Author + information
- Received March 11, 2019
- Accepted March 11, 2019
- Published online May 15, 2019.
- Manesh R. Patel, MDa,∗ (, )
- Bjarne Linde Nørgaard, MD, PhDb,
- Timothy A. Fairbairn, MB ChB, PhDc,
- Koen Nieman, MD, PhDd,
- Takashi Akasaka, MDe,
- Daniel S. Berman, MDf,
- Gilbert L. Raff, MDg,
- Lynne M. Hurwitz Koweek, MDa,
- Gianluca Pontone, MD, PhDh,
- Tomohiro Kawasaki, MDi,
- Niels Peter Rønnow Sand, MD, PhDj,
- Jesper M. Jensen, MD, PhDb,
- Tetsuya Amano, MDk,
- Michael Poon, MDl,
- Kristian A. Øvrehus, MD, PhDj,
- Jeroen Sonck, MDm,n,
- Mark G. Rabbat, MDo,
- Sarah Mullen, MBTp,
- Bernard De Bruyne, MD, PhDq,
- Campbell Rogers, MDp,
- Hitoshi Matsuo, MD, PhDr,
- Jeroen J. Bax, MD, PhDs and
- Jonathon Leipsic, MDt
- aDivision of Cardiology, Department of Medicine, Duke University Medical Center, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina
- bDepartment of Cardiology, Aarhus University Hospital, Aarhus, Denmark
- cDepartment of Cardiology, University of Liverpool, Liverpool Heart and Chest Hospital, Liverpool, United Kingdom
- dDepartments of Cardiovascular Medicine and Radiology, Stanford University, Stanford, California
- eDepartment of Cardiovascular Medicine, Wakayama Medical University, Wakayama, Japan
- fDivision of Nuclear Imaging, Department of Imaging, Cedars-Sinai Heart Institute, Los Angeles, California
- gDivision of Cardiology, Beaumont Academic Heart and Vascular Group, Royal Oak, Michigan
- hCentro Cardiologico Monzino, Milan, Italy
- iCardiovascular Center, Shin Koga Hospital, Fukuoka, Japan
- jCardiac Research Unit, Institute of Regional Health Research, University Hospital of SouthWest DK, University of Southern Denmark, Odense, Denmark
- kDepartment of Cardiology, Aichi Medical University, Aichi, Japan
- lDepartment of Noninvasive Cardiac Imaging, Northwell Health, New York, New York
- mCardiovascular Center Aalst, OLV Clinic, Aalst, Belgium
- nDepartment of Advanced Biomedical Sciences, University of Naples Federico II, Naples, Italy
- oDivision of Cardiology, Loyola University Chicago, Chicago, Illinois
- pHeartFlow Inc., Redwood City, California
- qCardiovascular Center, Onze-Lieve-Vrouwziekenhuis, Aalst, Belgium
- rDepartment of Cardiovascular Medicine, Gifu Heart Center, Gifu, Japan
- sDepartment of Cardiology, Leiden University Medical Center, Leiden, the Netherlands
- tDepartment of Radiology, Providence Health Care, St. Paul’s Hospital, University of British Columbia, Vancouver, Canada
- ↵∗Address for correspondence:
Dr. Manesh R. Patel, Duke Heart Center, Duke Clinical Research Institute, Duke University, 2400 Pratt Street, Durham, North Carolina 27710.
Objectives The 1-year data from the international ADVANCE (Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care) Registry of patients undergoing coronary computed tomography angiography (CTA) was used to evaluate the relationship of fractional flow reserve derived from coronary CTA (FFRCT) with downstream care and clinical outcomes.
Background Guidelines for management of chest pain using noninvasive imaging pathways are based on short- to intermediate-term outcomes.
Methods Patients (N = 5,083) evaluated for clinically suspected coronary artery disease and in whom atherosclerosis was identified by coronary CTA were prospectively enrolled at 38 international sites from July 15, 2015, to October 20, 2017. Demographics, symptom status, coronary CTA and FFRCT findings and resultant site-based treatment plans, and clinical outcomes through 1 year were recorded and adjudicated by a blinded core laboratory. Major adverse cardiac events (MACE), death, myocardial infarction (MI), and acute coronary syndrome leading to urgent revascularization were captured.
