Abstract
Background
Serotonin–norepinephrine reuptake inhibitors (SNRIs) inhibit the presynaptic reuptake of serotonin and norepinephrine and are widely used in the treatment of depression and neuropathic pain. We evaluate the impact of SNRIs use on the clinical outcomes of patients presenting with Takotsubo cardiomyopathy (TCM).
Methods
We analyzed data from the TriNetX research network. Patients aged ≥18 years with a principal diagnosis of TCM were identified from January 1, 2003, to December 31, 2023. After excluding patients on SSRIs, stimulants, 5‐fluorouracil, capecitabine, gemcitabine, and trastuzumab, patients with myocarditis, pheochromocytoma, acute myocardial infarction, cocaine‐related disorders, and opioid‐related disorders, were also excluded. The final study population was divided into 2 cohorts: patients on SNRIs and patients not on SNRIs. Propensity score matching was used to adjust for potential confounders. The primary outcome was all‐cause mortality, and the secondary outcome was Takotsubo cardiomyopathy‐related complications.
Results
A total of 16 853 patients with a diagnosis of TCM were identified, of which 2482 (14.7%) were on SNRIs. The SNRIs cohort were older (mean age 64.1 versus 63.4 years, P=0.012). After propensity score matching, patients on SNRI had lower 7‐day mortality (odds ratio [OR], 0.59 [95% CI, 0.41–0.84], P=0.004), cardiac arrest (OR, 0.58 [95% CI, 0.40–0.84], P=0.004), and ventricular fibrillation (OR, 0.496 [95% CI, 0.248–0.995], P=0.044). At 1 year, patients on SNRI had lower mortality (OR, 0.80 [95% CI, 0.67–0.96], P=0.015).
Conclusions
The use of SNRIs among patients presenting with TCM is associated with lower mortality and lower TCM‐related complications.
Keywords: antidepressants, cardiomyopathy, depression, mortality, serotonin–norepinephrine reuptake inhibitors (SNRIs), Takotsubo cardiomyopathy
Subject Categories: Heart Failure, Cardiomyopathy, Atrial Fibrillation, Arrhythmias, Cardiovascular Disease
Nonstandard Abbreviations and Acronyms
- PSM
propensity score matching
- SNRI
serotonin‐norepinephrine reuptake inhibitors
- TCM
Takotsubo cardiomyopathy
Clinical Perspective.
What Is New?
Our study suggests that the use of serotonin–norepinephrine reuptake inhibitors (SNRIs) is associated with lower all‐cause mortality and Takotsubo cardiomyopathy‐related complications among patients presenting with Takotsubo cardiomyopathy.
What Are the Clinical Implications?
These results warrant a prospective study to explore the underlying mechanisms of SNRIs; this may include the antiarrhythmic and anti‐inflammatory properties of SNRIs and their interactions with the autonomic nervous system in patients with and without depression and other current indications of SNRIs.
Additional research may explore the potential use of SNRIs in preventing the reoccurrence of Takotsubo cardiomyopathy.
Takotsubo cardiomyopathy (TCM) is a clinical entity characterized by acute transient multisegmental dysfunction of the left ventricle without the presence of obstructive coronary artery disease. 1 First described in 1990, TCM has become increasingly recognized and is currently estimated to affect 2% to 3% of all patients hospitalized with suspected acute coronary syndrome in 2022. 2 Despite often being self‐limited, TCM can be associated with adverse outcomes, including cardiogenic shock, life‐threatening arrhythmias, stroke, and cardiac arrest, and may have comparable mortality to acute coronary syndrome. 3 , 4 Although the exact mechanism of TCM has yet to be completely understood, most theories suggest the role of excess catecholamines and sympathetic nervous system activation due to acute stress. 5 Available data suggest increased prevalence and adverse cardiovascular outcomes among patients with mood disorders when presenting with acute coronary syndrome. 6 A study comparing the clinical features and outcomes of TCM using The International Takotsubo Registry revealed higher rates of psychiatric and neurologic disorders among patients with TCM compared with acute coronary syndrome. 7 Patients with underlying anxiety and depression may be more susceptible to TCM due to overactivity of the sympathetic nervous system. 8
Several case reports have implicated the use of serotonin–norepinephrine reuptake inhibitors (SNRIs) in the cause of TCM. 1 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 Specifically, the use of drugs such as venlafaxine, desvenlafaxine, and duloxetine has been implicated. 9 , 10 , 11 , 12 , 13 , 14 In theory, the presynaptic inhibition of neuronal uptake of catecholamines like norepinephrine may potentiate their impact on the activation of the sympathetic and adrenal systems. A large meta‐analysis of 42 placebo‐controlled trials evaluating the cardiovascular safety of duloxetine in 2005 showed statistically significantly higher rates of tachycardia, palpitation without ECG changes, increased blood pressure, and orthostatic hypotension in patients treated with duloxetine. These changes, however, were minimal and without clinical significance. 18 Additionally, at the time of the study, TCM was not widely recognized. 1 , 18 Our study sought to understand the impact of the use of SNRI on clinical outcomes among patients presenting with TCM.
METHODS
Data Accessibility Statement
Access to the TriNetX research network database can be granted via a request to TriNetX (https://live.trinetx.com). However, access is granted to only deidentified patient information, requires a data‐sharing agreement, and may incur some costs.
