CRCM Journal

Hemodynamic evaluation

FFR Results

The average basal FFR of our patients before hyperemia was 0.9 ± 0.05.

The average FFR after hyperemia was 0.79 ± 0.08.

A bit more than half of our patients (58.1%) had a positive FFR ≤0.8.

vFFR Results

The average vFFR of our patients was 0.8 ± 0.1.

A bit less than half of our patients (48.4%) had a positive vFFR ≤0.8.

Comparison of vFFR vs FFR

ROC curve and area under the curve

The ROC curve which represents the sensitivity in function of (1-specificity) was established.

The area under the ROC curve (AUC) thus was 0.891 (CI 95% 0.8 to 0.982) with p<0.0001.

The ROC curve for the vFFR in our study is illustrated in (Figure 2).

Figure 2: ROC curve for the vFFR

vFFR evaluation in the different sub-groups

Table 3 resumes the vFFR/FFR correlations in the different clinical and angiographic sub-groups in our study, using the vFFR threshold ≤0.8.

Table 3: Correlations between the vFFR and the FFR in the different sub-groups.

Discussion

Appeared in the mid-1990s, the FFR is an exploration technique that allows the study of the functional impact of a coronary lesion. If the problematic stays simple for the minimal stenoses (<40%) or very severe (>90 %), an important category of stenoses, classified as intermediate (40-90 %), has to benefit from a functional evaluation that will determine the benefit of a revascularization. The FFR allows a response to this question in the wake of coronarography, whereas, the non-invasive tests impose a strategy in two times.

We searched to clinically evaluate in our daily practice the calculation of the vFFR using the CAAS software, in comparison to the well validated reference which is the invasive measure of the FFR. Likewise, almost the majority of the studies have used the FFR as a reference diagnostic test for evaluating the diagnostic power of vFFR.

The vFFR allows to surpass complications of FFR technique given its non-invasive character apart from the specific risk of diagnostic coronarography. Another cause of insufficient use of FFR is its time-consuming nature. Indeed, Westra et al. in the FAVOR II Europe-Japan study have demonstrated a significant measurement time gain of the vFFR (5 minutes in average) in comparison with the FFR (7 minutes in average) with a p<0.001 [5]. All these arguments have pushed the development of software allowing the simulation of the FFR measurement based on the coronary angiography.

The reproducibility of the vFFR measurements has been well validated by Masjedi et al. in the FAST trial [3] and Pellicano et al. [6]. In fact, a low interindividual variability have been observed between operators to measure the vFFR, translating with an excellent correlation between the measured vFFR values by different operators at 0.95 (p<0.001) and 0.92 (p<0.0001) respectively in these two studies. Likewise, we have demonstrated a significant correlation between vFFR and FFR in our study at R²=0.692 with p<0.0001. Although significant, this correlation was a bit less powerful than the one found in literature which varied between 0.69 and 0.94 [7]. In 2 recent met analysis regrouping the prospective studies comparing the vFFR with the FFR, this correlation was on average at r=0.82 and r=0.8, and these were statistically significant [8,9]. In addition, according to the FAST study that studied the vFFR calculated by the help of the same software used in our study (CAAS), the correlation was excellent at 0.89(p<0.001) [3].

The area under the curve ROC found in our population was 0.891 (CI 95% 0.8-0.982). This testifies a very good diagnostic performance of the vFFR in the evaluation of the intermediate stenoses. These results are superimposable to big published studies as testified by the recent metanalysis of Westra and Cortés concluding to an AUC average of 0.92 (CI 95% 0.90-0.95) and 0.94 (CI 95% 0.91-0.95) respectively [8,9]. The FAST study evaluating the vFFR by means of the CAAS software concluded the same results with an AUC of 0.93 (CI 95% 0.88-0.97) [3].

The global performance in this study was improved with a threshold of 0.83 (optimal threshold found by the ROC curve). This new threshold increased the sensitivity of vFFR to above 80% (83.3%), which agrees with the majority of the published studies; this while always conserving a good specificity and diagnostic precision at 88.5% and 85.5% respectively.

Even better, the determination of both thresholds for the vFFR within the framework of a hybrid approach vFFR-FFR allowed us to suggest a new diagnostic strategy able to compensate the lack of sensitivity found in our study. The choice of these 2 thresholds vFFR ≤0.8 and vFFR ≥0.88, with recourse to the FFR if vFFR in grey zone, revealed an excellent diagnostic performance with a sensitivity and a specificity exceeding 90%. This approach has been studied and adopted by certain authors; Tar et al. have searched also to define ideal thresholds >0.88 to attain a PPV of 100% and ≤0.8 to obtain a NPV of 100%. These intervals defined in this study were found in 69% of the studied population.

In a review published in 2021, authors were searching the role of vFFR in ACS, 12 studies of angiography based vFFR trials involving patients with ACS were reviewed. A total of 2336 patients were enrolled. Accuracy of vFFR ranged from 86.8 to 95.7 and AUC ranged from 0.80 to 0.98. They concluded that vFFR should be used in intermediate lesions during an appropriate angiography to secure good results and increase the efficacy of vFFR [10]. 

