The Value of CT Fractional Flow Reserve

Overview

Coronary Computed Tomography Angiogram (CCTA) is a non-invasive imaging modality that has high sensitivity and negative predictive value for the detection of coronary artery disease (CAD). The main limitations of CCTA are its poor specificity and positive predictive value, as well as its inherent lack of physiologically relevant data on hemodynamic significance of coronary stenosis, a data that is provided either by non-invasive stress tests such as myocardial perfusion imaging (MPI) or invasively by measurement of the Fractional Flow Reserve (FFR). Recent advances in computational fluid dynamic techniques applied to standard CCTA are now emerging as powerful tools for virtual measurement of FFR from CCTA imaging (CT-FFR). These techniques correlate well with invasively measured FFR \[1-4\]. The primary purpose of this study is to evaluate the incremental benefit CT-FFR as compared to CCTA in triaging chest pain patients in outpatient settings who are found to have obstructive CAD upon CCTA (\> 30% and \< 90% stenosis). Invasive FFR and short term clinical outcomes (90 days) will be correlated with each diagnostic modality in order to evaluate positive and negative predictive value of each when used incrementally with CCTA. This will be an observational trial in which patients will undergo a CCTA, as part of routine care. If the patient consents to participate in the study and is found to have coronary stenosis of 30% to 90%, based on the cardiologist's reading, the CCTA study will be sent to HeartFlow, a vendor that will provide a computerized FFR reading, based on the CCTA study. If the noninvasive FFR diagnosis indicates obstructive disease, the patient will be recommended to undergo cardiac catheterization with invasive FFR. As CCTA utilization increases, the need to train additional imaging specialists will increase. This study will assess the capability of FFR-CT to enhance performance on both negative and positive predictive value for less experienced readers by providing feedback based on CT-FFR evaluation. CCTA readers will be grouped in two categories: those with more than 10 years reading experience and those with less than 10 years reading experience. Each CCTA will be read by a less experienced and a more experienced reader. Results from each reader will be correlated with each other and with the CT-FFR and invasive FFR results.

