Potential conflict of interest: Nothing to report.
Lamivudine is effective to control hepatitis B virus (HBV) reactivation in HBV-carrying cancer patients who undergo chemotherapy, but the optimal treatment protocol remains undetermined. In this study, HBV carriers with newly diagnosed non-Hodgkin's lymphoma (NHL) who underwent chemotherapy were randomized to either prophylactic (P) or therapeutic (T) lamivudine treatment groups. Group P patients started lamivudine from day 1 of the first course of chemotherapy and continued treatment until 2 months after completion of chemotherapy. Group T patients received chemotherapy alone and started lamivudine treatment only if serum alanine aminotransferase (ALT) levels elevated to greater than 1.5-fold of the upper normal limit (ULN). The primary endpoint was incidence of HBV reactivation during the 12 months after starting chemotherapy. During chemotherapy, fewer group P patients had HBV reactivation (11.5% versus 56%, P = 0.001), HBV-related hepatitis (7.7% versus 48%, P = 0.001), or severe hepatitis (ALT more than 10-fold ULN) (0 versus 36%, P < 0.001). No hepatitis-related deaths occurred during protocol treatment. Prophylactic lamivudine use was the only independent predictor of HBV reactivation. After completion of chemotherapy, the incidence of HBV reactivation did not differ between the 2 groups. Two patients, both in group P, died of HBV reactivation–related hepatitis, 173 and 182 days, respectively, after completion of protocol treatment. When compared with an equivalent group of lamivudine-naïve lymphoma patients who underwent chemotherapy, therapeutic use of lamivudine neither reduced the severity of HBV-related hepatitis nor changed the patterns of HBV reactivation. Conclusion: Prophylactic lamivudine use, but not therapeutic use, reduces the incidence and severity of chemotherapy-related HBV reactivation in NHL patients. (HEPATOLOGY 2008;47:844–853.)
Hepatitis B (HBV) reactivation and hepatitis flare induced by cytotoxic chemotherapy is common in cancer patients who have chronic HBV infection.1, 2 It is best characterized in patients with hematological malignancies but also can occur in patients with solid tumors. Hepatic decompensation and death attributable to HBV reactivation occurred to 5% to 40% of HBV carriers who underwent chemotherapy. Because most patients who develop HBV reactivation and hepatitis have to postpone their scheduled chemotherapy,3 prevention of HBV reactivation may improve the treatment outcome of cancer patients with chronic HBV infection.
Lamivudine is the first nucleoside analog to show efficacy in the treatment of chronic and acute HBV infection.4–6 The preventive effect of lamivudine on chemotherapy-induced HBV reactivation was demonstrated in a randomized clinical trial as well as in many prospective case series with or without historical control.7–10 The incidence of HBV reactivation for patients who received lamivudine prophylaxis in these studies ranged from 0 to 20%, compared with 33% to 67% in controls. The incidence and severity of hepatitis flare was also significantly reduced by lamivudine prophylaxis.
Although the preventive effect of lamivudine is well known, several important issues remain undetermined. First, the duration of lamivudine prophylaxis varied among different clinical series. Because HBV reactivation after withdrawal of lamivudine prophylaxis is common, prolonged lamivudine use is recommended, but the optimal duration is unknown. Second, long-term lamivudine treatment may induce mutation in the tyrosine-methionine-aspartate-aspartate motif of the HBV DNA polymerase gene (the YMDD mutation), which can cause hepatitis flare.11 Whether the induction of YMDD mutation by lamivudine prophylaxis jeopardizes the clinical outcome of the cancer patients remains unclear. Third, the effect of therapeutic use of lamivudine, starting after clinical evidence of hepatitis, is not fully established. Although early studies indicated that lamivudine treatment was effective for patients with developed hepatitis,12, 13 spontaneous recovery of viral reactivation and hepatitis was also common.
