Lack of Preferential Localization of Tumoral Mass in B-Cell Non-Hodgkin's Lymphoma Associated With Hepatitis C Virus Infection

To the Editor:

The possible association between hepatitis C virus (HCV) infection and B-cell–type non-Hodgkin's lymphoma (NHL) has previously been suggested.1-4 HCV tropism for hepatocytes and also, according to recent data, for the salivary gland ductular cells5 may imply the existence of some preferential sites of neoplasia.

We report here the clinical and virologic features of 150 patients with B-cell–type NHL (71 men) consecutively seen at either a hematology or internal medicine outpatient clinic at our university hospital. Only subjects at first diagnosis of lymphoma were included in the study. Biopsy material was classified according to the Working Formulation for Clinical Usage.6 The NHL grade was low, intermediate, or high in 37%, 49%, and 14%, respectively. All patients underwent clinical examination, routine blood analysis, chest x-ray or CT chest scan, abdominal ultrasonography and/or abdominal CT scan, and bone marrow biopsy. They were classified according to the Ann Arbor staging system7 and 27% of patients were stage I, 33% were stage II, 18% were stage III, and 21% were stage IV. One hundred and fifty healthy, anti-HCV–negative, age- and sex-matched subjects and 80 Hodgkin's lymphoma (HL) patients were also studied as controls. Twenty-six NHL patients were anti-HCV positive (RIBA II; Chiron, Emeryville, CA), 34 were HCVRNA positive using one-tube nested polymerase chain reaction (PCR),1 8 and 37 (25%) were anti-HCV and/or HCVRNA positive. The latter prevalence was significantly higher than that observed in healthy subjects (25% v 1%) and HL patients (25% v 8%; P < .01) as well as in the general Italian population (approximately 1.3%).

Samples of peripheral blood mononuclear cells (PBMC) were also tested by HCV PCR using previously described methods.1,9 HCV RNA sequences were detected in PBMC in 30 of 34 (90%) serum HCVRNA-positive, in 3 of 15 (20%) serum HCVRNA-negative (2 of them also being anti-HCV–negative) NHL patients and also in neoplastic lymph node and bone marrow samples in 5 cases in whom this material was available. The results of HCV genotyping, performed with previously described methods in serum and/or PBMC samples,10 11 are shown in Table 1. In particular, in 3 patients a mixed infection (1b + 2) was detected in PBMC and only HCV type 1b in serum. No significant differences were observed between HCV-positive and HCV-negative patients with respect to the extranodal sites of the neoplasia (Table 2) as well as to the range of age and the grade of lymphoma. In particular, among patients with low-grade, intermediate-grade, or high-grade NHL, HCV-positive subjects were 14 of 56 (25%), 18 of 73 (25%), and 5 of 21 (24%) respectively.

We previously showed the existence of a high prevalence of HCV infection in patients with both mixed cryoglobulinemia-associated and idiopathic B-cell–type NHL, suggesting that HCV may play a role in lymphomagenesis in humans.2,12 This study further confirms these previous results as well as the almost constant involvement of the lymphatic system by HCV infection. No specific or more probable extranodal tumoral localization were identified in HCV-positive patients when compared with non-HCV–associated forms of NHL, suggesting that systemic and not local factors may trigger malignant lymphoproliferation. In this respect, the possibility that direct infection of lymphatic cells, probably in association with extralymphatic HCV epitope stimulation, plays a role in lymphomagenesis ought to be taken into account and deserves further analysis.13 The demonstration of HCVRNA sequences in lymphoid cells from the great majority of HCV-positive NHL patients tested, even in the absence of serum HCV markers, is not sufficient to demonstrate the validity of this hypothesis. However, it does point in the same direction. The identification in lymphoid cells of genotypes not detectable in serum suggests clustering into extrahepatic sites of more lymphotropic types, which may have pathogenetic relevance. In synthesis, it is possible that HCV infection may induce a chronic B-cell proliferation that, in some cases, with the cooperation of host's and/or environmental cofactors, may dramatically affect the B lineage and favor the emergence of genetic changes necessary for malignant transformation. De Vita et al5 described the presence of viral sequences in some normal lymphatic cells and not in transformed ones. Because of the RNA nature of the HCV genome, integration of viral sequences in the host's cellular DNA is not possible. Consequently, it may be hypothesized that, even in the case of a neoplastic transformation initially related to HCV infection, resulting cells may no longer be permissive to HCV infection and replication.