Prognosis of CT-derived Fractional Flow Reserve in the Prediction of Clinical Outcomes

Abstract

Purpose

To examine the prognostic implication of fractional flow reserve (FFR) derived from coronary CT (FFR_CT) in routine clinical practice.

Materials and Methods

Patients referred for FFR_CT analysis at a single center between October 2015 and June 2017 were retrospectively included and followed up for rates of invasive angiography and clinical events. Two hundred seven patients underwent successful FFR_CT analysis with seven lost to follow-up, leaving 200 (mean age ± standard deviation, 62.4 years ± 10.0; 49 [24.5%] women) patients for analysis. At coronary CT angiography, patients were categorized as having significant stenosis (SS) in the presence of a diameter stenosis greater than or equal to 50% (hereafter, SS positive) and flow limitation in the presence of a postlesion (that is, FFR_CT measured 2 cm to the distal aspect of the lesion) FFR_CT less than 0.80 (hereafter, FFR_CT positive). Vessel-oriented clinical events (VOCEs) were defined as vessel-related late revascularization (>90 days), myocardial infarction, and cardiac mortality.

Results

At CT angiography, 130 (65%) studies were SS positive and 63 (31.5%) were FFR_CT positive. At median follow-up of 477 days (range, 252–859 days), there were 26 VOCE end points in 22 patients: 22 revascularizations and four nonfatal myocardial infarctions. VOCE end points occurred in zero of 58 (0%) of SS-negative and FFR_CT negative patients, in eight of 79 (10.1%) of SS-positive and FFR_CT-negative patients, in zero of 12 (0%) of SS-negative and FFR_CT-positive patients, and in 18 of 51 (35.3%) of SS-positive and FFR_CT-positive patients (log-rank χ² = 30.1; P < .001). At multivariable Cox regression, both FFR_CT (hazard ratio per 0.1 decrease, 1.54 [95% confidence interval: 1.1, 2.2] P = .013) and stenosis (hazard ratio per unit increase, 2.16 [95% confidence interval: 1.25, 3.72] P = .006) were independently associated with VOCE.

Conclusion

Stenosis and FFR_CT are independent predictors of intermediate-term outcomes. In the absence of a stenosis greater than 50%, a positive FFR_CT result is not associated with an increased intermediate risk.

© RSNA, 2019

Supplemental material is available for this article.

See also commentary by Fairbairn and Bull in this issue.

Summary

In this single-center real-world experience of intermediate-term outcomes, it was demonstrated that at a median follow-up of more than 1 year, a positive result at fractional flow reserve (FFR) derived from CT (or FFR_CT) analysis was associated with a significantly increased risk of late revascularization and myocardial infarction.

Key Points

• ■ Stenosis and fractional flow reserve (FFR) derived from CT (FFR_CT) are independent predictors of intermediate-term clinical outcomes.

• ■ At a median follow-up of more than 1 year, a positive result at FFR_CT analysis was associated with a significantly increased risk of late revascularization and myocardial infarction. There were no myocardial infarctions or deaths in the presence of a negative FFR_CT finding at follow-up.

Introduction

In stable coronary artery disease, percutaneous coronary intervention guided by angiographic stenosis has not been shown to be superior to optimal medical therapy for reduction of adverse events (1,2). However, superior health and economic outcomes have been achieved with ischemia-guided percutaneous coronary intervention when compared with both optimal medical therapy and angiographic-guided percutaneous coronary intervention (3–6). In spite of the benefit of a functional rather than an anatomically driven revascularization strategy, physiologic testing prior to percutaneous coronary intervention is low (7). Despite an 18-fold increase in fractional flow reserve (FFR) utilization following the publication of the original Angiography for Multivessel Evaluation (or FAME) trial, FFR has been reported to be used for only 6.1% of patients with intermediate coronary stenosis (8,9).

