Human leucocyte antigen and human T-cell lymphotropic virus type 1 associated diseases in Brazil.

Abstract

Although approximately 95% of the individuals infected with human T-cell lymphotropic virus type 1 (HTLV-I) remain as asymptomatic carriers (AC), there are at least two well-defined diseases associated with HTLV-I infection: adult T-cell leukaemia/lymphoma (ATL), and HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). AC status could represent an effective (ATL), a defective, and, in HAM/TSP, an aberrant immune response against the virus. Considering that HTLV-I-infected T-cells express a variety of T-cell epitopes that are recognized in the context of specific classes I and II human leucocyte antigen (HLA) molecules, HLA genes are attractive candidates for a role in the susceptibility to HTLV-I infection and related diseases. Several studies have been published, but there are still many controversies concerning association with HTLV-I and HLA. The objective of the present study was to investigate possible HLA-A, -B and -DR associations with HTLV-I infection and related diseases in the Brazilian population. The study was approved by the ethical committee. The HLA-A, -B and -DR specificities were determined by polymerase chain reaction with sequence-specific primers (PCR-SSP; One Lambda Inc.) in 71 HTLV-I-infected individuals from Brazil. For the analysis, the sample was stratified by ethnicity (White and Mestizo, an admixture of native Amerindian, Western Europeans and African descendents) and disease status: AC, ATL and HAM/TSP. The HLA phenotypic frequencies observed in these groups were compared (Fisher's exact test) with those determined in ethnically matched HTLV-I negative controls (68 white and 120 Mestizo individuals). Statistical significance was set at P < 0·05. HLA frequencies were not compared among the three groups of HTLV-I-infected individuals, because of the small the number of individuals in each category. A lower frequency of HLA-A2 was observed in white patients with ATL (15·0% vs. 47·0%), and a tendency (P < 0·09) to a lower HLA-A2 frequency was observed in HAM/TSP white patients (31·2% vs. 47·0%). A low frequency of HLA-A2 had already been reported in Japanese patients with HAM/TSP (Jeffery et al, 1999). Although it was not present in all racial groups, the negative association with HLA-A2 is of special interest considering that a highly dominant cytotoxic T-cell target antigen of HTLV-I contains a dominant A2-restricted epitope (Jeffery et al, 1999). Therefore, individuals lacking HLA-A2 would be expected to present a weaker cytotoxic response to HTLV-I, allowing the virus to reach a high-equilibrium provirus load. A positive association with HLA-A26 was observed in our sample of white ATL patients (35·0% vs. 10·3%), as was reported in the Japanese population, and is probably explained by the limited recognition of HTLV-I Tax peptide anchor motifs and epitopes capable of generating anti-HTLV-I Tax CD8(+) CTLs (Usuku et al, 1988; Sonoda et al, 1996; Yashiki et al, 2001). Concerning HLA class II antigens, our data show a positive association of HLA-DR8 with ATL, in both white and Mestizo patients (26·0% vs. 7·3% and 40·0% vs. 8·9% respectively). An association of DR8 with AC status was described in Africans (White et al, 1996) and in some Amerindian populations (Sonoda et al, 1996). The association of HLA-DR11 with HAM/TSP, observed only in the Mestizo patients (54·5% vs. 14·4%), has been previously described in Japanese patients (Usuku et al, 1988). In our Mestizo population, HLA-DR15 showed increased frequencies only in AC subjects (55·0% vs. 14·4%), in contrast to other studies that described association with ATL or HAM (Sonoda et al, 1996; Manns et al, 1998). In conclusion, the present investigation shows that several HLA molecules may influence the fate of HTLV-I infection and that these molecules are not necessarily the same in different ethnic groups, confirming the complexity of the influence of HLA genes upon HTLV-I infection.