CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

765 SERUM ALBUMIN AS A LONG TERM PREDICTOR OF SERIOUS NON-AIDS EVENTS Andreas Ronit 1 , Camilla Ingrid Hatleberg 1 , Lene Ryom 1 , Fabrice Bonnet 2 , Wafaa M. El-Sadr 3 , Peter Reiss 4 , Rainer Weber 5 , Christian Pradier 6 , Stephane de Wit 7 , Matthew Law 8 , Antonella D’Arminio Monforte 9 , Jens D. Lundgren 1 , Andrew N. Phillips 10 , Caroline Sabin 10 1 Rigshospitalet, Copenhagen, Denmark, 2 INSERM, Bordeaux, France, 3 ICAP at Columbia University, New York, NY, USA, 4 Academic Medical Center, Amsterdam, Netherlands, 5 University Hospital Zurich, Zurich, Switzerland, 6 Nice University Hospital, Nice, France, 7 Saint-Pierre University Hospital, Brussels, Belgium, 8 Kirby Institute, Sydney, NSW, Australia, 9 Azienda Ospedaliera San Paolo, Milan, Italy, 10 UCL Great Ormond Street Institute of Child Health, London, UK Background: Lower serum albumin (sAlb) has been associated with an increased risk of mortality and AIDS in HIV-positive people and may be associated with the development of serious non-AIDS events (SNAEs). We evaluated the association between sAlb and the risk of SNAEs. Methods: D:A:D participants free of SNAEs were followed from their first routine sAlb value (01/01 2004 onwards) to the first of a new centrally-validated SNAE (cardiovascular disease (CVD: myocardial infarction, stroke, invasive cardiovascular procedure), end-stage liver disease (ESLD), end-stage renal disease (ESRD), non-AIDS malignancy (NADM), death from any non-AIDS cause), AIDS death, 6 months after last clinic visit or 01/02/2016. Poisson regression models were used to determine associations between sAlb (fixed covariate) and a new i) SNAE, ii) CVD or iii) NADM event, with adjustment for potential confounders (including markers of HIV disease and other measures of organ system injury, Table 1). Models additionally tested whether the association varied with age, duration of follow-up, or smoking status. Results: Of 16,350 participants (71.8%male, median (interquartile range, IQR) age 44 (37, 51), 50.5%white, 90.1% antiretroviral therapy experienced) 1,463 developed a SNAE (371 CVD, 200 ESLD, 40 ESRD, 553 NADM, 299 deaths from other non-AIDS causes) over 80,264 person-years. Median (IQR) sAlb was 43 (40-46) g/L). The SNAE rate was 1.82 [95% confidence interval 1.73-1.92]/100 person-years. Lower sAlb was associated with an increased risk of each event (Table 1). While there was no clear threshold effect of sAlb, there was a clear reduced risk across sAlb levels above 35 g/L (Table 1). The association between sAlb (per 5 g/L higher) and a SNAE was attenuated with older age as event rates increased (age in years<30: adjusted rate ratio 0.73 [0.54-0.99]; 30-50: 0.77 [0.74-0.81]; >50: 0.82 [0.78-0.87]; p-interaction=0.001) but did not appear to wane with additional years of follow-up (0-2: 0.75 [0.72-0.78]; 3-4: 0.88 [0.81-0.95]; 5-6: 0.83 [0.74-0.94]; >6: 0.82 [0.74-0.91], p-interaction=0.79). The association between sAlb and SNAE appeared stronger for current smokers (0.76 [0.73-0.80] than for never smokers 0.87 [0.8-0.95], p<0.01). Conclusion: sAlb is an independent and long-term risk factor for SNAE. Future studies are needed to determine the mechanism underlying this association and risk-score studies should consider evaluating the potential role of sAlb in predicting SNAEs.

