A case-control study of hepatitis C virus infection and non-Hodgkin's lymphoma in Egypt.

摘要

A case-control study was conducted to investigate the association between non-Hodgkin's lymphoma and infection with hepatitis C virus (HCV) in a population of Egyptian adults. Between 1999 and 2001, I29 NHL cases and 18I controls with other cancers were recruited from a major cancer hospital in Cairo, Egypt. Controls were to be matched with cases by sex, age, and region of residence, but the matching scheme was relaxed to permit recruitment of an adequate number of controls, with the proviso that these variables were included in multivariate analyses. Each participant provided a blood sample and questionnaire data on residence, employment, and smoking histories; exposure to pesticides and other industrial or agricultural chemicals; and medical history. These factors were evaluated for confounding of the HCV-NHL association by statistically testing their associations with the main exposure (HCV) and the outcome (NHL), and potential confounders were included in the logistic regression along with sex, age, and region of residence. The prevalence of current HCV infection as measured by polymerase chain reaction (PCR) was 34.8% in cases and 21.5% in controls, corresponding to an unadjusted HCV-NHL odds ratio (OR) for the entire sample of 310 of 1.95 (95% confidence interval (CI) 1.18–3.23). The OR for the entire sample based on a multivariate logistic regression model adjusted for sex, age, region of residence (Upper vs. Lower Egypt), and place of birth (rural vs. urban) was 1.71 (95% CI 1.01–2.90). The association was confined to females (p for interaction = 0.04). Specifically, the OR for the 149 males, adjusted for sex, age, and current residence, was 1.08 (95% CI 0.52–2.24), but for the 16I females, further adjusting by place of birth and total number of deliveries, the final OR was 3.05 (95% CI 1.35–6.86). Breast cancer patients may have been over-sampled in the female control group, but a subgroup analysis showed that the OR in females remained elevated even after excluding breast cancer controls from analysis. Notwithstanding issues of selection bias and other potential limitations of the study, our data point to a real, positive association or HCV with NHL in females.