Worldwide, ischemic heart disease is the single most common cause of death.1,2 Recent studies have shown a decrease in acute and long-term mortality after ST-elevation myocardial infarction (STEMI), mostly driven by the widespread use of reperfusion therapies, mainly primary percutaneous coronary intervention (pPCI), modern antithrombotic therapy, and secondary prevention.3,4 The main cause of in-hospital mortality after STEMI is cardiogenic shock, which is responsible for> 50% of early deaths, followed by mechanical complications (MC), which account for approximately one-third of early deaths.5–8

MC include left ventricular free wall rupture (FWR), rupture of the interventricular septum (VSR), and papillary muscle rupture (PMR).9–11 Classically, reported risk factors for the development of MC were female sex, advanced age, hypertension, higher heart rate, delayed admission, and late thrombolysis.12 FWR of the infarcted ventricle is relatively common in patients with acute postinfarct death. Several large studies have found cardiac rupture in 14% to 26% of such patients.13,14 Data reported in the late 20th century indicates that thrombolysis reduced the incidence of FWR compared with no reperfusion.15–17 Relative to VSR, its incidence was found to be low (around 0.2%) in patients receiving thrombolysis in the GUSTO-I trial,15 although an increased risk of septal rupture was observed in patients with single-vessel disease (especially of the left anterior descending artery), extensive myocardial damage, and poor septal collateral circulation.16,17 The CIVIAM, a recent Spanish registry including VSR, showed a reduction in 1-year mortality in the last few years, although the relationship with the type of reperfusion therapy or specific changes in the incidence were not completely reported.18 Finally, PMR is a life-threatening STEMI complication that accounts for approximately 5% of deaths in these patients.11

Most of the data available on MC incidence and outcomes are from classic studies using thrombolysis. The impact of pPCI on MC within STEMI networks is unclear. Accordingly, we aimed to analyze changes in incidence and mortality trends (at 28 days and 12 months) of post-STEMI MC over 2 decades: a decade before and a decade after the establishment of pPCI networks.

A total of 6033 consecutive STEMI patients were included in this analysis. The mean age was 61.7 (± 12.8 years and 78.8% were men. The pre-pPCI period (1990-2000) included 2250 patients and the pPCI (2007-2017) 3783 patients. Baseline demographic and clinical characteristics are shown in table 1. Patients in the pPCI period were older (62.4 vs 60.4 years, P <.001) and more often had hypertension and dyslipidemia, with lower rates of peripheral disease or previous myocardial infarction than those in the pre-pPCI period. The diagnostic criteria for hypertension, diabetes and dyslipidemia changed during the years of the registry, which could influence differences between periods. Patients with Killip-Kimball class III-IV declined from the pre- to the pPCI period by 35% (14.7% vs 9.5%). In the pPCI period, more than 80% of patients progressed without any signs of heart failure (Killip-Kimball class I).

Table 1 shows the medical management of the studied patients in the 2 periods. In the pPCI period, there was a significant increase in the use of antiplatelet drugs, beta-blockers, statins, and renin-angiotensin-aldosterone system inhibitors. Nevertheless, key differences were found between the 2 study periods relative to the reperfusion strategy. First, reperfusion increased by ∼1.8 fold between the pre- and pPCI periods. Second, among reperfused patients, 99.1% received thrombolysis in the pre-pPCI period, whereas 95.7% received pPCI in the pPCI period.

MC developed in 135 patients (2.2%), including 73 patients (1.2%) with FWR, 38 (0.6%) with VSR, and 24 (0.4%) with PMR (table 2). The prevalence of MC as a whole declined by 41% from the pre- to the pPCI period (from 3.0% to 1.8%, respectively, P <.001), mainly due to reduction in FWR. FWR showed a relative reduction of 60% in the pPCI period (2.0% vs 0.8%, P <.001), without significant inter-period changes for VSR (0.7% vs 0.6%) and PMR (0.4% both) (figure 1).

Figure 1.

Central illustration. Prevalence of post-STEMI mechanical complications after the onset of a primary percutaneous coronary intervention network. FWR, free wall rupture; PMR, papillary muscle rupture; pPCI, primary percutaneous coronary intervention; VSR, ventricular septal rupture.

