Abstract Background Hypo-attenuating leaflet thickening (HALT) after transcatheter aortic valve replacement (TAVR) occurs in up to 25% of patients and is characterized by diverse temporal dynamics. Several studies investigated the natural history of HALT in transcatheter and surgical cohorts, but little is known about the imaging features of persistent thrombus. Purpose To examine the association of quantitative HALT volume and attenuation from computed tomography angiography (CTA) with its natural history. Methods We prospectively examined patients with HALT present on CTA at 30 days after TAVR and who underwent a follow-up CTA at one year. Hypoattenuation affecting motion (HAM) was defined as the presence of HALT with at least moderately reduced (>50%) leaflet motion. HALT was quantified by segmenting the inner volume of the bioprosthetic frame at the level of the leaflets and extracting voxels between a threshold of -200 to 200 HU. Thrombus volume, mean thrombus attenuation and thrombus heterogeneity measured with standard deviation of attenuation were automatically calculated (Figure 1). Results A total of 37 patients (70% men, 71±7 years old) undergoing TAVR with balloon-expandable SAPIEN3 (86%) and self-expanding EVOLUT (14%) valves were evaluated. Hypoattenuation affecting motion (HAM) at 30 days after TAVR was present in 29 (74%) patients. Patients with HAM had higher thrombus volume than patients without reduced leaflet motion (0.40cm3 [IQR 0.16 – 0.61cm3] vs 0.14 [IQR 0.06 – 0.19 cm3], p=0.009) but comparable mean attenuation values (145HU [IQR 141-151HU] vs 123 [IQR 114 – 141HU], p=0.079). Only 3 patients were switched to oral anticoagulation following HALT detection. At one year, HALT resolution was observed in 18/37 (48%) and HAM resolution in 19/29 (65%). No difference in quantitative thrombus features was noted between patients with persistent and resolved HALT. Notably, patients with persistent HAM had higher mean thrombus attenuation (141HU [IQR 123-156HU] vs 123HU [IQR 110 – 138 HU), p=0.045; Figure 2) and thrombus heterogeneity (58HU [IQR 51-63HU] vs 48HU [IQR 41- 54HU], p=0.021) than resolved HAM. Mean thrombus attenuation (area under the curve of 0.732, p=0.044) and thrombus heterogeneity (area under the curve of 0.763, p=0.022) were predictive of HAM resolution at one year. The optimal cutoff values as determined by Youden index was 140HU for mean attenuation and 52HU for heterogeneity. There was an increase in the mean transvalvular gradients between measurements at 30 days and one year in patients with persistent HAM as compared to patients with HAM resolution (Δ3mmHg [IQR -3 – 11mmHg] vs Δ-1mmHg [IQR -4 – 0mmHg], p=0.049). The volume of thrombus at one year was associated with the magnitude of change in transvalvular gradients (β = 10.3, 95% CI 1.6- 19.0, p=0.025). Conclusion Quantitative evaluation of HALT by CTA may aid the identification of patients at risk of persistent thrombus.
Read full abstract