766 MOBIILITY PREDICTS INCIDENT TB INFECTION IN CHILDREN & ADULTS IN RURAL UGANDA Carina Marquez 1 , Atukunda Mucunguzi 2 , Gabriel Chamie 1 , Laura B. Balzer 3 , Joel Kironde 4 , Theodore Ruel 1 , Bridget Nzarubara 4 , Dalsone Kwariisima 5 , Edwin D. Charlebois 1 , Maya L. Petersen 6 , Diane V. Havlir 1 1 University of California San Francisco, San Francisco, CA, USA, 2 Makerere University– University of California San Francisco Research Collaboration, Kampala, Uganda, 3 University of Massachusetts Amherst, Amherst, MA, USA, 4 Makerere Univ and Univ of California San Francisco Rsr Collab, Mbarara, Uganda, 5 Makerere University Joint AIDS Program, Kampala, Uganda, 6 University of California Berkeley, Berkeley, CA, USA Background: In sub-Saharan Africa, a significant portion of the large latent TB reservoir is established outside of the home. Data on TB risk among mobile children and adults who travel outside of the community for work or school is limited, but critical to targeting TB prevention efforts. Methods: We assessed the association between mobility and incident TB infection in an ongoing longitudinal cohort of tuberculin skin test (TST) negative children (≥5 years) and adults in the SEARCH trial (NCT:01864603) in Eastern Uganda, this sample was enriched for HIV-infection. Participants were included in the TST negative cohort if they had no induration at their baseline TST. A follow-up TST was placed one year from baseline. Incident TB infection was defined as a change in TST induration from 0mm at baseline to ≥10mm at their annual follow up test. Mobility was defined as living outside of the community for more than 1 month in the last year, such as to attend boarding school or work. We used multivariate logistic regression and adjusted for confounding by age, gender, lowest wealth tertile, BCG vaccination, HIV status, living in household with HIV-infected adult, and household TB contact within the last year. Results: One year follow-up TSTs were completed in 739 participants (age ≥5 years) from the TST negative cohort. Our definition of incident TB infection was met by 89 (12%) participants. Among those with incident TB infection: 48% were ages 5-14 years of age, 17%were 15-24 years, and 35%were older than 25 years, 65%were women, 93% had a BCG scar or record of vaccination, 10% were mobile, 6% reported a household contact within a year. In the adjusted model, mobility was associated with a 2.6-fold increased odds of incident TB infection (95% CI: 1.3-7.5, p<0.01). Reporting a household contact was also independently associated with TB infection (aOR 11.3, 95% CI: 2.1-62.2, p<0.01). HIV infection and living in a household with one or more HIV-infected adult was not associated with incident TB infection. Conclusion: In a rural Ugandan cohort of children and adults where population based treatment of HIV is ongoing, mobility was a predictor of increased risk of incident TB infection. TB exposure outside of the community, such as in boarding schools, may drive a portion of TB infections in rural communities. Casual TB contacts within rural communities and undiagnosed household contacts may explain the incident infections not associated with a known household contacts or mobility. 767 IMPROVED SENSITIVITY OF A NOVEL RECOMBINANT PROTEIN SKIN TEST FOR THE DIAGNOSIS OF TB Roberto Badaro 1 , Corey Casper 2 , Rhea Coler 2 , C A. Araujo-Neto 1 , Diana B. Pedral-Sampaio 1 , Maria Nakatani 1 , Antonio Campos Neto 3 , Steven Reed 2 1 Federal University of Bahia, Salvador, Brazil, 2 Infectious Disease Research Institute, Seattle, WA, USA, 3 Forsyth Institute, Boston, MA, USA Background: The tuberculin skin test (TST) or Mantoux reaction (MR) has been used worldwide in the diagnosis of M. tuberculosis infection for over 100 years. Many regions have experienced shortages of the purified protein derivative (PPD) TST in recent years as the number of suppliers has decreased. The accuracy of the PPD TST is limited by prior BCG vaccination, infection by other non- tuberculous mycobacteria, reduced sensitivity in immunocompromised hosts with a high risk of tuberculosis (TB) such as HIV, and limited ability to diagnose active TB. We previously described a recombinant protein (“DPPD”) derived from M. tuberculosis (MTB) that can replace the classical PPD antigen used in TST. Methods: We compared DDPD to PPD using MR in a population of 672 people in the state of Bahia, Brazil. Participants were divided among 5 specific groups: Group 1 - 232 BCG-vaccinated individuals with no signs or symptoms nor history of active TB; Group 2 - 40 HIV-negative patients with active TB as defined by a clinical isolate with a positive culture for MTB or positive acid fast stain; Group 3 - 38 HIV-positive individuals with active TB and MTB found on culture a positive acid fast stain; Group 4 - 182 HIV-positive patients with signs and symptoms compatible with active TB, but without microbiologic confirmation,

Poster Abstracts

CROI 2018 287