CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Methods: ART-experienced patients on first INSTI 2007-2016 were followed from INSTI initiation (baseline) until: death, loss to follow-up (LTFU-1 year without clinical visit), or August 2017. Time to CD4>500 was estimated among patients with baseline CD4<500 censoring patients at virologic failure, stratified by nadir CD4 and baseline VL. Virologic failure (VF) was defined as first of 2 consecutive viral loads (VL) >200 copies/mL >2 weeks apart, or 1 VL>200 before LTFU, 24 weeks post-baseline. Kaplan-Meier curves were fit and Cox proportional hazard models estimated hazard ratios (HR), adjusting for baseline age, race, sexual risk group, CD4, and VL. Patients were stratified by VL Results: The 777 patients were 32% female, 43%MSM, and 59% African American, with baseline median age of 47 years (IQR 38, 54), CD4 509 (274, 738), and prior exposure to 6 (4, 8) antiretrovirals. At baseline, 328 (42%) had VL>50, with median VL 4.2 log 10 copies/mL (3.0, 4.9). Among 247 patients with CD4<500 and VL>50 at baseline, patients with nadir CD4>200 had a shorter median time to CD4>500 (1.3 vs. 4.3 years, P<0.01). Among patients with VL<50 at INSTI start, 2 and 4% experienced VF after 1 and 2 years, compared to 23% and 34%, among patients with VL>50 (respectively, P<0.01). Time to VF differed by INSTI agent among patients with and without VL suppression at baseline (Figure 1 A/B, both P<0.05). In patients with baseline VL>50, time to VF was longer for raltegravir vs elvitegravir (EVG) (aHR 0.30, 95% CI, 0.14, 0.64). In patients with baseline VL<50, time to VF was longer for dolutegravir (DTG) vs EVG (aHR 0.13, 95% CI, 0.02, 0.97). Prior exposure to ≥8 drugs increased risk of VF among patients with baseline VL>50 (aHR 1.63, 1.12, 2.38). Conclusion: In suppressed patients who switch to INSTI-based therapy, INSTI regimens are highly effective, with a possible lower risk of VF with DTG-based therapy. Patients with viremia at INSTI start have high VF rates, associated with prior exposure to many drugs. Low nadir CD4 is associated with worse CD4 recovery while on INSTI therapy.

510 VIROLOGIC RESPONSE TO 2-DRUG ART REGIMENS AMONG TREATMENT- EXPERIENCED HIV+ PATIENTS Gerald Pierone 1 , Cassidy Henegar 2 , Jennifer Fusco 2 , Vani Vannappagari 3 , Michael Aboud 4 , Leigh Ragone 3 , Gregory Fusco 2 1 Whole Family Health Center, Vero Beach, FL, USA, 2 Epividian, Durham, NC, USA, 3 ViiV Healthcare, Research Triangle Park, NC, USA, 4 ViiV Healthcare, London, UK Background: Drug-sparing regimens have the potential to reduce complexity, toxicity, and cost of antiretroviral therapy (ART). Our objectives were to describe two-drug regimen (2-DR) use among ART-experienced HIV+ patients in a large clinical cohort, and to compare virologic outcomes of 2-DRs and three-drug regimens (3-DRs) following switch during the study period. Methods: Between 1/1/2010 and 6/30/2016, ART-experienced patients starting a new 2-DR or 3-DR, were selected from the OPERA cohort. Patients were observed from regimen start date (baseline) until regimen discontinuation (d/c), loss to follow-up, death, or study end (6/30/2017). Outcomes were stratified by viral load (VL) at baseline (switched while viremic: ≥50 copies/mL; stable switch: <50 copies/mL). Suppression during follow-up was defined as a VL <50 copies/mL; failure following suppression was defined as 2 consecutive VLs >200 copies/mL or a VL>200 copies/mL + d/c. Cox models for each outcome were fit to estimate adjusted hazard ratios (aHRs). Results: We identified 10,190 ART-experienced patients who switched during the study period; 1,337 (13%) switched to a 2-DR, and 8,853 (87%) to a 3-DR. At baseline, 2-DR patients were older, more likely to have AIDS, had been on ART longer and experienced more treatment lines, had more comorbidities, and were less likely to be a stable switch compared to 3-DRs (p<0.0001). The most common 2-DRs (55%) comprised a protease inhibitor and an integrase strand transfer inhibitor combination, regardless of baseline VL. Among those switching while viremic (2-DR: 612 (15%), 3-DR: 3566 (85%)), suppression during follow-up was comparable among patients on 2-DRs (61%) and 3-DRs (67%; aHR 1.00, 95% CI 0.88, 1.13) [Figure]. After achieving suppression during follow-up, 13% of 2-DR and 15% of 3-DR patients went on to experience a failure event. Among stable switch patients (2-DR: 723 (12%), 3-DR: 5285 (88%)), the difference in risk of virologic failure during follow-up was not statistically significant between 2-DR and 3-DR patients (10% vs. 11%; aHR 1.15, 95% CI 0.90, 1.48) [Figure]. Conclusion: Virologic outcomes were comparable between ART-experienced patients switching to two- and three-drug regimens, regardless of whether patients were virologically controlled at switch. These findings support the continued evaluation of 2-DRs in clinical trials and real-world settings. Long- term outcomes require further assessment.

Poster Abstracts

511 TREATING MULTICLASS-RESISTANT HIV+PATIENTS IN RWANDA USING A PUBLIC HEALTH APPROACH Sabin Nsanzimana 1 , Muhayimpundu Ribakare 2 , Remera Eric 1 , Amitabh B. Suthar 3 , Riedel J. David 4 , Edward J. Mills 5 , Heiner C. Bucher 6 1 Rwanda Biomedical Centre, Kigali, Rwanda, 2 Swiss Tropical and Public Health Institute, Basel, Switzerland, 3 Stellenbosch University, Stellenbosch, South Africa, 4 University of Maryland, Baltimore, MD, USA, 5 McMaster University, Hamilton, ON, Canada, 6 University Hospital Basel, Basel, Switzerland Background: While the vast majority (≥90%) of the 19 million individuals currently receiving ART are on first- and second-line regimens, the number of people requiring third-line regimens is rising. Most developing countries do not currently provide third-line ART because of limited resistance testing, alternative treatment options and high cost. We report outcomes from

CROI 2018 185