- Email: [email protected]
Steroid-free chemotherapy decreases risk of hepatitis B virus (HBV) reactivation in HBV-carriers with lymphoma
Steroid-Free Chemotherapy Decreases Risk of Hepatitis B Virus (HBV) Reactivation in HBV-Carriers With Lymphoma Ann-Lii Cheng,1,2,3 Chao A. Hsiung,4 Ih-Jen Su,5 Pei-Jer Chen,1,6 Ming-Chih Chang,7 Chao-Jung Tsao,8 Woei-Yao Kao,9 Wu-Ching Uen,10 Chih-Hung Hsu,2 Hwei-Fan Tien,1 Tsu-Yi Chao,9 Li-Tzong Chen,3 and Jacqueline Whang-Peng,3 representing the Lymphoma Committee of Taiwan Cooperative Oncology Group (TCOG) Reactivation of hepatitis is one of the most serious complications of chemotherapy in lymphoma patients who are carriers of the hepatitis B virus (HBV). Glucocorticoids are linked to increased risk of HBV reactivation. This study seeks to clarify whether removal of glucocorticoids from chemotherapy regimens may decrease the risk of HBV reactivation. Eligible patients were seropositive for hepatitis B surface antigen (HBsAg) and had histologically proven non-Hodgkin’s lymphomas for which intensive chemotherapy was indicated. Patients were randomized to receive either ACE (epirubicin, cyclophosphamide, and etoposide) or PACE (prednisolone ⴙ ACE). A total of 50 patients were enrolled, 25 each for the ACE and PACE arms. The cumulative incidence of HBV reactivation at 9 months after starting chemotherapy was 38% and 73% for ACE and PACE arm, respectively (P ⴝ .03). The degree of clinical hepatitis was significantly more severe in the PACE arm: 11 patients (44%) in the PACE and 3 patients (13%) in the ACE arm had ALT elevation more than 10-fold of normal (P ⴝ .025), and 7 patients (28%) in the PACE and 1 patient (4%) in the ACE arm had icteric hepatitis (P ⴝ .049). Complete remission of tumors occurred in 11 (46%) patients in the PACE and 8 (35%) patients in the ACE arm (P ⴝ .556). The estimated overall survival rate at 46 months was 68% in the PACE arm and 36% in the ACE arm, respectively (P ⴝ .18). In conclusion, steroid-free chemotherapy decreases the incidence and severity of HBV reactivation in HBsAg-positive lymphoma patients. However, further research is needed to evaluate whether steroid-free chemotherapy may confer a less satisfactory control of lymphoma. (HEPATOLOGY 2003;37:1320-1328.)
R
eactivation of hepatitis B virus (HBV) replication is a frequent and well-recognized complication in patients with chronic HBV infection who receive cytotoxic or immunosuppressive therapy.1-16 The risk of
Abbreviations: HBV, hepatitis B virus; NHL, non-Hodgkin’s lymphoma; HBsAg, hepatitis B surface antigen; TCOG, Taiwan Cooperative Oncology Group; ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; PACE, prednisolone/adriamycin/cyclophosphamide/etoposide; ACE, adriamycin/cyclophosphamide/ etoposide; IV, intravenously; nt, nucleotide. From the Departments of 1Internal Medicine, 2Oncology, and 5Pathology, National Taiwan University Hospital; Divisions of 3Cancer Research and 4Biostatics and Bioinformatics, National Health Research Institutes; 6Clinical Research Institute, National Taiwan University College of Medicine; 7Department of Internal Medicine, Mackay Memorial Hospital, Taipei; 8Department of Internal Medicine, National Cheng Kung University Hospital, Tainan; 9Department of Internal Medicine, Tri-Service General Hospital, Taipei; 10Department of Internal Medicine, Shin Kong Memorial Hospital, Taipei, Taiwan. Received December 4, 2002; accepted March 10, 2003. Supported by the National Health Research Institute (NHRI), Taiwan (to R.O.C.). Address reprint requests to: Pei-Jer Chen, M.D., Director, Hepatitis Research Center, National Taiwan University Hospital, No. 7 Chung-Shan South Rd., Taipei, Taiwan. E-mail: [email protected] or [email protected]; fax: (886) 2-23317624. Copyright © 2003 by the American Association for the Study of Liver Diseases. 0270-9139/03/3706-0016$30.00/0 doi:10.1053/jhep.2003.50220 1320
this complication is particularly high in HBV endemic areas where the carrier rate in the general population may be as high as 15% to 20%. In patients with nonHodgkin’s lymphoma (NHL), viral reactivation can result in clinical hepatitis that often jeopardizes the schedule of chemotherapy and, at its extreme, leads to hepatic failure and death. Glucocorticoid, an ingredient of most chemotherapeutic protocols for NHL, is implicated as an important predisposing factor for HBV reactivation. In patients with chronic hepatitis B, long-term prednisolone treatment increases levels of hepatitis B surface antigen (HBsAg), hepatitis B core antigen, and HBV DNA in hepatocytes,17-21 as well as decreases T-cell function.22 Upon withdrawal of prednisolone, there is an intense rebound in cytotoxic T-cell function that coincides with a surge in serum aminotransferases and a decrease in the level of HBsAg and HBV DNA.22 Histologically, there is acute hepatocyte destruction with piecemeal necrosis, bridging, and confluent necrosis and multilobular collapse.23 Immunosuppression is thought to enhance viral replication with a subsequent spread in hepatocyte infection. In addition,
HEPATOLOGY, Vol. 37, No. 6, 2003
glucocorticoids may cause a direct reactivation of the latent HBV infection.24,25 The findings of a glucocorticoidresponsive element in HBV DNA and the demonstration that glucocorticoids specifically activate HBV gene expression in cultured human hepatocellular carcinoma cells lend support to the direct reactivation theory.24,25 Sporadic retrospective clinical studies also indicate that steroid-free chemotherapeutic regimens decrease the risk of hepatitis flares of HBV carriers.13,26 One possible approach to reduce or prevent reactivation of HBV DNA and consequent hepatitis is to omit steroids in chemotherapy regimens. Although steroids are ingredients of almost all first-line chemotherapeutic regimens for NHL, they do not appear to be indispensable. Steroid-free protocols such as COMLA, VIM, MIME, and infusional CDE regimen and an oral combination regimen may be as effective as steroid-containing protocols in the treatment of fresh and recurrent lymphoma.27-31 One small-series retrospective study showed that steroid-free chemotherapy may control lymphoma as well as steroid-containing regimens while significantly decreasing the risk of hepatitis flares in HBV carriers.13 This prospective study was conducted to determine whether a steroid-free chemotherapy regimen may decrease the risk of HBV reactivation and hepatitis development in HBsAg-positive lymphoma patients. It also compared the efficacy of steroid-free and steroid-containing regimens in lymphoma control.
Patients and Methods In November 1995, the Taiwan Cooperative Oncology Group (TCOG) initiated a multicenter randomized clinical trial to determine whether steroid-free chemotherapeutic regimens decrease the risk of HBV reactivation and hepatitis development in HBsAg-positive carriers. Other objectives included comparing the efficacy of steroid-free chemotherapeutic regimens in lymphoma control to steroid-containing regimens, and to examine the activity of HBV replication during the course of chemotherapy. The study was approved by the ethical committees of the participating hospitals. Patient Characteristics. Patients eligible for study participation were 15 to 72 years old, HBsAg-positive, and had histologically proven NHL for which intensive chemotherapy was indicated. Patients were excluded from the study if they were pregnant or breast-feeding, had cirrhosis of Child’s class B or C, impaired cardiac function based on New York Heart Association classification grade II or above, concurrent glucocorticoid use for other reasons, previous or concurrent chemotherapy or radiotherapy, or had a history of central nervous system
CHENG ET AL.
1321
involvement. An acceptable hematological and biochemical profile was mandatory prior to chemotherapy: this was defined as absolute granulocyte count ⱖ2,000/mm3, platelet ⱖ100,000/mm3, total bilirubin ⱕ2.5 mg/dL, alanine aminotransferase (ALT) ⱕ200 IU/I, serum creatinine ⱕ1.5 mg/dL, and blood urea nitrogen ⱕ25 mg/dL. All patients had objectively measurable disease when entering the trial. Written informed consent was obtained from all participants. The following tests were performed in all patients: hepatitis B serology (HBsAg, antibody to HBsAg, hepatitis B e antigen [HBeAg], and antibody to HBe), hepatitis C and D (anti-HCV and anti-delta) serology, serum biochemistry (bilirubin, ALT, prothrombin time/partial thromboplastin time, lactate dehydrogenase, alkaline phosphatase), full blood count with leukocyte differentials, blood urea nitrogen and creatinine, electrolytes (sodium, potassium, chloride, and calcium), and serum immunoglobulin (IgG, IgA, and IgM). Bone marrow biopsy, chest X-ray, electrocardiogram, left ventricular ejection fraction, abdominal computed tomography, tumor measurement, and urinalysis were performed and patients’ performance status was determined prior to chemotherapy. Stratified randomization was done by serum ALT level (normal vs. abnormal) and the tumor stage (I/II vs. III/IV). A schematic of the study is shown in Fig. 1. Chemotherapy Regimens. PACE (prednisolone 60 mg/m2/d orally, day 1-7; epirubicin 60 mg/m2 intravenously [IV], day 1; cyclophosphamide 650 mg/m2 IV, day 1; etoposide 55 mg/m2/d IV, day 1-3) was given as first-line therapy for the steroid-containing group and ACE (epirubicin 60 mg/m2 IV, day 1; cyclophosphamide 650 mg/m2 IV, day 1; etoposide 55 mg/m2/d IV, day 1-3) for the steroid-free group. For patients who failed the first-line therapy, respective steroid-containing and steroid-free second-line salvage chemotherapy were used. VIMP (VP-16 100 mg/m2/d IV, day 1, 3, 5; ifosfamide 1 g/m2/d IV, day 1-5; methotrexate 30 mg/m2/d IV, day 1, 5; prednisolone 60 mg/m2/d orally, day 1-7) was given for the steroid-containing group. VIM (VP-16 100 mg/m2/d IV, day 1, 3, 5; ifosfamide 1 g/m2/d IV, day 1-5; methotrexate 30 mg/m2/d IV, day 1,5) was given for the steroidfree group. The decision to use these regimens was made by consensus vote of the members of the TCOG lymphoma committee. No crossing-over between steroidcontaining and steroid-free treatment was allowed. Hepatitis B Virus Serology. Serum samples were collected before and around the nadir of each cycle of chemotherapy and for at least 4 months after the end of the protocol treatment. Analysis of the serum samples was centralized at the Hepatitis Research Center of National
1322
CHENG ET AL.
