Risk of immune reconstitution inflammatory syndrome with integrase inhibitors versus other classes of antiretrovirals: a systematic review and meta-analysis of randomised trials

BACKGROUND: Integrase strand transfer inhibitors (InSTIs) decrease HIV plasma viral load faster than other antiretroviral classes. More rapid viral load decline has been associated with higher risk of immune reconstitution inflammatory syndrome (IRIS). There are conflicting reports on the association between InSTI and IRIS. We performed a systematic review and meta-analysis to compare the risk of IRIS among treatment-naïve HIV-positive patients starting InSTI versus non-InSTI regimens. METHODS: We searched PubMed, Scopus, Web of Science, Africa-Wide, and Cochrane databases from earliest available date to 26 November 2021, for randomised controlled trials (RCTs) having intervention arms with InSTI versus control arms without InSTI in patients initiating first-line antiretroviral therapy. The primary outcome was relative risk (RR) of IRIS, while the secondary outcome was RR of paradoxical tuberculosis-associated IRIS (TB-IRIS). Data were combined by random-effects meta-analysis according to the Mantel-Haenszel method. The protocol for this study is registered with PROSPERO, CRD42020213976. RESULTS: We included 14 RCTs comprising 8696 participants from six continents for the primary outcome of IRIS, and a subset of 674 participants (from three RCTs) for the secondary outcome of paradoxical TB-IRIS. Risk of IRIS was similar between InSTI and non-InSTI regimens (RR, 0.93, 95% confidence interval [CI], 0.75 – 1.14). There was a trend towards a lower risk of paradoxical TB-IRIS with InSTI versus efavirenz regimens that was not statistically significant (RR, 0.64, 95% CI, 0.34 – 1.19). CONCLUSIONS: In this meta-analysis among treatment-naïve patients commencing first-line antiretroviral therapy, InSTI regimens were not associated with higher risk of IRIS.