Data from 1485 patients enrolled in the INTERSTELAR registry cohort between 2007 and 2015 were reviewed. The INTERSTELLAR registry was designed to find out the prognosis of the patients after primary revascularization for STEMI in four regional hospitals (ClinicalTrials.gov identifier: NCT02804958) in Korea. Most of the participants in this study were Koreans, while approximately 3% were Chinese, Southeast Asian, Uzbekistani, and Mongolian. To compare the outcomes of separate strategies of RAAS blockade, we split the patients into two groups by two different standards. We excluded 58 patients with in-hospital death and 285 patients who were discharged without a RAAS blocking agent. Finally, 1142 patients were included in the first review.

Acute STEMI was diagnosed based on clinical information, including blood samples, chest pain, and 12-lead electrocardiography. An emergency primary revascularization call was made by $\ge$2 attending cardiologists at the time of patient presentation. Exclusion criteria were previously diagnosed coronary artery disease, valvulopathy (more than moderate), cardiomyopathy, pericardial disease, or congenital heart disease.

The 2016 European Society of Cardiology (ESC) Guidelines for the diagnosis and treatment of acute and chronic heart failure recommend ACEIs as class I drugs in patients with asymptomatic left ventricular (LV) systolic dysfunction and a history of MI to prevent or delay the onset of heart failure (HF) and prolong life [8]. However, in the 2017 ESC Guidelines for STEMI, the treatment indications for ACEIs have been broadened, and ACEIs should be considered in all patients with STEMI [9]. According to the guidelines, we defined adequate medication as the use of a dose equal to or higher than the recommended initiation dose of ACEIs with a concomitant use of beta-blockers. In contrast, we defined the following RAAS inhibition strategies as inadequate: (1) any use of ARB; (2) discharge with a lower-than-recommended ACEI initiation dose; and (3) use of an ACEI without a beta-blocker. Evidence-based recommended initiation doses are: captopril 6.25 mg thrice daily, enalapril (2.5 mg) twice daily, lisinopril (2.5 mg), and ramipril (2.5 mg) daily.

The primary endpoint was major adverse cerebrovascular and cardiovascular events (MACE), defined as a composite of MI, stoke, all-cause death, readmission due to symptoms of HF after PCI, and ischemia-driven target vessel revascularization. If the patients had multiple events, the first clinical event only was included in this analysis. Medical chart reviews and/or standardize phone call were included in this study. Medical status of all patients (RAAS blockade and beta-blockers) was assessed at discharge as well as at the first outpatient visit.

For set 1 comparison, we divided the patient population according to the type of RAAS blocking agent regardless of beta-blocker use: ACEI versus ARB. Among 1142 patients, 784 were prescribed an ACEI and 358 were prescribed an ARB at discharge. In set 2, we compared the adequate ACEI medication group with an inadequate RAAS blockade medication group. Among 1142 patients discharged with a RAAS blocking agent, data from 1064 patients were available for review. The study protocol was received from the institutional review board of Sejong General Hospital (SGH 2019-08-010). Written informed consent was obtained from all patients, and we complied with the Declaration of Helsinki (6th revision).

Patient characteristics were compared between the two groups using a two-sample t-test or Mann-Whitney U test for continuous variables and Pearson’s chi-square or Fisher’s exact tests for categorical variables. Continuous data are expressed as the mean $\pm$ standard deviation. Categorical variables are expressed as the number of objects and percentage. A Kaplan–Meier curve of MACE was created for each age group. The relationship between the time to MACE and use of ACEI and ARB was determined using Cox regression analysis with unadjusted and adjusted hazard ratios (HR).

For predisposition and to adjust the selection bias of RAAS inhibitor strategy, propensity score matching was used. The nearest neighbor method was used for the propensity matching algorithm, and 0.1 standard deviation was used for the caliper width with a matching ratio of 1:1. Co-variates for matching were chosen among those with p 0.10 in Table 1. To check the intergroup balance after matching, p value and the absolute standardized difference were used. The Matchit package in R software was used for the propensity score matching. Survival analysis was performed using the package suvminer in R (version 3.4.4; The corresponding software code for the open-source R project for statistical computing is provided; www.r-project.org) and Statistical Package for the Social Sciences (SPSS 18.0; SPSS Inc., Chicago, IL, USA), and p values 0.05 were considered statistically significant.