Results At 1 year, 449 patients did not have follow-up data. Revascularization occurred in 1,208 (38.40%) patients with an FFRCT ≤0.80 and in 89 (5.60%) with an FFRCT >0.80 (relative risk [RR]: 6.87; 95% confidence interval [CI]: 5.59 to 8.45; p < 0.001). MACE occurred in 55 patients, 43 events occurred in patients with an FFRCT ≤0.80 and 12 occurred in those with an FFRCT >0.80 (RR: 1.81; 95% CI: 0.96 to 3.43; p = 0.06). Time to first event (all-cause death or MI) occurred in 38 (1.20%) patients with an FFRCT ≤0.80 compared with 10 (0.60%) patients with an FFRCT >0.80 (RR: 1.92; 95% CI: 0.96 to 3.85; p = 0.06). Time to first event (cardiovascular death or MI) occurred cardiovascular death or MI occurred more in patients with an FFRCT ≤0.80 compared with patients with an FFRCT >0.80 (25 [0.80%] vs. 3 [0.20%]; RR: 4.22; 95% CI: 1.28 to 13.95; p = 0.01).
Conclusions The 1-year outcomes from the ADVANCE FFRCT Registry show low rates of events in all patients, with less revascularization and a trend toward lower MACE and significantly lower cardiovascular death or MI in patients with a negative FFRCT compared with patients with abnormal FFRCT values. (Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Wave [ADVANCE]; NCT02499679)
- clinical outcomes
- clinical practice
- coronary computed tomography angiography
- fractional flow reserve
- major adverse cardiac events
Evaluating patients for obstructive atherosclerotic coronary disease, which can lead to ischemia, remains a central part of clinical care in patients presenting with angina symptoms. The American Heart Association/American College of Cardiology Clinical Guidelines recommend noninvasive cardiac testing in patients with an intermediate cardiovascular disease risk based on the ability of available testing modalities to diagnose and predict downstream clinical outcomes (1). In the United States, coronary computed tomography angiography (CTA) is not currently recommended as a first-line test for patients with symptomatic chest pain (1), despite randomized trial data that have demonstrated similar or better prognostic outcomes compared with standard-of-care noninvasive testing (2). This is, in part, due to limitations in specificity and physician agreement when coronary atherosclerosis is moderate-severe (>50% degree stenosis) (3).
Fractional flow reserve derived from coronary CTA (FFRCT) is a noninvasive physiological test that assesses flow limitation across coronary stenoses with high diagnostic accuracy and good correlation to invasive FFR (4). The short-term 90-day outcomes data from the ADVANCE (Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care) Registry showed FFRCT modified treatment recommendations in a large majority of patients as compared with coronary CTA alone (5). In addition, a positive FFRCT was associated with revascularization and less negative invasive coronary angiography (ICA). Intermediate-term clinical utility and outcomes data with an FFRCT have been limited to single-center reports. Using 1-year data from the ADVANCE Registry, we evaluated the relationship of FFRCT with clinical outcomes.
The ADVANCE Registry (NCT02499679) has been described previously (6). Briefly, to understand the effect of FFRCT on clinical practice in the real world, a multicenter international registry of patients being investigated for clinically suspected coronary artery disease (CAD) with documented stenosis of at least 30% on coronary CTA was performed. Patients were prospectively enrolled at 38 sites in Europe, North America, and Japan from July 15, 2015, to October 20, 2017. Clinically stable patients were recruited following the documentation of the presence of CAD on coronary CTA by enrolling sites. All patients provided written informed consent following Institutional Review Board review and approval.
In an effort to understand clinical effect, site investigators reported an initial management plan and treatment strategy based on coronary CTA alone for each patient in accordance with local practice and interpretation of coronary CTA. As with invasive coronary assessment, the decision to further investigate coronary CTA results with an FFRCT was directed by the physician interpreting the scan with a recommendation to consider FFRCT for stenoses in the 30% to 90% range. All FFRCT analyses were performed in a single center (HeartFlow, Inc., Redwood City, California). Once the FFRCT result was made available, the site investigators were asked to report a treatment strategy based on the new information of the coronary CTA combined with the locally interpreted FFRCT result. A positive FFRCT was deemed to be a value ≤0.80 as the lowest per-patient value as previously published in the invasive and noninvasive literature (6). As with real-world studies, subsequent clinical management decisions, including revascularization or medical therapy, were at the discretion of the local referring physician. The study did not dictate interpretation or management decisions.