CCTA is increasingly becoming a preferred non-invasive imaging modality because of its high sensitivity and negative predictive value for the detection of CAD. It has been shown to be a robust imaging modality for evaluation of chest pain, and is associated with decreased unnecessary hospital admission, length of stay, major adverse cardiovascular event rates, recidivism rates, and downstream resource utilization compared to standard evaluation \[5\]. While findings so far are highly suggestive of CCTA's significance as a gatekeeper for Invasive Coronary Angiography (ICA) by ruling out obstructive CAD, fewer than half of obstructive stenosis identified by CCTA are ischemia-causing, signifying its poor positive predictive value and inherent lack of physiological information \[6-8\]. Consequently, utilization of CCTA has not entirely averted need for downstream testing for functional assessment of CCTA-detected obstructive lesions either by stress testing or ICA. Recently a major treatment modality, associated with the use of CCTA, has become available that offers promise for improving positive predictive value and physiological relevant hemodynamic data. Advances in computational fluid dynamic techniques applied to standard CCTA are now emerging as a powerful tool for virtual measurement of FFR from CCTA imaging (CT-FFR). This techniques correlate well with invasively measured FFR \[1-4\]. CT-FFR is not an investigational agent, having been approved by FDA in November, 2014. However, more work is necessary to delineate the patient population that could derive maximal benefit from this new technology. While few publications regarding the use of CT-FFR specifically address the cost of diagnostic work-up for obstructive disease, it is clear that the cost structure resulting from changes in diagnostic testing will also change. Deferral or avoidance of cardiac catheterization and nuclear stress testing will likely yield significant reductions in the cost of the diagnostic testing. From 1/1/2009 to 3/31/2015 our team introduced and operated a CCTA Chest Pain triage program for low to intermediate risk patients at Stony Brook University Hospital ED and non-emergency outpatient services, the only tertiary care hospital in Suffolk County. Concurrently, we established a registry to monitor patient outcomes for all patients receiving CCTA at Stony Brook Medicine. Our registry contained nearly 15,000 patient CCTA procedures. Our major registry study established the effectiveness of CCTA as an imaging modality for evaluating Emergency Department chest pain in a cost efficient manner with a false negative rate less than 1% \[5\]. However, our registry reflects the poorer positive predictive values documented by other industry studies \[6-8\]. False positive workup results in the necessity of performing cardiac catheterization on patients at risk for obstructive disease based on assessment with current standard of care (combined screening with CCTA and stress MPI). Reduction in the rate of false positive testing would lead to reduction in risk from invasive procedures and radiation exposure to patients and reduced cost to the health care system. Several medical institutions currently use HeartFlow CT-FFR as standard of care for evaluating obstructive disease. Generally, the standard of care at these institutions is to refer patients who are 30 to 90 percent obstructed by CCTA and who have reduction of flow \<= to 0.8 that is deemed to be medical significant by the attending cardiologist to Invasive FFR. HeartFlow has reported to us confidentially that this routine use of CT-FFR has resulted in a 54% reduction in false positive rate as compared to use of CCTA alone. The purpose of this study is to evaluate the incremental benefit of Fractional Flow Reserve derived from CCTA (FFR-CT) compared invasive FFR as the gold standard for patients with obstructive disease (\> 30% and \< 90% stenosis). This study will also assess the capability of CT-FFR to enhance performance on both negative and positive predictive value for less experienced readers by providing feedback based on CT-FFR evaluation. This will be a prospective observational clinical trial designed to evaluate the incremental benefit of virtual FFR measured from CCTA, compared to invasive FFR and CCTA for the detection of flow-limiting coronary stenosis, as defined by invasive FFR \<0.8 and vessel diameter of \>=2mm. 572 consecutive patients who present to Lenox Hill Hospital Outpatient Clinics for CCTA due to chest pain or stable angina over a ten month period and meeting the study inclusion criteria are eligible for the study (Figure 1). Our team will employ CCTA-appropriateness criteria to ensure proper selection of patients, derived from the Appropriate Use Criteria for Cardiac Computed Tomography published in 2010 and jointly authored by multiple societies including American College of Cardiology Foundation (ACCF), Society for Cardiovascular Computed Tomography (SCCT), and American College of Radiology (ACR) \[10\]. FFR-CT measurements will be performed by a core laboratory in a blinded fashion. All eligible patients will undergo 64-slice or greater multi detector CCTA and CT-FFR measurements. The severity of the stenosis will be determined on site by level III CCTA readers. Patients with obstructive lesions of (30% to 90% stenosis) will be referred for Stress-MPI, per SOC protocol, and CT-FFR. Patients with positive Stress-Myocardial Perfusion Imaging (MPI) and CCTA (50% - 70%%) or positive CCTA (71% to 90%) will be referred for undergo ICA with invasive FFR measurement in accordance to accepted guidelines and established practice standard. They will also received CT-FFR. Those patients with invasively measured FFR\<0.8 and with vessel diameter of \>= 2mm, or those who require revascularizations based on invasively estimated stenosis severity will be considered to have flow-limiting obstructive CAD, while the rest will be considered to have non-flow limiting obstructive CAD (if also \>50% stenosis on ICA). If stenosis severity turns out to be \< 50% after ICA (the gold standard), we will conclude that these patients have non-obstructive CAD. (Figure 1). Patients with 30% to 49% obstructive disease, according to CCTA, will be referred to optimal follow up care only. Any in this group who have positive CT-FFR will return for ICA with invasive FFR measurement, and follow the protocol for those with 50% to 90% obstruction. Defining an event as performance of ICA when no intervention is necessary, we expect to compare event rates for patients treated with CCTA and FFR-CT, using t-tests and a multivariate, risk adjusted 90 day hazard model with 95% confidence interval. Our null hypothesis is that outcomes will not vary regardless of which testing is used to assess obstructive disease. Our alternative hypothesis is that evaluation with FFR-CT as oppose to CTTA will change the event rate. We will also correlate noninvasive and invasive FFR studies. We will assess inter-observer reliability of the two reader cohorts' (\> 10 years experience and less than 10 years experience) readings for CCTA and CT-FFR.. We will also assess the nondiagnostic rate for CT-FFR exams as compared to independent quality ratings of CCTA scans by the scan reader. Scans will be rated as excellent, good, adequate, or non-diagnostic. Non-diagnostic exams will not be sent to CT-FFR. For those sent to CT-FFR we will compare the percentage of non-diagnostic exams to the percentage of non-diagnostic for CT-FFR, and we will also correlate exam results by level of obstruction.