Study Design and Participants
This retrospective observational study included data from TriNetX, a global federated health research network with access to electronic medical records (EMRs) from participating health care organizations (HCOs). The HCOs contributing de‐identified EMR data. The global collaborative network analysis was done using the EMR of >101 million patients from 113 HCOs across 21 countries with most of the data coming from the United States as of February 2, 2024. 19 TCM was identified using the International Classification of Diseases, Ninth and Tenth Revisions, Clinical Modification (ICD‐9‐CM and ICD‐10‐CM) code of I51.81 in patients' EMRs. TriNetX used an ICD‐9‐to‐10‐CM mapping based on general equivalence mappings plus custom algorithms and curation to transform data from ICD‐9‐CM to ICD‐10‐CM for HCOs that provided data in ICD‐9‐CM. We excluded patients with pheochromocytoma, myocarditis, acute myocardial infarction, cocaine‐related disorders, or opioid‐related disorders. Patients on chemotherapeutic agents were excluded, including 5‐fluorouracil, capecitabine, gemcitabine, and trastuzumab. Patients with active use of certain psychostimulants and antidepressants in the preceding 3 months were excluded (see Data S1 and Table S1 for details). The sample was then divided into 2 cohorts: those on selective norepinephrine reuptake inhibitors and the second cohort not on these medications. The selective norepinephrine reuptake inhibitors cohort comprised duloxetine, venlafaxine, and desvenlafaxine. Our primary outcome was short‐ and medium‐term all‐cause mortality. Secondary outcomes were TCM‐related complications.
This retrospective study did not involve human participants and was deemed exempt from review and informed consent by the Tower Health Reading Hospital under the Common Rule. TriNetX federated network does not share patient‐identifiable information for such research studies. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.
Data Collection
The TriNetX network was queried on February 2, 2024, and a de‐identified cohort of adult patients (18–89 years old) diagnosed with TCM from January 1, 2003, to December 31, 2023, was analyzed. For both cohorts, the follow‐up period to evaluate outcomes after the index event was analyzed at varying minimum time frames from 7 days, 30 days, 6 months, and 1 year. At the time of the query, 113 HCOs were queried; 57 had data available for patients who met the inclusion criteria for the SNRI cohort, and 69 had data available for patients who met the criteria for the control cohort.
Statistical Analysis
All statistical analyses were completed on the TriNetX online platform. Both cohorts' demographic and clinical baseline characteristics were summarized using descriptive statistics, reported as counts and percentages. Baseline characteristics were compared using Chi‐square tests for categorical variables and independent sample t tests for continuous variables, which were reported as mean, SD, SE, and 95% CI. Propensity score matching (PSM) was used to control for the baseline differences in the 2 cohorts. Patients on SNRIs and controls (not on SNRIs) were matched 1:1 based on propensity scores using logistic regression for demographic covariates (see Table 1).
Table 1.
Baseline Characteristics of Patients With Takotsubo Cardiomyopathy With and Without Serotonin–Norepinephrine Reuptake Inhibitors
| Before propensity score matching | After propensity score matching | |||||
|---|---|---|---|---|---|---|
| TCM with SNRI | TCM without SNRI | P value | TCM with SNRI | TCM without SNRI | P value | |
| Number, n (%) | 2482 | 14 368 | 2257 | 2257 | ||
| Age at diagnosis, y; mean±SD | 64.1±13 | 63.4 | 0.012 | 64.3±13 | 64.2±13 | 0.761 |
| Sex, n (%) | ||||||
| Male | 307 (12.4) | 3241 (22.6) | <0.001 | 300 (13.3) | 312 (13.8) | 0.602 |
| Female | 2064 (83.2) | 10 524 (73.2) | <0.001 | 1857 (82.3) | 1857 (82.2) | 0.969 |
| Race or ethnicity, n (%) | ||||||
| White | 1891 (76.2) | 9557 (66.5) | <0.001 | 1703 (75.5) | 1717 (76.1) | 0.627 |
| Black or African American | 201 (8.1) | 1173 (8.2) | 0.912 | 187 (8.3) | 186 (8.2) | 0.957 |
| Hispanic or Latino | 112 (4.5) | 721 (5.0) | 0.283 | 99 (4.4) | 102 (4.5) | 0.829 |
| Asian | 33 (1.3) | 461 (3.2) | <0.001 | 32 (1.4) | 38 (1.7) | 0.470 |
| Unknown | 281 (11.3) | 2531 (17.6) | <0.001 | 261 (11.6) | 247 (10.9) | 0.510 |
| SNRIs, n (%) | ||||||
| Duloxetine | 1453 (63.6) | … | … | 1409 | … | … |
| Venlafaxine | 763 (33.4) | … | … | 750 | … | … |
| Desvenlafaxine | 68 (3.0) | … | … | 64 | … | … |
| Comorbidities, n (%) | ||||||
| Hypertension | 1689 (68.0) | 5451 (37.9) | <0.001 | 1494 (66.2) | 1516 (67.2) | 0.487 |
| Diabetes | 795 (32.0) | 2374 (16.5) | <0.001 | 678 (30.0) | 692 (30.7) | 0.650 |