Design and outcomes assessment
The ADVANCE Registry was designed as a pragmatic real-world registry. Therefore, simple data forms and processes aimed at ease of site access and enrollment were implemented. Each site was asked to report outcomes at standard time points. For potential clinical events, sites were asked to complete case report forms. Pre-specified outcomes of interest included survival free from major adverse cardiovascular events (MACE) inclusive of myocardial infarction (MI), all-cause mortality, or unplanned hospitalization for acute coronary syndrome (ACS) leading to revascularization. In addition, cardiovascular death in combination with MI was assessed through 1 year as was the incidence of revascularization. Event adjudication was performed using standard definitions by an independent clinical events committee at the Duke Clinical Research Institute (DCRI) (Durham, North Carolina) whose members were blinded to clinical and coronary CTA data.
The study was funded by research grants from HeartFlow, Inc. The DCRI served as the core laboratory for FFRCT analysis as part of the site versus core analysis of care patterns. The trial database was managed by iMedNet (Minnetonka, Minnesota). Principal investigators, subinvestigators, and study coordinators at each site accessed iMedNet and were responsible for data entry. The DCRI had access to these databases for data entry, resolving queries, and locking data. HeartFlow did not have access to clinical adjudication forms. The ADVANCE Steering Committee had access to all of the data and drafted and finalized the manuscript.
The planned analysis included evaluation of the primary endpoint of MACE at 1 year and downstream clinical management. The analysis was planned to be evaluated based on FFRCT positivity and negativity and on-site recommendations following the receipt of the FFRCT results. Baseline demographics, risk factors, symptom status, and outcomes (MACE, cardiovascular death or MI, unplanned hospitalization for ACS leading to revascularization, and all revascularizations) at 1 year are presented based on FFRCT findings. Continuous variables are presented as median (interquartile range) and categorical variables are presented as percentage. Mann-Whitney-Wilcoxon nonparametric tests were used for continuous variables and chi-square tests were used for categorical variables. Survival curves for time-to-event analysis were constructed based on all available follow-up data with the use of Kaplan-Meier estimates and were compared with the use of a log-rank test. A 2-sided p value <0.05 was considered significant for all tests. All statistical analyses were performed using SAS software version 9.2 (SAS Institute, Cary, North Carolina).
The ADVANCE Registry enrolled a total of 5,083 patients, and 4,893 of them had coronary CTA submitted for FFRCT analysis. As previously reported from the ADVANCE Registry, of these submitted, 4,737 (96.80%) were of sufficient image quality for FFRCT analysis. Patients were followed by sites at 90 days and to 1 year for clinical events (Figure 1), with 449 patients without available 1-year follow-up data.
Baseline demographics and events
Patient demographics and coronary CTA and FFRCT findings in those who did and did not experience MACE and in those who did not have 1-year follow-up visits recorded are presented in Table 1. Events at 1 year of follow-up and all revascularizations within 90 days and from 90 days to 1 year are presented (Table 2). At 1 year, there were 55 MACE when evaluated as time to first event: 35 all-cause deaths, 12 MIs, and 8 ACS with unplanned hospitalization leading to revascularization. Of the 35 deaths, 15 were deemed cardiovascular related.
