Beta-blockers are considered the primary choice in long-term maintenance drug therapy in patients with coronary artery disease, based on positive evidence for improving clinical outcomes in patients with acute myocardial infarction (MI) (1, 2) or heart failure (3). Long-term beta-blocker maintenance is associated with reduced mortality after percutaneous coronary intervention (PCI) in patients with acute MI (4). However, there is a lack of evidence supporting the beneficial impact of long-term beta-blocker treatment in patients with stable coronary artery disease (CAD) (5). Randomized clinical studies, which usually enroll small patient numbers (6), and observational studies (7, 8) have found no significant benefit to beta-blocker treatment in reducing mortality or ischemic events among patients with stable CAD. Furthermore, published data evaluating the clinical benefits of beta-blocker treatment in patients with stable CAD under the specific situation of post-PCI with drug-eluting stents (DES) is very rare. Using a nationwide cohort database, we sought to investigate the clinical impact of long-term beta-blocker maintenance in patients with stable CAD after PCI with DES.

Baseline demographics and medical history of the cohort population before and after stabilized IPTW are presented in Table 1. After weighting, 78,380 DES-treated patients were included: 45,746 with and 32,634 without beta-blocker treatment. After stabilized IPTW, there was no evidence of inequality in baseline demographic characteristics or medical history between the two groups (all standardized mean difference <0.1, Supplementary Figures 1, 2). The incidence and relative hazards of primary and secondary outcomes are presented in Table 2, Figure 2, and Supplementary Figure 3. At 5 years after PCI with 6 months of quarantine, the adjusted incidence rate of MACE was significantly higher in patients treated with beta-blockers (10.0 vs. 9.1% in those without beta-blocker treatment; HR 1.11, 95% CI 1.06–1.16, p < 0.001, Figure 2) in an intention-to-treat analysis (Table 2 and Supplementary Table 5). There was no significant difference in all-cause death between patients treated with and without beta-blocker (8.1 vs. 8.2%; HR 0.99, 95% CI 0.94–1.04, p = 0.62, Supplementary Figure 3), As for the individual components of MACE, there was no significant association between beta-blocker treatment and risk of cardiovascular death (5.9 vs. 5.9% in those without beta-blocker treatment; HR 1.00, 95% CI 0.94–1.06, p = 0.88) or MI (3.8 vs. 3.6% in those without beta-blocker treatment; HR 1.03, 95% CI 0.96–1.11, p = 0.42), while the adjusted hospitalization rate due to heart failure was significantly higher in patients treated with beta-blockers (4.1 vs. 3.1% in those without beta-blocker treatment; HR 1.32, 95% CI 1.23–1.43, p < 0.001) (Table 2). Consistent findings were observed regardless of DAPT duration (Supplementary Table 6). In a subgroup analysis, there was no significant interaction between the baseline comorbidities and beta-blocker treatment for a 5-year occurrence of MACE (Figure 3) or all-cause mortality (Supplementary Figure 4). There was no significant difference in treatment effect according to the generation of beta-blockers (Supplementary Table 7). During a 5-year follow-up period, there were frequent changes in the prescribed beta-blocker status during follow-up (Figure 4). Among 45,326 patients initially treated with beta-blocker before stabilized IPTW weighting, administration of beta-blocker was discontinued in 6,162 (13.6%) without adverse cardiovascular events such as death, MI, or heart failure. Of the 32,813 patients initially treated without beta-blocker, 3,622 (11.0%) started taking beta-blocker during the 5-year follow-up period and showed no MI or heart failure. To take into account frequent cross-over between the treatment groups, we performed additional statistical analyses with time-varying Cox regression (Figure 5A) and RPSFT models (Figure 5B), which demonstrated no statistically significant impact of beta-blocker treatment on the occurrence of all-cause death, cardiovascular death, or MI, whereas beta-blocker treatment was associated with a higher incidence rate of MACE or hospitalization for heart failure. Consistent findings were obtained when patients treated with beta-blocker for more than 1 day during 180 days of quarantine were considered as a treatment group (Figure 5C) or when the prescription status of beta-blocker within a 30-day period, instead of a 180-day quarantine period, after index PCI was applied (Figure 5D).

This nationwide cohort analysis evaluated the association between long-term beta-blocker treatment and clinical outcomes including mortality in patients with stable CAD after DES implantation. Taking advantage of the unique feature of the Korean NHIS database that accurately tracks all medication information over the entire study period, we could analyze the clinical impact of long-term beta-blocker administration in real-world practice by emulating the intention-to-treat manner of a randomized controlled trial. Furthermore, we applied time-varying Cox regression analysis and the RPSFT model to account for switching between treatment strategies, which is typical in real-world practice. The major findings of our study are as follows: (1) in patients with stable CAD after DES implantation, long-term maintenance treatment with beta-blockers was not associated with improvement of clinical outcomes, and (2) sensitivity analyses that considered switching between treatment strategies revealed consistent findings of a negligible impact of beta-blocker on clinical outcomes.