| Dyslipidemia | 1431 (57.7) | 4365 (30.4) | <0.001 | 1258 (55.7) | 1288 (57.1) | 0.368 |
| Obesity and overweight | 684 (27.6) | 1537 (10.7) | <0.001 | 567 (25.1) | 571 (25.3) | 0.891 |
| Chronic kidney disease | 466 (18.8) | 1378 (9.6) | <0.001 | 397 (17.6) | 387 (17.1) | 0.694 |
| Atrial fibrillation and flutter | 443 (17.8) | 1620 (11.3) | <0.001 | 386 (17.1) | 385 (17.1) | 0.968 |
| Heart failure | 861 (34.7) | 2484 (17.3) | <0.001 | 735 (32.6) | 766 (33.9) | 0.327 |
| History of coronary artery disease | 812 (32.7) | 2272 (15.8) | <0.001 | 702 (31.1) | 717 (31.8) | 0.631 |
| Obstructive sleep apnea | 405 (16.3) | 726 (5.1) | <0.001 | 319 (14.1) | 301 (13.3) | 0.462 |
| Chronic obstructive pulmonary disease | 521 (21.0) | 1393 (9.7%) | <0.001 | 437 (19.4) | 462 (20.5) | 0.351 |
| Generalized anxiety disorder | 314 (12.7) | 381 (2.7) | <0.001 | 219 (9.7) | 214 (9.5) | 0.800 |
| Posttraumatic stress disorder | 112 (4.5) | 103 (0.7) | <0.001 | 67 (3.0) | 65 (2.9) | 0.860 |
| Depression | 1124 (45.3) | 1235 (8.6) | <0.001 | 903 (40.0) | 887 (39.3) | 0.626 |
| Panic disorder | 127 (5.1) | 121 (0.8) | <0.001 | 84 (3.7) | 78 (3.5) | 0.631 |
| Migraine | 305 (12.3) | 451 (3.1) | <0.001 | 229 (10.1) | 228 (10.1) | 0.961 |
| Polyneuropathy | 430 (17.3) | 561 (3.9) | <0.001 | 329 (14.6) | 318 (14.1) | 0.640 |
| Anemia | 809 (32.6) | 2044 (14.2) | <0.001 | 680 (30.1) | 678 (30.0) | 0.948 |
| Alcohol use disorder | 265 (10.7) | 607 (4.2) | <0.001 | 205 (9.1) | 211 (9.3) | 0.758 |
| Nicotine dependence | 307 (12.4) | 736 (5.1) | <0.001 | 256 (11.3) | 266 (11.8) | 0.642 |
| Thyroid disease | 797 (32.1) | 2068 (14.4) | <0.001 | 673 (29.8) | 683 (30.3) | 0.745 |
| Acute stress reaction | 45 (1.8) | 50 (0.3) | <0.001 | 32 (1.4) | 24 (1.1) | 0.282 |
| Medications, n (%) | ||||||
| Beta blocking agents | 1699 (68.5) | 5094 (35.5) | <0.001 | 1491 (66.1) | 1519 (67.3) | 0.377 |
| Angiotensin‐converting enzyme inhibitors | 1063 (42.8) | 2993 (20.8) | <0.001 | 919 (40.7) | 941 (41.7) | 0.506 |
| Angiotensin receptor blocker | 657 (26.5) | 1729 (12.0) | <0.001 | 563 (24.9) | 559 (24.8) | 0.890 |
| Statins | 1382 (55.7) | 3985 (27.7) | <0.001 | 1214 (53.8) | 1227 (54.4) | 0.698 |
| Aspirin | 1389 (56.0) | 3971 (27.6) | <0.001 | 1207 (53.5) | 1238 (54.9) | 0.354 |
| Clopidogrel | 389 (15.7) | 956 (6.7) | <0.001 | 331 (14.7) | 335 (14.8) | 0.867 |
| Laboratory results, mean±SD | ||||||
| Troponin I (ng/mL) | 0.8±3.8 | 1.4±20.5 | 0.491 | 0.8±4.0 | 2.6±36.5 | 0.256 |
| N‐terminal pro‐B‐type natriuretic peptide, pg/mL | 5186.5±10 862.7 | 4830.1±10 174.2 | 0.552 | 5198.2±10 549.0 | 5200.0±11 098.4 | 0.998 |
| Hemoglobin, g/dL | 12.0±2.2 | 12.1±2.3 | 0.189 | 12.0±2.2 | 12.0±2.2 | 0.813 |
| Echocardiography, mean LVEF | ||||||
| Mean LVEF±SD | 52.1±14.7 | 49.9±15.5 | 0.009 | 52.3±14.6 | 49.9±17.2 | 0.040 |
| LVEF ≤40%, n (%) | 151 (6.1) | 451 (3.1) | <0.001 | 125 (5.5) | 127 (5.6) | 0.897 |
| LVEF 41%–49%, n (%) | 76 (3.1) | 237 (1.6) | <0.001 | 65 (2.9) | 59 (2.6) | 0.585 |
| LVEF ≥50%, n (%) | 355 (14.3) | 977 (6.8) | <0.001 | 296 (13.1) | 280 (12.4) | 0.475 |
LVEF indicates left ventricular ejection fraction; SNRI, serotonin–norepinephrine reuptake inhibitor; and TCM, Takotsubo cardiomyopathy.
The TriNetX platform uses “greedy nearest neighbor matching” with a caliper of 0.1 pooled SDs. 20 Following PSM, logistic regressions produced odds ratios (ORs) with 95% CIs for all‐cause mortality and TCM‐related complications. Statistical significance was set at P<0.05.
For each outcome, we excluded patients with a prior history of the end point of interest (i.e., before the index TCM episode) from our primary and secondary analyses, consequently assessing only for newly diagnosed conditions.
RESULTS
Patient Characteristics
Before PSM, our global sample consisted of 16853 patients from 69 health care organizations with a diagnosis of TCM, of which 2482 (14.7%) were on SNRIs at the time of diagnosis. The SNRI cohort was slightly older (age 64.1±13.0 versus 63.4±14.1 years; SD 0.056), more likely to identify as White (76.2% versus 66.5%; SD 0.215), more likely to be female (83.2% versus 73.2%) and more likely to have comorbidities including cardiovascular, psychiatric, pulmonary, endocrine, and neurologic conditions. Both cohorts had more individuals who identified as White and female. Additionally, 1253 (50.1%) patients were on duloxetine, 685 (27.6%) patients were on venlafaxine, and 72 (2.9%) patients were on desvenlafaxine at least 12 months before the index event (TCM). After 1:1 PSM for 42 characteristics, a total of 2257 patients with TCM on SNRIs were compared with 2257 patients with TCM not on SNRIs; the SD for all matched cohorts was <0.1. (Refer to Table 1 and Table S12 for details.)
