CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

1 Univ of California San Francisco, San Francisco, CA, USA, 2 Stellenbosch Univ, Tygerberg, South Africa, 3 Translational Genomics Rsr Inst, Flagstaff, AZ, USA, 4 Univ of California San Diego, San Diego, CA, USA Background: Tuberculosis (TB) accounts for one in three deaths among persons living with HIV, and multidrug-resistant TB (MDR-TB) imposes major human and programmatic costs. Resistance to fluoroquinolones (FQ) and second-line injectable (SLI) medicines, the backbone of most MDR-TB regimens, occurs during treatment in up to 15% of patients. Yet, little attention has been given to the evolutionary events occurring before development of phenotypic resistance. We hypothesized that high-resolution quantification of “micro-heteroresistance” (resistant subpopulations <5% of the total M. tuberculosis (M.tb) population, beneath the threshold of current commercial molecular TB assays) would predict later acquisition of full FQ and SLI resistance. Methods: Next-generation sequencing (NGS) was performed on 145 serial MGIT 960 isolates obtained from 19 patients with MDR-TB acquiring additional resistance over 3-78 months in South Africa, 2008-2015. Single Molecule-Overlapping Reads (SMOR) is a targeted NGS approach able to reduce sequencing error by orders of magnitude, from~1% for a high GC content organism to an M.tb–specific resolution of ~0.01%. We used SMOR to monitor proportions of single-nucleotide polymorphisms (SNPs) at specific loci within three gene regions corresponding to FQ (gyrA) and SLI (rrs, eis) resistance. Phenotypic resistance was defined by Middlebrook 7H11 indirect proportion method with WHO-recommended critical concentrations. Results: At each serial time point, 1-10 resistance-conferring mutations were transiently detected in each target region in the same sputum specimen, though a single mutant ultimately became dominant in the majority (95%) of cases corresponding to acquisition of phenotypic resistance. Based on an average sequencing depth of 55,000X, 139 (39%) and 90 (25%) of 355 total SNPs detected within the target regions occurred at frequencies below 5% and 1%, respectively. Micro-heteroresistance (<5%) to FQs (gyrA) and SLIs (rrs, eis) was detected among 26% (5/19) and 42% (8/19) of patients, respectively, a median of 6 months prior to sputum sample collection for which phenotypic resistance was ultimately confirmed (Figure). Conclusion: Small, previously undetectable M.tb subpopulations may be early precursors to phenotypic resistance. Detection and monitoring of micro-heteroresistance with tabletop NGS platforms could transform clinical management through early individualized treatment regimens and prompt reassessment of ineffective treatments.

Poster and Themed Discussion Abstracts

716LB HUMAN AND MOUSE HEMATOPOIETIC STEM CELLS ARE A DEPOT FOR DORMANT M. TUBERCULOSIS Julia Tornack 1 , Stephen Reece 1 , Wolfgang Bauer 2 , Alexis Vogelzang 1 , Silke Bandermann 1 , Ulrike Zedler 1 , Georg Stingl 2 , Stefan Kaufmann 1 , Fritz Melchers 1 1 Max Planck Inst for Infection Bio, Berlin, Germany, 2 Med Univ of Vienna, Vienna, Austria Background: An estimated third of the world’s population is latently infected with Mycobacterium tuberculosis (Mtb), with no clinical signs of tuberculosis (TB), but lifelong risk of reactivation to active disease. The niches of persisting bacteria during latent TB infection remain unclear. We tested the hypothesis that in LTBI Mtb bacteria acquire a non- replicating state inside resting, long-term repopulating pluripotent hematopoietic stem cells (LT-pHSCs) Methods: qPCR was used on peripheral blood pHSC isolated from patients with LTBI to detect Mtb sequences MPB64 and IS6110. qPCR and CFU assays were also used to detect Mtb sequences in different organs as well as pHSC in a mouse model of latent TB infection. Cytospins were used to visualise Mtb in pHSC. RT-PCR was used to detect the expression of Mtb dormancy genes in human and mouse pHSC. Intratracheal administration of Mtb-infected pHSCs was employed to demonstrate reactivation of TB in mice. Results: We detect Mtb DNA in peripheral blood selectively in long-term repopulating pluripotent hematopoietic stem cells (LT-pHSCs) as well as in mesenchymal stem cells from latently infected human donors. In mice infected with low numbers of Mtb, that do not develop active disease we, again, find LT-pHSCs selectively infected with Mtb. In human and mouse LT-pHSCs Mtb are stressed or dormant, non-replicating bacteria. Intratracheal injection of Mtb-infected human and mouse LT-pHSCs into immune-deficient mice resuscitates Mtb to replicating bacteria within the lung, accompanied by signs of active infection. Conclusion: We conclude that LT-pHSCs, together with MSCs of Mtb-infected humans and mice serve as a hitherto unappreciated quiescent cellular depot for Mtb during latent TB infection. 717 RISK FACTORS FOR DEATH IN ADVANCED HIV-INFECTED ADULTS IN AN EMPIRIC TB THERAPY TRIAL Gregory Bisson 1 , Ritesh Ramchandani 2 , Sachiko Miyahara 2 , Xin Sun 2 , Jing Bao 3 , Amita Gupta 4 , Mina C. Hosseinipour 5 , Johnstone Kumwenda 6 , for the Adult AIDS ClinicalTrials Group A5274 (REMEMBER) StudyTeam 1 Univ of Pennsylvania, Philadelphia, PA, USA, 2 Harvard Univ, Boston, MA, USA, 3 NIH, Bethesda, MD, USA, 4 Johns Hopkins Univ, Baltimore, MD, USA, 5 Univ of North Carolina Proj–Malawi, Lilongwe, Malawi, 6 Malawi Coll of Med–Johns Hopkins Univ Rsr Proj, Blantyre, Malawi Background: Many HIV-infected individuals present for care with advanced HIV disease. These patients are at high risk of death after antiretroviral therapy (ART) initiation, but risk factors for death in those with advanced HIV are unclear. We used data from a trial of empiric TB therapy to evaluate pre-ART risk factors for early mortality and to test the hypothesis that decreased early CD4 count response to ART is associated with mortality between weeks 4 and 48. Methods: We used data from a multi-site randomized trial in HIV-infected adults initiating ART with CD4 counts <50 cells/mm3 to evaluate risk factors for death within 48 weeks after ART initiation. Cox proportional hazards models were fit to evaluate characteristics present at baseline and at 4 weeks after ART initiation, including the week 4 CD4 cell response and new opportunistic infections (OIs). Results: Of 850 enrolled, the median pre-ART CD4 count was 18 cells/mm 3 . Of 837 included in the baseline analysis, 66 died. Baseline risk factors for death included lymphadenopathy, marginally lower CD4 count, lower albumin, and lower hemoglobin (Table). Among 746 with complete data at week 4, the median change in CD4 count and

CROI 2017 313