Effect of human T-lymphotropic virus type I infection on non-Hodgkin's lymphoma incidence.

BACKGROUND: We previously reported from a case-control analysis that T-cell non-Hodgkin's lymphoma (NHL) was strongly associated with human T-lymphotropic virus type I (HTLV-I) infection in Jamaica and Trinidad and that the relative risk for HTLV-I infection was very high in younger patients. PURPOSE: The objective of this study was to estimate the age-specific incidence rates of NHL among HTLV-I-infected and HTLV-I-uninfected adults in Jamaica and Trinidad. METHODS: Population rates of HTLV-I infection were calculated from available census reports and serosurvey data. Incidence rates for NHL were calculated from all incident cases in Jamaica during 1984-1987 (n = 135) and from all incident cases in Trinidad during 1986-1990 (n = 117). Using biopsy material, we determined whether the immunophenotype of the tumor cells was T cell, B cell, or other. NHL incidence rates were computed according to HTLV-I status, age, sex, and tumor phenotype for each country separately and for both countries combined by weighting to the relative population size of each country. RESULTS: The age-standardized NHL incidence rate (mean +/- SE) in Jamaica was 1.9 +/- 0.2 per 100,000 person-years (PY). In Trinidad, the rate was 2.9 +/- 0.4 per 100,000 PY. Overall, the incidence of NHL increased with age and was higher in males than in females. In the HTLV-I-infected population, the incidence of NHL was inversely related to age, and age-specific rates were higher in males than in females. The NHL incidence in those estimated to have acquired HTLV-I infection in childhood, however, showed no sex difference, and one in 1300 such carriers (95% confidence interval: one in 1100 to one in 1600) per annum were estimated to be at such risk. For T-cell NHL, as proxy for adult T-cell lymphoma/leukemia, incidence was highest in those patients infected with HTLV-I early in life (perinatally or via breast milk), with high, sustained risk from early adulthood in both sexes. CONCLUSIONS: While overall NHL incidence rates reveal that HTLV-I endemicity does not impose an exaggerated lymphoma burden on these populations, the risk for lymphoma among carriers who acquire infection early in life is dramatic and is consistent with the hypothesis that virus exposure early in life is most important for lymphoma-genesis. IMPLICATIONS: Studies of HTLV-I carriers known to be infected in childhood may provide insight into markers intermediate in the lympho-magnetic process. Strategies to disrupt early-life transmission of HTLV-I, notably mother-infant transmission, may be critical in reducing the burden of lymphoreticular disease in these populations.

Role of HTLV-I in development of non-Hodgkin lymphoma in Jamaica and Trinidad and Tobago. The HTLV Lymphoma Study Group.

Human T-cell lymphotropic virus type I (HTLV-I) has been implicated in the aetiology of adult T-cell leukaemia/lymphoma in Japan and elsewhere, particularly the Caribbean. We have carried out parallel case-control studies in Jamaica and in Trinidad and Tobago to quantify the role of HTLV-I in the development of non-Hodgkin lymphoma (NHL). 135 cases of NHL were enrolled in Jamaica and 104 in Trinidad and Tobago. Controls were selected from patients treated in the same wards or clinics at the same time as the cases. Overall, patients with NHL were 10 times more likely than were controls to be seropositive for HTLV-I (Jamaica odds ratio 10.3 [95% CI 6.0-18.0], Trinidad and Tobago 14.4 [7.6-27.2]). In both countries the association between NHL and HTLV-I was greatest for T-cell lymphomas (18.3 [9.5-35.6] and 63.3 [25-167]). Among T-cell lymphomas especially, there was no significant difference between men and women in the association between NHL and HTLV-I, but there was a significant inverse relation between age and likelihood of HTLV-I seropositivity. B-cell lymphomas were predominant in the older age groups and were not associated with HTLV-I seropositivity. These findings are consistent with the hypothesis that early life exposure to HTLV-I is important for risk of subsequent ATL. Prevention of vertical transmission of HTLV-I could reduce by 70-80% cases of NHL in people under 60 years in this region.

Determinants of the geographic variation of invasive cervical cancer in Costa Rica.

The incidence of cervical cancer in Costa Rica is about twice as high in the coastal regions as in the interior. To study these regional variations, we used data from a 1986-1987 case-control study of 192 Costa Rican women with invasive cervical cancer and 372 controls. Risk factors identified included the following: The study participant's (1) number of sexual partners, (2) age at first sexual intercourse, (3) number of live births, (4) presence of type 16/18 human papillomavirus (HPV) DNA, (5) venereal disease (VD) history, (6) Pap smear history, and (7) socioeconomic status. The adjusted relative risks (RR) and 95% confidence intervals (CI) for each of these risk factors were as follows: (1) > or = 4 vs. 1 sexual partner: RR = 2.0, 95% CI = 1.1-3.5; (2) age of initiation < or = 15 vs. > or = 18 years: RR = 1.5, 95% CI = 0.9-2.5; (3) > or = 6 vs. < or = 1 live birth: RR = 1.7, 95% CI = 0.7-3.9; (4) HPV 16/18 DNA in cervix: RR = 2.8, 95% CI = 1.9-4.2; (5) VD history: RR = 2.2, 95% CI = 1.2-4.0; (6) no Pap smear: RR = 2.4, 95% CI = 1.5-3.8; and (7) low socioeconomic status: RR = 2.0, 95% CI = 1.2-3.2. The population-attributable risks related to HPV detection, four or more sexual partners, six or more live births, no prior Pap smear, and low socioeconomic status were 39%, 38%, 29%, 23%, and 22%, respectively. Several of the sexual and reproductive risk factors were relatively more prevalent in the high-risk region, but Pap screening and detection of HPV were equally prevalent in the high-risk and low-risk regions. Though differences in screening quality (laboratory and follow-up) may have been involved, we conclude that the observed regional differences reflect behavioral more than screening differences. This suggests that screening programs should be more aggressive in the high-risk area, given the more frequent occurrence of the disease there. Failure to detect a higher prevalence of HPV in the high-risk region could reflect weaknesses in the in situ hybridization test employed. Alternatively, cofactors may have to be present in order for HPV to exert its role in cervical carcinogenesis.