Fig. 1. Schema of the clinical trial. PACE, prednisolone/adriamycin/ cyclophosphamide/etoposide; ACE, adriamycin/cyclophosphamide/etoposide; CR, complete remission; PR, partial remission; SD, stable disease; PD, progressive disease; VIMP, VP-16/ifosfamide/methotrexate/prednisolone; VIM, VP-16/ifosfamide/methotrexate.
Taiwan University College of Medicine. Testing was performed for liver function, HBsAg, and HBeAg. HBV DNA was tested in duplicates using Chiron bDNA assay (VERSANT HBV DNA assay, Chiron diagnostics, Emeryville, CA). The sensitivity of the assay was 2.5 pg/mL. In cases with inconsistent results, polymerase chain reaction for HBV DNA detection was performed. Briefly, serum DNA was extracted using a commercially available kit (QIAamp DNA Blood Mini Kit, QIAGEN Inc, Valencia, CA). The presence of serum HBV DNA was detected by single-round or nested polymerase chain reaction using primer sets from surface region. The outer primers were S-1s (nucleotides [nt] 159-178: 5⬘AGAACATCGCATCAGGACTC-3⬘) and S-2a (nt 642623: 5⬘-CATAGGTATCTTGCGAAAGC-3⬘), and the inner primers were S-3s (nt 181-200: 5⬘-AGGACCCCTGFCGTGTTAC-3⬘) and S-4a (nt 619-600: 5⬘-AGATGATGGGATGGGAATAC-3⬘). Clinical hepatitis flare-up was defined as a 3-fold or greater increase in serum ALT level that exceeded 100 IU/L. The hepatitis or hepatitis flare-up was attributed to reactivation of chronic hepatitis B when there was a 10fold or higher elevation of serum HBV DNA compared with the previous baseline level or reappearance of HBV
HEPATOLOGY, June 2003
DNA or HBeAg in the serum. For cases of clinical hepatitis, serum samples were collected at 2-week intervals regardless of the schedule of chemotherapy. Regular schedule was resumed if hepatitis subsided and chemotherapy was restarted. Evaluation of Treatment-Related Toxicities, Tumor Response, and Survival. Hemogram was assessed weekly and blood chemistry was assessed biweekly during systemic chemotherapy. Imaging studies were performed monthly until tumor response could be documented. Subsequently, if no intercurrent events required earlier examinations, imaging studies were performed every 3 months for 1 year and every 6 months thereafter. The time to HBV DNA reactivation was the time from entry into the study until the first event of HBV DNA reactivation. Patients without evidence of HBV DNA reactivation were censored from the analysis at the time of most recent follow-up of HBV DNA. Complete remission was defined as the disappearance of all clinically detectable tumors for at least 4 weeks. Partial remission was defined as at least a ⱖ50% reduction in the sum of the products of all measurable tumors, without the appearance of any new lesions for at least 4 weeks. Stable disease was defined as a reduction of less than 50% or an increase of less than 25% of all measurable tumors with no appearance of new lesions for at least 4 weeks. Progression-free survival duration was defined as the interval from entry into the study to either death due to any cause or to disease progression (whichever occurred first). Patients alive without evidence of disease were censored at the time of most recent follow-up. Survival time was the interval from study entry until death, and surviving patients were censored at the time of their most recent follow-up. Statistical Design and Analysis. Patients were randomized to receive either steroid-containing or steroid-free chemotherapy. Stratification was performed according to status of transaminase (normal vs. abnormal) and tumor stage (I/II vs. III/IV). Abnormal transamimase level was defined as serum levels of ALT or aspartate aminotransferase being higher than their upper normal values at the prestudy evaluation. The primary end point was the incidence of HBV reactivation and associated hepatitis flare-up during treatment. The primary objective was to compare the incidence of HBV reactivation and associated hepatitis of the steroid-containing regimen with that of the steroid-free regimen. The objective hypothesis of the trial was to detect a difference of 35% in the incidence of HBV reactivation and associated hepatitis between the 2 arms from 45% to 10%. With a 2-sided binomial test at
HEPATOLOGY, Vol. 37, No. 6, 2003
the 5% significance level and 90% power, this objective required 76 evaluable patients, with 38 patients in each arm. Because of the relatively small number of patients, stratification was based only on 2 of the prognostic factors that were considered most relevant to this study. The planned accrual period was 4 years. All analyses were carried out according to the intentto-treat principle, and statistical significance was set at the .05 level. Time-to-event comparisons were performed using the log-rank test. Survival curves were estimated by the Kaplan-Meier technique. Comparison of 2 proportions was performed using Fisher’s exact test (2-tail) and comparison of continuous variables was performed using the t-test. The statistical software used for the analyses in this study was SAS V8.
Results For ethical reasons, the interim data were monitored by the Data and Safety Monitoring Committee (DSMC) of TCOG every 6 months. The trial was terminated early by the DSMC in February 2000 because of the high hepatitis flare-up rate of the PACE arm. The availability of potentially effective antiviral agents for prophylaxis of HBV reactivation in early 2000 was another major ethical concern for early termination. From November 1995 through February 2000, a total of 50 patients were enrolled, 25 patients in each arm. Forty-nine patients were evaluable for HBV reactivation, and 47 patients were evaluable for tumor response. One patient was excluded from any analysis because of wrong pathologic diagnosis, another 2 patients were excluded from evaluation of tumor response due to refusal of further treatment after 2 and 5 courses of scheduled chemotherapy. Both of the latter 2 patients complained of general malaise after chemotherapy although no grade II or higher toxicity was documented. The median duration of follow-up was 919 and 1,079 days for the ACE and PACE arm, respectively. The clinicopathologic features of the patients were well balanced between the 2 arms (Table 1). Four patients of the PACE arm and 3 patients of the ACE arm had received second-line steroid-containing and steroid-free chemotherapy, respectively. HBV Reactivation and Development of Clinical Hepatitis. A total of 27 patients, 18 of 25 (72%) from the PACE arm and 9 of 24 (37.5%) from the ACE arm (P ⫽ .022), developed HBV reactivation. The relative risk of PACE versus ACE was 1.9 (1.08-3.40, 95% confidence interval) (Table 2). The results were similar if the analyses were adjusted for stratification factors: status of transaminase (normal vs. abnormal) and tumor stage (I/II vs. III/ IV) (Table 3). As shown in Fig. 2, after starting
CHENG ET AL.
1323
Table 1. Clinicopathologic Characteristics of the Patients No. of eligible patients Sex Male (%) Female (%) Age (y) Mean Median Range Pathology Ibl/DL (%) Others (%)* Clinical stage III/IV (%) Elevated LDH (%) B symptoms (%) Bulky tumor, ⬎8 cm (%) Courses of chemotherapy Mean Median Range HBeAg Positive (%) Negative (%) HBV DNA Negative (%) Positive (%) HBV DNA (pg/mL) Mean Range
PACE
ACE
P Value
25
24
13 (52) 12 (48)
16 (67) 8 (33)
41.4 44.4 16.9-68.6
47.6 47.9 23.8-70.1
.142
19 (76) 6 (24) 15 (60) 11 (44) 9 (36) 5 (20)
19 (79) 5 (21) 13 (54) 13 (54) 6 (25) 8 (33)
1.000
5.8 6 2-10
5.7 6 3-8
.855
6 (24) 19 (76)
3 (13) 21 (87)
9 (36) 16 (64)
15 (63) 9 (37)
716.7 ⬍2.5-5,800
360.0 ⬍2.5-2,400
.387
.776 .572 .538 .345
.4635
.0887
.3314
Abbreviation: Ibl/DL, immunoblastic/diffuse large cell. *PACE arm: diffuse polymorphic T cell (1), diffuse mixed cell (2), follicular large cell (1), follicular mixed cell (1), and diffuse small cleaved cell (1); ACE arm: anaplastic large cell (2), diffuse mixed cell (2), and diffuse small cleaved cell (1).
chemotherapy, the likelihood of HBV reactivation was consistently higher for patients in the PACE arm (P ⫽ .03). Nine months after starting chemotherapy, the cumulative incidence of HBV reactivation was 73% for the PACE arm and 38% for the ACE arm. We also compared the cumulative incidence of HBV reactivation between the two groups. The hazard ratio of PACE versus ACE was 2.36 (1.05-5.29, 95% confidence interval, P ⫽ .037) with unadjusted analysis and was 2.62 (1.14-6.00, 95% confidence interval, P ⫽ .023) with Cox regression analysis adjusting for the stratification factors. Of the 27 patients with HBV reactivation, 23 (85.2%) developed clinical hepatitis flare-up as evidenced by a significant elevation of serum ALT closely associated or following the increment of HBV DNA. The degree of clinical hepatitis flare-up was significantly more severe among patients in the PACE arm (Table 2). The results were similar if the analyses were adjusted for stratification factors (Table 3). Lethal fulminant hepatitis developed in 1 patient in the PACE arm but in no patients in the ACE arm. Surprisingly, all clinically significant elevations of ALT during the period of chemotherapy were preceded
1324
CHENG ET AL.
