Streamlining the workflow: evaluation of ECG-gated computed tomography as a comprehensive imaging modality prior to intervention for mitral valve regurgitation.

BACKGROUND

ECG-gated computed tomography (CT) is not commonly utilized as the primary imaging modality for the diagnosis and management of degenerative mitral valve (MV) disease and preparation for MV intervention typically involves the use of several imaging studies. We hypothesized that CT can accurately identify MV pathology and exclude obstructive coronary artery disease (CAD) in this population potentially obviating the need for routine use of additional imaging modalities.

METHODS

CT was compared to invasive coronary angiography (ICA), transthoracic (TTE) and transesophageal (TEE) echocardiography for identification of coronary artery patency and mitral valve pathology respectively in patients undergoing MV evaluation between January 1, 2019, and December 31, 2023.

RESULTS

240 patients underwent TTE, TEE, and CT for mitral valve evaluation prior to intervention. Agreement was moderate when CT was compared to TTE for identifying prolapse (71%) and poor for flail segments (51%), with CT being superior. Compared to TEE, CT showed 96% agreement for prolapse (sensitivity 96.5%, CI 92.5-98.7% and specificity 82.3%, CI 71.2-90.5%) and 83% agreement for flail segments (sensitivity 75%, CI 67.8-81.8% and specificity 90.8%, CI 81.9-96.2%). Cohen's κ statistic was 0.81 and 0.65 for prolapse and flail respectively. 215 patients underwent ICA and CT evaluation followed by valve intervention. CAD prevalence was low at 6.0%. Compared to ICA, CT was 83.3% sensitive (CI 58.6-96.4%) and 97.9% specific (CI 94.9-99.4%) for CAD with excellent agreement (Cohen's κ = 0.81).

CONCLUSION

As compared to TEE, CT has excellent sensitivity and specificity for diagnosing prolapse and good agreement for flail segments. CT also shows strong agreement with ICA in evaluating CAD. CT provides a non-invasive alternative to TEE and ICA to accurately identify MV pathology and exclude significant CAD.