Early prognostic stratification and identification of irreversibly shocked patients despite primary percutaneous coronary intervention

Background Despite prognostic improvements in ST-elevation myocardial infarction (STEMI), patients presenting with cardiogenic shock (CS) have still high mortality. Which are the relevant early prognostic factors despite revascularization in this high-risk population is poorly investigated. Methods We analyzed STEMI patients treated with primary percutaneous coronary intervention (PCI) and enrolled at the University Hospital of Trieste between 2012 and 2018. A decision tree based on data available at first medical contact (FMC) was built to stratify patients for 30-day mortality. Multivariate analysis was used to explore independent factors associated with 30-day mortality. Results Among 1222 STEMI patients consecutively enrolled, 7.5% presented with CS. CS compared with no-CS patients had worse 30-day mortality (33% vs 3%, P < 0.01). Considering data available at FMC, CS patients with a combination of age ≥76 years, anterior STEMI and an expected ischemia time > 3 h and 21 min were at the highest mortality risk, with a 30-day mortality of 85.7%. In CS, age (OR 1.246; 95% CI 1.045–1,141; P = 0.003), final TIMI flow 2–3 (OR 0.058; 95% CI 0.004–0.785; P = 0.032) and Ischemia Time (OR = 1.269; 95% CI 1.001–1.609; P = 0.049) were independently associated with 30-day mortality. Conclusions In a contemporary real-world population presenting with CS due to STEMI, age is a relevant negative factor whereas an early and successful PCI is positively correlated with survival. However, a subgroup of elderly patients had severe prognosis despite revascularization. Whether pPCI may have an impact on survival in a very limited number of irreversibly critically ill patients remains uncertain and the identification of irreversibly shocked patients remains nowadays challenging.


Introduction
ST-segment elevation myocardial infarction (STEMI) incidence in Europe ranges from 43 to 144 per 100,000 per year 1 and despite modern STEMI network organization, reperfusion techniques and new antithrombotic therapies 2 is still burdened by significant mortality. 3 Primary percutaneous coronary intervention (pPCI) is the treatment of choice in STEMI patients, and it must be performed as soon as possible in order to improve complication rates and outcomes. 4,5 Several conditions are known to be associated with increased mortality: age, Killip class, time delay to treatment, treatment strategy, previous myocardial infarction, diabetes mellitus, etc. [6][7][8][9][10] Among these, cardiogenic shock (CS) presentation, defined as persistent hypotension despite adequate filling status with signs of hypoperfusion, remains a leading cause of death, with inhospital mortality rates up to 50% and over. 11 Moreover, the combination of these factors may exert an exponentially detrimental effect on survival, making pPCI futile in some irreversibly critically ill patients.
Some of the key points of the management of STEMI reported by guidelines are based on randomized trials. 7,12,13 However, one of the major limitations of clinical trials performed in the emergency setting is their applicability in the real world. Indeed real-world registry data may evaluate some specific populations which are often excluded by randomized trials, due to the complexity of the clinical scenario and also difficulties in obtaining informed consent, as in patients presenting with CS. It is important to understand challenges in clinical practice, so the aim of our study was to evaluate our registry data in order to understand the characteristics of the CS population, the outcomes, and to identify those patients who, based on data available at the first medical contact (FMC), have the highest mortality risk.

Population and outcome
We analyzed all the STEMI patients undergoing pPCI, consecutively enrolled in the Trieste University Hospital pPCI registry from January 2012 to June 2018, in order to have a homogeneous population regarding technical strategies and pharmacological treatment. Acute STEMI was defined as: (a) presentation within 12 h of typical symptoms onset; (b) 1 mm or greater ST-segment elevation in two contiguous leads of electrocardiogram (ECG) (!2 mm in precordial leads).
CS was defined as: Anterior STEMI were 41%, 42.6% had multivessel coronary disease at index angiography and 10.5% of STEMIs were complicated by cardiac arrest. Complete characteristics of the overall population are listed in Table 1. Among these patients, 91 (7.5%) presented with CS. Descriptive characteristics of the population according to CS status are presented in Table 2.
Cardiogenic shock population compared to no cardiogenic shock The CS population compared with the no-CS population had a lower percentage of males (62% vs 75%, P < 0.001) and was significantly older (71 vs 65 years old, P < 0.01). Moreover, the incidence of CS was higher in older patients ( Fig. 1). CS patients compared with no-CS patients had more frequently cardiac arrest presentation (30% vs 9%, P < 0.001), worse renal function (CKD prevalence 50% vs 21%, P < 0.001), had a more severe coronary artery disease (multivessel disease in 57% vs 42%, P ¼ 0.005) and worse final angiographic results (TIMI flow 2 or 3 in 88% vs 96%, P ¼ 0.007). Overall ischemia time was borderline significantly shorter in CS patients (2 h 50 min vs 3 h 10 min, P ¼ 0.079).

Outcomes
In the overall STEMI population in-hospital mortality was 5.1%, 30-day mortality was 5.4% and 10.8% at a median follow-up of 38 (Inter Quartile Range 16-57) months.
Descriptive analysis of patients with CS according to 30day mortality is presented in Table 3.

