Fractional Flow Reserve Derived from Computed Tomography Coronary Angiography in the Assessment and Management of Stable Chest Pain: FORECAST TRIAL

The FORECAST Trial will assess a new non-invasive test, which looks at the flow of blood in the vessels around the heart.

Currently, there are a number of well-recognised tests to help doctors decide whether patients’ symptoms are related to their heart arteries. Some tests describe the size and shape of the arteries (anatomy) - whilst others provide information as to whether there is sufficient blood flowing to the heart muscle to enable it to work properly. The most common test is called an invasive coronary angiogram. However, the angiogram does not tell the doctor how well the blood vessels are working (i.e. how easily the blood can flow through them, the physiology). A device known as a ‘pressure wire’ is used to assess the flow of blood. This involves threading a small tube into the heart artery to measure the pressure across a narrowing. This approach allows doctors to create a more tailored treatment plan for each patient.

More recently pictures of the heart arteries have been taken using a CT (computerised tomography) heart scanner. This is much more convenient and less invasive for patients. Until recently, CT has not been able to provide the additional information about whether the heart arteries are restricting the blood flow. A company called HeartFlow Inc, have now developed new software that allows a computer to generate information called FFRCT. This test can then not only give results about the size and shape (anatomy) of the blood vessels around the heart but can also be used to calculate how well they function, by measuring the flow of blood going through the vessels (physiology), in a non-invasive way.

This trial will compare two strategies for managing patients undergoing investigation for suspected problems in the heart arteries. Patients enrolled into the trial are randomly assigned (like the toss of a coin) to either:

• Group 1: Routine investigations following the local hospital procedures

or

• Group 2: Trial investigation undergoing CT scan (with FFRct as required)

The trial approached patients aged over 18, referred to a Rapid Access Chest Pain Clinic (RACPC) for chest pain, and whose care team felt the trial would be suitable for them. They were asked to consider taking part in the trial. This trial involves 1400 participants in total.

After discussing the trial and having read the participant information sheet, a member of the research team addressed any questions or concerns from patients regarding the trial. Patients were screened for eligibility and those who were agreeable, provided written informed consent.

Patients were asked a series of questions by the researchers to describe their medical history, medications and current health status. Patients were also asked to answer questions on 3 short forms about quality of life at this time.

Patients were then randomised to group 1 or group 2.

Group 1: Routine investigations following the local hospital procedures

Patients in group 1 were referred for routine investigation as per the hospital guidelines. Based on test results, the doctor recommended a treatment plan. Patients will be followed in the trial for 9 months. This will involve 2 telephone calls at 3 and 9 months. After the 9 months telephone call participation in the trial is complete.

Group 2: Trial investigation undergoing CT scan (with FFRct as required)

Patients in group 2 were referred for a CT scan of the heart arteries as their first investigation.

A CT scan involves X-ray and this scan requires the use of iodine based dye to show the heart arteries clearly. This is a non-invasive, non-surgical procedure.

If the results from the CT scan were negative, that is no blockages or narrowings are found, then the CT results are reviewed by the doctor alone and further trial analysis is not to be performed. If any potentially significant blockages are seen, the images collected during the CT scan are sent to the researchers in California, United States of America to special computers to perform the FFRCT evaluation. The FFRCT results are sent to the doctor within one working day to assist in evaluating treatment options. The FFRCT evaluation requires no participation on the part of patients. Either way, based on test results, the doctor recommends a treatment plan. Patents are followed in the trial for 9 months. This involves 2 telephone calls at 3 and 9 months. After the 9 months telephone call participation in the trial is complete.

No NHS Digital data will flow to the USA or any third party.

Patients are followed up at 3 months and at 9 months from the date they started on the trial. They are then be discharged back to the care of the GP as appropriate.

Participation in this trial does not present any extra risks other than the risks associated with having a CT scan. This is a test routinely used by doctors to evaluate the heart arteries and may be requested by their doctor regardless of patients participation in the trial. If patients are randomised to the test arm, they have a CT scan. If randomised to the routine arm, or if they choose not to take part in the trial patients may still require a CT scan to assess their condition.

