Long-term prognostic impact of complete revascularisation defined by CT-derived fractional flow reserve Authors: Kristian Taekker Madsen, Bjarne Linde Nørgaard, Kristian Altern Øvrehus, Jesper Møller Jensen, Søren Thorup Scheuer, Erik Parner, Erik Lerkevang Grove, Nadia Iraqi, Timothy Fairbairn, Koen Nieman, Manesh R Patel, Campbell Rogers, Sarah Mullen, Hans Mickley, Allan Rohold, Hans Erik Bøtker, Jonathon A Leipsic, Niels Peter Rønnow Sand Published: 2025 ABSTRACT AbstractBackgroundComplete revascularisation has been associated with improved short-term outcomes in patients with coronary artery disease, but whether these benefits persist long-term and can be defined non-invasively remains uncertain. We investigated the long-term prognostic impact of complete versus incomplete revascularisation determined by coronary CT angiography-derived fractional flow reserve (FFRCT).MethodsIn this prospective multicentre study, 900 patients with new-onset stable angina and at least one coronary stenosis of 30% or greater on coronary CT angiography were followed for a median of 7 years. FFRCT values were obtained for each vessel, and patients were categorised as completely revascularised (all vessels with FFRCT ≤0.80 revascularised), incompletely revascularised (one or more vessels with FFRCT ≤0.80 not revascularised), or with normal physiology (all vessels with FFRCT >0.80). Early revascularisation was defined as treatment within 90 days of the index scan. Quantitative coronary plaque burden was assessed using artificial intelligence-enabled plaque analysis. The primary endpoint was a composite of cardiovascular death or spontaneous myocardial infarction.ResultsOf 900 patients, 210 (23%) were classified as incompletely revascularised, 167 (19%) as completely revascularised and 523 (58%) as having normal physiology. The primary endpoint occurred in 34 of 210 (16.2%) incompletely revascularised patients, 13 of 167 (7.8%) completely revascularised patients and 30 of 523 (5.7%) with normal physiology. Incomplete revascularisation was associated with higher risk compared with complete revascularisation (HR 2.33, 95% CI 1.23 to 4.42; p=0.01) and normal physiology (HR 3.54, 95% CI 2.16 to 5.81; p<0.001). These associations remained significant after adjustment for total plaque burden, and the risk difference persisted beyond 3 years of follow-up.ConclusionComplete revascularisation defined by CT-derived fractional flow reserve was associated with a sustained reduction in cardiovascular death and spontaneous myocardial infarction over 7 years, supporting its potential role as a non-invasive tool to guide revascularisation strategies in stable coronary artery disease.Keywords: Coronary Artery Disease, Computed Tomography Angiography, Angina Pectoris WHAT IS ALREADY KNOWN ON THIS TOPIC Incomplete revascularisation is linked to worse short-term outcomes in coronary artery disease, but long-term effects and non-invasive definitions based on coronary CT angiography have been unclear. Incomplete revascularisation is linked to worse short-term outcomes in coronary artery disease, but long-term effects and non-invasive definitions based on coronary CT angiography have been unclear. WHAT THIS STUDY ADDS Over 7 years of follow-up in the Danish subset of patients from the Assessing Diagnostic Value of Non-invasive CT-fractional flow reserve in Coronary Care registry, complete revascularisation defined by CT-derived fractional flow reserve was associated with a sustained reduction in cardiovascular death and spontaneous myocardial infarction, independent of total plaque burden. Over 7 years of follow-up in the Danish subset of patients from the Assessing Diagnostic Value of Non-invasive CT-fractional flow reserve in Coronary Care registry, complete revascularisation defined by CT-derived fractional flow reserve was associated with a sustained reduction in cardiovascular death and spontaneous myocardial infarction, independent of total plaque burden. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings suggest that CT-derived fractional flow reserve can identify patients who benefit most from complete revascularisation and may serve as a non-invasive tool to guide long-term revascularisation strategies in stable coronary artery disease. These findings suggest that CT-derived fractional flow reserve can identify patients who benefit most from complete revascularisation and may serve as a non-invasive tool to guide long-term revascularisation strategies in stable coronary artery disease. Introduction It is widely accepted that preventive medical therapy constitutes the cornerstone of treatment in coronary artery disease (CAD),1 2 and that the presence of a haemodynamically significant stenosis is associated with increased risk of adverse outcomes.3 Recent studies have shown that complete, as opposed to incomplete, revascularisation of stable as well as unstable patients may confer improved cardiovascular outcomes.4–6 Current definitions of complete revascularisation primarily rely on invasive estimates.7 Consequently, the potential benefits of complete revascularisation cannot be directly applied to the routine clinical management of patients with stable CAD where catheter-laboratory referrals are increasingly guided by non-invasive test results.1 2 Moreover, patients with severe disease (eg, higher plaque burden or multiple significant stenoses) are more likely to undergo incomplete revascularisation, potentially influencing outcomes. It remains unclear whether complete revascularisation provides long-term benefits in stable patients with CAD when accounting for disease extent.6 We have previously reported on 3-year outcome data from a cohort of 900 patients with new onset stable angina pectoris included at three Danish centres participating in the Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care (ADVANCE-DK) in whom we observed a higher incidence of cardiovascular death or spontaneous myocardial infarction (MI) in patients classified as incompletely as opposed to completely revascularised by coronary CT angiography (CTA) derived fractional flow reserve (FFRCT).8 9 By extending the follow-up period of the ADVANCE-DK trial to 7 years the aim of the present study was to clarify whether FFRCT-determined completeness of revascularisation remained persistently associated with beneficial outcomes in a long-term perspective. Methods Study design and patient population The study design of ADVANCE-DK has previously been described.10 In brief, patients with new onset stable angina who had an estimated glomerular filtration rate ≥30 mL/min/1.73 m2 and no previous history of ischaemic heart disease or heart failure were eligible for inclusion if they had at least one coronary stenosis ≥30% by CTA and underwent subsequent FFRCT analysis. A total of 900 patients were included and all patients provided written informed consent. Coronary CT angiography CTA was performed using dual-source CT systems (Siemens Somatom Definition Flash or Force, Forchheim, Germany) according to best practice societal guidelines.11 Oral beta-blockers or ivabradine were, if necessary, administered targeting a heart rate of ≤60 beats/minute. On-site evaluation of CTA data was performed by imaging cardiologists with 6–12 years of experience. Vessels >2 mm in diameter were assessed and degree of stenosis (0%–29%, 30%–49%, 50%–69%, 70–89% and ≥90%) was graded visually. If the degree of stenosis could not be accurately assessed due to excessive calcification, the degree of stenosis was labelled ‘non-evaluable’. The location of stenoses was reported using a 17-segment model and classified as proximal if located in segments 1, 2, 5, 6, 7, 11 or 13; all other stenosis locations were classified as distal.12 Obstructive CAD was defined as a diameter reduction ≥50% in at least one vessel. Interpretation of FFRCT The methodology behind FFRCT has previously been described.13 For each vessel in which a stenosis was identified by CTA, the FFRCT value was obtained in a manually delineated region 2 cm distal from the point of maximal luminal stenosis and a value of ≤0.80 was considered abnormal, online supplemental figure 1. If a vessel contained two equally severe stenoses, the most proximally located stenosis was used for analysis. Quantified coronary plaque analysis Coronary plaque analysis data was obtained by Artificial-Intelligence Enabled Quantitative Coronary Plaque Analysis (AI-QCPA).14 AI-QCPA provides information on the vessel volume