Use of beta-blockers and in-hospital mortality in patients with Takotsubo cardiomyopathy: systematic review and meta-analysis

Abstract

Introduction

Takotsubo cardiomyopathy (TC) is a stress-induced catecholamine acute myocardial dysfunction in the absence of significant coronary disease. The pathophysiology of TC remains poorly understood, and the use of beta-blockers (β-Blockers) appears promising. However, the impact of β-blockers in acute-phase management remains uncertain.

Aims

We aimed to conduct a systematic review and meta-analysis evaluating the early use of β-Blocker in TC and its effects on in-hospital mortality.

Methods

PubMed, Embase and Cochrane were searched for studies that evaluated the use of BBs in TC patients and its short-term effects. Statistical analysis was performed using RevMan V,5.4.1. The results are expressed using HRs and 95% confidence intervals (CI) were extracted using a random-effects model. Heterogeneity was assessed with I².

Results

We included five cohort studies, with a total of 5428 patients. The vast majority were women (81%) with a mean age of 70.1±12.6 years. More than half of the patients (52.0%) had previous hypertension. ST-elevation in the first ECG was observed in 37.3% of the patients. Early administration of β-blockers was not associated with a statistically significant reduction in in-hospital mortality (HR 0.78, 95% CI 0.59 to 1.02, p=0.07), and this finding was consistent across different β-blocker types, doses and routes of administration.

Conclusion

Early β-Blocker therapy did not significantly influence in-hospital mortality in patients with TC. Future randomised studies are essential to clarify beta-blockers’ role in this setting.

PROSPERO registration number

CRD420251055617.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Takotsubo cardiomyopathy (TC) is not a benign condition and its mortality can be as high as in acute coronary syndrome. The most accepted pathophysiological hypothesis suggests a sympathetic hyperstimulation due to an excessive peak of catecholamines, and beta-blockers can be great allies in managing the acute phase.

WHAT THIS STUDY ADDS

• In this systematic review and meta-analysis of five observational studies, the use of beta-blockers in patients with TC did not significantly reduce in-hospital mortality, although the CI suggests a potential trend towards benefit.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• With these results, more randomised controlled trials need to be conducted to adequately elucidate this important question.

Introduction

First described in 1990,¹ Takotsubo cardiomyopathy (TC) is a stress-induced catecholamine acute myocardial dysfunction in the absence of significant coronary disease and is characterised by transient left ventricular (LV) dysfunction,² usually triggered by physical or emotional stress.^{3 4} It is worth noting that happy emotions are also associated with TC.⁵ Because of this transience, the pathophysiology of TC has been defined as benign conduction.⁶ However, there is a risk of recurrence⁷ and serious cardiac complications, such as cardiogenic shock,⁸ with a 5.6% mortality per patient a year.²

The acute in-hospital mortality of TC usually is low (<3%), but is more prevalent in elderly patients aged over 75 years, which can reach up to 6.3%,⁹ mainly because they have a greater number of non-cardiovascular comorbidities.⁵ Despite this, younger patients tend to have more in-hospital complications related to TC, such as cardiogenic shock or intracranial bleeding, but mortality is less.^{5 10} The young group also has distinct characteristics, with a male predominance and physical stress as the TC cause.⁵

The most widely accepted pathophysiological hypothesis for TC suggests that there is an excessive peak of catecholamines and sympathetic hyperstimulation.⁷ The excess of catecholamines could be responsible for apical ballooning and consequent cardiotoxicity, with a drop in the LV ejection fraction, due to microvascular dysfunction.¹¹ Therefore, the use of beta-blockers (β-blockers) in the treatment of TC has been proposed as a therapeutic strategy, reducing intraventricular pressure gradient.¹² To this date, there is no specific treatment for this condition and in the GEIST registry,³ the use of medications to treat heart failure secondary to TC did not alter the LV ejection fraction.

