Abstract
Background
Physiological assessment of intermediate coronary artery lesions to guide therapy is well established. Recently, the use of non-hyperaemic pressure ratios (NHPRs) has been suggested as a reliable alternative to hyperaemic assessment, such as fractional flow reserve (FFR). However certain patient / lesion subsets, including proximal lesion location, young patient age and assessment of non-LAD vessels, have been associated with discordance of resting and hyperaemic measurements which has led to some confusion over their interpretation and integration into routine clinical practice.
Purpose
To evaluate the frequency of discordant resting and hyperaemic measurements among patients undergoing assessment of intermediate coronary artery stenoses, and to identify common lesion-specific features within patients with discordant data.
Methods
From our coronary physiology database, we identified consecutive lesions which had been assessed between October 2020 and March 2021 with both resting (resting full-cycle ratio; RFR) and hyperaemic (wire-based FFR with peripheral adenosine) indices. Positive RFR defined as <0.89 (negative RFR >0.93, grey zone RFR 0.89–0.93). Positive FFR defined as <0.80 (negative FFR ≥0.80). Concordance between measurements was assessed: Group 1 (RFR -ve, FFR +ve: positive discordance); Group 2 (RFR -ve, FFR -ve: normal concordance), Group 3 (RFR +ve, FFR +ve: abnormal concordance) and Group 4 (RFR +ve, FFR -ve: negative discordance).
Results
100 lesions were identified as being assessed with both RFR and FFR, in 83 patients (67% male), mean age 67 (±12) years, vessel assessed; LAD 66, RCA 19, LCx 13, LMCA 1 and radial graft 1; with 45 being proximal lesions. 30 RFR measurements were in the grey zone. Of the remaining 70 lesions, 55 results (79%) were concordant (Group 2 = 31, Group 3 = 24), with 15 results (21%) being discordant (Group 1 = 3, Group 4 = 12). Negative predictive value (NPV) of RFR (for FFR <0.80) was 91%, when grey zone RFR measurements were excluded. Discordance was not related to age (69 vs 68 years, p=0.75), lesion location (proximal lesion with discordance (6/15) vs proximal lesion with concordance (27/55), p=0.91, figure 1) or non-LAD vessel (non-LAD with discordance (9/15) vs non-LAD lesion with concordance (20/55), p=0.77, figure 2).
Conclusion
Overall within our patient group, there appeared to be a good association of RFR to FFR. In particular, RFR had a high NPV for an FFR <0.80. The clinical relevance of discordant measurements requires further investigation. However, our data suggest that a positive RFR (<0.89) measurement may not always correlate with a significant FFR measurement (<0.80), and the mechanism for this is unclear. Consequently, caution should be applied when including these measures in every day practice, in particular within patients with a positive RFR measurement.
FUNDunding Acknowledgement
Type of funding sources: None.