in mm3, total plaque volume in mm3 (difference in lumen volume and vessel volume) and individual plaque constituent volumes in mm3, online supplemental figure 1. The individual plaque constituents were categorised based on attenuation density as previously described.14 In the present study, AI-QCPA data were analysed by assessing the plaque burden (plaque volume in mm3×100%/vessel volume in mm3). Role of core laboratory Standard acquired CTA data sets were transmitted to an independent core laboratory (HeartFlow, Mountain View, California, USA) for FFRCT and AI-QCPA analysis. The core laboratory was blinded to all patient and outcome data and was not involved in the planning, collection or analysis of the ADVANCE-DK follow-up data. Coronary angiography and revascularisation Invasive coronary angiography, percutaneous coronary intervention and coronary artery bypass grafting were performed according to international guidelines.15 Patients undergoing revascularisation (<90 days from index CTA) were categorised as early revascularised. Decision-making on revascularisation was at the discretion of the treating physician or heart teams who were unblinded, but not directly informed about the FFRCT result. Data on AI-QCPA were not available to the treating physician. Categorisation of completeness of revascularisation by FFRCT Patients were categorised according to the lesion-specific FFRCT test result and occurrence of early revascularisation as previously described.8 In brief, either: (1) normal FFRCT, all vessels with FFRCT >0.80 (including both revascularised and non-revascularised patients); (2) completely revascularised (CR-FFRCT), all vessels with FFRCT ≤0.80 revascularised; or (3) incompletely revascularised (IR-FFRCT), ≥1 vessel with FFRCT ≤0.80 left non-revascularised (including both revascularised and non-revascularised patients). Endpoint and follow-up The endpoint was a composite of cardiovascular death or spontaneous non-fatal MI.16 Endpoints and information on invasive procedures and revascularisation occurring during the follow-up were achieved by data from the Western Denmark Heart Registry17 and hospital electronic patient files. An independent committee adjudicated clinical events according to the Academic Research Consortium-2 consensus document.16 Follow-up began on the day of the index CTA scan and lasted until death or end of follow-up. No patients were lost to follow-up. Medical treatment Data regarding medication were obtained via medical records after 3 years of follow-up. Statistical methods Continuous variables were described using means and SD, and categorical variables using counts and percentages. Two-sample t-testing was used to compare means between two groups while linear regression was used to compare means between more than two independent groups. Patients who suffered ≥1 endpoint components during the follow-up period were registered based on time to first event for the composite endpoint analyses, while time to cardiovascular death and time to first spontaneous MI were registered for individual endpoint analyses. Event risk as a function of follow-up was estimated using the Aalen-Johansen method with all-cause death considered a competing risk for the event of interest. HRs for between-group comparisons of endpoints were performed using Cox proportional hazard model. Absolute risk differences (ARD) were calculated by transforming time-to-event data into pseudo-observations with an exit time of 6 years, which were then analysed using a generalised linear model. Propensity matching based on total plaque burden was used to create matched samples of patients classified as CR-FFRCT and IR-FFRCT using an exit time of 6 years which was the minimum follow-up period for all patients. In a landmark analysis with a landmark at 3 years, the follow-up was further divided into the periods 0–3 years and 3–7 years, where the analysis time in the 3–7 years period was time since 3 years. The two periods were compared using a Cox proportional hazard model. A p value of <0.05 was considered statistically significant. All statistical analyses were performed using Stata V.18.5 software (StataCorp, College Station, Texas, USA). Results A total of 927 patients were recruited in