The Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization: Rationale and Design of the PRECISE Trial

Abstract

Background:

Clinicians vary widely in their preferred diagnostic approach to patients with non-acute chest pain. Such variation exposes patients to potentially avoidable risks, as well as inefficient care with increased costs and unresolved patient concerns.

Objectives

The Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization (PRECISE) trial (NCT03702244) compares an investigational “precision” diagnostic strategy to a usual care diagnostic strategy in participants with stable chest pain and suspected coronary artery disease (CAD).

Study Design:

PRECISE randomized 2103 participants with stable chest pain and a clinical recommendation for testing for suspected CAD at 68 outpatient international sites. The investigational precision evaluation strategy started with a pre-test risk assessment using the PROMISE Minimal Risk Tool. Those at lowest risk were assigned to deferred testing (no immediate testing), and the remainder received coronary computed tomographic angiography (cCTA) with selective fractional flow reserve (FFR_CT) for any stenosis meeting a threshold of ≥30% and <90%. For participants randomized to usual care, the clinical care team selected the initial noninvasive or invasive test (diagnostic angiography) according to customary practice. The use of cCTA as the initial diagnostic strategy was proscribed by protocol for the usual care strategy. The primary endpoint is time to a composite of major adverse cardiac events (MACE: all-cause death or non-fatal MI) or invasive cardiac catheterization without obstructive CAD at one year. Secondary endpoints include health care costs and quality of life.

Conclusions:

PRECISE will determine whether a precision approach comprising a strategically deployed combination of risk-based deferred testing and cCTA with selective FFR_CT improves the clinical outcomes and efficiency of the diagnostic evaluation of stable chest pain over usual care.

BACKGROUND:

Patients without known coronary artery disease (CAD) often seek clinical evaluation for unexplained chest pain or exertional dyspnea. If the pattern of these symptoms is judged stable, clinicians use a varying combination of non-invasive and/or invasive diagnostic tests to clarify both the cause of the symptoms and the risk posed to the patient. However, current diagnostic practices are inefficient, in that many patients must be tested to identify the relatively few (10-20%) that have significant obstructive CAD. From a prognostic perspective, this inefficiency manifests as low annual composite event rates (1-2%/year) among those undergoing initial noninvasive testing (1-3). The root cause of this inefficiency is a combination of a) routine use of strategies that test many patients who are too low risk to benefit, and b) the use of tests that do not provide sufficiently prognostic, actionable information. The widespread use of testing strategies with low diagnostic yield wastes resources and increases expense in health systems already struggling with rising costs (4). However, the possibility of cost savings implied by this inefficiency are only relevant if new, more accurate, and less costly strategies can be developed and validated.

The Scottish Computed Tomography of the HEART (SCOT-HEART) (1) and PROspective Multicenter Imaging Study for Evaluation of chest pain (PROMISE) (2) trials together established coronary computed tomographic angiography (cCTA) as a reasonable, if not superior, first test for routine assessment of patients presenting with stable chest pain. Neither trial however examined whether some chest pain patients might be so low risk that no testing strategy could improve their outcomes, nor the potential added clinical value of selective FFR_CT in patients whose cCTA showed potentially obstructive CAD. The logical next step after SCOT-HEART and PROMISE, therefore, was to focus on making the pre-test clinical risk assessment process more efficient by developing strategies that reserved non-invasive and invasive diagnostic testing only for those with an intermediate or higher likelihood of having obstructive CAD or risk for events and utilized FFR_CT. The Prospective Randomized Trial of the Optimal Evaluation of Cardiac Symptoms and Revascularization (PRECISE) trial is testing the hypothesis that this “precision evaluation” strategy will improve clinical outcomes and diagnostic efficiency as compared with a usual care strategy.

METHODS:

Study Overview

PRECISE is a multicenter, randomized, trial comparing a risk-based precision evaluation strategy with usual care in stable symptomatic patients with suspected CAD (ClinicalTrials.gov Identifier NCT03702244). Figure 1 presents a diagrammatic representation of the trial design. Study activities were approved by the central Institutional Review Board at Duke University and the Institutional Review Boards/Ethics Committees at each of the enrolling sites. All participants enrolled in PRECISE provided written informed consent for participation in the trial.

Figure 1. Study Design.

Figure 1 presents a diagrammatic representation of the study design of the PRECISE trial.

Trial Objectives:

The primary objective of the PRECISE study is to assess clinical outcomes, patient-reported outcomes, decision-making regarding noninvasive testing and invasive coronary angiography (ICA), and costs with a precision evaluation strategy compared to a usual care strategy in participants with stable symptoms suggestive of CAD.