Seven‐Day All‐Cause Mortality and Cardiovascular Outcomes
After PSM, all‐cause mortality was 2.2% versus 3.7%, P=0.003 in the patients on SNRI versus patients not on SNRI cohorts, respectively. Logistic regression showed 41% lower odds of all‐cause mortality (OR, 0.59 [95% CI, 0.41–0.84], P=0.004) in the SNRI cohort compared with controls. The SNRI cohort was also found to have statistically significant lower odds of cardiac arrest (OR, 0.58 [95% CI, 0.40–0.84], P<0.004), ventricular tachycardia (OR, 0.56 [95% CI, 0.38–0.84], P=0.005), ventricular fibrillation (OR, 0.496 [95% CI, 0.248–0.995], P=0.044), atrial fibrillation and flutter (OR, 0.74 [95% CI, 0.57–0.97], P=0.028), ischemic stroke (OR, 0.67 [95% CI, 0.46–0.98], P<0.037), respiratory failure (OR, 0.72 [95% CI, 0.596–0.873], P<0.001) and acute kidney injury (OR, 0.79 [95% CI, 0.624–0.995], P=0.044). See Table 2 for details.
Table 2.
Seven‐Day In‐Hospital Outcomes of Patients With Takotsubo Cardiomyopathy With and Without SNRIs (Using Patients Without SNRI as Reference)
| Major adverse events | Odds ratio | CI (95%) | P value |
|---|---|---|---|
| All‐cause mortality | 0.591 | 0.414–0.844 | 0.004 |
| Cardiogenic shock | 0.818 | 0.603–1.109 | 0.195 |
| Cardiac arrest | 0.575* | 0.403–0.842* | 0.004 |
| Ventricular tachycardia | 0.564* | 0.377–0.843 | 0.005* |
| Ventricular fibrillation | 0.496 | 0.248–0.995* | 0.044 |
| Cardiopulmonary resuscitation | 0.947 | 0.512–1.752 | 0.862 |
| Acute kidney injury | 0.788 | 0.624–0.995 | 0.044* |
| Acute ischemic stroke | 0.672 | 0.461–0.978 | 0.037* |
| Acute respiratory failure | 0.721 | 0.596–0.873 | 0.001* |
| Acute pulmonary edema | 0.899 | 0.665–1.216 | 0.491 |
| Atrial fibrillation and flutter | 0.744* | 0.571–0.969* | 0.028 |
| Acute mitral regurgitation | 0.764 | 0.557–1.047 | 0.093 |
| Acute heart failure | 0.855 | 0.716–1.021 | 0.083 |
| Intracardiac thrombus | 0.726 | 0.380–1.385 | 0.329 |
SNRI indicates serotonin–norepinephrine reuptake inhibitor.
Statistically significant findings, p<0.05.
Thirty‐Day All‐Cause Mortality and Cardiovascular Outcomes
After PSM, all‐cause mortality was 5.0% versus 7.1%, P=0.003 in the patients on SNRI versus patients not on SNRI cohorts, respectively. Logistic regression showed 31% lower odds of all‐cause mortality (OR, 0.69 [95% CI, 0.53–0.88], P=0.013) in the SNRI cohort compared with controls. The SNRI cohort was also found to have statistically significant lower odds of cardiac arrest (OR, 0.60 [95% CI, 0.41–0.86], P=0.005), ventricular tachycardia (OR, 0.55 [95% CI, 0.38–0.82], P<0.001), atrial fibrillation and flutter (OR, 0.74 [95% CI, 0.57–0.97], P=0.022), and respiratory failure (OR, 0.72 [95% CI, 0.596–0.873], P=0.002). See Table 3 for details.
Table 3.
Thirty‐Day In‐Hospital Outcomes of Patients With Takotsubo Cardiomyopathy With and Without SNRIs (Using Patients Without SNRI as Reference)
| Major adverse events | Odds ratio | CI (95%) | P value |
|---|---|---|---|
| All‐cause mortality | 0.685* | 0.533–0.879* | 0.003* |
| Cardiogenic shock | 0.844 | 0.627–1.137 | 0.265 |
| Cardiac arrest | 0.597* | 0.413–0.861* | 0.005* |
| Ventricular tachycardia | 0.552* | 0.373–0.816* | 0.003* |
| Ventricular fibrillation | 0.516 | 0.263–1.012 | 0.050 |
| Cardiopulmonary resuscitation | 0.744 | 0.421–1.313 | 0.306 |
| Acute kidney injury | 0.807 | 0.645–1.009 | 0.059 |
| Acute ischemic stroke | 0.757 | 0.536–1.069 | 0.113 |
| Acute respiratory failure | 0.747* | 0.619–0.901* | 0.002* |
| Acute pulmonary edema | 0.921 | 0.699–1.212 | 0.556 |
| Atrial fibrillation and flutter | 0.741* | 0.572–0.959* | 0.022* |
| Acute mitral regurgitation | 0.787 | 0.586–1.057 | 0.110 |
| Acute heart failure | 0.882 | 0.743–1.047 | 0.149 |
| Intracardiac thrombus | 0.748 | 0.405–1.383 | 0.353 |
SNRI indicates serotonin–norepinephrine reuptake inhibitor.
Statistically significant findings, p<0.05.