ADVANCE-DK, of whom 27 were excluded due to inconclusive FFRCT analysis (12 due to calcium artefacts and 15 due to motion artefacts). Among the 900 included patients 523 (58%) had a normal FFRCT, 167 (19%) were classified as CR-FFRCT and 210 (23%) as IR-FFRCT. Invasive assessment within 90 days was performed in 183/523 (35%) of normal FFRCT, 167/167 (100%) of CR-FFRCT and 172/210 (82%) of IR-FFRCT patients, early revascularisation procedures in 19/523 (4%), 167/167 (100%) and 58/210 (28%) of which coronary arterial bypass grafting was performed in 0/19 (0%), 44/167 (26%) and 5/58 (9%), respectively. CR-FFRCT patients were more often males and presenting with typical angina, but other baseline characteristics were similar, table 1. Normal FFRCT patients had lower stenosis severity and total plaque burden than CR-FFRCT or IR-FFRCT. CR-FFRCT patients had more severe stenosis than IR-FFRCT but there were no differences in proximal disease, multivessel disease or total plaque burden, table 2. Statin adherence was higher in CR-FFRCT (93%) compared with normal FFRCT (75%, p<0.001) and IR-FFRCT (82%, p=0.002). Complete data on medical treatment is given in online supplemental table 1. The average follow-up for patients without events was 7.0 years (range: 6.3–8.2). Table 1Baseline characteristicsNormal FFRCT n=523CR-FFRCT n=167IR-FFRCT n=210P value*P value†Demographics Age, years64±1063±1166±90.970.004 Sex, male318 (61)127 (76)137 (65)0.0010.02 BMI, kg/m227.1±4.326.4±3.726.6±3.90.050.70Risk factors Diabetes64 (12)20 (12)25 (12)1.001.00 Hypertension278 (53)85 (51)106 (50)0.751.00 Dyslipidaemia239 (46)82 (49)107 (51)0.390.76 Smoking‡347 (66)116 (69)153 (73)0.220.49Biochemistry Low density lipoprotein, mmol/l3.2±1.13.4±1.73.1±1.00.150.10Baseline symptoms§0.040.06 Typical angina144 (28)71 (43)64 (30) CCS grade I25 (17)11 (15)6 (9) CCS grade II113 (78)51 (72)55 (86)0.080.12 CCS grade III6 (4)9 (13)3 (5) Atypical angina210 (40)51 (31)88 (42) Non-cardiac chest pain83 (16)21 (13)23 (11) Dyspnoea82 (16)23 (14)33 (16)Values are n (%) or mean±SD.FFRCT registered 2 cm distal to stenosis.Normal FFRCT=all vessels with FFRCT >0.80 (including both revascularised and non-revascularised patients). CR-FFRCT=all vessels with FFRCT ≤0.80 revascularised. IR-FFRCT=one or more vessels with FFRCT ≤0.80 non-revascularised (including both revascularised and non-revascularised patients).*P value represents difference between Normal FFRCT, CR-FFRCT and IR-FFRCT.†P value represents difference between CR-FFRCT and IR-FFRCT.‡Current and former smokers.§Seven patients with unknown symptom status at baseline.BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Baseline characteristics Values are n (%) or mean±SD. FFRCT registered 2 cm distal to stenosis. Normal FFRCT=all vessels with FFRCT >0.80 (including both revascularised and non-revascularised patients). CR-FFRCT=all vessels with FFRCT ≤0.80 revascularised. IR-FFRCT=one or more vessels with FFRCT ≤0.80 non-revascularised (including both revascularised and non-revascularised patients). P value represents difference between Normal FFRCT, CR-FFRCT and IR-FFRCT. P value represents difference between CR-FFRCT and IR-FFRCT. Current and former smokers. Seven patients with unknown symptom status at baseline. BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Table 2Non-invasive anatomical and physiological characteristics according to completeness of revascularisation by FFRCTNormal FFRCT n=523CR-FFRCT n=167IR-FFRCT n=210P value*P value†CTA, per-patient maximum stenosis‡<0.001<0.001 <50%133 (25)2 (1)15 (7) 50–69%321 (61)42 (25)115 (55) >69%58 (11)120 (72)74 (35) Proximal stenosis ≥50%336 (64)150 (90)177 (84)<0.0010.13 Multivessel disease§58 (11)77 (46)87 (41)<0.0010.46FFRCT, per-patient minimum value FFRCT value0.88±0.040.63±0.110.69±0.10<0.001<0.001AI-QCPA¶ Total plaque burden15.0±11.723.4±14.225.9±14.7<0.0010.11 Calcified plaque burden3.3±3.35.0±5.16.1±5.0<0.0010.06 Non-calcified plaque burden11.7±8.818.4±10.019.9±10.30.010.19 Low-attenuation plaque burden0.44±0.450.76±0.540.79±0.61<0.0010.55Values are n (%) or mean±SD.FFRCT registered and patients classified as in table 1. Early revascularisation defined as revascularisation <90 days of index CTA.