Abstract Rationale Immune restoration following antiretroviral therapy (ART) in people with HIV (PWH) and pulmonary tuberculosis (TB) can adversely affect lung function. Although HIV increases TB risk, PWH exhibit less long-term pulmonary impairment than those without HIV, suggesting an immunomodulatory effect of HIV. Studies characterizing the evolution of radiographic patterns in PWH and mycobacterial pulmonary disease are lacking. Methods We conducted a secondary analysis of two prospective cohorts of ART-naive PWH with CD4100 cells/µL, identifying 47 participants with mycobacterial pulmonary disease (TB or NTM) with the objective of comparing qualitative radiographic findings between two groups: those with mycobacterial IRIS (“IRIS group”) and those without mycobacterial IRIS (“non-IRIS group”). IRIS was defined using established clinical criteria. Qualitative evaluation of chest CT was performed at three timepoints: before initiation of ART (“pre-ART”), 6-8 weeks after starting ART (“IRIS or equivalent timepoint”), and 96 weeks after starting ART (“96-week timepoint”). Two pulmonologists (M.C, S.G), blinded to IRIS status of participants, independently evaluated the CT scans, with discrepancies resolved by consensus. Five radiographic patterns were identified, representing a spectrum from favorable to unfavorable: resolved, improved, mixed, persistent, and worse (Figure 1). Radiographic patterns were defined as: resolved - complete resolution of abnormalities; improved - decreased but residual findings; mixed - improvement in some areas with new or worsening lesions in other areas; persistent - no change; and worse -progression without improvement. Results There were no differences in the baseline clinical characteristics (age, sex, smoking, opportunistic infection) between the IRIS and non-IRIS groups. No difference was found when comparing CD4 and CD8 count between the two groups at the three timepoints. Ordinal logistic regression was performed to assess differences in radiographic patterns between IRIS and non-IRIS groups. When comparing scans between baseline and the IRIS (or equivalent) timepoint, unadjusted analysis showed that patients with IRIS were more likely to have unfavorable radiographic pattern (OR = 3.98; 95% CI 1.2 to 12.23; p = 0.016). After adjusting for age, sex, OI and smoking, similar results were observed (aOR = 4.98; 95% CI 1.3 to 17.9; p = 0.014). Comparison of scans between IRIS (or equivalent) and 96-week timepoints showed no significant differences between the groups (aOR=2.29, 95% CI -2.4 to 4, p = 0.6). Most patients (91%) demonstrated favorable radiographic patterns at 96 weeks. Conclusions PWH and mycobacterial IRIS demonstrate unfavorable radiographic patterns during immune reconstitution but not at 96 weeks, highlighting that IRIS drives transient lung injury that resolves over time. This abstract is funded by: National Institutes of Allergy and Infectious Diseases
Chaturvedi et al. (Fri,) studied this question.