HEPATOLOGY, June 2003
Table 2. Incidence of HBV Reactivation and Severe Clinical Hepatitis P Value
No. of eligible patients HBV reactivation HBV reactivation and hepatitis flare-up HBV reactivation and ALT ⬎10 ⫻ N HBV reactivation and bilirubin ⬎1.5 ⫻ N
PACE
ACE
25 18 (72%) 15 (60%) 11 (44%) 7 (28%)
24 9 (37.5%) 8 (33%) 3 (13%) 1 (4%)
RR (95% CI)
P Value
(Unadjusted*)
.022 .088 .025 .049
1.9 (1.08, 3.40) 1.8 (0.94, 3.45) 3.5 (1.12, 11.08) 6.7 (0.89, 50.61)
RR (95% CI) (Adjusted†)
.021 .069 .020 .025
1.9 (1.06, 3.42) 1.8 (0.93, 3.51) 3.5 (1.07, 11.44) 6.4 (0.84, 49.25)
Abbreviations: ALT, alanine transaminase; N, normal; RR, Relative Risk (PACE vs. ACE); CI, confidence interval. *Using 2 test. †Using Cochran-Mantel-Haenszel test by adjusting stratifying factors status of transaminase and tumor stage.
by an increase in serum HBV DNA. No other causes of clinical hepatitis could be detected in this group of patients during this period. The magnitude of HBV reactivation, as reflected by the increase in serum HBV DNA, was similar between patients in the PACE and ACE arm (Table 3). As shown in Fig. 3, we characterized the patterns of HBV reactivation during systemic chemotherapy of HBsAg-positive lymphoma patients. The most common HBV reactivation, which occurred in 12 of the 27 (44.4%) reactivators, was characterized by a rapid and transient surge of HBV DNA that resolved within 4 to 5 weeks after ALT elevation (Fig. 3A). Four patients developed multiple HBV reactivation, characterized by repeated HBV increment after apparent resolution from previous episodes (Fig. 3B). Another 10 patients developed protracted HBV reactivation, characterized by a prolonged period of persistently high titer of HBV DNA. Four of these 10 patients finally achieved resolution of HBV reactivation after several bouts of clinical hepatitis (Fig. 3C), while the other 6 patients stayed in the status of high-titer of HBV DNA throughout the entire period of chemotherapy (Fig. 3D). These patterns of HBV reactivation showed no preference of occurrence in either the PACE or ACE arm. The timing of HBV reactivation in relation to the start of chemotherapy was interesting. As illustrated in Fig. 3B-D, reactivation of HBV DNA occurred within 2 weeks of the first dose of chemotherapy in 13 of the 27 (48.1 %) cases of reactivation. However, this early reactivation occurred with similar frequency among patients in
both the PACE (9 of 18, 50%) and ACE arm (4 of 9, 44%), which suggests that direct promoting of HBV replication by glucocorticoids was not the major mechanism for this phenomenon.24,25 All except one patient in the PACE arm developed HBV reactivation during the use of first-line chemotherapy. Tumor Response and Survival. Tumor response to first-line chemotherapy was evaluable in 24 and 23 patients of the PACE and ACE arm, respectively. Complete remission was found in 11 patients (46%) of the PACE arm and 8 patients (35%) of the ACE arm (P ⫽ .556, Fisher’s exact test). As shown in Fig. 4A, the estimated overall survival at 46 months was 68% for patients in the PACE arm and 36% for patients in the ACE arm (P ⫽ .18, log-rank test). The cause of death was attributed to uncontrolled lymphoma in 4 and 9 patients of the PACE and ACE arm, respectively. No significant difference was found in progression-free survival between patients in the PACE and ACE arm (P ⫽ .85) (Fig. 4B). Other Serologic Studies. At the last serologic sampling, all 49 patients remained HBsAg-positive; while 1 HBeAg-positive patient had become HBeAg-negative
Table 3. Magnitude of HBV Reactivation of the Patients No. of Patients Increment of HBV DNA 10⬃⬍102 ⫻ N 102⬃⬍103 ⫻ N 103⬃⬍104 ⫻ N
PACE (%)
ACE (%)
18
9
5 (28) 9 (50) 4 (22)
4 (45) 3 (33) 2 (22)
P Value
.776
Fig. 2. Development of HBV DNA reactivation during systemic chemotherapy.
HEPATOLOGY, Vol. 37, No. 6, 2003
CHENG ET AL.
1325
Fig. 3. Classification of HBV reactivation according to duration and frequency of HBV DNA increment. Four representative cases are shown. The units of the 3 parameters at 1X are ALT, 40 U/L; Bil, 1.0 mg/dL; and HBV DNA, 2.5 pg/mL.
and 2 HBeAg-negative patients had become HBeAg-positive. Patients with prechemotherapy HBeAg-positive status appeared to have a lower risk (2 of 9, 22%) of HBV reactivation than those with HBeAg-negative status (25 of 40, 63%) (P ⫽ .06, Fisher’s exact test). HBV reactivation was not observed in any of the 9 patients with prechemotherapy HBV DNA titer more than 1,000 pg/mL.
Discussion In this prospective study, we demonstrated that steroid-free chemotherapy significantly decreases the incidence and severity of HBV reactivation in HBsAgpositive lymphoma patients. We also found that all ALT elevations of clinical significance during systemic chemo-
therapy in our patients was due to HBV reactivation. Other possible causes of hepatic damage, such as lymphoma infiltration in liver, anti-cancer drug–induced hepatotoxicity, or infection by other hepatotropic viruses, did not occur in this series. The reported frequency of HBV reactivation in HBsAg-positive lymphoma patients undergoing cancer chemotherapy has ranged from 14% to 72%.8,10 This wide variation in the reported reactivation rate is partly due to incomplete observation of the HBV DNA in many of the studies. As shown in our study, HBV DNA titer usually decreased rapidly soon after ALT elevation, and ALT elevation tended to be transient and minor in many patients. Even a minor transient elevation of ALT, which is
Fig. 4. (A) Overall survival and (B) progression-free survival of the patients.
1326
CHENG ET AL.
often overlooked clinically, was frequently preceded by more than a 1,000-fold increment of HBV DNA. These features have led to the underestimation of the true incidence of HBV reactivation in retrospective studies that rely on detecting ALT elevation as the first clue of HBV reactivation. Several factors are linked to the incidence and severity of hepatitis flare-up in HBsAg-positive cancer patients undergoing chemotherapy, including the proliferation status of the virus in hepatocytes, the intensity of the chemotherapeutic regimens, and the use of glucocorticoids.16 In this study, we confirm that hepatitis flares occur more frequently in patients who are in the nonreplicative phase of infection (i.e., HBeAg negative and serum HBV DNA negative).14,16,32 During these episodes of hepatitis flares, serum aminotransferase levels increase in response to the sudden reemergence of viral replication. However, none of our patients with chemotherapy-induced hepatitis flares seroconverted to HBsAgnegative status, suggesting that most of these episodes were abortive attempts of cytotoxic T cells to wipe out the reproliferating viruses. Results of several retrospective studies suggest that the severity of hepatitis flare is roughly correlated with the intensity of chemotherapy.5,12,14,15 The PACE and ACE regimens used in our study had an intensity similar to or slightly higher than that of the first-generation regimens such as CHOP (cyclophosphamide, adriamycin, vincristine, and prednisolone). However, the complete response rates of patients in the PACE (46%) and ACE (35%) arms are slightly lower than those usually expected for CHOP. This may be partly due to the relatively stringent criteria of complete remission adopted by this study, in which the existence of residual tiny lesions of unknown significance was considered partial remission in most patients. The treatment intensity of PACE, with an additional drug, prednisolone, may be higher than that of ACE. However, the dramatically increased incidence and severity of hepatitis flares among patients in the PACE arm cannot be fully explained by the slight increase of treatment intensity of PACE. One of the earliest observations suggesting a relationship between glucocorticoids and HBV replication was that long-term treatment resulted in increased expression of hepatitis B core antigen in the serum.33 It was subsequently discovered that HBV DNA polymerase activity increased during a short course of glucocorticoid therapy.34 In vitro studies also indicate that glucocorticoids may activate HBV replication via a steroid-responsive element located at the promoter of HBV DNA.24,25 Results of sporadic clinical studies suggest that removal of steroids
HEPATOLOGY, June 2003
from chemotherapy of hematologic malignancies significantly decreases the risk of hepatitis flares in HBV carriers.13,26 The current study confirms that the use of glucocorticoids in HBsAg-positive lymphoma patients is one of the predominant risk factors for HBV reactivation. One key issue that remains unanswered is whether steroid-free chemotherapy is less effective in the control of lymphoma. Results of several single-arm studies suggest that steroid-free chemotherapeutic regimens control lymphoma as well as steroid-containing regimens.27-31 In the current study, although the steroid-free arm showed a trend toward less satisfactory overall survival and more patients in the ACE arm died of uncontrolled lymphoma, these differences were not significant. However, further study is needed to determine if the benefit of decreased hepatitis flares with use of steroid-free chemotherapy is offset by less satisfactory lymphoma control. In this prospective study, we described the various patterns of HBV reactivation during systemic chemotherapy of HBsAg-positive lymphoma patients. We were surprised to notice that nearly half of the HBV reactivation developed within 2 weeks of the first dose of systemic chemotherapy. At least some of these “early reactivators” developed HBV reactivation before the development of neutropenia, suggesting that mechanisms of HBV reactivation other than immune suppression are involved. One possible mechanism is direct activation of HBV replication by the anti-cancer drugs. Although a steroid-responsive element was found in the promoter region of the HBV genome,24 our results suggest that steroids are not the activator since both ACE and PACE arms had the same proportion of early reactivators. The possibility that cytotoxic agents other than steroids may activate the replication of HBV genome should also be considered. This study identified several clinically relevant patterns of HBV reactivation during systemic chemotherapy. Ten of our patients developed protracted HBV reactivation characterized by a prolonged high titer of HBV DNA after ALT elevation, which was often associated with repeated episodes of clinical hepatitis. One of these 10 patients died of fulminant hepatitis. Four of our patients had multiple episodes of reactivation (i.e., repeated HBV reactivation after apparent resolution of the previous episode). Multiple episodes of reactivation and remission have been shown to accelerate the progression of chronic hepatitis.35,36 Further, repeated attacks of clinical hepatitis often jeopardize the scheduled chemotherapy. Longer follow-up is needed to clarify the long-term effect on the liver by the various patterns of HBV reactivation. Although antiviral agents are now available for the prevention and treatment of HBV reactivation,16,37-39 pro-
HEPATOLOGY, Vol. 37, No. 6, 2003
spective randomized studies are needed to confirm efficacy. In this study, we confirmed that steroid-free chemotherapy significantly decreases the risk of hepatitis flare-up in HBsAg-positive lymphoma patients. However, whether steroid-free chemotherapy may jeopardize the efficacy of lymphoma control needs to be evaluated in future research to determine the efficacy of combining potentially effective antiviral agents with either steroidcontaining or steroid-free chemotherapy in the treatment of HBsAg-positive lymphoma patients. An ongoing prospective randomized study of TCOG is currently studying the efficacy of lamivudine in either the prevention or treatment of HBV reactivation in HBV carriers receiving steroid-containing chemotherapy. Preliminary data of this study suggest a good efficacy of lamivudine in both arms. Before the final reports of all these clinical studies are available, it may be prudent to advise the use of standard steroid-containing regimens for HBV-carrier lymphoma patients and treat HBV reactivation with effective antiviral agents as early as possible.
