Hepatitis C Virus Infection in Patients with B-Cell Non-Hodgkin Lymphoma

Methods

Patients

Between October 1994 and May 1996, 120 consecutive patients with B-cell non-Hodgkin lymphoma were evaluated at the outpatient hematology clinic of the Los Angeles County-University of Southern California (LAC-USC) Medical Center. During the same period, we enrolled 268 additional patients as controls: 154 unselected patients with malignant hematologic conditions other than B-cell non-Hodgkin lymphoma who were seen at the same clinic (control group 1, which comprised 44 patients with acute leukemia, 45 with Hodgkin disease, 11 with T-cell lymphoma, 23 with chronic myelogenous leukemia, 20 with multiple myeloma, 10 with chronic lymphocytic leukemia, and 1 with hairy-cell leukemia) and 114 unselected patients without malignant hematologic conditions who were attending the general medicine clinic at LAC-USC (control group 2, which comprised 69 patients with systemic hypertension or ischemic heart disease, 35 with diabetes mellitus, and 10 with primary hypothyroidism). All potential study participants were tested for antibodies to HIV. Markers for HCV infection (anti-HCV and HCV RNA) and hepatitis B virus infection (hepatitis B surface antigen, antibodies to hepatitis B core antigen, and antibodies to hepatitis B surface antigen) were also determined. Patients who tested positive for antibodies to HIV (2 patients) or hepatitis B surface antigen (6 patients) were excluded. We did not exclude patients with known HCV infection or liver disease. Identical inclusion and exclusion criteria were used to enroll patients and controls.

All patients were given a questionnaire that asked about demographic characteristics and potential risk factors for viral hepatitis and chronic liver disease. A patient's ethnicity was determined on the basis of self-report. A liver panel, which included prothrombin activity and levels of alkaline phosphatase, albumin, globulin, total and direct bilirubin, lactic dehydrogenase, alanine aminotransferase, and aspartate aminotransferase, was obtained for all patients at the time of evaluation. All patients gave written informed consent, and the study was approved by the institutional review board at LAC-USC Medical Center.

Pathologic Evaluation of Lymphoma

Lymphoma was diagnosed on the basis of morphologic evaluation of lymph node tissue or extranodal tissues, including bone marrow specimens. Immunophenotypic analysis for surface B- and T-lymphocytic markers was performed by using monoclonal antibodies, as described elsewhere [18, 19]. Non-Hodgkin lymphoma was graded by using the Modified Working Formulation classification [20, 21]. All slides were independently rereviewed by two expert hematopathologists who were not aware of the HCV infection status of the patients.

Assessment of Hepatitis C Virus Infection

We detected HCV antibodies by using a second-generation enzyme-linked immunosorbent assay (Anti-HCV EIA 2, Ortho Diagnostic Systems, Raritan, New Jersey). Blood samples for HCV RNA determination and HCV genotyping were processed and stored under the optimal conditions described by Davis and colleagues [22]. Serum HCV RNA was measured by a reverse-transcription polymerase chain reaction (RT-PCR) assay (Amplicor HCV test, Roche Molecular Systems, Somerville, New Jersey), as described elsewhere [23]. Briefly, total RNA was extracted from 100 µL of serum with guanidinium thiocyanate and precipitated in isopropanol with poly(A) carrier RNA. An equivalent of 5 µL of serum was reverse transcribed and amplified in a master mixture containing rTth DNA polymerase, biotinylated primers KY80 and KY78, buffer salts, unguentum, deoxyadenosine triphosphate, deoxycytidine triphosphate, 2-deoxyguanosine-5′-triphosphate, and deoxyuridine triphosphate. Deoxyuridine triphosphate was incorporated into each amplification product to serve as a substrate for unguentum (AmpErase, Roche Molecular Systems); this prevented carryover contamination of previously amplified DNA. The reaction was optimized for the use of rTth that, in the presence of manganese, performs both reverse-transcription and DNA-polymerase functions. After RT-PCR was performed, biotin-labeled PCR products were chemically denatured; captured by a solid-phase, HCV-specific probe that was bound to microwell plates; and detected by using an avidin-horseradish peroxidase system with a conventional microtiter plate reader (450 nm). Optical density readings of more than 0.500 were considered positive for HCV, readings of less than 0.300 were considered negative, and readings of 0.300 to 0.500 were considered equivocal.

