Abstract
E
In the context of African countries, the without situation is more frequently used and feasible than the with situation because of the limited resources. In addition, the specificity of African countries in terms of HIV-1 genetic diversity is rarely considered when developing treatment guidelines, although it can highly impact monitoring strategies and HIV treatment outcome. HIV-1 group O (HIV-1O) viruses represent one of the four known groups of HIV-1 that also includes HIV-1 groups M, N, and P. 3 The majority of identified cases of HIV-1O worldwide are detected in Cameroon and West-Central Africa, where it accounts for about 1% of all HIV-1 infections (≈10,000 persons), but other countries in Africa and in Europe have reported the presence of HIV-1O in lower proportions. 4,5 HIV-1O variants are genotypically divergent form HIV-1 group M viruses, and even within the O group, a significant genetic variation is observed, leading to important consequences in diagnosis and monitoring.
We recently reported inaccurate diagnosis of HIV-1O in Cameroon and other studies highlighted monitoring issues in terms of viral load testing as well, 6,7 but little is still known about the treatment outcome of HIV-1O-infected patients. Current antiretroviral (ARV) drugs are designed for the major group M, and in vitro and clinical evaluations rarely involve HIV-1O isolates. The use of nonnucleoside reverse transcriptase inhibitors (NNRTIs) is not recommended because of the natural resistance of most HIV-1O isolates to this drug class, 8 and extensive assessments of the virological response and resistance profiles to other drug classes including protease inhibitors (PIs), nucleosides reverse transcriptase inhibitors (NRTIs), and newly developed integrase and fusion inhibitors are still very limited. We report here a case of a multidrug-class-resistant HIV-1O-infected patient whom we monitored for 36 months with a salvage ARV regimen including integrase inhibitors (INIs).
The patient, a 45-year-old woman, was evaluated in February 2009 at the reference HIV treatment center of Yaoundé Central Hospital in Cameroon because of the suspicion of failing treatment. She was transferred from a peripheral treatment site where she initiated her treatment in February 2001 and reported previous exposure to lamivudine, stavudine, and nelfinavir. Her ARV regimen at the time of the visit included didanosine, nevirapine, and lopinavir/ritonavir. All previous treatment changes were based on clinical evaluations and biological data were not used. She presented with cutaneous lesions indicative of Kaposi sarcoma and reported no treatment discontinuation.
To thus evaluate her virological status, EDTA whole blood was collected for CD4 T cell count, viral load testing, and drug resistance evaluation. The viral load test was performed using the Abbott realtime PCR method (Abbott, Des Plaines, IL), and the viral load level was 7,539 copies/ml (3.88 log10 copies/ml), indicating virological failure. Initial genotyping attempts were unsuccessful using a home brew method optimized for non-B HIV-1 group M strains. Subsequent analyses with a discriminatory serotyping assay showed that the patient's virus was a group O HIV-1 strain, and we thus performed a specific group O RT-PCR in the viral pol region, covering the protease (PR) and reverse transcriptase (RT) regions using homebrew primers: outer primers HIVGrO-P3s 5′-GCYTTACAAGTRYTRAARGAAGTAATCAATGA-3′ and HIVGrO-P6as 5′-CTCTAGTRAYTGGCAGCTTRAATTTRGGYAAT-3′ and inner primers HIVGrO-P4s 5′-GAMCCAACAGGAAGTGACATTGCTGGRACAA-3′ and HIVGrO-P5as 5′-ATTCWGGAATCCAGGTGGCTTGCCAATA-3′. The obtained nucleotide sequence confirmed infection with HIV-1O virus and also showed a significant number of drug resistance mutations (DRMs) in the viral PR and RT regions.