*P value represents difference between Normal FFRCT, CR-FFRCT and IR-FFRCT.†P value represents difference between CR-FFRCT and IR-FFRCT.‡Degree of stenosis non-evaluable in 20 patients.§Multivessel disease defined as ≥1 coronary stenosis ≥50%.¶Plaque analysis not available in 59 patients.AI-QCPA, Artificial Intelligence-Enabled Quantitative Coronary Plaque Analysis; BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Non-invasive anatomical and physiological characteristics according to completeness of revascularisation by FFRCT Values are n (%) or mean±SD. FFRCT registered and patients classified as in table 1. Early revascularisation defined as revascularisation <90 days of index CTA. P value represents difference between Normal FFRCT, CR-FFRCT and IR-FFRCT. P value represents difference between CR-FFRCT and IR-FFRCT. Degree of stenosis non-evaluable in 20 patients. Multivessel disease defined as ≥1 coronary stenosis ≥50%. Plaque analysis not available in 59 patients. AI-QCPA, Artificial Intelligence-Enabled Quantitative Coronary Plaque Analysis; BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Clinical outcomes according to completeness of revascularisation by FFRCT During follow-up, cardiovascular death occurred in 41/900 (4.6%) patients, spontaneous MI in 39/900 (4.3%) and 3 patients experienced both, resulting in a total of 77/900 (8.6%) patients with the composite endpoint. The endpoint occurred more frequently in IR-FFRCT (34/210, 16.2%) than in CR-FFRCT (13/167, 7.8%; HR: 2.33; 95% CI 1.23 to 4.42, p=0.01 (ARD: 8.0%; 95% CI 1.8% to 14.3%, p=0.01)) or normal FFRCT (30/523 5.7%; HR: 3.54; 95% CI 2.16 to 5.81, p<0.001 (ARD: 9.1%; 95% CI 3.9% to 14.3%, p=0.001)), figure 1. No significant difference was observed between CR-FFRCT and normal FFRCT (HR: 1.46; 95% CI 0.76 to 2.80, p=0.26 (ARD: 1.3%; 95% CI −3.1% to 5.6%, p=0.57)). The elevated risk in IR-FFRCT compared with CR-FFRCT and normal FFRCT persisted after adjusting for statin adherence (HR: 2.42; 95% CI 1.24 to 4.74, p=0.01 and HR: 3.39; 95% CI 2.05 to 5.60, p<0.001, respectively) and CAD extent as determined by the number of vessels with obstructive CAD and total plaque burden (HR 1.86; 95% CI 1.04 to 3.72, p=0.04 and HR: 3.22; 95% CI 1.75 to 5.94, p<0.001, respectively). Univariate adjusted analyses are available in table 3. In a 1:1 propensity match based on total plaque burden a persistent higher risk in IR-FFRCT compared with CR-FFRCT (p<0.05) was demonstrated, online supplemental table 2. Data on individual endpoint components according to FFRCT classifications are displayed in online supplemental figure 2 and a full list of all registered spontaneous MIs is available in online supplemental table 3. Data on all-cause death according to FFRCT classification are displayed in online supplemental figure 3. 7-year outcomes of the ADVANCE-DK study population according to completeness of revascularisation by FFRCT. Completeness of revascularisation by FFRCT was defined according to the vessel-specific FFRCT test result and occurrence of early revascularisation as either: (1) normal FFRCT, all vessels with FFRCT >0.80 (including both revascularised and non-revascularised patients); (2) CR-FFRCT, all vessels with FFRCT ≤0.80 revascularised; or (3) incompletely revascularised (IR-FFRCT), ≥1 vessel with FFRCT ≤0.80 left non-revascularised (including both revascularised and non-revascularised patients). ADVANCE-DK, Danish patients participating in Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Table 3Univariately adjusted HRs of clinical outcomes between FFRCT groupsAdjustedHR (95% CI)Baseline symptoms*GenderAgeMaximum stenosis by CTA†Total plaque burdenIR-FFRCT vs CR-FFRCT Composite endpoint‡2.25 (1.19 to 4.27)2.21 (1.16 to 4.20)1.91 (1.01 to 3.66)2.42 (1.23 to 4.74)2.08 (1.06 to 4.07) Cardiovascular death1.68 (0.72 to 3.93)1.78 (0.76 to 4.16)1.31 (0.55 to 3.12)1.92 (0.79 to 4.69)1.39 (0.58 to 3.33) Spontaneous MI3.57 (1.34 to 9.55)3.34 (1.24 to 8.97)3.19 (1.18 to 8.63)3.78 (1.36 to 10.51)3.84 (1.29 to 11.40)IR-FFRCT vs normal FFRCT Composite endpoint‡3.55 (2.17 to 5.83)3.69 (2.24 to 6.06)3.13 (1.91 to 5.14)3.32 (1.98 to 5.56)2.84 (1.60 to 5.05) Cardiovascular death2.96 (1.49 to 5.89)2.93 (1.47 to 5.84)2.47 (1.24 to 4.91)2.33 (1.15 to 4.75)1.89 (0.84 to 4.25) Spontaneous MI4.40 (2.21 to 8.75)4.81 (2.40 to 9.62)4.12 (2.07 to 8.23)4.97 (2.42 to 10.22)3.82 (1.75 to 8.34)CR-FFRCT vs normal FFRCT Composite endpoint‡1.47 (0.76 to 2.83)1.41 (0.73 to 2.73)1.55 (0.81 to 2.98)1.71 (0.78 to 3.79)1.09 (0.53 to 2.23) Cardiovascular death1.59 (0.68 to 3.70)1.53 (0.66 to 3.59)1.85 (0.79 to 4.34)1.08 (0.42 to 2.82)1.09 (0.44 to 2.73) Spontaneous MI1.21 (0.43 to 3.37)1.15 (0.41 to 3.23)1.18 (0.43 to 3.28)3.78 (1.08 to 13.21)0.90 (0.28 to 2.90)FFRCT registered and patients classified as in table 1.Age and total plaque burden were treated as continuous variables.