We did HCV genotyping on all HCV RNA-positive specimens by using genotype-specific primers from the HCV core region under conditions described elsewhere [24]. Two rounds of amplification were done: We used genotype-nonspecific primers during the first round and genotype-specific primers during the second round. The amplification products were resolved on 2% agarose gels stained with ethidium bromide. Hepatitis C virus genotype was determined by genotype-specific bands of complementary DNA. Control samples of each detectable genotype (1a, 1b, 2a, 2b, 3a, 3b, 4, 5a, and 6a) were evaluated in parallel.

Radioimmunoassay (Ausria II, Abbott Laboratories, North Chicago, Illinois) was used to detect hepatitis B surface antigen and antibodies to hepatitis B surface antigen, and enzyme-linked immuno-assay (Corab, Abbott Laboratories) was used to detect antibodies to hepatitis B core antigen.

Statistical Analysis

The primary comparisons among groups for categorical variables were performed by using the Fisher exact test. We calculated the relative risk and 95% CIs that patients with HCV infection compared with patients without HCV infection would have extranodal lymphoma. We also computed 95% CIs for the prevalence of HCV infection in the B-cell lymphoma group and the two control groups. All CIs were calculated by using StatXact3 for Windows (Cytel Software Corp., Cambridge, Massachusetts). All other analyses were performed by using SAS 6.08 for Windows (SAS Institute, Inc., Cary, North Carolina).

Results

Patients in the B-cell lymphoma group and the two control groups were similar with respect to age, sex, ethnicity, and risk factors for viral hepatitis (Table 1). The prevalence of anti-HCV and HCV RNA in the three groups are shown in Table 2. Infection with HCV was detected in 26 of 120 patients with B-cell lymphoma (22% [95% CI, 15% to 30%]) compared with 7 of 154 patients (4.5%) in control group 1 and 6 of 114 patients (5%) in control group 2 (P < 0.001).

In the B-cell non-Hodgkin lymphoma group, both HCV RNA and anti-HCV were detected in 21 patients; HCV RNA was the only detectable marker in 4 patients, and HCV RNA was not detected in 1 anti-HCV-reactive patient. All control patients who were considered to be infected with HCV had detectable HCV RNA, although 3 of these patients were anti-HCV negative (Table 2). The most common HCV genotypes in patients with B-cell non-Hodgkin lymphoma were genotype 1a (12 of 26 patients [46%]), genotype 1b (8 of 26 patients [31%]), mixed genotype 1a and 1b (2 of 26 patients [8%]), and genotype 2b (2 of 26 patients [8%]) This incidence is similar to those reported elsewhere [25, 26] for patients with chronic HCV infection in the United States.

Chronic liver disease had been diagnosed in 4 HCV-positive patients with B-cell lymphoma (15%) 2 to 6 years before lymphoma was diagnosed. Although 17 of 26 HCV-positive patients with lymphoma had increased aminotransferase activity, only 4 had levels of alanine aminotransferase or aspartate aminotransferase that exceeded 100 U/L. Similarly, most HCV-positive patients in the control groups had only mildly elevated aminotransferase levels.

At least one percutaneous risk factor for HCV exposure was identified in 15 of 26 HCV-positive patients (58%) with B-cell non-Hodgkin lymphoma. The period during which patients were at risk for percutaneous exposure to HCV preceded the diagnosis of B-cell lymphoma by a median of 15 years (range, 5 to 35 years). Patients who had been given blood transfusions had received them before their malignant hematologic condition was diagnosed (Table 1). In fact, 39 of 44 patients with acute leukemia in control group 1 had received transfusions after leukemia was diagnosed but before testing for HCV was done. Nevertheless, the prevalence of HCV infection was still considerably lower in control group 1 than in the B-cell lymphoma group.

