Prognostic relevance of β1-adrenergic receptor polymorphisms on cardiovascular outcome in patients treated with metoprolol after acute myocardial infarction - an observational study Authors: Lina Dorthea Bruun, Geir Øystein Andersen, Peder Langeland Myhre, Sigrun Halvorsen, Marianne Kristiansen Kringen, Charlotte Holst Hansen, Espen Molden, Erik Øie Published: 2026 ABSTRACT PurposeTo investigate the association between β1-adrenergic receptor (ADRB1) polymorphisms on cardiovascular events in patients treated with metoprolol after acute myocardial infarction (AMI).MethodsPatients hospitalized for AMI (n = 1584) who were treated with metoprolol at discharge were genotyped for the ADRB1 variants Arg389Gly and Ser49Gly. The association between the ADRB1 variants and major adverse cardiovascular events (MACE) and cardiovascular death were assessed by the Cox proportional hazards model, using both a dominant and a codominant genetic model, as well as by haplotype copy number.ResultsAfter three years of follow-up, MACE occurred in 105 and cardiovascular death in 47 patients. Gly389 homozygotes (n = 105, 11 events; hazard ratio [HR] 2.02 [1.01, 4.02], p = 0.046), but not heterozygotes (n = 588, 45 events, HR 1.40 [0.90, 2.13], p = 0.12) were at increased risk of MACE, compared with Arg389 homozygotes (n = 891, 49 events). The risk was not significantly different when all Gly389 carriers were merged (p = 0.052). No primary events occurred in Gly49 homozygotes (n = 28); however, when merged with Gly49 heterozygous carriers (n = 383, 6 events) we found a numerical, but not statistically significant difference in cardiovascular deaths relative to Ser49 homozygotes (n = 1173, 41 events) (p = 0.07). The Arg389Ser49 haplotype was not associated with a primary outcome.ConclusionIn post-AMI patients treated with metoprolol, Gly389 homozygotes, but not heterozygotes were associated with an increased risk of MACE compared with Arg389 homozygotes. Our observations raise the possibility of a protective association of the Gly49 variant on cardiovascular death, but this hypothesis requires validation in larger, future studies.Graphical Abstract Supplementary InformationThe online version contains supplementary material available at 10.1007/s00228-026-04091-x. Introduction The β1-adrenergic receptor (ADRB1) is the predominant adrenergic receptor expressed in the heart, mediating inotropic, chronotropic, and dromotropic responses upon catecholamine activation. As the main target of β1-selective β-blockers, ADRB1 antagonism mitigates the unfavorable effects of chronic adrenergic stimulation [1, 2]. Long-term β-blocker therapy following acute myocardial infarction (AMI) has traditionally been considered standard of care for most patients since the seminal β-blocker trials in the 1980s [3–5]. However, the magnitude and consistency of benefit across all post-AMI patient subgroups have been debated, particularly among patients with a left ventricular ejection fraction (LVEF) ≥ 40% [6, 7]. Recent randomized controlled trials (RCTs) have demonstrated beneficial effects in patients with mildly reduced EF (40–49%) [8, 9], whereas no clear benefit has been observed in those with preserved EF ≥ 50% [10–13]. Nevertheless, since study results reflect effects at the group level, there may still be individual responders within this population. Pharmacogenetic variability may influence β-blocker response, but the clinical impact of single nucleotide polymorphisms (SNPs) in the ADRB1 gene on cardiovascular events is unclear [1, 14]. Two common SNPs which may influence receptor function, are the missense variants rs1801253 (Arg389Gly) and rs1801252 (Ser49Gly) (Fig. 1). In approximately 27% of Caucasians, arginine is replaced by glycine at position 389 (Gly389), and in around 15%, serine is replaced by glycine at position 49 (Gly49) [15, 16]. In vitro, receptor stimulation of the Gly389 variant mediates reduced signaling by cyclic adenosine monophosphate (cAMP) compared with the Arg389 variant, while stimulation of the Gly49 variant increases receptor downregulation compared with Ser49 [17–20]. Consequently, Gly389 and Gly49 are associated with decreased ADRB1 receptor activity, also implying that patients carrying these variants might be less responsive to β-blocker treatment. While it has been suggested that these SNPs could influence individual therapeutic responses to β-blocker treatment or the required dose, no ADRB1 genotype-based clinical recommendations have yet been established [14]. Fig. 1A. Schematic representation of the SNPs Ser48Gly and Arg389Gly in the β1-adrenergic receptor. *Serine is replaced by glycine at position 49; **arginine is replaced by glycine at position 389. B. Functional effects of ADRB1 variants (14-15,38). AC, adenylyl cyclase; cAMP, cyclic adenosine monophosphate; Gs, G-protein stimulatory subunits; GRKs: G protein-coupled receptor kinases; PKA, protein kinase A; NE, norepinephrine. Red bar-headed line indicates inhibitory mechanism, whereas blue arrow indicates stimulatory process A. Schematic representation of the SNPs Ser48Gly and Arg389Gly in the β1-adrenergic receptor. *Serine is replaced by glycine at position 49; **arginine is replaced by glycine at position 389. B. Functional effects of ADRB1 variants (14-15,38). AC, adenylyl cyclase; cAMP, cyclic adenosine monophosphate; Gs, G-protein stimulatory subunits; GRKs: G protein-coupled receptor kinases; PKA, protein kinase A; NE, norepinephrine. Red bar-headed line indicates inhibitory mechanism, whereas blue arrow indicates stimulatory process Previous studies investigating the hemodynamic response to β-blockers in acute coronary syndrome (ACS) patients [21–23] or the cardiovascular risk in individuals with coronary artery disease (CAD) [24–27] have reported conflicting results. Among post-AMI patients, previous studies found no association between ADRB1 variants and survival or major adverse cardiovascular events (MACE) [28–30]; however, these studies included both patients with and without β-blocker treatment, involved multiple β-blockers, and did not consider the level of drug exposure. The prognostic impact of pharmacogenetic variants is however most relevant in treated individuals, and the impact of ADRB1 variants on cardiovascular events has not previously been investigated in a cohort of AMI patients uniformly treated with the β1-selective blocker metoprolol. Metoprolol is the main β-blocker in use for AMI in Norway [31], as well as the only β-blocker with genotype-based recommendations related to drug exposure [14]. Therefore, the aim of this study was to investigate the association of the ADRB1 variants Arg389Gly and Ser49Gly with cardiovascular events in a cohort of patients treated with metoprolol after AMI during a three-year follow-up. Methods Biobanking in Acute Myocardial Infarction (BAMI) was a prospective, observational study that included 2150 patients admitted with AMI to Oslo University Hospital, Ullevål (Oslo, Norway) from 2007 to 2017. All patients gave written informed consent to participate. Patients considered unable to consent, due to an unstable condition, or age younger than 18 years were excluded. Additional criteria for inclusion in the present substudy were metoprolol treatment upon discharge from hospital and the availability of a blood sample for genotyping. Given the limited use of other β-blockers in the cohort, these individuals were excluded. See study flow chart in Figure S1. Baseline characteristics were collected from medical records, including history of previous MI, heart failure (HF), cerebrovascular disease, hypertension, chronic kidney disease (serum creatinine ≥ 150 µmol/L), smoking, peripheral artery disease, dyslipidemia, and diabetes mellitus. LVEF was available for approximately half of the cohort, and categorized as < 40%, 40–50%, or > 50%. HF was defined as a composite of established HF and/or serum N-terminal pro-B-type natriuretic peptide (NT-proBNP) > 500 ng/mL and/or LVEF < 40% at baseline). We also recorded the angiographic findings and revascularization procedures including the type of stent used. Metoprolol treatment status and dosing at discharge were recorded. As metoprolol is extensively metabolized by the cytochrome p450 2D6 enzyme (CYP2D6), where genetics strongly predict metabolizer phenotype and metoprolol dose exposure, we also recorded concomitant treatment with CYP2D6 inhibitors (paroxetine, fluoxetine, fluvoxamine, and bupropion) as these agents can cause phenoconversion. Metoprolol initiation, continuation, and dose were determined by the attending physician according to routine practice. Follow-up started at the time of inclusion and was restricted a priori to 3 years. The 10th International Classification of Diseases system (ICD-10) was used for all outcomes reported. All-cause mortality, date of death, and classification of causes of death (cardiovascular or non-cardiovascular) were obtained from the Norwegian Cause-of-Death Registry. Clinical outcomes and dates of recurrent MI (ICD-10, I21), HF (I50) and stroke (I60-64) were obtained by linkage to the National Patient Registry. The dual primary outcomes were first occurrence of MACE (a composite of cardiovascular death, recurrent MI, HF, and stroke) or cardiovascular death. Secondary outcomes included all-cause mortality, recurrent MI, HF, and stroke. The study was approved by the Regional Committee for Medical and Health Research Ethics (Case number: 107832). Genotyping Blood samples were collected at the time of enrollment, biobanked, and stored at -80∘C until genotyping. Genomic DNA was extracted from whole blood. ADRB1 genotyping Genotyping of ADRB1 were performed using predesigned TaqMan-based real-time PCR assays; rs1801252 (C___8898508_10) and rs1801253 (C___8898494_10) (Thermo Fischer Scientific, Waltham, USA). The PCR reactions were run on QuantStudio12K Flex or QuantStudio Dx Real-Time PCR Instruments (Thermo Fischer Scientific, Waltham, USA). CYP2D6 genotyping CYP2D6 genotyping was performed using TaqMan-based real-time PCR assays implemented for routine pharmacogenetic analyses, and the patients were divided into four genotype-predicted metabolizer groups (UM, ultrarapid metabolizers; NM, normal metabolizers; IM, intermediate metabolizers; and PM, poor metabolizers) as previously described in detail [32]. Genotype subgroups For the analyses, the ADRB1 variants were grouped as major allele homozygotes (Arg389Arg, Ser49Ser), heterozygotes (Arg389Gly, Ser49Gly), and minor allele homozygotes (Gly389Gly, Gly49Gly). Due to the expected low frequency of the minor ADRB1 alleles (Gly389 and Gly49), homozygous and heterozygous carriers of these variants were also merged into groups of “Gly389 carriers” and “Gly49 carriers” for separate analyses. The ADRB1 variants were also combined into the haplotypes Arg389Ser49, Gly389Ser49, and Arg389Gly49, and we assigned patients into carriers of 0, 1, or 2 copies of each haplotype. Double heterozygotes were assigned to the haplotypes Gly389Ser49 and Arg389Gly49, based on previous observations that the minor alleles do not form a haplotype due to linkage disequilibrium [26]. Patients were also subgrouped based on CYP2D6 genotype, i.e. (i) a low exposure group consisting of normal and ultrarapid metabolizers (NM + UM), and (ii) a high exposure group consisting of