According to the facts above, the impact of the use of β-blockers in-hospital mortality of patients with TC remains incompletely uncertain. To fill this gap, a systematic review and meta-analysis were carried out, seeking to demonstrate the effects on in-hospital mortality of using β-Blockers in the acute phase of TC.

Methods

Search strategy and selected criteria

The study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines¹³ and registered in the International Prospective Register of Systematic Reviews (PROSPERO ID CRD420251055617).

We searched PubMed, Embase and Cochrane databases until March 2025 for studies that evaluated the use of β-blockers in TC patients and their short-term effects. The search initial strategy consisted of a combination of the following keywords: “Takotsubo” OR “Acute stress-induced cardiomyopathy” OR “broken heart syndrome” OR “stress cardiomyopathy” OR “apical ballooning syndrome”) AND (“beta blocker” OR “beta-blocker” OR “carvedilol” OR “bisoprolol” OR “metoprolol” OR “esmolol” OR “adrenergic beta-antagonists”). Two independent authors (COFB and BS) did the selection of the studies and the screening processes were performed in the Rayyan systematic review management platform.

Inclusion and exclusion criteria

Studies were included if they matched the following characteristics: adults (> 18 years) patients with TC, reporting in-hospital use of β-Blockers, with in-hospital data. Those that evaluated long-term patients without medication use, case reports, studies with the paediatric population and experimental animal studies were excluded. There was no language restriction.

Study selection

As shown in figure 1, the initial search found 1474 studies. After careful reading of the data presented in the abstracts and deleting duplicate studies, 11 articles were read in full. After applying the inclusion and exclusion criteria, five cohort studies were selected,14,18 comprising 5428 patients.

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 flow diagram for new systematic reviews that included searches of databases and registers only.

Statistical analysis

Data extraction was independently conducted by two investigators (COFB and BS), who systematically collected the relevant variables and synthesised the outcomes of interest. To ensure accuracy and consistency, all extracted data were subsequently reviewed and validated by a third author (VAG).

Statistical analyses were performed using the Review Manager software (RevMan, V.5.4.1). HRs and corresponding 95% CIs were pooled using a random-effects model in order to account for potential interstudy variability. Between-study heterogeneity was quantitatively assessed using the I² statistic, with values closer to 0% indicating minimal heterogeneity and greater consistency among the included studies.

Cohort studies and registry analyses in this field report effect estimates using different statistical models. Some studies use ORs when outcomes are analysed with logistic regression (typically in-hospital or short-term events), others use relative risks when proportions are compared directly, and long-term follow-up studies often use HRs derived from time-to-event analyses such as Cox proportional hazards models. For consistency across studies, we extracted the reported effect measure as presented by the authors and interpreted them within the context of their study design and statistical method, without converting between measures.

Results

We included five cohort studies, with a total of 5428 patients. The vast majority were women (81%) with a mean age of 70.1±12.6 years. More than half of the patients (52.0%) had previous hypertension. ST-elevation in the first ECG was observed in 37.3% of the patients. In-hospital use of β-Blockers was a therapeutic strategy in 48% of the patients. All the baseline characteristics are demonstrated in table 1.

Table 1. Baseline characteristics of included studies.

Study	Study design	Follow-up (days)	Number of patients	Female sex N (%)	Age, mean±SD	Previous hypertension N (%)	ST-elevation in initial ECG N (%)	In-hospital use of β-B N (%)
Almendro-Delia et al18	Prospective cohort	284	711	616 (87)	70.0±12.0	453 (64)	403 (57)	440 (62)
Gopalakrishnan et al 17	Prospective cohort	2190	56	45 (80.4)	65.8±14.1	32 (57.1)	7 (12.5)	46 (82)
Isogai et al14	Retrospective cohort	30	2672	2177 (81.5)	71.4±11.1	N/A	N/A	422 (15.7)
Kim et al16	Prospective cohort	2117	265	252 (95)	69.9±11.8	186 (70)	172 (65)	235 (89)
Petursson et al15	Retrospective cohort	877	1724	1327 (77)	66.0±14.0	764 (44.3)	446 (25.9)	1485 (86)

.ST, segment ST ; β-B, beta-blockers.

