Determination of the optimal measurement point for fractional flow reserve derived from CTA using pressure wire assessment as reference

BACKGROUND. For clinical decision making, it was recently recommended that values of fractional flow reserve (FFR) derived from coronary CTA (FFR_CT) be measured 1–2 cm distal to the stenosis, given the potential for overestimation of ischemia when FFR_CT values at far distal segments are used. Supporting data are, however, lacking. OBJECTIVE. The purpose of the present study was to evaluate the diagnostic performance of FFR_CT values measured 1–2 cm distal to the stenosis and at more distal locations relative to invasive FFR values. METHODS. FFR_CT and invasive FFR values for 365 vessels in 253 patients identified from the Assessing Diagnostic Value of Noninvasive FFR_CT in Coronary Care (ADVANCE) registry were prospectively assessed. FFR_CT values were measured 1–2 cm distal to the stenosis and at the pressure wire position and far distal segments. The diagnostic accuracy of FFR_CT was assessed on the basis of the ROC AUC. The AUC of FFR_CT was calculated using FFR_CT as an explanatory variable and an invasive FFR of 0.80 or less as the dichotomous dependent variable. RESULTS. The AUC of FFR_CT values measured 1–2 cm distal to the stenosis (0.85; 95% CI, 0.80–0.88) was higher (p = .002) than that of FFR_CT values measured at far distal segments (0.80; 95% CI, 0.76–0.84) and similar (p = .16) to that of FFR_CT values measured at the pressure wire position (0.86; 95% CI, 0.81–0.89). FFR_CT values measured 1–2 cm distal to the stenosis and at far distal segments had sensitivity of 87% versus 92% (p = .003), specificity of 73% versus 42% (p < .001), PPV of 75% versus 59% (p < .001), and NPV of 86% versus 85% (p = .72), respectively. Subgroup analyses of lesions of the left anterior descending coronary artery, left circumflex coronary artery, and right coronary artery all showed improved specificity and PPV (all p < .005) for FFR_CT values measured 1–2 cm distal to the stenosis compared with values measured at the pressure wire position. However, the AUC was higher for measurements obtained 1–2 cm distal to the stenosis versus those obtained at far distal segments, for left anterior descending coronary artery lesions (p < .001) but not for left circumflex coronary artery lesions (p = .27) or right coronary artery lesions (p = .91). CONCLUSION. The diagnostic performance of FFR_CT values measured 1–2 cm distal to the stenosis was higher than that of FFR_CT values measured at far distal segments and was similar to that of FFR_CT values measured at the pressure wire position in evaluating ischemic status, particularly for left anterior descending coronary artery lesions. CLINICAL IMPACT. The present study supports recent recommendations from experts to use FFR_CT measured 1–2 cm distal to the stenosis, rather than measurements obtained at far distal segments, in clinical decision making.