Recurrent myocardial infarction (re-MI) is one of the most common adverse cardiovascular (CV) events that may occur after an episode of acute coronary syndrome (ACS). According to the 4th Universal Definition of Myocardial Infarction, re-MI is defined as the MI that occurs after 28 days following the index MI event. Differently, an MI that occurs within 28 days of the first index event is defined as reinfarctions [1]. However, current studies investigating ACS rarely use this distinction and adopt generic definitions as “MI” or “new MI”. In the present review, re-MI is applied to any acute coronary events following the index MI. The main objectives of this review are to summarize evidence on the incidence and prognosis of re-MI and to discuss the most effective strategies that proved successful in significantly reducing re-MI rates in the current revascularization era.

Hospitalization rates due to MI have steadily decreased over the last 30 years [2, 3]. Obviously, re-MI incidence has dramatically changed as well.

From 1980s to the early 1990s, re-MI incidence appeared to be extremely high, without significant changes over time [4]. Later, a number of studies documented a progressive and slow decrease in re-MI occurrence, though case fatality tended to be stable [5, 6]. Buch et al. [6] compared two different cohorts of first MI survivors developing re-MI included in the National Danish Patient Registry between 1985–1989 and 2000–2002. In 1985–1989, early re-MI (within 30 days) and late re-MI (31–365 days after the index MI) occurred in 2.5% and 9% of patients, respectively. In 2000–2002, early re-MI and late re-MI occurred in 4.4% and 6.6% of patients, respectively, with a significant decline in related mortality.

In the era of percutaneous coronary intervention (PCI), the overall incidence of re-MI dramatically decreased over time, though rates were not consistent across registries [7, 8, 9]. In a sample of 48,688 US patients of Medicare beneficiaries who suffered MI between 2001 and 2009, a progressive decline in 1-year re-MI occurrence from 7.6% to 5.8% was observed [7]. In a larger cohort of patients hospitalized for acute MI from 1999 to 2010 (2.3 millions), re-MI rates declined from 12.1% in 1999 to 8.9% in 2010, with a relative reduction of 26.4% [8]. These data were confirmed in a recent study that reported a reduction in re-MI rates at 3 years from 7.1% to 5.1% in 4169 patients during a more contemporary period [10].

Regarding the incidence of in-hospital re-MI, i.e., re-MI occurring during the first hospitalization and often related to a PCI procedure (Type 4 MI), the FAST-MI registry showed comparable rates of re-MI for ST-elevation MI (STEMI) and non-ST-elevation MI (NSTEMI) with values approaching 1% [10]. Similar data from other European registries have recently been published [11, 12].

Differently, after the in-hospital phase, re-MI incidence seems to be strongly related to readmission rates during the first year after the index MI. Based on the results from several studies evaluating this event in contemporary cohorts, 30-day rehospitalization rates after MI vary from 12% to 20%, with almost 70% occurring within the first 2 weeks [13, 14]. As reported by Kim et al. [14], 11.3% of these patients may experience re-MI. Despite evidence that re-MI may represent one of the major causes of readmission during the first 30 days after the index MI [15], there are discordant data regarding the relative burden of re-MI in readmission causes at follow-up longer than 30 days. Culler et al. [16] retrospectively evaluated readmissions rates and causes of re-hospitalization in 143,286 patients discharged alive after MI in US during 2014. At 90 days, 28% experienced at least one readmission and 8% had more than one readmission. The main reasons for readmission were heart failure (HF) and need for cardiac surgery (15.3% and 10.1%, respectively), while re-MI occurred in 2.1% of patients with an average number of days to MI recurrence of 35.3 $\pm$ 29.7. In another study investigating readmissions in 296,965 US patients discharged after NSTEMI, 58.4% of total readmissions at 90 days were due to CV causes, with re-MI being the most frequent [17]. Notably, the risk of re-MI persists even at longer follow-up and re-MI rates actually tend to rise over the years after the index MI (13–16% at 7 years), though more than half of events occur within the first year [18, 19].

The prognostic impact of re-MI may be dramatic in patients surviving a first coronary event. Significantly higher mortality rates both at 30 days and 1 year have been reported in patients suffering re-MI compared to patients with no re-MI [20, 21]. In a substudy of TRITON-TIMI-38 (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel–Thrombolysis in Myocardial Infarction) including 13,608 patients with ACS, those who experienced a new MI had a significantly higher rate of CV death at 6 months compared to patients who had no re-MI (6.5% vs 1.3%; p 0.001) [21]. The prognostic impact of re-MI may be related to its timing. In a recent prospective cohort of 3387 patients, the average mortality rate at 1 year in re-MI patients was 32.2%, reaching 53.3% in those with early recurrences. In this population, re-MI was associated with a 25-fold increased risk of death at 1 year compared to patients with a single acute coronary event [19].

