Introduction

Statins are widely prescribed for lipid-lowering management in the setting of primary and secondary prevention of coronary artery disease (CAD). Among patients with the acute coronary syndrome (ACS) such as those with ST-elevation myocardial infarction (STEMI) and non-ST-elevation ACS (NSTE-ACS), statins hold IA class of recommendation (CoR) in major guidelines and are stipulated to be initiated as early as possible, in the absence of contraindications and regardless of baseline cholesterol levels at the time of the acute event [1–3]. In North-American guidelines for STEMI management from 2013, high-dose atorvastatin of 80 mg daily is recommended in all STEMI patients without contraindications (IB CoR) [4].

However, the exact timing of statin initiation and clinical benefit of high-dose statin loading (pretreatment) before primary or delayed percutaneous coronary intervention (PCI) among patients presenting with ACS is unclear. The biological rationale supporting the early or very early [5] use of statins during the initial (unstable) phase of the acute thrombotic event is based on pleiotropic effects beyond lipid lowering that might act early while awaiting PCI [6–10]. Data corroborates that, among patients with both stable angina and ACS, a high-dose statin loading before PCI was associated with improvement in clinical outcomes, mainly driven by the reduction in periprocedural MIs and major adverse cardiovascular and cerebrovascular events (MACCE) during the vulnerable short-term period [11–13]. The largest-to-date randomized controlled trial (RCT) evaluating the effect of high-dose atorvastatin loading before planned PCI in the population of ACS patients provided mixed results; however, significant MACCE reduction was observed among patients that received PCI and had STEMI [14, 15].

Due to the present evidence gap for this problem and the fact that none of the previous systematic reviews qualified evidence using Grading of Recommendations Assessment, Development, and Evaluation (GRADE), we performed a meta-analysis with GRADE evidence qualification. We included all relevant RCTs to date that assessed the efficacy of early high-dose statin loading vs. no statin/low-dose statin/placebo before PCI among statin-naive patients with ACS on endpoints of MACCE, recurrent MIs, and all-cause death during 30-day follow-up. Secondarily, we underwent to determine the effects of high-dose statin loading before PCI on 30-day MACCE concerning statin used (atorvastatin or rosuvastatin) and ACS subtype.

Methods

The PRISMA [16] (preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were followed. Methods used to construct this analysis are elaborated to detail and provided in Appendix A.

Results

Search results

A total of 1520 records were identified from electronic databases. After duplicated were removed, 931 records were screened for eligibility, of which 893 were considered irrelevant. We obtained full texts for the 38 records and excluded 27 for reasons provided (Supplementary Table S1). This resulted in 11 randomized controlled trials being included in this review. The flowchart of study selection is shown in Fig. 1 according to updated PRISMA 2020 statement [17].

Fig. 1
figure 1

Flow diagram of study selection

Full size image

Included trials

We included 11 parallel design RCTs in this review, of which 5 were conducted in China [18–22], one in China and Korea [23], two in Italy [24, 25], one in the Netherlands [26], one in Korea [27], and one in Brazil [14]. Detailed information about the included studies is provided in Table 1. All included RCTs took place at the hospital setting and were reported between 2007 and 2018 while studies were conducted from 2005 to 2017. All studies were powered for at least 30-day follow-up.

Table 1 Baseline characteristics of included studies encompassing PICOS components and other elements
Full size table

Participants and clinical features

Overall, 11 RCTs enrolled a total of 6291 statin-naïve patients with ACS that were scheduled to undergo PCI, of which 3132 were randomized to high-dose statin loading before the procedure and 3159 comprised control group. Trials that examined rosuvastatin loading enrolled a total of 1300 patients of which 649 were randomized to high-dose rosuvastatin and 651 received no statin or placebo while trials involving atorvastatin loading enrolled a total of 4991 patients of which 2483 were randomized to high-dose atorvastatin while 2508 comprised control group. A total of 4743 (75.4%) patients received PCI while percutaneous coronary revascularization was planned in all enrolled patients. Among the remaining 1548 patients that did not receive PCI, the most common cause for not utilizing PCI was an indication for conservative medical treatment or referral to coronary artery bypass graft (CABG) surgery. In 9 out 11 studies, 100% of patients received PCI, while in one study, PCI receipt was 66% [25], and in another one, PCI receipt was 64.7% [14].