FFR derived from CT (FFR_CT) is a noninvasive alternative for physiologic evaluation of coronary artery disease. Rather than direct pressure measurements as performed in invasive FFR or instantaneous wave-free ratio, FFR_CT uses computational fluid dynamics to estimate the FFR. It has a high diagnostic accuracy for the detection of flow-limiting stenosis and correlates well with invasive FFR, although FFR_CT has not yet been directly compared with the instantaneous wave-free ratio (10). FFR_CT allows accurate identification of ischemia-inducing stenosis to guide management and has been shown to safely allow deferral of invasive angiography in the setting of clinical trials, as well as real-world clinical practice (11–13). Although the short-term outcomes and impact of FFR_CT on revascularization and resource utilization has been examined, the prognosis of FFR_CT on intermediate-term (1–2 years) outcomes has not been previously reported in a real-world environment.

The aim of the current study was to evaluate the intermediate-term prognostic importance of FFR_CT for vessel-oriented clinical events (VOCEs) with the hypothesis that FFR_CT would predict the occurrence of adverse clinical outcomes.

Materials and Methods

Study Cohort

This was a single-center, retrospective, nonrandomized observational study. Patients who underwent coronary CT angiography examination at St Paul’s Hospital (Vancouver, British Columbia, Canada) with a subsequent clinically indicated FFR_CT analysis between October 2015 and June 2017 were included in the study. Patients were excluded if they had undergone prior coronary artery bypass graft surgery or percutaneous coronary intervention, or if they were found to have an anomalous coronary origin. The study was approved by the institutional research ethics board of British Columbia (H16–02450 “FFR_CT in Patient Populations’’). The research ethics board waived the need for consent for this retrospective observational study. Patient demographics, past medical history, cardiovascular risk factors, and patient outcomes were collected from a self-reported questionnaire and from medical records. The study was partly funded by HeartFlow (Redwood City, Calif) through the provision of the FFR_CT analysis. The funders had no role in the data collection, analysis, preparation of the manuscript, or the decision to submit. Authors who are not consultants to HeartFlow had control of inclusion of any data and information that might present a conflict of interest.

Coronary CT Angiography Acquisition and FFR_CT

Coronary CT angiography examinations were performed by using either a 64-row (Discovery 750HD) or 256-row (Revolution; GE Healthcare, Waukesha, Wis) CT scanner. Scan acquisition was performed in accordance with Society of Cardiovascular Computed Tomography (or SCCT) best practice guidelines with tube voltage altered according to body mass index (14). The tube current was modulated by using Auto mA (GE Healthcare). Sublingual nitroglycerin spray (0.8 mg) and optimal heart rate control were strictly adhered to. The contrast agent used was iodixanol (320 mg I/mL), using a triphasic contrast material injection protocol (administered at 5.5 mL/sec) with the contrast material volume given typically at 70 mL/sec (reduced dose used in those with chronic kidney disease). Coronary CT angiography data sets were sent for FFR_CT analysis at the discretion of the reporting clinician. Patients with an intermediate stenosis (50%–69%) were routinely referred for FFR_CT; however, stenosis less than 50% and greater than or equal to 70% could be referred if deemed appropriate by the reporting physician or at the request of the referring clinician.

Coronary CT Angiography Evaluation

Each coronary segment was evaluated for the presence of coronary artery disease by using the 18-segment model, and the degree of luminal stenosis was graded as 0%, less than 25%, 25%–49%, 50%–69%, 70%–99%, and occluded in accordance with SCCT guidelines (15). Lesions causing a reduction greater than or equal to 50% in luminal diameter were considered obstructive. Patients were graded by using the Coronary Artery Disease Reporting and Data System (CAD-RADS) based on the most severe segmental stenosis, where 1 is 1%–24% stenosis, 2 is 25%–49% stenosis, 3 is 50%–69% stenosis, 4 is either 70%–99% stenosis or at least 50% stenosis in the left main, and 5 is at least one occluded segment (16). Results were further stratified into positive or negative for significant stenosis (SS) by using a diameter stenosis threshold of greater than or equal to 50% (CAD-RADS 3–5).