As shown in Table 1, we observed DRMs to NRTIs that included M184V and thymidine analog mutations (TAMs) (D67N + K70R + K219Q), leading to high-level resistance to lamivudine/emtricitabine, zidovudine, and stavudine. Only one major DRM to NNRTIs was found, Y181C, but a few minor mutations (A98G+V106I) were identified as well, and together led to high-level resistance to this first generation drug class, nevirapine and efavirenz, and also possible resistance to newly developed molecules such as etravirine and rilpivirine. Observed protease resistance mutations included M46I, I54V, V82A, and L90M and several accessory mutations (L10V, I15V, K20R, M36I, Q58E, I62V, H69R, A71V, and L89I), leading to resistance to atazanavir, fosamprenavir, indinavir, lopinavir, nelfinavir, saquinavir, and tipranavir. As a result, only a few drugs were still found to be effective using the Agence Nationale de Recherches sur le Sida et les hépatites virales algorithm (
ANRS (Agence Nationale de Recherches sur le Sida et les hépatites virales) interpretation algorithm.
Black and gray boxes indicate drug resistance and possible drug resistance, respectively. Boxes with no color indicate effective drugs.
PIs, protease inhibitors; NRTIs, nucleoside reverse transcriptase inhibitors; NNRTIs, nonnucleoside reverse transcriptase inhibitors; DRPs, drug resistance positions.
We thus designed a salvage ART regimen that included tenofovir/lamivudine, darunavir/ritonavir, and raltegravir. The new treatment was accompanied by monthly clinical assessments and quarterly biological monitoring including CD4 T cells count and HIV-1 RNA quantification from February 2009 to February 2012. Two months after initiation of the new treatment, the CD4 T cell count increased to 298 cells/μl and the viral load dropped to <40 copies/ml. We replaced darunavir with atazanavir at month 10 because of darunavir stock-out in country, and that did not affect the effectiveness of the treatment. Overall, over the 36 months of follow-up, the median CD4 T cell value (interquartile interval) was 422 cells/μl (386–473), the viral load remained undetected, and all signs of Kaposi sarcoma disappeared.
The following results indicate that the salvage treatment worked well for this patient despite her background of resistance to almost all PIs, NRTIs, and NNRTIs. Although few studies have evaluated the susceptibility of HIV-1O viruses to PIs, NRTIs, NNRTIs, INIs, and fusion inhibitors, very little is known about in vivo susceptibility to these drugs, especially for treatment-experienced HIV-1O-infected patients. A study assessing the in vitro susceptibility of B and non-B HIV-1 strains to a DRV/r-based regimen showed good response of HIV-1O primary isolates, 9 but clinical data are rare. We here showed that despite previous exposure to several PIs and a selection of numerous PI DRMs, darunavir and atazanavir remained effective, probably because of their high genetic barrier. Recent in vitro data showed that raltegravir is effective against HIV-1O, 10 but clinical evaluations are still very limited. Our results demonstrated the effectiveness of a treatment combination that includes raltegravir, with a sustained immunological and virological response over 36 months.
This observation indicates that a raltegravir-based regimen can be successfully used for naive and treatment experienced HIV-1O-infected patients, although additional data including more patients are needed to confirm that finding. A few natural polymorphisms in the integrase region of HIV-1O, including V72I, L74I, S153A, E157Q, V201I, and T206S, which can be associated with resistance to raltegravir, have been reported, but with no impact on raltegravir in vitro efficacy. 10 Most HIV-1O isolates naturally carry the Y181C mutation, leading to high-level resistance to nevirapine and efavirenz, and also to a new generation of NNRTIs that includes etravirine and rilpivirine.
This is a major threat in the context of developing countries such as Cameroon, since neither viral identification nor baseline drug resistance evaluation is conducted prior to ART initiation. HIV-1O-infected patients are therefore treated according to WHO recommendations, i.e., two NRTIs plus one NNRTI as first-line ART, and stay on failing regimens for several years or simply die. Our results and previous reports indicate that adequate ARV regimens exist for HIV-1O-infected patients, 9 –12 but efforts are needed to improve ART initiation and monitoring in low-income settings, including uncommon HIV strains.
Footnotes
Acknowledgments
We thank all contributors who directly or indirectly participated in the successful completion of this study, especially the patient and all health care personnel. This study was supported by grants from the Institut de Recherche pour le Développement-France (IRD–UMI 233).
Author Disclosure Statement
No competing financial interests exist.