Effect of Paroxetine-Mediated G-Protein Receptor Kinase 2 Inhibition vs Placebo in Patients With Anterior Myocardial Infarction: A Randomized Clinical Trial

Key Points

Question  Does treatment with paroxetine mitigate adverse left ventricular remodeling in patients presenting with acute anterior myocardial infarction?

Findings  In this double-blind, placebo-controlled randomized clinical trial allocating patients after acute anterior myocardial infarction to treatment with paroxetine or placebo, no difference was found in recovery of left ventricular ejection fraction at 12 weeks. In contrast, patients in the experimental arm experienced a greater reduction in late gadolinium enhancement compared with those receiving placebo, indicating attenuation of myocardial fibrosis.

Meaning  In this trial, in patients with acute myocardial infarction, treatment with paroxetine did not improve left ventricular ejection fraction compared with placebo.

Importance  Left ventricular remodeling following acute myocardial infarction results in progressive myocardial dysfunction and adversely affects prognosis.

Objective  To investigate the efficacy of paroxetine-mediated G-protein–coupled receptor kinase 2 inhibition to mitigate adverse left ventricular remodeling in patients presenting with acute myocardial infarction.

Design, Setting, and Participants  This double-blind, placebo-controlled randomized clinical trial was conducted at Bern University Hospital, Bern, Switzerland. Patients with acute anterior ST-segment elevation myocardial infarction with left ventricular ejection fraction (LVEF) of 45% or less were randomly allocated to 2 study arms between October 26, 2017, and September 21, 2020.

Interventions  Patients in the experimental arm received 20 mg of paroxetine daily; patients in the control group received a placebo daily. Both treatments were provided for 12 weeks.

Main Outcomes and Measures  The primary end point was the difference in patient-level improvement of LVEF between baseline and 12 weeks as assessed by cardiac magnetic resonance tomography. Secondary end points were changes in left ventricular dimensions and late gadolinium enhancement between baseline and follow-up.

Results  Fifty patients (mean [SD] age, 62 [13] years; 41 men [82%]) with acute anterior myocardial infarction were randomly allocated to paroxetine or placebo, of whom 38 patients underwent cardiac magnetic resonance imaging both at baseline and 12 weeks. There was no difference in recovery of LVEF between the experimental group (mean [SD] change, 4.0% [7.0%]) and the control group (mean [SD] change, 6.3% [6.3%]; mean difference, −2.4% [95% CI, −6.8% to 2.1%]; P = .29) or changes in left ventricular end-diastolic volume (mean difference, 13.4 [95% CI, −12.3 to 39.0] mL; P = .30) and end-systolic volume (mean difference, 11.4 [95% CI, −3.6 to 26.4] mL; P = .13). Late gadolinium enhancement as a percentage of the total left ventricular mass decreased to a larger extent in the experimental group (mean [SD], −13.6% [12.9%]) compared with the control group (mean [SD], −4.5% [9.5%]; mean difference, −9.1% [95% CI, −16.6% to −1.6%]; P = .02).

Conclusions and Relevance  In this trial, treatment with paroxetine did not improve LVEF after myocardial infarction compared with placebo.

Trial Registration  ClinicalTrials.gov Identifier: NCT03274752

Left ventricular remodeling after acute myocardial infarction is the result of a cascade of intracellular signaling processes modulating adaptive and reparative changes that invokes scar formation, ventricular dilatation, and deterioration of contractile function.¹ Dysregulation of G-protein–coupled receptor kinases (GRK) plays a central role in the progression of left ventricular remodeling after ischemic injury.^2,3 Competitive inhibition of GRK2 has been associated with attenuation of the maladaptive response, reduced myocardial fibrosis, and improved left ventricular function in animal studies, thus offering a potential target for pharmacologic intervention.^4-6

The selective serotonin reuptake inhibitor paroxetine has been shown to selectively inhibit GRK2 as an off-target effect.⁷ In a mouse model,⁸ paroxetine treatment has been demonstrated to mitigate sympathetic overdrive, reverse myocardial remodeling, and improve left ventricular function. The aim of the present study was to prospectively investigate the efficacy of paroxetine-mediated GRK2 inhibition to mitigate adverse left ventricular remodeling in patients presenting with acute myocardial infarction.