FFR_CT Evaluation

FFR_CT analyses were performed by a central core laboratory blinded to the CT angiography assessment (HeartFlow). On completion of FFR_CT analysis, a three-dimensional patient-specific coronary model was provided from which FFR_CT values were available at any given point along the length of the modeled coronary vessel. Coronary modeling with FFR_CT data was limited to coronary vessels with a minimum luminal diameter of greater than 1 mm.

For each coronary lesion identified at coronary CT angiography, the location and severity of stenosis was correlated with the three-dimensional coronary model to ensure accurate anatomic modeling of the stenosis. For any anatomic stenosis greater than 25%, FFR_CT values were obtained at 2 cm distal to the stenosis. An FFR_CT threshold of less than 0.80 was prospectively set as being considered positive (hereafter, FFR_CT positive) in accordance with previously published literature (17). FFR_CT results were added as an addendum to the original coronary CT angiography report. Summary images were provided on the picture archiving and communication system, available at the time of treatment decision making.

Patient Follow-up

Follow-up of the study cohort was conducted through February 2018. The undertaking and timing of invasive coronary angiography (ICA), presence of SS at ICA (≥50% diameter stenosis), and coronary revascularization through percutaneous coronary intervention or coronary artery bypass graft surgery were recorded. A VOCE was defined as the presence of late vessel revascularization (defined as revascularization occurring >90 days after the initial diagnostic CT angiography), myocardial infarction (excluding periprocedural myocardial infarctions), or cardiac-related death.

Statistical Analysis

Continuous variables are presented as means ± standard deviations and categorical variables are presented as numbers. Normality of distribution of data were tested by using a Shapiro-Wilk test. Analyses were performed on a per-patient basis. Where more than one stenosis was present, the more severe of the two was used to categorize the patient. Continuous data were compared between groups by using an independent t test. Categorical data were compared between groups by using Pearson χ² test or Fisher exact test as appropriate. Kaplan-Meier plots were generated by using VOCE as the end point, with censoring of the patients at their first event. A log-rank test was used to test the null hypothesis that the four groups were drawn from the same population. A Cox regression model was used to determine hazard ratios for VOCE for the presence of severe stenosis and FFR_CT status. A multivariable Cox regression model was performed containing both stenosis status and FFR_CT status, both as categorical (positive vs negative) and as continuous variables. As with the Kaplan-Meier analysis, patients were censored at their first event. Statistical analysis was performed by using software (SPSS Statistics, version 22; IBM, Armonk, NY).

Results

Patient Population and Demographics

Two hundred seventeen patients were included, with 207 (95.4%) undergoing successful FFR_CT analysis. Ten of 217 (4.6%) patients were unable to be processed due to motion and misalignment, and an additional seven were excluded because the patients were lost to follow-up (3.4%) (Fig 1). A total of 200 patients (mean age ± standard deviation, 62.4 years ± 10.0; 49 [24.5%] women) were included in the final analysis. Among those referred for FFR_CT, 28 of 200 (14%) were for typical chest pain, 92 of 200 (46%) were for atypical chest pain, 36 of 200 (18%) were for positive risk factors for coronary artery disease, 30 of 200 (15%) were following a positive stress test result, 10 of 200 (5%) were for dyspnea, and two of 200 (1%) were for preoperative work-up. Full patient demographics are provided in Table 1. The median length of follow-up was 477 days (interquartile range, 252–859 days).

Figure 1:

Flowchart shows management strategy of all patients since undergoing coronary CT angiography, as stratified by using CT angiography and fractional flow reserve (FFR) derived from CT (FFR_CT) status. ICA = invasive coronary angiography, MI = myocardial infarction, revasc = revascularization.