1-year clinical events by FFRCT
Among patients undergoing FFRCT analysis, there were 55 MACE at 1 year (1.16%). Patient demographics and coronary CTA and FFRCT findings in those who did and did not experience MACE and in those who did not have 1-year follow-up visits recorded are presented in Table 1. Of the 55 MACE, 43 occurred in patients with an FFRCT ≤0.80 and 12 occurred in those with an FFRCT >0.80 (RR: 1.81; 95% confidence interval [CI]: 0.96 to 3.43; p = 0.06) (Central Illustration). Time to first all-cause death or MI occurred in 38 (1.20%) patients with an FFRCT ≤0.80 as compared with 10 (0.60%) patients with an FFRCT >0.80 (RR: 1.92; 95% CI: 0.96 to 3.85; p = 0.06) (Central Illustration). Time to first cardiovascular death or MI was significantly more common among patients with an FFRCT ≤0.80 compared with an FFRCT >0.80 (25 [0.80%] vs. 3 [0.20%]; RR: 4.22; 95% CI: 1.28 to 13.95; p = 0.01) (Central Illustration). At 1 year, 3 of 1,592 (0.19%) patients with an FFRCT >0.80 had an MI and no patients experienced cardiovascular death. Of those with a negative FFRCT and MI, 1 patient had known disease and had undergone prior percutaneous coronary intervention (PCI) during the study period, and 1 patient with an FFRCT of 0.85 had diffuse aneurysmal atherosclerotic disease with thrombus. In addition, patients who experienced MACE had lower mean FFRCT values than did those who did not (0.70 ± 0.13 vs. 0.74 ± 0.12; p = 0.02), and event free survival for MACE, all-cause death or MI, and cardiovascular death or MI tended to increase as FFRCT values decreased (Figure 2).
Relationship of FFRCT, stenosis, and downstream clinical management
Downstream clinical management was significantly different among groups depending on stenosis severity and FFRCT value. Timing of revascularization during follow-up was most often within 90 days for those patients with an FFRCT ≤0.80, and 1,208 (38.4%) patients with an FFRCT ≤0.80 were revascularized (34.90% within 90 days, 3.50% between days 91 and 365) compared with 89 (5.6%) with an FFRCT >0.80 (4.4% within 90 days, 1.20% between days 91 and 365) (RR: 6.87; 95% CI: 5.59 to 8.45; p < 0.001) (Figure 3A). When further categorized based on stenosis severity, those with 50% stenosis and an FFRCT ≤0.80 were significantly more likely to be revascularized compared with patients with ≥50% stenosis and an FFRCT >0.80 (p < 0.001) and those with <50% stenosis and an FFRCT ≤0.80 (Figure 3B, Table 3). The revascularization-to-ICA ratio was 66.70% (1,155 of 1,733) for those with ≥50% stenosis and an FFRCT ≤0.80, while the revascularization-to-ICA ratio was 29.5% (81 of 273) in those with ≥50% stenosis and an FFRCT >0.80.
When FFRCT was evaluated in a categorical fashion, there was an inverse relationship between FFRCT and the rate of revascularization (FFRCT <0.71: 57.5%, FFRCT 0.71 to 0.75: 27.5%, FFRCT 0.76 to 0.80: 15.9%, FFRCT 0.81 to 0.85: 6.9%, FFRCT 0.86 to 0.90: 3.9%, FFRCT >0.90: 4.6%; p < 0.001) (Table 4).
Relationship of site post–FFRCT treatment recommendation and 1-year actual clinical management
The majority of patients (n = 2,679) in whom medical therapy was the recommended treatment strategy following FFRCT continued on only medical therapy at 1 year (n = 2,490 [92.9%]), with 189 (7.1%) undergoing revascularization (166 PCI and 23 coronary artery bypass grafting). Similarly, when the site recommendation was for revascularization (n = 1,416), the majority (n = 975 [68.9%]) were revascularized, with 842 (59.5%) undergoing PCI and 133 (9.4%) undergoing coronary artery bypass grafting. The revascularization-to-ICA ratio was 77.00% (975 of 1,267) in patients for whom revascularization was recommended.
The 1-year clinical outcomes following FFRCT are reported from a large prospective multicenter registry that included patients from Japan, North America, and Europe. These findings build on the prior report of 90-day outcomes and other single-center datasets (6–8), and highlight a number of findings that have the potential to inform clinical practice. In patients with clinically suspected CAD and at least 30% stenosis observed on coronary CTA, the 1-year outcomes show low rates of MACE events in all registry patients. In patients with an FFRCT in all vessels >0.80, there was a trend toward lower MACE. In addition, among those with an FFRCT >0.80 as compared with ≤0.80, the exploratory outcome of cardiovascular death and MI was significantly lower, and revascularization was less frequent (5.80% vs. 38.40%) and unlikely after 90 days. In keeping with prior invasive physiology studies, the rate of revascularization and cardiovascular death or MI showed an inverse correlation to FFRCT values.