*Baseline symptoms were; typical angina, atypical angina, non-cardiac chest pain or dyspnoea.†Degree of stenosis (%) 30–49, 50–69, 70–89 or ≥90.‡A composite of cardiovascular death and non-fatal MI.BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT; MI, myocardial infarction. Univariately adjusted HRs of clinical outcomes between FFRCT groups FFRCT registered and patients classified as in table 1. Age and total plaque burden were treated as continuous variables. Baseline symptoms were; typical angina, atypical angina, non-cardiac chest pain or dyspnoea. Degree of stenosis (%) 30–49, 50–69, 70–89 or ≥90. A composite of cardiovascular death and non-fatal MI. BMI, body mass index; CCS, Canadian Cardiovascular Society grading of angina pectoris; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT; MI, myocardial infarction. 3-year landmark clinical outcome analysis according to completeness of revascularisation During the >3-year follow-up period, 839 patients remained at risk (CR-FFRCT: 155 (18%), IR-FFRCT: 175 (21%) and normal FFRCT: 507 (60%)), with 55/77 (71%) of the total registered endpoints occurring in this period: 19/177 (6%) in CR-FFRCT, 19/177 (11%) in IR-FFRCT and 27/507 (5%) in normal FFRCT, p=0.04. Comparison between time periods showed similar HRs for CR-FFRCT and IR-FFRCT, but a significantly increased risk in normal FFRCT >3 years versus 0–3 years, figure 2. HRs for the time periods 0–3 and >3 years were; IR-FFRCT versus CR-FFRCT (HR: 3.02; 95% CI 1.01 to 9.1, p=0.04 and HR: 1.98; 95% CI 0.89 to 4.38, p=0.09); IR-FFRCT versus normal FFRCT (HR: 12.38; 95% CI 3.6 to 42.8, p<0.001 and HR: 2.33; 95% CI 1.29 to 4.20, p=0.005); CR-FFRCT versus normal FFRCT (HR: 4.08; 95% CI 0.91 to 18.23, p=0.07 and HR: 1.13; 95% CI 0.53 to 2.41, p=0.75). Of 16 endpoints occurring in patients who were non-adherent to statin treatment, 15/16 (94%) occurred >3 years of follow-up; 1/15 (7%) CR-FFRCT, 5/15 (33%) IR-FFRCT and 9/15 (60%) normal FFRCT. Landmark analysis of outcomes in ADVANCE-DK from 0 to 3 years and >3 years according to completeness of revascularisation by FFRCT. In this landmark analysis, the HR for suffering an event within the previously reported follow-up period (0–3 years) compared with the extended follow-up period (>3 years) was assessed within each of the three FFRCT categories. Patients were classified according to completeness of revascularisation as in figure 1. ADVANCE-DK, Danish patients participating in Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Prognostic value of coronary plaque burden Patients with high (>50th percentile) versus low (≤50th percentile) total plaque burden had a higher risk of the endpoint, HR: 2.38, 95% CI 1.44 to 3.95, p=0.001. Mean total plaque burden was greater in patients with an endpoint compared with those without (26.1±17.1 vs 18.5±13.4, p<0.001). A high plaque burden was associated with an increased endpoint risk in CR-FFRCT and normal FFRCT (although not statistically significant in the latter), but not in IR-FFRCT, figure 3. Data on mean total plaque burden for each FFRCT category are detailed in online supplemental table 4. Survival curves for the endpoint in different FFRCT categories dichotomised by the 50% centile value for total plaque burden. The endpoint was cardiovascular death or spontaneous myocardial infarction. The 50% centile was calculated separately within each of the three FFRCT categories. FFRCT category definitions as in figure 1. ADVANCE-DK, Danish patients participating in Assessing Diagnostic Value of Non-invasive FFRCT in Coronary Care; CR-FFRCT, completely revascularised-FFRCT; CTA, coronary CT