Table 1:

Baseline, CT Angiography, and Fractional Flow Reserve Characteristics of Study Participants

Note.—Unless otherwise specified, data are numbers, with percentages in parentheses. CAD-RADS = Coronary Artery Disease Reporting and Data System. Source.—Reference 15.

*Data are means ± standard deviation.

^†CAD-RAD status (stenosis severity): 1 = 1%–24% stenosis, 2 = 25%–49% stenosis, 3 = 50%–69% stenosis, 4 = 70%–99% stenosis or greater than 50% stenosis in the left main coronary artery, and 5 = 100% stenosis.

Rates of Anatomic Stenosis and Flow Limitation at Coronary CT Angiography

At coronary CT angiography, 130 of 200 (65%) patients had a diameter stenosis greater than or equal to 50% (hereafter, SS positive) and 70 of 200 (35%) had a diameter stenosis less than 50% (hereafter, SS negative). A total of 63 of 200 (31.5%) patients were FFR_CT positive. Among the SS-positive patients, 51 of 130 (39%) were FFR_CT positive and 79 of 130 (61%) were FFR_CT negative. Among the SS-negative patients, 12 of 70 (17%) were FFR_CT positive and 58 of 70 (83%) were FFR_CT negative (Table 1; Figs 2, 3).

Figure 2a:

Images in a 40-year-old man with hyperlipidemia and ST changes noted in inferior leads at exercise stress test demonstrate relative disconnect between percentage stenosis and fractional flow reserve (FFR) derived from CT (FFR_CT) status. (a) CT angiography demonstrates large ramus lesion graded as 70%–99% stenosis, which was (b) FFR_CT negative (0.87). Patient was successfully treated with medical management with no vessel-oriented clinical events during follow-up.

Figure 3a:

Images in a 52-year-old woman with atypical chest pain and strong family history of coronary artery disease demonstrate relative disconnect between percentage stenosis and fractional flow reserve (FFR) derived from CT (FFR_CT) status. (a) CT angiography demonstrates 25%–49% stenosis in midleft anterior descending artery; however, it was (b) FFR_CT positive (0.77). Despite this finding, the patient was managed with medical therapy.

Figure 2b:

Images in a 40-year-old man with hyperlipidemia and ST changes noted in inferior leads at exercise stress test demonstrate relative disconnect between percentage stenosis and fractional flow reserve (FFR) derived from CT (FFR_CT) status. (a) CT angiography demonstrates large ramus lesion graded as 70%–99% stenosis, which was (b) FFR_CT negative (0.87). Patient was successfully treated with medical management with no vessel-oriented clinical events during follow-up.

Figure 3b:

Images in a 52-year-old woman with atypical chest pain and strong family history of coronary artery disease demonstrate relative disconnect between percentage stenosis and fractional flow reserve (FFR) derived from CT (FFR_CT) status. (a) CT angiography demonstrates 25%–49% stenosis in midleft anterior descending artery; however, it was (b) FFR_CT positive (0.77). Despite this finding, the patient was managed with medical therapy.

By using a CT angiography diameter stenosis of 70% as the definition of obstructive disease, 45 of 200 (22.5%) patients had a diameter stenosis greater than or equal to 70% and 155 of 200 (77.5%) had a diameter stenosis less than 70%. Among those with stenosis greater than 70%, 25 of 45 (56%) were FFR_CT positive and 20 of 45 (44%) were FFR_CT negative (Fig E1 [supplement]).

Those who were FFR_CT positive were significantly more likely to have diameter stenosis greater than 70% reported at coronary CT angiography, and less likely to have mild stenosis when evaluated by using FFR_CT status (P = .001 for trend). Stratification of patients by using CAD-RADS score and FFR_CT status demonstrated that the frequency of patients classified as FFR_CT positive increased with increasing CAD-RADs score (P = .001 for trend) (Fig 4); 12 of 64 (18.8%) of patients classified as CAD-RADS 2 were FFR_CT positive, while 26 of 85 (30.6%) of patients classified as CAD-RADS 3 were FFR_CT positive. Meanwhile, 23 of 43 (53.5%) of patients classified as CAD-RADS 4 were FFR_CT positive.