With regard to clinical adoption and utility of FFRCT within the registry, we found good adherence to the post-FFRCT site recommendations in clinical practice. Importantly, 92.90% of patients in whom medical therapy alone was recommended following receipt of FFRCT results remained on medical therapy without revascularization or MACE at 1 year. These data together with the broader outcomes data indicate the safety of patient management following the incorporation of FFRCT into a decision pathway and in most patients with a negative FFRCT, the avoidance of invasive evaluation.
Our findings align with previously published single-center data as well as the invasive physiology literature (7,8). Those patients with CAD but absence of significant pressure loss as manifest by a low FFR, enables the discrimination of those who are more likely to be adequately managed with medical therapy alone versus those who are more likely to require invasive evaluation and revascularization. These findings help confirm the findings from smaller patient populations, highlighting the capacity of FFRCT to better determine which patients could be considered for ICA with the intention of revascularization. Notably, the event rate of MACE in the ADVANCE Registry was 1.50% for patients with reduced FFRCT compared with 0.19% for patients with negative FFRCT >0.80. A corollary is that the incorporation of FFRCT may mitigate the high reported rates of ICA following the detection of moderate atherosclerosis by coronary CTA. It is important to highlight that the presented work is strictly observational in nature and the extent to which those patients with an FFRCT >0.80 would have been deferred for ICA on the basis of CTA alone is not known.
With increasing health care expenditures, there is ever-expanding focus on appropriate resource utilization. To that end, the historically high rates of nonobstructive disease at the time of ICA (9) are being increasingly scrutinized. Prior studies such as the PLATFORM (Prospective Longitudinal Trial of FFRCT: Outcome and Resource Impacts Study) (10) and CONSERVE (Coronary Computed Tomographic Angiography for Selective Cardiac Catheterization) (11) studies have highlighted the potential for CTA and FFRCT to help inform ICA referral in an effective and safe fashion. Our data build on these experiences, with a rate of nonobstructive disease during ICA in the setting of an FFRCT <0.80 of <15% and a PCI-to-ICA ratio, a marker of catheterization laboratory efficiency, of 77%. These findings underscore a prior hypothesis-generating post hoc analysis of the PROMISE (Prospective Multicenter Imaging Study for Evaluation of Chest Pain) trial (12,13). Lu et al. (13) reported that introducing FFRCT following a “positive” coronary CTA may have reduced ICA after CTA by 28% (50 of 181), decreased the rate of ICA without 50% stenosis by 44% (from 27% [49 of 181] to 15% [20 of 131]), and increased the rate of ICA leading to revascularization by 24% (from 49% [88 of 181] to 61% [80 of 131]). These results add to the current knowledge and practical utility and safety of an FFRCT-driven patient pathway in the “real-world” management of CAD with the ability to discriminate risk, improve invasive catheterization to revascularization rates, and potentially act as a gatekeeper to the catheterization lab. Given the observational nature of these findings and the absence of randomized trial evidence, 2 multicenter randomized trials have begun with the intention to test the hypothesis that a CTA–FFRCT strategy will inform clinical practice in a fashion that improves clinical outcomes and is cost effective in diverse health care environments (FORECAST [Fractional Flow Reserve Derived from Computed Tomography Angiography in the Assessment and Management of Stable Chest Pain] [NCT03187639] and PRECISE [Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization] [NCT03702244] studies).