In set 1, patients in the ARB group were older (60.7 $\pm$ 13.4 years vs. 59.2 $\pm$ 12.8 years, p = 0.065) and had a higher prevalence of hypertension (56.3% vs. 43.1%, p 0.001), diabetes (30.6% vs. 23.1%, p = 0.007), hypercholesterolemia (31.2% vs. 13.3%, p 0.001) and ever-smoking (53.8% vs. 46.6%, p = 0.024) compared to those in the ACEI group. The initial Killip class was higher in the ARB group (Killip class 3–4, 14.0% vs. 8.5%, p = 0.005). Although the mean ejection fraction was lower (48.0 $\pm$ 9.6% vs. 49.5 $\pm$ 12.0%, p = 0.028), LV dysfunction was less common in the ARB group (18.7% vs. 23.7%, p = 0.063). The rate of single-vessel coronary disease was similar in the two groups (43.5% vs. 39.7%, p = 0.227) and the left anterior descending artery was the most common infarct-related artery in both groups (50.4% vs. 50.6%, p = 0.945). The stent diameters were also similar between the groups (3.36 $\pm$ 1.28 mm vs. 3.36 $\pm$ 0.49 mm, p = 0.961). Beta-blockers were prescribed to most patients, and no intergroup difference was noted (94.7% vs. 94.9%, p = 0.886). Newer potent antiplatelet therapy, including ticagrelor or prasugrel, was prescribed more frequently for patients in the ARB group (17.0% vs. 4.2%, p 0.001).

A total of 1064 patients were enrolled for the set 2 comparison, with 412 and 652 patients assigned to the inadequate medication and adequate medication groups, respectively. Patients in the inadequate group had a higher age (61.2 $\pm$ 13.9 vs. 58.7 $\pm$ 12.5, p = 0.003) and prevalence of diabetes (31.8% vs. 22.4%, p = 0.001), hypercholesterolemia (26.9% vs. 13.7%, p 0.001), ever-smoking (63.6% vs. 44.3%, p 0.001), and initial Killip class (Killip class 3–4, 13.4% vs. 8.1%; p = 0.005).

The stent diameters were similar in the 2 groups (3.27 $\pm$ 0.45 mm vs. 3.37 $\pm$ 0.50 mm, p = 0.001). Newer potent antiplatelet therapy, including ticagrelor or prasugrel, was more commonly prescribed for patients in the inadequate group (17.0% vs. 4.2%, p 0.001).

In this study, we analyzed physicians’ ARB preference for patients with renal dysfunction. As shown in Table 1, eGFR was lower in the ARB group (69.1 $\pm$ 26.4 vs. 78.0 $\pm$ 31.0; p 0.001), and there was a trend toward patients with chronic kidney disease (CKD) $>$grade 3 to be included in the ARB group (32.7% vs. 27.3%; p = 0.068). Similarly, the mean eGFR was lower (71.2 $\pm$ 29.4 vs. 78.8 $\pm$ 30.4; p 0.001) and a higher number of patients with CKD $>$grade 3 were included in the inadequate medication group (32.8% vs. 26.1%; p = 0.018).

In set 1, 278 pairs of patients were obtained after 1:1 propensity score matching. The parameters are listed in Table 2. Table 3 shows the comparison between the groups with adequate use of RAAS inhibitors and inadequate use of ACEI. There were no differences in age, body mass index, ejection fraction, and prevalence of diabetes, dyslipidemia, ever-smoking, hypertension, CKD $>$grade 3, multivessel coronary artery disease, use of ticagrelor and prasugrel, Killip class 3–4, and the left anterior descending artery as the culprit vessel between the groups. We categorized stent diameter in four groups as follows: (1) 2.5–3.0 mm; (2) 3.25–4 mm; (3) $\ge$4.5 mm; and (4) 2.25 mm. There was no intergroup difference in the beta-blockers prescription rate (94.9% vs. 93.8%, p = 0.712). There was no difference in the stent diameter between the groups. Likewise, there were no baseline differences between the continuous and categorical variables after matching for set 2 comparison (Supplementary Fig. 1. The figure shows the standardized mean differences before and after propensity matchings In set 1 and set 2 comparisons.)

In set 1, the incidence of MACE in the whole population during the follow-up was lower in the ACEI group than in the ARB group (Cox regression HR, 0.645; 95% confidence interval [CI], 0.440–0.944; p = 0.024). The outcome was similar after propensity matching (Cox regression HR, 0.535; 95% CI, 0.296–0.967; p = 0.039). The Kaplan–Meier survival curves for MACE for the whole and propensity-matched population are shown in Fig. 1.

Fig. 1.

The Kaplan–Meier survival curves for the composite of MACE (major adverse cardiac events; composite of myocardial infarction, stroke, all-cause death, readmission for heart failure, or target vessel revascularization) according to the use of angiotensin receptor blocker (ARB) or angiotensin-converting enzyme inhibitor (ACEI). (A) The left Kaplan–Meier curves shows comparison in the whole population. (B) The right Kaplan–Meier curves for the above medication groups in the matched population.

In set 2, the incidence of MACE during follow-up was higher in the inadequate medication group than in the adequate medication group (HR, 0.673; 95% CI, 0.459–0.985; p = 0.042). The outcome did not differ between the two groups after propensity score matching (HR, 1.023; 95% CI, 0.654–1.602; p = 0.919) (Fig. 2).