(0.14MB).

The baseline characteristics of patients with and without MC are shown in table 3. Patients with post-STEMI MC were 10 years older and were more often women and hypertensive, with a lower prevalence of prior infarct. The use of reperfusion therapy was less frequent in patients who developed a MC (48.2% vs 75.0%; P <.001). Analyses by strata showed that among patients with MC, in the pre-pPCI period, 48.5% received thrombolysis and the remaining 51.5% did not receive reperfusion treatment, whereas in the pPCI period only 4.5% received thrombolysis and 43.3% received pPCI reperfusion therapy. To compare the prevalence of FWR in the 2 periods, logistic regression analysis modeling was performed with the following covariates: age, sex, hypertension, previous myocardial infarction, infarct location and reperfusion. After multivariate adjustment, the prevalence of FWR remained higher in the first period than in the pPCI period (OR, 1.93; 95%CI, 1.10-3.41; P=.023) (table 4).

Table 5 illustrates the in-hospital characteristics and management of MC during the 2 study periods. Surgical repair increased in the pPCI period in all MC (7.4% vs 37.3%; P <.001), but this difference was statistically significant only for FWR (4.5% vs 34.6%; P=.037). Overall, 28-day mortality was higher in patients with MC treated conservatively in both periods (pre-pPCI: 79.4% vs 20.0%, P=.012; pPCI: 88.1% vs 24.0%; P <.001). Lower 28-day mortality was observed in patients treated with cardiac surgery in the pPCI period for FWR (90.0% vs 33.3%; P <.004) and PMR (75.0% vs 8.3%; P=.027), with no significant differences in either the pre-pPCI period or in VSR. The use of pericardial puncture in patients with FWR was similar in both periods (∼45% of patients). The use of intra-aortic balloon pump was limited in the pre-pPCI period, but was used in half of the patients with VSR and in one-third of those with PMR in the pPCI period.

The 28-day mortality of all STEMI patients showed a relative reduction of 40% in the pPCI period (9.4% vs 5.6%; P <.001). Nevertheless, there were no significant changes in MC mortality as a whole, although a trend was observed for PMR (table 2).

One-year mortality of all STEMI patients showed a relative reduction of 25% (10.9% vs 8.2%; P <.001) in the pPCI period. No relevant changes were found in PMR (62.5% vs 43.8%; P=.386), VSR (56.3% vs 81.8%; P=.086) or FWR (86.4% vs 75.9%; P=.251) between the 2 periods (table 2).

The present study reports a large series of post-STEMI MC spanning over 2 decades before (pre-pPCI: 1990-2000) and after the implementation of a pPCI reperfusion network (post-pPCI: 2007-2017). Several findings may be highlighted. First, overall STEMI mortality showed a dramatic reduction both at 28 days and at 12 months. These data are in agreement with reports from other European regional networks.22,23 Second, the development of post-STEMI MC declined in the pPCI period, mainly driven by a reduction in FWR (figure 1). Third, short-term mortality related to the studied MC did not significantly benefit from more aggressive management with pPCI and cardiac surgery.

In this study, FWR was the only MC that showed a significant decline in prevalence with pPCI over thrombolysis. Early studies found a decrease in the prevalence of FWR with the early administration of fibrinolysis over no reperfusion, with the caveat that late fibrinolysis increased the rate of FWR.24–26 This latter association was speculated to be due to a major increase in the rate of myocardial hemorrhage. More recently, Puerto et al.27 reported MC in a cohort of elderly STEMI patients with mixed used of thrombolysis and pPCI and identified a 48% relative reduction in the incidence of FWR. Here, we found a 60% relative reduction in FWR in the pPCI decade with universal use of pPCI compared with thrombolysis. In our series, most patients in both periods obtained the benefit of reperfusion, which confers restoration of microvascular perfusion, limits the size of the myocardial infarction and the proportion of patients with transmural infarction, and favors remodeling of the left ventricle. Nevertheless, there are several possible explanations for the reduced prevalence of FWR in the pPCI period. First, the efficacy of thrombolysis to achieve TIMI grade 3 flow decreases when time to treatment increases, whereas pPCI has been shown to achieve infarct-related artery patency in> 90% of patients regardless of prehospital delays.28 Accordingly, the increased patency rate associated with pPCI may result in greater myocardial salvage and a reduced area of transmural infarction. Second, animal studies have demonstrated that there is a greater amount of myocardial edema, hemorrhage and injury after reperfusion with thrombolytic therapy than with reperfusion by pPCI.29