References 1. Galbraith RM, Eddleston AL, Williams R, Zuckerman AJ. Fulminant hepatic failure in leukaemia and choriocarcinoma related to withdrawal of cytotoxic drug therapy. Lancet 1975;2:528-530. 2. Wands JR, Chura CM, Roll FJ, Maddrey WC. Serial studies of hepatitisassociated antigen and antibody in patients receiving antitumor chemotherapy for myeloproliferative and lymphoproliferative disorders. Gastroenterology 1975;68:105-112. 3. Hoofnagle JH, Dusheiko GM, Schafer DF, Jones EA, Micetich KC, Young RC, Costa J. Reactivation of chronic hepatitis B virus infection by cancer chemotherapy. Ann Intern Med 1982;96:447-449. 4. Thung SN, Gerber MA, Klion F, Gilbert H. Massive hepatic necrosis after chemotherapy withdrawal in a hepatitis B virus carrier. Arch Intern Med 1985;145:1313-1314. 5. Lau JYN, Lai CL, Lin HJ, Lok ASF, Liang RHS, Wu PC, Chan TK, Todd D. Fatal reactivation of chronic hepatitis B virus infection following withdrawal of chemotherapy in lymphoma patients. Q J Med 1989;73:911917. 6. Pariente EA, Goudeau A, Dubois F, Degott C, Gluckman E, Devergie A, Brechot C, et al. Fulminant hepatitis due to reactivation of chronic hepatitis B virus infection after allogeneic bone marrow transplantation. Dig Dis Sci 1988;33:1185-1191. 7. Liang RHS, Lok ASF, Lai CL, Chan TK, Tood D, Chiu EKW. Hepatitis B infection in patients with lymphomas. Hematol Oncol 1990;8:261-270. 8. Lok ASF, Liang RHS, Chiu EKW, Wong KL, Chan TK, Todd D. Reactivation of hepatitis B virus replication in patients receiving cytotoxic therapy—report of a prospective study. Gastroenterology 1991;100:182-188. 9. Liaw YF. Hepatitis viruses under immunosuppressive agents. J Gastroenterol Hepatol 1998;13:14-20. 10. Alexopoulos CG, Vaslamatzis M, Hatzidimitriou G. Prevalence of hepatitis B virus marker positivity and evolution of hepatitis B virus profile, during chemotherapy, in patients with solid tumors. Br J Cancer 1999;81: 69-74. 11. Markovic S, Drozina G, Vovk M, Fidler-Jenko M. Reactivation of hepatitis B but not hepatitis C in patients with malignant lymphoma and immunosuppressive therapy—a prospective study in 305 patients. Hepatogastroenterology 1999;46:2925-2930.
CHENG ET AL.
1327
12. Liang R, Lau GKK, Kwong YL. Chemotherapy and bone marrow transplantation for cancer patients who are also chronic hepatitis B carriers: a review of the problem. J Clin Oncol 1999;17:394-398. 13. Cheng AL. Steroid-free chemotherapy decreases the risk of hepatitis flare-up in hepatitis B virus carriers with non-Hodgkin’s lymphoma. Blood 1996;87:1202. 14. Kumagai K, Takagi T, Nakamura S, Sawada U, Kura Y, Kodama F, Shimano S, et al. Hepatitis B virus carriers in the treatment of malignant lymphoma: an epidemiological study in Japan. Ann Oncol 1997:8(Suppl 1):107-109. 15. Faggioli P, Paschale M, Tocci A, Luoni M, Fava S, De Paoli A, Tosi A, Cassi E. Acute hepatic toxicity during cyclic chemotherapy in nonHodgkin’s lymphoma. Haematologica 1997;82:38-42. 16. Perrillo RP. Acute flares in chronic hepatitis B: the natural and unnatural history of an immunologically mediated liver disease. Gastroenterology 2001;120:1009-1022. 17. Sagnelli E, Manzillo G, Mario G, Pasquale G, Felaco FM, Filippini P, Izzo CM, et al. Serum levels of hepatitis B surface and core antigens during immunosuppressive treatment of HBsAg-positive chronic active hepatitis. Lancet 1980;2:395-397. 18. Scullard GH, Smith CI, Merigan TC, Roinson WS, Gregory PB. Effects of immunosuppressive therapy on viral markers in chronic active hepatitis B. Gastroenterology 1981;81:987-991. 19. Lam KC, Ching LL, Ng RP, Trepo C, Wu PC. Deleterious effect of predinosolone in HBsAg-Positive chronic active hepatitis. N Engl J Med 1981;304:380-386. 20. Hoofnagle JH, Davis GL, Pappas SC, Hanson RG, Peters M, Avigan MI, Waggoner JG, et al. A short course of prednisolone in chronic type B hepatitis. Ann Intern Med 1986;104:12-17. 21. Lee SD, Tong MJ, Wu JC, Lin HC, Tsai YT, Lo KJ. A randomized double-blind placebo-controlled trial of prednisolone therapy in HBeAg and HBV DNA positive Chinese patients with chronic active hepatitis B. J Hepatol 1991;12:146-150. 22. Hanson RG, Peters MG, Hoofnagle JH. Effects of immunosuppressive therapy with prednisolone on B and T lymphocyte function in patients with chronic type B hepatitis. HEPATOLOGY 1986;6:173-179. 23. Czaja AJ, Wolf AM, Baggenstoss AH. Laboratory assessment of severe chronic active liver disease during and after corticosteroid therapy: correlation of serum transaminase and gamma globulin levels with histologic features. Gastroenterology 1981;80:687-692. 24. Tur-Kaspa R, Burk RD, Shaul Y, Shafrits DA. Hepatitis B virus DNA contains a glucocorticoid responsive element. Proc Natl Acad Sci U S A 1986;83:1627-1631. 25. Chou CK, Wang LH, Lin HM, Chi CW. Glucocorticoid stimulates hepatitis B viral gene expression in cultured human hepatoma cells. HEPATOLOGY 1992;16:13-18. 26. Ohtsu T, Sai T, Oka M, Sugai Y, Tobinai K. Activation of hepatitis B virus infection by chemotherapy containing glucocorticoid in hepatitis B virus carriers with hematologic malignancies. Jpn J Clin Oncol 1991;21:360365. 27. Sparano JA, Wiernik PH, Leaf A, Dutcher JP. Infusional cyclophosphamide, doxorubicin, and etoposide in relapsed and resistant non-Hodgkin’s lymphoma: evidence for a schedule-dependent effect favoring infusional administration. J Clin Oncol 1993:11:1071-1079. 28. Cabanillas F. Non-Hodgkin’s lymphomas: a review of the M.D. Anderson experience. Semin Oncol 1993;19:11-13. 29. Herbrecht R, Damonte JC, Dufoure P, Maloisel F, Liuk L, Ortiz S, Bergerat JP, et al. Etoposide, ifosfamide and methotrexate combination chemotherapy for aggressive non-Hodgkin’s lymphoma after failure of the NLH 84 regimen. Semin Oncol 1992;197:7-10. 30. Berd D, Cornog J, Decont RC, Levitt M, Bertino JR. Long-term remission in diffuse histiocytic lymphoma treated with combination sequential chemotherapy. Cancer 1975;35:1050. 31. Remick SC, McSharry JJ, Wolf BC, Blanchard CG, Eastman AY, Wagher H, Portuese E, et al. Novel oral combination chemotherapy in the treat-
1328
32.
33.
34.
35.
CHENG ET AL.
ment of intermediate-grade and high grade AIDS-related non-Hodgkin’s lymphoma. J Clin Oncol 1993;11:1691-1702. Davis GL, Hoofnagle JH, Waggoner JG. Spontaneous reactivation of chronic hepatitis B virus infection. Gastroenterology 1984;86:230235. Sagnelli E, Manzio G, Maio G, Pasquale G, Felaco FM, Filippini P, Izzo CM, et al. Serum levels of hepatitis B surface and core antigens during immunosuppressive treatment of HBsAg-positive chronic active hepatitis. Lancet 1980;2:395-397. Nair PV, Tong MJ, Stevenson D, Roskamp D, Boone C. Effects of shortterm, high-dose prednisone treatment of patients with HBsAg-positive chronic active hepatitis. Liver 1985;5:8-12. Perrillo RP, Campbell CR, Sanders GE, Regenstein FG, Bodicky CJ. Spontaneous clearance and reactivation of chronic hepatitis B virus
HEPATOLOGY, June 2003
36. 37.
