptimal control of lymphoma, and death from fulminant hepatitis.6,7 This analysis assumes that effective strategies for screening and prophylaxis exist. The authors conclude that routine screening of all patients with DLBCL before chemotherapy is cost effective because, in comparison with alternatives, it is least costly and associated with superior 1-year survival. Setting aside the above assumptions and implications of the authors' conclusions, this analysis follows published rules for reporting an economic evaluation,8 as the study question includes clear alternatives (screen all, screen high-risk patients, screen none), the perspective of the analysis is provided (a Canadian province's Ministry of Health and Long Term Care), costing appears to be comprehensive and credibly valued, and results were provided using incremental differences and with associated sensitivity analyses. A minor criticism is that conclusions may be overstated because the differences in costs and life-years saved between alternatives are marginal; stating that a screen-all strategy falls well within standard benchmarks for cost effectiveness would be a more conservatively stated conclusion. Screening all patients with DLBCL would be expected to be even more cost effective in jurisdictions with HBV prevalence rates that exceed those observed in Canada.
At issue is how this economic analysis informs decisions to adopt a strategy to screen all patients with DLBCL for HBV. Other information helps to frame Zurawska et al's5 conclusions. First, the problem is important: reactivation of HBV in patients with cancer receiving chemotherapy is recognized and available data show that this risk is increased with the more profound immunosuppression associated with lymphoma and treatment that includes steroids and now rituximab.6,7,9 Second, authors of a meta-analysis7 that included two RCTs and additional observational data concluded that while these data are associated with important limitations, a reasonable interpretation is that prophylaxis with lamivudine is preferred over a no-treatment approach when patients with cancer testing positive for hepatitis B surface antigen (HBsAg) are treated with chemotherapy. Third, synthesis of these two points has resulted in publication of numerous guidelines recommending screening for HBV and antiviral prophylaxis for cancer patients with positive testing10–15 (Table 1), especially when rituximab is used to treat lymphoma. Screening has also been recommended for patients who are to receive rituximab for treatment of benign conditions.16 Finally, despite these guidelines, practice variation exists with poor uptake of these recommendations: Zurawska et al cite data reporting routine screening in their geographic region of only 14% of patients with cancer who are to receive chemotherapy,17 an Australian survey that included oncologists who treat lymphoma reported that 47% never screen for HBV before initiating chemotherapy18 and in a North American teaching hospital only 36.6% of patients treated with rituximab between 1997 and 2009 had HBV testing although rates increased to 67.4% after introduction of guidelines.19
Selected Guidance Documents With Recommendations for Hepatitis B Screening
A previous survey has suggested that one reason for variable adoption may be concern about cost effectiveness.18 Zurawska et al's5 conclusions might thus promote adoption of screening for patients with lymphoma, especially if concerns about cost effectiveness were accentuated by a recent publication in JCO by Day et al,20 which deemed routine prechemotherapy screening of patients with solid tumors cost ineffective. In that analysis, the cost per life-year saved was $88,224 for patients receiving adjuvant therapy and $1,344,251 for patients receiving palliative chemotherapy (values in Australian dollars). Review of the methodologies of the Day et al and Zurawska et al economic evaluations prompt speculation that both effectiveness and cost effectiveness of screening and prophylaxis are amplified for patients with DLBCL. In comparison with populations of other patients with cancer receiving chemotherapy, those with DLBCL may have greater background risks of HBV infection, experience more severe immunosuppression, and are treated with curative rather than palliative intent and in a manner that is associated with effect sizes that exceed those anticipated when treating most patients with solid tumors. These differences are now even greater because, in comparison with CHOP, R-CHOP is associated with superior disease control and overall survival, but also is more immunosuppressive and appears to be associated with greater risks of HBV reactivation, and clinical and severe HBV-related hepatitis.21 Thus, preventing HBV reactivation is now more important and success should be associated with opportunities for greater benefit. These suppositions also support an argument that the 1-year survival end point used by Zurawska et al is conservative, as long-term survival and the morbidity associated with HBV reactivation are undervalued.