Six‐Month All‐Cause Mortality and Cardiovascular Outcomes
Following PSM, all‐cause mortality was 9.7% versus 12.0%, P=0.013 in the patients on SNRI versus patients not on SNRI cohorts, respectively. Logistic regression showed 21% lower odds of all‐cause mortality (OR, 0.79 [95% CI, 0.65–0.95], P=0.013) in the SNRI cohort compared with controls. The SNRI cohort was also found to have statistically significant lower odds of cardiac arrest (OR, 0.63 [95% CI, 0.46–0.87], P=0.005), ventricular tachycardia (OR, 0.47 [95% CI, 0.33–0.67], P<0.001) ventricular fibrillation (OR, 0.50 [95% CI, 0.26–0.945], P=0.029), and atrial fibrillation and flutter (OR, 0.68 [95% CI, 0.54–0.86], P=0.013). See Table 4 for details.
Table 4.
Six‐Month In‐Hospital Outcomes of Patients With Takotsubo Cardiomyopathy With and Without SNRIs (Using Patients Without SNRI as Reference)
| Major adverse events | Odds ratio | CI (95%) | P value |
|---|---|---|---|
| All‐cause mortality | 0.786* | 0.651–0.951* | 0.013* |
| Cardiogenic shock | 0.862 | 0.646–1.152 | 0.316 |
| Cardiac arrest | 0.632* | 0.457–0.874* | 0.005* |
| Ventricular tachycardia | 0.472* | 0.331–0.672* | <0.001* |
| Ventricular fibrillation | 0.496* | 0.260–0.945* | 0.029* |
| Cardiopulmonary resuscitation | 0.724 | 0.450–1.164 | 0.181 |
| Acute kidney injury | 0.822 | 0.667–1.013 | 0.066 |
| Acute ischemic stroke | 0.775 | 0.571–1.052 | 0.101 |
| Acute respiratory failure | 0.758* | 0.633–0.908* | 0.003* |
| Acute pulmonary edema | 0.881 | 0.681–1.139 | 0.334 |
| Atrial fibrillation and flutter | 0.680* | 0.537–0.862* | 0.001* |
| Acute mitral regurgitation | 0.750* | 0.576–0.977* | 0.033* |
| Acute heart failure | 0.888 | 0.753–1.047 | 0.156 |
| Intracardiac thrombus | 0.813 | 0.462–1.432 | 0.474 |
SNRI indicates serotonin–norepinephrine reuptake inhibitor.
Statistically significant findings, p<0.05.
One‐Year All‐Cause Mortality and Cardiovascular Outcomes
After PSM, all‐cause mortality was 11.5% versus 13.9%, P=0.016 in the SNRI and no‐SNRI cohorts, respectively. Logistic regression showed a persistent survival benefit with 20% lower odds of all‐cause mortality (OR, 0.80 [95% CI, 0.67–0.96], P=0.016) in the SNRI cohort compared with controls. The benefits in the SNRI cohort persisted from 6 months to 1 year after the index event with significantly lower odds of cardiac arrest (OR, 0.61 [95% CI, 0.44–0.84], P=0.002), ventricular tachycardia (OR, 0.47 [95% CI, 0.34–0.67], P<0.001), ventricular fibrillation (OR, 0.53 [95% CI, 0.283–0.998], P=0.05), and atrial fibrillation and flutter (OR, 0.73 [95% CI, 0.58–0.91], P=0.045). At the 12‐month follow‐up after the index event, 1074 (43.3%) patients were on duloxetine, 587 (23.7%) patients were on venlafaxine, and 30 (1.2%) patients were on desvenlafaxine. See Table 5 for details.
Table 5.
One‐Year In‐Hospital Outcomes of Patients With Takotsubo Cardiomyopathy With and Without SNRIs (Using Patients Without SNRI as Reference)
| Major adverse events | Odds ratio | CI (95%) | P value |
|---|---|---|---|
| All‐cause mortality | 0.804* | 0.674–0.960* | 0.016* |
| Cardiogenic shock | 0.847 | 0.636–1.128 | 0.255 |
| Cardiac arrest | 0.605* | 0.439–0.835* | 0.002* |
| Ventricular tachycardia | 0.472* | 0.335–0.665* | <0.001* |
| Ventricular fibrillation | 0.532* | 0.283–0.998* | 0.046* |
| Cardiopulmonary resuscitation | 0.754 | 0.474–1.199 | 0.232 |
| Acute kidney injury | 0.872 | 0.713–1.068 | 0.185 |
| Acute ischemic stroke | 0.783 | 0.585–1.049 | 0.100 |
| Acute respiratory failure | 0.77* | 0.645–0.920* | 0.004* |
| Acute pulmonary edema | 0.885 | 0.689–1.137 | 0.341 |
| Atrial fibrillation and flutter | 0.727* | 0.579–0.913* | 0.006* |
| Acute mitral regurgitation | 0.732* | 0.567–0.945* | 0.016* |
| Acute heart failure | 0.905 | 0.770–1.064 | 0.228 |
| Intracardiac thrombus | 0.826 | 0.479–1.423 | 0.491 |
SNRI indicates serotonin–norepinephrine reuptake inhibitor.
Statistically significant findings, p<0.05.
Subgroup Analyses
Subgroup logistic regression analysis compared 7‐day outcomes between patients on duloxetine and venlafaxine. The duloxetine cohort contained 1453 patients. Following PSM, 1416 were matched to controls, and duloxetine was associated with lower odds of all‐cause mortality (OR, 0.57 [95% CI, 0.37–0.88], P=0.010), see Figure 1 for details. The venlafaxine cohort contained 763 patients. Following PSM, 756 were matched to controls, and venlafaxine was associated with lower odds of all‐cause mortality (OR, 0.50 [95% CI, 0.28–0.89], P=0.017), see Figure 2 for details. A more detailed sex‐stratified analysis revealed that female patients receiving SNRIs experienced significantly lower 7‐day and 30‐day mortality rates (OR, 0.532 [95% CI, 0.35–0.79]; P=0.002 and OR, 0.651 [95% CI, 0.65–0.88]; P=0.005, respectively). Additionally, these patients had a reduced risk of cardiac arrest, atrial fibrillation, ventricular tachycardia, ischemic stroke, and respiratory failure at multiple time points (P<0.05) compared with those not on SNRIs. In contrast, no significant difference in mortality was observed among male patients receiving SNRIs versus those who were not. See Tables S2 through S11 for details.
