Finding a gatekeeper to coronary angiography: a step in the right direction

In recent years, coronary computed tomography angiography (CCTA) has come to the forefront in the evaluation of patients with chest pain.1 Multiple large scale randomized clinical trials (RCT) confirmed its high sensitivity and negative predictive value.2,3In addition, multicentre registries and RCTs confirmed its ability to triage patients with acute and chronic chest pain.4,5 Its high negative predicative value resulted in a low rate of referral to coronary angiography over a 3-year period in patients with normal or non-obstructive disease.6 However, its ability to accurately quantify stenotic lesions and obstructive CAD, particularly in intermediate lesions, is limited.7,8 This is expected to increase post-CCTA resource utilization especially in patients with intermediate lesions. Data from the Medicare claims data in the USA suggested that using CCTA as the assessment tool of Medicare patients was associated with increased referral to the coronary angiography compared to other non-invasive modalities.9 Similarly, a meta-analysis of RCT of CCTA in acute chest pain showed increased rate of angiography and revascularization without impacting outcomes.10 This could be in part due to the fact that anatomical stenosis does not necessarily correlate with haemodynamically significant stenosis. In the FAME study, 65% of angiographically intermediate lesions were not haemodynamically significant while 20% of initially thought obstructive (71–90%) anatomical lesions had normal fractional flow reserve (FFR).11 Recent technological advances with CCTA-derived FFR (FFRCT) or CT myocardial perfusion imaging Computed Tomography Perfusion (CTP) allows to partially overcome these limitations. The diagnostic accuracy of FFRCT is high and is comparable to angiography alone.12 However, despite its incremental diagnostic and prognostic value, FFRCT has some limitations including off-site performance, extra cost, and other logistic issues limiting its widespread use.13,14 In contrast, CTP can be performed on-site, but requires vasodilator stress agent. Adding CTP to CCTA has been associated with improved diagnostic accuracy of CCTA similar to a combination of angiography and single photon emission tomography (SPECT).15 Similarly, major adverse clinical events (MACE) and event-free survival from combination of CCTA and CTP were not different from the combination of angiography and SPECT.16 In this issue of the journal, Van Rosendael et al.17 report on the impact of adding CTP to CCTA on downstream referral to Invasive coronary angiography (ICA), revascularization and outcome in patients with stable new-onset chest pain. As per protocol, patients with normal CCTA did not undergo CTP. However, in patients with intermediate or high grade stenosis, CTP was performed. Thus, nearly one-third of the 384 patients referred for CCTA were found to have ≥ 50% stenosis and subsequently underwent CTP. Similar to previously reported data from nuclear imaging, almost half of the patients with significant anatomical stenosis had normal CTP. The rate of ICA and revascularization was much higher in patients with abnormal CTP compared to normal CTP. However, major adverse cardiovascular events rate in patients with obstructive CAD and normal CTP were low. Thus, supporting the concept that in patients with anatomical but no physiological disease, deferring angiography and revascularization is safe. Performing CCTA and sequential CTP reduced the rate of referral to ICA (five-fold reduction) and revascularization. This well conducted study demonstrated the feasibility of 1-day CCTA/CTP protocol and affirmed the role of CCTA in accurately ruling out CAD with limited radiation dose (7.5 mSv). It has also shown that the positive productive value of CCTA can be improved with CTP resulting in selective referral to angiography. Adopting such technique as a gate keeper would potentially improve post-CCTA angiography yield. The study also confirmed prior observations that nearly 50% of obstructive lesions on CCTA are not associated with ischaemia on physiologic testing. Adopting this approach would also provide an opportunity to evaluate the patient prognosis based on the presence of ischaemia and high-risk plaques. However, there are still many unanswered questions prior to a wide adoption of this approach in routine clinical practice. It appears that compared to invasive angiography, CTP was normal in 15% of patients requiring clinically indicated revascularizations and nearly 32% of patients with multi-vessel disease had normal CTP. This might even be an underestimate since CTP was not performed on around 17% of patients with significant disease by CCTA. Whether this is due to limited sensitivity of CTP, image quality or reproductivity of the technique is yet to be determined. Another important unanswered question is the comparison between CT-FFR and CTP in the rate of resource utilization post-testing. Weather either technique performs better in special populations needs further evaluation. Finally, despite having a 1-year follow-up duration, the event rate in this study is low which may affect the overall power of the study. Thus, major conclusions about patient outcomes cannot be drawn and longer follow-up of larger cohorts of CTP is needed. This study brings us a step in the right direction to increase the yield of coronary angiography after CCTA through providing a comprehensive assessment of anatomy and physiology. Once fine-tuned, CCTA will be the best gatekeeper to the catheterization lab. Conflict of interest: None declared.