The study's objective is to compare the efficacy of prophylactic and therapeutic use of lamivudine for chemotherapy-induced HBV reactivation in non-Hodgkin's lymphoma (NHL) patients. NHL has been considered a high-risk population for chemotherapy-induced HBV reactivation.14 To further explore the potential efficacy of therapeutic use of lamivudine, the results of this study were compared with those of our previous study in which the natural course of HBV reactivation in NHL patients was prospectively followed during and after chemotherapy.15
ALT, alanine aminotransferase; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisolone; CI, confidence interval; HBeAg: hepatitis B e antigen; HBV, hepatitis B virus; IPI, International Prognostic Index; NHL, non-Hodgkin's lymphoma; ULN, upper limit of normal; YIDD, tyrosine-isoleucine-aspartate-aspartate; YMDD, tyrosine-methionine-aspartate-aspartate; YVDD, tyrosine-valine-aspartate-aspartate.
Patients and Methods
Study Design and Patient Eligibility.
This study, designed by the Taiwan Cooperative Oncology Group and approved by the Ethics Committee of the National Health Research Institutes, Taiwan, in 2001, was conducted in 10 clinical centers in Taiwan. After obtaining informed consent, the patients were checked for eligibility for this clinical trial. Patients who fulfilled all the eligibility criteria were then randomized in a 1:1 ratio to receive either prophylactic or therapeutic lamivudine treatment. The Statistical Center of the Division of Biostatistics of the National Health Research Institutes generated the randomization list by permuted block randomization. The randomization code was given only when the patient passed the eligibility check.
Participants were patients with newly diagnosed, histologically proven intermediate-grade or high-grade NHL and positive serum HBV surface antigen. Key eligibility criteria included: age 16 to 75 years, alanine aminotransferase (ALT) less than 5 times the upper limit of normal (ULN), bilirubin < 2.5 mg/dL, neutrophil ≥ 2000/mm3, platelet ≥ 100,000/mm3, creatinine ≤ 1.5 mg/dL, urea nitrogen ≤ 25 mg/dL, Eastern Cooperative Oncology Group performance score 0 to 2, and measurable tumors on clinical imaging. Key exclusion criteria included Child-Pugh class B or C cirrhosis, grade 2 or greater heart failure by the New York Heart Association classification, previous chemotherapy or radiotherapy, concurrent glucocorticoid therapy for other reasons, or other primary liver diseases, such as chronic hepatitis C, hepatitis D, autoimmune hepatitis, or Wilsons' disease.
All participants received the chemotherapy regimen consisting of cyclophosphamide, 750 mg/m2, doxorubicin, 50 mg/m2, and vincristine 1.4 mg/m2 intravenously on day 1 and prednisolone, 60 mg/m2/day orally on days 1 to 7 (CHOP regimen). The treatment cycles were repeated every 21 days. Participants who achieved a complete response were given 2 more cycles of chemotherapy, for a minimum of 6 cycles. Participants who achieved a maximal response of partial response, stable disease, or progressive disease were changed to second-line chemotherapy. The regimens were at the discretion of individual investigators in different hospitals. Radiation therapy for residual localized tumors was allowed.
Participants who were randomized to the prophylactic group started lamivudine treatment, 100 mg/day orally, on day 1 of the first course of chemotherapy. Lamivudine treatment was continued until 2 months after the completion of chemotherapy, when chemotherapy-induced myelosuppression resolved, as confirmed by follow-up hemogram examinations. If second-line chemotherapy was used, lamivudine at the same dosage was continued until 2 months after completion of the second-line chemotherapy. Patients who were randomized to the therapeutic group started lamivudine therapy, 100 mg/day orally, when elevation of ALT was noted during follow-up, and continued lamivudine treatment until hepatitis resolved. For patients whose baseline ALT was normal, the cutoff ALT level for starting lamivudine treatment was greater than 1.5-fold ULN. For patients whose baseline ALT was abnormal, the cutoff level was greater than 2.0-fold ULN. After resolution of hepatitis, prophylactic lamivudine was not given in subsequent chemotherapy.
Outcomes and Follow-up.