Figure 1. Forest plot showing 7‐days clinical outcomes of patients with Takotsubo cardiomyopathy on duloxetine compared with patients not on duloxetine.
OR indicates odds ratio.
Figure 2. Forest plot showing 7‐days clinical outcomes of patients with Takotsubo cardiomyopathy on venlafaxine compared with patients not on venlafaxine.
OR indicates odds ratio.
DISCUSSION
This study highlights the association between the use of SNRIs and clinical outcomes among patients with TCM and its complications. To our knowledge, this is the first study to evaluate these outcomes. We found a statistically significant association between the use of SNRIs and a 41% reduction in 7‐day all‐cause mortality among patients with TCM. The observed beneficial effect persisted over a 1‐year follow‐up period. There was also a substantial reduction in the odds of TCM‐related complications such as cardiac arrest, ventricular tachycardia, ventricular fibrillation, atrial flutter and fibrillation, ischemic stroke, respiratory failure, and acute kidney injury among patients on SNRIs. Additionally, some of these beneficial effects, including atrial and ventricular arrhythmias, persisted at 1‐year follow‐up. Further subanalysis of individual SNRIs, including duloxetine and venlafaxine, showed a significant reduction in 7‐day mortality.
Approximately 80% to 90% of patients who present with TCM are women. 21 Women constituted 82% of our study population after PSM. It is largely unknown why women are TCM disproportionately affects women. Several hypotheses implicating estrogen deprivation after menopause and its role in the regulation of the sympathetic nervous system have been proposed. 2 QTc prolongation is often a concern with the use of this medication class in patients with cardiovascular diseases. However, in a meta‐analysis of 42 placebo‐controlled studies evaluating the cardiovascular safety of duloxetine, no significant changes in QTc were observed using Fridericia's formula. 18 Cardiac autonomic dysfunction associated with depression is characterized by decreased parasympathetic activity, enhanced sympathetic activity, and increased propensity for arrhythmogenesis. 22 In theory, modulating the activity of the cardiac autonomic system with SNRIs may indirectly reduce cardiac arrhythmias.
The exact mechanism underlying TCM is not fully understood. Among many theories, the adrenergic hypothesis, which describes the role of catecholamines, is the most well‐established theory and the focus of our study. 23 In brief, the locus coeruleus is the main adrenergic nucleus that secretes norepinephrine and is responsible for the regulation of arousal and autonomic activity, and receives input from the limbic system. 24 When triggered by emotional stimuli, the locus coeruleus releases norepinephrine, which activates the hypothalamic–pituitary–adrenal axis and the consequent secretion of epinephrine and norepinephrine. 25 Wittstein et al. demonstrated increased levels of plasma catecholamines in the acute phase of TCM. 5 In animal studies, the exogenous administration of epinephrine, norepinephrine, and isoprenaline produced apical left ventricular ballooning, which is reversible. 26 SNRIs, on the other hand, act by blocking the presynaptic neuronal uptake of 5‐hydroxytryptamine (serotonin) and norepinephrine, thereby increasing the synaptic concentration of monoamines and prolonging their effect on postsynaptic receptors. 27 Our study did not account for serum plasma catecholamine; elevation in plasma catecholamines has been demonstrated in patients on SNRIs such as duloxetine. 28 We postulate that the use of SNRIs and a possible consequent increase in serum catecholamines may result in the downregulation of cardiac adrenoreceptors, which may dampen myocardial response to a surge in catecholamines in patients presenting with TCM. This may explain the cardiovascular benefit of using SNRIs as demonstrated in this study.
Our findings can potentially benefit a significant number of patients with TCM. Depression is estimated to affect 20% of patients with cardiovascular disease and represents an independent risk factor for cardiovascular events and mortality. 29 With >37 million prescriptions issued in the United States in 2021, SNRIs are widely used in the treatment of major depressive disorder, generalized anxiety disorder, panic disorder, peripheral neuropathy, chronic pain, and fibromyalgia. 30 Contrary to case reports implicating the use of SNRIs as a provoking factor in TCM, 1 , 9 , 10 , 11 , 12 , 13 , 14 using a large sample size and robust methodology, we have been able to demonstrate improved clinical outcomes among patients using SNRIs when presenting with TCM. This benefit was significant, representing up to a 50% reduction in 7‐day mortality for patients on duloxetine and venlafaxine in subanalysis. In a multicenter analysis of 749 patients from the GEIST (German Italian Stress Cardiomyopathy) registry, the reoccurrence rate of TCM was 4%, mostly occurring in the first 5 years. 3 There are no known clinical or psychological predictors of reoccurrence but it is often precipitated by another stressful event. 2 Although the mechanism by which SNRIs affect patients with TCM is not yet understood, these findings may have implications for preventing the reoccurrence of TCM. Knowledge gained from this study could potentially influence the choice of antidepressants in high‐risk patients. Future work is needed to verify these results.
Such a study may also evaluate the impact of SNRI on patients presenting with TCM across various mental health disorders.