Study Population

Trial enrollment completed on May 19, 2021 with 2103 (target 2100) patients enrolled at sites in the US, Canada, and Europe and across a diverse array of specialties and settings, including physician offices, clinics, hospital outpatient departments, and diagnostic testing facilities. A total of 2146 patients were screened, with 43 screen failures. Symptomatic patients with suspected CAD and a stable clinical course who were recommended by their treating clinician to have a non-invasive diagnostic test or ICA were considered potentially eligible for the study. Patients were excluded if they had a history of documented obstructive CAD (including revascularization and/or myocardial infarction [MI]), or any degree of CAD proven by imaging (with the exception of minor unscored or incidentally noted CAD), or had diagnostic cardiovascular testing for suspected CAD within the last year. Patients were also excluded if their symptoms were not clearly stable (such as patients with acute coronary syndromes) or if their managing clinician felt testing was needed on an urgent or emergent basis. In the control arm we anticipated that 10% of participants would undergo direct-to-invasive catheterization. Specific inclusion and exclusion criteria for PRECISE are listed in Table 1, and are compared with the two other major trials of CTA in stable chest pain patients, PROMISE and SCOT-HEART.

Table 1:

Inclusion & Exclusion Criteria for PRECISE, PROMISE & SCOT-HEART

Inclusion Criteria
PRECISE	PROMISE	SCOT-HEART
Age ≥18 years Stable typical or atypical symptoms suggesting possible significant coronary artery disease (CAD) with further non-emergent testing or elective catheterization recommended to evaluate the presence of suspected significant CAD. Stable chest pain (or equivalent) includes those who have fully been ruled out for Acute Coronary Syndrome (ACS) and for whom elective testing is recommended, regardless of the venue in which they are seen. If prior CV testing has occurred, it must have been performed greater than one year prior to randomization, and the following must be met: cCTA or invasive coronary angiography (ICA) with stenosis < 50% Quantified coronary artery calcium (CAC) < 100 AU Safe performance of cCTA: Creatinine clearance ≥45 ml/min per most recent measurement within 90 days For a female participant of childbearing potential (those who have not been surgically sterilized or are not postmenopausal), a pregnancy test must be performed with negative results known within 7 days prior to randomization Willingness to comply with all aspects of the protocol, including adherence to the assigned strategy and follow-up visits regardless of actual testing performed Ability to provide written informed consent	Men age > 55 and women age ≥65 years If age in men 45 -54 or women 50 -64 years, then must have increased probability of CAD due to EITHER: Diabetes Mellitus (DM) requiring medical treatment Peripheral Arterial Disease (PAD) Cerebrovascular disease OR d) At least one of the following CV risk factors: Ongoing tobacco use Hypertension Abnormal ABI Dyslipidemia 3.New or worsening chest pain syndrome or equivalent symptoms suspicious for clinically significant CAD 4.No prior evaluation for this episode of symptoms 5.Planned non-invasive testing for diagnosis 6.Negative urine/serum pregnancy test for female subjects of child-bearing potential 7.Serum creatinine < 1.5 mg/dL within the past 90 days	Age 18 to 75 years. Referred by a primary-care physician to a dedicated cardiology chest pain clinic with stable suspected angina due to coronary heart disease
Exclusion Criteria
PRECISE	PROMISE	SCOT-HEART
Acute chest pain (in patients who have not been ruled out for ACS) Unstable clinical status Noninvasive or invasive CV testing for CAD within 1 year. CV testing for CAD refers to any stress tests, invasive coronary angiography (ICA), and cCTA (including calcium scoring) only. Resting ECG, resting echocardiogram and resting CMR (MRI) are not exclusionary regardless of when they were performed. Lifetime history of known obstructive CAD (prior myocardial infarction, CABG or PCI, stenosis ≥50%), known EF≤40% or other moderate to severe valvular or congenital cardiac disease Contraindications to cCTA including but not limited to creatinine clearance (GFR) <45 ml/min as per most recent measurement taken within 90 days Exceeds the site’s weight or size limit for cCTA or cardiac catheterization Any condition leading to possible inability to comply with the protocol procedures and follow-up Any condition that might interfere with the study procedures or follow-up Enrolled in an investigational trial that involves a non-approved cardiac drug or device which has not reached its primary endpoint Life expectancy less than 2 years due to non-cardiovascular comorbidities	Diagnosed or suspected ACS requiring hospitalization or urgent or emergent testing Elevated troponin or CK-MB2 Hemodynamically or clinically unstable condition Any invasive coronary angiography or non-invasive anatomic or functional CV test for detection of CAD, including CTA and exercise ECG, within the previous twelve (12) months. Known CAD with prior MI, PCI, CABG or any angiographic evidence of CAD ≥50% lesion in a major epicardial vessel Known significant congenital, valvular (> moderate) or cardiomyopathic process which could explain cardiac symptoms. Contraindication to undergoing a CTA, including but not limited to: Allergy to iodinated contrast agent Unable to receive beta blockers unless heart rate < 65 beats per minute or Pregnancy Unable to provide written informed consent or participate in long–term follow-up Life expectancy < 2 years	Inability to undergo CT scanning Renal failure (serum creatinine >250 μmol/L or estimated glomerular filtration rate <30 mL/min) Previous recruitment to the trial Major allergy to iodinated contrast media Inability to give informed consent Known pregnancy Acute coronary syndrome within 3 months.

