CROI 2018 Abstract eBook

Abstract eBook

Oral Abstracts

39LB A RANDOMIZED CONTROLLED TRIAL OF HIGH-DOSE RIFAMPIN FOR PULMONARY TUBERCULOSIS Gustavo E. Velásquez 1 , Meredith B. Brooks 2 , Julia M. Coit 2 , Dante Vargas Vásquez 3 , Epifanio Sánchez Garavito 4 , Roger I. Calderón 5 , Judith Jiménez 5 , Karen Tintaya 5 , Charles A. Peloquin 6 , Elna Osso 2 , Dylan B. Tierney 1 , Kwonjune J. Seung 1 , Leonid Lecca 5 , Geraint R. Davies 7 , Carole D. Mitnick 2 1 Brigham and Women’s Hospital, Boston, MA, USA, 2 Harvard University, Boston, MA, USA, 3 Hospital Nacional Hipólito Unanue, Lima, Peru, 4 Hospital Nacional Sergio Bernales, Lima, Peru, 5 Partners in Health, Lima, Peru, 6 University of Florida, Gainesville, FL, USA, 7 University of Liverpool, Liverpool, UK Background: The standard of care for patients with pulmonary tuberculosis is a 6-month, 4-drug regimen that includes rifampin throughout. This blinded, randomized, controlled Phase II clinical trial (ClinicalTrials.gov NCT01408914) systematically examined the concept that increased rifampin doses could shorten standard therapy for tuberculosis and improve treatment outcomes without increased toxicity. Methods: We randomized 180 adults with new, smear-positive, drug- susceptible pulmonary tuberculosis in equal numbers to receive 10, 15, or 20 mg/kg/day of rifampin during the 8-week intensive phase. The primary endpoints were: [1] change in elimination rate of M. tuberculosis log 10 colony forming units (log 10 CFU) on 7H11 Middlebrook medium during the first 8 weeks of treatment (efficacy); and [2] frequency of grade 2 or higher rifampin-related adverse events occurring up to 4 weeks after intensive phase completion (safety). Safety was evaluated in the intention-to-treat (ITT, participants who received at least one dose of study medication) population. Efficacy was evaluated in the modified intention-to-treat (mITT, participants whose CFU data permitted sputum sterilization modeling) and per-protocol (PP, participants whose intensive phase rifampin dose was not altered by a protocol-defined study halt) populations. Results: The ITT, mITT, and PP analyses included 180 (100%), 174 (96.7%), and 132 (73.3%) participants, respectively. Each 5 mg/kg/day increase in rifampin dose resulted in differences of -0.011 (95% confidence interval [CI], -0.025 – +0.002; P=0.230) and -0.022 (95% CI, -0.046 – -0.002; P=0.022) log 10 CFU/ mL/day in the mITT and PP analyses, respectively. Faster count declines with rifampin AUC 0-6 (P=0.011) were borderline significant with AUC 0-6 /MIC 99.9 (P=0.053). The frequency of grade 2 or higher rifampin-related adverse events was similar across the three treatment arms: 26 (43.3%), 31 (51.7%), and 23 (38.3%) participants had at least one event (P=0.7092). The frequency of rifampin-related serious adverse events was also similar across arms: 1 (1.7%), 1 (1.7%), and 2 (3.3%) participants had at least one event (P=0.2679). Conclusion: This is the first controlled study to show both dose- and exposure- response of rifampin on sputum culture sterilization. Rifampin doses of up to 20 mg/kg/day were safe compared to the standard dose of 10 mg/kg/day, with similar frequencies of rifampin-related adverse events.