The Paroxetine-Mediated GRK2 Inhibition to Reduce Cardiac Remodeling After Acute Myocardial Infarction (CARE-AMI) trial was an investigator-initiated, double-blind, placebo-controlled randomized clinical trial that compared treatment with paroxetine with placebo in addition to guideline-directed medical treatment in patients with acute anterior ST-segment elevation myocardial infarction with a left ventricular ejection fraction (LVEF) of 45% or less. Detailed inclusion and exclusion criteria are listed in the eMethods in Supplement 1. All patients gave written informed consent for participation in the trial before randomization. The study protocol complied with the Declaration of Helsinki and was approved by the cantonal ethics committee of Bern. The research team of Bern University Hospital conducted the study and managed all study data. The Clinical Trials Unit Bern (University of Bern, Switzerland) performed statistical analyses. Additional information can be found in the Trial Protocol in Supplement 2.

Participants were randomly allocated between October 26, 2017, and September 21, 2020, in a 1:1 ratio by use of sequentially numbered boxes to daily treatment with 20 mg of paroxetine or a placebo. Placebo tablets were specifically prepared for this trial and identical to the real drug in color, appearance, smell, and taste as well as packaging and labeling. The allocation schedule was based on computer-generated random numbers. Study participants, treating physicians (T.P., R.V., S.D., S.S., M.F., G.C.M.S., and J.L.), outcome assessors (C.G., A.W.S., S.A.E., and K.F.), and data analysts (F.B.B. and D.H.) were blinded to treatment allocation.

Paroxetine is a selective serotonin reuptake inhibitor approved for the treatment of major depressive disorder. As an off-target effect, paroxetine directly binds to the catalytic domain of GRK2 and increases β-adrenergic receptor–mediated myocardial contractility.⁷

Left ventricular ejection fraction was assessed at baseline and 12 weeks by use of transthoracic echocardiography and cardiac magnetic resonance (CMR) imaging. Methods used for CMR measurements are provided in the eMethods in Supplement 1. A telephone follow-up was performed at 4 weeks, and a clinical visit including measurement of paroxetine levels was conducted at 12 weeks.

The primary end point was the difference in mean patient-level change in LVEF between baseline and follow-up in the experimental and control arms as assessed by CMR imaging. The primary analysis was performed in the intention-to-treat population. Secondary end points were the change in left-ventricular end-diastolic and end-systolic volumes and the change in late gadolinium enhancement between baseline and 12 weeks. A secondary, per-protocol analysis included all participants who underwent baseline and follow-up CMR imaging and had paroxetine levels of 6 nmol/L or more or returned less than 20% of the study medication.

We hypothesized that paroxetine is superior to placebo with respect to the primary end point of LVEF improvement between baseline and 12 weeks. Based on preclinical results, we assumed a change in LVEF of 10% and anticipated a study attrition of 10%.⁸ A sample of 50 individuals was required to show superiority of paroxetine with greater than 90% power and a 2-sided type I error of .05.

Continuous variables were summarized as means (SDs) and categorical variables as frequencies (percentages), as appropriate. The P values have been computed using χ² or Fisher tests for categorical variables and t tests for normally distributed continuous variables. Patient-level changes in values within each group were compared using paired t tests, with equal variances assumed. The comparisons of paroxetine vs placebo arm are reported using mean differences with 95% CIs and P values (2 sided) of unpaired t tests, with equal variances assumed. Data were analyzed with R software version 4.0.3 (R Foundation for Statistical Computing). Additional details are in the Statistical Analysis Plan in Supplement 3.

A total of 50 patients (mean [SD] age, 62 [13] years; 41 men [82%]) presenting with acute anterior myocardial infarction were randomly assigned to the experimental group or control group, of whom 38 patients underwent CMR imaging both at baseline and 12 weeks (eFigure 1 in Supplement 1). Selected baseline characteristics are provided in Table 1; comprehensive characteristics, including information on discharge medications and treatment regimen at 12 weeks, are summarized in eTables 1, 2, and 3 in Supplement 1. Door-to-balloon times and intervals between symptom onset and first balloon inflation were comparable between the groups (eTable 2 in Supplement 1). Intervals between primary percutaneous coronary intervention and CMR imaging at baseline were similar (Table 1). There were no significant differences in baseline characteristics between patients who underwent CMR imaging at follow-up compared with those who did not (eTable 4 in Supplement 1).