Study Design

Endpoints:

The primary endpoint is a composite of three outcomes analyzed as time to first event: all-cause death, non-fatal MI, or invasive cardiac catheterization without obstructive CAD (no FFR≤0.80 or instantaneous wave free ratio (iFR) ≤0.89, or, if hemodynamic information is not available, no coronary stenosis ≥50% by angiographic core laboratory-assessed quantitative coronary angiography (QCA) or site interpretation if QCA is not available). The primary study hypothesis will be tested in all patients at 12 months following randomization in the intention-to-treat (ITT) population.

Secondary endpoints include:

1. Hierarchical analysis of primary endpoint (giving priority to clinical importance of the components of the composite outcome rather than time to first event) (5,6).

2. Resource use patterns (all participants) and medical costs (US participants): resources to be assessed include index testing, follow-up testing, diagnostic and other cardiac procedures and hospitalizations. Primary analysis will be performed at 12 months.

3. Quality of life (QoL): the Seattle Angina Questionnaire (SAQ) will be used to assess angina-specific Quality of Life; the EuroQoL 5D (EQ-5D-5L) survey will be used for a brief assessment of overall (generic) health status; patient satisfaction with diagnostic process will be assessed once at Day 45 using a 4-item instrument created for this trial.

4. Death: All-cause, cardiovascular, non-cardiovascular.

5. MI: All, procedural, spontaneous.

6. Hospitalizations: All, cardiovascular, non-cardiovascular, and for progressive or unstable angina.

7. Rates of preventive medication use.

8. Cumulative estimated radiation exposure at 1 year related to cardiovascular care.

9. Proportion of invasive coronary angiogram patients who undergo revascularization (PCI or CABG) within 6 months of enrollment (catheterization efficiency). Rates will also be calculated for each randomized strategy arm overall.

An independent clinical event committee (CEC) is responsible for the blinded review and adjudication of the primary endpoint components. Catheterization endpoints will first be determined by computerized algorithmic analysis of all angiographic data with output reviewed by a medical monitor. Ambiguous cases will be referred to the CEC, including review of the original cardiac catheterization images and records. Cardiovascular hospitalizations and revascularization procedures will also be confirmed by CEC review.

Study Enrollment, Randomization & Follow-up

Prior to randomization, eligible consenting patients underwent blinded risk stratification with the PROMISE Minimal-Risk Tool, and the site clinician specified their preferred first diagnostic strategy (noninvasive stress test vs. direct to catheterization). Participants were randomized 1:1 at each enrolling center to a precision strategy or usual care, using a random permuted block design with variable block size. Randomization was stratified by the intended first test and by classification of low risk (cut point based on the lowest risk 10% of the PROMISE trial cohort) vs. elevated risk by the PROMISE Minimal Risk Tool. Participants with known non-obstructive plaque or coronary calcification randomized to the precision care arm were assigned to CTA with selective FFR_CT regardless of estimated risk. The number of participants with a non-invasive intended first test (vs. direct to catheterization) was capped at 90% of the total sample size. All participants were followed at 45 (±14) days and at 6 and 12 months (±30 days) after enrollment. For US participants, clinical follow-up at 45 days was done at the enrolling site and QoL assessment was done by structured phone interviews conducted by the DCRI Outcomes Call Center. The entire follow up visit at 6 and 12 months was done by phone interviews conducted by the DCRI Outcomes Call Center in the US and by site coordinators outside of the US.

Study Assessments

Study assessments conducted at enrollment and each follow up contact are shown in Table 2.

Table 2.

Study Assessments in the PRECISE Trial

	Baseline	Day 45	6-month	12-month
Demographics including age and sex	X
General Medical history	X
Cardiovascular risk factors and comorbidities as well as prior testing or events*	X
Cardiovascular update**		X	X	X
Concomitant cardiovascular medications	X	X	X	X
Pregnancy test***	X
Creatinine (if not done in last 90 days)	X
PROMISE Minimal-Risk Tool Assessment	X
Physical Exam	X
Laboratory Testing	X
Resting 12-lead ECG (optional, clinical care only)	X
QoL evaluation: SAQ, EQ5D-5L	X	X	X	X
Participant Satisfaction Questionnaire		X
Initial diagnostic invasive or noninvasive test performed (if assigned)		Prior to 45 day visit
Collection of cardiovascular tests & imaging		X	X	X
Interval assessment for MACE events and testing		X	X	X

BMI, body mass index; CV, cardiovascular

^*Includes history of elevated blood pressure/hypertension, diabetes, cholesterol (including low-density lipoprotein [LDL], high-density lipoprotein [HDL]) if available, smoking status, family history, sedentary life style, obesity (BMI, waist hip ratio), cerebrovascular and peripheral arterial disease (PAD), and history of abnormal ankle brachial index (ABI)

