RESUMEN
Potential pitfalls of fractional flow reserve (FFR) measurements are well-known drawbacks of invasive physiology measurement, e.g., significant drift of the distal pressure trace may lead to the misclassification of stenoses. Thus, a simultaneous waveform analysis of the pressure traces may be of help in the quality control of these measurements by online detection of such artefacts as the drift or the wedging of the catheter. In the current study, we analysed the intracoronary pressure waveform with a dedicated program. In 130 patients, 232 FFR measurements were performed and derivative pressure curves were calculated. Local amplitude around the dicrotic notch was calculated from the distal intracoronary pressure traces (δdPn/dt). A unidimensional arterial network model of blood flow was employed to simulate the intracoronary pressure traces at different flow rates. There was a strong correlation between δdPn/dt values measured during hyperaemia and FFR (r = 0.88). Diagnostic performance of distal δdPn/dt ≤ 3.52 for the prediction of FFR ≤ 0.80 was 91%. The correlation between the pressure gradient and the corresponding δdPn/dt values obtained from all measurements independently of the physiological phase was also significant (r = 0.80). During simulation, the effect of flow rate on δdPn/dt further supported the close correlation between the pressure ratios and δdPn/dt. Discordance between the FFR and the δdPn/dt can be used as an indicator of possible technical problems of FFR measurements. Hence, an online calculation of the δdPn/dt may be helpful in avoiding some pitfalls of FFR evaluation.
RESUMEN
AIMS: The aim of this study was to develop a simplified model of FFR calculation (FFRsim) derived from three-dimensional (3D) coronary angiographic data and classic fluid dynamic equations without using finite element analysis. METHODS AND RESULTS: Intracoronary pressure measurements were performed by pressure wire sensors. The lumens of the interrogated vessel segments were reconstructed in 3D. The coronary artery volumetric flow was calculated based on the velocity of the contrast material. Pressure gradients were computed by classic fluid dynamic equations. The diagnostic power of the simplified computation of the FFR (FFRsim) was assessed by comparing the results with standard invasive FFR measurements (FFRmeas) in 68 vessels with a single stenosis. We found a strong correlation between the FFRsim and the FFRmeas (r=0.86, p<0.0001). The sensitivity and specificity for predicting the abnormal FFR of ≤0.80 (indicating haemodynamically significant stenosis) were 90% and 100%, respectively. The area under the curve (AUC) was 0.96. To achieve 100% negative and positive predictive values we defined the FFRsim >0.88 and the FFRsim ≤0.8 ranges. In our patient population, these ranges were found in 69% of the cases. CONCLUSIONS: According to our simplified model, the invasive FFR measurement can be omitted without misclassification in pre-specified ranges of the calculated FFRsim.