Abstract Body

Background: Cotrimoxazole (CTX) prophylaxis reduces morbidity and mortality in HIV-infected children onlong-term ART, despite high-level pathogen resistance. Children taking CTX also have improved growth and reduced anemia, for reasons that are unclear. We hypothesized that benefits of CTX may be partly driven by reductions in inflammation. Methodology: 758 children/adolescents (median (IQR) age 7 (4,11) years) in the ARROW trial in Uganda/Zimbabwe were randomized to stop (n=382) vs continue (n=376) daily CTX after median(IQR) 2.1(1.8,2.2) years on ART.Inflammatory biomarkers (CRP, TNF-α, IL-6, sCD14) were measured by ELISA on cryopreserved plasma in a subset of 304 children at randomization (155 stop, 149 continue), then at 12, 24, and 96 weeks. Those co-enrolled in a crosssectional metabolic substudy also had albumin, total protein and lipid profile measured 3 years after ART initiation.Change in log10 cytokine levels after randomization were modelled using normal interval regression, adjusting for site performing the assay. Results: CRP was significantly higher in those randomized to stop CTX at 12, 24 and 96 weeks after ART initiation (2.22-fold [95% CI 1.23-4.01], p=0.008; 1.98-fold [1.26-3.14], p=0.003; and1.95-fold [1.10-3.46], p=0.02, higher, respectively; N=193, 260, 192 to date). There was a trend towards higher TNF-α in the stop CTX group at week 12 (1.16-fold higher vs continue [0.99-1.36], p=0.07), but this attenuated at 24 weeks (p=0.19) and had disappeared at 96 weeks (p=0.30). For sCD14 and IL-6, there was no evidence of a difference between stop vs continue CTX groups at week 12 (p=0.16 and p=0.17 respectively); by 24 weeks differences had widened (difference stop vs continue: sCD14 1.10-fold higher [1.00-1.21], p=0.06; IL6 1.35-fold higher [0.99-1.84], p=0.06), but were absent at week 96 (sCD14 p=0.91; IL-6 p=0.34). At a single measurement taken at median 48 (IQR 36-60) weeks after randomization in 169 children, albumin was also significantly lower in those randomized to stop CTX (difference -1.77 g/L vs continue [-2.97,-0.57] p=0.004), with non-significant trends towards higher total protein (+1.76 g/L vs continue [-0.13,+3.66] p=0.07) and LDL cholesterol (+0.17mmol/L vs continue [-0.04,+0.38] p=0.11). Conclusions: Children on long-term ART randomized to stop CTX showed a rise in inflammatory biomarkers compared to those continuing CTX, in particular with increases in CRP persisting for two years after stopping prophylaxis. Previously reported benefits of CTX, particularly for anemia and growth, may therefore be partly driven by reductions in chronic inflammation, either directly through the well-described immunomodulatory properties of CTX, or indirectly through modulation of microbial translocation.