^**Review and documentation of any cardiovascular diagnostic test, cardiovascular procedure, or hospitalizations/clinic visits due to cardiovascular symptoms and complications since enrollment

^***For females of child-bearing potential – those who have not been surgically sterilized or are not postmenopausal)

COVID-19 study modifications

In the spring of 2020, due to the COVID-19 pandemic, all sites in PRECISE were forced to pause enrollment, with most sites beginning phased reopening as circumstances permitted 6 weeks later. Some sites were able to institute remote screening and consenting of participants. With the sudden closing of outpatient testing centers, some enrolled participants had planned testing postponed or cancelled. Multiple efforts spread out over time, including participant, physician, and hospital contacts, were made to capture all tests, hospitalizations and events occurring within the 12-month trial follow-up period. With most site coordinators working remotely, sites and/or the DCRI call center were able to pivot and perform scheduled 45 day and subsequent follow-up visits via telephone and to add medical records collection when direct patient contact was not possible. Under the guidance of the Steering Committee, PRECISE undertook a multipronged approach to maintain site engagement throughout the pandemic, including bi-monthly newsletters, weekly email updates, monthly calls to retain compliance, bi-weekly study calls and email outreach to individual sites. Event reporting and adjudication employed teleconferencing and was uninterrupted. In addition, the relationship of any definite adverse event to COVID19 was determined by the CEC from May 19^th, 2020 onwards.

Tested Strategies

Precision Arm

Participants randomized to the precision evaluation arm were assigned to diagnostic assessment based on use of the PROMISE Minimal-Risk Tool. PRECISE participants who would have had the lowest decile of risk in the PROMISE cohort were assigned to guideline-recommended care focused on symptom and risk factor management without planned cardiac diagnostic testing (6). As PRECISE inclusion and exclusion criteria were more liberal than those in PROMISE, the cut point was finalized to achieve enrollment of approximately 20% of precision evaluation participants (n=181) in the no immediate testing group. The remaining participants, determined to be at intermediate or elevated risk, underwent cCTA with selective FFR_CT. Enrolled participants who had known non-obstructive coronary atherosclerosis or coronary calcification underwent cCTA as the initial diagnostic strategy regardless of risk score results, according to current best practice standards. Since trial leadership remains blinded to randomization, cCTA feasibility, quality, and coronary artery calcium scores will be reported in the primary results paper once unblinding occurs. Selective FFR_CT analysis was initiated by the site based on the presence of at least one 30-90% stenosis in an epicardial vessel ≥2 mm in diameter. Results were provided by HeartFlow to sites within a few hours. For participants in the precision evaluation arm, the study protocol recommended that invasive coronary angiography (ICA) not be performed unless at least one of the following criteria were met: 1) A stenosis ≥90% identified by cCTA; 2) Left main stenosis ≥30% identified by cCTA; 3) Plaque rupture identified by cCTA; or 4) Lesion-specific FFR_CT ≤0.85 in vessels with reference vessel diameter of 2.0 mm or greater. This guidance was intended to allow considerable leeway for clinical judgment in keeping with the pragmatic design.

Usual Care Arm

Participants assigned to usual care underwent either noninvasive stress testing of the site’s choice (exercise electrocardiogram (ECG), stress echocardiogram, stress nuclear (including positron emission tomography (PET)) or stress cardiovascular magnetic resonance imaging (cMR) or ICA as the initial test at the discretion of the participant’s clinician. cCTA (± FFR_CT) was not allowed within the first 45 days following randomization into the usual care arm and strongly discouraged thereafter.

Equipment & Interpretation

All participating sites used standard equipment and procedures for usual care testing, as defined by current practice guidelines (8-15). Sites also used at least 64-slice multi-detector computed tomography (MDCT) for coronary cCTA (15,16). All testing protocols and interpretations were in accordance with current best-practice standards (8-14,16). The site interpretation of all diagnostic tests was verified by central review of the clinical reports for all diagnostic imaging tests performed, with review of the primary images as needed. The angiographic core laboratory performed QCA on all coronary angiograms.

Guideline-directed medical care

Regardless of randomization arm or testing, all participants received guideline directed medical therapy. To facilitate this, informational materials outlining standards of care for all noninvasive test interpretation and guideline recommendations for risk factor and symptom management were provided to sites in both physician and. participant-friendly versions (Supplemental Materials).

In addition to interpretation of all tests, subsequent care was provided by the individual care team at their own discretion, with encouragement to follow guideline-based approaches. Precision evaluation participants determined to be at low risk and assigned to the strategy of guideline-recommended medical management with no immediately planned testing were highly unlikely (by definition) to have significant obstructive CAD. Thus, it is expected that symptoms would resolve with guideline-recommended preventive care, anti-anginal medications, and other medical therapy in nearly all cases. In cases where symptoms were judged to be intractable or accelerating despite optimal guideline-recommended medical treatment, additional non-cardiac diagnostic testing could be pursued as an escalation of testing. In those cases, a cCTA followed by selective FFR_CT was recommended by protocol unless there was an urgent or emergent indication to perform invasive testing.