Across studies, mean age of enrolled patients ranged from 58 to 66.2 years, and a majority of those in the treatment group were men while similar distribution was present in the control group. Two trials enrolled only patients with STEMI, and eight trials enrolled only patients with NSTE-ACS while one trial included both patients with STEMI and NSTE-ACS (Table 2). Patients in both the treatment and control group were well-balanced in terms of distribution of comorbidities and global systolic function at baseline (Supplementary Table S2; Supplementary Figure S1). Likewise, the use of beta-blockers, ACE inhibitors and ARBs, and GPIIb/IIIa inhibitors during hospitalization was similar between two studied groups and across trials (Supplementary Table S3; Supplementary Figure S2).

Table 2 Clinical characteristics of included studies indicating study size, period of enrollment, follow-up period, distribution of acute coronary syndrome subtypes, mean age, sex distribution, PCI receipt proportion, and number of patients evaluated vs. randomized to treatment
Full size table
Table 3 Summary of findings showing quality of evidence with respect to studied outcomes, as assessed by the GRADE methodology
Full size table

Interventions

Patients were randomized to either receive high-dose atorvastatin or rosuvastatin. In all studies, statin was administered per os and before PCI. Atorvastatin was administered in at least 80-mg cumulative dose before PCI in all trials while rosuvastatin was administered in 40 mg cumulative dose before PCI in all trials except in a trial by Wang et al. [21] in which 20 mg rosuvastatin was administered at 2 to 4 h prior to PCI. Regarding the control group, patients in 6 trials [18, 20, 23–26] received no statin, and patients in 4 trials [14, 19, 21, 22] received placebo, while in one study [27], a low-dose statin (10 mg atorvastatin) was used as a control. Details of intervention vs. comparator across trials are shown in Table 1.

The timing of oral statin loading varied across trials—in most trials, loading dose was administered at 12 h before invasive management, whereas in some trials, this was commenced within few hours before the procedure. Furthermore, the concomitant standard of care antithrombotic treatment for ACS consisting of P2Y12 receptor antagonist (dominantly clopidogrel) and acetylsalicylic acid was administered to all patients before invasive management with the elective use of glycoprotein IIb/IIIa inhibitors (GPI) at the discretion of the operator. Maintenance antithrombotic and statin therapy were similar across trials and prescribed to all patients post-PCI (Supplementary Table S4).

Outcomes

The primary outcomes of this review were rates of MACCE, recurrent MI, and all-cause mortality at 30 days after ACS. MACCE at 30 days was explicitly stated or eligible for calculation in all included trials, as well as rates of recurrent MI and all-cause mortality, in a whole ACS population. In a secondary analysis stratified by ACS type, MACCE at 30 days in the setting of NSTE-ACS was available for analysis in all trials except in the Post et al. [26] and Kim et al. [27] that both investigated high-dose atorvastatin loading. On the other hand, in the setting of STEMI, 30-day MACCE could be analyzed only in 3 studies [14, 26, 27] that investigated high-dose atorvastatin loading while no studies examining the use of high-dose rosuvastatin in STEMI were available.

Funding

A majority of trials were free of industry-sponsored grants or funds. Six trials [14, 18, 21, 22, 25–27] reported receiving full funding, or at least partial grants from the government, universities, or private foundations. Two trials [19, 24] reported no external funding received while one trial [20] did not declare funding within the manuscript. Two trials [23, 26] declared receiving funds from the pharmaceutical industry (Pfizer Inc.).