Another review showed that although FAST trial demonstrated a good correlation between FFR and vFFR in both stable and NSTE-ACS patients (r=0.89 vs 0.89 respectively) with a good agreement between them (mean difference =0.01 SD =0.0356), no conclusions was made due to small sample size.

Therefore, a larger FAST EXTEND trial was conducted retrospectively on 294 patients with stable angina or NSTE-ACS to evaluate the performance of vFFR in complex lesions. Still, it confirmed a strong correlation between vFFR and FFR in different coronary vessels and lesion subsets (bifurcations, tortuous vessels, calcified lesions, tandem lesions, and diffuse disease). Furthermore, FAST II study was conducted to overcome previous limitations. However, it is a prospective trial involving six centers and 334 patients (diagnosed with stable angina or NSTE-ACS), aimed for the same purpose and showed same results. Similarly to our study, it confirmed a good correlation of vFFR with FFR but r was equal to 0.74, p < 0.001 [11].

Limitations

Our study contains a certain number of limits

The number of patients included is relatively low (59 patients with 62 evaluated stenoses).

We did not evaluate by 2 operators the reproducibility of the vFFR measures.

Conclusion

The main interest of the vFFR is the non-invasive and instantaneous functional evaluation of the intermediate stable coronary stenoses. Besides this important contribution, some additional applications of the vFFR were described in literature, of which can even present advantages in the future compared to the FFR. Among these ones are the evaluation of the non-culprit coronary stenoses within the framework of the acute coronary syndrome with ST segment elevation by vFFR.

This work, certainly with several limits, represents a draft allowing to evaluate the interest of adopting new technologies in current practice.

Finally, an evaluation as part of a new larger study, multicentric, with a prognostic analysis, and ideally with a randomization protocol, will be necessary to be able to retain the vFFR as a needed tool in our arsenal diagnostic of non-invasive and instantaneous evaluation of intermediate stenoses and of guidance of coronary revascularizations.

References

1. Neumann FJ, Sousa-Uva M, Ahlsson A, Alfonso F, Banning AP, et al. (2018) 2018 ESC / EACTS Guidelines on myocardial revascularization The Task Force on myocardial revascularization of the European Society of Cardiology ( ESC ) and European Association for CardioThoracic Surgery (EACTS). G Ital Cardiol (Rome). 220:1S-61S.
2. Härle T, Zeymer U, Hochadel M, Zahn R, Kerber S, et al. (2016) Real-world use of fractional flow reserve in Germany : results of the prospective ALKK coronary angiography and PCI registry. Clin Res Cardiol 106: 140-150.
3. Masdjedi K, Zandvoort LJC Van, Balbi MM, Gijsen FJH, Ligthart JMR, et al. (2020) Validation of a three-dimensional quantitative coronary angiography-based software to calculate fractional flow reserve : the FAST study. Eurointervention. 16:591–599.
4. Xu B, Tu S, Shubin, Qiao S,  Qu X,  Chen Y, et al. (2017) Diagnostic Accuracy of Angiography-Based Quantitative Flow Ratio Measurements for Online Assessment of Coronary Stenosis. Journa Am Coll Cardiol. 70: 3077–3087.
5. Westra J, Andersen BK, Campo G, Matsuo H, Koltowski L, Eftekhari A, et al. (2018) Diagnostic Performance of InProcedure AngiographyDerived Quantitative Flow Reserve Compared to Pressure-Derived Fractional Flow Reserve: The FAVOR II Europe - Japan Study. J Am Hear Assoc. 7:e009603.
6. Pellicano M, Lavi I, Bruyne B De, Vaknin-Assa H, Assali A, et al. (2017) Validation Study of Image-Based Fractional Flow Reserve During Coronary Angiography. Circ Cardiovasc Interv. 10:1–8.
7. Xing Z, Pei J, Huang J, Hu X, Gao S (2019) Diagnostic Performance of QFR for the Evaluation of Intermediate Coronary Artery Stenosis Confirmed by Fractional Flow Reserve. Brazilian J Cardiovasc Surg. 34: 165–172.
8. Westra J, Tu S, Campo G, Qiao S, Matsuo H, et al. (2019) Diagnostic performance of quantitative flow ratio in prospectively enrolled patients : An individual patient-data meta-analysis. Catheter Cardiovasc Inerv. 94:693-701.
9. Cortés C, Carrasco-Moraleja M,  Aparisi A,  Rodriguez-Gabella T, Campo A, et al. (2020)  Quantitative flow ratio — Meta-analysis and systematic review. Catheter Cardiovasc Inerv. 97:807-814.
10. Haley HA, Ghobrial M, Morris PD, Gosling R, Williams G, et al. (2021) Virtual (Computed) Fractional Flow Reserve: Future Role in Acute Coronary Syndromes. Front Cardiovasc Med. 8: 735008.
11. Dobrić M, Furtula M, Tešić M, Timčić S, Borzanović D, et al. (2023) Current status and future perspectives of fractional flow reserve derived from invasive coronary angiography. Front Cardiovasc Med. 10: 1181803.