Statistics and Data Analysis

Data version

The data presented in this report are from a data extract on June 1st, 2021. Thus, the data are subject to change and are not final until trial completion and database lock.

Sample Size

Based on the clinical assumption that 10% of usual care participants would receive angiography as a first test, we arrived at an 8% primary endpoint event rate at 1 year in the usual care group and 5% event rate in the precision care group (i.e., 37.5% relative risk reduction) (2,3,17-19). We estimated that approximately 20% of the overall trial cohort (half of whom will be in the precision arm) will be low risk as assessed by the PROMISE Minimal Risk Score, and that 30% will cross over from no immediate testing to cCTA with selective FFR_CT in the precision evaluation arm. Based on those assumptions, enrolling 1050 participants per group (2100 total participants) provides at least 90% power to detect a primary endpoint relative risk reduction of 37.5%. Sample size calculations are based on the log-rank test (20) with 12-month follow-up in all participants, 10% attrition rate (i.e., lost to follow-up, withdrawals) and a two-sided type I error rate of 0.05.

Statistical Analysis Overview

All analyses will be performed by statisticians at DCRI. The primary endpoint is the time-to-first occurrence of any of the components of the composite of MACE (all-cause death or non-fatal MI) or invasive cardiac catheterization without obstructive CAD during 12-month follow-up. Primary and secondary endpoint comparisons between the precision and usual care groups will be performed using an ITT approach, with statistical testing performed using the log-rank test (21). A secondary per-protocol analysis will also be conducted. Treatment effect sizes will be estimated as hazard ratios (HR) and 95% confidence intervals (CI) from unadjusted and adjusted Cox proportional hazards models (22). The Kaplan-Meier method (23) will be used to calculate time to first event rates for the precision and usual care arms, respectively. Competing risks will be accounted for using the Fine and Gray method (24).

We will conduct a sensitivity analysis for the primary composite endpoint using the unmatched “win-ratio” method which order-ranks the composite endpoint based on clinical importance of the components (death > MI > ICA for non-obstructive CAD) (6). Pre-specified subgroup analyses will assess whether the intervention effect is consistent across multiple participant characteristics. Subgroup HR and 95% CIs will be displayed using a forest plot and subgroup x treatment group interaction terms will be tested in the Cox model. Pre-specified sub-groups include those with low risk vs. elevated risk by PROMISE Minimal Risk Tool or presence of known non-obstructive CAD, intended first test (functional vs. invasive), sex, age, history of diabetes, symptom presentation (chest pain vs. other), SAQ angina score at presentation, and geographic region. All secondary endpoints will utilize the same general statistical approach described above for the primary efficacy endpoint.

Resource Use and Cost Analyses

The primary comparison for economic outcomes will be at 12 months between the usual care and precision arms using an ITT approach. All-cause hospitalizations, cardiovascular hospitalizations, ER visits not resulting in hospitalization, and major outpatient procedures will be enumerated. In addition, we will examine length of stay by intensity of care, numbers of cCTAs, noninvasive stress tests (stress perfusion imaging, stress echocardiography, exercise electrocardiography, stress cMR, invasive tests (ICA, invasive FFR or equivalent, optical coherence tomography, intravascular ultrasound), coronary revascularization procedures (coronary artery bypass graft surgery (CABG), percutaneous coronary intervention (PCI), number of coronary stents), and cardiac medications (beta blockers, aspirin, statins, antiplatelet medications).

The primary economic analyses will be performed from the perspective of the US healthcare system. The primary cost comparison will be made for cumulative costs at 12 months. Comparisons between the two testing strategies will be made using a normal approximation with standard errors estimated using bootstrapping. The primary effect size will be the mean cost difference between the two arms with 95% confidence intervals. In secondary analyses, we will compare resource use and costs between treatment groups at 45 days and 6 months. In addition, resource use patterns and costs for participants enrolled in the US and in the rest of the world will be compared by intention-to-treat. Differences in cost will be interpreted in the context of the trial clinical results, looking for both consistency and plausibility.

Analyses of Quality of Life Outcomes

We will conduct descriptive and comparative analyses for the QOL measures assessed in the trial. The SAQ Summary Score will serve as the primary QOL outcome. We will use a repeated-measures mixed model (SAS Proc MIXED) with the baseline score as a covariate, Day 45, Month 6, and Month 12 responses included as outcome variables, and time as a fixed variable. Restricted maximum likelihood estimation will be used to model all available data from each participant without imputing missing values. An unstructured covariance matrix will be used. The prespecified clinical subgroups for the primary endpoint will also be examined for differences in QOL by treatment arm.