RoB in included trials

Results of the RoB assessment in the included trials along with judgments and the supporting explanations are shown in Supplementary Table S5. Included trials generally had a low risk of bias with respect to performance, detection, attrition, reporting, and other potential biases while, on the other hand, most trials had unclear risk with respect to selection bias (random sequence generation and allocation concealment). Supplementary Figure S3 presents the summary of the RoB for each included trial. The percentage of high, low, or unclear risk of bias judgments across included trials is presented in Supplementary Figure S4.

Publication bias assessment and funnel plot asymmetry analysis were performed separately for the endpoints of MACCE, MI, and all-cause mortality at 30 days that included all 11 trials. This analysis revealed funnel plot asymmetry (Supplementary Figure S5) that was significant for the endpoint of MACCE at 30 days (Egger’s Z = -3.23, p = 0.001; Kendall’s τ = -0.24, p = 0.359) and MI at 30 days (Egger’s Z = -2.34, p = 0.020; Kendall’s τ = -0.31, p = 0.218) thus indicating a possibility of publication bias. Analysis of all-cause mortality at 30 days did not reveal significant funnel plot asymmetry (Egger’s Z = -0.64, p = 0.521; Kendall’s τ = -0.16, p = 0.54). Funnel plot asymmetry appeared to be largely driven by the absence of negative or neutral trials with respect to designated outcomes in trials that examined rosuvastatin loading (Supplementary Figure S6).

GRADE qualification of evidence

The quality of evidence for each primary and secondary outcome was qualified using GRADE. High, moderate, and low quality of evidence were determined for the primary outcomes of MI, MACCE, and all-cause death at 30 days, respectively. MACCE at 30 days endpoint was downgraded from high to moderate quality due to inconsistency—47% heterogeneity was detected and effects estimates from studies differed, especially concerning the study size. For the all-cause death endpoint, the level of certainty was downgraded to low quality due to imprecision—a very low number of events was observed; thus, the confidence intervals were large. Finally, findings that high-dose statin loading before PCI reduces the risk of MACCE at 30 days in NSTE-ACS and STEMI were based on the high and moderate quality of evidence, respectively. For the latter endpoint, the level of certainty was downgraded from high to moderate due to imprecision—only three studies were included in the analysis. The summary of findings based on GRADE qualification is provided in Table 3.

Effects of interventions

MACCE at 30 days

All studies contributed to effect estimates with an overall of 6291 patients. High-dose statin loading was associated with an overall 43% risk reduction of MACCE at 30 days (RR 0.57, 95% CI 0.44–0.71) in the whole ACS population (Fig. 2). This was predominantly driven by the effects of rosuvastatin (54% risk reduction) while atorvastatin reduced the risk of MACCE by 31%, although this was of borderline significance (p = 0.08). Low heterogeneity was detected while no differences between subgroups were observed (p = 0.15). Taken together, by combining the results of the meta-analysis and considering the level of certainty of the evidence, we found that high-dose statin loading before PCI reduces MACCE at 30 days.

Fig. 2
figure 2

Cumulative risk ratio (RR) of high-dose statin loading vs. control before PCI for major adverse cardiovascular events (MACCE) at 30 days

Full size image

Myocardial infarction at 30 days

All studies contributed to effect estimates with an overall of 6291 patients. High-dose statin loading before PCI was associated with an overall 39% risk reduction of MI at 30 days (RR 0.61, 95% CI 0.46–0.80) in the whole ACS population compared to the control group (Fig. 3). No significant heterogeneity across studies and no subgroup differences were detected (p = 0.33). Results of meta-analysis along with the certainty of evidence qualification together show that high-dose statin loading before PCI reduces the risk of MACCE at 30 days.