The primary endpoint was the incidence of HBV reactivation during and within 12 months after chemotherapy. Serum samples were collected at baseline, before the start, and at the nadir of every course of chemotherapy to check hemogram, liver function, and HBV DNA levels. Hepatitis flare was defined as a greater than 3-fold increase of serum ALT level that exceeded 100 IU/L. The hepatitis flare was attributed to HBV reactivation if it was preceded or accompanied by a greater than 10-fold increase, compared with previous nadir levels, of HBV DNA or by the reappearance of hepatitis B e antigen (HBeAg) in the serum for patients whose baseline HBeAg was negative. For patients who developed hepatitis, serum samples were collected every week until the hepatitis resolved, regardless of the chemotherapy schedule.
After completing the protocol treatment (CHOP chemotherapy plus lamivudine for the prophylactic group, CHOP chemotherapy alone for the therapeutic group), serum samples were collected every month during follow-up for 12 months for the examination of liver function and HBV DNA levels. For patients who underwent second-line chemotherapy, serum samples were collected every month until 3 months after the last course of second-line chemotherapy.
All the HBV DNA quantification and genotyping tests were performed by the Hepatitis Research Center of National Taiwan University Hospital. Serum DNA was extracted by using the QIAamp DNA Blood Mini Kit (Qiagen Inc., Valencia, CA). Methods of HBV DNA quantification and HBV genotyping were described previously.16 Detection of the YMDD mutation was confirmed by sequencing.
During protocol treatment, patients were followed up by computed tomography after every 2 cycles of chemotherapy to document the response to chemotherapy. The response was evaluated according to the standardized response criteria recommended by the National Cancer Institute.17
To further explore the effect of therapeutic use of lamivudine on the outcome of hepatitis, data from our previous study of chemotherapy-induced HBV reactivation was used as historical control.15 In the previous study, patients with newly diagnosed NHL were regularly followed up for HBV reactivation and HBV-related hepatitis flares during and after chemotherapy. The severity of hepatitis flare and the types of HBV reactivation were compared.
The sample size calculation was based on the assumption that the incidence of HBV reactivation was 50% and 80% for the prophylactic group and the therapeutic group, respectively. Based on 2-sided tests with a significance level of 0.05 and a statistical power of 80%, recruiting of 90 patients in the 2 arms was planned. Secondary endpoints included the incidence of hepatitis, hepatic failure, the objective response rate to chemotherapy, and overall survival. The Data and Safety Monitoring Committee of the Taiwan Cooperative Oncology Group reviewed the safety data of this study every 6 months.
Survival was calculated by the Kaplan-Meier method and compared by the log-rank test. Cox regression and unconditional logistic regression were applied to estimate the hazard ratios and odds ratios, respectively. Categorical variables were compared by the Fisher's exact test. Continuous variables were compared by the Wilcoxon rank sum test. The statistical analyses were done using the SAS software (v.8.2; SAS Institute, Inc., NC).
The study started patient enrollment on October 1, 2001. After the publication7 of a randomized study demonstrating the efficacy of lamivudine prophylaxis, the Data and Safety Monitoring Committee performed an interim analysis. The conditional power, calculated by the method of stochastic curtailment, of rejecting the null hypothesis was 0.95651, given the data already observed. Therefore, the Committee decided that early termination of patient recruitment was indicated. At the time of the early termination, a total of 52 patients were randomized (Fig. 1).
The baseline characteristics of the participants are summarized in Table 1. Patients in the therapeutic group appeared more likely to have positive HBeAg (8 versus 2, P = 0.04) and HBV DNA of more than 1,000,000 copies/mL (9 versus 3, P = 0.05). Other clinical characteristics were similar between the 2 groups. The patients received a median of 6 cycles of chemotherapy (range, 1-8 cycles).
All of the 26 patients in the prophylactic group and 17 patients in the therapeutic group received lamivudine treatment. The median duration of lamivudine treatment was 190 days (range, 85–385 days) for the prophylactic group and 139 days (range, 17–276 days) for the therapeutic group.