Figure 4:

Graph shows relative percentage of positive and negative fractional flow reserve (FFR) derived from CT (FFR_CT) patients as per Coronary Artery Disease Reporting and Data System (CAD-RADS) score classification.

Among the 63 FFR_CT-positive patients, 21 of 63 (33%) were in the “gray zone” of 0.75–0.80. Among these, one case was in a stenosis less than 25% (see Fig 3), one case was in a stenosis of 25%–49%, 16 patients were in stenosis of 50%–69%, and three patients were in stenosis of greater than or equal to 70%.

Rates of ICA and Revascularization in FFR_CT-Positive and FFR_CT-Negative Patients

Patients with lesion-specific ischemia, indicated by a positive poststenosis FFR_CT value measured 2 cm to the distal aspect of the lesion, were significantly more likely to undergo ICA (FFR_CT positive, 63.5% vs FFR_CT negative, 30.0%; P < .001) and revascularization (FFR_CT positive, 41% vs FFR_CT negative, 15%; P < .001) (Figs 5, E2 [supplement]).

Figure 5:

Graphs show frequency of obstructive (diameter stenosis [DS] >50%) and nonobstructive disease (DS <50%) at invasive coronary angiography (ICA) and frequency of subsequent percutaneous coronary intervention (PCI) at ICA with respect to lesion-specific fractional flow reserve (FFR) derived from CT (FFR_CT) status.

Coronary CT Angiography Findings and VOCE Stratified by FFR_CT Results

In total, there were 26 VOCE end points in 22 patients, composed of 22 revascularizations and four nonfatal myocardial infarctions. VOCE occurred in 18 of 63 patients (29%) classified as FFR_CT positive. All myocardial infarctions were reported in the FFR_CT-positive group (n = 4), with no events in the FFR_CT-negative group (n = 0) (P = .009). Those who were FFR_CT positive had a significantly higher incidence of late coronary revascularization compared with those who were FFR_CT negative (FFR_CT positive, 14 of 63 [22%] vs FFR_CT negative, eight of 137 [6%]; P = .009) (Table 2).

Table 2:

Coronary CT Angiography Findings and Vessel-oriented Clinical Event End Points Stratified by Postlesion FFR_CT Results

Note.—Data are numerators and denominators, with percentages in parentheses and 95% confidence intervals in brackets. CABG = coronary artery bypass graft surgery, CAD-RADS = Coronary Artery Disease Reporting and Data System, FFR = fractional flow reserve, FFR_CT = FFR derived from CT, PCI = percutaneous coronary intervention.

Of the VOCE occurring in FFR_CT-negative patients, six percutaneous coronary intervention and two coronary artery bypass graft procedures were performed. All patients exhibited stenosis greater than 50% at CT angiography. None of the six percutaneous coronary intervention patients had invasive FFR performed at ICA. Of the two coronary artery bypass graft procedures that occurred in the setting of a negative postlesion FFR_CT, the first was due to the presence of a left main lesion with 50%–60% stenosis and the second was due to severe diffuse atherosclerosis that had no lesion-specific flow drop-off. Both were FFR_CT positive at the distalmost aspect of the vessels but did not demonstrate translesional ischemia (postlesion FFR_CT <0.80).

At Kaplan-Meier analysis (Fig 6), there was a significant difference in outcomes between the four groups (log-rank χ² = 30.1; P <.001), with the combination of SS with flow limitation at FFR_CT having the numerically poorest event-free survival curves for VOCE. Considering a CT angiography stenosis of 70% as being obstructive produced similar results (log-rank χ² = 28.5; P < .001 for between-group differences) (Fig E3 [supplement]), with the exception that VOCE including a myocardial infarction occurred in those with a stenosis less than 70%, which was not the case for a stenosis less than 50%.

