The genetic diversity of the HIV virus may impair immune responses to vaccine candidates, cause false-negative results in diagnostic/monitoring laboratory tests, and lead to differences in the disease progression. On the other hand, little is known about the genetic diversity of HIV and in vivo response to antiretrovirals. It has been demonstrated so far that HIV-2 is not susceptible to nonnucleoside reverse transcriptase inhibitors (NNRTIs) and presents little susceptibility to zidovudine. 1

One retrospective analysis of a prospective study demonstrated that clade F-infected individuals presented a poorer virologic response to antiretrovirals as compared to clade B. 2 It is interesting to note that clade F protease almost always presents the methionine in codon 89 followed by a leucine in codon 90 among protease inhibitor (PI)-exposed and PI-unexposed individuals. On the other hand, wild type clade B strains present leucine at codon 89 followed by a leucine at codon 90, whereas some resistant clade B strains present leucine at 89 followed by methionine at codon 90 (L90M). The sequence L89M-L90M is exceedingly rare among clade B or non-B strains (http://www. ncbi. nlm. nih. gov/Genbank/). It has also been demonstrated in vitro that the L89M polymorphism in clade F decreases the virus's susceptibility to PIs, presenting the same impact as the L90M mutation does for clade B. 3 In Brazil, HIV-1 clade B and clade F viruses cocirculate, with a significant proportion of B/F recombinants; 4 and the B/F circulating recombinant forms 28 and 29 present protease F. 5 Therefore, we attempted to prospectively compare the antiretroviral response among previously naive clade B-infected and clade F-infected individuals to a boosted PI-containing regimen in a longitudinal, open-label trial, conducted in 3 centers in Brazil.