CROI 2015 Program and Abstracts

Abstract Listing

Oral Abstracts

Conclusions: Although there was a significant reduction in grade 3 or 4 haematological adverse events in patients allocated to stopping CTX non-inferiority of this strategy with respect to CTX preventable events was not demonstrated. 95LB High-Dose Rifampin, SQ109 and Moxifloxacin for Treating TB: The PanACEA MAMS-TB Trial

Martin J. Boeree 1 ; Michael Hoelscher 2 On behalf of the PanACEA consortium 1 Radboudumc, Nijmegen, Netherlands; 2 University of Munich, Munich, Germany

Background: Shorter regimens are urgently needed for the treatment of TB. The PanACEA MAMS-TB trial was conducted to evaluate whether 12-week combinations of high-dose rifampin, SQ109 and moxifloxacin with standard drugs reduced time to culture conversion on liquid media sufficiently to select for a phase III treatment-shortening trial. Methods: Adult patients with drug-sensitive smear-positive TB were randomly allocated in the ratio 1:1:1:1:2 to be treated for 12 weeks of 1) Q: SQ109 together with standard dose rifampin (R), isoniazid (H) and pyrazinamide (Z), 2) 20RQ: SQ109 and 20 mg/kg R together with ZH, 3) 20RM: moxifloxacin and 20mg/kg R together with ZH, 4) 35R: 35mg/kg R together with ethambutol (E) and HZ, and a control arm for 8 weeks with standard RHZE. All patients then received standard RH to complete a total of 26 weeks of treatment, and were followed for treatment failure and relapse. The trial had a multi-armmulti-stage (MAMS) design with one interim analysis where recruitment to arms could be stopped due to lack of benefit based on pre-specified stopping rules. Results: 365 patients were randomised from 7 sites in Tanzania and South Africa, of whom 25 (7%) were HIV positive. Recruitment to both SQ109 arms was terminated after the interim analysis; patients on these arms remained on treatment and in follow-up. At the final analysis, covariate-adjusted hazard ratios compared to control over 12 weeks were 0.82 (Q, 95% CI 0.55-1.24), 0.73 (20RQ, 0.48-1.13), 1.42 (20RM, 0.98-2.05), and 1.75 (35R, 1.21-2.55). For comparison to previous TB trials, covariate-adjusted hazard ratios compared to control over 8 weeks were 1.69 (1.02-2.80) for 20RM and 1.99 (1.21-3.29) for 35R. This is the largest reduction in time to culture conversion seen in any previous TB trial to our knowledge. Grade 3 or higher adverse events were experienced by 7(12%) Q, 7(12%) 20RQ, 9 (14%) 20RM, 9(14%) 35R and 12(10%) control patients, of which 1, 5, 7, 4 and 5 were considered at least possibly related to treatment. Hepatic adverse events leading to a change in treatment were experienced by 10 (2.7%) patients. Conclusions: These data suggest that 35mg/kg rifampin may reduce the time to culture conversion and may be an important component in future treatment-shortening regimens. For 20mg/kg rifampin and moxifloxacin there was a modest reduction; there was no reduction with SQ109. Adaptive designs such as MAMS are feasible for multi-centre TB clinical trials and could speed regimen development.

Session O-8 Oral Abstracts

Room 6E

10:00 am– 12:.30 pm Factors Affecting HIV Care and Outcome: Global Perspective 96 Special Presentation: PEPFAR 3.0: Controlling the Epidemic and Delivering on the Promise of an AIDS-Free Generation Ambassador Deborah L. Birx US Department of State, Washington, DC, US 97 Joint Estimation of HIV Progression and Survival: A Pooled Analysis of 25 Countries Tara D. Mangal On behalf of the UNAIDSWorking Group on CD4 Progression and Mortality Among HIV Seroconverters including the CASCADE Collaboration in EuroCoord Imperial College London, London, United Kingdom

