Lung function tracking in children with perinatally acquired HIV following early antiretroviral therapy initiation

Introduction Lung disease remains a frequent complication in children with perinatal HIV infection (CHIV) and exposure without infection (CHEU), resulting in diminished lung function. In CHIV, early antiretroviral therapy (ART) initiation improves survival and extrapulmonary outcomes. However, it is unknown if there is benefit to lung function. Methods Cohorts of CHIV (ART initiated at median 4.0 months), CHEU and HIV-unexposed children (CHU) prospectively performed pulmonary function testing (PFT) consisting of spirometry, plethysmography and diffusing capacity from 2013 to 2020. We determined lung function trajectories for PFT outcomes comparing CHIV to CHU and CHEU to CHU, using linear mixed effects models with multiple imputation. Potential confounders included sex, age, height, weight, body mass index z-score, urine cotinine and Tanner stage. Results 328 participants (122 CHIV, 126 CHEU, 80 CHU) performed PFT (ages 6.6–15.6 years). Spirometry (forced expiratory volume in 1 s, FEV1, forced vital capacity (FVC), FEV1/FVC) outcomes were similar between groups. In plethysmography, the mean residual volume (RV) z-score was 17% greater in CHIV than CHU (95% CI 1% to 33%, p=0.042). There was no difference in total lung capacity (TLC) or RV/TLC z-scores between groups. Diffusing capacity for carbon monoxide was similar in all groups, while alveolar volume (VA) differed between HIV groups by sex. Conclusion Our study indicates that early ART initiation can mitigate the loss of lung function in CHIV with lasting benefit through childhood; however, there remains concern of small airway disease. CHEU does not appear to disrupt childhood lung function trajectory.

Supplementary Figure 3: Effect display for the interaction of HIV group and age in the linear mixed model for TLC z-score This figure is a graphical depiction of the terms from the linear mixed model, after adjusting for age, height, weight, BMI z-score, Tanner stage, and the interaction between HIV group and age.The figure has been constructed using the Effects package in R, and therefore the analysis only reflects the fixed effects of the model (not random effects) and assumes that other variables in the model are constant and take on their average values.The graphs are constructed using one random dataset from a set of 10 datasets constructed using multiple imputation, and the results do not differ across imputations.95% Confidence intervals are shown, with standard errors computed from the covariance matrix of the fitted regression coefficients.BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)  BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Effect display for the interaction of HIV group and gender in the linear mixed model for VA z-score This figure is a graphical depiction of the terms from the linear mixed model, after adjusting for gender, height, weight, cotinine group, and the interaction between HIV group and gender.The figure has been constructed using the Effects package in R, and therefore the analysis only reflects the fixed effects of the model (not random effects) and assumes that other variables in the model are constant and take on their average values.95% Confidence intervals are shown, with standard errors computed from the covariance matrix of the fitted regression coefficients.Predicted mean value of VA z-score: HIV group by gender effect plot 06 (-0•53 to 0•41) 0•14 (-0•05 to 0•32) Association between each HIV group and each lung function outcome using linear mixed effects models.The models are based on three datasets: Black children only; Mixed race children only; and the combined dataset.All results are after adjusting for relevant confounders (p<0.20) and after multiple imputation.Only the results for the HIV group variable are displayed, with results showing the mean difference (with 95% CI) in z-score between the exposed group (CHIV or CHEU) and the unexposed group, CHU.Results highlighted in bold indicate significant results.

Table 3 :
FEV1Association between each HIV group and FEV1 z-score using linear mixed effects model.Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Thorax doi: 10.1136/thorax-2023-220197 -1239.Results show the adjusted differences in z-scores between HIV groups.The model with random intercept only was optimal.Multi-predictor analysis is shown after adjusting for sex, age and BMI z-score.CHU: Children who are HIV unexposed; CHEU: Children who are HIV exposed uninfected; CHIV: Children with perinatally-acquired HIV BMJ

Table 4 :
FVCAssociation between each HIV group and FVC z-score using linear mixed effects model.

Table 5 :
FEV1/FVC z-score Association between each HIV group and FEV1FVC z-score using linear mixed effects model.Unadjusted estimates show single-predictor analysis of each variable on FEV1FVC z-score.Adjusted results show the mean difference (with 95% CI) in FEV1FVC z-score between the exposed group (CHIV/CHEU) and the unexposed group, CHU.

Table 6 :
TLC z-score Association between each HIV group and TLC z-score using linear mixed effects model.Unadjusted estimates show single-predictor analysis of each variable on TLC z-score.Adjusted results show the mean difference (with 95% CI) in TLC z-score between the exposed group (CHIV/CHEU) and the unexposed group, CHU.

Table 7 :
RV z-score Association between each HIV group and RV z-score using linear mixed effects model.Unadjusted estimates show single-predictor analysis of each variable on RV z-score.Adjusted results show the mean difference (with 95% CI) in RV z-score between the exposed group (CHIV/CHEU) and the unexposed group, CHU.Results show the adjusted differences in z-scores between HIV groups.The model with random intercept only was optimal.Multi-predictor analysis is shown after adjusting for sex, age and height.CHU: Children who are HIV unexposed; CHEU: Children who are HIV exposed uninfected; CHIV: Children living with perinatally-acquired HIV BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)