Abstract Body

Although most people with HIV (PWH) experience robust CD4 recovery after achieving virologic suppression on antiretroviral therapy (ART), immunologic non-responders (INR) have persistently low CD4 T cell counts. Studies suggest INRs have an increased risk of mortality and serious morbidity, but these events are uncommon in ART-treated PWH and may not be feasible trial endpoints. Patient advocates urged FDA to encourage drug development for INRs by providing guidance on acceptable endpoints for INR trials.Therefore, we investigated the feasibility of a composite endpoint designed to capture non-serious and serious adverse events associated with CD4 lymphopenia.

Among Phase 3 clinical trial datasets submitted to the FDA (2005-2016) in support of ART approval for ART-naïve adults, we identified datasets with 144 wks of HIV RNA, CD4, and safety data. We excluded subjects with virologic failure between Wks 24 and 144 and, based on the Week 96 CD4 value, categorized subjects as INR (CD4 <200 cells/µL), immunologic responders (IR, CD4 200 – 349 cells/µL) or optimal immunologic responders (OIR, CD4 ≥ 350 cells/µL). Using safety data between Wks 96 and 144 and descriptive statistics, we evaluated differences in our composite endpoint, which included 1993 CDC HIV Classification System events (Categories A, B and C [not limited by duration, response to tx, or recurrence]), non-AIDS related events included in the START study, HPV-related disease, skin and soft tissue infections, and neurocognitive events.

79 (1.7%) participants met criteria for INR, 481 (10.3%) for IRs, and 4110 (88%) for OIRs. INRs were older (41.8 yrs) compared to IRs (39.6 yrs) and OIRs (36.7 yrs). INRs had lower baseline CD4 (64 c/µL) compared to IRs (152 c/µL) and OIRs (381 c/µL) and were more likely to have enrolled in a trial that started before 2010 (63% of IRs) than IRs (52.2%) and OIRs (20.2%). The composite endpoint occurred in 17 (21.5%) INRs, 92 (19.1%) IRs, and 709 OIRs (17.2%).

INRs were uncommon among ART-naïve adults starting ART in the 2000s, and even more uncommon after 2010.  Like previous studies, INRs were older with lower baseline CD4 counts. The proportion of INRs experiencing the composite endpoint was slightly higher compared to IRs and OIRs. Our results suggest our composite endpoint is not a feasible endpoint for clinical trials evaluating drugs to treat INRs.