CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

the questionnaires. Inferential statistics (with false discovery rate) compared demographic and clinical factors between clusters (e.g., ELS). Results: We identified four MH phenotypes (Fig. 1). Cluster 1 (n=76; PTSD phenotype) endorsed clinically significant PTSD symptoms, with an average PCL-C total score >33. Clusters 2 (n=32; anxiety phenotype) and 3 (n=130; mixed anxiety/depression phenotype) reported minimal PTSD symptoms, with modest BAI (Cluster 2) and PHQ-9 (Cluster 3) elevations. Cluster 4 (n=39; normative MH phenotype) reported no clinical MH symptom elevations. Comparisons of explanatory factors between MH phenotypes revealed a modestly higher rate of physical (14.5% vs. 5.1%; P=0.13) and sexual (27.6% vs. 12.8%; P=0.07) abuse among the PTSD phenotype (Cluster 1) vs. the normative MH phenotype (Cluster 4 ). Conclusion: We identified unique MH phenotypes among PWH and confirmed the importance of ELS, particularly sexual abuse, as an early risk determinant for unfavorable MH among PWH in adulthood. Notably, PWH in the normative MH phenotype also reported a history of sexual abuse, consistent with resilience. Follow-up analyses will investigate these groups to further examine mechanisms of risk vs resi. The figure, table, or graphic for this abstract has been removed. Asymptomatic CSF HIV-1 Escape: Incidence and Effects on CNS Biomarkers Gustaf Ulfhammer 1 , Arvid Edén 1 , Aylin Yilmaz 1 , Lars Hagberg 1 , Erik Sörstedt 1 , Erika Tyrberg 1 , Åsa Mellgren 1 , Staffan Nilsson 1 , Kristina Nyström 1 , Henrik Zetterberg 1 , Dietmar Fuchs 2 , Johanna Gostner 2 , Magnus Gisslén 1 1 Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden, 2 Medical University of Innsbruck, Innsbruck, Austria Background: Elevated levels of HIV RNA can occasionally be detected in cerebrospinal fluid (CSF) despite viral suppression in plasma by antiretroviral therapy (ART), a condition termed CSF HIV-1 escape (CSF-E). In rare cases CSF-E is associated with clinical neurological disease (symptomatic CSF-E) but more commonly, no neurological symptoms are evident (asymptomatic CSF-E). Previous studies on asymptomatic CSF-E have been limited in size, leaving its incidence and clinical relevance unclear. Our objectives were to determine the incidence of CSF-E in a large, well-characterized Swedish cohort of people with HIV (PWH), and to evaluate its impact on biomarkers indicative of CNS inflammation and injury. Methods: We retrospectively included neuroasymtomatic PWH from a longitudinal cohort who had been successfully treated with ART for >6 months (plasma HIV RNA <50 copies/mL prior to inclusion) and had undergone CSF examinations for research purposes between 2016–2022. CSF-E was defined as either increased CSF HIV RNA with concurrent plasma suppression or as CSF HIV RNA exceeding that in plasma when both were detectable. HIV RNA, neopterin, neurofilament light protein (NfL) and albumin ratio were analyzed in paired CSF and plasma samples. Results: In total, 452 samples from 174 PWH (65% male) were included. Mean number of samples per participant was 2.6 (range 1–7) where ≥2 samples were collected from 115 PWH. CSF-E was present in 15/174 PWH (8.6%) with a median (interquartile range, IQR) CSF HIV RNA of 34 (26–52) copies/mL. Of 452 samples, CSF-E was present in 4% (1% with the cut-off ≥50 copies/mL) and 13% had detectable HIV RNA in plasma (3% with the cut-off ≥50 copies/mL). One individual had ≥2 consecutive samples with an isolated increase in CSF HIV RNA but remained clinically stable and did not require any treatment adjustment. None of the PWH with CSF-E developed subsequent viral failure (defined as a single plasma viral load >500 copies/mL or two consecutive samples ≥50 copies/mL drawn >6 weeks apart). CSF HIV RNA levels were weakly, but significantly, correlated with CSF neopterin concentrations (r = 0.21, p< 0.001). No significant association was found between increased CSF viral load and elevated levels of either CSF NfL or albumin ratio. Conclusion: In this cohort, CSF-E was rare and usually resolved without treatment adjustments. A weak association exists between CSF-E and intrathecal immune activation, but no correlations were found with biomarkers of neuronal or blood-brain barrier damage.