intermediate and poor metabolizers (IM + PM). Statistical analysis Categorical data are presented as numbers (%) and continuous data as mean ± standard deviation (SD) and range (minimum-maximum). Statistical comparison of baseline characteristics between genotypes was not performed. Hardy-Weinberg equilibrium for each SNP was tested using the χ2 test. Clinical outcomes are visualized in cumulative incidence plots with risk tables. Cox proportional hazard regression models were used to investigate the associations between genotypes and events, presented as population level hazard ratios (HR) with 95% confidence intervals (CIs) and p-values. The HRs are cause-specific as competing events (non-cardiovascular death for primary outcomes and death for secondary outcomes) were censored. For the primary outcomes we also include a Fine-Gray model, including non-cardiovascular death as a competing event. The proportional hazard assumption was assessed by log-log plots and tested using Schoenfeld residuals. To investigate the association between ADRB1 variants and a primary outcome we used two different models: a dominant model comparing minor allele carriers (heterozygotes and homozygotes) versus the reference major allele homozygotes, and a codominant model referencing both heterozygotes and minor allele homozygotes to the major allele homozygotes. Multivariable Cox models included the following baseline covariates: age, sex, treated hypertension, previous AMI, previous cerebrovascular disease, previous HF, chronic kidney disease, and treated diabetes mellitus. Missing data are reported (Table 2), and only complete case individuals are included in adjusted models. As the carriers of the Gly389Gly49 and Arg389Gly haplotype copy numbers were identical to the dominant model, only the associations between the Arg389Ser49 haplotype and primary outcomes are reported. We estimated HRs for carriers of 0 and 1 copies of Arg389Ser49 using carriers of 2 copies as the reference. Although the study only included on-treatment patients, and as such cannot examine β-blocker effectiveness, we performed several exploratory analyses to examine the effectiveness of β-blockers within ADRB1 genotypes. As these analyses included several small subgroups, they were considered underpowered and therefore hypothesis-generating. To increase power, subgroups were only investigated using the dominant model: multiplicative interactions for sex and age (< 70 years vs. ≥70 years), AMI type (ST-elevation myocardial infarction (STEMI) vs. non-ST-elevation myocardial infarction (NSTEMI)), HF, peak troponin T level (< 1000 ng/L vs. ≥1000 ng/L), CYP2D6 metabolizer subgroup (NM + UM vs. IM + PM), and metoprolol dose (≤ 50 mg, > 50 mg) were added to the adjusted model, using homozygous major allele carriers as the reference. Pre-specified stratified subgroup analyses were performed by (i) AMI type (STEMI, NSTEMI), (ii) CYP2D6 metabolizer group (NM + UM, IM + PM), (iii) excluding patients already treated with β-blocker before the AMI, and (iv) excluding patients with HF at baseline. Post-hoc, we also stratified by metoprolol dose (≤ 50 mg, > 50 mg). All statistical analyses were performed using STATA versions 18.0 and 19.5, and the level of significance was pre-specified as p < 0.05. As this was an exploratory study, power calculations were not conducted and adjustments for multiple testing were not performed. Results Baseline characteristics by ADRB1 genotypes are presented in Tables 1 and 2. Among the 1584 individuals included, 1269 (80.1%) were males, and the mean age was 60.9 years. The minor allele frequencies were 25% for Gly389 and 14% for Gly49. The observed frequencies did not deviate from the Hardy-Weinberg equilibrium [33]. Eight hundred and ninety-one (56.2%) were Arg389 homozygotes, 588 (37.1%) were Arg389Gly heterozygotes, and 105 (6.6%) were Gly389 homozygotes, whereas 1173 (74.1%) were Ser49 homozygotes, 383 (24.2%) were Ser49Gly heterozygotes, and 28 (1.8%) were Gly49 homozygotes. The mean daily metoprolol dose (extended-release metoprolol succinate) in the cohort was 51.3 ± 34.0 mg, and this was comparable across all ADBR1 genotypes, as was the proportions belonging to the CYP2D6 predicted metabolizer groups (Tables 1 and 2). The baseline characteristics of the Arg389Ser49 haplotype are presented in Table S1 and the distribution of ADRB1 genotype combinations in the cohort in Table S2. Table 1Baseline characteristics by Arg389Gly genotypeTotalArg389 homozygotesArg389Gly heterozygotesGly389 homozygotesGly389 carriersN (%)1584891 (56.2)588 (37.1)105 (6.6)693 (43.7)Age, mean (SD) (range)60.9 (± 11.8) (24–94)61.1 (± 11.5) (24–94)60.4 (± 12.4) (29–90)61.7 (± 11.7) (33–88)60.6 (± 12.3) (29–90)Male/female, n (%)1269/315 (80.1/19.9)716/173 (80.4/19.6)467/121 (79.4/20.6)86/19 (81.9/18.1)553/140 (79.8/20.2)CYP2D6 NM, n (%)877 (55.4)481 (53.9)328 (55.8)68 (64.8)396 (57.1)CYP2D6 UM, n (%)32 (2.0)13 (1.5)19 (3.2)019 (2.7)CYP2D6 IM, n (%)477 (30.1)280 (31.4)178 (30.2)19 (18.1)197 (28.4)CYP2D6 PM, n (%)198 (12.5)117 (13.1)63 (10.1)18 (17.1)81 (11.7)STEMI/NSTEMI, n (%)1323/261 (83.5/16.5)742/149 (83.3/16.7)489/99 (83.2/16.8)92/13 (87.6/12.4)581/112 (83.8/16.2)Hypertension, n (%)521 (32.9)297 (33.3)195 (33.3)29 (27.6)224 (32.3)Diabetes mellitus, n (%)195 (12.3)103 (11.6)81 (13.8)11 (10.5)92 (13.3)Previous MI, n (%)217 (13.7)123 (13.8)83 (14.1)11 (10.5)94 (13.6)Stroke, n (%)56 (3.5)29 (3.3)23 (3.9)4 (3.8)27 (3.9)Heart failure, n (%)28 (1.8)17 (1.9)10 (1.7)1 (1.0)11 (1.6)Chronic kidney disease (creatinine > 150 µmol/L), n (%)27 (1.7)14 (1.6)12 (2.0)1 (1.0)13 (1.9)Dyslipidemia, n (%)162 (10.7)108 (12.7)50 (8.8)4 (4.4)54 (9.1)Peripheral artery