Fig. 3
figure 3

Cumulative risk ratio (RR) of high-dose statin loading vs. control before PCI for myocardial infarction at 30 days

Full size image

All-cause death at 30 days

All studies contributed to effect estimates with an overall of 6291 patients. Overall results obtained in the analysis suggested an 8% reduction of death at 30 days; however, due to imprecise results, no significant difference between high-dose statin loading and control group was found (RR 0.92, 95% CI 0.67–1.26; Fig. 4). No significant differences between subgroups were found (p = 0.34), and no heterogeneity was detected across included studies. This result of meta-analysis along with qualification of certainty of evidence shows that high-dose statin loading before PCI does not reduce the risk of all-cause death at 30 days, compared to no statin loading.

Fig. 4
figure 4

Cumulative risk ratio (RR) of high-dose statin loading vs. control before PCI for all-cause death at 30 days

Full size image

MACCE at 30 days in NSTE-ACS

Nine studies contributed to the effect estimates with an overall of 4955 patients. High-dose statin loading before PCI was associated with a cumulative 45% risk reduction of MACCE at 30 days in the NSTE-ACS (RR 0.55, 95% CI 0.38–0.80; Fig. 5A). This effect was significantly driven by rosuvastatin (52% risk reduction) while atorvastatin was associated with 35% risk reduction, although of borderline significance. No significant difference between subgroups treated with two different statins was detected in this population (p = 0.38, I2 = 0%). There was a moderate heterogeneity (60%) present among studies in this setting, largely due to the study of Yu et al.; however, we excluded this study in the sensitivity analysis, and similar effect estimates were obtained. Such results combined with a high level of certainty of evidence show that high-dose statin loading, particularly that with rosuvastatin, reduces the risk of MACCE at 30 days in the NSTE-ACS population.

Fig. 5
figure 5

Risk ratios (RR) of rosuvastatin or atorvastatin high-dose loading vs. control before PCI for MACCE at 30 days in (A) patients with non-ST-elevation acute coronary syndrome (NSTE-ACS) and B patients with ST-elevation myocardial infarction (STEMI)

Full size image

MACCE at 30 days in STEMI

Three studies of which all studied atorvastatin as intervention contributed to the effect estimates with an overall of 1255 patients. No studies for rosuvastatin were eligible for the analysis. No significant heterogeneity across studies was detected. Taken together with qualification of certainty of the evidence, a result of this meta-analysis shows that high-dose atorvastatin loading before PCI reduces MACCE at 30 days in the STEMI population, and this was attributed to 33% relative risk reduction (RR 0.67, 95% CI 0.48–0.94; Fig. 5B).

Exploratory PCI analysis

The analysis of PCI-only cohorts included a total of 4306 patients of which 2144 received statin loading while 2162 received no statin loading before PCI. We observed that, among patients with ACS, of which all received PCI, statin loading was associated with a 44% relative risk reduction of MACCE at 30 days (RR 0.56, 95% CI 0.43–0.74) (Fig. 6). This effect was consistent and significant for both rosuvastatin (RR 0.46, 95% CI 0.33–0.65) and atorvastatin (RR 0.65, 95% CI 0.44–0.97). Finally, a low degree of overall heterogeneity (I2 = 33%) across studies was detected, and there was no significant difference observed between trials concerning statin type used (p = 0.20).

Fig. 6
figure 6

Exploratory analysis of statin loading in population of patients with ACS that received PCI (100% receipt of PCI) with respect to the primary outcome of MACCE at 30 days