Due to its frequent occurrence and prognostic implications, re-MI is routinely included in the composite outcome of major adverse CV events (MACE) of studies conducted in patients with ACS [22]. Therefore, re-MI and global MACE tend to share many common risk factors. Conditions frequently associated with MACE in the ACS setting include age, female sex, prior MI, prior stroke, diabetes, left ventricular dysfunction, failed or not attempted revascularization, high Killip class, low systolic blood pressure, and renal failure [23, 24, 25, 26, 27].

Older age has been shown to be significantly associated with re-MI in several studies [19, 28] and is occasionally considered as one of the most important predictive factors of re-MI [29, 30]. Similar considerations may apply to diabetes [29, 30, 31], smoking status [32], female sex [28, 31, 33] or socio-demographic status [34]. In patients with STEMI and a prior history of stroke that account for 9% of all-comers STEMI, a two-fold increased risk of suffering re-MI has been observed during the first 30 days after the index coronary event when compared to other STEMI patients [35]. In addition, re-MI is not just a major cause of readmission in patients surviving the first MI, but it can also occur in patients hospitalized for other clinical reasons. As recently demonstrated by Wang and colleagues in a retrospective wide population of Medicare fee-for-service beneficiaries, there are at least 11 disease categories causing readmission that are significantly associated with re-MI, including diabetes, anemia, hypertension, coronary artery disease and HF [20]. This evidence supports the relevance of specific medical strategies designed to prevent all-cause readmissions in order to reduce re-MI rates and improve global patient’s prognosis.

Coronary revascularization has been one of the most debated topics regarding MI recurrence. Although PCI is universally recognized as the most effective strategy in reducing MACE and mortality following ACS [36, 37], some authors have suggested that PCI itself may be a risk factor for re-MI [19, 28, 31, 32, 33]. In a prospective cohort of 3283 ACS patients, prior coronary artery bypass grafting (CABG) and prior PCI were respectively the first and second strongest predictors of re-MI at 1-year follow-up [28]. Similarly, prior CABG and PCI were significantly associated with re-MI in a prospective population of 9615 patients [31]. It should be noted, however, that both studies enrolled patients starting from 2004–2005, long before the modern antiplatelet strategies and second-generation drug-eluting stents (DES) have become available [16, 17, 19, 38, 39, 40, 41, 42].

Growing evidence suggesting a protecting role of PCI on re-MI events comes from modern randomized clinical trials (RCTs) investigating complete revascularization of non-infarct-related artery (IRA) vessels in STEMI patients [39]. Since 2013, a number of studies [40, 41, 42, 43, 44] have been conducted with the purpose of demonstrating how routine revascularization of significant non-culprit lesions may improve outcomes, i.e., mortality and MACE (including re-MI). All these trials differ significantly by sample size and methods, and this has led to varying results on the effectiveness of PCI in reducing re-MI rates. Gupta et al. [45] included data from two trials, DANAMI-3-PRIMULTI (Primary PCI in Patients with ST-elevation Myocardial Infarction and Multivessel Disease: Treatment of Culprit Lesion Only or Complete Revascularization) (n = 627) and Compare-Acute (Comparison Between FFR guided Revascularization versus Conventional Strategy in Acute STEMI Patients with MVD) (n = 885), with the aim to assess if a fractional flow reserve (FFR)-guided strategy of complete revascularization could improve outcomes during a follow-up of 12 to 44 months. A similar analysis design, adding another minor RCT to the previous two ones, was used by Wang and colleagues in a more recent meta-analysis [46]. Both studies demonstrated a significant reduction in MACE rates and unplanned or ischemia-driven coronary interventions, with no evidence of of a higher risk of re-MI in patients undergoing complete revascularization.

