CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

785 EVALUATION OF A CRYPTOCOCCAL SCREENING AND TREATMENT PROGRAM IN HIV CLINICS IN UGANDA

Poster Abstracts

Elizabeth Nalintya 1 , David Meya 1 , Sarah Lofgren 2 , Kathy Huppler Hullsiek 2 , David R. Boulware 2 , Radha Rajasingham 2 1 Infectious Disease Institute, Kampala, Uganda, 2 University of Minnesota, Minneapolis, MN, USA Background: Screening for cryptococcal antigen (CRAG) among those with a CD4 cell count <100 cells/µL, and treating asymptomatic CRAG+ with fluconazole saves lives and is recommended by the WHO. However, implementation and uptake have been slow outside of clinical trials. Programs with high rates of loss to follow up or delayed ART initiation do not demonstrate the same survival benefit as seen in clinical trials. Few studies have evaluated optimal programmatic implementation to maximize survival amongst CRAG+. We designed a CRAG screening program, and evaluated implementation and clinical outcomes in routine care in Kampala. Methods: At 11 HIV clinics, CD4 results<100 cells/µL were reflexively tested for CRAG using remaining plasma. Preemptive fluconazole treatment was given to asymptomatic CRAG+ persons. Our screening program included a) clinic-wide education, b) laboratory staff training, c) identification of a responsible clinic point-person, trained in operational aspects, and d) a system for ongoing review and feedback. If participants lost to follow up were not found after active tracing, they were presumed dead. Results: Between December 2015 and January 2017, 1446 persons with a CD4<100 cells/µL were screened for CRAG, with a median CD4 cell count of 40 (IQR: 17 to 70) cells/µL. Prevalence of CRAG+ was 6.5% (n=94/1446) (Figure 1). Seven CRAG+ persons died or were lost to follow up prior to further clinic evaluation. Of the 53 asymptomatic CRAG+, median CD4 cell count was 20 cells/ µL (IQR: 5 to 45), and median plasma CRAG titer was 1:40 (IQR: 1:20 to 1:160). 100%were prescribed fluconazole therapy and ART, a median of 14 days and 28 days after CRAG screening, respectively. Six-month survival was 87% (46/53). Of 26 symptomatic CRAG+, 17 consented to a lumbar puncture, and 16 had confirmed cryptococcal meningitis with a median CD4 cell count of 27 cells/ µL (IQR: 10 to 45) and median plasma CRAG titer of 1:160 (IQR: 1:40 to 1:1120). For those with confirmed meningitis, mortality was 44% at 6 months despite amphotericin-based treatment. Of those symptomatic who declined lumbar puncture (n=9), 78% (7/9) died. Conclusion: Our CRAG screening program resulted in 87% survival for asymptomatic CRAG+ persons who received timely fluconazole and ART, however 13% CRAG+ died before returning for evaluation. Mortality from symptomatic meningitis was 44% despite amphotericin therapy. Access to standard meningitis treatment is desperately needed to further reduce mortality.

786 CLINIC-BASED LATERAL FLOW CRYPTOCOCCAL ANTIGEN TESTING AT HIV DIAGNOSIS, SOUTH AFRICA Paul K. Drain 1 , Ting Hong 1 , Hilary Thulare 2 , Mahomed-Yunus Moosa 3 , Connie L. Celum 1 1 University of Washington, Seattle, WA, USA, 2 AIDS Healthcare Foundation, Durban, South Africa, 3 University of KwaZulu-Natal, Durban, South Africa Background: The World Health Organization (WHO) recommends screening HIV-infected people for cryptococcal antigens to identify cryptococcosis, a major cause of AIDS-related deaths. A point-of-care assay to detect cryptococcal antigens may be beneficial, but has not been validation at the clinical point of care following HIV testing. We sought to determine the feasibility and validation of a lateral flow assay for cryptococcal antigenemia among newly-diagnosed HIV-infected adults in South Africa. Methods: We conducted a cross-sectional study of newly-diagnosed HIV- infected adults who received voluntary HIV testing in an outpatient clinic. A trained nurse obtained urine, finger-prick and venous whole blood and performed clinic-based testing with a rapid cryptococcal antigen lateral flow assay (Immy Inc., Norman, USA) per manufacturer’s specifications. We also performed lab-based serum cryptococcal antigen testing with an ELISA-based assay, as developed by the same manufacturer, as the gold standard. We assessed diagnostic accuracy of the rapid cryptococcal assay, and stratified results by CD4 categories. Results: We enrolled 5,618 participants, among whom 1,588 were HIV- infected and screen for cryptococcal antigenemia. The mean age was 33.6 (SD ±9.2) years, 1,184 (59.8%) were female, and median CD4 count was 309 cells/ mm 3 (interquartile range: 162–477 cells/mm 3 ). Among those, 133 (8.4%), 48 (3.0%), and 31 (2.0%) were cryptococcal antigen positive by the lateral flow assay for urine, finger-prick blood, and venous whole blood samples, respectively. The urine, finger-prick blood, and venous whole blood had a overall sensitivity of 50% (95% CI 23-77%), 36% (95% CI 13-65%), and 43% (18- 71%), and likelihood ratio positive values of 4.90, 10.8, and 22.3, respectively. When frozen serum samples were tested using the lateral flow assay, the assay sensitivity was 93% (95% CI 66-100%) and specificity was 100% (95% CI 88- 100%). Two independent readers had very high agreement for lateral flow assay results (p<0.0001). Conclusion: The performance of the lateral flow cryptococcal antigen assay was moderately sensitive for both finger-prick and venous whole blood, but was too non-specific for urine samples, among untreated HIV-infected adults in South Africa. However, the performance of the cryptococcal antigen assay was excellent for serum samples. While clinic-based cryptococcal antigenemia

CROI 2018 295