Back

Optimal Thresholds for Cardiac CT Parameters in Patients with Aortic Stenosis Referred for Transcatheter Intervention

Mittal, T. K.; Sabarwal, K.; Ariff, B.; Mirsadraee, S.; Hothi, S. S.

2026-01-09 cardiovascular medicine
10.64898/2026.01.07.26343641 medRxiv
Show abstract

BackgroundComputed tomography (CT) is routinely performed for planning transcatheter aortic valve implantation (TAVI) procedures. Aortic valve calcium score (AVCS) is recommended to ascertain the severity of degenerative aortic stenosis (AS) in low-flow states, with the role of CT-AVA (aortic valve area) being uncertain. In this study, we determined the best AVCS and CT-AVA thresholds in patients referred for TAVI. MethodsA retrospective multicentre study evaluated 1162 patients undergoing TAVI CT for severe AS. Two inclusion criteria for normal cardiac output by transthoracic echocardiography (TTE) were considered: Group A: left ventricular ejection fraction [≥]50% and indexed stroke volume 28-48 ml/m2, Group B: cardiac index 1.9-4.3 L/min/m2. The predictive ability of CT parameters for severe AS (mean gradient [≥]40 mmHg) was assessed using ROC curves and optimal thresholds were determined. Results428 patients in Group A and 685 patients in Group B fulfilled the inclusion criteria. Best thresholds for AVCS and CT-AVA in group A (mean age 81{+/-}7.5 years; 54% women) were [≥]2034 AU (AUC 0.79) and [≤]0.90 cm2 (AUC 0.71) for women and [≥]3046 AU (AUC 0.78) and [≤]0.97 cm2 (AUC 0.69) for men; while in Group B (mean age 81{+/-}7.5 years; 45% women), they were [≥]1962 AU (AUC 0.80) and [≤]0.99 cm2 (AUC 0.73) in women, and [≥]2983 AU (AUC 0.78) and [≤]1.04 cm2 in men. ConclusionsThis study defines optimal AVCS thresholds for severe AS in a TAVI patient cohort, which are much higher than previously reported. CT-AVA could also be an additional parameter to establish AS severity.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.