Percutaneous Coronary Intervention in Stable Coronary Artery Disease: Still in Equipoise?

Yusuke Yoshikawa; Takeshi Kimura


Eur Heart J. 2019;40(2):187-189. 

Appropriate indications and prognostic benefits of percutaneous coronary intervention (PCI) remain controversial in patients with stable coronary artery disease (CAD). In the COURAGE trial, addition of PCI to optimal medical treatment for patients with stable CAD was not associated with reduction in all-cause death or myocardial infarction (MI).[1] Based on the COURAGE trial results, the number of PCI procedures for patients with stable CAD has substantially decreased in US. More recently, the ORBITA trial has reported that PCI as compared with the sham procedure was not associated with improvement in exercise capacity in stable patients with angiographically significant coronary artery stenosis.[2] On the other hand, in the FAME 2 trial, fractional flow reserve (FFR)-guided PCI as compared with medical therapy alone was more effective in reducing the primary endpoint, a composite of all-cause death, MI, and urgent revascularization, and also it was shown that FFR-guided PCI more effectively relieved angina.[3,4] Similarly, in the DANAMI-3–PRIMULTI trial and in the Compare-Acute trial, FFR-guided PCI for non-culprit lesions in stabilized ST-segment elevation myocardial infarction (STEMI) patients was associated with lower risk for the primary composite endpoints including coronary revascularization than medical therapy alone.[5,6] However, the results favouring FFR-guided PCI in all these three trials have been criticized, because the differences in the primary outcomes between the FFR-guided PCI and medical therapy groups were mainly driven by revascularization without a large difference in mortality or MI.[3–6] The open-label trial designs with a possible placebo effect of PCI might well have influenced the number of revascularization events in both groups in opposite directions. Also, by the trial designs, the total number of the first revascularization procedures was greater in the FFR-guided PCI group than in the medical therapy group, further making it problematic to include revascularization as a component of the primary endpoint.

In this context, Zimmermann et al. have reported a meta-analysis using patient-level data from the three randomized trials comparing FFR-guided PCI using new-generation drug-eluting stent (DES) vs. medical therapy in patients with stable CAD or stabilized STEMI (the FAME 2 trial,[3,4] the DANAMI-3–PRIMULTI trial,[5] and the Compare-Acute trial[6]) in the current issue of the European Heart Journal.[7] After a median follow-up of 35 months, FFR-guided PCI was associated with a substantially lower risk for the composite endpoint of cardiac death or MI than medical therapy only, which was mainly driven by the significant reduction of MI (Table 1). The authors and the investigators of the three trials should be congratulated on providing the crucially important finding from this sophisticated meta-analysis of individual patient data. Certainly, MI is a more clinically relevant and relatively 'hard' endpoint compared with revascularization. This result would be a good reason for us to recommend PCI in many patients with stable CAD who have a significant lesion with FFR ≤0.80 (Figure 1). Furthermore, in the DEFER study, PCI as compared with medical therapy alone in patients with FFR ≥0.75 was associated with a numerically higher cumulative 5-year incidence of cardiac death or MI.[8] Therefore, FFR is a reliable gatekeeper for the decision-making for PCI in patients with stable CAD.

Figure 1.

Percutaneous coronary intervention in stable coronary artery disease. FFR is useful in decision-making for PCI but is still in equipoise due to many issues to be elucidated.

However, in real clinical practice, penetration of lesion assessment based on 'physiology' such as FFR is still very limited, and decision-making for PCI often involves visual angiographic assessment only. It is important to note that there were a substantial proportion of patients without any FFR-positive lesion (27% in FAME 2, 31% in DANAMI-3–PRIMULTI, and 50% in Compare-Acute) in whom PCI might have been performed without any benefit if there was no FFR information.

There might be a caveat in advocating FFR-guided PCI in patients with stable CAD. There was no signal suggesting mortality benefit with an FFR-guided PCI strategy despite substantial risk reduction for MI. The present study did not have adequate statistical power for evaluating the mortality difference. Nevertheless, it would be important to consider the reasons for the absence of mortality benefit with an FFR-guided PCI strategy. In this respect, it might be important to consider the clinical significance of MI as a clinical trial endpoint. The present study did not report the types and severity of MI, which would be important to understand the present study results fully with its clinical implications. We previously reported that patients who had a small post-discharge MI after PCI with a creatine kinase level <3 times the upper limit of normal were not associated with excess mortality risk as compared with patients without a post-discharge MI.[9] Therefore, it would be clinically important to know how many large MIs were prevented by FFR-guided PCI. At the end of the day, we could accept the decision-making for medical therapy alone according to the preference of the patients because of the absence of signal suggesting mortality benefit with an FFR-guided PCI strategy (Figure 1).

The important study of Zimmermann et al. raises more questions than answers. First, the 5-year mortality rate was relatively low in the present study (7.0% in the FFR-guided PCI group and 6.5% in the medical therapy group), suggesting selective inclusion of low-risk patients. What would be the mortality outcomes of FFR-guided PCI vs. medical therapy alone in patients with more severe CAD with higher mortality risk who were unlikely to be enrolled in these types of randomized clinical trials and were likely to have been treated by revascularization? Secondly, improvement of angina is an important goal of PCI in patients with stable CAD. Should we have different thresholds for PCI between symptomatic and asymptomatic patients? Thirdly, what would be the appropriate management strategy in patients with so-called 'grey zone' FFR values (0.75–0.80)? Fourthly, should we have different thresholds for FFR-guided PCI between elderly and non-elderly patients? There remain many questions about the indications of PCI in stable CAD.

A large ongoing randomized controlled trial, the ISCHEMIA trial, has recently finished recruitment of >5000 patients.[10] Unlike the three trials in the present meta-analysis, invasive coronary angiographic assessment is reserved until failure of optimal medical therapy in the conservative strategy arm of the ISCHEMIA trial, while in the invasive strategy arm, FFR is recommended for all lesions with a site-assessed diameter stenosis of <80% unless non-invasive evidence of ischaemia is already present in that myocardial territory. We are eagerly awaiting the results of this crucially important trial. Furthermore, ongoing improvement of FFR derived from computer tomography might obviate the need for invasive angiography and invasive FFR in a substantial proportion of patients with stable CAD.[11,12] Many more studies would be warranted to establish the optimal management of patients with stable CAD, which is a moving target in cardiology.