Sex-based Differences in Outcomes Related to Thoracic Aorta Dimensions

See also the article by Rueda-Ochoa and Bons et al in this issue.

Download as PowerPoint

Despite advances in detection and treatment, thoracic aortic dissection remains an important cause of morbidity and mortality. Identifying patients at risk for acute thoracic aortic dissection is important because they are usually asymptomatic prior to presentation, and prehospital mortality is high. Notably, outcomes for women with thoracic aortic dissection are worse than those for men, with less frequent diagnoses within 24 hours of symptom onset and higher in-hospital mortality (1). The thoracic aorta can be visualized on every chest CT image. Thus, there is an opportunity to help detect patients at risk who undergo chest CT for other reasons. However, age, sex, body habitus, and cardiovascular risk factors can all impact aortic dimensions (2–5), and the clinical implications of such assessments remain uncertain.

In this issue of Radiology, Rueda-Ochoa and Bons et al (6) assessed the clinical implications of thoracic aorta diameter in 2178 patients who underwent CT as part of the Rotterdam Study. The patients had a mean age of 69 years ± 7 (SD) and 55% were women (n = 1186). The Rotterdam Study is a prospective population-based cohort study of people older than 45 years that has been ongoing since the 1990s (7). This study reports on a subset of the 15 000 participants of the Rotterdam Study who underwent noncontrast electrocardiogram-gated cardiac CT as part of the study protocol and were followed up over a 9-year period. The authors have previously reported important sex-based differences in the thoracic aorta dimensions in this population. In particular, the authors noted the presence of an ascending aorta diameter greater than 40 mm in 19% of men and 6% of women (2). In the current article (6), they present the clinical implications of the ascending and descending thoracic aorta diameters in terms of all-cause mortality, cardiovascular mortality, stroke, coronary heart disease events (including fatal or nonfatal myocardial infarction, other coronary heart disease mortality, coronary artery bypass grafting, or percutaneous transluminal angioplasty), and heart failure.

For the ascending thoracic aorta, in models corrected for multiple cardiovascular risk factors and coronary artery calcification, they found that body mass index dimensions were an independent predictor of cardiovascular mortality (hazard ratio [HR], 1.33; 95% CI: 1.03, 1.73; P < .03) and all-cause mortality (HR, 1.22; 95% CI: 1.07, 1.38; P < .002) for women, but only all-cause mortality for men (HR, 1.30; 95% CI: 1.11, 1.51; P < .001). Similar findings were observed for the descending thoracic aorta measurements. The only body mass index–adjusted dimension associated with a specific cardiovascular outcome was the descending thoracic aorta, which was associated with stroke in women (HR, 1.38; 95% CI: 1.07, 1.78; P < .01). Interestingly, nonindexed ascending thoracic aorta dimensions were associated with heart failure in women and stroke in men, and nonindexed descending thoracic aorta dimensions were associated with stroke in men. However, no ascending or descending thoracic aorta dimensions were associated with coronary heart disease events in either men or women.

It is well established, with a variety of imaging modalities, that women have smaller blood vessels than men do, even when adjusting for body size. However, the clinical implications of this finding are not well established. Large population-based studies, such as the Framingham Heart Study (8), have shown that the aortic root diameter measured with echocardiography is associated with a variety of cardiovascular events. In 3318 patients followed over a 9-year period, the Framingham Heart Study found that descending but not ascending thoracic aorta diameter at CT was associated with composite cardiovascular disease events. These events included cardiovascular disease death, myocardial infarction, coronary insufficiency, heart failure, or stroke (9). Similarly, in a smaller case-control study, the ascending and descending thoracic aorta diameters at CT were associated with a composite cardiovascular disease end point (10). The end points used in these previous studies represented a composite of different outcomes, so it is useful in the current article (6) that these outcomes are reported separately. However, the mechanisms behind the associations found are uncertain. For example, is the association of body mass–indexed dimensions with all-cause mortality due to a general change in the vasculature or correlation with another unknown etiology? This article is one of the few population-based studies to report on the clinical implications of thoracic aorta dimensions. To my knowledge, it is also the first study to assess sex-based differences in thoracic aorta diameter as a risk factor for a variety of separate cardiovascular outcomes.

An important limitation of this study was that the CT scans were not obtained with contrast material. The outer-edge to outer-edge method used in this study to assess the aorta leads to an overestimation of diameters compared with contrast-enhanced methods using the inner edge. Also, the arch of the aorta was not covered on these CT scans because they were focused on the heart (principally performed to assess coronary artery calcification). Aortic arch dimensions, geometry, blood flow, and atherosclerotic plaque are all potential predictors of cardiovascular events, particularly stroke. However, none of these potential predictors were assessed in this study. The HRs presented in this article were adjusted for coronary artery calcification, which has been shown to be a key predictor of a variety of cardiovascular events. The results were not adjusted for the presence of atherosclerotic plaque in the aorta, which has also been associated with a variety of cardiovascular events. The lack of association of the ascending or descending aorta dimensions with coronary artery disease outcomes in this study may be due to the variety of soft outcomes that were combined, including fatal or nonfatal myocardial infarction, other coronary heart disease mortality, coronary artery bypass graft, or percutaneous transluminal angioplasty. Alternatively, it may confirm that these coronary and aortic diseases are distinct despite sharing similar risk factors, as suggested by previous studies.

Noncontrast chest CT is widely performed for a variety of conditions and will become even more common with the increasing use of CT for lung cancer screening. However, unlike the electrocardiogram-gated CT performed in this study, lung cancer screening CT is performed without electrocardiogram gating and with a low radiation dose. Both factors can affect image quality and the accuracy of aortic measurements. Therefore, translating the findings of this study into clinical practice will require further research. Of several available age- and sex-based cutoff values, it is currently unclear which should be used to identify patients who would benefit from follow-up imaging or surgical referral. Also, a key remaining question is what is the most appropriate method to index aortic measurements—body mass index, body surface area, height, or other methods? A variety of different indexed and nonindexed dimensions have been reported in different studies. Other factors may also be important for predicting outcomes in patients with an enlarged thoracic aorta: rate of growth, aorta geometry, PET measures of disease activity, perivascular fat attenuation, and radiomic characteristics of the aortic wall or plaque. Thus, a variety of other thoracic aorta parameters need to be investigated to better understand the natural history in thoracic aortic disease. Future automation of the measurement of aortic dimensions at chest CT with machine learning techniques will hopefully aid rapid routine assessments.

This study (6) has established that thoracic aorta diameters from noncontrast CT are independent predictors of all-cause mortality in men and women, but of cardiovascular mortality and stroke only in women. Notably, this was independent of other key cardiovascular risk factors, including coronary artery calcium score and body mass index. The sex-based differences in outcomes highlighted in this study will be important for future research and clinical management. Further work is needed to help translate these findings into management recommendations for individual patients. Unanswered essential questions remain, including which measurements we should perform and which patients benefit from further assessment.