CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

who responded to drug therapy for TB; Group 5 - 115 HIV-negative patients with a clinical diagnosis of active TB, without microbioloigc confirmation of MTB, who responded to drug therapy for active TB. PPD and DPPD were applied intradermally in contralateral arms using a standard Mantoux test technique. Results: Among healthy people previously vaccinated with BCG (Group 1), 57% had positive PPD and 41% had positive DPPD with a good concordance index (Kappa = 0.90). In the context of active TB, DPPD was positive in 93% of patients with active co-infection HIV and TB (Groups 3 and 4), compared to 50% of PPD (p <0.001). All (100%) of HIV negative patients with active TB (Groups 2 and 5) were both DPPD and PPD positive. No differences were noted when microbiologically-confirmed cases were compared to clinically-confirmed cases of active TB. Among HIV-positive patients with active TB, DPPD was significantly more sensitive than PPD when stratified by CD4 cell counts (P=0.0001). Conclusion: The recombinant DPPD antigen may be an important tool in the diagnosis and control of TB, with improved sensitivity and enhanced specificity compared with PPD. 768 HIV-ASSOCIATED TB IN A LOW BURDEN COUNTRY: IS SCREENING FOR LATENT TB STILL NEEDED? Enrico Girardi 1 , Alessandro Cozzi-Lepri 2 , Andrea Gori 3 , Annalisa Saracino 4 , Antonio Di Biagio 5 , Giuseppe Lapadula 6 , Giancarlo Orofino 7 , Miriam Lichtner 8 , Sergio Lo Caputo 4 , Andrea Antinori 1 , Antonella D’Arminio Monforte 3 , Cristina Mussini 9 1 Lazzaro Spallanzani National Institute for Infectious Diseases, Rome, Italy, 2 University College London, London, UK, 3 University of Milan, Milan, Italy, 4 University of Bari, Bari, Italy, 5 University of Genoa, Genoa, Italy, 6 San Gerardo Hospital, Monza, Italy, 7 Amedeo di Savoia Hospital, Torino, Italy, 8 Sapienza University of Rome, Rome, Italy, 9 University of Modena and Reggio Emilia, Modena, Italy Background: Combination antiretroviral therapy (cART) reduces the risk of developing tuberculosis (TB) in persons with HIV both in low and high TB burden countries. Nonetheless, some studies suggest that this risk may remain above that of background HIV -uninfected populations. It is not clear whether in low TB burden countries the recommendation to screen and treat for latent TB infection (LTBI) all persons with newly diagnosed HIV infection is presently justified Methods: We analyzed data of persons enrolled in the ICONA cohort in 2006-2016 in Italy, diagnosed with HIV within 6 months of enrolment. We considered cases of TB diagnosed at first presentation or during follow-up. Factors associated with the risk of developing TB at enrolment were identified by multivariable logistic regression. Incidence rates of TB from enrolment were calculated, and Poisson regression model was used to identify factors associated with the incidence of TB in the study population Results: Among 5,555 persons included in the analysis, 98 (2%) were diagnosed with TB: 72 (73%) were diagnosed at HIV diagnosis, 14 (14%) at least 3 month after HIV diagnosis and before cART start; 8 (8%) during the first 6 months of cART, and 4 (4%) more than 6 months after cART initiation. Factors associated with being diagnosed with TB at enrollment were being immigrants (OR 6.99 vs. Italians; 95%CI: 3.76- 13.02) and having a nadir CD4 cells count< 200/mm 3 (OR 3.17 vs. CD4 cells count > 200, 95%CI: 1.87-5.39). Incidence rates of TB per 1,000 person-years of follow-up were 12.8 (95%CI: 10.4 - 15.8) before cART 3.5 (95%CI: 1.8 - 7.0) during the first 6 months of cART and 0.3 (95%CI: 0.1, 0.8) more than 6 months after cART initiation. Incidence of TB decreased in more recent calendar years and was significantly associated with being immigrants and with a low CD4 cells count (Table) Conclusion: Most of the cases of TB in our study were diagnosed before cART initiation, and the risk of presenting with TB significantly decreased over time. Promoting early HIV diagnosis and immediate initiation of cART appear to be the most important interventions to further decrease the risk of HIV-associated TB in a low TB burden country. Additional prevention efforts should be focused on persons born in high TB burden countries and on these with CD4 nadir below 200/mm 3 , while the effectiveness of screening and treatment of LTBI among those from low TB burden countries and with less advanced immunosuppression should be reevaluated.

769 A PILOT RCT OF GENEXPERT MTB/RIF ON A MOBILE HIV TESTING UNIT IN SOUTH AFRICA Ingrid V. Bassett 1 , Leah S. Forman 2 , Sabina Govere 3 , Hilary Thulare 3 , Simone C. Frank 1 , Bright Mhlongo 3 , Elena Losina 4 1 Massachusetts General Hospital, Boston, MA, USA, 2 Boston University, Boston, MA, USA, 3 AIDS Healthcare Foundation, Durban, South Africa, 4 Brigham and Women’s Hospital, Boston, MA, USA Background: TB is a leading cause of death in South Africa, yet often goes undiagnosed. Community-based GeneXpert MTB/RIF screening may expedite TB diagnosis and treatment. Methods: We conducted a proof-of-concept randomized trial to evaluate the yield of GeneXpert screening on a mobile HIV testing unit at community venues in Umlazi Township, Durban. Eligible adults (≥18y) underwent rapid HIV testing and a TB symptom screen and were randomized to usual care or intervention. Participants with TB symptoms in the usual care armwere asked to produce a sputum sample, for GeneXpert testing at the provincial hospital laboratory; participants were contacted ~7 days later with results and treatment referral. In the “Test & Treat TB” intervention, all HIV-infected or HIV-uninfected/ TB symptomatic participants able to produce sputum underwent GeneXpert testing on the mobile unit. GeneXpert-positive participants received expedited TB treatment initiation on the mobile unit as well as monthly SMS reminders and non-cash incentives for picking up test results, linking to an initial clinic visit, and TB treatment completion. We assessed linkage to TB care and 6-month TB treatment outcomes. Results: We screened 7,361 people over 20 months. 4,815 were eligible and enrolled; median age was 27 (IQR 22 to 35), 51%were male, and 95% reported prior HIV testing. TB symptoms included cough (5%), weight loss (4%), night sweats (4%), and fever (3%). Overall, 42% of eligible participants could produce sputum samples (intervention: 55%; usual care: 26%). Among intervention participants, 41% exhibiting no TB symptoms successfully produced sputum compared to those with 1 (48%), 2 (71%), 3 (72%), or 4 (89%) symptoms. Seven participants tested GeneXpert-positive, six in the intervention arm (3%, 95% CI 1%, 5%) and one in the usual care arm (1%, 95% CI 0%, 6%). All 6 in the intervention arm linked to care within 6 months, and 5 of the 6 completed treatment; the GeneXpert-positive participant in the control arm did not link to care and did not complete TB treatment. Conclusion: Screening for TB on a mobile HIV testing unit in the community using GeneXpert is feasible, though likelihood of specimen production is higher among those with more TB symptoms. Overall yield for GeneXpert-positive TB was low, however, the expedited “Test & Treat TB” strategy led to rates of TB treatment completion comparable to clinic-initiated treatment in South Africa.

Poster Abstracts

CROI 2018 288