Incidence and Clinical Parameters of HIV-1 Paradoxical and Unmasking Immune Reconstitution Syndrome in Antiretroviral Naïve Pregnant Women attending selected facilities in Kenya

Background Following ART initiation, a spectrum of HIV-associated immune reconstitution inflammatory syndrome (IRIS), as opportunistic infections occurs, presenting as either paradoxical or unmasking IRIS. There are benefits of ART for maternal health and prevention of perinatal transmission, however, some risks may be possible with IRIS. The study sought to estimate the incidence, compare survival times to, and investigate the baseline clinical predictors of HIV-1 paradoxical and unmasking IRIS as well as identify the homogeneity defined by combinations of specific covariates in Art Naïve Pregnant Women. Methods A prospective, active records observational study [P609/08/-2018-E] was conducted among ART-naive pregnant women attending antenatal care unit (ANCu) from Kenyatta National and Mbagathi Hospitals in Kenya. Participants enrolled were between the ages of 20 - 49 years with a clear documentation of HIV status and, an independent review committee adjudicated IRIS diagnosis. Associations between baseline pre-ART characteristics, biomarkers, and IRIS diagnosis were assessed with Cox models. Decision-tree analysis was finally performed to identify homogeneity defined by combinations of specific parameters. Results Of the initial total 532 women, (24.8%) (n=133) participants experienced IRIS by the 12 th week post ART initiation, 27.3 % (n= 36) and 72.7% (n = 96) paradoxical and unmasking respectively. Maternal Basal Metabolic Index (MBMI) of (25 -29.9) significantly predicted unmasking IRIS [(β)=0.907, Wald test (β^) = 6.550, (HR = 2.478, 95% C.I 1.237 – 4.965, P = 0.010] same case to gravidity of above 5 [(β)=0.743, P = 0.338] while that of 2-3 predicted paradoxical IRIS [(β)= - 0.542, P = 0.037. WHO-HIV clinical stage 1 and 2 showed a link towards paradoxical IRIS, [(β)=- 0.111 and (β)= - 0.276 (P < 0.05)], clinical stage 4 with unmasking IRIS [(β)= 0.047, HR = 1.048, P = 0.941]. CD4 count > 500 cells/mm 3 skewed towards unmasking IRIS diagnosis [(β)= 0.192, HR = 1.211, P = 0.416], while RNA-HIV viral loads > 50 copies/ml towards paradoxical IRIS [(β)= - 0.199, HR = 0.820, P = 0.360. On decision tree analysis, 76% (P = 0.729) of ART naïve pregnant women aged 20-29 and 40-49 years of age presented with unmasking IRIS and a gravidity of 4-5 and 1 predicted 88% (P = 0.045) unmasking IRIS as compared to that of 2-3 62 % (P = 0.045). Conclusion Unmasking IRIS was the common type of IRIS post-ART initiation. Maternal BMI and RNA-HIV viral loads level predicted paradoxical IRIS while the baseline CD4 count and WHO-HIV clinical infection stage 1 and 4 were associated with unmasking IRIS. Gravidities of 1 and 4 – 5 predicted unmasking IRIS.