Patients who were HCV positive and HCV negative did not differ with respect to the prevalence of low-grade (35% and 33%), intermediate-grade (38% and 48%), or high-grade lymphoma (27% and 19%) (P > 0.2 for all comparisons). Four HCV-positive patients with lymphoma initially presented with low- or intermediate-grade lymphoma but had relapse with high-grade lymphoma 2 to 7 years after receiving the initial diagnosis.

Monocytoid B-cell lymphoma [27-31] was the most common type of lymphoma found in HCV-infected patients (23% of HCV-infected patients compared with 7% of patients not infected with HCV; P = 0.034) (Table 3). Furthermore, monocytoid B-cell lymphoma accounted for 67% of all lowgrade lymphomas in HCV-positive patients and 23% of all low-grade lymphomas in HCV-negative patients (P = 0.04).

Extranodal lymphoma occurred more often in HCV-positive patients than in HCV-negative patients (10 of 26 HCV-positive patients compared with 15 of 94 HCV-negative patients; relative risk, 2.4 [CI, 0.85 to 6.6]; P = 0.12). Extranodal disease was seen primarily in patients with intermediate-grade lymphoma (9 of 10 patients) or high-grade lymphoma (7 of 15 patients) (relative risk, 1.9 [CI, 0.7 to 7.0]; P > 0.2). Extranodal lymphoma in the HCV-positive patients was found in the stomach (n = 5), thyroid (n = 1), bone (n = 2), soft tissues of the face (n = 1), and central nervous system (n = 1). No HCV-positive patient with lymphoma was found to have lymphoma of the liver. Extranodal lymphoma in the HCV-negative patients were found in the lip (n = 1), stomach (n = 8), colon and rectum (n = 4), orbit (n = 1), and soft tissues of the head (n = 1).

Discussion

We show that the incidence of HCV infection among patients with B-cell lymphoma was higher than that in two groups of controls that included patients with other malignant hematologic conditions and patients with general medical conditions. A high prevalence of HCV infection among patients with B-cell lymphoma has been reported in some Italian studies [12-15] but has not been confirmed by data from the United Kingdom [16, 17]. To our knowledge, our study is the first report of this kind from the United States. The 22% prevalence of HCV infection that we found in our cohort of patients with B-cell lymphoma concurs with the results reported from Italy. However, the prevalence of HCV infection in our control groups (4.5% and 5%) was higher than that reported in healthy blood donors in the United States [32, 33] and in control groups in various studies from Italy [12-15]. In our study, all patients and controls were of the same socioeconomic level, and the distribution of age, sex, and ethnicity was similar across groups. Furthermore, the prevalence of risk factors for HCV infection was similar among the three groups; this indicates that the control groups were appropriate. The prevalence of anti-HCV in healthy blood donors from LAC-USC Medical Center is 3.2% [34], a rate higher than those found in other studies [32, 33]. Because this center serves a largely indigent urban population, this high prevalence presumably reflects the known association between low socioeconomic status and HCV infection [35]. Because persons with parenteral risk factors are discouraged from donating blood, the higher prevalence of HCV infection in our control groups probably accurately reflects the prevalence of HCV infection in the general population served by our institution. Although the prevalence of HCV infection among the controls was high, the prevalence of HCV infection among the patients with B-cell non-Hodgkin lymphoma was higher still. The distribution of HCV genotypes in our series is similar to that reported in HCV-infected patients in the United States [25, 26]. B-cell lymphoma was not associated with any specific HCV genotype.

