CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

628 BREAST CANCER RISK AMONG WOMEN WITH HIV IN NORTH AMERICA (2000-2015) Sally B. Coburn 1 , Michael J. Silverberg 2 , Richard D. Moore 1 , W. C. Mathews 3 , Julia L.Marcus 4 , Jessica L. Castilho 5 , Mari M. Kitahata 6 , Heidi M. Crane 6 , Marina Klein 7 , Sonia Napravnik 8 , Todd T. Brown 1 , Kala Visvanathan 1 , Bryan Lau 1 , Keri N. Althoff 1 , for the NA-ACCORD of IeDEA 1 Johns Hopkins University, Baltimore, MD, USA, 2 Kaiser Permanente Northern California, Oakland, CA, USA, 3 University of California San Diego, San Diego, CA, USA, 4 Harvard University, Cambridge, MA, USA, 5 Vanderbilt University, Nashville, TN, USA, 6 University of Washington, Seattle, WA, USA, 7 McGill University, Montreal, QC, Canada, 8 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: Breast cancer burden is poorly characterized in women with HIV regarding incidence and risk factors. Some studies suggesting reduced risk in women with versus without HIV further support assessing breast cancer among women with HIV. We estimated breast cancer cumulative incidence and risk factors in the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD). Methods: We included women ≥35 years old who were prescribed antiretroviral therapy (ART), observed in the NA-ACCORD from 1/1/2000- 12/31/2015, had no cancer history, and had ≥6 months of follow-up. Study entry was the latest date of: 1/1/2000, age 35, ART initiation, or NA-ACCORD enrollment. Study exit was the earliest date of: invasive breast cancer diagnosis, death, loss to follow-up (≥2-year gap after CD4/HIV RNA measurement), or administrative censoring. Standardized case validation included abstraction, linkages with cancer registries and/or record review of cancer site/pathology. With age as the time metric, we used non-parametric estimators accounting for the competing risk of death to assess breast cancer cumulative incidence. We used Fine & Gray regression to quantify breast cancer risk by time-varying HIV viral load, obesity (body mass index [BMI] ≥30 kg/m 2 ) at study entry, race, smoking status, and CD4 count at ART initiation. We calculated adjusted subdistribution hazard ratios (aSDHR) with 95% confidence intervals (95% CI). Results: We included 8,242 women contributing 53 breast cancer cases with 55,113 years of follow-up. Median follow-up was 6 years (IQR: 3 to 10) and median age at study entry was 42 (IQR: 37 to 48). Median age at diagnosis was 52 (IQR: 46 to 56). Breast cancer cumulative incidence was 3.7%. Breast cancer risk was associated with smoking (aSDHR 3.0, 95% CI 1.2, 7.6) and marginally non-significantly associated with obesity (aSDHR 1.8, 95% CI 1.0, 3.2) (Table 1). Though non-significant, there was a 39% reduced risk of breast cancer per log 10 increase in viral load (p=0.6), and 17% increased risk with lower CD4 count (p=0.6). There were no differences by race. Conclusion: We observed breast cancer risk of 3.7% in women with HIV and compelling associations with smoking, obesity, and HIV clinical factors. Findings should be interpreted cautiously given our sample size and limited follow-up after age 65. Further investigation is merited as more follow-up is accrued incorporating traditional/reproductive risk factors and direct comparisons to the general population.

627LB OUTCOMES AFTER PEGYLATED LIPOSOMAL DOXORUBICIN FOR KAPOSI SARCOMA IN MOZAMBIQUE Matthew E. Coldiron 1 , Ana Gabriela Gutierrez Zamudio 2 , Rolanda Manuel 3 , Gilda Luciano 3 , Barbara Rusch 4 , Iza Ciglenecki 4 , Rebecca F. Grais 1 , Laurence Toutous Trellu 5 , Lucas Molfino 2 1 Epicentre, Paris, France, 2 MSF, Maputo, Mozambique, 3 Ministry of Health, Maputo, Mozambique, 4 MSF, Geneva, Switzerland, 5 University Hospitals of Geneva, Geneva, Switzerland Background: Standard care for Kaposi’s sarcoma in Mozambique consists of ARV and a combination of doxorubicin, bleomycin and vincristine. Historically, patient outcomes have been poor, with Mozambique’s largest case series reporting 43% death or loss to follow-up after two years. To improve outcomes and retention in care, Médecins Sans Frontières introduced pegylated liposomal doxorubicin (PLD) in 2016. Methods: A prospective observational study was implemented between March 2016 and December 2018. Patients aged >15 years were eligible if they had T1 KS or T0 KS not responding to ≥6 months of ART with significant impact of KS on their quality of life. Exclusion criteria included prior receipt of chemotherapy and Karnofsky score <50. Patients received PLD on three-week cycles until meeting stopping criteria based on clinical response or toxicity. HIV status and KS were monitored at regular follow-up visits, and PHQ-9 and SF-12 questionnaires were administered to evaluate depression and quality of life. Adverse events were monitored passively. Primary outcomes were vital status and KS progression at 6, 12 and 24 months of follow-up. Survival was estimated using Cox proportional-hazards regression. Results: 183 KS patients were screened and 116 participants were enrolled. At baseline, patients presented with advanced KS (72% had lymphedema) and moderately advanced immunosuppression, with 53% of patients on ART ≤6 months. 23 participants (20%) died and 15 (13%) were lost to follow-up. Participants with CD4<100 at the time of enrolment were more likely to die (HR 2.7, 95%CI [1.2-6.2], p=0.016) (Figure), as were participants with T1S1 disease at enrollment compared to those with T1S0 disease (HR 2.7, 95%CI [1.1-6.4], p=0.023). Most deaths occurred in patients with advanced immunosuppression and multiple social problems. 92 participants achieved complete or partial remission at any point during follow-up (overall response 80%), including 15 (13%) who achieved complete remission. Of those achieving CR or PR, 26 (28%) eventually restarted PLD because of recurrent disease or worsening symptoms. The most common AEs were due to neutropenia and anemia. Quality of life improved significantly after 6 months. Conclusion: PLD was safe, well-tolerated and effective for the treatment of KS in Mozambique. The high mortality rate is likely due to advanced immunosuppression at baseline and underscores the need to provide earlier screening and referral for treatment of KS. Efforts should be made to increase access to PLD in Mozambique.

Poster Abstracts

CROI 2020 227