The optimal algorithm with which to assess and then manage patients who present with stable chest pain (CP) that might represent angina (i.e. ischaemia (restriction in blood supply to tissues) secondary to coronary artery disease (CAD) is not clear. In such patients, there may be both symptomatic and prognostic benefit to knowing whether they have either;

(a) significant coronary artery atheroma (build-up of plaque, i.e. the anatomy of their coronary arteries)

(b) reversible myocardial ischaemia (RMI) (i.e. assessing the physiology of the artery(ies) as a surrogate)

(c) both.

Evidence demonstrates that it is useful to know about the presence of atheroma because the medium and long-term natural history of this disease process can be altered by the administration of disease-modifying medical therapy. However, it is now accepted, based upon both observational data and randomised trials, that it is the presence and extent of RMI that most closely correlates with the near-term risk of acute myocardial infarction (MI) (heart attack) and that mechanical intervention by coronary revascularisation (restoration of blood flow) is effective both for symptoms and prognosis in the population with RMI both at patient and lesion-level. Thus, RMI is generally considered to represent the principal target for revascularisation strategies whereas optimal medical therapy (OMT) can improve prognosis regardless of the existence of RMI in some patients.

Based on this, it is clear that both coronary anatomy and physiological assessment (as a surrogate of RMI) are of value in the diagnosis and management of patients presenting with stable cardiac-sounding CP. Current tests used in patients presenting with stable CP assess either coronary anatomy or physiology: non-invasive stress tests (echo/nuclear perfusion/CMR) assess for RMI. By contrast, invasive coronary angiography (IC) and CT coronary angiography (CTCA) are tests of anatomy.

The advent of the invasive pressure wire has provided an accurate tool for the direct assessment of coronary physiology at the time of IC and until recently was therefore unique in allowing for assessment of both anatomy and RMI at the same time. However, the wide range of choice offered by this menu of noninvasive and invasive tests cannot hide the fact that there are important limitations to them all. Specifically, the noninvasive stress tests are highly specialised, expensive, not freely available to patients in all hospitals and associated with significant waiting time in the UK. Invasive angiography carries small but definite risk, is expensive and again not freely available. Furthermore, the ͞yield͟ in terms of finding significant CAD when IC is used as the front-line screening tool for patients with new onset stable CP is relatively low. For example, in 398,978 patients in one large US registry undergoing diagnostic coronary angiography, only 37.6% had obstructive CAD (defined as 50% or more of the diameter of the left main coronary artery or 70% or more of the diameter of a major epicardial coronary artery). Even in patients in whom CAD is identified, there is a well described discrepancy between the anatomical appearance and physiological evidence of RMI in a variety of clinical studies and this effectively calls into question whether an anatomy-based test is an appropriate tool with which to define the management of such patients.

The inherent limitations of both non-invasive and invasive tests for the accurate diagnosis and management of each patient presenting in this context is reflected in the diversity of management in the real world. This is well demonstrated by the 2010 NICE Guideline for Chest Pain of Recent Onset in which the recommended algorithms for assessment and management of this cohort are complex and disparate. Furthermore, one consequence of a true negative stress test is that clinicians do not establish if the patient has a coronary blockage which may represent an indication for disease-modifying therapy. By contrast, a true positive stress test inevitably then results in referral to the waiting list for IC, thus delaying the time from assessment to definitive diagnosis and management plan significantly. Conversely, a referral directly for IC in the absence of evidence of RMI subjects the patient to a stressful procedure that carries potential for serious complication merely to establish in many cases that there is no underlying cause for the CP presentation.

The ideal screening test for patients with stable CP with which to determine the coronary anatomy and physiology would be non-invasive, quick, simple, widely accessible, accurate (using a combination of IC + FFR as the gold standard reference) and as cheap as possible. However, it is possible to refine the desirable features of such a screening test even more precisely.

Firstly, the ideal test would avoid the expense, inconvenience and risk of invasive investigation in patients who turn out to have no important atheroma, whilst being able, conversely, to identify even non-flow limiting atheroma in patients who may therefore benefit from disease-modifying medical therapy.

