Approximately 30–50% patients with acute ST-segment elevation myocardial infarction (STMEI) were found to have non-infarct-related coronary artery (IRA) disease, which was significantly associated with worse prognosis [1, 2]. However, challenges still remain for these patients: which non-infarct-related lesion should be treated and when should the procedure be performed? Previously, many STEMI guidelines from AHA/ACC/ESC didn’t recommend to offer complete revascularization for STEMI patients with multi-vessel disease during primary percutaneous coronary intervention (PCI) without hemodynamic instability, which could increase the rate of mortality [3‐5]. 2015 ACC/AHA guideline declared IIb recommendation for complete revascularization in selected STEMI patients with multi-vessel disease. Recently, for these patients, the updated 2017 ESC STEMI management guideline recommended complete revascularization that non-infarcted related artery lesion should be treated during either index procedure or index admission, following culprit lesion revascularization. This recommendation was based on the data from PRAMI, DANAMI-3-PRIMULTI, CVLPRIT and COMPARE-ACUTE trials, which favored the reductions in the risk of major adverse cardiovascular event (MACE) and repeat revascularization, not in all-cause or cardiovascular mortality rate. It is important to note that most of the time the evaluation of non-culprit lesion by angiography may not be accurate, because of underestimating or overestimating the lesion. Fractional flow reserve (FFR) can functionally evaluate the pathophysiological significance of the non-culprit lesion by using pressure wire in favor of functional angioplasty [6]. Moreover, the data from Fraction Flow Reserve Versus Angiography for Multi-vessel Evaluation (FAME) study at two years, showed FFR-guided PCI in patients with stable coronary artery disease lowered mortality rate and the rate of re-infarction, compared to angiography-guided PCI [7]. Therefore, we aimed to investigate whether FFR-guided functionally complete revascularization with PCI in patients with STEMI and multi-vessel disease could further improve the prognosis, especially the hard end point.

For this meta-analysis, we searched the PubMed/Medline, EMBASE, and the Cochrane library /CENTRAL databases and selected published RCTs which compared FFR-guided CR and COR in STEMI patients with multi-vessel disease up to May 12, 2018. The search terms included “Acute ST-segment elevation Myocardial Infarction”, “STEMI”, “Fractional flow reserve”, “FFR”, “Percutaneous Coronary Intervention”, “PCI”, “complete revascularization”, “culprit-only revascularization”, “Multi-vessel disease”, “culprit lesion” and “non-culprit lesion”. Additionally, presentations and abstracts were also searched from major cardiovascular conferences.

The inclusion criteria were: (1) published Randomized Controlled Trials (RCTs); (2) comparing FFR-guided complete revascularization PCI with culprit only revascularization PCI; (3) a study population of STEMI patients with multi-vessel disease.

The exclusion criteria were: (1) hemodynamic instability, such as cardiogenic shock; (2) previous meta-analysis or overlapping data.

Independently, according to the inclusion and exclusion criteria, three authors (L.J.W, S.H and X.H.Z) assessed RCTs eligibility and bias risk (Additional file 1: Figure S1 and Additional file 2: Figure S2), and extracted data. Their disagreements would be resolved by consensus.

The outcomes involved in the current study were major adverse cardiovascular events (MACE), all-cause mortality, myocardial infarction (MI) and repeated revascularization.

The current meta-analysis followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) study guideline [8]. Heterogeneity among the subgroups was estimated by the Cochrane Q-statistic test and I²-statistic test [9], whereby a P-value > 0.05 implied no statistically different result and a I² value <50% suggested a fixed effect model by using funnel plots assessed publication bias. Meta-analysis were carried out with Review Manager (RevMan) version 5.3(Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) software, obtaining Odds Ratios (OR) and 95% Confidence Intervals (CIs). a P value <0.05 indicated a statistically significant result.