LIMITATIONS
We used the TriNetX research network database, which relies on data collected from the EMRs of organizations with variations in reporting standards of various health conditions. Second, using ICD‐9‐CM and ICD‐10‐CM codes in administrative data sets may contain variations in age, comorbidities, illness severity, length of stay, and laboratory parameters. There is also the potential for coding errors and given the retrospective and observational nature of this study, the data should be interpreted with caution. In addition, results represent the association of health outcomes, and causality cannot be established. Third, although the TriNetX database contains data from multiple health care systems globally, the majority of the data are from US health systems, and hence, the applicability of our results in a global population may be limited. Fourth, we could not account for variations in drug dosing of SNRIs and the interactions of other medications on the activity and serum concentrations of SNRIs. Several factors have been found to influence the pharmacokinetics of SNRIs; they include genetic variation in metabolism, use of fluoroquinolones, sex, age, smoking status, ethnicity, hepatic and renal function, and, notably, the activity of the cytochrome P450 system. Fifth, due to the nature of our database, relevant clinical history, information regarding clinical symptoms, and the trigger event or inciting factor in the development of TCM were not accounted for. To mitigate this, acute stress reaction was used as a surrogate for a recent traumatic or emotional event (Table 1). Sixth, as demonstrated in our subanalysis by sex, there appears to be a gender‐specific cardioprotective effect of SNRIs in favor of female subjects. The applicability of our findings to male subjects is rather limited, given that they constitute only 18% of our study population. Additionally, due to the inherent limitations of the TrinetX database analysis platform, we were unable to use the interaction term P value to assess potential effect modification. Lastly, our study lacks laboratory data on serum levels of catecholamines in patients taking and patients not taking SNRIs. The use of such markers would add to the objectivity of our data.
CONCLUSIONS
We evaluated the association of SNRIs among patients presenting with TCM using a global federated health research network and found that SNRI use is associated with lower 7‐day mortality and lower Takotsubo‐related complications including cardiac arrest, atrial fibrillation, ventricular tachycardia, ventricular fibrillation, acute kidney injury, acute ischemic stroke, and acute respiratory failure. The mortality benefit and associated complications were persistent at 1‐year follow‐up. Although these findings are reassuring, further research is needed to understand the mechanism of action of SNRIs and their impact on the clinical outcomes of patients with TCM.
Sources of Funding
None.
Disclosures
None.
Supporting information
Data S1: Supplemental Methods
Tables S1–S12
This article was sent to Sula Mazimba, MD, MPH, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.124.037951
For Sources of Funding and Disclosures, see page 9.
REFERENCES
- 1. Selke KJ, Dhar G, Cohn JM. Takotsubo cardiomyopathy associated with titration of duloxetine. Tex Heart Inst J. 2011;38:573–576. [PMC free article] [PubMed] [Google Scholar]
- 2. Singh T, Khan H, Gamble DT, Scally C, Newby DE, Dawson D. Takotsubo syndrome: pathophysiology, emerging concepts, and clinical implications. Circulation. 2022;145:1002–1019. doi: 10.1161/CIRCULATIONAHA.121.055854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. El‐Battrawy I, Santoro F, Stiermaier T, et al. Incidence and clinical impact of recurrent Takotsubo syndrome: results from the GEIST Registry. J Am Heart Assoc. 2019;8:e010753. doi: 10.1161/JAHA.118.010753 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Schneider B, Athanasiadis A, Schwab J, Pistner W, Gottwald U, Schoeller R, Toepel W, Winter KD, Stellbrink C, Müller‐Honold T, et al. Complications in the clinical course of tako‐tsubo cardiomyopathy. Int J Cardiol. 2014;176:199–205. doi: 10.1016/j.ijcard.2014.07.002 [DOI] [PubMed] [Google Scholar]
- 5. Wittstein IS, Thiemann DR, Lima JAC, Baughman KL, Schulman SP, Gerstenblith G, Wu KC, Rade JJ, Bivalacqua TJ, Champion HC. Neurohumoral features of myocardial stunning due to sudden emotional stress. N Engl J Med. 2005;352:539–548. doi: 10.1056/NEJMoa043046 [DOI] [PubMed] [Google Scholar]
- 6. Li J, Ji F, Song J, Gao X, Jiang D, Chen G, Chen S, Lin X, Zhuo C. Anxiety and clinical outcomes of patients with acute coronary syndrome: a meta‐analysis. BMJ Open. 2020;10:e034135. doi: 10.1136/bmjopen-2019-034135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Templin C, Ghadri JR, Diekmann J, Napp LC, Bataiosu DR, Jaguszewski M, Cammann VL, Sarcon A, Geyer V, Neumann CA, et al. Clinical features and outcomes of Takotsubo (stress) cardiomyopathy. N Engl J Med. 2015;373:929–938. doi: 10.1056/NEJMoa1406761 [DOI] [PubMed] [Google Scholar]
- 8. Compare A, Bigi R, Orrego PS, Proietti R, Grossi E, Steptoe A. Type D personality is associated with the development of stress cardiomyopathy following emotional triggers. Ann Behav Med. 2013;45:299–307. doi: 10.1007/s12160-013-9474-x [DOI] [PubMed] [Google Scholar]
