Objective: To describe the distribution of HIV-1 subtypes in two cities wit
h high HIV prevalence (Kisumu, Kenya and Ndola, Zambia) and two with relati
vely low prevalence (Cotonou, Benin and Yaounde, Cameroon), and to examine
whether the differences in prevalence of HIV infection could be due to the
predominance within the infected populations of subtypes with differing eff
iciency of heterosexual transmission.
Methods: For around 100 randomly selected HIV-positive sera from the genera
l population and 60 from sex workers in each city, the HIV-1 subtype was de
termined in the env fragment. For between 19 and 52 of the sera from the ge
neral population and 20-32 sera from sex workers, the subtype was also dete
rmined in the gag fragment.
Results: Over 70% of infections in Cotonou, Yaounde and Kisumu were with su
btype A (by env). However, around one-half of subtype A infections in Coton
ou and Yaounde were found to be the circulating recombinant form CRF02(-)AG
when the gag fragment was also examined. A large number of different HIV s
trains were found in Yaounde, including some belonging to group O. Over 20%
of infections in Kisumu and around 10% in Yaounde were with isolated inter
subtype recombinant forms. All but a few infections in Ndola were with subt
ype C and no recombinants were found.
Conclusions: The pattern of distribution of subtypes that we found does not
suggest that differences in circulating subtypes play a major role in expl
aining the differences in prevalence of HIV-1 infection between the four ci
ties. The emergence and spread of recombinants requires close surveillance
to adapt testing strategies if needed, to inform vaccine development and to
ascertain their role in the future spread of HIV. (C) 2001 Lippincott Will
iams & Wilkins.