Interim analyses

Due to the pause in enrollment at the onset of the pandemic, the DSMB charter was amended to allow for an adaptive trial design, with the inclusion of interim data analyses to assess the primary endpoint for efficacy and futility. Statistical boundary for the efficacy evaluation was based on a truncated O’Brien-Fleming type boundary generated using the Lan-DeMets α-spending function approach (25). Conditional power (CP) associated with the observed effect size at the interim was assessed utilizing the Chen-DeMets-Lan method (26) for unblinded sample size re-estimation. This would allow the DSMB to make recommendation for sample size re-estimation (upward to 3000) to be conducted by the independent statistician, if the interim result is promising, that is if the CP is greater than 0.5 but less than 0.8. Following each of two interim analyses, the DSMB recommended to continue with the originally planned enrollment of 2100 subjects. The type I error rate for the primary trial results will be tested at α* level, i.e. α*=0.05 – actual alpha spent at the interim looks.

Study Oversight, Organization & Funding

PRECISE is funded by HeartFlow, Inc. The trial is directed by an independent steering committee composed of international experts in cardiology, radiology, clinical trials, and outcomes research. The steering committee developed the study protocol, oversees trial execution, analysis and interpretation of study results, and prepares all study publications. The sponsor, in collaboration with DCRI, is responsible for operational oversight. The Cardiovascular Research Foundation (CRF) serves as the QCA core laboratory, supports the independent Clinical Events Committee providing adjudication of clinical events, and supports the DSMB monitoring study safety and conduct. The steering committee has the final authority regarding any scientific decisions related to the trial. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper and its final contents.

RESULTS

2103 study participants were enrolled from November 27^th, 2018 to May 19, 2021. Baseline characteristics of the study cohort, as well as those from the previously completed PROMISE and SCOT-HEART trials, are described in Table 3. At baseline, participants in PRECISE were 58 years old and 49.8% (N=1047/2103) were female. The majority reported chest pain as their primary symptom (N=1744/2103, 82.9%). PRECISE subjects were comparable to participants from PROMISE and SCOT-HEART in terms of age, sex, and overall risk profile.

Table 3.

Characteristics of PRECISE Trial Participants at Baseline

	PRECISE N=2103	PROMISE N=10,003	SCOT-HEART N=4146
Mean age – years ± SD	58.4 ± 11.5	60.8 ± 8.3	57.1 ± 9.7
Female sex – no. (%)	1047/2103 (49.8%)	5270 (52.7%)	1821 (43.9%)
Caucasian/non-Hispanic- no. (%)	1765/2103 (83.9%)	7693/9941 (77.4%)	N/A
Geographic origin	US + CA +EU	All US+CA	All Scotland
Cardiovascular Risk Factors
Diabetes- no. (%)	373/2103 (17.7%)	2144 (21.4%)	441 (10.6%)
Hypertension- no. (%)	1247/2103 (59.3%)	6501 (65.0%)	1395 (33.6%)
Mean body-mass index- mean ± SD	30.0 ± 6.5	30.5 ± 6.1	29.7 ± 5.9
Dyslipidemia- no. (%)	1341/2103 (63.8%)	6767/10002 (67.7%)	2176 (52.5%)
Sedentary lifestyle	771/2102 (36.7%)	4866/9982 (48.7%)	N/A
Family history of premature CAD (<55 years old)	796/2103 (37.9%)	3202/9970 (32.1%)	1716 (41.4%)
Peripheral arterial disease	21/2103 (1.0%)	182 (1.8%)	53 (1.3%)
Cerebrovascular disease	102/2103 (4.9%)	410 (4.1%)	139 (3.4%)
Coronary Heart Disease	0%	0%	372 (9.0%)
Current or past tobacco use- no./total no. (%)	1197/2103 (56.9%)	5104/10000 (51.1%)	2185 (52.7%)
10-year risk of CV events- mean ± SD	12.0 ± 12.0*	14.7 ± 11.7*	17 ± 12% **
Relevant Medications- no./total no. (%)
Beta-blocker	494/2103 (23.5%)	2399/9589 (25.1%)	1357 (32.7%)
ACE inhibitor/ARB	747/2103 (35.5%)	4194/9569 (43.8%)	685 (16.5%)
Statin	807/2103 (38.4%)	4389/9569 (45.9%)	1786 (43.0%)
Aspirin	595/2103 (28.3%)	4280/9569 (44.7%)	1993 (48.1%)***
P2Y12 inhibitor	43/2103 (2.0%)	N/A	N/A
Calcium channel blocker	384/2103 (18.3%)	N/A	377 (9.1%)
Nitrates	238/2103 (11.3%)	N/A	1160 (28.0%)
Primary presenting symptom-no./total no. (%)
Chest pain	1744/2103 (82.9%)	7272/996 (72.7%)	N/A
Dyspnea	213/2103 (10.1%)	1490/9996 (14.9%)	N/A
Other	146/2103 (6.9%)	1234/9996 (12.3%)	N/A
Type of angina – no./total no. (%)
Typical	502/2103 (23.9%)	1166 (11.7%)	1462 (35.3%)
Atypical	1196/2103 (56.9%)	7773 (77.7%)	988 (23.8%)
Non-anginal	24/2103 (1.1%)	1064 (10.6%)	1692 (40.8%)
Preenrollment testing allowed	No	No	Yes