Between baseline and follow-up at 12 weeks, mean LVEF as assessed by CMR imaging improved in both the experimental group (mean [SD] change, 4.0% [7.0%]) and the control group (mean [SD] change, 6.3% [6.3%]), with no difference between the 2 treatment arms (mean difference, −2.4% [95% CI, −6.8% to 2.1%]; P = .29; Figure, A). There were no differences in the change in left ventricular end-diastolic diameter (mean difference, −0.2 [95% CI, −4.8 to 4.3] mm; P = .92) or end-systolic diameter (mean difference, 3.6 [95% CI, −1.7 to 8.8] mm; P = .18) between baseline and follow-up (Table 2). Late gadolinium enhancement as a percentage of total left-ventricular mass decreased more in the experimental group (mean [SD], −13.6% [12.9%]) compared with the control group (mean [SD], −4.5% [9.5%]; mean difference, −9.1% [95% CI, −16.6% to −1.6%]; P = .02; Figure, B). Illustrative examples of temporal changes in late gadolinium enhancement observed by CMR imaging are provided in eFigure 2 in Supplement 1.

Echocardiographic data at baseline and follow-up (eTable 5 in Supplement 1) and clinical end points (eTable 6 in Supplement 1) were comparable between the experimental and control groups. None of the study participants died. Reported adverse effects are provided in eTable 7 in Supplement 1.

Thirty-five patients were considered for the per-protocol analysis. The results of the per-protocol analysis were consistent with the intention-to-treat analysis and are summarized in eTable 8 in Supplement 1.

In this double-blind, placebo-controlled randomized clinical trial, the extent of LVEF recovery 12 weeks after acute anterior myocardial infarction was comparable in patients treated with paroxetine or placebo. There were no significant differences in changes of left ventricular dimensions and volumes. In contrast, patients treated with paroxetine had greater reduction in late gadolinium enhancement, indicating attenuation of myocardial fibrosis and scar formation after myocardial infarction.

Preclinical evidence⁸ has suggested a causal role of GRK2 in maladaptive cardiac remodeling and proposed it as a potential target for pharmacologic intervention after myocardial infarction. In an animal model,⁸ wild-type mice started on paroxetine at a dosage of 5 mg/kg/day after induced myocardial infarction exhibited robust improvement of left ventricular function and structure compared with mice treated with a control drug or fluoxetine. In contrast, in the present study, which is to our knowledge the first conducted in humans, we found no difference in CMR imaging–determined LVEF at 3 months after myocardial infarction in patients with paroxetine treatment compared with control participants.

The observation of reduced late gadolinium enhancement, a marker for scar formation, in patients treated with paroxetine is consistent with data from preclinical studies⁸ indicating a reduction of fibrosis in the border zone of the infarct area. The finding is relevant against the background of an established association^9,10 of infarct size as determined by CMR imaging with an increased risk of death at 1 year and warrants further validation in larger studies. In contrast with the mouse model, in which treatment was started 2 weeks after inducing acute myocardial infarction, patients in this study were included 48 to 96 hours after primary percutaneous coronary revascularization and paroxetine was started on top of guideline-directed optimal medical therapy less than 7 days after percutaneous coronary revascularization. Furthermore, study medication may have been underdosed, because GRK2 inhibition mediated by paroxetine is an off-target effect; however, serum levels in the experimental arm were comparable with those observed in studies with positive results conducted in animal models.⁸

Our findings need to be interpreted in light of several limitations. First, only 38 patients underwent CMR imaging at the time of follow-up, and the actual attrition rate (24%) was larger than anticipated. Hence, the study was underpowered, resulting in an increased likelihood of both type I and type II errors. As a consequence, the observed reduction in late gadolinium enhancement in patients treated with paroxetine is hypothesis generating. Second, despite confining the study population to patients with anterior wall myocardial infarction, study participants had on average only a moderately reduced LVEF. This limited the sensitivity of the primary end point and further decreased the chances to demonstrate a difference in LVEF recovery between groups.

In conclusion, in this double-blind, placebo-controlled randomized clinical trial, a 3-month course of paroxetine treatment following acute myocardial infarction did not lead to an improvement in left ventricular ejection fraction compared with placebo. Further studies are needed to investigate the effect of paroxetine-mediated GRK2 inhibition on the development of myocardial fibrosis after myocardial infarction.

Accepted for Publication: May 7, 2021.