The specific characteristics of lymphoma in the HCV-infected patients were noteworthy. In the studies from Italy, lymphoplasmacytoid immunocytoma occurred most frequently [12, 13], but our HCV-positive cohort had an even distribution of low-, intermediate-, and high-grade lymphoma. However, the most common single type of lymphoma in our series was nodal monocytoid B-cell lymphoma, which accounted for 23% of all lymphomas and 67% of all low-grade lymphomas among the HCV-positive patients. When extranodal disease occurred among HCV-positive patients, it was not found in the liver.

Because of the striking association between HCV infection and mixed cryoglobulinemia [4-6, 10], it has been hypothesized that HCV may play a role in the pathogenesis of clonal proliferations of B-cells. The association of HCV infection and B-cell lymphoma that we found raises the possibility that HCV plays a pathogenic role. Because HCV RNA genomic sequences cannot be integrated into the host genome, an indirect mechanism of malignant transformation has been considered. In this regard, HCV is known to be both hepatotrophic and lymphotrophic [3, 10]. The persistence of HCV and its replicative intermediate in peripheral-blood mononuclear cells may result in chronic stimulation of B-cells, which may lead to polyclonal and, later, to monoclonal expansion of these cells. Alternatively, immune defects that predispose patients to B-cell lymphoma may also render them more susceptible to HCV infection.

Additional support for the possible association of clonal B-cell expansion and HCV infection has been provided by Franzin and colleagues [36], who found a high frequency of clonal B-cell expansion in HCV-infected patients, even in the absence of cryoglobulinemia. On the basis of these observations, one may speculate that chronic HCV infection alone or in the presence of other factors may lead to B-lymphoid expansion. The occurrence of a subsequent transforming event may lead to malignant lymphoma [37].

Although two studies have shown a high prevalence of HCV infection in patients with benign or malignant monoclonal gammopathy [38, 39], others have not. In the study by Silvestri and associates [13], only 3 of 78 patients (4%) with multiple myeloma were positive for anti-HCV. In our study, which included fewer patients with multiple myeloma, HCV infection was found in only 1 of 20 patients (5%). Analysis of all patients with B-cell neoplasms other than B-cell lymphoma (20 with multiple myeloma, 10 with chronic lymphocytic leukemia, 1 with hairy-cell leukemia, and 10 with B-cell acute lymphoblastic leukemia) showed that only 1 of 41 patients (2.4%) was infected with HCV; this rate is lower than that found in our controls. This suggests that clonal B-cell expansion in patients with chronic HCV infection may be only one of many factors necessary for the full expression of malignant conditions. Other factors may include genetic, immunologic, or environmental triggers that can vary widely across geographic regions. Geographic variability is also suggested by the difference in the prevalence of HCV infection in patients with B-cell non-Hodgkin lymphoma from Italy and from the United Kingdom.

In conclusion, we show an increased prevalence of HCV infection among patients from the United States with B-cell lymphoma. Because most of our patients were of Hispanic ethnicity, the generalizability of our findings to other populations in the United States may be uncertain. Further work is necessary to elucidate the full pathogenetic mechanisms of lymphoma in the setting of HCV.

Dr. T. Zuckerman: Department of Internal Medicine A, Carmel Medical Center, 7 Michal Street, Haifa 34362, Israel.

Drs. Levine and Douer: Division of Hematology, University of Southern California, 1441 Eastlake Avenue, NOR 3431, Los Angeles, CA 90033.

Dr. Gutekunst: Roche Molecular Systems, 1080 Highway 202, Somerville, NJ 08876.

Dr. Mizokami: Second Department of Internal Medicine, Nagoya City University Medical School Hospital, 1-1 Kawasumi Mizuho, Nagoya 467, Japan.

Mr. Qian: Department of Preventive Medicine, University of Southern California Medical School, CHP 218, 1540 Alcazar Street, Los Angeles, CA 90033.

Drs. Velankar and Nathwani: University of Southern California, 1200 North State Street, Room 2422, Los Angeles, CA 90033.

Dr. Fong: Transplant Hepatology, Cedars-Sinai Medical Center, 8635 West Third Street, Suite 590W, Los Angeles, CA 90048.