^*Calculated using the 2013 ACC/AHA ASCVD pooled cohort risk calculator (27)

^**Calculated using the ASSIGN score to estimate 10-year CHD risk (28)

^***Includes all antiplatelet agents, not just aspirin

The calculated 10-year risk of cardiovascular events in PRECISE was 12.0% compared with 14.7% in PROMISE (both using the ACC/AHA 10-year pooled cohort calculator (27) and 17% in SCOT-HEART (using the ASSIGN score (28)). Patients reporting typical anginal symptoms were more common in PRECISE than in PROMISE (23.9% vs. 11.7%), but less common than in SCOT-HEART (35.3%).

DISCUSSION

Despite the large population of patients presenting with stable chest pain or equivalent symptoms and suspected CAD, the optimal diagnostic strategy is still uncertain (29). The PRECISE trial will define the optimal management approach by assessing clinical outcomes, patient-reported outcomes, and medical costs, in participants randomized to an investigational precision strategy including cCTA plus selective FFR_CT, compared with a usual care strategy. PRECISE will be the first randomized trial to prospectively test the safety and efficacy of a guideline-recommended medical therapy with no immediately planned testing approach in very low risk patients presenting with stable chest pain or equivalent symptoms. Once completed, the results of PRECISE will potentially support an improved evaluation and diagnostic testing paradigm for patients presenting with stable chest pain, informing future key policy and practice guideline decisions.

Large contemporary randomized trials of cCTA in stable chest pain populations include the SCOT-HEART (1) and PROMISE (2) trials. In SCOT-HEART, patients undergoing cCTA in addition to usual care (including stress testing in most) had a significantly lower rate of death from coronary heart disease and nonfatal MI at 5 years (30). Patients undergoing cCTA in the PROMISE trial had a lower frequency of catheterization showing no obstructive CAD compared with the functional testing group (3.4% vs. 4.3%, p=0.02) and a lower rate of death and MI during the first 12 months of follow-up (HR 0.66; p=0.049) (though neutral at 25 months) (1,2). Furthermore, in both trials, the use of cCTA led to improved use of key preventive medical treatments such as aspirin and statins (31,32), which tracked with the reduction in adverse outcomes, and was consistent with multicenter registry data (33,34). The totality of data supporting the use of cCTA have resulted in the United Kingdom’s National Institute for Health and Care Excellence (NICE) 2016 guidelines recommending cCTA as the first-line investigation for patients presenting with new-onset chest pain felt to be due to CAD and establishment of reimbursement standards by the Centers for Medicare and Medicaid Services (CMS) (35,36). Similarly the 2019 European Society of Cardiology (ESC) Guidelines for the diagnosis and management of chronic coronary syndromes included cCTA as one of the first line non-invasive testing options to diagnose CAD in symptomatic patients (Class IB) (37). Most recently, the 2021 ACC/AHA Guideline for the Evaluation and Diagnosis of Chest pain includes cCTA as a first line diagnostic test for intermediate-risk patients with stable chest pain (Class 1A) and FFR_CT for those patients with 40% to 90% stenosis in a proximal or middle coronary artery on CCTA (class 2a) (38).

Despite this progress, there is no current consensus among clinicians and researchers on the preferred diagnostic strategy for stable patients with suspected CAD. The stable chest pain population enrolled in PRECISE appears to be similar to both the PROMISE and SCOT-HEART populations, with comparable demographic and presentation characteristics. The estimated 10-year risk profiles across the three trials also appears to be consistent. Thus, the PRECISE trial will build upon the foundation of evidence laid by the PROMISE and SCOT-HEART trials for this commonly encountered patient population by tailoring an anatomic testing strategy based on individual risk.

In recent years, non-invasive computationally-derived fractional-flow reserve computed tomography (FFR_CT) has been developed to calculate FFR values from an cCTA-derived anatomical model of the coronary arteries coupled with computational fluid dynamics and artificial intelligence, a feature that was not included as part of the diagnostic testing strategies in either PROMISE or SCOT-HEART (39-43). By providing both functional and anatomic data, the introduction of FFR_CT to cCTA may reduce the number of false positive cCTA studies that result in invasive angiography for hemodynamically insignificant disease. Considering the favorable long-term outcomes of patients with invasive FFR-negative lesions who do not undergo revascularization (44), the addition of FFR_CT to a broader cCTA-based diagnostic strategy has significant potential to improve the workup and management of stable chest pain patients (45,46). A cCTA± FFR_CT strategy was first prospectively evaluated in the observational Prospective LongitudinAl Trial of FFR_CT: Outcome and Resource Impacts (PLATFORM) trial, which assessed rates of invasive coronary angiography (ICA) without obstructive CAD in patients referred for angiography who were then assigned to either undergo cCTA ± FFR_CT followed by selective ICA use based on those results versus those undergoing ICA alone. PLATFORM identified a significant reduction in the rates of coronary angiography without obstructive disease in the cCTA±FFR_CT arm (12% vs. 73%) and deferral of 61% of ICA procedures in those undergoing cCTA±FFR_CT, which led to a substantially reduced cost with equivalent safety (47). Data from the ADVANCE registry demonstrated excellent long-term outcomes in individuals with negative FFR_CT, with <1% of patients experiencing death or MI at 1 year (48). Thus, we expect the integration of FFR_CT into the anatomic testing strategy in PRECISE to improve efficiency of an anatomic testing strategy, expanding upon the currently available evidence-base supporting an anatomic testing approach.