rapid plasma reagin, Treponema pallidum particle agglutination assay, and metagenomic sequencing. Results: Sequential CSF samples from 465 people living with HIV with suspected CNSI underwent analysis with FilmArray-ME and 454 had analysis with Xpert MTB/RIF Ultra. The median age of study participants was 40 (IQR 34-48), 54.4% were male, and the median CD4 count was 114 cells/µL (IQR 43-325); 49.0% (220/465) were established on ART the remainder were either ART naïve or had cycled out of treatment services. Through routine CSF analysis alone there were 99 microbiologically confirmed CNSIs; with the addition of enhanced diagnostics this increased to 155 (Figure 1), a relative increase of 57%. Detection of Mycobacterium tuberculosis, T. gondii and viral pathogens was only achieved through the addition of enhanced diagnostics. Cryptococcal meningitis was the most commonly diagnosed CNSI in 97/465 patients with 92/97 diagnoses made through cryptococcal antigen (CrAg) testing or India ink, figure 1. Restricting analysis to CrAg negative patients, only 11 CNSI diagnoses were made through routine analysis compared to 62 diagnoses made with the addition of enhanced diagnostics, a relative increase of 464%. Conclusion: Enhanced diagnostic platforms substantially increased diagnostic yield in patients with suspected CNSI in Botswana and Zimbabwe. There was significant variation between population groups suggesting targeted use may be possible. The most clinically and cost-effective diagnostic algorithms need to be defined.

574

Poster Abstracts

573

Mental Health Phenotypes of Well-Controlled HIV in Uganda Leah H Rubin 1 , Kyu Cho 2 , Jacob Bolzenius 2 , Julie Mannarino 2 , Aggrey Anok 3 , Stephen Tomusange 3 , Raha M. Dastgheyb 1 , Eran F. Shorer 1 , Deanna Saylor 1 , Maria J. Wawer 1 , Noeline Nakasujja 4 , Gertrude Nakigozi 3 , Robert Paul 2 1 The Johns Hopkins University, Baltimore, MD, USA, 2 University of Missouri St Louis, St Louis, MO, USA, 3 Rakai Health Sciences Program, Kalisizo, Uganda, 4 Makerere University, Kampala, Uganda Background: HIV and mental health (MH) disorders, particularly depression, anxiety, and post-traumatic stress disorder (PTSD), are among the top 10 causes of disability among people with HIV (PWH) in Uganda. Most studies of PWH have focused on MH disorders as unidimensional constructs. However, the phenotypic expression and clinical course of MH conditions among PWH in Uganda and worldwide are heterogeneous. Accordingly, there has been a shift towards identifying MH phenotypes using data driven methods capable of identifying novel insights into mechanisms of divergent MH phenotypes among PWH. We leverage the analytic strengths of machine learning combined with inferential methods to identify novel MH phenotypes among PWH and the underlying explanatory features, with a particular interest in early life stress (ELS) as a determinant of MH phenotypes. Methods: 277 PWH (46% female, median age=44; 93% undetectable viral load [<50copies/mL]) were included in the analyses. Participants were enrolled in an observational community-based cohort residing in the Rakai region of Uganda. Participants completed the Patient Health Questionnaire (PHQ-9), Beck Anxiety Inventory (BAI), and the PTSD Checklist-Civilian (PCL-C). Hierarchical clustering was used to identify MH subtypes using total symptom scores on

CROI 2024 157