Published Online: July 14, 2021. doi:10.1001/jamacardio.2021.2247

Corresponding Author: Thomas Pilgrim, MD, MSc, Department of Cardiology, Inselspital, Bern University Hospital, Freiburgstrasse 18, CH-3010 Bern (thomas.pilgrim@insel.ch).

Correction: This article was corrected on August 18, 2021, to fix errors in Table 2, Limitations, and the visual abstract.

Author Contributions: Drs Babongo Bosombo and Heg had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Pilgrim and Vollenbroich contributed equally to this article.

Concept and design: Pilgrim, Vollenbroich, Hunziker.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Pilgrim, Deckarm, Gräni, Babongo Bosombo, Reusser, Lanz.

Critical revision of the manuscript for important intellectual content: Vollenbroich, Deckarm, Gräni, Dobner, Stark, Erne, Fischer, Stortecky, Fürholz, Siontis, Heg, Hunziker, Windecker, Lanz.

Statistical analysis: Deckarm, Gräni, Babongo Bosombo, Heg.

Obtained funding: Vollenbroich, Lanz.

Administrative, technical, or material support: Pilgrim, Vollenbroich, Dobner, Stortecky, Reusser, Fürholz, Siontis, Hunziker, Windecker.

Supervision: Pilgrim, Gräni, Heg, Lanz.

Conflict of Interest Disclosures: Dr Pilgrim reports research grants to his institution from Edwards Lifesciences, Boston Scientific, and Biotronik; personal fees from Biotronik and Boston Scientific as a speaker and HighLife SAS for clinical event committee participation; and proctoring with Medtronic outside of this study. Dr Stortecky reports research grants to the institution from Edwards Lifesciences, Medtronic, Abbott Vascular, and Boston Scientific and personal fees from Boston Scientific, BTG, and Teleflex outside the submitted work. Dr Windecker reports research and educational grants to his institution from Abbott, Amgen, AstraZeneca, BMS, Bayer, Boston Scientific, Biotronik, Cardinal Health, CardioValve, CSL Behring, Daiichi Sankyo, Edwards Lifesciences, Guerbet, InfraRedx, Johnson & Johnson, Medicure, Medtronic, Novartis, Pfizer, Polares, OrPha Suisse, Regeneron, Sanofi Aventis, Terumo, Sinomed, and V-Wave. Dr Windecker serves as an unpaid advisory board member and/or unpaid member of the steering or executive groups of trials funded by Abbott, Abiomed, Amgen, AstraZeneca, Bristol Myers Squibb, Boston Scientific, Biotronik, Cardiovalve, Edwards Lifesciences, MedAlliance, Medtronic, Novartis, Polares, Sinomed, V-Wave, and Xeltis (but has not received personal payments by pharmaceutical companies or device manufacturers); is a member of the steering or executive committee groups of several investigator-initiated trials that receive funding by industry without personal remuneration; is an unpaid member of the Pfizer Research Award selection committee in Switzerland of the Women as One Awards Committee; is a member of the Clinical Study Group of the Deutsches Zentrum für Herz Kreislauf-Forschung; is a member of the advisory board of the Australian Victorian Heart Institute; is the chairperson of the European Society of Cardiology Congress Program Committee; is a former chairperson of the European Society of Cardiology Clinical Practice Guidelines Committee; and is a deputy editor of JACC CV Interventions. Dr Vollenbroich reports grants from Clinical Trials Unit of the University of Bern during the conduct of the study. Dr Gräni reported receiving funding from the Swiss National Science Foundation. Drs Babongo Bosombo and Heg reported affiliations with the Clinical Trials Unit Bern of the University of Bern, which has a staff policy of not accepting honoraria or consultancy fees; Clinical Trials Unit Bern is involved in design, conduct, or analysis of clinical studies funded by not-for-profit and for-profit organizations, and pharmaceutical and medical device companies provide direct funding to some of these studies. An up-to-date list of Clinical Trials Unit Bern’s conflicts of interest is at http://www.ctu.unibe.ch/research/declaration_of_interest/index_eng.html. Dr Lanz reported grants from Bangerter-Rhyner Foundation and grants from University of Bern during the conduct of the study. No other disclosures were reported.

Funding/Support: The study was supported by dedicated grants from the Clinical Trial Unit of the University of Bern and the Bangerter-Rhyner Foundation.

Role of the Funding/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 4.

Additional Contributions: Dominik Günsch, MD, provided assistance in the analysis of cardiac magnetic resonance data. He was not compensated for his contribution.