The primary endpoint of this study is HBV reactivation during chemotherapy and within 12 months of ending chemotherapy. The HBV reactivation rate was 30.8% [95% confidence interval (CI), 14.3%–51.8%] in the prophylactic group and 60.0% (95%CI, 38.7%–78.9%) in the therapeutic group (P = 0.05). Patients in the prophylactic group had a significantly longer time-to-HBV reactivation (hazard ratio, 0.35; 95%CI, 0.15–0.84; P = 0.018).
The relationship between protocol treatment, HBV reactivation, and hepatitis flare is summarized in Table 2. During protocol treatment, patients in the prophylactic group had a significantly lower risk of HBV reactivation [11.5% (95%CI, 2.5%–30.2%) versus 56% (95%CI, 34.9%–75.6%); P = 0.001]. In the prophylactic group, hepatitis flare occurred in 4 patients; 2 episodes were attributed to HBV reactivation. None of these patients developed severe hepatitis, defined as an increase of ALT to more than 10-fold ULN or bilirubin to more than 1.5-fold ULN. By contrast, in the therapeutic group, hepatitis flare occurred in 15 patients (60.0%; 95% CI, 38.7%–78.9%) and 12 of these episodes were attributed to HBV reactivation. The incidence of severe hepatitis was significantly higher in the therapeutic group. Logistic regression analysis was done to determine the predictive value of the following risk factors on HBV reactivation during chemotherapy: patient age, sex, baseline ALT level (normal versus elevated), HBeAg status (negative versus positive), baseline HBV DNA level (<1,000,000 copies/mL versus 1,000,000 copies/mL), lymphoma staging (by Ann Arbor staging and the International Prognostic Index [IPI]),18 and lamivudine treatment (prophylactic versus therapeutic). The only 2 significant predictors identified were prophylactic lamivudine use (P = 0.002) and high baseline HBV DNA level (P = 0.008). Multivariate analysis indicated that prophylactic lamivudine use was the only independent predictor of HBV reactivation (odds ratio, 0.04; 95% CI, 0.005–0.344; P = 0.003)
Table 2. Incidence of HBV Reactivation and Hepatitis
During protocol treatment
n = 26
n = 25
HBV reactivation and hepatitis flare
HBV reactivation and ALT10 × ULN
HBV reactivation and bilirubin > 1.5 × ULN
After protocol treatment
n = 26
n = 21
HBV reactivation and hepatitis flare
HBV reactivation and ALT>10 x ULN
HBV reactivation and bilirubin > 1.5 x ULN
The incidence of HBV reactivation and hepatitis after completion of protocol treatment did not differ significantly between the 2 groups. HBV reactivation was found in 8 patients (5 in the prophylactic group and 3 in the therapeutic group) (Figs. 2, 3). The median duration from the completion of protocol treatment to the occurrence of HBV reactivation for these 8 patients was 93.5 days (range, 25–182 days). The cumulative risk of HBV reactivation-associated hepatitis after the completion of protocol treatment was 15.4% (95%CI, 4.4%–34.9%) for the prophylactic group and 4.8% (95%CI 0.1%–23.8%) for the therapeutic group (P = 0.36). No risk factors, including patient age, sex, HBeAg status, baseline ALT or HBV DNA levels, and lymphoma staging, were found to be independently predictive of HBV reactivation after the completion of chemotherapy.
For the 5 patients in the prophylactic group who developed HBV reactivation after protocol treatment, 2 died of hepatitis flare at 173 and 182 days after completion of protocol treatment, respectively (Fig. 2A, B). A third patient (Fig. 2C) died of tumor progression. All of these 3 patients received lamivudine treatment at the time of diagnosis of hepatitis flare. The remaining 2 patients did not have hepatitis flares, and lamivudine treatment was not given. For the 3 patients in the therapeutic group who developed HBV reactivation after protocol treatment, none received lamivudine treatment. No hepatitis-related death occurred.