angiography; FFRCT, CTA-derived fractional flow reserve; IR-FFRCT, incompletely revascularised-FFRCT. Discussion In this substudy of the prospective ADVANCE-DK trial, we extended the previously reported 3-year outcome data and evaluated the association between 7-year risk of cardiovascular outcomes and FFRCT determined completeness of revascularisation in stable symptomatic patients with coronary stenosis by CTA. We observed that complete (CR-FFRCT) as opposed to incomplete revascularisation (IR-FFRCT) was consistently associated with a lower risk of cardiovascular death or spontaneous MI, and similar to the risk observed in patients with normal FFRCT despite the latter exhibiting more favourable baseline anatomical and physiological characteristics. The presented extended prognostic data from the ADVANCE-DK trial indicate that a complete revascularisation strategy guided by FFRCT may confer long-term improved outcomes. This was underscored by an unchanged risk difference between 0–3 years and >3 years of follow-up within both CR-FFRCT and IR-FFRCT, indicating that the risk of events continued to be reduced by complete as opposed to incomplete revascularisation during the entire follow-up period. Moreover, the risk difference between CR-FFRCT and IR-FFRCT persisted when adjusted for extent of CAD. This is noteworthy because: (1) extent of anatomical CAD18 and coronary plaque burden19 are both strongly associated with adverse outcomes in stable patients, (2) the majority of previous studies favouring complete revascularisation did not adjust for extent of CAD as a confounding factor when assessing outcomes,5 and (3) the effect of complete versus incomplete revascularisation in an outcome substudy of the Initial Invasive or Conservative Strategy for Stable Coronary Disease (ISCHEMIA) trial20 was significantly attenuated when accounting for extent of CAD.6 In the present study, the risk difference between CR-FFRCT and IR-FFRCT was primarily driven by a lower rate of spontaneous MIs in the former. Our finding is in accordance with results from the ISCHEMIA20 and the 5-year follow-up of the Fractional Flow Reserve versus Angiography for Multivessel Evaluation (FAME 2)21 trials in which revascularisation resulted in numerical reductions of MIs compared with medical treatment. Notably, in the ISCHEMIA trial, it was subsequently debated whether the potential beneficial effects of revascularisation had been attenuated by incomplete revascularisation among patients in the invasive arm.6 In FAME 2, the decision not to interrogate stenoses with a pressure wire if visually deemed to be <50% during angiography may have had a similar attenuating impact on the effects of revascularisation, especially as a sizeable number of 30–50% stenoses in proximal coronary segments may be FFR positive.22 Our study indicates that non-revascularised visually intermediate, but FFRCT documented, haemodynamically significant stenoses may impact patient outcomes by demonstrating that most MIs (89%) in the IR-FFRCT group occurred in vessels with a degree of stenosis by CTA <70%. Given the modest utilisation of FFR (approximately 20%) in intermediate stenoses in clinical practice,23 the introduction of a more accessible and non-invasive approach to identify haemodynamically significant intermediate stenoses, may be of clinical importance. The totality of information by CTA and FFRCT, which not only displays a strong correlation to FFR, but also yields data on FFR calculations for all identified stenoses in the entire coronary tree,24 highlights the potential advantages of applying an FFRCT guided revascularisation strategy. Moreover, the morphological data obtained from the CTA allows early initiation of statin treatment based on visual detection of all stages of CAD, which was thought to be a major determining factor for the improved prognosis in patients undergoing evaluation by CTA versus standard of care in the Coronary CT Angiography and 5-Year Risk of Myocardial Infarction (SCOT-HEART) study.25 In our study, artificial intelligence enabled quantified coronary plaque analysis, extended prognostic information beyond FFRCT in selected patients as demonstrated by the associations between total plaque burden and risk of events in CR-FFRCT