GASTROENTEROLOGY

1991;100:162-188

Reactivation of Hepatitis B Virus Replication in Patients Receiving Cytotoxic Therapy Report of a Prospective

Study

ANNA S. F. LOK, RA’YMOND H. S. LIANG, EDMOND K. W. CHIU, KEE-LAM WONG, TAI-KWONG CHAN, and DAVID TODD Department

of Medicine, University

of Hong Kong, Queen Mary Hospital, Hong Kong

One hundred Chinese patients who received induction cytotoxic therapy for malignant lymphoma were prospectively studied to determine the incidence, morbidity, mortality, and predisposing factors for reactivation of hepatitis B virus replication during cytotoxic therapy. In 18 (67%) hepatitis B surface antigen-positive and 10 (14%) hepatitis B surface antigen-negative patients, hepatitis developed during cytotoxic therapy (P < 0.0001). Hepatitis could be attributed to exacerbation or reactivation of chronic hepatitis B in 13 (72%) hepatitis B surface antigen-positive patients but in only 2 (20%) hepatitis B surface antigen-negative patients (P < 0.0001). Sudden increase or reactivation of hepatitis B virus replication gave rise to icteric hepatitis, nonfatal hepatic failure, and death in 22.3%,3.7%,and 3.7% of patients who were positive for hepatitis B surface antigen; in 2%, 2%, and 0% of those positive for hepatitis B antibodies; and in none of those who were seronegative. Among the hepatitis B surface antigen-positive patients, male sex was the only factor that was associated with an increased risk of reactivation of hepatitis B virus replication. We recommend that hepatitis B surface antigen-positive patients with malignancies receiving cytotoxic therapy be closely monitored.

R

eactivation of hepatitis B virus (HBV) replication is a well-recognized complication in patients with chronic HBV infection who receive cytotoxic or immunosuppressive therapy (1-16). Immunosuppression enhances viral replication with consequent increase in hepatocyte infection. Withdrawal of cytotoxic or immunosuppressive drugs leads to restoration of immune function, resulting in rapid destruction of infected hepatocytes. This may give rise to a clinical picture of acute hepatitis and even hepatic failure

(17,lB). In theory, hepatic necrosis under these circumstances can be ameliorated by reinstitution of immunosuppressive therapy (6). Nevertheless, fatal hepatic failure had been reported even in patients who were treated with corticosteroids at the first sign of clinical hepatitis (1,2,8,10,15). Some investigators have questioned whether patients with malignancies and chronic HBV infection should be treated with less aggressive chemotherapeutic regimens. Unfortunately, data on this subject has come from anecdotal reports only. In a retrospective analysis of 484 patients with malignant lymphoma over a 15year period, we found that in a significantly higher proportion of hepatitis B surface antigen (HBsAg)-positive patients, hepatic dysfunction developed during cytotoxic therapy (2 1% vs.5.2% P < 0.001) (18a). The incidence of hepatic dysfunction was probably underestimated since serum biochemistry was regularly monitored in a small proportion of patients only. The disease-free survival and overall survival rates were similar in HBsAgpositive and HBsAg-negative patients, but the mortality from hepatic failure was significantly higher in HBsAg-positive patients: 6% vs. 0.5% in HBsAgnegative patients (P < 0.005). Because of the lack of serial serological data, we were not able to conclude whether the high incidence of hepatic dysfunction in HBsAg-positive patients was attributable to reactivation of HBV replication. A prospective study was conducted to determine

Abbreviations used in this paper: BACOP,bleomycin, doxorubitin, cyclophosphamide, vincristine, prednisone; mBACOD, methotrexate, bleomycin, doxorubicin, cyclophosphamide, vincristine, dexamethasone. o 1991by the American Gastroenterological Association