Two patients, both in the prophylactic group, developed YMDD mutation (1 YVDD [tyrosine-valine-aspartate-aspartate] and 1 YIDD [tyrosine-isoleucine-aspartate-aspartate]) after 173 days and 196 days of prophylactic lamivudine treatment, respectively (Fig. 4). The mutations were detected during episodes of hepatitis flares, which occurred after withdrawal from lamivudine prophylaxis. HBV DNA elevation was also detected, but the magnitude was less than 10-fold that of previous nadir levels. The hepatitis flares subsided spontaneously in both patients.
Comparison With the Historical Control Group.
The baseline clinical characteristics of patients in the current study and in the historical control group are compared in Table 3. More patients in the current study had advanced tumor status (IPI high or high-intermediate groups), whereas more patients in the historical control group had elevated HBV DNA levels at baseline. Other demographic features were similar between the 2 groups.
Table 3. Comparison of Baseline Clinical Characteristics Between the Present Study and the Historical Control Group
The incidence of HBV reactivation and hepatitis and the patterns of HBV reactivation are compared between the therapeutic group of the current study and the historical control group in Table 4. No significant difference was found between the 2 groups of patients, and the patterns of HBV reactivation did not appear to be changed by lamivudine treatment. Three patients in the historical group died of HBV reactivation–related hepatitis, whereas no patients in the therapeutic group of the current study had hepatitis-related death. However, the difference is not statistically significant.
Table 4. The Effects of Therapeutic Use of Lamivudine: Comparison with the Historical Group
Therapeutic Group of the Current Study (n = 25)
The Historical Control Group (n = 48)P
Incidence of HBV reactivation and hepatitis
HBV reactivation and hepatitis flare
HBV reactivation and ALT > 10 × ULN
HBV reactivation and bilirubin > 1.5 × ULN
Patterns of HBV reactivation
To further clarify the role of lamivudine use in the clinical course o HBV carriers who undergo chemotherapy for NHL, the incidence, severity, and patterns of HBV reactivation are compared between the whole group of the current study and the historical control group (Appendix Table 1). Again, no significant difference was detected.
Because ALT elevation that occurs in chemotherapy-induced HBV reactivation is usually preceded by HBV DNA elevation,3, 9 it is important to determine whether the timing of lamivudine treatment may be delayed significantly by using our criteria of ALT elevation as the starting point. We compared the timing of HBV DNA reactivation with that of ALT elevation for the 12 patients in the therapeutic group who developed HBV reactivation and hepatitis during protocol treatment. HBV DNA surge was preceded by ALT elevation for a median of 31.5 days (range, 4–84 days) in 4 of these patients and coincided with ALT elevation in another 2 patients. In the remaining 6 patients, HBV DNA reactivation preceded ALT elevation by a median of 23 days (range, 7–97 days). In 4 of these 6 patients, HBV DNA levels were still elevated when lamivudine treatment was started, whereas in the remaining 2 patients HBV DNA levels had decreased at the start of lamivudine treatment.
Response to Chemotherapy and Clinical Outcome.
The response rate to CHOP chemotherapy was 88.5% (95%CI, 69.9%–97.6%) (12 complete and 11 partial responses) in the prophylactic group and 84.0% (95%CI, 63.9%–95.5%) (12 complete and 9 partial responses) in the therapeutic group (P = 0.70). Reduction of chemotherapy doses was required, mostly because of myelosuppression, in 5 patients of the prophylactic group and 3 patients in the therapeutic group (P = 0.70). Nine patients in the prophylactic group and 14 in the therapeutic group had to delay chemotherapy because of serious adverse events (P = 0.16).
The median overall survival had not been reached in either group as of October 31, 2006, after a median follow-up duration of 33.2 months for the prophylactic group and 38.6 months for the therapeutic group. The 3-year overall survival rate was 69% (95%CI, 51%–87%) in the prophylactic group and 72% (95%CI, 54%–90%) in the therapeutic group (P = 0.98). Ten patients in the prophylactic group and 7 in the therapeutic group died during follow-up. The most common causes of death were tumor progression (15 patients) and hepatitis-induced liver failure (2 patients).