38.
39.
infection among male homosexuals with chronic type B hepatitis. Ann Intern Med 1984;100:43-46. Hoofnagle JH, Seeff LB. Natural history of chronic type B hepatitis. Prog Liver Dis 1982;7:469-479. Rossi G, Pelizzari A, Motta M, Puoti M. Primary prophylaxis with lamivudine of hepatitis B virus reactivation in chronic HbsAg carriers with lymphoid malignancies treated with chemotherapy. Br J Haematol. 2001;115:58-62. Liao CA, Lee CM, Wu HC, Wang MC, Lu SN, Eng HL. Lamivudine for the treatment of hepatitis B virus reactivation following chemotherapy for non-Hodgkin’s lymphoma. Br J Haematol 2002;116:166-169. Marcello P, Fedele DM, Giovanni DGR, Alessandro S, Bruno P, Marco P, Aristide M, et al. Efficacy of lamivudine to prevent hepatitis reactivation in hepatitis B virus-infected patients treated for non-Hodgkin lymphoma. Blood 2002;99:724-725.
R
eactivation of hepatitis B virus (HBV) replication is a frequent and well-recognized complication in patients with chronic HBV infection who receive cytotoxic or immunosuppressive therapy.1-16 The risk of
Abbreviations: HBV, hepatitis B virus; NHL, non-Hodgkin’s lymphoma; HBsAg, hepatitis B surface antigen; TCOG, Taiwan Cooperative Oncology Group; ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; PACE, prednisolone/adriamycin/cyclophosphamide/etoposide; ACE, adriamycin/cyclophosphamide/ etoposide; IV, intravenously; nt, nucleotide. From the Departments of 1Internal Medicine, 2Oncology, and 5Pathology, National Taiwan University Hospital; Divisions of 3Cancer Research and 4Biostatics and Bioinformatics, National Health Research Institutes; 6Clinical Research Institute, National Taiwan University College of Medicine; 7Department of Internal Medicine, Mackay Memorial Hospital, Taipei; 8Department of Internal Medicine, National Cheng Kung University Hospital, Tainan; 9Department of Internal Medicine, Tri-Service General Hospital, Taipei; 10Department of Internal Medicine, Shin Kong Memorial Hospital, Taipei, Taiwan. Received December 4, 2002; accepted March 10, 2003. Supported by the National Health Research Institute (NHRI), Taiwan (to R.O.C.). Address reprint requests to: Pei-Jer Chen, M.D., Director, Hepatitis Research Center, National Taiwan University Hospital, No. 7 Chung-Shan South Rd., Taipei, Taiwan. E-mail: [email protected] or [email protected]; fax: (886) 2-23317624. Copyright © 2003 by the American Association for the Study of Liver Diseases. 0270-9139/03/3706-0016$30.00/0 doi:10.1053/jhep.2003.50220 1320
this complication is particularly high in HBV endemic areas where the carrier rate in the general population may be as high as 15% to 20%. In patients with nonHodgkin’s lymphoma (NHL), viral reactivation can result in clinical hepatitis that often jeopardizes the schedule of chemotherapy and, at its extreme, leads to hepatic failure and death. Glucocorticoid, an ingredient of most chemotherapeutic protocols for NHL, is implicated as an important predisposing factor for HBV reactivation. In patients with chronic hepatitis B, long-term prednisolone treatment increases levels of hepatitis B surface antigen (HBsAg), hepatitis B core antigen, and HBV DNA in hepatocytes,17-21 as well as decreases T-cell function.22 Upon withdrawal of prednisolone, there is an intense rebound in cytotoxic T-cell function that coincides with a surge in serum aminotransferases and a decrease in the level of HBsAg and HBV DNA.22 Histologically, there is acute hepatocyte destruction with piecemeal necrosis, bridging, and confluent necrosis and multilobular collapse.23 Immunosuppression is thought to enhance viral replication with a subsequent spread in hepatocyte infection. In addition,
HEPATOLOGY, Vol. 37, No. 6, 2003
glucocorticoids may cause a direct reactivation of the latent HBV infection.24,25 The findings of a glucocorticoidresponsive element in HBV DNA and the demonstration that glucocorticoids specifically activate HBV gene expression in cultured human hepatocellular carcinoma cells lend support to the direct reactivation theory.24,25 Sporadic retrospective clinical studies also indicate that steroid-free chemotherapeutic regimens decrease the risk of hepatitis flares of HBV carriers.13,26 One possible approach to reduce or prevent reactivation of HBV DNA and consequent hepatitis is to omit steroids in chemotherapy regimens. Although steroids are ingredients of almost all first-line chemotherapeutic regimens for NHL, they do not appear to be indispensable. Steroid-free protocols such as COMLA, VIM, MIME, and infusional CDE regimen and an oral combination regimen may be as effective as steroid-containing protocols in the treatment of fresh and recurrent lymphoma.27-31 One small-series retrospective study showed that steroid-free chemotherapy may control lymphoma as well as steroid-containing regimens while significantly decreasing the risk of hepatitis flares in HBV carriers.13 This prospective study was conducted to determine whether a steroid-free chemotherapy regimen may decrease the risk of HBV reactivation and hepatitis development in HBsAg-positive lymphoma patients. It also compared the efficacy of steroid-free and steroid-containing regimens in lymphoma control.
Patients and Methods In November 1995, the Taiwan Cooperative Oncology Group (TCOG) initiated a multicenter randomized clinical trial to determine whether steroid-free chemotherapeutic regimens decrease the risk of HBV reactivation and hepatitis development in HBsAg-positive carriers. Other objectives included comparing the efficacy of steroid-free chemotherapeutic regimens in lymphoma control to steroid-containing regimens, and to examine the activity of HBV replication during the course of chemotherapy. The study was approved by the ethical committees of the participating hospitals. Patient Characteristics. Patients eligible for study participation were 15 to 72 years old, HBsAg-positive, and had histologically proven NHL for which intensive chemotherapy was indicated. Patients were excluded from the study if they were pregnant or breast-feeding, had cirrhosis of Child’s class B or C, impaired cardiac function based on New York Heart Association classification grade II or above, concurrent glucocorticoid use for other reasons, previous or concurrent chemotherapy or radiotherapy, or had a history of central nervous system
CHENG ET AL.
1321
involvement. An acceptable hematological and biochemical profile was mandatory prior to chemotherapy: this was defined as absolute granulocyte count ⱖ2,000/mm3, platelet ⱖ100,000/mm3, total bilirubin ⱕ2.5 mg/dL, alanine aminotransferase (ALT) ⱕ200 IU/I, serum creatinine ⱕ1.5 mg/dL, and blood urea nitrogen ⱕ25 mg/dL. All patients had objectively measurable disease when entering the trial. Written informed consent was obtained from all participants. The following tests were performed in all patients: hepatitis B serology (HBsAg, antibody to HBsAg, hepatitis B e antigen [HBeAg], and antibody to HBe), hepatitis C and D (anti-HCV and anti-delta) serology, serum biochemistry (bilirubin, ALT, prothrombin time/partial thromboplastin time, lactate dehydrogenase, alkaline phosphatase), full blood count with leukocyte differentials, blood urea nitrogen and creatinine, electrolytes (sodium, potassium, chloride, and calcium), and serum immunoglobulin (IgG, IgA, and IgM). Bone marrow biopsy, chest X-ray, electrocardiogram, left ventricular ejection fraction, abdominal computed tomography, tumor measurement, and urinalysis were performed and patients’ performance status was determined prior to chemotherapy. Stratified randomization was done by serum ALT level (normal vs. abnormal) and the tumor stage (I/II vs. III/IV). A schematic of the study is shown in Fig. 1. Chemotherapy Regimens. PACE (prednisolone 60 mg/m2/d orally, day 1-7; epirubicin 60 mg/m2 intravenously [IV], day 1; cyclophosphamide 650 mg/m2 IV, day 1; etoposide 55 mg/m2/d IV, day 1-3) was given as first-line therapy for the steroid-containing group and ACE (epirubicin 60 mg/m2 IV, day 1; cyclophosphamide 650 mg/m2 IV, day 1; etoposide 55 mg/m2/d IV, day 1-3) for the steroid-free group. For patients who failed the first-line therapy, respective steroid-containing and steroid-free second-line salvage chemotherapy were used. VIMP (VP-16 100 mg/m2/d IV, day 1, 3, 5; ifosfamide 1 g/m2/d IV, day 1-5; methotrexate 30 mg/m2/d IV, day 1, 5; prednisolone 60 mg/m2/d orally, day 1-7) was given for the steroid-containing group. VIM (VP-16 100 mg/m2/d IV, day 1, 3, 5; ifosfamide 1 g/m2/d IV, day 1-5; methotrexate 30 mg/m2/d IV, day 1,5) was given for the steroidfree group. The decision to use these regimens was made by consensus vote of the members of the TCOG lymphoma committee. No crossing-over between steroidcontaining and steroid-free treatment was allowed. Hepatitis B Virus Serology. Serum samples were collected before and around the nadir of each cycle of chemotherapy and for at least 4 months after the end of the protocol treatment. Analysis of the serum samples was centralized at the Hepatitis Research Center of National
1322
CHENG ET AL.
Fig. 1. Schema of the clinical trial. PACE, prednisolone/adriamycin/ cyclophosphamide/etoposide; ACE, adriamycin/cyclophosphamide/etoposide; CR, complete remission; PR, partial remission; SD, stable disease; PD, progressive disease; VIMP, VP-16/ifosfamide/methotrexate/prednisolone; VIM, VP-16/ifosfamide/methotrexate.
