Abstract Body

Background:

Children living with HIV (CLWH) are at high risk of developing tuberculosis (TB). Here we compare clinical outcomes between CLWH vs HIV-uninfected children with non-severe TB in the SHINE trial; and describe viral suppression among CLWH

Methods:

SHINE was an open-label, randomized controlled trial conducted in 3 African countries and India. Children aged < 16 years of known HIV status, with smear-negative, non-severe TB were randomized to receive 4 vs 6 months anti-tuberculosis treatment (ATT) and followed for 72 weeks. Among CLWH, CD4 and viral load (VL) were measured at 24 and 48 weeks. Logistic regression and Cox proportional hazards models were fitted to investigate predictors of mortality, hospitalization and VL suppression

Results:

Of 1204 enrolled, 127(11%) were CLWH and of similar age (median[IQR] 3.6 [1.2, 10.3] vs 3.5[1.5, 6.9] years), but more underweight (WAZ;-2.3 vs -1.0, p< 0.01) and anemic (hemoglobin 9.5 vs 11.5 g/dl, p< 0.01), and fewer had bacteriological confirmation of TB (6 vs 15%, p=0.01), compared to HIV-uninfected children. Baseline median CD4% and counts were 16% (10-26) and 719 (241-1134) cells/mm3, and 68(54%) were ART-naïve. The SHINE trial found no difference in TB outcomes between 4 vs 6 months of treatment by HIV status (NEJM 2022). There were 31(3%) deaths during the trial, many occurring early (median time to death; 4.5 vs 6.5 months in CLWH vs HIV-uninfected). Mortality risk in the trial was higher in CLWH (aHR [95%CI] 2.6 [1.2,5.8]). Age < 3 years (aHR 6.3 [1.5,27.3]), malnutrition (aHR 6.2 [2.4,15.9]) and anemia (Hb < 7g/dl) (aHR 3.8 [1.3,11.5]) independently predicted mortality. Overall, CLWH were more likely to be hospitalized (aOR=2.4 [1.3-4.6]) compared to HIV-uninfected children. Viral suppression was suboptimal with 45% and 61% with VL < 1000 copies/ml at week 24 and 48. Children initiating EFV-based ART were more likely to have VL suppression vs LPV/r-based by week 48 (71 vs 50%, p=0.056). There was no effect of randomized duration of treatment (4 vs 6 months ATT) on mortality and hospitalization by HIV-status

Conclusions:

In children with non-severe TB, deaths occurred early during ATT with higher mortality in CALWH. Young age, malnutrition and anemia independently predicted mortality in both CALWH and HIV-uninfected children. Suboptimal viral suppression was common among CLWH, with worst suppression on LPV/r. There was no evidence that CALWH needed longer ATT than HIV-uninfected children, and therefore can receive 4 months ATT for non-severe TB