Since the discovery of human immunodeficiency virus (HIV) as the causative agent in 1983, the HIV-1 pandemic is characterized by an extensive variability and the circulation of numerous genotypes and recombinant forms (CRFs)[1, 2]. Subtype C is the most prevalent, estimated to represent 48% infections in the HIV-1 pandemic[3]. Subtype C is predominant particularly in south and east Africa[4]. This strongly contrasts with the situation in central Africa where a marked genetic heterogeneity is observed.

HIV-1 remains a crucial global health problem. Delivery of ART in low and middle income countries has required utilization of a public health approach in which standardized, rather than individualized, regimens are prescribed to very large numbers of HIV-1-infected individuals[5]. At present, the majority of individuals in these countries are initiating first-line therapy with a non-nucleoside reverse transcriptase inhibitor (NNRTI) and two nucleoside reverse transcriptase inhibitors (NRTI)[6].

The introduction of antiretroviral (ARV) therapy in resource-poor settings is effective in suppressing HIV-1 replication and prolonging life of infected individuals. Nevertheless, patients in routine clinical care experience sustain detectable viral replication even under potent combination therapy. Various factors underlie this failure to achieve viral suppression, but resistance to current medications has been increasingly recognized as an important factor. Rational choice of the NRTI component of second-line therapy should be based on patterns of resistance developed during first-line therapy[7].

Djibouti is the smallest country in east Africa. The estimated prevalence of HIV-1 infection was around 2.5% (2009). Since the introduction of antiretroviral therapy in 2004, no data exist on the epidemiology of drug resistance mutations and little was known about HIV-1 genotypes circulating in Djibouti. Data on drug resistance in the routine care setting are urgently required to check ART efficacy. The present study aims to characterize circulating HIV-1 subtypes and to identify drug resistance mutations in patients who first-line therapy failed (>1000 copies/mL) and antiretroviral-naive patients from Djibouti.

The ART naïve patients showed only NNRTI-associated mutations which were K103N (2 strains) and K101Q (1). No mutations of NRTI were detected. The minor PI mutation L10V was detected in one ART naïve patient[8].

Drug susceptibility results were classified into three categories: susceptible, intermediate resistance and resistance. Of the 16 patients with first-line ART failure, 56.25% showed resistance to at least one antiretroviral molecule: two patients (12.5%) were resistant to nucleoside (NRTI), one (6.25%) to non-nucleoside (NNRTI) reverse transcriptase inhibitors, and six (37.5%) to both. We found that the resistance rates to Lamivudine and Nevirapine in the 16 patients with first-line ART failure were 37.5% and 43.75% respectively. Forty four percent (44%) of sequences from viral strains did not show any mutation associated to ARV drug resistance- 7out of 16[10]. Therefore, therapeutic solutions using second line drugs could be proposed to our patients with therapeutic failure and drug resistance mutations.

Minimizing drug resistance to highly active antiretroviral therapy (HAART) is important. First, it maximizes the opportunity for successful second-line and subsequent therapies after viral rebound during first-line treatment; second, it limits the transmission of drug-resistant viruses. We report results of the HIV-1 genetic diversity and prevalence of mutations associated with antiretroviral drugs among ART naïve as well as patients who failed first line therapy. The HAART was introduced in Djibouti in 2004. Djibouti, surrounded by Eritrea, Ethiopia, and Somalia, is an important crossroad for trade and exchange.

Our result revealed a majority of subtype C (~66%). In fact, Subtype C prevalence has been described in the Horn of Africa[11–14] and a similar prevalence was previously reported in Djibouti in 2005[15, 16]. However, our study describes the subtype B in Djibouti for the first time. It is the predominant subtype in the Western world and the clade that has circulated in North America and Europe. The detection of CRF02_AG strains indicates that they are still circulating in Djibouti, the only country in East Africa in which this recombinant virus was found some years ago[17, 18]. CRF02_AG recombinant isolates were primarily described in West and Central Africa[19, 20]. The discordances observed between RT and PR sequences in 2 isolates (CRF02_AG/C, K/C) are probably attributable to the high variability of the genome of HIV in particularly intra subtype recombination within group M subtype C[21].

For the ART naïve patients, the low rate of CD4 and the high rate of VL can be explain by the fact that 80% of patients infected by HIV in Djibouti was already detected at an advanced stage of the epidemic according to WHO classification[22, 23]. Among the 19 strains, 2 were resistant only to non-nucleoside reverse transcriptase inhibitors (NNRTIs) while one strain showed a minor mutation to PI. Detection of RT and protease mutations associated with drug resistance in individuals who have not been exposed to antiretroviral therapy is thought to result from either transmission from a treated individual.

Concerning first-line ART failures, 38% of 16 strains sequenced in this group had the M184V mutation. This mutation is selected by lamivudine (3TC), emtricitabine (FTC) or abacavir (ABC) use and confers high-level resistance to 3TC (Figure 1)[24]. Patients receiving incompletely suppressive 3TC regimens usually develop M184V as their first mutation[25, 26]. Accumulation of the other mutations observed included thymidine associated mutations (including M41L, D67N, K70R, L210W, T215Y/F, K219Q) results in increasing resistance to AZT, Tenofovir, D4T, Abacavir, and DDI[27, 28]. Two multidrug resistance mutation profiles in general, confer high level drug resistance to all NRTIs. The most common occurring nucleotide excision mutations (NEMs) that confer resistance by enhanced excision[29] were at position 215. The T215Y mutation was observed in 13% of the sequences while T215F was detected in 6%. Clinical studies have shown that the NEMs, particularly mutations at position 215, interfere with the clinical response to zidovudine, stavudine, abacavir, didanosine, and most dual NRTI combinations[30]. Mutations conferring resistance to NNRTIs were also seen, with one strain showing resistance only to NNRTIs. The NNRTI resistance-conferring mutations included K103N (25%), Y188L (12.5%), Y181C (6%), G190A (6%), H221Y (6%), M230L (6%). Our data suggest the non-epidemic circulation of resistant viruses, and the absence of multi-class drug resistant viral strains.

It is important to appreciate that resistance at failure is only one factor to consider when choosing an initial HAART regimen. Coformulation, simplicity of administration, price, drug interactions (particularly with tuberculosis therapy), toxicity and adverse events are all important considerations and will differ between patient populations. We noticed seven patients (~44%) did not show resistance-conferring mutations and the therapeutic failure may have been due to other factors[31–35]. Information on treatment from their clinic folders allowed us to confirm that for five of these seven cases the ART failure profile was due to poor adherence and for the remaining two cases the detectable viral level was due to coinfection with tuberculosis and malnutrition[36–38].

The escalating rates of transmission of drug resistant virus observed in the past few years, coupled with the poorer response to treatment in persons with drug resistant virus, are being considered as a basis for a recommendation that resistance tests should be performed routinely for persons with new HIV infection[39]. This evaluation, both at the baseline and in follow-up, will be crucial to assess the impact of therapy regimens on resistance development and also their adequation.

This study reveals primary mutations that confer resistance to NRTI, NNRTI in HIV-1 non-B and B subtype strains, isolated from ART naïve as well as patients who failed first line therapy in Djibouti. It also illustrates the importance of preliminary surveys of drug-resistance mutations in resource-poor countries. Therefore, we suggest the use of resistance testing to check the prevalence of the drug resistance mutation that arises following failure of the first line regimen. This will help in establishing guidelines for second line regimens in Djibouti. This allows the selection of adapted antiretroviral regimens in these regions. Our results provide a basis for repeated epidemiological studies to measure the population effects of HIV treatment programs over time.