Taiwan University College of Medicine. Testing was performed for liver function, HBsAg, and HBeAg. HBV DNA was tested in duplicates using Chiron bDNA assay (VERSANT HBV DNA assay, Chiron diagnostics, Emeryville, CA). The sensitivity of the assay was 2.5 pg/mL. In cases with inconsistent results, polymerase chain reaction for HBV DNA detection was performed. Briefly, serum DNA was extracted using a commercially available kit (QIAamp DNA Blood Mini Kit, QIAGEN Inc, Valencia, CA). The presence of serum HBV DNA was detected by single-round or nested polymerase chain reaction using primer sets from surface region. The outer primers were S-1s (nucleotides [nt] 159-178: 5⬘AGAACATCGCATCAGGACTC-3⬘) and S-2a (nt 642623: 5⬘-CATAGGTATCTTGCGAAAGC-3⬘), and the inner primers were S-3s (nt 181-200: 5⬘-AGGACCCCTGFCGTGTTAC-3⬘) and S-4a (nt 619-600: 5⬘-AGATGATGGGATGGGAATAC-3⬘). Clinical hepatitis flare-up was defined as a 3-fold or greater increase in serum ALT level that exceeded 100 IU/L. The hepatitis or hepatitis flare-up was attributed to reactivation of chronic hepatitis B when there was a 10fold or higher elevation of serum HBV DNA compared with the previous baseline level or reappearance of HBV
HEPATOLOGY, June 2003
DNA or HBeAg in the serum. For cases of clinical hepatitis, serum samples were collected at 2-week intervals regardless of the schedule of chemotherapy. Regular schedule was resumed if hepatitis subsided and chemotherapy was restarted. Evaluation of Treatment-Related Toxicities, Tumor Response, and Survival. Hemogram was assessed weekly and blood chemistry was assessed biweekly during systemic chemotherapy. Imaging studies were performed monthly until tumor response could be documented. Subsequently, if no intercurrent events required earlier examinations, imaging studies were performed every 3 months for 1 year and every 6 months thereafter. The time to HBV DNA reactivation was the time from entry into the study until the first event of HBV DNA reactivation. Patients without evidence of HBV DNA reactivation were censored from the analysis at the time of most recent follow-up of HBV DNA. Complete remission was defined as the disappearance of all clinically detectable tumors for at least 4 weeks. Partial remission was defined as at least a ⱖ50% reduction in the sum of the products of all measurable tumors, without the appearance of any new lesions for at least 4 weeks. Stable disease was defined as a reduction of less than 50% or an increase of less than 25% of all measurable tumors with no appearance of new lesions for at least 4 weeks. Progression-free survival duration was defined as the interval from entry into the study to either death due to any cause or to disease progression (whichever occurred first). Patients alive without evidence of disease were censored at the time of most recent follow-up. Survival time was the interval from study entry until death, and surviving patients were censored at the time of their most recent follow-up. Statistical Design and Analysis. Patients were randomized to receive either steroid-containing or steroid-free chemotherapy. Stratification was performed according to status of transaminase (normal vs. abnormal) and tumor stage (I/II vs. III/IV). Abnormal transamimase level was defined as serum levels of ALT or aspartate aminotransferase being higher than their upper normal values at the prestudy evaluation. The primary end point was the incidence of HBV reactivation and associated hepatitis flare-up during treatment. The primary objective was to compare the incidence of HBV reactivation and associated hepatitis of the steroid-containing regimen with that of the steroid-free regimen. The objective hypothesis of the trial was to detect a difference of 35% in the incidence of HBV reactivation and associated hepatitis between the 2 arms from 45% to 10%. With a 2-sided binomial test at
HEPATOLOGY, Vol. 37, No. 6, 2003
the 5% significance level and 90% power, this objective required 76 evaluable patients, with 38 patients in each arm. Because of the relatively small number of patients, stratification was based only on 2 of the prognostic factors that were considered most relevant to this study. The planned accrual period was 4 years. All analyses were carried out according to the intentto-treat principle, and statistical significance was set at the .05 level. Time-to-event comparisons were performed using the log-rank test. Survival curves were estimated by the Kaplan-Meier technique. Comparison of 2 proportions was performed using Fisher’s exact test (2-tail) and comparison of continuous variables was performed using the t-test. The statistical software used for the analyses in this study was SAS V8.
Results For ethical reasons, the interim data were monitored by the Data and Safety Monitoring Committee (DSMC) of TCOG every 6 months. The trial was terminated early by the DSMC in February 2000 because of the high hepatitis flare-up rate of the PACE arm. The availability of potentially effective antiviral agents for prophylaxis of HBV reactivation in early 2000 was another major ethical concern for early termination. From November 1995 through February 2000, a total of 50 patients were enrolled, 25 patients in each arm. Forty-nine patients were evaluable for HBV reactivation, and 47 patients were evaluable for tumor response. One patient was excluded from any analysis because of wrong pathologic diagnosis, another 2 patients were excluded from evaluation of tumor response due to refusal of further treatment after 2 and 5 courses of scheduled chemotherapy. Both of the latter 2 patients complained of general malaise after chemotherapy although no grade II or higher toxicity was documented. The median duration of follow-up was 919 and 1,079 days for the ACE and PACE arm, respectively. The clinicopathologic features of the patients were well balanced between the 2 arms (Table 1). Four patients of the PACE arm and 3 patients of the ACE arm had received second-line steroid-containing and steroid-free chemotherapy, respectively. HBV Reactivation and Development of Clinical Hepatitis. A total of 27 patients, 18 of 25 (72%) from the PACE arm and 9 of 24 (37.5%) from the ACE arm (P ⫽ .022), developed HBV reactivation. The relative risk of PACE versus ACE was 1.9 (1.08-3.40, 95% confidence interval) (Table 2). The results were similar if the analyses were adjusted for stratification factors: status of transaminase (normal vs. abnormal) and tumor stage (I/II vs. III/ IV) (Table 3). As shown in Fig. 2, after starting
CHENG ET AL.
1323
Table 1. Clinicopathologic Characteristics of the Patients No. of eligible patients Sex Male (%) Female (%) Age (y) Mean Median Range Pathology Ibl/DL (%) Others (%)* Clinical stage III/IV (%) Elevated LDH (%) B symptoms (%) Bulky tumor, ⬎8 cm (%) Courses of chemotherapy Mean Median Range HBeAg Positive (%) Negative (%) HBV DNA Negative (%) Positive (%) HBV DNA (pg/mL) Mean Range
PACE
ACE
P Value
25
24
13 (52) 12 (48)
16 (67) 8 (33)
41.4 44.4 16.9-68.6
47.6 47.9 23.8-70.1
.142
19 (76) 6 (24) 15 (60) 11 (44) 9 (36) 5 (20)
19 (79) 5 (21) 13 (54) 13 (54) 6 (25) 8 (33)
1.000
5.8 6 2-10
5.7 6 3-8
.855
6 (24) 19 (76)
3 (13) 21 (87)
9 (36) 16 (64)
15 (63) 9 (37)
716.7 ⬍2.5-5,800
360.0 ⬍2.5-2,400
.387
.776 .572 .538 .345
.4635
.0887
.3314
Abbreviation: Ibl/DL, immunoblastic/diffuse large cell. *PACE arm: diffuse polymorphic T cell (1), diffuse mixed cell (2), follicular large cell (1), follicular mixed cell (1), and diffuse small cleaved cell (1); ACE arm: anaplastic large cell (2), diffuse mixed cell (2), and diffuse small cleaved cell (1).
chemotherapy, the likelihood of HBV reactivation was consistently higher for patients in the PACE arm (P ⫽ .03). Nine months after starting chemotherapy, the cumulative incidence of HBV reactivation was 73% for the PACE arm and 38% for the ACE arm. We also compared the cumulative incidence of HBV reactivation between the two groups. The hazard ratio of PACE versus ACE was 2.36 (1.05-5.29, 95% confidence interval, P ⫽ .037) with unadjusted analysis and was 2.62 (1.14-6.00, 95% confidence interval, P ⫽ .023) with Cox regression analysis adjusting for the stratification factors. Of the 27 patients with HBV reactivation, 23 (85.2%) developed clinical hepatitis flare-up as evidenced by a significant elevation of serum ALT closely associated or following the increment of HBV DNA. The degree of clinical hepatitis flare-up was significantly more severe among patients in the PACE arm (Table 2). The results were similar if the analyses were adjusted for stratification factors (Table 3). Lethal fulminant hepatitis developed in 1 patient in the PACE arm but in no patients in the ACE arm. Surprisingly, all clinically significant elevations of ALT during the period of chemotherapy were preceded
1324
CHENG ET AL.
