Abstract
AbstractHIV associated neurocognitive disorder (HAND) remains an important HIV-associated comorbidity despite antiretrovirals (ARVs). Cerebrospinal fluid (CSF) escape/discordance is now recognised in the context of individuals with a reconstituted immune system with an estimated prevalence of 10%. However, the contribution of CSF escape/discordance to HAND remains uncertain. Furthermore, a latent reservoir of HIV in the brain has implications for lasting cure strategies. Little is known about the prevalence of CSF escape/discordance amongst people living with HIV (PLWH) in sub-Saharan Africa (SSA).We conducted a longitudinal cohort study of PLWH who were 18 years or older on ARVs for at least 1 year who reported neurocognitive complaints. We obtained paired CSF and blood at baseline, 6, 12 and 24 months. Viral load (VL) testing was done with the Abbott m2000 RealTime System. HIV genotyping was done by Sanger sequencing and next generation sequencing (NGS) by Illumina MiSeq. Resistance calling was done using Stanford HIV drug resistance database. Random drug levels were done on plasma and CSF using mass spectrometry.We present the results at baseline. Seven hundred and eight adult PLHIV attending a HIV treatment centre were screened using the Simioni symptom questionnaire and in addition asked “do you have a chronic, persistent headache?” Fifty-nine PLHIV answered yes to at least one of the screening questions and were considered for enrolment. Thirty consented to participate. The median age was 37.6 (IQR 33.2 to 48.3) years. The majority were women (98.0%, 28/30). Headache was the most common symptom (93.3%), then memory impairment (56.4%), attention deficit (48.3%) and impairment in executive functioning (46.7%). All participants had Karnofsky performance scale > 70% and were able to perform their activities of daily living independently. Symptoms of depression were common, with 82.8% scoring a CESD-R-10 >10. The median duration of ART was 9.9 (IQR 5.7 to 11.9) years. 72.4% (n=21/29) were on tenofovir/emtricitabine/efavirenz. The rest were on second line ARVs (ritonavir boosted lopinavir plus zidovudine/lamivudine or tenofovir/emtricitabine). The median nadir CD4 count was 193 (IQR 98 to 301) cells/mm3and the current median CD4 count was 547 (IQR 384 to 856) cells/mm3. At baseline 86.2% had an undetectable plasma HIV viral load (<40 copies/ml) (25/29). Eighteen participants had paired CSF and blood successfully sampled. Of these 4 had detectable virus in the blood with VL ranging from 82 to 38,992 copies/ml. Two participants had CSF escape/discordance (9.0% (2/22) and 2 others had detectable VL in CSF but lower than the VL in blood. We found an association between detectable CSF VL and viraemia (p 0.001) and boosted protease inhibitor (PI) based ART (p 0.02). All participants who had undetectable VL in blood and CSF had detectable blood and CSF drugs levels corresponding to their prescribed ARVS. Two participants with detectable VL in the CSF had drug levels measured, which was below the limit of detection of the drug assays in both blood and CSF. Both participants were on second line boosted PI based ARVs. Viral sequencing revealed NNRTI resistance mutation G190A detected in both blood and CSF in the participant with CSF discordance and NRTI M184V, NNRTI K103N and P225H in the blood but not CSF of the second participant. No minority variants were identified below 20% by NGS.Virological failure appears to be driven by poor treatment adherence. The results suggest that the prevalence of CSF escape/discordance in HIV positive neurosymptomatic persons is consistent with previously published prevalence from resource rich settings however this needs to be explored in the larger study. Symptoms of depression were common and may bias self-reported neurocognitive impairment and needs to be explored further. This ongoing longitudinal study will also investigate the evolution of drug resistant variants in CSF and the relationship with plasma viral quasispecies.
Publisher
Cold Spring Harbor Laboratory