disease, n (%)58 (3.8)32 (3.8)24 (4.2)2 (2.0)26 (3.9)Smoking no/yes/previous, n (%)518/702/357 (32.7/44.4/22.6)300/385/291 (33.7/43.3/22.6)188/263/135 (32.0/44.7/23.0)30/54/21 (28.6/51.4/20.0)218/317/56 (21.5/45.7/22.5)β-blockers at admission, n (%)346 (21.8)189 (21.2)132 (22.5)25 (23.8)157 (22.7)Metoprolol dose mg, mean (SD) Overall (n = 1584)51.3 (± 34.0)52.2 (± 34.5)50.2 (± 33.6)50.1 (± 32.5)50.2 (± 33.4) If STEMI (n = 1292)49.2 (± 31.4)51.1 (± 33.3)47.2 (± 29.5)45.0 (± 23.7)46.8 (± 28.6) If NSTEMI (n = 262)62.1 (± 43.6)57.9 (± 39.6)65.2 (± 46.4)86.5 (± 57.4)67.6 (± 48.0) If heart failure (n = 26)87.5 (± 54.6)88.2 (± 59.4)85.0 (± 51.6)10086.4 (± 49.2) If previous MI (n = 215)67.1 (± 47.3)68.7 (± 48.4)66.0 (± 47.7)56.8 (± 29.8)64.9 (± 45.9) If hypertension (n = 514)59.8 (± 40.1)62.0 (± 40.2)56.9 (± 40.3)58.2 (± 37.2)57.1 (± 39.8) ≤ 50 mg, n (%)1296 (81.8)720 (80.9)489 (83.5)87 (82.9)576 (83.4) > 50 mg, n (%)285 (18.0)170 (19.1)97 (16.6)18 (17.1)115 (16.6)CYP2D6 enzyme inhibitors a, n64202LVEF, mean (SD)49.0 (± 8.6)49.1 (± 8.6)48.9 (± 8.8)47.7 (± 8.0)48.7 (± 8.7)LVEF ≤ 40%, n (%)164 (10.4)93 (10.4)59 (10.0)12 (11.4)71 (10.3)LVEF 40–50%, n (%)459 (28.9)262 (29.4)169 (28.7)28 (26.7)197 (28.4)LVEF > 50%, n (%)231 (14.5)129 (14.5)90 (15.3)12 (11.4)102 (14.7)Revascularization, n (%) No procedure102 (6.8)54 (6.4)42 (7.5)6 (6.1)48 (7.3) PCI/CABG1340/69 (88.7/4.6)758/37 (89.3/4.4)494/27 (84.1/4.6)88/5 (83.8/4.8)582/32 (87.9/4.8)Type of stent, n (%) No54 (3.)31 (4.0)21 (4.1)2 (2.2)23 (3.8) Metal/DES/Metal + DES645/663/15 (46.8/48.2/1.1)368/368/7 (47.6/47.6/0.9)236/249/7 (46.2/48.7/0.9)41/46/3 (44.6/50.0/3.3)277/295/8 (45.9/48.9/1.3)Diseased vessels: 1/2/3/other, n (%)779/477/270/15 (49.2/28.2/17.1/1.0)430/255/156/9 (50.5/27.0/18.3/1.1)297/165/94/6 (50.5/28.1/16.0/1.0)52/27/20/0 (49.5/25.7/19.1/0)349/192/114/6 (52.6/29.0/17.2/0.9)Numbers in table are mean (standard deviation (SD)) or number (%), as specifieda Bupropion, fluoxetine, fluvoxamine, paroxetineAbbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Table 1Baseline characteristics by Arg389Gly genotypeTotalArg389 homozygotesArg389Gly heterozygotesGly389 homozygotesGly389 carriersN (%)1584891 (56.2)588 (37.1)105 (6.6)693 (43.7)Age, mean (SD) (range)60.9 (± 11.8) (24–94)61.1 (± 11.5) (24–94)60.4 (± 12.4) (29–90)61.7 (± 11.7) (33–88)60.6 (± 12.3) (29–90)Male/female, n (%)1269/315 (80.1/19.9)716/173 (80.4/19.6)467/121 (79.4/20.6)86/19 (81.9/18.1)553/140 (79.8/20.2)CYP2D6 NM, n (%)877 (55.4)481 (53.9)328 (55.8)68 (64.8)396 (57.1)CYP2D6 UM, n (%)32 (2.0)13 (1.5)19 (3.2)019 (2.7)CYP2D6 IM, n (%)477 (30.1)280 (31.4)178 (30.2)19 (18.1)197 (28.4)CYP2D6 PM, n (%)198 (12.5)117 (13.1)63 (10.1)18 (17.1)81 (11.7)STEMI/NSTEMI, n (%)1323/261 (83.5/16.5)742/149 (83.3/16.7)489/99 (83.2/16.8)92/13 (87.6/12.4)581/112 (83.8/16.2)Hypertension, n (%)521 (32.9)297 (33.3)195 (33.3)29 (27.6)224 (32.3)Diabetes mellitus, n (%)195 (12.3)103 (11.6)81 (13.8)11 (10.5)92 (13.3)Previous MI, n (%)217 (13.7)123 (13.8)83 (14.1)11 (10.5)94 (13.6)Stroke, n (%)56 (3.5)29 (3.3)23 (3.9)4 (3.8)27 (3.9)Heart failure, n (%)28 (1.8)17 (1.9)10 (1.7)1 (1.0)11 (1.6)Chronic kidney disease (creatinine > 150 µmol/L), n (%)27 (1.7)14 (1.6)12 (2.0)1 (1.0)13 (1.9)Dyslipidemia, n (%)162 (10.7)108 (12.7)50 (8.8)4 (4.4)54 (9.1)Peripheral artery disease, n (%)58 (3.8)32 (3.8)24 (4.2)2 (2.0)26 (3.9)Smoking no/yes/previous, n (%)518/702/357 (32.7/44.4/22.6)300/385/291 (33.7/43.3/22.6)188/263/135 (32.0/44.7/23.0)30/54/21 (28.6/51.4/20.0)218/317/56 (21.5/45.7/22.5)β-blockers at admission, n (%)346 (21.8)189 (21.2)132 (22.5)25 (23.8)157 (22.7)Metoprolol dose mg, mean (SD) Overall (n = 1584)51.3 (± 34.0)52.2 (± 34.5)50.2 (± 33.6)50.1 (± 32.5)50.2 (± 33.4) If STEMI (n = 1292)49.2 (± 31.4)51.1 (± 33.3)47.2 (± 29.5)45.0 (± 23.7)46.8 (± 28.6) If NSTEMI (n = 262)62.1 (± 43.6)57.9 (± 39.6)65.2 (± 46.4)86.5 (± 57.4)67.6 (± 48.0) If heart failure (n = 26)87.5 (± 54.6)88.2 (± 59.4)85.0 (± 51.6)10086.4 (± 49.2) If previous MI (n = 215)67.1 (± 47.3)68.7 (± 48.4)66.0 (± 47.7)56.8 (± 29.8)64.9 (± 45.9) If hypertension (n = 514)59.8 (± 40.1)62.0 (± 40.2)56.9 (± 40.3)58.2 (± 37.2)57.1 (± 39.8) ≤ 50 mg, n (%)1296 (81.8)720 (80.9)489 (83.5)87 (82.9)576 (83.4) > 50 mg, n (%)285 (18.0)170 (19.1)97 (16.6)18 (17.1)115 (16.6)CYP2D6 enzyme inhibitors a, n64202LVEF, mean (SD)49.0 (± 8.6)49.1 (± 8.6)48.9 (± 8.8)47.7 (± 8.0)48.7 (± 8.7)LVEF ≤ 40%, n (%)164 (10.4)93 (10.4)59 (10.0)12 (11.4)71 (10.3)LVEF 40–50%, n (%)459 (28.9)262 (29.4)169 (28.7)28 (26.7)197 (28.4)LVEF > 50%, n (%)231 (14.5)129 (14.5)90 (15.3)12 (11.4)102 (14.7)Revascularization, n (%) No procedure102 (6.8)54 (6.4)42 (7.5)6 (6.1)48 (7.3) PCI/CABG1340/69 (88.7/4.6)758/37 (89.3/4.4)494/27 (84.1/4.6)88/5 (83.8/4.8)582/32 (87.9/4.8)Type of stent, n (%) No54 (3.)31 (4.0)21 (4.1)2 (2.2)23 (3.8) Metal/DES/Metal + DES645/663/15 (46.8/48.2/1.1)368/368/7 (47.6/47.6/0.9)236/249/7 (46.2/48.7/0.9)41/46/3 (44.6/50.0/3.3)277/295/8 (45.9/48.9/1.3)Diseased vessels: 1/2/3/other, n (%)779/477/270/15 (49.2/28.2/17.1/1.0)430/255/156/9 (50.5/27.0/18.3/1.1)297/165/94/6 (50.5/28.1/16.0/1.0)52/27/20/0 (49.5/25.7/19.1/0)349/192/114/6 (52.6/29.0/17.2/0.9)Numbers in table are mean (standard deviation (SD)) or number (%), as specifieda Bupropion, fluoxetine, fluvoxamine, paroxetineAbbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Baseline characteristics by Arg389Gly genotype Numbers in table are mean (standard deviation (SD)) or number (%), as specified a Bupropion, fluoxetine, fluvoxamine, paroxetine Abbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Table 2Baseline characteristics by Ser49Gly genotypeSer49 homozygotesSer49Gly heterozygotesGly49 homozygotesGly49 carriersMissingN (%)1173 (74.1)383 (24.2)28 (1.8)411 (25.9)-Age, mean (SD) (range)61.1 (± 12.1) (24–94)60.6 (± 11.0) (31–90)57.4 (± 9.7) (31–77)60.4 (± 10.9) (31–90)-Male/female, n (%)945/228 (80.6/19.4)301/82 (78.6/21.4)23/5 (82.1/17.9)334/87 (78.8/21.2)-CYP2D6 NM, n (%)651 (55.5)212 (55.4)14 (50.0)226 (54.9)-CYP2D6 UM, n (%)27 (2.3)5 (1.3)05 (1.2)-CYP2D6 IM, n (%)343 (29.2)123 (32.1)11 (39.2)134 (32.6)-CYP2D6 PM, n (%)152 (13.0)43 (11.2)3 (10.7)46 (11.2)-STEMI/NSTEMI, n (%)977/196 (83.3/16.7)322/61 (84.1/15.9)24/4 (85.7/14.3)346/65 (84.2/15.8)-Hypertension, n (%)382 (32.6)133 (34.7)6 (21.4)139 (33.8)68 (4.3)Diabetes mellitus, n (%)144 (12.2)48 (12.5)3 (10.7)53 (12.4)-Previous MI, n (%)164 (14.0)51 (13.3)2 (7.1)53 (12.4)65 (4.1)Stroke, n (%)41 (3.5)15 (3.9)015 (3.7)67 (4.2)Heart failure, n (%)21 (1.8)7 (1.8)07 (1.7)1Chronic kidney disease (creatinine > 150 µmol/L), n (%)20 (1.7)7 (1.8)07 (1.7)-Dyslipidemia, n (%)112 (10.0)46 (12.5)4 (14.8)50 (12.7)67 (4.2)Peripheral artery disease, n (%)44 (3.9)14 (3.8)014 (3.6)67 (4.2)Smoking no/yes/previous, n (%)364/533/270 (31.1/45.5/23.0)145/155/82 (37.9/40.5/21.4)9/14/5 (32.7/44.4/22.6)154/169/87 (37.5/41.1/21.2)7 (0.4)β-blockers at admission, n (%)259 (22.1)84 (21.9)3 (10.7)87 (21.2)14 (0.9)Metoprolol dose mg, mean (SD) Overall (n = 1584)50.2 (± 33.2)54.6 (± 35.8)53.6 (± 41.6)54.6 (± 36.2)3 (0.2) If STEMI (n = 1292)47.6 (± 29.4)53.5 (± 35.6)55.7 (± 42.3)53.6 (± 36.0)- If NSTEMI (n = 262)62.9 (± 45.7)60.6 (± 36.6)40.6 (± 40.0)59.4 (± 36.8)- If heart failure (n = 26)89.3 (± 54.5)82.1 (± 59.0)-82.1 (± 59.0)- If previous MI (n = 215)64.4 (± 45.7)72.3 (± 49.3)150 (± 70.7)75.2 (± 51.6)- If hypertension (n = 514)57.6 (± 38.9)57.6 (± 38.9)81.3 (± 68.8)66.2 (± 42.7)- ≤ 50 mg, n (%)973 (83.2)302 (78.9)21 (75.0)323 (78.6)- > 50 mg, n (%)197 (16.8)81 (21.2)7 (25.0)88 (21.4)-CYP2D6 enzyme inhibitors a, n5011-LVEF, mean (SD)48.7 (± 8.8)49.6 (± 8.2)52.6 8(± 6.3)(± 40.1)730 (46.1)LVEF < 40%, n (%)129 (11.0)34 (8.9)1 (3.5)35 (8.5)-LVEF 40–50%, n (%)333 (28.4)122 (31.9)4 (14.3)126 (30.7)-LVEF > 50%, n (%)13.9)63 (16.5)5 (17.9)68 (16.6)-Revascularization procedure, n (%)73 (4.6) No procedure78 (7.0)20 (5.2)4 (14.3)24 (6.1)- PCI/CABG991/50 (88.6/3.5)328/17 (85.6/4.4)21/2 (75.0/7.1)349/19 (89.0/4.9)-Type of stent, n (%)207 (13.1) No stent35 (3.4)17 (5.0)2 (8.7)19 (5.3)- Metal/DES/Metal + DES486/484/1.3 (47.7/47.5/1.3)147/170/2 (43.8/50.6/0.6)12/9/0 (52.2/39.1/0)159/179/15.1 (44.3/43.5/1.1)-Diseased vessels: 1/2/3/Other, n (%)563/333/210/14 (50.0/26.7/18.8/1.0)199/110/54/1 (52.0/28.7/14.1/0.3)17/4/6/0 (60.7/14.3/21.4/0)216/114/60/1 (55.0/29.1/15.3/0.3)-73 (4.6)Numbers in table are mean (standard deviation (SD)) or number (%), as specified a Bupropion, fluoxetine, fluvoxamine, paroxetineAbbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Table 2Baseline characteristics by Ser49Gly genotypeSer49 homozygotesSer49Gly heterozygotesGly49 homozygotesGly49 carriersMissingN (%)1173 (74.1)383 (24.2)28 (1.8)411 (25.9)-Age, mean (SD) (range)61.1 (± 12.1) (24–94)60.6 (± 11.0) (31–90)57.4 (± 9.7) (31–77)60.4 (± 10.9) (31–90)-Male/female, n (%)945/228 (80.6/19.4)301/82 (78.6/21.4)23/5 (82.1/17.9)334/87 (78.8/21.2)-CYP2D6 NM, n (%)651 (55.5)212 (55.4)14 (50.0)226 (54.9)-CYP2D6 UM, n (%)27 (2.3)5 (1.3)05 (1.2)-CYP2D6 IM, n (%)343 (29.2)123 (32.1)11 (39.2)134 (32.6)-CYP2D6 PM, n (%)152 (13.0)43 (11.2)3 (10.7)46 (11.2)-STEMI/NSTEMI, n (%)977/196 (83.3/16.7)322/61 (84.1/15.9)24/4 (85.7/14.3)346/65 (84.2/15.8)-Hypertension, n (%)382 (32.6)133 (34.7)6 (21.4)139 (33.8)68 (4.3)Diabetes mellitus, n (%)144 (12.2)48 (12.5)3 (10.7)53 (12.4)-Previous MI, n (%)164 (14.0)51 (13.3)2 (7.1)53 (12.4)65 (4.1)Stroke, n (%)41 (3.5)15 (3.9)015 (3.7)67 (4.2)Heart failure, n (%)21 (1.8)7 (1.8)07 (1.7)1Chronic kidney disease (creatinine > 150 µmol/L), n (%)20 (1.7)7 (1.8)07 (1.7)-Dyslipidemia, n (%)112 (10.0)46 (12.5)4 (14.8)50 (12.7)67 (4.2)Peripheral artery disease, n (%)44 (3.9)14 (3.8)014 (3.6)67 (4.2)Smoking no/yes/previous, n (%)364/533/270 (31.1/45.5/23.0)145/155/82 (37.9/40.5/21.4)9/14/5 (32.7/44.4/22.6)154/169/87 (37.5/41.1/21.2)7 (0.4)β-blockers at admission, n (%)259 (22.1)84 (21.9)3 (10.7)87 (21.2)14 (0.9)Metoprolol dose mg, mean (SD) Overall (n = 1584)50.2 (± 33.2)54.6 (± 35.8)53.6 (± 41.6)54.6 (± 36.2)3 (0.2) If STEMI (n = 1292)47.6 (± 29.4)53.5 (± 35.6)55.7 (± 42.3)53.6 (± 36.0)- If NSTEMI (n = 262)62.9 (± 45.7)60.6 (± 36.6)40.6 (± 40.0)59.4 (± 36.8)- If heart failure (n = 26)89.3 (± 54.5)82.1 (± 59.0)-82.1 (± 59.0)- If previous MI (n = 215)64.4 (± 45.7)72.3 (± 49.3)150 (± 70.7)75.2 (± 51.6)- If hypertension (n = 514)57.6 (± 38.9)57.6 (± 38.9)81.3 (± 68.8)66.2 (± 42.7)- ≤ 50 mg, n (%)973 (83.2)302 (78.9)21 (75.0)323 (78.6)- > 50 mg, n (%)197 (16.8)81 (21.2)7 (25.0)88 (21.4)-CYP2D6 enzyme inhibitors a, n5011-LVEF, mean (SD)48.7 (± 8.8)49.6 (± 8.2)52.6 8(± 6.3)(± 40.1)730 (46.1)LVEF < 40%, n (%)129 (11.0)34 (8.9)1 (3.5)35 (8.5)-LVEF 40–50%, n (%)333 (28.4)122 (31.9)4 (14.3)126 (30.7)-LVEF > 50%, n (%)13.9)63 (16.5)5 (17.9)68 (16.6)-Revascularization procedure, n (%)73 (4.6) No procedure78 (7.0)20 (5.2)4 (14.3)24 (6.1)- PCI/CABG991/50 (88.6/3.5)328/17 (85.6/4.4)21/2 (75.0/7.1)349/19 (89.0/4.9)-Type of stent, n (%)207 (13.1) No stent35 (3.4)17 (5.0)2 (8.7)19 (5.3)- Metal/DES/Metal + DES486/484/1.3 (47.7/47.5/1.3)147/170/2 (43.8/50.6/0.6)12/9/0 (52.2/39.1/0)159/179/15.1 (44.3/43.5/1.1)-Diseased vessels: 1/2/3/Other, n (%)563/333/210/14 (50.0/26.7/18.8/1.0)199/110/54/1 (52.0/28.7/14.1/0.3)17/4/6/0 (60.7/14.3/21.4/0)216/114/60/1 (55.0/29.1/15.3/0.3)-73 (4.6)Numbers in table are mean (standard deviation (SD)) or number (%), as specified a Bupropion, fluoxetine, fluvoxamine, paroxetineAbbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Baseline