Measurement of Aortic Luminal Diameter Challenging Using Systematic Workflow for CTA

The measurement of the aorta’s luminal diameter after type A aortic dissection (TAAD) was found to be challenging when using a systematic workflow for measurements based on computed tomography (CT) imaging, according to study results published in the Journal of Cardiac Surgery.

Patients who have experienced TTAD need to be monitored to assess the occurrence of aortic aneurysmal degeneration of the native aorta. This can be achieved using reproducible serial CT angiography (CTA) measurements that allow clinicians measure aortic dimensions and assess potential remodeling using the aorta’s luminal diameter — the current gold standard — as well as area and volume measurements at several locations.

The CTA scans of 20 patients (n=40) with TAAD (mean age, 60.4±12.1 years; 50% men) from the University of Michigan cardiac surgery database were retrospectively analyzed. In the study, 2 independent observers assessed aortic geometry in all scans prior to and 3 months after surgery. Inter- and intra-observer agreement scores were calculated and analyzed with the concordance correlation coefficient (CCC).

A center line of the entire aorta — including the true and false lumens — was drawn manually and luminal diameter (mm) and area (cm²) were measured along this line at 6 points (mid ascending aorta, mid aortic arch, proximal descending aorta, mid descending aorta, distal descending aorta, and infrarenal aorta). Volumetric measurements of the descending thoracic aorta were performed for the true and false lumens.

The mean post-surgical follow-up duration was 104.0±24.5 days. Measurements of the luminal diameter, area, and volume had acceptable interobserver agreement, with area scores achieving the highest agreement and yielding CCCs that ranged from 0.909 to 0.984. Agreement scores were lower for diameter vs area, with the lowest reproducibility observed for the mid aortic arch (CCC <.886). Volumetric agreement was moderate, with CCC ranging from 0.908 to 0.941. Volume measurements performed worse in general than measurement of diameter and area.

Overall, intraobserver scores (combined CCC, 0.894) were higher than interobserver scores (combined CCC, 0.881), with less variability for luminal diameter (mean difference standard deviation [SD], 1.89 vs 1.94, respectively) and luminal area (SD, 0.61 vs 0.66, respectively) measurements. Although no correlation was established between measurement variability and ellipticity (P =.225), ellipticity was found to be negatively correlated with interobserver agreement (correlation factor, R −0.693; P =.026).

Study limitations include its small sample size, the inability to account for false lumen contrast enhancement differences, lack of assessment of the relationship between false lumen growth and clinical events, and possible human error due to manual segmentation.

“The addition of luminal area and possibly volumetric measurements to the standard diameter-based assessment of aortic dimensions in patients with TAAD may significantly

improve the reproducibility of aortic growth measurements, and therefore alter clinical decision making in specific cases,” noted the authors.

Disclosures: Dr Himanshu J. Patel was supported by the Joe D. Morris Collegiate Professorship, the David Hamilton Fund, and the Phil Jenkins Breakthrough Fund.

Reference

Houben IB, Bakel TMJ, Burris NS, Moll FL, Herwaarden JA, Patel HJ. Critical appraisal of multidimensional CT measurements following acute open repair of type A aortic dissection. J Card Surg. February 2020:1-11. doi:10.1111/jocs.14446