Fig. 2.

The Kaplan–Meier survival curves for the composite of MACE (major adverse cardiac events; composite of myocardial infarction, stroke, all-cause death, readmission for heart failure, or target vessel revascularization) according to the inadequate use of renin-angiotensin-aldosterone system (RAAS) antagonist and its adequate use. An inadequate use of RAAS antagonist is defined as: any use of ARB; discharge with an initiation dose lower than the guideline recommendation; use of ACEI without a beta-blocker. (A) The left Kaplan–Meier curves shows comparison in the whole population. (B) The right Kaplan–Meier curves for the above medication groups in the matched population.

Although both ACEI and ARB act by blocking a similar pathway in RAAS, their mechanisms of action differ. Theoretically, the beneficial effects of ACEI are strongly associated with increased levels of bradykinin. Bradykinin acts as a potent vasodilator by activating endothelial B${}_2$ receptors. Prostacyclin, nitric oxide, and endothelium-derived hyperpolarizing factor are stimulated by its release [10, 11]. The risk of acute MI was significantly lower with ACEI than that with placebo in large randomized trials. In the HOPE study, ramipril had a 20% lower event rate, and in the EUROPA study, perindopril had a 22% lower event rate [12, 13].

ARBs work by selectively blocking angiotensin II type 1 (AT1) receptors. The compensatory increase in angiotensin II levels may mediate harmful vascular effects via other angiotensin II receptors (AT2, 3, and 4), resulting in plaque instability or rupture [14, 15]. The risk of acute MI was slightly higher with ARB (candesartan and irbesartan) than with placebo [16, 17].

Ferrari et al. [18, 19] compared the vascular and endothelial functions between ACEIs and ARBs and found that ACEIs produce a greater reduction in endothelial dysfunction, inflammation, cell adhesion, and apoptosis, and have better antithrombotic and anti-atherosclerotic effects. ARB has only partial anti-thrombotic and inflammatory benefits. The ONTARGET study compared ACEI and ARB and showed a lower tendency for the risk of MI with ramipril as compared to telmisartan (relative risk, 1.07; 95% CI, 0.94–1.22).

In addition to choosing between ACEI and ARB, the clinically difficult decision about hemodynamic drug prescription would be which medicine to start with, and at what dose. A combination strategy of prescribing RAAS antagonists with beta-blockers might be a clinical concern when the patient’s blood pressure is low. Some physicians start with one type of hemodynamic drug and increase the dose quickly, while others start with two drugs (ACEI and beta-blocker) and gradually increase the dose. Interestingly, previous studies or guidelines do not conclusively favor one strategy over the other. Importantly, high-risk patients with LV dysfunction might require both classes of medicine before discharge; nevertheless, they usually have a higher chance of developing hemodynamic instability during hospital stay. Thus, we attempted to analyze set 2 for these clinical concerns.

Patients undergoing primary PCI usually stabilize after pain control. Some patients may experience hypotension or compensatory tachycardia after the procedure. The treatment strategies at this stage may differ depending on the physician’s experience. We did not include patients who either died before discharge or did not receive RAAS inhibitors or beta-blockers due to instability (343/1485, 23.0%). The 2017 ESC and 2013 American College of Cardiology/American Heart Association guidelines for STEMI recommend that ACEI should be considered in all patients with STEMI [20, 21]. Unlike the recommendation that beta-blockers should be used before discharge, ACEI are recommended within the first 24 h. According to the guidelines, a RAAS inhibitor should be considered first, followed by a beta-blocker. However, due to unstable vital parameters, a lower-than-recommended starting dose may be prescribed and not titrated until discharge. Similarly, unstable patients may not receive beta-blockers until discharge. There is evidence that, in practice, a small number of patients receive suboptimal doses of ACEI [22]. Interestingly, the use of a dose equal to or higher than the recommended initiation dose of ACEI according to the HF guideline with a mandatory addition of beta-blockers was not necessarily associated with a good prognosis.

There are several reasons that may explain the observed results. First, the role of a beta-blocker is not as strong as that of an ACEI in the primary PCI era. Beta-blockers are known to reduce mortality in patients with acute MI; however, most of this evidence comes from the pre-reperfusion era [23]. We must be aware that hospitalization rates for HF may increase further with the use of beta-blockers in the reperfusion era [23]. Second, in the absence of HF or systolic dysfunction, an equal risk of death was observed, regardless of the use of beta-blockers [24]. Indeed, we had less than a quarter of the population with LV dysfunction (ejection fraction 40%) in our study. Third, the impact of the ACEI dose at discharge might not be as high as the use of ACEI itself. It is apparent that a higher dose of ACEI is more beneficial for patients with a reduced ejection fraction [25]. Fourth, the correlation between the higher dose of a RAAS inhibitor and better prognosis seems to become weaker in the reperfusion era [26].