Arriving Now at the Crossroads of Technology: Sex and Gender Differences in COPD

Phenotypic heterogeneity is an oft cited reason for non-replication of epidemiologic and genetic findings in COPD studies. As COPD includes emphysema, small airways disease and chronic bronchitis, there is inherent heterogeneity when relying solely on spirometry for phenotyping. The COPDGene study (1) has addressed this issue directly by including chest computed tomography(CT) and spirometry in the careful phenotyping of all enrolled subjects. The assessment and “reading” of such a large number of CT scans has evolved from the radiologists’ workspaces to output from a highly automated quantitative assessment software. As a result of computer algorithms and batch analysis approaches, the number of high-resolution high-fidelity pulmonary phenotypes has multiplied, allowing for a variety of approaches to dissecting the anatomical features of lung disease.

In a comprehensive review, Drs. Becklake and Kauffmann presented a summary of sex and gender difference in airway biology over the human life course. In this missive they suggest that “the differential influence of sex and gender on airway behavior should always be taken into account and the consequences of these gender differences in airway behavior on the occurrence of airway diseases, acute and chronic , should be the target of investigation.” (2) The article by Kim and colleagues in this issue of COPD is such an investigation. The authors used imaging software to analyze CT scans from 2047 current and former smokers, and assessed the sex specific differences for quantitative measures of segmental, subsegmental and subsubsegmental bronchi. For each airway, the authors compared luminal area, inner diameter, wall thickness and wall area percent (WA%) between male and female subjects. They concluded that women with a history of smoking had higher WA%, but lower airway thickness, luminal area and internal diameter in airways matched to the CT findings of male smokers. This work is novel in both its size and scope, as it includes systematically acquired CT scans assessed by a computer algorithm for thousands of subjects. However, as suggested in prior studies, it is unclear whether these CT findings are just statistically significant or clinically meaningful—the male-female differences are generally small (around 1%).

Appropriately, the authors convey that although the sex/gender differences in this analysis are small, “gender is just one factor for airway wall” and “that a complex background of airway dimensions exist.” Interestingly, in a detailed analysis of CT scan findings integrated with lung histology, Martinez and colleagues demonstrated that the airways of women had thicker walls(3)—a different finding than presented here and potentially more consistent with the apparent higher prevalence of chronic bronchitic type symptoms in women compared to men.

We need to be circumspect about the findings highlighted by Kim and colleagues, as they may be due to normal baseline sex-specific anatomical variations in the airways.

As such, we must ask a series of questions:

1. What is the within and between sex variation in quantitative CT phenotypes that reflect normal variation in lung and airway development?

2. What are the age-related variations in these phenotypes and how will difference in longitudinal CT phenotypes vary by sex with aging?

3. What are the parameters of statistical noise germane to these measures?

4. What is the impact of current and former smoking on these measurements?

The list of questions could go on but the real crux is that to better apply these phenotypes to lung research we need a catalogue of normal values for quantitative CT metrics in non-smokers. This is the kind of atlas that is needed to both catalogue normal within and between sex variation in airway dimensions, as well as to identify sex differences in COPD.

High throughput quantitative CT analysis currently has it main application for research, but the integration of objective CT scan phenotypes with genetic profiles (such as those being generated in the COPDGene study) could have future relevance for sex-specific personalized medicine approaches to lung disease. The article by Kim and the COPDGene investigators provides further support that studying sex-specific manifestations of COPD is highly productive and rewarding, as it not only leads to new insights, but inevitably highlights the panoply of unanswered questions that remain regarding sex and gender-based features of COPD.