Secondly, an ideal screening test would incorporate the value, in terms of patient outcome, of physiological assessment that has been demonstrated in the clinical trial data accumulated using the coronary pressure wire. Specifically, two basic principles of management can be applied at the diagnostic stage if it includes a physiological assessment. Firstly, that stenting an anatomical lesion deemed significant which is FFR (fractional flow reserve) negative is associated with a worse outcome than medical therapy alone. Secondly, that treating pressure wire positive lesions with OMT alone is associated with a worse clinical outcome than PCI (percutaneous coronary intervention) plus OMT is just as important. In summary, the ideal screening test for patients with stable chest pain needs to be able to deliver information about RMI (or a physiological surrogate of it), as well as anatomy. FFRCT (Fractional Flow Reserve Computerised Tomography -A non-invasive technique using CT to determine Fractional Flow Reserve) represents a plausible candidate for a screening test in this population. FFRCT is a novel method for computer-derived estimation of coronary physiology from CT angiography data, and early validation studies with regard to its diagnostic performance are promising. The original development and availability of computerised tomographic coronary angiography (CTA) has introduced this non-invasive assessment of the presence and extent of CAD as a useful additional option for the investigation of patients with suspected stable angina. Recent data from the SCOT-HEART randomised trial, for example, highlights the diagnostic value of this technique in such a population.

However, in the PROMISE trial, which randomised 10,003 patients with symptoms suggestive of CAD to stress testing or CTA, there was almost 50% more ICA performed in the CTA group without any difference in clinical outcome. Importantly, most revascularisations performed in the CTA group had no objective evidence of myocardial ischaemia. This highlights the potential limitation of anatomy-based screening in patients with symptoms thought to be due to myocardial ischaemia.

Recently, using sophisticated image assessment, computational fluid dynamics, and computer modelling, it has become possible to model FFR from the data obtained from CTA, thus known as FFRCT. A series of validation studies have assessed the diagnostic accuracy of this technique, including most recently the NXT trial, which demonstrated superior accuracy of CTA plus FFRCT versus CTA alone, using an invasive FFR of <0.80 as the reference.

RATIONALE AND RISK BENEFITS FOR CURRENT STUDY

This trial aims to test the hypothesis that FFRCT, used as the default screening tool for patients presenting with recent onset stable chest pain, would be associated with

(i) shorter rapid time period between initial consultation and definitive

management plan

(ii) better patient experience;

(iii) lower overall use of resources.

The UK is well suited to test this hypothesis because of its well-established system of Rapid Access Chest Pain Clinics (RACPC). The majority of patients presenting with stable CP that is of suspected cardiac origin are referred to such clinics, with a mandated access time within two weeks. The majority of such clinics work to the algorithm recommended in the NICE guidelines for Chest Pain of Recent Onset (March 2010). Within this guideline, patients are stratified according to their risk profile and pre-test likelihood of CAD to outcomes that include discharge, stress test, CTCA, CT coronary calcium score and IC.

Given the relative streamlining of this initial assessment of such patients throughout the country, it facilitates a comparison of strategies in FORECAST.

The 2 strategies for the FORECAST trial are:

A) TEST: all patients undergo CTCA+FFRCT as the default test, assuming they have no pre-specified contradictions to CT angiography. The result of the CTCA will determine whether the FFRCT will be performed. Results of the FFRCT will be conveyed to the supervising physician within 24 hours and will be used to determine the subsequent management plan.

B) REFERENCE: all patients will be assessed and managed exactly as they are usually treated by that centre and that RACPC using the local algorithms interpreted from the NICE Chest Pain of Recent Onset Guidance.

The study aim - To determine whether in a population of patients presenting to the rapid access chest pain clinic routine FFRct is better in terms of resource utilisation, (i.e. number and cost of investigational procedure, number of hospital visits etc.) when compared to routine clinical investigations recommended by NICE.

Primary endpoint - Resource utilisation at 9 months.

Secondary endpoints include -

- the comparison of clinical outcomes between the two groups at nine months.