0016~5065/91/$3.00

January 1991

the incidence, morbidity, mortality, and predisposing factors for reactivation of HBV replication in patients with malignant lymphoma receiving cytotoxic therapy. Patients with one type of malignancy were studied so that most patients would be treated with similar chemotherapeutic regimens. We chose to study patients with malignant lymphoma because most of them survive long enough to complete the induction treatment. We included patients who were HBsAgpositive as well as those who were HBsAg-negative, because it has been shown that reactivation of HBV replication can also occur in patients who have recovered from previous HBV infection (3,4,7,13). Patients and Methods One hundred Chinese patients (53 men and 47 women), aged 14-74 years [median, 52 years), who received induction cytotoxic therapy for malignant lymphoma were prospectively studied during a 4-year period. All patients had non-Hodgkin’s lymphoma. Nineteen (19%) patients had hepatic involvement at the time of diagnosis. The chemotherapeutic regimens included: BACOP (bleomycin, doxorubicin, cyclophosphamide, vincristine, and prednisone) (19) in 53 (53%) patients; mBACOD (methotrexate, bleomycin, doxorubicin, cyclophosphamide, vincristine, and dexamethasone) (20) in 32 (32%) patients; and a variety of other chemotherapeutic regimens in the remaining 15

(15%) patients. The number of courses of cytotoxic therapy ranged from 2-t 5 (median 51. All the patients were tested for hepatitis B serology: [HBsAg, antibody to HBsAg (anti-HBs) and antibody to hepatitis B core antigen (anti-HBc)]; serum biochemistry [albumin, globulin, bilirubin, alkaline phosphatase, gamma glutamyl transpeptidase, aspartate aminotransferase (AST), and alanine aminotransferase (ALT)]; and HBV DNA before treatment and again after each course of cytotoxic therapy (i.e., every 3-5 weeks). Patients who were HBsAg-positive were also tested for hepatitis B e antigen (HBeAg) and antibody to HBeAg (anti-HBe). Additional tests for immunoglobulin M (IgM) antibody to hepatitis A virus (IgM antiHAV) and antibody to hepatitis D virus (anti-HDV) were performed when patients’ serum ALT levels were elevated. Hepatitis B markers, IgM anti-HAV and total anti-HDV were tested by enzyme-linked immunoassays (Auszyme, Ausab, Corzyme, Corzyme-M, HBe, Havab-M, and Antidelta; Abbott Laboratories, North Chicago, IL). Serum biochemistry was determined by sequential multiple autoanalyzer. Serum HBV DNA was detected by direct spot hybridization of 20-t~L serum samples onto nylon membranes (Hybond; Amersham, Buckinghamshire, England) according to the method of Matsuyama et al. (21). Serial samples from the same patient were tested in the same run to minimize interassay variations. Hepatitis was defined as a threefold or greater increase in serum ALT level that exceeded 100 IU/L (normal, to-fold) in serum HBV DNA level or reappearance of HBV DNA and HBeAg in the serum in the absence of evidence of hepatic infiltration by lymphoma, hepatotoxic drugs, recent transfusion, and other systemic infections. Statistical analyses were performed by x2 test or Fisher’s exact test. Results

At presentation, 27 (27%) patients were positive for HBsAg, 33 (33%) for anti-HBs with or without anti-HBc, 18 (18%) for anti-HBc alone, and 22 (22%) were seronegative for HBV markers. The sex distribution, age, histological type of lymphoma (according to Working Formulation) (22), and incidence of hepatic involvement were similar in HBsAg-positive and HBsAg-negative patients. The two groups of patients were treated with similar chemotherapeutic regimens and received similar number of courses of cytotoxic patherapy [Table 1). Eight (30%) HBsAg-positive tients had elevated serum ALT levels at presentation: in four, the elevated AL,T levels were caused by hepatic involvement by lymphoma. Of the remaining four, two had liver biopsies which showed reactive hepatitis _and inactive cirrhosis. Five patients had hypoalbuminemia because of gut involvement by lymphoma, disseminated disease, and/or poor nutrition. Prothrombin time was not checked in all patients before cytotoxic therapy. None of the patients had clinical evidence of cirrhosis (ascites, esophageal varices, or hypersplenism).

Table 1. Comparison of the Demographic Characteristics, Histological Type, and Chemotherapeutic Regimen in Hepatitis B Surface Antigen Positive and Negative Patients With Malignant Lymphoma HBsAg+

HBsAg-

n Sex

27

73

Male (%) Female (%) Age lyr); median (range) Histological type (%)’ Low grade Intermediate grade High grade Others Elevated serum ALT level (%)” Hepatic involvement (%) Chemotherapeutic regimens (%) BACOP mBACOD Others No. of courses of treatment; median (range)

ll(41) 16 (59) 48(19-68)

42 (58) 31(42) 53(14-74)

2 (7.5) 23 (85) 0 2 (7.5)

5 (7) 51(70)

8 (30) 4 (151

6 (8) 15 (21)

13 (48)

40 (55) 22 (30) ll(15)

10 (37) 4 (15)

5 (2-15)

“Classified according to Working Formulation. *ALT normal level, < 40 IU/L.

