We are pleased with the responses to our review [1], and we thank the readers for bringing to our attention several conditions, some of which may eventually be established as causally related to HCV infection. These reports testify to the many clinical conditions that appear to be associated with HCV infection. The evidence for the association of autoimmune thrombocytopenia, polyarteritis nodosa, and progressive multifocal leukoencephalopathy, however, remains preliminary. Thus far, the data have appeared in isolated case reports, letters to the editor, or abstracts. Still other data are unconvincing. Although these conditions should be considered for their potential association with HCV infection, more evidence is needed before routine testing for HCV can be recommended in patients with these conditions.

We do agree that some of the data on B-cell non-Hodgkin lymphoma support an association between this neoplastic disorder and HCV infection. In a carefully controlled study, Ferri and colleagues [2] determined the prevalence of HCV infection in a series of 50 unselected patients with B-cell non-Hodgkin lymphoma. They noted a 32% prevalence by HCV RNA compared with a prevalence of 0% in age-matched, healthy controls and a 3% seropositivity in controls with Hodgkin disease. Ten of the 50 patients with non-Hodgkin lymphoma did report previous exposure to blood products, but they were equally distributed among patients with and without HCV markers. An 8% prevalence of cryoglobulinemia was found in these patients. In an uncontrolled retrospective study, Pozzato and colleagues [3] found a 38.7% prevalence of low-grade non-Hodgkin lymphoma by bone marrow biopsy in 31 selected patients with known mixed cryoglobulinemia who had been followed for at least 10 years. Hepatitis C virus RNA was documented in 96% of these patients. Two of these patients, who received interferon, had resolution of cryoglobulinemia and remained asymptomatic. Although the HCV genome has been found in peripheral mononuclear cells [4], no attempt has been made to document HCV RNA in the lymphoid aggregates of the bone marrow or the lymph nodes [2, 3]. A direct role for HCV in non-Hodgkin lymphoma, therefore, cannot yet be determined. Although it is reasonable to test for HCV infection in patients with non-Hodgkin lymphoma, a routine bone marrow biopsy in all patients with chronic HCV infection or mixed cryoglobulinemia is not currently indicated.