Background:  HIV-1 group O (HIV-O), rare and highly divergent variants, are endemic in Cameroon. Management of infected patients is compromized by limitations of serological and viral load commercial assays and HIV-O natural resistance to NNRTI. As HIV serotyping is not routinely used in Cameroon, HIV-O identification is often belated, leading to inadequate treatment and selection of resistance. The objective of this work was to set up specific tools for diagnosis and monitoring of HIV-O-infected patients in Cameroon.
Methods:  Patients consulting at Centre Pasteur du Cameroon in Yaounde for HIV screening, based on 2 enzyme immunoassay (EIA) tests (AxSym Ag/Ac VIH Combo [Abbott] and Genscreen v.2 [BioRad]), or for viral load quantification were screened for HIV-O infection. Serotyping was performed with an EIA using gp41 and V3 antigens representative of HIV-1 groups M, N, and O and HIV-2. Plasma RNA quantification was realized with an in-house real-time polymerase chain reaction (RT-PCR) targeting the LTR gene. Samples with a detectable viral load were submitted to reverse transcriptase, protease and gp41 sequencing for resistance genotyping and phylogenetic analyses.
Results:  Between July 2006 and August 2007, 2.571 HIV⁺ and 4553 viral load samples were serotyped. 27 (1.3%) and 47 (1.0%) HIV-O infections were identified respectively, with seven (9,4%) M/O dual infections. In 2 HIV-O samples, Genscreen v.2 was not reactive. CD4 counts (when available) and viral load ranges were 23 to 872 cells/mm³ and 2.5 to 5.5 log₁₀ copies/mL, respectively. Among the 25 pol sequences available, 3 (27.2%) naive samples out of 11 harbored the 181C mutation conferring NNRTI resistance in HIV-M; 8 patients with documented HAART presented resistance mutations described for HIV-M, corresponding to the antiretrovirals used. Minor mutations leading to possible resistance to tipranavir (TPV) and saquinavir (SQV) according to the French National AIDS Trial Group (ANRS) algorithm were observed for 100% and 88% of the samples respectively. Strains cluster with phylogenetic clades A (majority), B, and C, with some unclassified.
Conclusions:  In our study, prevalence of HIV-O infections was low (1.3%) with nearly 10% of M/O dual infections. Monitoring of HIV-O infections can be easily done with specific tools. However, the high natural polymorphism of HIV-O pol gene underlines the need for an adapted resistance interpretations algorithm. Efficacy of NNRTI on 181Y strains must be studied, especially in Cameroon where nevirapine (NVP) is widely used in first-line regimen and therapeutic options remain limited.