and normal FFRCT patients. The information gained from coronary plaque analysis may be of particular interest in patients with normal FFRCT because coronary events remained almost completely absent until >3 years of follow-up. Supplementary coronary plaque analysis may, among patients without flow-limiting CAD, identify individuals at increased risk of late-onset events who require increased attention regarding risk factor management and preventive medical care. Although the potential benefits of complete revascularisation are increasingly acknowledged, classifications of complete revascularisation currently rely primarily on invasive estimates.7 A non-invasive definition of completeness of revascularisation may optimise the pathway to the catheter laboratory for stable patients by separating patients who may benefit from revascularisation from those who may benefit the most from medical therapy. Our study complements recent data from the Precise Percutaneous Coronary Intervention Plan (P3) study,26 demonstrating that virtual percutaneous coronary intervention based on FFRCT provides precise estimates of post-percutaneous coronary intervention FFR values, by indicating that a novel non-invasive recognition of complete revascularisation based on FFRCT may be feasible and prognostically beneficial. More studies are needed to further delineate the diagnostic and prognostic value of FFRCT to guide interventional procedures in patients with stable CAD. Limitations The risk of confounding by indication is a potential concern, because our observational design does not allow us to specify whether the observed differences in outcomes are due to the completeness of revascularisation itself or underlying differences in disease severity and patient characteristics. However, we applied hard, independently adjudicated endpoints and our results remained consistent after adjustment for our proxy measures of disease severity. Even though risk factors were similar, patients in the IR-FFRCT group were slightly older and more frequently females, which may have influenced our findings.27 Inherent to the observational, non-randomised study design, our results may have been affected by unmeasured residual confounding. Consequently, the results do not imply causality and should be considered exploratory. The inability to account for peri-procedural MIs, which were not routinely assessed by biochemical monitoring following successful revascularisation, may be considered a limitation as major biomarker elevation periprocedural MIs has been associated with an increased risk of mortality.28 Information on lesion-specific coronary plaque characteristics was not available, which might have further delineated any relationship between focal haemodynamics, interventions and future risk of events. Information on the use of intracoronary imaging and physiologically interrogated for decision-making on revascularisation was not captured. Statin adherence was solely assessed at 3 years of follow-up, and discontinuation of statins may increase the risk of adverse events.29 However, we observed similar HRs for the CR-FFRCT and IR-FFRCT group, in the time period >3 years versus 0–3 years. Moreover, it seems unlikely that a major proportion of patients tolerating statins for years would discontinue treatment.29 Conclusion In patients with stable angina pectoris and suspected significant coronary stenosis by CTA, complete as opposed to incomplete revascularisation defined by FFRCT, remained consistently associated with a lower risk of cardiovascular death and spontaneous MI. The long-term risk observed in completely revascularised patients was similar to patients with normal FFRCT, despite the latter exhibiting more favourable baseline anatomical and physiological characteristics. Supplementary material Acknowledgements The entire financial support for the study was delivered by the participating departments. No external funding was provided. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: The study was approved by the Danish Data Protection Agency (2008-58-0035; 1563 and 1-16-02-633-20). Participants gave informed consent to participate in the study before taking part. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.