Article Text
PDF
Abstract
Distal coronary to aortic pressure ratio (Pd/Pa) and instantaneous wave-free ratio (iFR) are resting indices of the functional significance of a coronary stenosis measured without inducing hyperaemia. It has been suggested that iFR has superior diagnostic accuracy to Pd/Pa when both are compared to FFR.
Hypotheses In comparison to an FFR for all strategy, revascularisation decisions based on either binary cut-off values for iFR or Pd/Pa or hybrid strategies incorporating iFR or Pd/Pa will result in similar levels of disagreement.
Methods A prospective study in consecutive patients undergoing FFR assessment for clinical indications using proprietary software to calculate iFR. We measured Pd/Pa, iFR, FFR and hyperaemic iFR (HiFR). Diagnostic accuracy vs FFR was calculated firstly using binary cut-off values of <0.90 for iFR and ≤0.92 for Pd/Pa and again using the adenosine zones for iFR of 0.86–0.93 and Pd/Pa of 0.87–0.94 in the hybrid strategy. The pre-determined sample size established prior to the start of the study was 254 vessels.
Results 197 patients with 257 moderate stenoses (mean DS 48%) were studied. 127 (49.4%) vessels were in patients with stable angina and 79 (31%) vessels in patients with recent (>72 h) acute coronary syndromes. Using binary cut-off values diagnostic accuracy was similar for iFR and resting Pd/Pa with misclassification rates of 20.6% vs 19.8%, p = 0.86. In the hybrid analysis, 54% of iFR cases and 53% of Pd/Pa cases were outwith the adenosine zone. Rates of misclassification were 9.4% vs 11.9%, p = 0.55. Sensitivity analyses showed no impact of a variety of angiographic measures of stenosis severity or myocardial area at risk. Comparing proximal stenoses (Syntax segments 1, 11, 5 and 6) to all other lesions and using the RESOLVE cutoff of ≤0.90 for iFR the level of misclassification was 27.7% vs 15.2%, p = 0.014 (Table 1). Using the iFR cutoff of <0.90 the level of misclassification was 26.3% vs 16.2%, p = 0.05.
Sensitivity analyses for iFr in proximal vs distal segments using ≤ 0.9 cut-off compared with FFR. Proximal vs Distal – comparision of 2 proportions: 95% CI = (0.0248, 0.225), Pearson Chi-sq test of assoc: p = 0.014 Fisher’s exact test: p = 0.021
Conclusion When compared to FFR, binary cut-offs for iFR and Pd/Pa results in misclassification of 1 in 5 lesions. Using a hybrid strategy approximately half of the patient do not receive adenosine but 1 in 10 lesions is still misclassified. Neither resting index or strategy can be recommended for decision making in the cath lab. Operators wishing to use resting indices of stenosis severity should be particularly cautious when interpreting data from proximal stenoses in prognostically important vessels.
Receiver operator characteristic curves demonstrating diagnostic accuracy of iFR, Pd/Pa and HiFR in reference to FFR <0.80 Delong’s test p-values for 2 correlated ROC curves: IFR vs iFR: 0.0001, HiFR vs Pd/Pa: 0.0003
Total number of inappropriate PCI’s and incomplete revascularisations when binary cut-offs are utilised for Pd/Pa and iFR. Numbers shown are % of total, n = 257
- FFR
- IFR
- Resting Indices