Recently, the results from the COMPLETE trial (Complete vs Culprit-only Revascularization to Treat Multi-Vessel Disease after Early PCI for STEMI) have been published. In this trial, 4041 patients were randomized to an IRA-only strategy versus complete revascularization. At a mean follow-up of 36.2 months, complete revascularization significantly reduced the co-primary outcome of CV death and new MI (hazard ratio [HR] 0.74, 95% confidence interval [CI] 0.60–0.91). This result was mainly driven by the lower incidence of new MI in the IRA-only PCI group compared to the complete-revascularization group (HR 0.68, 95% CI 0.53–0.86), whereas no significant differences were reported in CV death (HR 0.93, 95% CI 0.65–1.32) or all-cause death (HR 0.91, 95% CI 0.69–1.20) [44]. After the publication of the COMPLETE trial, other meta-analyses showed significantly improved outcomes associated with complete revascularization. In an analysis including 10 RCTs for a total of more than 7000 patients, this strategy was found to be effective in reducing both CV death (OR 0.69, 95% CI 0.48–0.99) and re-MI (OR 0.68, 95% CI 0.49–0.96). In patients undergoing complete revascularization, re-MI incidence was 5.1% at a median follow-up of 29.5 months (compared to 6.9% in the IRA-only group, p = 0.03) [47]. Consistent results were reported in two additional meta-analyses with a comparable study design [48, 49, 50]. Notably, the recent FLOWER-MI trial (FLOW Evaluation to Guide Revascularization in Multi-vessel ST-elevation Myocardial Infarction) (n = 1163 with STEMI) suggested that an FFR-guided strategy may not be superior to an angiography-guided strategy in STEMI patients undergoing complete revascularization (primary outcome: HR 1.32, 95% CI 0.78–2.23; non fatal re-MI: HR 1.77, 95% CI 0.82–3.84) [51]. All these data strongly support the hypothesis that PCI can significantly reduce re-MI after ACS and that the increased risk of type 4 MI associated with aggressive revascularization is largely counterbalanced by a reduction in type 1 re-MI incidence [52, 53, 54, 55, 56, 57] (Table 1, Ref. [45, 46, 47, 48, 49, 50]). In the STEMI setting, strict monitoring of ST-elevation resolution following PCI can be an effective tool in predicting the risk of recurrent events, including re-MI, at short and long-term follow-up [53].

Dual antiplatelet therapy (DAPT) is the cornerstone of pharmacological treatment of ACS. Since the early stages of both pharmaco-invasive and percutaneous treatment of MI, DAPT combining clopidogrel and aspirin showed to be strongly effective in reducing all-cause death, CV death, stent thrombosis (ST), and re-MI [54, 55]. Recently, new oral antiplatelet drugs and DAPT strategies have been investigated and approved in the MI setting, and currently either prasugrel or ticagrelor are strongly recommended by international guidelines [36, 37, 56].

In TRITON-TIMI 38, prasugrel proved superior to clopidogrel in reducing both re-MI (7.4% vs 9.7%; HR 0.76, 95% CI 0.67–0.85) and ST (1.1% vs 2.4%; HR 0.48, 95% CI 0.36–0.64) at 15 months, with no differences in overall mortality between treatment groups [57, 58]. In PLATO (Platelet Inhibition and Patient Outcomes) (n = 18,624), ticagrelor at a maintenance dose of 90 mg twice daily significantly reduced the rates of all-cause mortality, CV death, and MACE in patients with ACS compared with clopidogrel [59]. Re-MI incidence at 12 months was also significantly lower in the ticagrelor arm (5.8% vs 6.9%; HR 0.84, 95% CI 0.75–0.95). In PEGASUS-TIMI 54 (Prevention of Cardiovascular Events in Patients with Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin) (n = 21,162 with previous MI), a prolonged DAPT duration with aspirin and ticagrelor 60 mg twice daily vs. placebo significantly reduced re-MI rates at 3-year follow-up (HR 0.84, 95% CI 0.72–0.98) [60]. These findings have been confirmed in real-world registries suggesting that both prasugrel and ticagrelor are highly effective in reducing CV outcomes compared to clopidogrel, without major safety concerns [61, 62].

The ISAR-REACT 5 trial (Prospective, Randomized Trial of Ticagrelor versus Prasugrel in Patients with Acute Coronary Syndrome) was designed to compare the efficacy of ticagrelor and prasugrel in reducing all-cause death, cardiac death, and MACE. More than 4000 patients with ACS were randomized to receive ticagrelor or prasugrel, with PCI performed in more than 80% of cases. At 1 year, the composite primary endpoint (death, MI, or stroke) was significantly reduced in the prasugrel vs. the ticagrelor group (6.9% vs 9.3%; HR 1.36, 95% CI 1.09–1.70). The lower incidence of the primary endpoint was primarily driven by a reduction in re-MI incidence (3.0% vs 4.8%; HR 1.63, 95% CI 1.18–2.25), while the other individual components of the composite outcome were not significantly different between the treatment groups [63].

Cangrelor is a strong P2Y${}_{12}$ parental inhibitor and the latest to have been approved for clinical use. Differently from oral antiplatelet agents, cangrelor provides a prolonged inhibition of platelet activity with extremely rapid onset (2 min) and offset (60–90 min) periods [64]. Cangrelor efficacy in reducing ischemic endpoints was compared to clopidogrel standard treatment in the CHAMPION Program (Cangrelor versus Standard Therapy to Achieve Optimal Management of Platelet Inhibition), consisting of three RCTs enrolling both stable and unstable patients and differing each other by sample size and timing of clopidogrel administration. Both CHAMPION-PLATFORM (n = 5362 with stable and unstable CAD) and CHAMPION-PCI (n = 8877 with stable and unstable CAD) failed to demonstrate a significant reduction of the primary endpoint (death, MI, and ischemia-driven

