[Using epidemiology and bioinformatics for studying HTLV-1 infection-associated adult T-cell leukemia].

OBJECTIVES

Establishing a correlation between the number of HTLV-1 provirus and the characteristics of the genomic environment in ATL cases.

METHODOLOGY: A systematic search was made of publications as well as a meta-analysis of the pertinent literature considering proviruses per chromosome and structural and functional characteristics of flanking chromatin regions as variables. The concordance of experts' study was evaluated by Spearman Rho correlation. Publication bias was analysed by funnel plot and the Egger statisgrapher. A fixed effects model was applied according to heterogeneity evaluation to combine the results of integration occurring in coding sequences as well as coding sequences according to their molecular function.

RESULTS

The expert concepts' Kappa index was 0.7 and no publication bias was observed. The meta-analysis result was homogeneous (p>0.05). HTLV-1 integration was directed towards several chromosomes' telomeric and subtelomeric regions. The combination of published results in the articles which were analysed supported the hypothesis of integration events being site-directed towards coding regions of the human genome (p<0.05). Moreover, the groups of genes having enzymatic and receptor functions was statistically significant.

CONCLUSION

The results led to concluding that HTLV-I integration in the ATLL cases analysed here was not random but was directed towards regulatory regions. Such results could help to explain the role of some processes involved in leukemogenesis.