CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

817 HIV AND CANCER RISK IN ADOLESCENTS AND YOUNG ADULTS IN SOUTH AFRICA, 2004-2014 Tafadzwa G. Dhokotera 1 , Julia Bohlius 2 , Adrian Spoerri 2 , Matthias Egger 2 , Jabulani Ncayiyana 3 , Gita Naidu 4 , Victor Olago 1 , Elvira Singh 1 , Mazvita Sengayi 1 1 National Health Laboratory Service, Johannesburg, South Africa, 2 Institute of Social and Preventive Medicine, Bern, Switzerland, 3 University of Cape Town, Cape Town, South Africa, 4 University of the Witwatersrand, Johannesburg, South Africa Background: The risk of developing cancer is increased amongst Adolescents and Young Adults (AYA) living with HIV compared to uninfected AYA. Low retention in care, poor virologic control, low CD4 counts and a high prevalence of other oncogenic viruses may further increase cancer risk in AYA living with HIV. We aimed to determine the association between cancer and HIV as well as the spectrum of cancers amongst AYA living with and without HIV. Methods: This was a record linkage study of all AYA aged between 10 and 25 who had a cancer diagnosed between 2004 and 2014 in the South African public health sector laboratories. HIV data were retrieved from the National Health Laboratory Service and linked using probabilistic methods to cancer records in the South African National Cancer Registry database. The linkage variables included names, surnames, age and gender. We further extracted additional HIV data from the clinical history section of the cancer pathologists’ report using text searching methods. To determine the association between HIV status (infected vs uninfected) and different cancers in AYA , we fitted logistic regression models adjusting for age (adolescents vs young adults), gender (as appropriate), ethnicity (black vs non-black) and calendar period (early vs mid vs late-ART). Results: From 2004 to 2014, 8,472 AYA were diagnosed with incident cancer. The HIV status was known for 45% (n=3,812) of the AYA cancer population and the remainder was not tested for HIV. About half of those with a known HIV status were HIV positive (n=1,853; 48.6%). Female AYA with cancer were more frequently HIV positive as compared to male AYA with cancer and black AYA with cancer more frequently than non-black ethnic groups. Young adult cancer patients were more frequently HIV-positive compared to adolescent cancer patients. Adjusted odds ratios for AYA living with compared to only those without HIV were 219 (95% CI 90-530) for Kaposi sarcoma, 2.18 (95% CI 1.23- 3.88) for cervical cancer, 2.09 (95% CI 1.66-2.63) for non-Hodgkin’s lymphoma and 2.73 (95% CI 1.27-5.86) for anogenital cancers other than cervix. Conclusion: The elevated risk of different AIDS defining and non AIDS defining cancers in this age group points to a possible gap in the ART programme for AYA living with HIV. The elevated risk for cervical cancer in young women indicates the need for enhanced continuous screening for pre-cancerous cervical lesions amongst AYA living with HIV as well as improved primary prevention strategies.

children, adolescents and young adults who received public sector health care in Gauteng province, South Africa, and assessed cancer incidence. Methods: We retrieved laboratory records for HIV and CD4 tests from the National Health Laboratory Service for the years 2004-2014. We used privacy preserving probabilistic record linkage using names, date of birth/age, sex, and facility information as linkage variables to identify records belonging to the same person. Next, we linked the cohort to records of the National Cancer Registry using similar methods. We included children, adolescents and young adults with ≥2 HIV-related laboratory test records in the analysis. We estimated cancer incidence rates per 100,000 person-years (pys), stratified by current age group (<10, 10-17, and 18-25 years). Time at risk was calculated from the 1st positive HIV test or CD4 measurement to 6 months after the last laboratory test record or to cancer diagnosis. Results: Based on 907,171 laboratory records we created a cohort of 49,946 children (<10 years), 17,707 adolescents (10-17 years) and 120,706 young adults aged 18-25 years at their 1st HIV-related laboratory test. Half (51%) of children, 70% of adolescents and 89% of young adults were female. Median 1st CD4 count in children ≥5 years, adolescents, and young adults was 349 cells/ µl (IQR 210-521); in children <5 years, median 1st CD4%was 9% (IQR 5-15). Over 576,389 pys 414 incident cancer cases were recorded. Median follow-up time was 2.3 years (IQR 1.0-4.6). Overall cancer incidence rates per 100,000 pys increased with current age group from 77 (95% confidence interval [CI] 63-94) in children, 90 (95% CI 70-117) in adolescents to 121 (95% CI 108-137) in young adults (Figure). In children, cancer-specific incidence rates were similarly high for Kaposi sarcoma (KS), Burkitt lymphoma, other non-Hodgkin Lymphoma (NHL), and leukemia. KS incidence rates increased with age. Specific incidence rates were higher for KS than for other cancers in adolescents and young adults. Conclusion: Record linkage methods using laboratory test records can facilitate the assessment of cancer risk in children and young people living with HIV in South Africa. Incidence rates for cancers increased with age which was mainly driven by cancers associated with oncogenic viruses. Dedicated prevention strategies are required.

Poster Abstracts

819 YOUTH PSYCHIATRIC TRAJECTORIES PREDICT PERINATALLY HIV INFECTED YOUNG-ADULT VIREMIA

Nadia Nguyen 1 , Chien-Wen J. Choi 1 , Reuben Robbins 1 , Curtis Dolezal 1 , Cheng- Shiun Leu 1 , Rehema Korich 1 , Elaine J. Abrams 2 , AndrewWiznia 3 , Claude A. Mellins 1 1 New York State Psychiatric Institute, New York, NY, USA, 2 ICAP at Columbia University, New York, NY, USA, 3 Albert Einstein College of Medicine, Bronx, NY, USA Background: Little is known about the relationship between patterns of psychiatric functioning over time and HIV outcomes, especially among perinatally HIV-infected (PHIV) adolescents and young adults (AYA). Methods: Using data from one of the few cohort studies of PHIV AYA, we identified longitudinal trajectories of psychiatric disorders among 130 PHIV AYA living in New York City, and examined their associations with sociodemographic factors at enrollment and experiencing a viremic event in young adulthood. Psychiatric disorders (mood, anxiety, behavioral, substance use) were assessed using the Diagnostic Interview Schedule for Children at enrollment (ages 9-16) and 4 follow-up (FU) visits. At last FU (ages 18-28), a viremic event was defined as any past year viral load >200 copies/mL. Multivariate longitudinal latent class

818 CANCER INCIDENCE IN YOUNG PEOPLE WITH HIV: PRIVACY-PRESERVING RECORD-LINKAGE STUDY Eliane Rohner 1 , Lina Bartels 1 , Victor Olago 2 , Wenlong C. Chen 2 , Elvira Singh 2 , Adrian Spoerri 1 , Matthias Egger 1 , Mazvita Sengayi 2 , Julia Bohlius 1 1 University of Bern, Bern, Switzerland, 2 National Health Laboratory Service, Johannesburg, South Africa Background: Data on cancer incidence in children and young people living with HIV are sparse. We used laboratory test records to create a cohort of

CROI 2019 318