Figure 6:

Kaplan-Meier plot for vessel-oriented clinical event (VOCE) according to presence or absence of severe stenosis (SS) and positive result at fractional flow reserve (FFR) derived from CT (FFR_CT).

SS was associated with a significantly elevated risk of VOCE (hazard ratio, 47.1 [95% confidence interval {CI}: 1.4, 1594]; P = .03) compared with those with no SS. Positive FFR_CT was also associated with a significantly elevated risk of VOCE (hazard ratio, 4.83 [95% CI: 2.0, 11.5]; P < .001) compared with those with a negative FFR_CT. The combination of an SS and positive FFR_CT was associated with an increased risk of VOCE (hazard ratio, 7.01 [95% CI: 3.0, 16.9]; P <.001) compared with those with only an SS, only a positive FFR_CT, or neither. In a multivariable Cox regression model containing both SS and FFR_CT, positive FFR_CT (hazard ratio, 3.50 [95% CI: 1.47, 8.36]; P =.005) but not SS (hazard ratio, 318170 [95% CI: 0, 2.58 × 10¹⁴⁶]; P = .94) was associated with VOCE. However, the model was unstable with extremely wide CIs due to the absence of events in those with a stenosis less than 50%. When a multivariable Cox regression model was run by using a stenosis threshold of greater than or equal to 70% as indicating an SS, both FFR_CT positive (hazard ratio, 3.55 [95% CI: 1.44, 8.75]; P = .006) and stenosis greater than or equal to 70% (hazard ratio, 3.73 [95% CI: 1.56, 8.91]; P = .003) were independently associated with a significantly increased risk of VOCE. Similar results were achieved when the analysis was performed with FFR_CT and the CAD-RADS score inserted as continuous variables (FFR_CT hazard ratio per 0.1 decrease in FFR_CT, 1.54 [95% CI: 1.1, 2.2]; P =.013; stenosis hazard ratio per unit increase in CAD-RADS score, 2.16 [95% CI: 1.25, 3.72]; P = .006).

Discussion

In this single-center real-world experience of intermediate-term outcomes, we demonstrated that at a median follow-up of more than 1 year, a positive finding at FFR_CT analysis was prognostic for a significantly increased risk of late revascularization and myocardial infarction. Importantly, there were no myocardial infarctions or deaths in the presence of a negative FFR_CT finding at follow-up. Combining the lesion-specific ischemia information from FFR_CT with the presence of SS from CT angiography resulted in a further increase in the risk of VOCE compared with either measure alone.

The finding that FFR_CT was a significant predictor of VOCE is concordant with the recent analysis of the optimal medical therapy arm of the FAME-2 trial, wherein invasive FFR positivity was the strongest predictor of a need for urgent revascularization (17). We also found the combination of severe stenosis with flow restriction to be a stronger predictor than either measure alone. Moreover, all myocardial infarctions occurred in patients with an SS and positive FFR_CT finding, demonstrating the adverse implications of the combination of both stenosis and flow restriction. This is also consistent with a substudy of the FAME-2 trial showing that an increased risk of hard end points was only observed in the presence of both severe stenosis and flow limitation at invasive FFR (18). This may therefore highlight a group who are at highest risk for myocardial infarction and thus most likely to derive benefit from revascularization.