characteristics by Ser49Gly genotype Numbers in table are mean (standard deviation (SD)) or number (%), as specified a Bupropion, fluoxetine, fluvoxamine, paroxetine Abbreviations: CYP2D6 cytochrome p450 enzyme 2D6, NM normal metabolizer, UM ultrarapid metabolizer, IM intermediate metabolizer, PM poor metabolizer, STEMI ST-elevation myocardial infarction, NSTEMI non-ST-elevation myocardial infarction, MI myocardial infarction, LVEF left ventricular ejection fraction, PCI percutaneous coronary intervention, CABG coronary artery bypass grafting, DES drug-eluting stent Primary outcomes During the 3-year follow-up, there was an increased risk of MACE among Gly389 homozygotes (11 events, 10.5%), relative to Arg389 homozygotes (49 events, 5.5%) [HR 2.02 (1.01, 4.02), p = 0.046]. This difference was no longer statistically significant in the merged Gly389 carrier group (p = 0.052). There was no difference in cardiovascular death between Arg389Gly variants (Table 3; Figs. 2 and 3). Fig. 2Cumulative incidence plots for ADRB1 variants and MACE. A. Arg389 homozygotes vs. Gly389 carriers. B. Arg389 homozygotes vs. Arg389Gly and Gly389 homozygotes. C. Ser49 homozygotes vs. Gly49 carriers. D. Ser49 homozygotes vs. Ser49Gly heterozygotes and Gly49 homozygotes Cumulative incidence plots for ADRB1 variants and MACE. A. Arg389 homozygotes vs. Gly389 carriers. B. Arg389 homozygotes vs. Arg389Gly and Gly389 homozygotes. C. Ser49 homozygotes vs. Gly49 carriers. D. Ser49 homozygotes vs. Ser49Gly heterozygotes and Gly49 homozygotes Fig. 3Cumulative incidence plots for ADRB1 variants and MACE. A. Arg389 homozygotes vs. Gly389 carriers. B. Arg389 homozygotes vs. Arg389Gly and Gly389 homozygotes. C. Ser49 homozygotes vs. Gly49 carriers. D. Ser49 homozygotes vs. Ser49Gly heterozygotes and Gly49 homozygotes Cumulative incidence plots for ADRB1 variants and MACE. A. Arg389 homozygotes vs. Gly389 carriers. B. Arg389 homozygotes vs. Arg389Gly and Gly389 homozygotes. C. Ser49 homozygotes vs. Gly49 carriers. D. Ser49 homozygotes vs. Ser49Gly heterozygotes and Gly49 homozygotes Among Gly49 homozygote carriers no events occurred during follow-up, precluding the estimation of a meaningful HR. However, there was a numerical, but not statistically significant difference in cardiovascular death when merged with Gly49 heterozygous carriers (6 events, 1.6%), relative to Ser49 homozygotes (41 events, 3.5%, p = 0.07). The Ser49Gly polymorphism was not associated with MACE (Table 3; Figs. 2 and 3). There was no association between the number of Arg389Gly copies (0, 1 or 2) and MACE or cardiovascular death (Table 3; Fig. 4). Unadjusted HRs for the primary outcomes are reported in Table S3. The estimates from the Fine-Gray competing risk models were consistent with the Cox model (Table S4). Fig. 4Cumulative incidence plots for the number of copies of the Arg389Ser49 haplotype. A. MACE. Arg389Ser49 2 copies vs. 1 and 0 copies. B. Cardiovascular death. Arg389Ser49 2 copies vs. 1 and 0 copies Cumulative incidence plots for the number of copies of the Arg389Ser49 haplotype. A. MACE. Arg389Ser49 2 copies vs. 1 and 0 copies. B. Cardiovascular death. Arg389Ser49 2 copies vs. 1 and 0 copies Table 3Number of events and associations between genotypes and clinical outcomes at 3-year follow-up MACE Cardiovascular death Genotype n (%)n (%)HR* (95% CI)p-valuen (%)HR* (95% CI)p-value Arg389Arg 891 (56.3)49 (5.5)1.0021 (2.4)1.00 Arg389Gly 588 (37.1)45 (7.7)1.40 (0.92,2.13)0.12023 (3.9)1.70 (0.91,3.20)0.097 Gly389Gly 105 (6.6)11 (10.5)2.02 (1.01,4.02)0.0463 (2.9)1.04 (0.24,4.50)0.962 Gly389 carriers 693 (43.8)56 (8.1)1.48 (1.00,2.22)0.05226 (3.8)1.61 (0.87,2.99)0.127 Ser49Ser 1173 (74.1)84 (7.2)1.0041 (3.5)1.00 Ser49Gly 383 (98.2)21 (5.5)0.75 (0.45,1.24)0.266 (1.6)0.44 (0.17, 1.12)0.09 Gly49Gly 28 (1.8)0 (0.0)……0 (0.0)…… Gly49 carriers 411 (26.0)21 (5.1)0.71 (0.43,1.17)0.186 (1.5)0.41 (0.16,1.07)0.07 Arg389Ser49 copies 2593 (37.4)38 (6.1)1.0016 (2.7)1.00 1745 (47.0)46 (6.2)0.99 (0.63,1.55)0.9527 (3.6)1.48 (0.76,2.88)0.24 0246 (15.5)21 (8.5)1.40 (0.80,2.45)0.244 (1.6)0.55 (0.16,1.93)0.35*Adjusted for age, sex, treated hypertension and previous cerebrovascular disease, myocardial infarction, heart failure, chronic kidney disease, and diabetes mellitus, including only complete cases.Abbreviations: MACE major adverse cardiovascular events, HR hazard ratio, CI confidence interval Table 3Number of events and associations between genotypes and clinical outcomes at 3-year follow-up MACE Cardiovascular death Genotype n (%)n (%)HR* (95% CI)p-valuen (%)HR* (95% CI)p-value Arg389Arg 891 (56.3)49 (5.5)1.0021 (2.4)1.00 Arg389Gly 588 (37.1)45 (7.7)1.40 (0.92,2.13)0.12023 (3.9)1.70 (0.91,3.20)0.097 Gly389Gly 105 (6.6)11 (10.5)2.02 (1.01,4.02)0.0463 (2.9)1.04 (0.24,4.50)0.962 Gly389 carriers 693 (43.8)56 (8.1)1.48 (1.00,2.22)0.05226 (3.8)1.61 (0.87,2.99)0.127 Ser49Ser 1173 (74.1)84 (7.2)1.0041 (3.5)1.00 Ser49Gly 383 (98.2)21 (5.5)0.75 (0.45,1.24)0.266 (1.6)0.44 (0.17, 1.12)0.09 Gly49Gly 28 (1.8)0 (0.0)……0 (0.0)…… Gly49 carriers 411 (26.0)21 (5.1)0.71 (0.43,1.17)0.186 (1.5)0.41 (0.16,1.07)0.07 Arg389Ser49 copies 2593 (37.4)38 (6.1)1.0016 (2.7)1.00 1745 (47.0)46 (6.2)0.99 (0.63,1.55)0.9527 (3.6)1.48 (0.76,2.88)0.24 0246 (15.5)21 (8.5)1.40 (0.80,2.45)0.244 (1.6)0.55 (0.16,1.93)0.35*Adjusted for age, sex, treated hypertension and previous cerebrovascular disease, myocardial infarction, heart failure, chronic kidney disease, and diabetes mellitus, including only complete cases.Abbreviations: MACE major adverse cardiovascular events, HR hazard ratio, CI confidence interval Number of events and associations between genotypes and clinical