Discussion
In an all-comers contemporary series of real-world STEMI patients, the principal findings of our study that focused on CS patients are: (1) a nonnegligible number of STEMI patients (7.5%) presented with CS; (2) we confirmed that the incidence of CS increases with age, 8,15 which is a major independent predictor of mortality; (3) successful PCI and short total time of ischemia are factors which have impact on prognosis, highlighting the importance of STEMI network organization, and application of current reperfusion techniques; (4) a subgroup of elderly patients with CS presentation due to anterior STEMI and with long ischemia time presented a severe prognosis despite revascularization.
Mortality rate after STEMI has decreased over the past two decades, 16,17,18 and we confirmed very good outcomes especially in young STEMI patients without CS (1.1% inhospital mortality). Conversely, STEMI presenting with CS have still high 30-day mortality (33%). As expected, CS patients compared with no-CS patients were older, had more often intercurrent cardiac arrest and their renal function was worse compared with no-CS patients, probably reflecting both a more severe comorbidity profile before the event, and the results of severe end-organ hypoperfusion. Interestingly we have found an increased presence of female patients in the CS setting compared with the no-CS setting. Women are on average older than men, they have a worse risk profile (more hypertension, diabetes mellitus, peripheral vascular disease), are more prone to developing mechanical complications and they have more often microvascular dysfunction. 19 Ischemia time was shorter in CS patients compared with no-CS patients, probably because of shorter 'patient delay', an important component of ischemia time, 20 due to the more severe symptoms. Indeed, the presence of higher peak troponin and lower LVEF indicates more extensive damage. In addition, CS patients have more often multivessel coronary artery disease.
Our registry included a high rate of right ventricular MI in the CS population (1/4 of CS patients, and 4% of the total registry). This may explain also the better LVEF and the lower heart rate (due to bradyarrhythmia during the acute phase) compared with other major studies. 7,21 Importantly RV dysfunction in STEMI has been shown to be an important predictor of mortality. 22 However, patients with transient RV dysfunction have a better prognosis compared with patients who have persistent RV dysfunction after revascularization. 22 Strategies of revascularization in CS have been recently evaluated by the CULPRIT-SHOCK 7,12 trial. Complete revascularization in multivessel patients undergoing pPCI was found to be associated with worse outcome (a composite of mortality and need for renal replacement therapy), and therefore in patients with CS multivessel PCI during pPCI it is no longer recommended. Usual practice in our center is to not perform the nonculprit lesions during the index procedure of STEMI presentation but to defer the procedure. 23 Thus all patients with CS and multivessel disease were treated with PCI of the culprit lesion only.
More evidence regarding the potential role of mechanical circulatory supports [Impella, TandemHeart, extracorporeal membrane oxygenation (ECMO), etc.] in CS patients is necessary also considering the downgrading of the intraaortic balloon pump (IABP) recommendation in the guidelines. 24 Indeed IABP failed to show a significant benefit on 30-day mortality in patients with CS complicating acute myocardial infarction. 13 In our population    7 However, our registry included patients treated before theIABP usewas downgraded by guideline in the CS setting. To date, Impella did not show superiority in terms of short-term mortality against IABP in several different trials, 25,26,27 whereas there are some favorable data in support of ECMO. 28 These devices, however, are mainly reserved for young patients. In this regard, age was found as a major factor influencing prognosis. As can be observed in Fig. 3, the stratification according to age (cutoff 75 years old) and the presence or absence of CS is itself capable of affecting heavily the prognosis: patients younger than 75 years old and without CS have an excellent prognosis whereas patients with the combination of age>75 years and CS presentation have a dramatic increase in 30-day mortality of up to 50%. Elderly patients have various reasons to have a high mortality: the aging process is associated with frailty and higher prevalence of comor-bidities; 29 moreover elderly patients have often atypical symptoms, which may delay prompt diagnosis. However, a decline in mortality of elderly CS patients has been observed in the last few years, possibly because of more extensive use of PCI in these patients. 30 Even the data from the SHOCK trial confirmed the advantage of PCI in elderly patients. 21 As shown in our registry, an effective revascularization in CS patients is paramount, indeed obtaining a final TIMI flow of 2 or 3 was an independent prognostic factor associated with lower 30-day mortality. Despite current STEMI guidelines 24

Limitations
We must acknowledge limitations to our study. First, the results of our findings should be interpreted in light of the common limitations of a registry-based cohort study. Second, the population is relatively small, in particular when subgroups are taken into account. We did not evaluate the role of acute kidney injury during hospitalization on outcomes. Finally, our pPCI registry included patients considered amenable to pPCI. Nevertheless, some of these limitations are also the strength of these data which provide a pure representation of real-world all-comer STEMI patients treated with pPCI.

Conclusion
In CS patients, age is a relevant negative factor whereas a successful PCI is positively correlated with survival, confirming that prompt revascularization is pivotal in the management of STEMI complicated by CS. However, a subgroup of elderly patients with CS presentation due to anterior STEMI and with long ischemia time presented a severe prognosis despite revascularization. Whether pPCI may have an impact on survival in a very limited number of irreversibly critically ill patients remains uncertain and the identification of these irreversibly shocked patients is nowadays challenging.