Roughly one fourth of a group of 510 HIV-positive adults had a signal of emphysema on quantitative chest CT scans in a cross-sectional U.S. study, while 29% had fibrosis-like lung changes. A higher viral load independently predicted fibrosis-like changes in the three-center Lung-HIV study, the results of which were published in the April 1 issue of JAIDS.
As HIV-positive people live longer with antiretroviral therapy, chest CT has documented non-AIDS pulmonary complications including chronic obstructive pulmonary disease and interstitial inflammation or fibrosis. But only a few studies have performed quantitative pulmonary CT analysis. Because a quantitative approach may be more sensitive in detecting correlations between lung disease and other measures and may detect lung abnormalities at an earlier stage, Lung-HIV researchers conducted this study.
The study focused on adults at three centers participating in the National Heart, Lung and Blood Institute Lung-HIV study. Participants could not have acute respiratory illness and were not required to have chronic pulmonary disease. All participants from one center smoked, while smoking was not an entry criterion at the other two centers. Participants completed standardized questionnaires and underwent pulmonary function testing and chest CT. One center analyzed CT images by segmenting the lung then quantifying the percentage of lung voxels associated with emphysema (>2.5% of voxels affected) or fibrosis (>5% of voxels affected). These voxel cutoffs indicate early disease-associated changes that are probably not yet clinically relevant.
Of the 510 study participants, 412 (81%) were men, 350 (69%) were taking antiretroviral therapy and age averaged 48.9 years. The study group was nearly evenly divided between whites (50%) and blacks (46%). Almost two-thirds of the group (64%) smoked at the time of the study, while 21% were former smokers. Almost one-quarter (24%) had injected drugs. Median CD4 count stood at 466 cells/mm3, median viral load measured 48 copies/mL and 61% had a viral load below 500 copies/mL.
There were 128 study participants (25.1%) with trace level or worse emphysema. Emphysema correlated with older age (Pearson correlation coefficient [PCC] 0.283) and with smoking pack-years (PCC 0.164), but not with current CD4 count or viral load. Four multivariable linear regression models found no association between quantitative emphysema and current CD4 count, viral load or antiretroviral use. Emphysema correlated positively with lung function as measured by percent predicted forced vital capacity (FVC%) (PCC +0.146) and negatively with forced expiratory volume in one second (FEV1)/FVC (PCC -0.301).
There were 150 study participants (29.4%) with trace level or worse fibrosis-like changes in the lung. Viral load correlated with fibrosis-like changes (PCC 0.210), while age had a significant negative correlation with fibrosis (PCC -0.239). In the same analysis, two lung function measures had significant negative correlations with fibrosis-like changes: FEV1% (PCC -0.133) and FVC% (PCC -0.173). In multivariable analysis, viral load and current smoking correlated positively with fibrosis-like changes, while age had a significant negative correlation.
Only six study participants (1.2%) had at least trace levels of emphysema and fibrosis-like changes.
The authors note that the 25% prevalence of at least trace level emphysema is not surprising in an HIV population, but the 29% prevalence of fibrosis-like changes did come as a surprise because "pulmonary fibrosis has been a relatively uncommon chronic lung disease in HIV-infected persons." The correlation between higher viral load and fibrosis-like changes -- after controlling for age, race, smoking and antiretroviral therapy -- is also a novel finding. The researchers suggest that areas of increased computed attenuation may indicate early interstitial lung disease or inflammatory changes. Such changes, they add, "are early events in the pathogenesis of other HIV-related pulmonary disorders."
Mark Mascolini writes about HIV infection.