Incremental Role of Cardiac CT Angiography in the Assessment of Left Ventricular Diastolic Function in Patients With Suspected Coronary Artery Disease With Normal Ejection Fraction: A Comparison With Transthoracic Echocardiography

ABSTRACT Objective Diastolic dysfunction (DD) is defined as impaired left ventricular (LV) relaxation, caused by structural or functional heart diseases. We sought to assess the role of cardiac CT angiography (CCTA) as a tool to evaluate LV DD in patients with normal EF using the diastolic expansion index (DEI), as compared to transthoracic echocardiography (TTE) as the gold standard. Methods Patients presenting with atypical chest pain with suspected coronary artery disease (CAD) and having a normal LV ejection fraction on TTE underwent CCTA using a dual source CT scanner. Assessment of diastolic function on CCTA was done using DEI, and it was compared with TTE‐derived diastolic function as per standard ASE/EACVI guidelines. DEI was measured using a dedicated postprocessing software (Syngo.via VB20, Siemens Healthcare, Germany) and was calculated as a ratio of maximum LV volume (in LV end‐diastole) to maximum left atrial (LA) volume (in LV end‐systole), with the mean value of readings from three independent readers taken for each calculation. Results The study included 141 patients (60 male; mean age 50.82 ± 10.20 years; mean heart rate 84.90 ± 16.15 beats per minute). CCTA‐derived mean indexed LA volumes correlated well with TTE measurements but were larger when compared to TTE ( r = 0.53, p < 0.0001). A progressive increase in LA volume index was noted with worsening DD grades. Excellent inter‐observer correlation was seen between the three observers for LA volume index (intra‐class correlation coefficient: 0.999 [95% CI: 0.999–0.999] and for DEI 0.996 [95% CI: 0.995–0.997]). A normal LV diastolic function was predicted by a DEI > 1.57 with a specificity and sensitivity of 88.3% and 81.6%, respectively (area under curve (AUC)‐0.87), whereas an abnormal LV diastolic function was predicted by DEI <1.46 with a specificity and sensitivity of 84.8% and 62.5%, respectively (AUC‐0.848). There was a significant correlation between the presence of obstructive CAD and an increase in calcium score with DD grade ( p = 0.006 and 0.001, respectively). Conclusion CCTA‐derived DEI is a reliable metric for assessing LV diastolic dysfunction in patients with suspected CAD with normal ejection fraction. The presence of obstructive CAD and an increase in calcium score correlated with worsening DD grade.