Importantly, within our analysis, FFR_CT negativity provided a strong clinical warranty with no myocardial infarction or death occurring within 1 year. This suggests ICA can be safely deferred in the presence of a negative FFR_CT result. Despite the wide CIs in the current study, the results are concordant with the Prospective Longitudinal Trial of FFR_CT: Outcome and Resource Impacts (or PLATFORM) study, which observed no major adverse cardiovascular events in those in whom ICA was deferred in the setting of a negative FFR_CT finding (19). However, a small number of VOCE occurred in those who were FFR_CT negative in our study, with all these events being late revascularization. Among these patients who underwent revascularization despite being FFR_CT negative, all exhibited a stenosis greater than 50%. Notably, the decision to revascularize was at the discretion of the clinical physician with many factors—both patient and physician centered—impacting this decision-making pathway (20). Furthermore, the relative novelty of FFR_CT compared with a long-held practice of stenosis-based decision making likely contributed to this finding. Notably, of the two coronary artery bypass graft procedures that occurred in the setting of a negative postlesion FFR_CT, both were FFR_CT positive at the distalmost aspect of the vessels but did not demonstrate translesional ischemia (postlesion FFR_CT negative) demonstrated by using FFR_CT. Use of the most distal FFR_CT substantially increases the sensitivity for flow limitation, but at a cost of a significant loss of specificity (21). At present, postlesional FFR_CT is the recommended approach, but in the setting of diffuse atherosclerosis afflicting the length of the vessel or left main disease, it may be that distal FFR_CT is the more appropriate metric to use. However, future prospective studies will be required to address this issue.

The current study demonstrated high rates of revascularization following ICA in those who were FFR_CT positive, with 60% of those undergoing ICA being revascularized (comparable to the PLATFORM study where a revascularization rate of 50% was achieved [19]). That both the current study and a study by Jensen et al (12) have found a revascularization rate of 60% in those undergoing ICA following a positive FFR_CT supports the translatability of these prior trial observations into routine clinical practice. It is notable that not all patients progress to the catheter laboratory for invasive testing even in the presence of stenosis and a positive FFR_CT finding. This same phenomenon was observed in the Assessing Diagnostic Value of Noninvasive FFR_CT in Coronary Care (or ADVANCE) registry, where 54% of patients with FFR_CT-positive stenosis were treated medically (22). Given that a mortality benefit from FFR-driven revascularization is not present above an FFR of 0.67 (23), it is possible that clinicians are emboldened to follow a conservative approach when the FFR does not fall below this threshold; however, dedicated studies are required to specifically address the nuances of this issue. Furthermore, the threshold currently used for FFR_CT for the definition of ischemia is based on the cutoff used for the invasive reference standard. However, it may be that an alternate threshold specific to FFR_CT will more accurately define symptoms and risk, with this as the subject of ongoing investigation. We also observed a high acceptance rate of CT angiography studies for FFR_CT analysis by the core laboratory of 95%, with a low rejection rate of 5%. Rejection rates as high as 33% have been reported such as in the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (or PROMISE) trial (24); however, our rates are far closer to the rejection rate of 13.2% in the Analysis of Coronary Blood Flow Using CT Angiography: Next Steps (or NXT) trial (13.2%) (10), and the rate of 2% reported in another study performed at a single high-throughput center (12,25). The concordance between our high rate of acceptance and the high rate of acceptance in this previous study from another academic center demonstrates the potential for high utilization rates of FFR_CT when there is strict adherence to best image acquisition practices.

There were several limitations in our study. This was a single-center, retrospective, nonrandomized observational study. Thus, the results will be inherently affected by selection bias of patients referred for FFR_CT evaluation, and physician bias for both ICA referral and decisions on revascularization. The lack of mandatory invasively measured FFR among those who had undergone ICA means no conclusions can be drawn on the patients with discordance between FFR_CT findings and decisions regarding revascularization strategy. After all, although FFR_CT correlates well with invasive FFR, there is variability between the techniques that may have accounted for some of the observed variance in outcomes. Because of an absence of events in the SS-FFR arm even at 2 years, the CIs are wide for the hazard ratio estimates and should thus be interpreted with caution. Larger conformational studies will be of benefit in this.

In conclusion, stenosis and FFR_CT are independent predictors of intermediate-term outcomes. In the absence of SS, a positive result at FFR_CT analysis is not associated with an increased intermediate risk.