Viral diversity in some immunodominant epitopes: possible implications for retroviral immunopathogenesis.

it is intuitively more appealing in such a situation to use coinfection as the origin from which time to KS is measured, a complete analysis that uses both origins (i.e., HHV-8 infection and coinfection) provides the elements to fully characterize how HHV-8, HIV, and HIV-related immunosuppression interact for the development of KS. In our publication [2], we presented both analyses and showed that HHV-8 alone carried a negligible risk for KS and that, after dual infection, those who acquired HHV-8 after HIV had an elevated hazard for KS (relative ). Alhazard p 1.75 though this association did not reach nominal statistical significance, it confirmed the finding of Renwick et al. [4] but did not correspond to the simplistic example of the 2 persons whose experience demonstrates a relative hazard of 1, as provided by Cannon and Pellet [1]. Of more importance, the elevated relative hazard of persons infected with HHV-8 after HIV is reinforced by the effect of measures of immunosuppression (CD4 cell count) and viral replication (HIV RNA) on the risk of KS, as we documented [2]. We welcome the consensus of Cannon and Pellett on our primary inference; however, we believe that, for prevention of KS, the most important aspect is not to become infected with HIV if one is infected with HHV-8. If one does, it is imperative to treat the HIV infection and to maintain a competent immune system.