There are few reports on the prevalence of VSR, as most studies combine FWR and VSR in the same category under the term “cardiac rupture”12,30 due to the low rate of VSR. Early registries including information on the trend in VSR showed a reduction in prevalence with thrombolysis compared with no reperfusion (0.2%-0.3% vs 1%-2%).15,31–33 In these studies, a very low rate of pPCI was observed: in the GUSTO-1 trial15 all patients received thrombolysis, in the MIDAS registry31 19% of patients were reperfused with pPCI, and in the GRACE registry32 38% of patients were reperfused with pPCI. The more contemporary APEX-AMI study,33 with a pPCI rate of 94%, found an incidence of VSR of 0.17%, similar to previous data with thrombolysis. Collectively, these data indicate that the incidence of VSR has declined with the use of reperfusion therapies, although the widespread use of pPCI has failed to reduce this complication compared with thrombolysis, in line with the findings reported here.

Several factors may contribute to the pathogenesis of VSR. At the biomolecular level, in the context of reperfusion injury, VSR autopsy data reveal the presence of a myocardial hemorrhage in 83% of patients treated with pPCI, in 71% of patients who received fibrinolysis, and in 18% of patients without reperfusion therapy.34 Previous studies have demonstrated that myocardial hemorrhage can create dissections in the infarcted myocardium and delay the healing process. At the biophysical level, it is known that the distribution of the mechanical load in the septum differs from that in the anterior free wall, which could increase the stress on the infarcted area in the interventricular septum. Although in-hospital prognosis is still ominous, the management with extracorporeal membrane oxygenation in selective patients could be an option to improve prognosis. 18,35

The prevalence of PMR reported in the literature is so low that addressing PMR prevalence trends with management changes remains elusive. Here we did not observe changes in PMR with thrombolysis or pPCI. This finding could be due to a very low prevalence of PMR and that we only included confirmed papillary muscle ruptures after ruling out acute papillary muscle dysfunction, since, in the absence of rupture, conservative medical management is the best option with favorable outcome. Despite stable PMR prevalence rates in the 2 study periods, pPCI together with a broader use of intra-aortic balloon pumps and cardiac surgery tend to improve survival at 12 months of follow-up.

The natural history of post-STEMI MC is catastrophic, and urgent surgical repair provides the best chance of survival in patients with MC. The present study shows increasing use of emergent surgery in the pPCI period, together with better medical management, but nonetheless the survival of these patients at 28 days and 12 months remains poor. Accordingly, there is an unmet need in the management of post-STEMI patients with MC. On the one hand, preventing reperfusion injury might be a novel target in future adjunctive STEMI treatment to reduce inflammation and its associated hemorrhage.36 On the other hand, mechanical circulatory support, including ECMO, as a bridge to surgery, is being evaluated.35 Such circulatory support could improve patients’ hemodynamic status, allowing end-organ recovery for a few days before the surgical repair. The dark side of this is that VA-ECMO should not be extended more than 4 to 5 days to prevent its hemorrhagic and thrombotic complications.

Two decades were examined to better understand prevalence trends and outcomes of post-STEMI MC. The establishment of regional pPCI networks has modified the landscape of MC in STEMI. Relative to the pre-pPCI period, FWR declined by 60% in the pPCI era, likely due to greater myocardial salvage and to pPCI reducing the area of transmural infarction. The prevalence of VSR and PMR remained unchanged in the 2 decades studied. Despite increasing use of emergent surgery in the pPCI period, 28-day and 12-month mortality remain exceedingly high. Adjunctive cardioprotection against reperfusion injury and mechanical support prior to surgery require further research if we are to improve current suboptimal results.

None.

All authors contributed to the study according to international consensus on authorship and have approved the final version.

The authors have no relationship with industry related to this paper, and no disclosures.

.