Although the ADVANCE Registry enrolled real-world patients from across the globe and represents a large sampling of those undergoing FFRCT from different health care systems, it remains a registry and as such is subject to potential referral bias inherent in local practices. Additionally, given the structure of the registry design and clinical practice, the testing or management post-imaging was not randomized, and local physicians interpreted and acted on test results. This does not allow this observational registry to make treatment conclusions and randomized trials are needed to address this gap. This analysis does not attempt to colocalize the segment with a stenosis to a given FFRCT value. As a result, discordant vessels could have been combined in this patient-level analysis. One-year patient visits were not mandated if follow-up could be obtained via other mechanisms, such as phone calls or electronic health review, subsequently leaving some patients with further follow-up not available in the locked database. As a result, there could be a number of events not captured in this observation. The baseline characteristics of patients without 1-year follow up are presented in Table 1, along with those for whom 1-year follow-up data were available.
The 1-year outcomes from the ADVANCE FFRCT Registry show low rates of events in all patients, with less ICA and revascularization and a trend toward lower MACE and significantly lower cardiovascular death or MI in patients with a negative FFRCT compared with patients with abnormal FFRCT values.
COMPETENCY IN MEDICAL KNOWLEDGE: FFRCT is used to evaluate patients with suspicion of CAD. The noninvasive FFRCT value is generated based on computational fluid dynamics and has been shown to correlate to invasive FFR.
COMPETENCY IN PATIENT CARE: The 1-year outcomes from the ADVANCE Registry trend toward lower MACE and significantly lower cardiovascular death and MI in patients with negative FFRCT (>0.80) compared with those with positive FFRCT.
TRANSLATIONAL OUTLOOK: This intermediate-term outcome analysis from a real-world registry provides the safety data for deferral of invasive evaluation in patients with negative FFRCT.
The ADVANCE Registry was funded by HeartFlow, Inc. Dr. Patel has received research grants from HeartFlow, Bayer, Janssen, and the National Heart, Lung, and Blood Institute; and has served on the advisory board for HeartFlow, Bayer, and Janssen. Dr. Nørgaard has received unrestricted institutional research grants from Siemens and HeartFlow. Dr. Fairbairn has served on the Speakers Bureau for HeartFlow. Dr. Nieman has received institutional research support from Siemens Healthineers, HeartFlow, GE Healthcare, and Bayer Healthcare. Dr. Berman has received unrestricted research support from HeartFlow. Dr. Hurwitz Koweek has received research support and speaking fees from HeartFlow and Siemens. Dr. Pontone has received institutional research grant and/or honorarium as consultant/speaker from GE Healthcare, Bracco, Medtronic, Bayer, and HeartFlow. Dr. Sonck has received research grant support from the Cardiopath PhD program. Dr. Rabbat has served as a consultant for HeartFlow. Dr. Mullen is an employee of and owns equity in HeartFlow. Dr. De Bruyne has received consulting fees from Abbott, Opsens, and Boston Scientific; and is a shareholder for Siemens, GE Healthcare, Bayer, Philips, HeartFlow, Edwards Lifesciences, and Sanofi. Dr. Rogers is employee of and owns equity in HeartFlow. Dr. Leipsic has served as a consultant for and owns stock options in Circle CVI and HeartFlow. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute coronary syndrome
- coronary artery disease
- confidence interval
- computed tomography angiography
- Duke Clinical Research Institute
- fractional flow reserve derived from coronary computed tomography angiography
- invasive coronary angiography
- major adverse cardiac events
- myocardial infarction
- relative risk
- Received March 11, 2019.
- Accepted March 11, 2019.
- 2019 The Authors
- Fihn S.D.,
- Blankenship J.C.,
- Alexander K.P.,
- et al.
- Newby D.E.,
- Adamson P.D.,
- Berry C.,
- et al.
- Williams M.C.,
- Moss A.J.,
- Dweck M.,
- et al.
- Driessen R.S.,
- Danad I.,
- Stuijfzand W.J.,
- et al.
- Fairbairn T.A.,
- Nieman K.,
- Akasaka T.,
- et al.
- Chinnaiyan K.M.,
- Akasaka T.,
- Amano T.,
- et al.
- Nørgaard B.L.,
- Terkelsen C.J.,
- Mathiassen O.N.,
- et al.
- Douglas P.S.,
- Pontone G.,
- Hlatky M.A.,
- et al.
- Chang H.J.,
- Lin F.Y.,
- Gebow D.,
- et al.
- Lu M.T.,
- Ferencik M.,
- Roberts R.S.,
- et al.