HEPATOLOGY, June 2003
Table 2. Incidence of HBV Reactivation and Severe Clinical Hepatitis P Value
No. of eligible patients HBV reactivation HBV reactivation and hepatitis flare-up HBV reactivation and ALT ⬎10 ⫻ N HBV reactivation and bilirubin ⬎1.5 ⫻ N
PACE
ACE
25 18 (72%) 15 (60%) 11 (44%) 7 (28%)
24 9 (37.5%) 8 (33%) 3 (13%) 1 (4%)
RR (95% CI)
P Value
(Unadjusted*)
.022 .088 .025 .049
1.9 (1.08, 3.40) 1.8 (0.94, 3.45) 3.5 (1.12, 11.08) 6.7 (0.89, 50.61)
RR (95% CI) (Adjusted†)
.021 .069 .020 .025
1.9 (1.06, 3.42) 1.8 (0.93, 3.51) 3.5 (1.07, 11.44) 6.4 (0.84, 49.25)
Abbreviations: ALT, alanine transaminase; N, normal; RR, Relative Risk (PACE vs. ACE); CI, confidence interval. *Using 2 test. †Using Cochran-Mantel-Haenszel test by adjusting stratifying factors status of transaminase and tumor stage.
by an increase in serum HBV DNA. No other causes of clinical hepatitis could be detected in this group of patients during this period. The magnitude of HBV reactivation, as reflected by the increase in serum HBV DNA, was similar between patients in the PACE and ACE arm (Table 3). As shown in Fig. 3, we characterized the patterns of HBV reactivation during systemic chemotherapy of HBsAg-positive lymphoma patients. The most common HBV reactivation, which occurred in 12 of the 27 (44.4%) reactivators, was characterized by a rapid and transient surge of HBV DNA that resolved within 4 to 5 weeks after ALT elevation (Fig. 3A). Four patients developed multiple HBV reactivation, characterized by repeated HBV increment after apparent resolution from previous episodes (Fig. 3B). Another 10 patients developed protracted HBV reactivation, characterized by a prolonged period of persistently high titer of HBV DNA. Four of these 10 patients finally achieved resolution of HBV reactivation after several bouts of clinical hepatitis (Fig. 3C), while the other 6 patients stayed in the status of high-titer of HBV DNA throughout the entire period of chemotherapy (Fig. 3D). These patterns of HBV reactivation showed no preference of occurrence in either the PACE or ACE arm. The timing of HBV reactivation in relation to the start of chemotherapy was interesting. As illustrated in Fig. 3B-D, reactivation of HBV DNA occurred within 2 weeks of the first dose of chemotherapy in 13 of the 27 (48.1 %) cases of reactivation. However, this early reactivation occurred with similar frequency among patients in
both the PACE (9 of 18, 50%) and ACE arm (4 of 9, 44%), which suggests that direct promoting of HBV replication by glucocorticoids was not the major mechanism for this phenomenon.24,25 All except one patient in the PACE arm developed HBV reactivation during the use of first-line chemotherapy. Tumor Response and Survival. Tumor response to first-line chemotherapy was evaluable in 24 and 23 patients of the PACE and ACE arm, respectively. Complete remission was found in 11 patients (46%) of the PACE arm and 8 patients (35%) of the ACE arm (P ⫽ .556, Fisher’s exact test). As shown in Fig. 4A, the estimated overall survival at 46 months was 68% for patients in the PACE arm and 36% for patients in the ACE arm (P ⫽ .18, log-rank test). The cause of death was attributed to uncontrolled lymphoma in 4 and 9 patients of the PACE and ACE arm, respectively. No significant difference was found in progression-free survival between patients in the PACE and ACE arm (P ⫽ .85) (Fig. 4B). Other Serologic Studies. At the last serologic sampling, all 49 patients remained HBsAg-positive; while 1 HBeAg-positive patient had become HBeAg-negative
Table 3. Magnitude of HBV Reactivation of the Patients No. of Patients Increment of HBV DNA 10⬃⬍102 ⫻ N 102⬃⬍103 ⫻ N 103⬃⬍104 ⫻ N
PACE (%)
ACE (%)
18
9
5 (28) 9 (50) 4 (22)
4 (45) 3 (33) 2 (22)
P Value
.776
Fig. 2. Development of HBV DNA reactivation during systemic chemotherapy.
HEPATOLOGY, Vol. 37, No. 6, 2003
CHENG ET AL.
1325
Fig. 3. Classification of HBV reactivation according to duration and frequency of HBV DNA increment. Four representative cases are shown. The units of the 3 parameters at 1X are ALT, 40 U/L; Bil, 1.0 mg/dL; and HBV DNA, 2.5 pg/mL.
and 2 HBeAg-negative patients had become HBeAg-positive. Patients with prechemotherapy HBeAg-positive status appeared to have a lower risk (2 of 9, 22%) of HBV reactivation than those with HBeAg-negative status (25 of 40, 63%) (P ⫽ .06, Fisher’s exact test). HBV reactivation was not observed in any of the 9 patients with prechemotherapy HBV DNA titer more than 1,000 pg/mL.
Discussion In this prospective study, we demonstrated that steroid-free chemotherapy significantly decreases the incidence and severity of HBV reactivation in HBsAgpositive lymphoma patients. We also found that all ALT elevations of clinical significance during systemic chemo-
therapy in our patients was due to HBV reactivation. Other possible causes of hepatic damage, such as lymphoma infiltration in liver, anti-cancer drug–induced hepatotoxicity, or infection by other hepatotropic viruses, did not occur in this series. The reported frequency of HBV reactivation in HBsAg-positive lymphoma patients undergoing cancer chemotherapy has ranged from 14% to 72%.8,10 This wide variation in the reported reactivation rate is partly due to incomplete observation of the HBV DNA in many of the studies. As shown in our study, HBV DNA titer usually decreased rapidly soon after ALT elevation, and ALT elevation tended to be transient and minor in many patients. Even a minor transient elevation of ALT, which is
Fig. 4. (A) Overall survival and (B) progression-free survival of the patients.
1326
CHENG ET AL.
often overlooked clinically, was frequently preceded by more than a 1,000-fold increment of HBV DNA. These features have led to the underestimation of the true incidence of HBV reactivation in retrospective studies that rely on detecting ALT elevation as the first clue of HBV reactivation. Several factors are linked to the incidence and severity of hepatitis flare-up in HBsAg-positive cancer patients undergoing chemotherapy, including the proliferation status of the virus in hepatocytes, the intensity of the chemotherapeutic regimens, and the use of glucocorticoids.16 In this study, we confirm that hepatitis flares occur more frequently in patients who are in the nonreplicative phase of infection (i.e., HBeAg negative and serum HBV DNA negative).14,16,32 During these episodes of hepatitis flares, serum aminotransferase levels increase in response to the sudden reemergence of viral replication. However, none of our patients with chemotherapy-induced hepatitis flares seroconverted to HBsAgnegative status, suggesting that most of these episodes were abortive attempts of cytotoxic T cells to wipe out the reproliferating viruses. Results of several retrospective studies suggest that the severity of hepatitis flare is roughly correlated with the intensity of chemotherapy.5,12,14,15 The PACE and ACE regimens used in our study had an intensity similar to or slightly higher than that of the first-generation regimens such as CHOP (cyclophosphamide, adriamycin, vincristine, and prednisolone). However, the complete response rates of patients in the PACE (46%) and ACE (35%) arms are slightly lower than those usually expected for CHOP. This may be partly due to the relatively stringent criteria of complete remission adopted by this study, in which the existence of residual tiny lesions of unknown significance was considered partial remission in most patients. The treatment intensity of PACE, with an additional drug, prednisolone, may be higher than that of ACE. However, the dramatically increased incidence and severity of hepatitis flares among patients in the PACE arm cannot be fully explained by the slight increase of treatment intensity of PACE. One of the earliest observations suggesting a relationship between glucocorticoids and HBV replication was that long-term treatment resulted in increased expression of hepatitis B core antigen in the serum.33 It was subsequently discovered that HBV DNA polymerase activity increased during a short course of glucocorticoid therapy.34 In vitro studies also indicate that glucocorticoids may activate HBV replication via a steroid-responsive element located at the promoter of HBV DNA.24,25 Results of sporadic clinical studies suggest that removal of steroids
HEPATOLOGY, June 2003
from chemotherapy of hematologic malignancies significantly decreases the risk of hepatitis flares in HBV carriers.13,26 The current study confirms that the use of glucocorticoids in HBsAg-positive lymphoma patients is one of the predominant risk factors for HBV reactivation. One key issue that remains unanswered is whether steroid-free chemotherapy is less effective in the control of lymphoma. Results of several single-arm studies suggest that steroid-free chemotherapeutic regimens control lymphoma as well as steroid-containing regimens.27-31 In the current study, although the steroid-free arm showed a trend toward less satisfactory overall survival and more patients in the ACE arm died of uncontrolled lymphoma, these differences were not significant. However, further study is needed to determine if the benefit of decreased hepatitis flares with use of steroid-free chemotherapy is offset by less satisfactory lymphoma control. In this prospective study, we described the various patterns of HBV reactivation during systemic chemotherapy of HBsAg-positive lymphoma patients. We were surprised to notice that nearly half of the HBV reactivation developed within 2 weeks of the first dose of systemic chemotherapy. At least some of these “early reactivators” developed HBV reactivation before the development of neutropenia, suggesting that mechanisms of HBV reactivation other than immune suppression are involved. One possible mechanism is direct activation of HBV replication by the anti-cancer drugs. Although a steroid-responsive element was found in the promoter region of the HBV genome,24 our results suggest that steroids are not the activator since both ACE and PACE arms had the same proportion of early reactivators. The possibility that cytotoxic agents other than steroids may activate the replication of HBV genome should also be considered. This study identified several clinically relevant patterns of HBV reactivation during systemic chemotherapy. Ten of our patients developed protracted HBV reactivation characterized by a prolonged high titer of HBV DNA after ALT elevation, which was often associated with repeated episodes of clinical hepatitis. One of these 10 patients died of fulminant hepatitis. Four of our patients had multiple episodes of reactivation (i.e., repeated HBV reactivation after apparent resolution of the previous episode). Multiple episodes of reactivation and remission have been shown to accelerate the progression of chronic hepatitis.35,36 Further, repeated attacks of clinical hepatitis often jeopardize the scheduled chemotherapy. Longer follow-up is needed to clarify the long-term effect on the liver by the various patterns of HBV reactivation. Although antiviral agents are now available for the prevention and treatment of HBV reactivation,16,37-39 pro-
HEPATOLOGY, Vol. 37, No. 6, 2003
spective randomized studies are needed to confirm efficacy. In this study, we confirmed that steroid-free chemotherapy significantly decreases the risk of hepatitis flare-up in HBsAg-positive lymphoma patients. However, whether steroid-free chemotherapy may jeopardize the efficacy of lymphoma control needs to be evaluated in future research to determine the efficacy of combining potentially effective antiviral agents with either steroidcontaining or steroid-free chemotherapy in the treatment of HBsAg-positive lymphoma patients. An ongoing prospective randomized study of TCOG is currently studying the efficacy of lamivudine in either the prevention or treatment of HBV reactivation in HBV carriers receiving steroid-containing chemotherapy. Preliminary data of this study suggest a good efficacy of lamivudine in both arms. Before the final reports of all these clinical studies are available, it may be prudent to advise the use of standard steroid-containing regimens for HBV-carrier lymphoma patients and treat HBV reactivation with effective antiviral agents as early as possible.