distribution of all rapidly growing clusters identified during the first 15 months of implementation. Methods: At quarterly intervals during December 31, 2015– December 31, 2016, we analyzed partial HIV-1 pol sequences reported to the NHSS for persons with HIV diagnosed during the 3 years prior. We calculated genetic distance for each pair of sequences and inferred clusters using a pairwise threshold of 0.005 substitutions/site. Rapidly growing clusters were defined as those with ≥ 5 diagnoses during the most recent 12-month period. We used node ages inferred by molecular clock phylogenetic analysis to estimate HIV transmission rates for these clusters and compared demographic characteristics and transmission category of persons in these clusters to other persons with HIV diagnosed during January 1, 2013 – December 31, 2016 who had sequences available. The Rao-Scott correction to the Pearson Chi-Square test was used to account for correlation between cases in the same cluster. Results: Among 51,750 persons with sequences analyzed, 60 rapidly growing transmission clusters were identified. Rapidly growing clusters were 5–42 persons in size, with transmission rates ranging from 21–132 transmission events/100 py (median: 44 per 100 py). Compared with the 50,847 persons not in rapidly growing clusters, the 903 persons in rapidly growing clusters were disproportionately young men who have sex with men (MSM) (61% vs 32% p<.0001), and particularly young Hispanic/Latino MSM (26% vs 10%, p<.0001). Clusters were identified in all regions of the country and involved 20 states. Conclusion: Routine surveillance for rapidly growing clusters consistently identifies clusters across the United States with transmission rates far exceeding the estimated national rate. These findings suggest rapid transmission in networks involving young MSM, especially young Hispanic MSM. Prioritizing these clusters for public health intervention may have increased potential to reduce future infections. 41 CLUSTER ANALYSIS REVEALS IMPORTANT SHIFT OF DRIVERS OF THE HIV EPIDEMIC IN SWISS MSM Nadine Bachmann 1 , Claus Kadelka 2 , Teja Turk 1 , Katharina Kusejko 1 , Jürg Böni 2 , Vincent Aubert 3 , Sabine Yerly 4 , Thomas Klimkait 5 , Huldrych F. Günthard 1 , Roger Kouyos 1 1 University Hospital Zurich, Zurich, Switzerland, 2 University of Zurich, Zurich, Switzerland, 3 Lausanne University Hospital, Lausanne, Switzerland, 4 University Hospitals of Geneva, Geneva, Switzerland, 5 University of Basel, Basel, Switzerland Background: Transmission clusters in phylogenetic trees constructed from densely sampled HIV-1 sequencing data reflect recent or ongoing transmission and can be used to identify local outbreaks. In this study, we aimed to illustrate methods to identify those phylogenetic clusters relevant for targeted prevention in a real-world clinical setting. We hypothesized that cluster growth in high-risk HIV-1 subpopulations can be predicted using a combination of phylogenetic methods, clinical and behavioral data. Methods: We used HIV-1 pol sequence data from the Swiss HIV Cohort Study (SHCS) and Los Alamos background sequences to construct eight phylogenetic trees including all patients enrolled in the SHCS by the end of the years 2007 to 2014. The SHCS is highly representative of the HIV epidemic in Switzerland and contains sequences for approximately 60% of all ~19’000 patients ever enrolled. We identified HIV-1 transmission clusters of Swiss MSM in eight consecutive years, assigned annual per-cluster infectivity scores as the fraction of cluster members who had a viral load measurement above 1’000 copies/ml and annual per-cluster risk scores as the fraction who reported condom-less sex with occasional partners, and then studied the cluster growth in the subsequent years. Results: Our analysis revealed that, over the course of the study, the infectivity score became less predictive of new infections within MSM clusters, while the risk score gained predictive power. We quantified the fraction of new infections within pre-existing transmission clusters and compared cluster characteristics of growing and non-growing clusters (Figure). Between 2008 and 2014, 35-65% of the newly infected MSM appeared within pre-existing MSM transmission clusters (p for linear trend=0.6). Uni- and multivariate Poisson regressions with per capita growth as outcome and infectivity and risk scores as dependent variables exhibited that infectivity significantly predicted the per capita growth of a cluster from 2007 to 2012, while the risk behavior was only a significant predictor in 2011, 2012 and 2014. Conclusion: Our results demonstrate the effectiveness of treatment as prevention but also highlight that in recent years there was an epidemiologically important shift from the diagnosed to the undiagnosed population as the driver

Oral Abstracts

40 RAPIDLY GROWING HIV TRANSMISSION CLUSTERS IN THE UNITED STATES, 2013–2016 Anne Marie France 1 , Nivedha Panneer 1 , Cheryl B. Ocfemia 1 , Neeraja Saduvala 2 , Ellsworth Campbell 1 , WilliamM. Switzer 1 , Joel O. Wertheim 3 , Alexandra M. Oster 1 1 CDC, Atlanta, GA, USA, 2 ICF International, Atlanta, GA, USA, 3 University of California San Diego, San Diego, CA, USA Background: In early 2016, CDC began routinely analyzing molecular sequence data reported to the National HIV Surveillance System (NHSS) to identify transmission clusters suggestive of recent and rapid growth in the United States. An assessment of the initial 13 clusters identified demonstrated transmission rates greater than the 4/100 person-years (py) estimated for the entire United States. Here, we assessed transmission rates, characteristics, and geographic

17

CROI 2018