Will Zurawska et al's5 analysis alter physician behaviors, strengthen recommendations from professional societies and health care agencies, and reduce practice variation? While a commonly recommended practice policy is now associated with favorable economic parameters, a larger question relates to a core principle of economic evaluations and the assumptions on which this analysis was based: is there sufficient evidence demonstrating effectiveness of screening and prophylaxis strategies for HBV? In 2010, the American Society of Clinical Oncology (ASCO) provided a Provisional Clinical Opinion12 (PCO) addressing these questions (Table 1); the PCO included different conclusions than those recommended by other agencies and the reason for these differences may help explain why incomplete adoption has occurred. In contrast with other bodies, such as the U.S. Center for Disease Control,10 the ASCO PCO concluded insufficient evidence exists to determine net benefits and harms for routine screening in individuals with cancer and recommends screening only be considered (as opposed to universally performed) for patients with lymphoma who are to receive rituximab; the conclusion that there was inadequate evidence applied to both HBsAg testing as a screening tool and to prophylactic treatment for those with a positive test. The conceptual basis of the PCO process includes recognition that new scientific evidence is complex, develops rapidly and that the label of “Provisional” necessitates regular review in order to assure that a goal to answer “does this change my practice?” is addressed.22 So, at least with respect to treating DLBCL, should the ASCO PCO be updated? If updated, Zurawska et al's analysis indicates that effectiveness of prophylaxis should be emphasized; while not minimizing the importance of what constitutes optimum screening, their data suggest that even for populations with relatively low prevalence rates, screening is likely to be cost effective.
The ASCO PCO statements could be considered for three populations: patients with nonlymphoma cancer, patients with DLBCL who were previously treated with CHOP and current patients with DLBCL undergoing active treatment with R-CHOP. Policies for those with solid tumors require separate consideration as these patients are less likely to develop severe hepatitis, and patients receiving palliative chemotherapy for metastatic disease are more likely to have cancer-related risks that affect morbidity and mortality. The uncertainties associated with limited evidence referred to in the ASCO PCO most directly apply to this population. Zurawska et al's5 analysis requires that up-to-date lymphoma-specific evidence be considered. The only two RCTs evaluating patients with DLBCL23,24 included treatment with CHOP, and not R-CHOP, and contributed to the assumptions used by Zurawska et al. Results of these trials showed reduced rates of HBV reactivation and HBV-related clinical hepatitis and severe hepatitis. The magnitude of reduction of clinically apparent HBV hepatitis (45% versus 5%) was substantial, but the pooled sample size was only 84 patients. These trials provide insufficient data for conclusions about risks of clinical hepatitis following cessation of prophylactic therapy and long-term survival, but were interpreted by developers of most guidelines as sufficient for adoption of prophylaxis. Use of R-CHOP creates new complexities. Risks of reactivation in patients with HBsAg positivity are increased and a new risk group is recognized that includes those with testing that is negative for HBsAg but positive for hepatitis B core antibody (anti-HBc).9,21 Furthermore, R-CHOP is associated with reactivation risks that persist after completing prophylactic therapy,25,26 a finding that coincides with discovery of the YMDD mutation,27 which is associated with lamivudine resistance. Thus, guidelines have moved beyond whether these patients should be screened and receive prophylaxis and onto debate about which screening tests to employ, which prophylactic agent to use and the optimum duration of therapy. These debates assume that added risks associated with rituximab-related HBV reactivation create a larger population in need of prophylaxis and that the efficacy of prophylaxis observed in two small RCTs evaluating patients treated with CHOP will not be compromised by the more severe and prolonged immunosuppression associated with R-CHOP.
It is unlikely that practitioners and policy makers will be informed with definitive data from RCTs in the near future. As indicated by Zurawska et al,5 a RCT evaluating screening will probably never be performed. Thus, communications to practitioners and policy makers about today's best practices need to emphasize the importance and yet remaining uncertainties associated with this decision-making process. The economic analysis of Zurawska et al provides helpful information addressing one of these uncertainties. Their work contributes to available information supporting as a standard of care routine HBV screening of patients with DLBCL who are to receive R-CHOP and provision of prophylactic therapy to those with a positive test.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory Role: None Stock Ownership: None Honoraria: Ralph M. Meyer, Eli Lilly, Celgene Research Funding: Ralph M. Meyer, Amgen, ARIAD Pharmaceuticals, Astex Therapeutics, AstraZeneca, Boston Biomedical, Bristol-Myers Squibb, Celgene, Geron, GalxoSmithKline, Janssen Pharmaceuticals, Eli Lilly, Merck Frosst Canada, Novartis, Oncolytics Biotech, Oncothyreon, Pfizer, Roche, sanofi-aventis, Schering-Plough Canada Expert Testimony: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Provision of study materials or patients: Annette E. Hay
Manuscript writing: All authors
Final approval of manuscript: All authors