revascularization) at 48 h in the cangrelor arm, despite some evidence suggesting a reduction in isolated death and ST [65, 66]. In a pooled analysis of these two trials that used a more precise definition of periprocedural MI, cangrelor was found to be associated with a significant reduction in early ischemic events when compared with clopidogrel [67]. Later results from the large CHAMPION PHOENIX trial (n = 11,145 with stable and unstable CAD) demonstrated a significant reduction of the primary efficacy endpoint (death, MI, ischemia-driven revascularization, and ST) in patients treated with cangrelor compared with patients treated with clopidogrel (4.7% vs 5.9%; OR 0.78, CI 95% 0.66–0.93). Notably, the benefit from cangrelor was mainly driven by lower rates of MI and ST (MI: 3.8% vs 4.7%; OR 0.80, 95% CI 0.67–0.97) [68]. A pooled analysis of these three trials, including $\sim$25,000 patients, definitely confirmed these findings [69]. It should be considered that the CHAMPION trials (particularly CHAMPION PHOENIX) included a large proportion of stable patients, so that most of the reported MIs during follow-up cannot be properly reclassified as re-MIs. Thus, even if more than half of patients enrolled in these trials experienced an ACS, specific trials investigating cangrelor efficacy in the acute setting are lacking [70] (Table 2).

Additional pharmacological treatments are currently recommended to further reduce CV events following MI. In the ACS setting, several RCTs and meta-analyses have proven that statins dramatically reduce short- and long-term outcomes, including re-MI [71, 72, 73]. Indeed, an intensive statin regimen may decrease non-fatal MI rates by 15% [74, 75]. Moreover, in the acute setting, a high-dose statin pretreatment may be associated with a reduced rate of procedural MI and injury [76, 77, 78]. At present, the European Society of Cardiology ACS guidelines strongly recommend statin therapy (if not contraindicated) to reduce MACE, MI, and CV death, regardless of baseline LDL-cholesterol levels [36, 37]. Other lipid-lowering agents have shown to be effective in reducing CV events and MI in patients with ACS. In IMPROVE-IT (Improved Reduction of Outcomes: Vytorin Efficacy International Trial), more than 18,000 ACS patients were randomized to simvastatin vs. simvastatin plus ezetimibe. At 7-year follow-up, the combined lipid-lowering therapy effectively reduced the composite endpoint of CV death, nonfatal MI, unstable angina requiring hospitalization, reintervention of coronary revascularization, and nonfatal stroke. Notably, re-MI was strongly reduced by ezetimibe combined therapy (HR 0.87, 95% CI 0.80–0.95; p = 0.002) [79]. As for PCSK9 inhibitors, both evolocumab and alirocumab have been shown to significantly reduce CV events (including MI) in patients with established atherosclerotic CV disease or ACS, respectively [80, 81]. Besides lipid-lowering therapies, in the recent REDUCE-IT trial (n = 8179 with multiple CV risk factors), icosapent ethyl (targeting triglycerides) proved effective in reducing MACE (including MI) and CV death in high-risk patients (HR 0.75, 95% CI 0.68–0.83) [82].

The effects of non-antiplatelet drugs in the post-MI setting are critically influenced by other pathological conditions that may frequently coexist. Among these, left ventricular systolic dysfunction confers a much higher risk of re-MI and drugs such as beta-blockers and angiotensin-converting enzyme inhibitors can dramatically improve outcomes after ACS, including re-MI [83].

The relationship between beta-blockers and CV events in patients with ACS without HF is still a matter of debate. A large meta-analysis on this topic including 16 observational studies failed to demonstrate any relationship between beta-blocker therapy and survival improvement [84]. Conversely, in a recent prospective study enrolling more than 13,000 Asian patients, beta-blocker treatment was associated with reduced CV death at 1-year follow-up (HR 0.70, 95% CI 0.589–0.834; p 0.001) [85]. Accordingly, there are data supporting the hypothesis that these agents may further reduce coronary events following an index ACS [89, 90, 91].

Re-MI is one of the most frequent complications occurring after an ACS episode.

Few dedicated epidemiological studies have tried to systematically evaluate the incidence and prognosis of re-MI in contemporary cohorts. It seems that the risk of re-MI persists for many years after the index event and approved secondary prevention strategies are able to only partially reduce its incidence, especially in subgroups at high risk. Novel pharmacological therapies and non-pharmacological strategies are highly needed in order to reduce the burden of re-MI and the overall residual risk after ACS.

LDL and LP drafted the paper; all other authors revised it critically.

Not applicable.

Thanks to all the authors and peer reviewers for their opinions and suggestions.

This research received no external funding.

The authors declare no conflict of interest.