In this study we confirmed the efficacy of lamivudine in preventing chemotherapy-associated HBV reactivation in NHL patients during chemotherapy. By contrast, lamivudine therapy started at the time of ALT elevation did not appear to change the natural course of chemotherapy-associated HBV reactivation.
The current study is so far the largest prospective, randomized trial of lamivudine prophylaxis for chemotherapy-associated HBV reactivation in NHL patients. The patient population and the chemotherapy regimen used in this study are more homogenous than in previously reported studies.7–10 Lamivudine prophylaxis was found to be the only factor that independently predicted the risk of HBV reactivation during chemotherapy. The previous randomized trial by Lau et al.7 indicated that lamivudine prophylaxis completely prevented HBV reactivation during chemotherapy.7 However, our data indicated that HBV reactivation-associated hepatitis can still occur during chemotherapy for patients with lamivudine prophylaxis, but the incidence of HBV reactivation and the severity of hepatitis flare were both significantly reduced.
Clinical data necessary to establish the optimal duration of lamivudine prophylaxis are lacking. Most previous studies continued lamivudine prophylaxis for 1 to 3 months after completion of chemotherapy, but HBV reactivation after withdrawal from lamivudine prophylaxis was frequently found, especially with longer follow-up.3, 19 In 2004 the American Association for the Study of Liver Diseases recommended continuation of lamivudine prophylaxis for 6 months after the completion of chemotherapy, based on level III evidence (evidence based on clinical experience, descriptive studies, or reports of expert committees).20 In 2007, American Association for the Study of Liver Diseases updated its recommendation, suggesting that lamivudine prophylaxis for more than 6 months may be required for patients with high baseline HBV DNA.21 However, when the current study was designed, the optimal duration of lamivudine prophylaxis was unknown.22 In addition, our data indicate that HBV reactivation and life-threatening hepatitis may occur more than 6 months after completion of chemotherapy, irrespective of baseline HBV DNA levels. Therefore, a longer duration of antiviral prophylaxis should be considered for all patients. The incidence of YMDD mutation after lamivudine prophylaxis was 7.7% in the current study and may increase with more prolonged lamivudine therapy. The emergence of YMDD mutant usually resulted in clinical hepatitis; thus, agents with more favorable resistance profiles, such as adefovir or entecavir, may be better prophylactic agents.21
The fact that no hepatitis-related death occurred in the therapeutic group of this study while 3 of the 48 patients in the historical control group died of hepatitis suggests some effects of therapeutic use of lamivudine. The difference is not significant, probably because of the limited sample size of the 2 studies. Previous studies indicated that the efficacy of therapeutic use of lamivudine may depend on an early start of treatment.2, 23, 24 For chemotherapy-associated hepatitis, monitoring of HBV DNA levels is now the gold standard for diagnosis of HBV reactivation, but quantitative HBV DNA assays are not routinely available in clinical laboratories. Therefore, in this study we chose to use follow-up ALT levels as the guide for starting lamivudine treatment and set a more stringent cutoff point. Although most of the patients in the therapeutic group who had HBV reactivation can start lamivudine treatment, timely HBV DNA quantification may still be the best assay for monitoring in the near future.
Two patients in the prophylactic group died of HBV reactivation–related hepatitis after stopping of lamivudine prophylaxis. Both of them received lamivudine treatment at the diagnosis of hepatitis. However, because of less stringent follow-up of liver function tests (once every month, according to the study protocol), these patients may not be able to receive lamivudine treatment early enough to prevent liver damage. Besides, these patients might already have unrecognized liver diseases, rendering them more vulnerable to liver damage induced by HBV reactivation. Therefore, patients should be monitored weekly or biweekly by HBV DNA quantification so that antiviral therapies could be given in time.95
The authors thank Mei-Hsing Chuang and Yueh-Ling Ho of the National Health Research Institutes of Taiwan, ROC, for their help in data management and statistical analysis and Microbial Genomics Core at National Taiwan University for HBV study. The authors also thank the investigators and research nurses from all centers who participated in this study for their assistance with patient recruitment, clinical care, and follow-up.