outcomes at 3-year follow-up *Adjusted for age, sex, treated hypertension and previous cerebrovascular disease, myocardial infarction, heart failure, chronic kidney disease, and diabetes mellitus, including only complete cases. Abbreviations: MACE major adverse cardiovascular events, HR hazard ratio, CI confidence interval Secondary outcomes No differences were observed between ADRB1 genotypes or haplotypes on all-cause mortality, recurrent MI, HF, or stroke (Table S5 and S6). Subgroup analyses An interaction between the presence of HF and genotype risk for MACE and CV death was present across ADRB1 genotype groups (all Pint < 0.05). There were no significant interactions between ADRB1 group and the other covariates, but a potentially heterogenous influence of Arg389Gly groups was observed across the peak level of troponin T (cardiovascular death: Pint= 0.052 and MACE: Pint =0.082) (Figs. 5 and 6). Fig. 5Subgroup analyses by Arg389Gly genotype group. A. Major adverse cardiovascular events (MACE). B. Cardiovascular death. Abbreviations: CYP2D6 NM+UM, cytochrome p450 enzyme 2D6 normal + ultrarapid metabolizer group; CYP2D6 IM+PM, cytochrome p450 enzyme 2D6 poor + intermediate metabolizer group, MI, myocardial infarction; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction Subgroup analyses by Arg389Gly genotype group. A. Major adverse cardiovascular events (MACE). B. Cardiovascular death. Abbreviations: CYP2D6 NM+UM, cytochrome p450 enzyme 2D6 normal + ultrarapid metabolizer group; CYP2D6 IM+PM, cytochrome p450 enzyme 2D6 poor + intermediate metabolizer group, MI, myocardial infarction; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction Fig. 6Sugroup analyses by Ser49Gly genotype group. A. Major adverse cardiovascular events (MACE). B. Cardiovascular death. Abbreviations: CYP2D6 NM+UM, cytochrome p450 enzyme 2D6 normal + ultrarapid metabolizer group; CYP2D6 IM+PM, cytochrome p450 enzyme 2D6 poor + intermediate metabolizer group, MI, myocardial infarction; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction Sugroup analyses by Ser49Gly genotype group. A. Major adverse cardiovascular events (MACE). B. Cardiovascular death. Abbreviations: CYP2D6 NM+UM, cytochrome p450 enzyme 2D6 normal + ultrarapid metabolizer group; CYP2D6 IM+PM, cytochrome p450 enzyme 2D6 poor + intermediate metabolizer group, MI, myocardial infarction; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction In the stratified subgroup analyses (Table S7 and S8), STEMI patients carrying the Gly389 variant allele had an increased risk of MACE (p = 0.016) and cardiovascular death (p = 0.038), but there was no association between ADRB1 genotype group and a primary outcome in other subgroups. Discussion In the present study of post-AMI patients treated with metoprolol, homozygous carriers of the minor variant allele Gly389 were associated with an increased risk of MACE, relative to homozygous carriers of the major Arg389 variant. However, when merged with heterozygous Gly389 carriers, this difference was no longer statistically significant. This observation may indicate a potential difference in outcomes between carriers of the Gly389 variant allele compared with Arg389 homozygous carriers among patients discharged on metoprolol treatment. Although not statistically significant, a trend towards less favourable outcome for carriers of the Gly389 allele was observed across several subgroups. This pattern appeared somewhat more evident among those with HF or markers of transmural myocardial injury (STEMI or a higher peak level of troponin T); however, these findings should be interpreted with caution. Given the exploratory nature of the study, these subgroup findings should be considered as hypotheses-generating only. Although no primary events were observed among homozygous Gly49 carriers, we found no significant association between Ser49Gly variants and MACE or cardiovascular death. These analyses were likely limited by insufficient power, due to the low prevalence of minor allele carriers. Nonetheless, we observed a potentially protective influence of the Gly49 variant on cardiovascular death. This trend was not significant but consistent throughout several subgroups. The Arg389Ser49 haplotype was not associated with a primary outcome. Previous studies have not supported a relationship between ADRB1 variants and cardiovascular events or death in post-AMI patients [28–30]. A recent case-control study described an increased risk of ventricular fibrillation (VF) in STEMI patients without β-blocker treatment carrying the Gly389 variant, but not in Gly389 carriers on β-blocker treatment. This study also reported a lower incidence of VF among both treated and untreated Gly49 carriers [26]. In a study investigating hypertensive patients with CAD, the Arg389Ser49 haplotype was associated with an increased mortality risk that was dissipated in patients randomized to treatment with atenolol, but not in those randomized to verapamil treatment [24]. The studies mentioned above differ from the present in several aspects by including both patients with and without β-blocker treatment, the duration of follow-up, and patient characteristics such as ethnicity, the proportion of females, or the proportion of STEMI patients. Also, the SNPs investigated diverged. Nonetheless, we believe our results support these previous findings suggesting that ADRB1 variants may not be of significance in the overall AMI population. The observations from the present study could indicate that the impact of ADRB1 variants on β-blocker response in AMI patients might depend on the severity of myocardial injury. This is in line with recent RCTs suggesting