The UK National Health Service mandates the addition of FFR_CT to cCTA as a cost savings measure and CMS approved a New Technology Ambulatory Payment Classification (APC) for HeartFlow FFR_CT analysis (35,49). Notably, these data have not yet been reflected in American appropriate use documents (50), with functional modalities such as exercise ECG testing and exercise stress testing with nuclear MPI or echocardiography receiving strong recommendations for use above that of cCTA and FFR_CT. PLATFORM’s positive findings are confirmed by the more recent FORECAST randomized trial. In FORECAST, although participants randomized to selective FFR_CT vs usual care had a reduction in ICA of 22% with no adverse clinical consequences, the study failed in the UK NHS system to show cost savings of CTA ± FFR_CT (51). This may have been partially due to 60% of the control group receiving CTA as a first test, as well as differences in cost implications between the United Kingdom and United States, and underpowering. Economic findings aside, an adequately powered pragmatic randomized trial with clinical endpoints is necessary to demonstrate improvements in clinical outcomes. This highlights the need for the PRECISE trial to demonstrate the broader impact of a CTA ± FFR_CT first strategy on clinical outcomes to drive clinical practice changes.

SCOT-HEART and PROMISE both highlighted the low event rates in participants with stable symptoms, and therefore the need to identify appropriate candidates for conservative management without sacrificing clinical outcomes. The low risk for obstructive CAD and paucity of future events of the stable chest pain population overall suggests that the lowest risk patients may not derive benefit from any diagnostic testing (1,2). This is the driving rationale behind including cardiac catheterization without obstructive disease in the primary composite endpoint of the trial, since cardiac catheterization in patients without obstructive CAD puts patients at risk of unnecessary invasive testing without any tangible benefit, all while introducing excess cost. An additional part of the rationale for deferring testing in this population is that multiple randomized trials now suggest that medical therapy is comparable to revascularization even in those patients who do have obstructive CAD (52-54). The majority of patients with stable CAD and angina eventually have resolution of their symptoms without revascularization or cardiac events (55). These trials highlight the need for a patient-centered strategy that identifies patients who can be expected to derive minimal benefit from further testing. The PRECISE trial will address whether such a strategy is safe and effective for patients, as well as cost effective.

While COURAGE (52), ORBITA (53), and ISCHEMIA (54) were all conducted in patients who already had diagnostic testing, the PROMISE Minimal Risk Tool was developed to identify even lower risk patients with stable chest pain prior to testing who are unlikely to benefit from non-invasive diagnostic testing (56,57). The risk model was developed using the PROMISE cohort and employs 10 readily available clinical variables to identify patients with stable chest pain who had no coronary plaque or calcification by cCTA and no cardiac events over 2 years, who may be safely managed with guideline-recommended medical therapy alone (56,57). The risk score has been validated in the SCOT-HEART and Danish Study of Non-Invasive Diagnostic Testing in Coronary Artery Disease 3(DAN NICAD) populations with no decrement in predictive power (56,58) and recently in a “real world” North American cohort (59). The PRECISE trial will prospectively evaluate the hypothesis that the validated PROMISE Minimal-Risk Tool can be safely and effectively implemented to identify low risk patients for a strategy of guideline-recommended medical care with no planned testing. Such an approach may reduce cost, radiation exposure, and unnecessary procedures, while improving quality of life, all of which will be measured in PRECISE. At the other end of the risk spectrum, PRECISE is uniquely enrolling participants whose clinicians favor a direct to cardiac catheterization approach for evaluation of their symptoms. While this strategy was excluded from both SCOT-HEART and PROMISE, in the PLATFORM invasive cohort, approximately 55% of participants in both the usual care (cardiac catheterization) and CTA ± FFR_CT groups had not undergone prior noninvasive diagnostic testing. The cardiac catheterization cancellation rate and economic savings were similar in this group to those with prior testing (4).

SUMMARY

The PRECISE trial will define the optimal evaluation and management strategy of stable, symptomatic patients with suspected CAD while incorporating an investigational, patient-centric risk-based strategy. PRECISE will fill key knowledge gaps around the optimal testing strategy in patients with stable chest pain, the safety and efficacy of a deferred testing strategy in selected very low-risk patients and the impact of a precision strategy on clinical outcomes, QOL, and cost.