Full size image

Discussion

Data obtained from 6291 patients from 11 RCTs that were included in this meta-analysis showed that high-dose statin loading before PCI reduces the risk of MACCE and reduces the risk of MI at 30 days in the whole ACS population. High-dose loading of either statin does not reduce the risk of all-cause death at 30 days. These findings are based on evidence of moderate certainty concerning MACCE, high certainty with respect to MI, and low certainty concerning all-cause death. In the context of STEMI, high-dose atorvastatin loading before PCI reduces the risk of MACCE at 30 days, and this is backed by evidence of moderate certainty while no studies with rosuvastatin were available in this setting. In patients with NSTE-ACS, evidence of high certainty shows that high-dose rosuvastatin loading before PCI reduced the risk of MACCE at 30 days while atorvastatin had a borderline impact on this endpoint. These findings suggest that high-dose atorvastatin loading before PCI likely confers benefits in STEMI while high-dose rosuvastatin loading reduces adverse events in NSTE-ACS patients. These effects seem to be primarily driven by the significant reduction of recurrent MIs during the vulnerable 30-day period which translates to an overall reduction of MACCE, however, without impact on mortality. Of note, trials studying atorvastatin loading had larger patient enrollment compared to those studying rosuvastatin loading thus inferring its use in ACS potentially more valid. Finally, our exploratory analysis focused on patients with ACS, of whom all received PCI, confirmed our main findings since both atorvastatin and rosuvastatin loading consistently reduced MACCE at 30 days in this cohort of patients.

Our findings were derived from RCTs that included patients with ACS that we typically see in real-world clinical practice. Of note, we only included trials that enrolled patients naive to statin treatment and with the unstable disease. Equally important, three-quarters of all included patients underwent PCI while those that did not were in most instances referred to CABG surgery or conservative treatment. However, our exploratory analysis of PCI-only patients was in line with the main findings and showed that both statins reduced MACCE at 30 days.

We also focused our analyses on trials using the two most commonly used high-potency statins in clinical practice, atorvastatin, and rosuvastatin. Previous meta-analyses [28–30] were heterogeneous concerning such clinical parameters—some studies mixed stable and unstable CAD patients or included both statin-naive and chronic statin users. Furthermore, actual receipt of PCI widely differed in a substantial number of included trials or analyses were focused on one particular or multiple statin types. Thus, the quality of evidence in this setting was never critically evaluated before. Therefore, findings obtained from previous systematic reviews might not entirely apply to real-world ACS patients.

Cardioprotective effects of early high-dose statin loading in ACS patients with planned percutaneous revascularization are not fully elucidated, but it is suggested that statins exert pleiotropic effects that are beneficial to the cardiovascular system and are beyond the dominant lipid-lowering mechanism [31]. The seminal CANTOS trial validated the inflammatory hypothesis of atherothrombosis since monoclonal antibody blocking inflammatory interleukin-1β pathway managed to reduce rates of recurrent cardiovascular events, compared to placebo, among patients that had a history of MI and increased systemic inflammation as evidenced by high-sensitivity C-reactive protein (CRP) levels ≥ 2 mg/L [32]. Drugs interfering with inflammatory and immune pathways such as colchicine, methotrexate as well as drugs antagonizing IL-6 receptors, were tested for the prevention of adverse cardiovascular events with mixed success in clinical trials [33]. Since statins act anti-inflammatory and reduce CRP levels independent of low-density lipoprotein cholesterol (LDL-C) reduction [34], it is plausible that synergistic anti-inflammatory and lipid-lowering mechanisms of statins administered in high dose during the early phase of ACS may yield beneficial cardiovascular effects, although such mechanisms are yet to be fully explained.

Earlier clinical studies showed that intensive statin use can reduce MACCE in patients with ACS undergoing PCI [35] and prevent procedural myocardial injury in elective PCI among patients with stable CAD [36]. However, questions such as how early to initiate these agents and in which dose during the acute presentation of the coronary event have been less clear [37]. The signal of benefit was more clearly revealed when Patti and colleagues published a collaborative patient-level meta-analysis showing that preprocedural use of statins led to a significant reduction of periprocedural MIs and adverse events at 30 days in a heterogeneous patient cohort receiving PCI [11]. More recent meta-analyses suggested the benefits of atorvastatin [28] and rosuvastatin [29] loading on periprocedural myocardial injury and MACCE among ACS patients undergoing PCI.

