Switching from efavirenz to rilpivirine improves sleep quality and self-perceived cognition but has no impact on neurocognitive performances

Background: Efavirenz (EFV) association with neurocognitive impairment is debated. Whether switching away from EFV improves neurocognitive performances is still controversial.Methods:In a randomized open-label controlled trial, patients under effective treatment with tenofovir disoproxil-fumarate (TDF), emtricitabine (FTC) and EFV, who had altered neurocognitive assessment (z-transformed score below -1 in at least one cognitive domain), depression, anxiety or low sleep-quality, were randomized 1:1 to immediate or delayed (24-weeks) switch to TDF/FTC/rilpivirine (RPV). Treatment efficacy, neurocognitive function, symptoms and quality of life were evaluated 12, 24 and 48 weeks after randomization.Findings:Seventy-four patients were randomized to immediate (36 patients) or delayed switch (38 patients). At baseline, 63 and 25% of patients had z-scores below -1 in at least one or two neurocognitive domains, 31.1, 17.6 and 44.6% had significant depression or anxiety symptoms or low sleep quality. At week 24 (primary end-point), overall neurocognitive improvement was observed, with no statistically significant differences between arms, neither considering the global z score (between arms difference +0.1; P=0.458), nor domain-specific z scores. Patients switching away from EFV had significant greater improvement of sleep quality index (between-arm difference -1.5; P=0.011), self-reported cognitive failures (-6.2; P=0.001) and CNS symptoms score (-5; P=0.002), but not of anxiety or depression. No protocol defined virological failure, grade at least 3 lab abnormalities or drug-related serious adverse events were reported.Conclusion:Our results do not support the hypothesis that switching to RPV improves cognitive function in patient under stable treatment with EFV. Nonetheless, improvements in neuropsychiatric symptoms, sleep quality and self-perceived cognition were observed.