Long-term maintenance beta-blocker treatment after PCI in patients with MI or heart failure is recommended highly based on a large body of evidence that the treatment reduces mortality and morbidity (4, 16). This benefit mainly relies on the heart rate lowering property that decreases oxygen requirements, and negative inotropic effects that mitigate adverse cardiac remodeling and ventricular arrhythmia (17). Furthermore, in a COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial, which demonstrated comparable effects of optimal medical therapy to PCI in stable CAD patients, beta-blockers were a mainstay drug treatment prescribed in 87% of patients enrolled in the trial (18).

There are two published studies that evaluated the association between beta-blocker treatment at discharge and clinical outcomes in stable CAD patients undergoing PCI without prior history of MI or heart failure. One registry study (n = 5,288) reported that beta-blocker treatment at discharge was associated with a significantly increased risk of cardiac death/MI during a 3-year follow-up after index PCI (HR 1.48, 95% CI 1.05–2.10, p =0.02) (19). Another registry study with a larger number of patients (n = 122,734) reported no significant association between beta-blocker treatment at discharge and mortality or MI at 3-year follow-up (8). In addition, discharge with beta-blocker treatment was associated with more frequent readmission due to heart failure (8). However, these two studies did not provide detailed information on the prescribed beta-blocker status during the 3-year follow-up (8, 19). Because continuous beta-blocker prescription status after discharge during long-term follow-up was not evaluated clearly, the impact of drug switch during the follow-up period was not addressed (8, 19).

To minimize potential sources for bias, we excluded patients with concomitant indications for beta-blocker treatment such as MI, heart failure, or atrial fibrillation. Next, we emulated randomized controlled trials using intention-to-treat analysis with a 180-day quarantine period to assign treatment groups and the stabilized IPTW model to adjust for baseline differences. Furthermore, taking advantage of the unique strength of the NHIS database of Korea, which enables tracking and tracing of complete medication information during an entire follow-up period, we compared the main results with those of sensitivity analyses using the time-varying Cox regression and RPSFT model, which consider switches between treatment groups during follow-up. Finally, to compensate for the possible immortal time bias caused by 6 months of quarantine without clinical events, we set a quarantine period of 1 month as a sensitivity analysis. Results from observational studies cannot be used to establish causality, and residual perturbations can persist after propensity score weighting. However, despite the heterogeneity of treatment groups, various sensitivity analyses confirmed consistency compared with the main analysis.

In our analysis, 14% of patients initially treated with beta-blocker after DES implantation eventually discontinued the drug without the occurrence of significant clinical events. In fact, patients who are prescribed beta-blocker can complain of numerous side effects such as fatigue, bradycardia, depression, hypotension, bronchospasm, peripheral vasoconstriction, or postural hypotension, which usually leads to discontinuation of beta-blocker treatment. Furthermore, chronic beta-blocker use has been associated with lipid profile deterioration and new-onset diabetes (20–22). One study reported that beta-blocker treatment increased serum triglyceride level, decreased HDL cholesterol level, and increased plasma small dense LDL, resulting in an atherogenic lipoprotein phenotype (23). A meta-analysis that included 94,492 hypertensive patients treated with beta-blocker has suggested a positive association between beta-blocker treatment and new-onset diabetes (21). Furthermore, non-selective beta-blockers can cause coronary artery spasms by inhibiting β-adrenergic mediated vasodilation (24). Concerns for possible side effects of long-term beta-blocker administration weaken the rationale for routine use of beta-blocker in specific patients with stable CAD after PCI with DES.

This study has several limitations. First, in this nationwide cohort based on claims data, the systolic function of the left ventricle before PCI and during the follow-up period are not included; thus, patients with borderline left ventricular systolic function or reduced left ventricular systolic function without heart failure diagnosis could be included. Furthermore, in the administrative database, heart failure with preserved ejection fraction could have been underdiagnosed (25). Second, considering the nature of retrospective data based on claims, the findings presented in this study cannot be used to establish causal associations, and residual confounding variables could persist even after stabilized IPTW. Third, we adopted the RPSFT model to correct for frequent changes in beta-blocker prescription. Since the RPSFT is a statistical model typically applied to the analysis of randomized controlled trials, caution is needed in interpreting the findings obtained through this analysis despite the 1:1 propensity score matching for the establishment of RPSFT.

Among patients with stable CAD undergoing DES implantation, long-term maintenance with beta-blocker treatment was associated with an increased occurrence of MACE. Beta-blocker treatment may not be recommended as a maintenance drug therapy in specific patients with stable CAD after index PCI.

The datasets generated for the analyses are not publicly available because of strict government restrictions. Requests to access these datasets should be directed to M-KH, mkhong61@yuhs.ac.

The studies involving human participants were reviewed and approved by Yonsei University Health System Institutional Review Board. The Ethics Committee waived the requirement of written informed consent for participation.

S-JL, D-WC, YS, CK, S-JH, C-MA, J-SK, B-KK, Y-GK, DC, E-CP, YJ, C-MN, and M-KH contributed to the conception and design. S-JL and M-KH wrote the study protocol. D-WC and C-MN performed the programming to extract the data from the NHIS database. C-MN and M-KH had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. S-JL, D-WC, C-MN, and M-KH verified the data and conducted all analyses. YS, CK, S-JH, C-MA, J-SK, B-KK, Y-GK, DC, E-CP, and YJ provided a critical review of the manuscript. All authors read and approved the final publication.

This work was supported by the Cardiovascular Research Center, Seoul, Korea.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.