Current European and American guidelines for the management and diagnosis of STEMI [2, 4] and NSTE-ACS [1, 3] recommend the initiation of statins as early as possible; however, no specific remarks to high-dose loading before PCI and exact timing of statin initiation are made. To answer these questions, the largest-to-date, double-blind, placebo-controlled, multicenter RCT (SECURE-PCI) that evaluated the use of 80 mg of atorvastatin before and 24 h after coronary angiography among patients with ACS was neutral with respect to the primary outcome of MACCE at 30 days. However, when the data were analyzed concerning those patients that in the end received PCI, it was shown that high-dose atorvastatin loading was associated with a significant 34% reduction in 30-day MACCE (HR 0.66, 95% CI 0.48–0.98), and this effect was sustained in STEMI, but not NSTE-ACS patients [15] which is in line with our findings. Similarly, the observed greater impact of rosuvastatin on MACCE reduction in NSTE-ACS compared to atorvastatin could be explained by the more robust reduction in systemic and microvascular inflammation by rosuvastatin compared to atorvastatin among patients with ACS [38], and this might translate to improved clinical outcomes. Of note, elevated hs-CRP levels were shown to be a prognostic indicator of new MACE and cardiovascular and all-cause mortality in patients with ACS [39]. However, the mechanistic pathways explaining the potential advantage of rosuvastatin in NSTE-ACS have not been fully elucidated yet.

This meta-analysis has some limitations. No protocol was published before this analysis and we only included RCTs in the English language without searching grey literature. While trials were overall at low risk of bias, the majority of studies were of unclear risk concerning selection bias. Similarly, due to the subjective nature of the risk of bias assessment and lack of some information provided by the authors of the studies, uncertainty regarding some evaluations may exist. Moreover, most of the studies had discrepancies concerning a number of analyzed cases vs. those that were randomized to treatment; however, this is in line with the clinical practice since it is expected that not all of the patients randomized to PCI will receive PCI since they might have other indications. Also, it should be acknowledged that most of the evidence of the benefit of statin loading comes from smaller trials whereas the largest trial included (SECURE-PCI) was neutral concerning the benefit of statin loading in ACS. Due to the predominant weight of this trial and the great number of events compared to other trials, it is likely that it significantly impacted on overall results, although our sensitivity analyses in which this trial was removed confirmed our initial findings. Similarly, the body of evidence in terms of patient size was more robust for NSTE-ACS than STEMI population, and this might have an effect on obtained results. For some endpoints, we encountered moderate heterogeneity and discrepancies between effects of larger vs. small studies; however, to account for this, we used a random-effects statistical model that provides more conservative and generalizable results as opposed to a fixed model. Publication bias might also be present due to the detected funnel plot asymmetry, and this observation was mostly driven by trials examining rosuvastatin loading since there appears to be a lack of studies reporting neutral or negative effects with respect to rosuvastatin loading on outcomes of interest. Finally, we should account for a potential geographical bias since most of the included studies were conducted in East Asia.

Conclusions

The results of this meta-analysis, based on moderate and high certainty of the evidence, suggest that early high-dose loading of atorvastatin and rosuvastatin before planned PCI was in association with a significant reduction of MACCE and recurrent MI, in particular, during the short-term period, among statin-naive patients with ACS. This effect seemed to be pronounced among STEMI patients loaded with high-dose atorvastatin and NSTE-ACS patients loaded with high-dose rosuvastatin before planned PCI. However, the results from this meta-analysis should be interpreted with caution due to the potential presence of publication bias and limitations such as small studies contributing to most of the observed effects and the largest trial demonstrating an overall neutral effect of statin loading on clinical outcomes, although more granular analyses of this trial suggested a benefit of early atorvastatin loading in patients receiving PCI and particularly if they had STEMI.

Taken together, these findings suggest a personalized, rather than “one size fits all” approach regarding the early high-potency statin loading in ACS. Therefore, to identify which specific ACS populations would benefit the most from early statin loading strategy, future RCTs should enroll a large number of patients receiving high-dose rosuvastatin and atorvastatin loading in NSTE-ACS, as well as high-dose rosuvastatin loading in the context of STEMI for which the largest evidence gap currently exists.