References 1. Galbraith RM, Eddleston AL, Williams R, Zuckerman AJ. Fulminant hepatic failure in leukaemia and choriocarcinoma related to withdrawal of cytotoxic drug therapy. Lancet 1975;2:528-530. 2. Wands JR, Chura CM, Roll FJ, Maddrey WC. Serial studies of hepatitisassociated antigen and antibody in patients receiving antitumor chemotherapy for myeloproliferative and lymphoproliferative disorders. Gastroenterology 1975;68:105-112. 3. Hoofnagle JH, Dusheiko GM, Schafer DF, Jones EA, Micetich KC, Young RC, Costa J. Reactivation of chronic hepatitis B virus infection by cancer chemotherapy. Ann Intern Med 1982;96:447-449. 4. Thung SN, Gerber MA, Klion F, Gilbert H. Massive hepatic necrosis after chemotherapy withdrawal in a hepatitis B virus carrier. Arch Intern Med 1985;145:1313-1314. 5. Lau JYN, Lai CL, Lin HJ, Lok ASF, Liang RHS, Wu PC, Chan TK, Todd D. Fatal reactivation of chronic hepatitis B virus infection following withdrawal of chemotherapy in lymphoma patients. Q J Med 1989;73:911917. 6. Pariente EA, Goudeau A, Dubois F, Degott C, Gluckman E, Devergie A, Brechot C, et al. Fulminant hepatitis due to reactivation of chronic hepatitis B virus infection after allogeneic bone marrow transplantation. Dig Dis Sci 1988;33:1185-1191. 7. Liang RHS, Lok ASF, Lai CL, Chan TK, Tood D, Chiu EKW. Hepatitis B infection in patients with lymphomas. Hematol Oncol 1990;8:261-270. 8. Lok ASF, Liang RHS, Chiu EKW, Wong KL, Chan TK, Todd D. Reactivation of hepatitis B virus replication in patients receiving cytotoxic therapy—report of a prospective study. Gastroenterology 1991;100:182-188. 9. Liaw YF. Hepatitis viruses under immunosuppressive agents. J Gastroenterol Hepatol 1998;13:14-20. 10. Alexopoulos CG, Vaslamatzis M, Hatzidimitriou G. Prevalence of hepatitis B virus marker positivity and evolution of hepatitis B virus profile, during chemotherapy, in patients with solid tumors. Br J Cancer 1999;81: 69-74. 11. Markovic S, Drozina G, Vovk M, Fidler-Jenko M. Reactivation of hepatitis B but not hepatitis C in patients with malignant lymphoma and immunosuppressive therapy—a prospective study in 305 patients. Hepatogastroenterology 1999;46:2925-2930.
CHENG ET AL.
1327
12. Liang R, Lau GKK, Kwong YL. Chemotherapy and bone marrow transplantation for cancer patients who are also chronic hepatitis B carriers: a review of the problem. J Clin Oncol 1999;17:394-398. 13. Cheng AL. Steroid-free chemotherapy decreases the risk of hepatitis flare-up in hepatitis B virus carriers with non-Hodgkin’s lymphoma. Blood 1996;87:1202. 14. Kumagai K, Takagi T, Nakamura S, Sawada U, Kura Y, Kodama F, Shimano S, et al. Hepatitis B virus carriers in the treatment of malignant lymphoma: an epidemiological study in Japan. Ann Oncol 1997:8(Suppl 1):107-109. 15. Faggioli P, Paschale M, Tocci A, Luoni M, Fava S, De Paoli A, Tosi A, Cassi E. Acute hepatic toxicity during cyclic chemotherapy in nonHodgkin’s lymphoma. Haematologica 1997;82:38-42. 16. Perrillo RP. Acute flares in chronic hepatitis B: the natural and unnatural history of an immunologically mediated liver disease. Gastroenterology 2001;120:1009-1022. 17. Sagnelli E, Manzillo G, Mario G, Pasquale G, Felaco FM, Filippini P, Izzo CM, et al. Serum levels of hepatitis B surface and core antigens during immunosuppressive treatment of HBsAg-positive chronic active hepatitis. Lancet 1980;2:395-397. 18. Scullard GH, Smith CI, Merigan TC, Roinson WS, Gregory PB. Effects of immunosuppressive therapy on viral markers in chronic active hepatitis B. Gastroenterology 1981;81:987-991. 19. Lam KC, Ching LL, Ng RP, Trepo C, Wu PC. Deleterious effect of predinosolone in HBsAg-Positive chronic active hepatitis. N Engl J Med 1981;304:380-386. 20. Hoofnagle JH, Davis GL, Pappas SC, Hanson RG, Peters M, Avigan MI, Waggoner JG, et al. A short course of prednisolone in chronic type B hepatitis. Ann Intern Med 1986;104:12-17. 21. Lee SD, Tong MJ, Wu JC, Lin HC, Tsai YT, Lo KJ. A randomized double-blind placebo-controlled trial of prednisolone therapy in HBeAg and HBV DNA positive Chinese patients with chronic active hepatitis B. J Hepatol 1991;12:146-150. 22. Hanson RG, Peters MG, Hoofnagle JH. Effects of immunosuppressive therapy with prednisolone on B and T lymphocyte function in patients with chronic type B hepatitis. HEPATOLOGY 1986;6:173-179. 23. Czaja AJ, Wolf AM, Baggenstoss AH. Laboratory assessment of severe chronic active liver disease during and after corticosteroid therapy: correlation of serum transaminase and gamma globulin levels with histologic features. Gastroenterology 1981;80:687-692. 24. Tur-Kaspa R, Burk RD, Shaul Y, Shafrits DA. Hepatitis B virus DNA contains a glucocorticoid responsive element. Proc Natl Acad Sci U S A 1986;83:1627-1631. 25. Chou CK, Wang LH, Lin HM, Chi CW. Glucocorticoid stimulates hepatitis B viral gene expression in cultured human hepatoma cells. HEPATOLOGY 1992;16:13-18. 26. Ohtsu T, Sai T, Oka M, Sugai Y, Tobinai K. Activation of hepatitis B virus infection by chemotherapy containing glucocorticoid in hepatitis B virus carriers with hematologic malignancies. Jpn J Clin Oncol 1991;21:360365. 27. Sparano JA, Wiernik PH, Leaf A, Dutcher JP. Infusional cyclophosphamide, doxorubicin, and etoposide in relapsed and resistant non-Hodgkin’s lymphoma: evidence for a schedule-dependent effect favoring infusional administration. J Clin Oncol 1993:11:1071-1079. 28. Cabanillas F. Non-Hodgkin’s lymphomas: a review of the M.D. Anderson experience. Semin Oncol 1993;19:11-13. 29. Herbrecht R, Damonte JC, Dufoure P, Maloisel F, Liuk L, Ortiz S, Bergerat JP, et al. Etoposide, ifosfamide and methotrexate combination chemotherapy for aggressive non-Hodgkin’s lymphoma after failure of the NLH 84 regimen. Semin Oncol 1992;197:7-10. 30. Berd D, Cornog J, Decont RC, Levitt M, Bertino JR. Long-term remission in diffuse histiocytic lymphoma treated with combination sequential chemotherapy. Cancer 1975;35:1050. 31. Remick SC, McSharry JJ, Wolf BC, Blanchard CG, Eastman AY, Wagher H, Portuese E, et al. Novel oral combination chemotherapy in the treat-
1328
32.
33.
34.
35.
CHENG ET AL.
ment of intermediate-grade and high grade AIDS-related non-Hodgkin’s lymphoma. J Clin Oncol 1993;11:1691-1702. Davis GL, Hoofnagle JH, Waggoner JG. Spontaneous reactivation of chronic hepatitis B virus infection. Gastroenterology 1984;86:230235. Sagnelli E, Manzio G, Maio G, Pasquale G, Felaco FM, Filippini P, Izzo CM, et al. Serum levels of hepatitis B surface and core antigens during immunosuppressive treatment of HBsAg-positive chronic active hepatitis. Lancet 1980;2:395-397. Nair PV, Tong MJ, Stevenson D, Roskamp D, Boone C. Effects of shortterm, high-dose prednisone treatment of patients with HBsAg-positive chronic active hepatitis. Liver 1985;5:8-12. Perrillo RP, Campbell CR, Sanders GE, Regenstein FG, Bodicky CJ. Spontaneous clearance and reactivation of chronic hepatitis B virus
HEPATOLOGY, June 2003
36. 37.
38.
39.
infection among male homosexuals with chronic type B hepatitis. Ann Intern Med 1984;100:43-46. Hoofnagle JH, Seeff LB. Natural history of chronic type B hepatitis. Prog Liver Dis 1982;7:469-479. Rossi G, Pelizzari A, Motta M, Puoti M. Primary prophylaxis with lamivudine of hepatitis B virus reactivation in chronic HbsAg carriers with lymphoid malignancies treated with chemotherapy. Br J Haematol. 2001;115:58-62. Liao CA, Lee CM, Wu HC, Wang MC, Lu SN, Eng HL. Lamivudine for the treatment of hepatitis B virus reactivation following chemotherapy for non-Hodgkin’s lymphoma. Br J Haematol 2002;116:166-169. Marcello P, Fedele DM, Giovanni DGR, Alessandro S, Bruno P, Marco P, Aristide M, et al. Efficacy of lamivudine to prevent hepatitis reactivation in hepatitis B virus-infected patients treated for non-Hodgkin lymphoma. Blood 2002;99:724-725.