Multiply spliced HIV RNA is a predictive measure of virus production ex vivo and in vivo following reversal of HIV latency

Background: One strategy being pursued to clear latently infected cells that persist in people living with HIV (PLWH) on antiretroviral therapy (ART) is to activate latent HIV infection with a latency reversing agent (LRA). Surrogate markers that accurately measure virus production following an LRA are needed. Methods: We quantified cell-associated unspliced (US), multiply spliced (MS) and supernatant (SN) HIV RNA by qPCR from total and resting CD4+ T cells isolated from seven PLWH on ART before and after treatment ex vivo with different LRAs, including histone deacetylase inhibitors (HDACi). MS and plasma HIV RNA were also quantified from PLWH on ART (n-11) who received the HDACi panobinostat. Findings: In total and resting CD4+ T cells from PLWH on ART, detection of US RNA was common while detection of MS RNA was infrequent. Primers used to detect MS RNA, in contrast to US RNA, bound sites of the viral genome that are commonly mutated or deleted in PLWH on ART. Following ex vivo stimulation with LRAs, we identified a strong correlation between the fold change increase in SN and MS RNA, but not the fold change increase in SN and US RNA. In PLWH on ART who received panobinostat, MS RNA was significantly higher in samples with detectable compared to non0detectable plasma HIV RNA. Interpretation: Following administration of an LRA, quantification of MS RNA is more likely to reflect an increase in virion production and is therefore a better indicator of meaningful latency reversal. Funding: NHMRC, NIH DARE collaboratory.