High HIV-1 Virological Failure and Drug Resistance among Adult Patients Receiving First-Line ART for At least 12 Months at a Decentralized Urban HIV Clinic Setting in Senegal before the Test-and-Treat.

BACKGROUND: The feasibility of antiretroviral therapy (ART) monitoring remains problematic in decentralized HIV clinic settings of sub-Saharan Africa. We assessed the rates and correlates of HIV-1 virological failure (VF) and drug resistance (DR) in 2 pre-test-and-treat urban clinic settings of Senegal. METHODS: Consenting HIV-1-infected adults (⩾18 years) receiving first-line ART for ⩾12 months were cross-sectionally enrolled between January and March 2015, at the referral outpatient treatment center of Dakar (n = 151) and decentralized regional hospital of Saint-Louis (n = 127). In the 12 months preceding plasma specimens' collection patients at Saint-Louis had no viral load (VL) testing. Significant predictors of VF (VL ⩾ 1000 copies/ml) and DR (clinically relevant mutations) were determined using binomial logistic regression in R software. RESULTS: Of the 278 adults on EFV-/NVP-based regimens, 32 (11.5% [95%CI: 8.0-15.9]) experienced VF. Failing and non-failing patients had comparable median time [interquartile] on ART (69.5 [23.0-89.5] vs 64.0 [34.0-99.0] months; P = .46, Mann-Whitney U-test). Of the 27 viraemic isolates successfully genotyped, 20 (74.1%) carried DR mutations; most frequent were M184VI (55.6%), K103N (37.1%), thymidine analog mutations (29.6%), Y181CY (22.2%). The pattern of mutations did not always correspond to the ongoing treatment. The adjusted odds of VF was significantly associated with the decentralized clinic site (P < .001) and CD4 < 350 cells/mm3 (P < .006). Strong correlates of DR also included Saint-Louis (P < .009), CD4 < 350 cells/mm3 (P <. 001), and nevirapine-based therapies (comparator: efavirenz-based therapies; P < .027). In stratification analyses by site, higher rate of VF at Saint-Louis (20.5% [95%CI: 13.8-28.5] vs 4.0% [95%CI: 1.5-8.5] in Dakar) was associated with nevirapine-based therapies (OR = 3.34 [1.07-11.75], P = .038), self-reported missing doses (OR = 3.30 [1.13-10.24], P = .029), and medical appointments (OR = 2.91 [1.05-8.47], P = .039) in the last 1 and 12 months(s), respectively. The higher rate of DR at Saint-Louis (12.9% [95%CI: 7.6-20.1] vs 2.7% [95%CI: 0.7-6.7] in Dakar) was associated with nevirapine-based therapies (OR = 5.13 [1.12-37.35], P = .035). CONCLUSION: At decentralized urban settings, there is need for enhanced virological monitoring and adherence support. HIV programs in Senegal should intensify early HIV diagnosis for effective test-and-treat. These interventions, in addition to the superiority of efavirenz-based therapies provide a favorable framework for transitioning to the recommended potent drug dolutegravir, thereby ensuring its long-term use.

A machine learning approach for viral genome classification.

BACKGROUND: Advances in cloning and sequencing technology are yielding a massive number of viral genomes. The classification and annotation of these genomes constitute important assets in the discovery of genomic variability, taxonomic characteristics and disease mechanisms. Existing classification methods are often designed for specific well-studied family of viruses. Thus, the viral comparative genomic studies could benefit from more generic, fast and accurate tools for classifying and typing newly sequenced strains of diverse virus families. RESULTS: Here, we introduce a virus classification platform, CASTOR, based on machine learning methods. CASTOR is inspired by a well-known technique in molecular biology: restriction fragment length polymorphism (RFLP). It simulates, in silico, the restriction digestion of genomic material by different enzymes into fragments. It uses two metrics to construct feature vectors for machine learning algorithms in the classification step. We benchmark CASTOR for the classification of distinct datasets of human papillomaviruses (HPV), hepatitis B viruses (HBV) and human immunodeficiency viruses type 1 (HIV-1). Results reveal true positive rates of 99%, 99% and 98% for HPV Alpha species, HBV genotyping and HIV-1 M subtyping, respectively. Furthermore, CASTOR shows a competitive performance compared to well-known HIV-1 specific classifiers (REGA and COMET) on whole genomes and pol fragments. CONCLUSION: The performance of CASTOR, its genericity and robustness could permit to perform novel and accurate large scale virus studies. The CASTOR web platform provides an open access, collaborative and reproducible machine learning classifiers. CASTOR can be accessed at http://castor.bioinfo.uqam.ca .