CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Results: 3,504 HIV mono-infected patients [66%male, age=39 (IQR 32-47) years, ALT=32 (23-45) U/L, 73% under combined antiretroviral therapy (c-ART), CD4=469 (249-703) cells/mm3 and 50%with detectable HIV viral load] were included. A total of 274 patients (7.8%) died during a mean follow-up of 5.3 (range, 0.2-18.1) years. The mortality rate (95%CI) was 14.7 (95%CI 13.1-16.5) per 1000 person-year. Patients with high FIB-4 score (> 3.25) had a significantly lower 5-year overall survival (95%CI) than those with intermediate/low score [87% (82-90) vs 94% (92-95), p=0.001] (Figure). In a multivariate Cox model the following factors [Hazard Ratio (95%CI)] were independently associated with mortality : age [per year; HR=1.02 (1.01-1.03), p=0.003], CD4 count [< 200 cells/ mm3; HR=2.13 (1.66 2.75), p<0.001], detectable HIV viral load [> 40 copies/ mm3; HR=1.66 (1.15-2.42), p=0.008], intermediate FIB-4 score [FIB-4=1.45- 3.25; HR=1.36 (1.01-1.84), p=0.048] and high FIB-4 score [FIB-4>3.25; HR=1.72 (1.07-2.79), p=0.027]. Conclusion: Simple serological biomarker, such as FIB-4, can be used to predict overall mortality in HIV mono-infected patients. Intermediate to high FIB-4 score (>1.45) was associated with higher risk of mortality adjusted for age, sex and classic HIV risk factors.

1.32]) (Table). Compared to CD4 > 350, CD4 51-200 (aIRR 1.13 [1.03, 1.23]) and CD4 ≤ 50 (aIRR 1.37 [1.22, 1.54]) were associated with increased LOS. Health care coverage with Medicaid and Medicare were each associated with increased LOS. Age ≥ 60 (aIRR vs. age 18-29, 1.17 [1.02, 1.34]) and Southern region (aIRR vs. Eastern region, 1.10 [1.01 1.20]) were associated with increased LOS, while Western region had lower LOS (0.87 [0.80,0.95]). Sex, race, and HIV risk were not associated with LOS. Conclusion: Higher inpatient utilization in patients with ADI and low CD4 highlights the potential importance of early ART initiation to reduce LOS. The association of public health care coverage and geographic region suggests structural factors (poverty, inefficiencies in health care delivery, regulatory policies, etc.) that may be difficult to modify. Older age was generally associated with greater health care utilization, independent of diagnosis. Attention to preventive efforts to decrease the need for hospitalization is particularly important.

Poster Abstracts

901 LONG-TERM EFFECTIVENESS OF HIV TREATMENT IN ZAMBIA

Elvin Geng 1 , Ingrid Eshun-Wilson 1 , Kombatende Sikombe 2 , Carolyn Bolton Moore 3 , Sandra Simbeza 2 , Paul Somwe 3 , Nancy Padian 4 , David Glidden 1 , Izukanji Sikazwe 3 , Charles B. Holmes 5 1 University of California San Francisco, San Francisco, CA, USA, 2 Centre for Infectious Disease Research in Zambia, Lusaka, Zambia, 3 Center for Infectious Disease Research in Zambia, Lusaka, Zambia, 4 University of California Berkeley, Berkeley, CA, USA, 5 Johns Hopkins University, Baltimore, MD, USA Background: Differences in mortality between HIV-infected patients on antiretroviral treatment (ART) and HIV-uninfected persons indicate the effectiveness of public health HIV treatment programs. We combine data from routine clinical care, intensive tracing of a sample of those lost to follow up from care, as well as external data from the Global Burden of Disease Project (GBD) to assess excess mortality up to 10 years after ART initiation in Zambia. Methods: Between 1 August 2013 and 31 July 2015 we followed patients on ART – including both new initiators as well as those already on treatment in 64 clinics in Zambia supported by the CIDRZ program. Sociodemographic, clinical and visit information were obtained from the electronic medical record system. A probability sample of lost patients were intensively traced to ascertain vital status and inverse probability weights were used to incorporate these outcomes into estimates of mortality. We compared age-standardized mortality in HIV patients to age-standardized mortality among HIV-uninfected Zambians in 2015 provided by GBD estimates. Results: Among 165,464 persons on ART followed for 217,849 person-years (pyrs), we observed an age-standardized death rate of 2.8 deaths/100 pyrs among all persons after ART initiation and a age-standardized mortality ratio 3.37 fold higher than HIV uninfected persons (95% CI:3.35-3.66). After excluding the first year of therapy, HIV patients experienced 2.28 deaths/100 pyrs, still 2.78 fold higher than the 0.82 deaths/100 pyrs in HIV-uninfected individuals with same age distribution (95%CI: 2.68-2.90). After one year of treatment, the

900 FIB-4 SCORE PREDICTS OVERALL MORTALITY IN HIV MONOINFECTED: A PROSPECTIVE STUDY Hugo Perazzo 1 , Antonio G. Pacheco 1 , Sandra W. Cardoso 1 , Carolyn Yanavich 1 , Ricardo Santos 1 , Ursula B. Chaves 1 , Mario Sergio Pereira 1 , Ronaldo I. Moreira 1 , Estevao P. Nunes 1 , Valdilea Veloso 1 , Beatriz Grinsztejn 1 , for the GPC-HEPATOL 1 Oswaldo Cruz Foundation - Fiocruz, Rio de Janeiro, Brazil Background: The prognostic value of FIB-4 score, a serological biomarker for detection of advanced fibrosis, has been validated in HIV patients co-infected by viral hepatitis. We aimed to evaluate the prognostic value of FIB-4 score to predict overall mortality in HIV mono-infected patients. Methods: A total of 3,989 HIV patients were prospectively followed from January 2000 to December 2016 at HIV-INI/FIOCRUZ cohort. The exclusion criteria were viral hepatitis co-infection (n=362); missing information of death (n=22); mortality up to 6 months after baseline (n=80) and absence of follow- up (n=21). Deaths were validated by two investigators. FIB-4 was calculated using parameters of the baseline visit by the following formula: FIB-4=(age [years]*AST [U/L])/(platelet [109/L] * sqr (ALT [U/L]). Published cut-off values were used to define low (FIB-4 < 1.45), intermediate (FIB-4=1.45-3.25), and high (FIB-4 > 3.25) probability of advanced fibrosis. Mortality rates were calculated, Kaplan-Meier curves were plotted and multivariate Cox models adjusted for age, gender and classic HIV factors were performed.

CROI 2019 351