- 9. Bairashevskaia AV, Belogubova SY, Kondratiuk MR, Rudnova DS, Sologova SS, Tereshkina OI, Avakyan EI. Update of Takotsubo cardiomyopathy: present experience and outlook for the future. IJC Heart Vasc. 2022;39:100990. doi: 10.1016/j.ijcha.2022.100990 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Christoph M, Ebner B, Stolte D, Ibrahim K, Kolschmann S, Strasser RH, Schön S. Broken heart syndrome: Takotsubo cardiomyopathy associated with an overdose of the serotonin–norepinephrine reuptake inhibitor Venlafaxine. Eur Neuropsychopharmacol. 2010;20:594–597. doi: 10.1016/j.euroneuro.2010.03.009 [DOI] [PubMed] [Google Scholar]
- 11. Schroeder I, Zoller M, Angstwurm M, Kur F, Frey L. Venlafaxine intoxication with development of Takotsubo cardiomyopathy: successful use of extracorporeal life support, intravenous lipid emulsion and Cytosorb®. Int J Artif Organs. 2017;40:358–360. doi: 10.5301/ijao.5000595 [DOI] [PubMed] [Google Scholar]
- 12. Rotondi F, Manganelli F, Carbone G, Stanco G. “Tako‐tsubo” cardiomyopathy and duloxetine use. South Med J. 2011;104:345–347. doi: 10.1097/SMJ.0b013e318213f3e5 [DOI] [PubMed] [Google Scholar]
- 13. Gurunathan U. Takotsubo cardiomyopathy and intraoperative cardiac arrest: is desvenlafaxine a contributing factor? J Cardiothorac Vasc Anesth. 2018;32:e16–e18. doi: 10.1053/j.jvca.2017.08.031 [DOI] [PubMed] [Google Scholar]
- 14. Neil CJA, Chong C‐R, Nguyen TH, Horowitz JD. Occurrence of Tako‐Tsubo cardiomyopathy in association with ingestion of serotonin/noradrenaline reuptake inhibitors. Heart Lung Circ. 2012;21:203–205. doi: 10.1016/j.hlc.2011.12.004 [DOI] [PubMed] [Google Scholar]
- 15. Vasudev R, Rampal U, Patel H, Patel K, Bikkina M, Shamoon F. Selective serotonin–norepinephrine reuptake inhibitors‐induced Takotsubo cardiomyopathy. N Am J Med Sci. 2016;8:312–315. doi: 10.4103/1947-2714.187153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bergman BR, Reynolds HR, Skolnick AH, Castillo D. A case of apical ballooning cardiomyopathy associated with duloxetine. Ann Intern Med. 2008;149:218–219. doi: 10.7326/0003-4819-149-3-200808050-00021 [DOI] [PubMed] [Google Scholar]
- 17. Daly MJ, Harbinson MT, Dixon LJ, Spence MS. An unusual case of mid‐ventricular Takotsubo cardiomyopathy. QJM. 2010;103:695–696. doi: 10.1093/qjmed/hcq008 [DOI] [PubMed] [Google Scholar]
- 18. Wernicke J, Lledó A, Raskin J, Kajdasz DK, Wang F. An evaluation of the cardiovascular safety profile of duloxetine: findings from 42 placebo‐controlled studies. Drug Saf. 2007;30:437–455. doi: 10.2165/00002018-200730050-00007 [DOI] [PubMed] [Google Scholar]
- 19. Anon . Real‐world data for the life sciences and healthcare. TriNetX. Accessed May 9, 2024. https://trinetx.com/.
- 20. Anon . Publication guidelines. TriNetX. Accessed February 19, 2024. https://trinetx.com/real‐world‐resources/publications/trinetx‐publication‐guidelines/.
- 21. Anon . Pathophysiology of Takotsubo Syndrome | Circulation . Accessed May 12, 2024. 10.1161/CIRCULATIONAHA.116.027121. [DOI]
- 22. Koschke M, Boettger MK, Schulz S, Berger S, Terhaar J, Voss A, Yeragani VK, Bär KJ. Autonomy of autonomic dysfunction in major depression. Psychosom Med. 2009;71:852–860. doi: 10.1097/PSY.0b013e3181b8bb7a [DOI] [PubMed] [Google Scholar]
- 23. Khalid N, Ahmad SA, Shlofmitz E, Chhabra L. Pathophysiology of Takotsubo syndrome. StatPearls. Treasure Island (FL): StatPearls Publishing; 2024. [PubMed] [Google Scholar]
- 24. Samuels ER, Szabadi E. Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function part I: principles of functional organisation. Curr Neuropharmacol. 2008;6:235–253. doi: 10.2174/157015908785777229 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Jakob MO, Murugan S, Klose CSN. Neuro‐immune circuits regulate immune responses in tissues and organ homeostasis. Front Immunol. 2020;11:308. doi: 10.3389/fimmu.2020.00308 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Heather LC, Catchpole AF, Stuckey DJ, Cole MA, Carr CA, Clarke K. Isoproterenol induces in vivo functional and metabolic abnormalities: similar to those found in the infarcted rat heart. J Physiol Pharmacol. 2009;60:31–39. [PubMed] [Google Scholar]
- 27. Fanelli D, Weller G, Liu H. New serotonin–norepinephrine reuptake inhibitors and their anesthetic and analgesic considerations. Neurol Int. 2021;13:497–509. doi: 10.3390/neurolint13040049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Vincent S, Bieck PR, Garland EM, Loghin C, Bymaster FP, Black BK, Gonzales C, Potter WZ, Robertson D. Clinical assessment of norepinephrine transporter blockade through biochemical and pharmacological profiles. Circulation. 2004;109:3202–3207. doi: 10.1161/01.CIR.0000130847.18666.39 [DOI] [PubMed] [Google Scholar]
- 29. Shi S, Liu T, Liang J, Hu D, Yang B. Depression and risk of sudden cardiac death and arrhythmias: a meta‐analysis. Psychosom Med. 2017;79:153–161. doi: 10.1097/PSY.0000000000000382 [DOI] [PubMed] [Google Scholar]
- 30. Anon . Duloxetine—Drug Usage Statistics, ClinCalc DrugStats Database . Accessed March 19, 2024. https://clincalc.com/DrugStats/Drugs/Duloxetine.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1: Supplemental Methods
Tables S1–S12
Data Availability Statement
Access to the TriNetX research network database can be granted via a request to TriNetX (https://live.trinetx.com). However, access is granted to only deidentified patient information, requires a data‐sharing agreement, and may incur some costs.