Oral Abstracts

Background: National estimates of antiretroviral treatment (ART) need and coverage, based on CD4 cell count thresholds, are generated by most countries using the Spectrum model. This uses model parameters for survival and CD4 progression that are derived from separate analyses. We developed a model which simultaneously estimates CD4 decline and survival in HIV-positive individuals by age, sex, and geographic region. We use the largest pooled dataset to date, collating data from 50 HIV seroconverter cohorts and collaborations in Africa, Europe, North America and Asia. Methods: We used a hidden Markov model to describe survival and CD4 decline following seroconversion in the absence of ART. The model estimates forward-only transition rates through six transient states characterised by ranges of decreasing CD4 levels, towards two absorbing states, corresponding to ART initiation and death. We stabilised variability in CD4 counts using a log transformation. Natural short-term fluctuations along with measurement errors in CD4 cell counts were accounted for using a normal measurement error model. Covariates were included on the transition rates and survival probabilities were estimated for each subgroup. Results: The 27,511 eligible individuals had a median follow-up time of 3.55 (IQR 1.71, 6.20) years, were predominantly male (74.1%) and had a median age at seroconversion of 29.1 (IQR 24.4, 35.4) years. Of these individuals, 23,031 had recorded CD4 counts pre-ART, 1943 died, and 14,519 initiated ART. Median (95% CI) survival for males aged 20 years at seroconversion was 11.6 (10.6-12.1), 11.3 (10.5-11.7) and 9.0 (8.7-9.4) years in Africa, Europe/North America and Asia, respectively. The mean time from seroconversion to CD4 counts <500 cells/mm 3 for males in Africa aged 20 years was 3.02 years, conditional on starting above 500 cells/mm 3 . Transition rates and all-cause mortality rates progressively increase with increasing age (hazard ratio [HR] 1.02, 95% CI 1.02-1.03 for every five years of age for CD4 stages, HR 1.17, 95% CI 1.17-1.18 for mortality). Conclusions: CD4 decline and mortality after HIV infection in the absence of ART was similar in African, European and North American cohorts, but significantly faster for all ages in Asian cohorts. Older age is associated with an increased hazard for progression. Joint estimation of these parameters reveals heterogeneities between regions and ages, which should be incorporated into future HIV models to accurately estimate ART need and coverage. 98 Impact of Emergency Department HIV Testing and Linkage to Care: 25 Years’Experience Gabor Kelen 3 ; Eshan U. Patel 2 ;Yu-Hsiang Hsieh 3 ; Oliver B. Laeyendecker 2 ; Judy Shahan 3 ;William Clarke 3 ; Jordyn L. Manucci 3 ; Richard Rothman 3 ; Thomas C. Quinn 2 1 Johns Hopkins University School of Medicine, Baltimore, MD, US; 2 National Institute of Allergy and Infectious Diseases (NIAID), Baltimore, MD, US; 3 Johns Hopkins University School of Medicine, Baltimore, MD, US Background: The Johns Hopkins Hospital (JHH) Emergency Department (ED) has served as both an observational window on the HIV-epidemic for over 25 years, and as a pioneer in ED-based testing and linkage to care programs. We document the changing nature of the local epidemic in inner city Baltimore and the success of ED-based strategic approaches to curtail the epidemic. Methods: We analyzed 7 discrete identity-unlinked serosurveys conducted on 18,240 untargeted adult JHH-ED patients between 1987-2013 for demographic trends in HIV prevalence, cross-sectional incidence estimates, viral load and HCV prevalence. JHH ED HIV testing and linkage to care programs were initiated in 2005 and continue to date. Results: HIV prevalence in 1987 was 5.2%, peaked at >11% from 1992-2003, and then declined to 5.6% in 2013. While seroprevalence was highest for black males, (initial 10%, peak 20%, last 10%), and lowest for white females, the time trend for prevalence was consistent for all groups. Proportion of undiagnosed HIV declined over time from 77% in

132

CROI 2015