(a) Major adverse cardiac events including: non-fatal MI, all coronary revascularisation, cardiac-related

hospitalisation.

(b) Requirement for non-invasive cardiac investigations

(c) Requirement for invasive coronary angiography

(d) Revascularisation

(e) Procedural complications

- the comparison of the effect on general wellbeing between the two groups at 9 months.

(a) QoL questionnaire

(b) Patient satisfaction questionnaire

(c) Angina status

(d) Time to definitive management plan

(e) Time to completion of initial management

(f) Number of hospital attendances

(g) Working days lost

Data minimisation:

The number of data items sent to NHS Digital for linkage is limited to:

i) name

ii) date of birth

iii) NHS Number

With these determinants, the research team can accurately follow consented patients tracking specified clinical events and resource utilisation as stipulated in the protocol (e.g. cardiac investigations, hospitalisations, revascularisation procedures, outpatient appointments). Using these determinants, all relevant data can be adequately linked to each patient's correct and unique study identifier, to ensure pseudonymisation of all patients.

The research team at the University of Southampton do not propose to collect any data that does not contribute to either the primary or secondary endpoints. Eleven UK sites have enrolled participants into the Forecast trial. On receiving the data from NHS Digital, the data processor (The University of Southampton) will collate and process in preparation for publication.

The study team have excluded any mortality, mental health and maternity data as this is not required.

The legal basis for processing the information requested in that it is necessary for the performance of a task carried out in the public interest (GDPR Article 6(1)(e) and under Article 9(2)(j) necessary for archiving purposes in the public interest, scientific or historical research purposes in accordance with Article 89 (1).

The data requested under this agreement is carefully selected and filtered to support the primary and secondary endpoints. Study results will be of interest to both potential patients undergoing diagnosis in the future and the wider NHS in ensuring the effective and efficient use of resource whilst diagnosing, managing and treating patients in the most cost effective, clinically effective way. Study participants have consented to the use of their data for this purpose. There are no foreseen ethical issues.

Recruiting Trusts provide Southampton Clinical Trials Unit (CTU) who are based within the University of Southampton and situated at the University Hospital Southampton, with a copy of the patient consent form that is accompanied by a flagging form for those participants who agree to long term follow-up. The flagging form provides the source for identifiable information for the linkage of participants to long-term clinical and resource data in the format of - name, date of birth and NHS number. These data will be collated by the CTU and submitted as a single application to NHS Digital with a request for processing.

The University of Southampton (CTU) will send the cohorts name, date of birth and NHS number to NHS Digital for linkage to HES data.

NHS Digital will link the cohort to HES patient index and supply HES data.

- filter Admitted Patient Care data by episode start date >= consent date (start date)

further filter by OPCS codes which will be supplied.

- filter Outpatient data by appointment date >= consent date (start date)

further filter by OPCS codes which will be supplied.

No identifiable data will be returned to the University of Southampton (CTU) as the Study ID will be sufficient and no re-identification will take place.

Processing will require the collation of data to support both the primary and secondary endpoints for this study -

Primary endpoint - objective

To determine whether, in a population of patients presenting to RACPC, routine FFRct as a default test is superior, in terms of resource utilisation, when compared to routine clinical pathway algorithms recommended by NICE.

Endpoint used to evaluate resource utilisation at 9 months -

a) Requirement for non-invasive cardiac investigations

b) Invasive angiography

c) Revascularisation (including PCI & CABG)

d) Hospitalisation for cardiac event (including MI/unstable angina/revascularisation/heart failure/arrhythmia)

e) Cardiac medications

Secondary endpoint - objectives

To compare clinical outcomes between the two groups at 9 months.

Endpoint used to evaluate -

(a) Major adverse cardiac events including: non fatal MI, all coronary revascularisation, cardiac-related hospitalisation.

(b) Requirement for non-invasive cardiac investigations

(c) Requirement for invasive coronary angiography

(d) Coronary revascularisation

(e) Revascularisation procedural complications

- the comparison of the effect on general well being between the two groups at 9 months.