a more pronounced effect on cardiovascular prevention from β-blockers in patients with reduced LVEF after AMI [8–10, 12, 13]. In agreement with previous reports from HF patients, the impact of ADRB1 variants might also depend upon β-blocker dose [34–37]. Although the initial trial supporting metoprolol treatment after AMI investigated a target dose of 200 mg daily [5], data from current clinical practice indicates that the metoprolol doses commonly used are substantially lower [8, 10, 12, 31, 32]. The actual metoprolol exposure is however largely dependent upon CYP2D6 genotype [38]. Using CYP2D6 genotype as a proxy of metoprolol dose, we did not find an influence of high (CYP2D6 IM + PM) vs. low (CYP2D6 NM + UM) exposure on the cardiovascular risk associated with the ADRB1 SNPs in the present study, with similar findings when comparing the dose at discharge (≤ 50 mg vs. > 50 mg). Thus, the influence of ADRB1 SNPs did not appear to be dose-dependent, supporting previous investigations indicating that increasing β-blocker dose (exposure) might not add treatment benefit in AMI patients [12, 32, 39, 40]. As far as we are aware of, this is the first study investigating the impact of both ADRB1 variants and metoprolol exposure on cardiovascular outcomes in post-AMI patients and is not necessarily contradictory to HF studies demonstrating that Arg389 carriers might benefit from a higher dose of metoprolol than Gly389 carriers [34–36]. While AMI patients might recover with only minor myocardial injury following contemporary treatment, HF is a complex systemic syndrome characterized by chronic adrenergic activation which can lead to a selective decrease in myocardial ADRB1 receptor density and a changed β1:β2-adrenergic receptor ratio [1, 41, 42]. Such adaptations might influence the responsiveness to β-blocker treatment as well as the importance of ADRB1 variants depending on indication. This is also supported by in vitro dose-response curves, which suggest a more obvious impact of receptor variants under conditions with high sympathetic activation, greater receptor occupancy, or within a failing myocardium [17, 20, 43]. The observations from the present study are supported by functional effects of the ADRB1 variants Arg389Gly and Ser49Gly seen in vitro [17, 20, 43]. In the presence of an antagonist, a greater decrease in cAMP generation has been demonstrated for the Arg389 variant compared with Gly389 [17, 43], while metoprolol has been demonstrated to decrease cAMP generation more for Gly49 than the Ser49 variant. Also, the Ser49 variant, but not Gly49 is considered resistant to receptor downregulation upon agonist stimulation [20]. Limitations Our cohort represents an unselective AMI population, including patients with a wide range of disease severities, ages, and comorbidities. Based on the number of patients included, any major influence of ADRB1 variants in the overall post-AMI population on metoprolol treatment appears unlikely. However, findings in subgroups are likely constrained by limited sample size and a low number of events and should be interpreted cautiously. Our study cannot inform on β-blocker treatment effectiveness, as we only included patients who were on treatment and excluded those who were discharged without β-blockers. Moreover, as this was an observational study, the presence of unmeasured confounding cannot be excluded. As metoprolol treatment status was determined upon hospital discharge, any changes in treatment continuation or metoprolol dose throughout the observation period could not be taken into consideration. We did not have information on blood pressure, heart rate, or treatment tolerance during follow-up that might have impacted treatment continuation. However, we have previously demonstrated that persistence with β-blocker treatment following AMI in Norway is high, and that the prescribed dose remains largely unchanged after discharge [31]. In addition, the use of CYP2D6 metabolizer subgroup as a proxy for metoprolol exposure is imprecise as there is considerable interindividual variation within each subgroup. Also, drug exposure may be affected by concomitant use of CYP2D6-inhibiting drugs [14]. Even if the use of CYP2D6 inhibitors were negligible at baseline, we could not account for changes in concomitant drug treatment during follow-up. LVEF was not available for more than half of the cohort, limiting further analyses stratified by EF. Another limitation is that we only included ADRB1 variants, and not polymorphisms in other adrenergic receptors or regulators which combined might have been more informative [14]. Given the statistical model used, our findings may have been influenced by the duration of follow-up. Furthermore, the optimal duration of β-blocker treatment after AMI remains unknown, and as pathophysiological responses following an AMI may alter adrenergic signaling over time, the influence of ADRB1 variants may also vary accordingly. Despite these limitations, our study identified potential subgroups that may derive greater benefit from metoprolol treatment and could generate hypotheses for future studies on individualized treatment post-AMI. Conclusion In this study of post-AMI patients treated with metoprolol, homozygous carriers of the ADRB1 variant Gly389 were associated with an increased risk of major adverse cardiovascular events, but not cardiovascular death. Our observations raise the possibility of a protective association of the Gly49 variant on cardiovascular death, but this hypothesis requires validation in larger, future studies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (DOCX 75 KB) Supplementary Material 1 (DOCX 75 KB)