6 (8) ll(15)

5 (Z-12)

184

GASTROENTEROLOGY Vol. 100, No. 1

LOK ET AL.

Table 2. Incidence and Causes of Hepatitis in Patients With Malignant Lymphoma Receiving Cytotoxic Therapy HBsAgAnti-HBs t , anti-HBck”

HbsAg+

33

27

No. of patients No. developing hepatitis (%) Cause of hepatitis Reactivation of HBV replication Hepatic infiltration by lymphoma Transfusion related non-A, non-B hepatitis Hepatotoxic drugs Systemic sepsis Indeterminate

Anti-HBc+ alone 18

Anti-HBs-, anti-HBc22

18(67)

7 (21)

2 (11)

l(5)

13

1

1

0

1

0

0

2 1 1 1

0 0 0 1

1 0 0 0

“Six patients tested positive for anti-HBs only.

In 18 (67%) HBsAg-positive and 10 (14%) HBsAgnegative patients, hepatitis developed during cytotoxic therapy (P < 0.0001) (Table 2). Hepatitis could be attributed to exacerbation or reactivation of chronic hepatitis B in 13 (72%) HBsAg-positive (Figure 1) but only 2 (20%) HBsAg-negative patients (P < 0.0001) (Figure 2). One of the latter 2 patients tested positive for anti-HBs and anti-HBc, whereas the other tested positive for anti-HBc only. Both became HBsAgpositive during the hepatitis. Non-HBV-related hepatitis occurred with similar frequencies in HBsAgpositive and HBsAg-negative patients (18% vs. 11% P = NS). Among the patients who developed hepatitis, hepatitis was transient, lasting 1 week to 3 months, in 9 (50%) HBsAg-positive and 8 (80%) HBsAg-negative patients. Recurrent (2-4) episodes of hepatitis associated with fluctuations in serum HBV DNA levels were

observed in 7 (38%) HBsAg-positive patients during an observation period of 4-12 months. Recurrent hepatitis was not observed in the HBsAg-negative patients. In one patient in each group, persistent hepatitis developed (lasting more than 6 months). In approximately 50% of patients, the hepatitis was anicteric. Icteric hepatitis was more often seen in HBsAg-positive patients, and in 7 of 8 patients it was caused by sudden increase or reactivation of HBV replication. Hepatic failure with ascites and encephalopathy developed in 2 HBsAg-positive and 3 HBsAgnegative patients, giving an overall incidence of hepatic failure of 7% for HBsAg-positive and 4% for HBsAg-negative patients (P = NS). Hepatic failure was attributed to exacerbation or reactivation of chronic hepatitis B in both HBsAg-positive and 1 of the 3 HBsAg-negative patients. The cause of hepatic failure in the other 2 HBsAg-negative patients was

Chemotherapy HBsAg

+

HBeAg

-

Anti-HBe

+ 600

r

o_

I 0

1

3

2

MONTHS

4

5

Figure 1. Profile of a HBsAg positive patient receiving cytotoxic therapy. He experienced an attack of hepatitis after the third course of cytotoxic therapy. The hepatitis was preceded by elevation in serum HBV DNA level, but the patient remained HBeAgnegative, anti-HBe-positive. ALT level (normal, < 40 IU/L) and serum HBV DNA level scored from 0 to 4 + .

REACTIVATION OF HBV AND CYTOTOXIC THERAPY

January 1991

Chemotherapy HBsAg

-

HBeAg

NT

Anti-HBe

NT

Anti-HBs

+

Figure 2. Profile of a patient who tested negative for HBsAg but positive for anti-HBs and anti-HBc. He became HBsAgpositive but HBeAg- and HBV DNA-negative just before commencement of cytotoxic therapy. He experienced an attack of hepatitis after the second course of cytotoxic therapy. This was associated with the appearance of HBeAg and high levels of HBV DNA in the serum. ALT level (normal,