(a) Number of hospital attendances

Data is held securely within Southampton Clinical trials Unit which is based at the University of Southampton. The unit has controlled access to the building and to individual rooms within the building. Access is limited to individuals that have been vetted and approved through the allocation of swipe cards. All guests to the unit are required to sign in and out and are accompanied at all times. Computer access is password protected to limited individuals employed within the unit. Storage areas are securely maintained and all reasonable protection from floods, fire and physical damage are in place.

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract ie: employees, agents and contractors of the Data Recipient who may have access to that data.

Data will only be accessed and processed by substantive employees of The University of Southampton and will not be accessed or processed by any 3rd parties not named in this DSA, or for any other purpose.

Patient data processed by NHS Digital will be used to clarify the role of CTCA +/- FFRct in the investigation pathway for patients presenting to rapid access chest pain clinics (RACPC) across the UK. Current guidance from NICE will be reviewed in the light of the Forecast study results.

The data from the NHS Digital download will be used for the main study analysis and it is this final report that will be available for NICE to review and consider, using aggregated results in line with NHS Digital suppression guidance.

At no point during the processing and publication of this data will individual identifiable patient information be released. Aggregate data will be used to address the principle study question of whether in a population of patients presenting to RACPC, routine FFRct as a default test is superior in terms of resource utilisation at 9 months when compared to routine clinical pathway algorithms recommended by NICE CG95.

Manuscripts will be submitted to high impact journals for consideration of publication in addition to presentations at national and international conferences.

It is planned to submit to the Lancet and will also submit a request to present at the European Society of Cardiology (ESC) Conference 2020.

Both the publication and presentation are designed to target clinical colleagues within the field of cardiovascular medicine. Southampton CTU will provide study results for research sites in a format suitable for participants. A lay study report will also drafted. This text will be supplied to all participating sites who will be able to pass this on to study participants. This may be via email or letter and will be at the discretion of the local PI and research team.

The current NICE guidelines recommend CT coronary angiography as the first line investigation for typical and atypical stable angina. NICE medical technologies guidance site HeartFlow FFRct as an additional assessment that can be performed in conjunction with the CT coronary angiogram as this may avoid the need for invasive coronary angiography and revascularization. NICE predict the use of FFRct may lead to cost savings of £214 per patient and by adopting this new technology nationally an overall saving of £9.1million as a minimum may be made to the NHS by 2020. The Forecast study will provide NICE with evidence to either support or question this assumption.

No NHS Digital data will be provided to researchers running other research studies within University Hospital Southampton NHS Foundation Trust or the University of Southampton.

All outputs will contain only data that is aggregated with small numbers suppressed in line with NHS Digital Guidance.

The results of the Forecast study have the potential to change the current investigational pathway for patients presenting to RACPC according to the current NICE guidelines.

The study aims to determine whether routine use of CTCA +/- FFRct is superior in terms of resource utilisation when compared to current pathways.

Healthcare Benefits

A positive finding for this study would offer a potential fiscal benefit to individual Trusts and the NHS as a whole by reducing overall costs for the routine investigation of patients presenting to RACPC with cardiac sounding chest pain.

Clinicians will be alerted to this information through publication and presentations. Should the NICE guidelines be amended in the light of the Forecast results, this information will be disseminated to Trusts and clinicians in the standard fashion for NICE recommendations.

A competitive tariff for CTCA +/- FFRct as a standalone study providing both anatomical and physiological information in a single test would reduce the overall number of tests required to finalise a treatment plan, leading to potentially significant cost savings.

Patient benefits

From a patient perspective the opportunity to undergo a single non-invasive test to acquire both anatomical and physiological information is appealing, with significant clinical risk reduction, speedier diagnosis and for a number of patients the avoidance of an unnecessary invasive procedure.

Patients are offered the optimum test that provides clinicians with the most sensitive and accurate information from which a diagnosis and treatment plan can be generated. Clinical care teams will continue to follow the recommendations released by NICE to maximise patient safety and comfort whilst ensuring the most appropriate use of valuable NHS resource.

These potential benefits will be measured through the collection of data to support the primary and secondary endpoints of the study.

Yielded Benefits is not a requirement for new applications.