In this meta-analysis from the comparison between FFR-guided CR and COR in patients with STEMI and multi-vessel disease, we found that FFR-guided CR resulted in low rate of MACE, including all-cause mortality, non-fatal myocardial infarction and repeat revascularization. The reduction of repeat revascularization was similar to that of MACE, which suggested the decreased need for revascularization favored the low incidence of MACE during the follow up period. Moreover, additional FFR-guidance did not significantly increase the rate of all-cause mortality and non-fatal myocardial infarction.

The results of our meta-analysis were in accordance with some previous studies. PRAMI trial [10] showed that angiography guided complete revascularization in patients with STEMI and multi-vessel disease during the primary PCI, using the criteria of percentage diameter stenosis>50% in one view for non-culprit lesion treatment, significantly decreased the rate of MACE and repeat revascularization without the reduction of the rate of all-cause mortality, as compared to COR. With the treatment criteria of non-culprit lesion changed into percentage diameter stenosis>70% in one view or>50% in two views, CVLPRIT trial [21] reported that angiography guided complete revascularization (during primary PCI: 73%; during staged PCI: 27%) was merely associated with low rate of MACE and repeat revascularization, compared with COR. Nevertheless, as the same as PRAMI trial, the risk of all-cause mortality was not changed.

In consideration of the Dissociation Between Angiographic results and Clinical outcomes in coronary artery Disease [22], FFR was gradually considered to be a good measurement for making decision on treating or not treating the coronary artery lesion, based on its high sensitivity and specificity in identification of ischemia [23]. In stead of the FFR threshold value of 0.75 [23], a FFR value of >0.8 suggested a non-ischemic lesion and good clinical outcome [24]. The results from FAME II trial at 3 years follow-up, demonstrated FFR-guided PCI in patients with stable coronary artery disease to lower mortality rate and the rate of re-infarction, compared to angiography-guided PCI [25]. A meta-analysis showed that FFR-guided PCI in patients with stable coronary artery disease was associated with significantly lower rate of re-infarction when compared to angiography-guided PCI [26]. Beside the patients with stable coronary artery disease, FFR measurement could be used effectively and safely in patients with acute myocardial infarction [14]. The data from a meta-analysis further confirmed the effectiveness and safety of FFR measurement in patients with acute coronary syndrome [27].

Ghani trial was the first RCTs of FFR measurement in patients with STEMI and multi-vessel disease. However, the result of this trial did not support FFR-guided CR early after primary PCI (described in Table 1) because of the high rate of the mortality and re-infarction [17]. Subsequently, as a relatively large trial, DANAMI-3-PRIMULTI [19] showed FFR-guided CR with staged PCI strategy (2 days after primary PCI and before discharge) might favor the reduction of MACE not the all-cause mortality. However, the latest large and multi-center trial, COMPARE ACUTE trial [20] revealed FFR-guided CR (during index PCI procedure: 83%; during index hospitalization: 17%) could not only reduce the rate of MACE, also numerically decrease the incidence of death from any cause without statistically difference.

Our meta-analysis was consistent to another two meta-analyses [28, 29], they confirmed that CR (including angiography-guided and FFR-guided) could merely decrease the incident of MACE and repeat revascularization, not the hard end point (all-cause mortality). Intriguingly, the percentage of all-cause mortality in FFR guided complete revascularization group was 1.3% in the Compare Acute trial and 4.7% in the DANAMI-3–PRIMULTI trial, with 50 and 31% of non–infarct-related lesions with a negative FFR value respectively. Probably, these good results could be derived from the accurate choose of revascularization. Additionally, our data also supported the reliability, feasibility and safety of FFR-guided complete revascularization (PCI) during acute phase of STEMI.

In comparison to COR, among patients with acute STEMI and MVD, FFR-guided functionally CR favored the reduction of the risk of MACE and ischemia-driven repeat revascularization, without a reduction in the rate of all-cause mortality. In the future, further large RCTs are required to investigate whether FFR guidance of complete revascularization significantly affects hard end point (all-cause mortality).