AIMS:

The haemodynamic changes due to severe aortic stenosis raise concerns on the reliability of both hyperaemic and resting pressure-wire indexes. This study sought to provide a comprehensive assessment of epicardial and microvascular function by means of single-view angiography-derived physiology in patients with aortic stenosis undergoing TAVI.

METHODS AND RESULTS:

Coronary vessels with pre- and post-TAVI FFR assessment were analysed by means of Murray’s law-based quantitative flow ratio (μQFR) and angiography microvascular resistance (AMR) from a single angiographic projection. A total of 198 coronary arteries (123 patients) were analysed pre-TAVI. Of these, instantaneous wave-free ratio (iFR) was available in 148 cases and index of microvascular resistance (IMR) was available in 42 cases. Post-TAVI angiographic projections were available in 91 vessels (64 patients), with μQFR and AMR computation feasible in 82 vessels (55 patients). Median pre-TAVI FFR, μQFR and iFR values were 0.89 (interquartile range [IQR], 0.82–0.95), 0.88 (IQR, 0.82-0.92) and 0.91 (IQR, 0.84-0.97), respectively. The number of ischaemia-causing stenoses as judged by FFR ≤0.80 or μQFR ≤0.80 did not differ (19.7% vs 19.2%; p=0.899), while was significantly higher as judged by iFR ≤0.89 (44.6%; p=0.001). The area under the curve (AUC) of μQFR in predicting pre-TAVI FFR ≤ 0.80 was 0.967 [95% CI 0.875–0.965] with a sensitivity of 82.5%, a specificity of 96.8%, a NPV of 95.6%, and a PPV of 86.8%, while the AUC of iFR in predicting pre-TAVI FFR ≤ 0.80 was 0.967 [95% CI 0.924–0.989] and did not significantly differ from that of μQFR (p = 0.163). A sensitivity of 97.1%, a specificity of 71.1%, a NPV of 98.8%, and a PPV of 50.0% were reported. The accuracy of iFR ≤ 0.89 in predicting pre-TAVI FFR ≤ 0.80 was significantly lower than the accuracy of μQFR ≤ 0.80 (77.0% vs. 93.4%; p = 0.001). μQFR values remained stable after TAVI (0.84 ± 0.11 vs 0.85 ± 0.10; p = 0.558) with a low rate of lesions significance reclassification (9.9%), similarly to FFR (0.86 ± 0.11 vs 0.85 ± 0.14; p = 0.206; lesion reclassification rate of 9.9%). AMR had a median value of 2.10 (IQR, 1.75-2.51) mmHg*s/cm and IMR had a median value of 16.0 (IQR, 10.8-25.0). AMR demonstrated a significant correlation with IMR (r = 0.494, p < 0.001), showing an AUC of 0.887 [95% CI 0.752–0.964] in predicting IMR ≥ 25. AMR cut-off value ≥ 2.5 mmHg*s/cm showed an accuracy of 90.5%, a sensitivity of 70.0%, a specificity of 87.5%, a NPV of 90.3%, and a PPV of 63.6% .

CONCLUSION:

Firstly, pre-TAVI μQFR demonstrated an optimal agreement with FFR assessment and showed a higher accuracy than iFR, which was flawed by a low specificity. Secondly, μQFR values remained stable before and after TAVI with a low rate of lesions significance reclassification. Lastly, AMR showed good correlation and diagnostic accuracy in predicting wire-based IMR ≥ 25. The use of angiography-based computational technologies offers the net advantage of a systematic and comprehensive physiological assessment of both epicardial and microvascular determinants of the coronary circulation, in a fast, simple, and reproducible manner also in a complex scenario such as severe aortic stenosis.