Abstract 4364098: Long-term Growth Trajectories of Ascending Aortic Aneurysm Demonstrate Low Utility of Extended Imaging Surveillance.

Abstract

Background: Imaging surveillance for ascending thoracic aortic aneurysm (aTAA) relies on serial diameter measurements to guide management. However, measurement variability can lead to uncertain growth assessments and inaccurate risk stratification. Research Question: What is the prevalence and stability of distinct growth trajectories among patients with ascending aortic dilation undergoing routine surveillance? Methods: Single-center, retrospective study of 3,363 patients with ≥2 chest CT/MR angiograms. Latent Profile Analysis (LPA), an unsupervised clustering technique, was applied to double-oblique measurements of the mid-ascending aorta to classify growth trajectories based on each patient’s first 3 scans (“Early” n=1997) and first 5 scans (“Extended” n=757). We analyzed patient characteristics, outcomes (i.e., TAAD, repair) and rates of growth re-classification. Results: Median surveillance was 3.5 years, and the median aortic growth rate was 0.08 mm/y. Only 1.1% of patients met guideline criteria for repair based on growth rate alone. LPA at the Early time-point (3-scans) identified 4 classes: Stable (74%, median 0.05 mm/y), Growth (23%, median 0.58 mm/y), Dramatic Growth (2%, median 2.0 mm/y), and Non-physiologic (1%). For those at Extended surveillance, growth re-classification was observed in 38% of patients ( Figure 1 ). The Non-physiologic (noise-predominant) class grew from 1.3% to 8.3% between Early and Extended timepoints (p<0.001), with 59% originating from the Early Stable group. Most of the population was Stable at 5-scans, with this subgroup composed of 81% Stable and 18% Growth from the 3-scan time point. The Dramatic Growth class had smaller baseline diameters than the Stable class (34.0 vs 41.0 mm, p<0.001). Re-classification from Stable to Dramatic Growth was rare (<0.3%) and associated with Marfan syndrome (p=0.006). Among the patients that experienced type A dissection, the median pre-dissection diameter was 42.0 mm and 50% had been classified as Stable. Conclusion: A large majority of patients with ascending aortic dilation have indolent disease and outside of heritable causes, extended surveillance shows little added value for identifying those with rapid growth or at risk for type A dissection. Our findings raise question of the utility of life-long imaging surveillance for initially stable patients who are far from repair size thresholds and without high-risk features.