The definition of ‘early β-Blocker use’ varies among published studies. In the largest registry analysis, early administration was explicitly defined as initiation of oral or intravenous therapy within the first 48 hours of hospitalisation.¹⁴ In contrast, a single-centre study reported medications prescribed during hospitalisation without specifying timing.¹⁷ Other registry studies, such as the Mayo Clinic registry,¹⁶ the SWEDEHEART cohort¹⁵ and the RETAKO registry¹⁸, primarily reported β-blocker use at discharge rather than at admission. Accordingly, when interpreting the literature, ‘early administration’ should be understood as treatment begun within 24–48 hours of admission, whereas ‘during hospitalisation’ and ‘at discharge’ refer to progressively later points in the care pathway.

The route of administration of β-blockers varied considerably between studies. In the study by Isogai et al,¹⁴ 89.1% of patients received only oral β-blockers, with bisoprolol being the most commonly used. Only 6.4% of patients received only intravenous formulations, which included propranolol, landilol and esmolol. In the SWEDEHEART registry,¹⁵ 77.8% of patients used β-Blockers orally and 8.3% intravenously.

In the SWEDEHEART trial,¹⁵ there were 3% in-hospital deaths, and half of them were due to cardiac causes (cardiogenic shock and ventricular arrhythmia). The largest cohort in this meta-analysis¹⁴ also revealed 3% in-hospital mortality, but with no significant difference between the group that used β-blockers and the control group without this medication. In the RETAKO trial,¹⁸ in-hospital mortality was higher in patients who developed cardiogenic shock due to TC (13.6%). Even in this scenario with possible cardiovascular complications, the overall in-hospital mortality was 2.4% and mostly in-hospital deaths were attributable to non-cardiovascular conditions, such as post-anoxic encephalopathy and pulmonary sepsis.

Analysing data from these five studies, early administration of β-blockers was not associated with a statistically significant reduction in in-hospital mortality (HR 0.78, 95% CI 0.59 to 1.02, p=0.07), and this finding was consistent across different β-blocker types, doses and routes of administration, as shown by the low heterogeneity of the sample (I² 0%), figure 2. This result means that any variation between studies data is likely due to random sampling error, not any real differences. However, this I² value is only an estimate, and since the CI exceeds the 95% threshold (p=0.57), some substantial degree of real heterogeneity may still be present.

Figure 2. In-hospital mortality of patients with Takotsubo cardiomyopathy using β-blockers (BB).

Risk of Bias

Risk of bias was assessed using the ROBINS-1 tool¹⁹ for non-randomised studies. Two independent authors (COFB and GLS) collected all the domains and the results are presented in figure 3. All the studies had some concerns, due to the fact that they are observational cohorts. Most concerns were about bias due to confounding (D1 domain) and classification of interventions (D3 domain). Two articles had serious bias to confounding^{16 17} and received a serious overall risk of bias. The final analysis evaluates all studies with moderate overall risk of bias.

Figure 3. Risk of bias summary for non-randomised studies.¹⁹.

Discussion

We conducted a systematic review and meta-analysis comprising 5428 patients with TC to evaluate the impact of early use of β-blockers on in-hospital mortality. Our main result showed that administration of β-blockers was not significantly associated with a reduction in in-hospital mortality (figure 4).

Figure 4. Central illustration of the meta-analysis. β-B, beta blocker.

All studies shared similar demographic characteristics, with a predominance of elderly female patients. The InterTAK²⁰ and the GEIST registry⁵ indicated a higher prevalence of TC in women, with a female-to-male ratio of 9:1 and 80% of them older than 50 years,²¹ affecting especially postmenopausal women.²² However, greater attention should be given to male patients, as they are more likely to present with TC in association with cardiogenic shock and higher in-hospital mortality.^{4 18}

The presence of comorbidity varied significantly. The prevalence of hypertension ranged from 44.3%¹⁵ to 70%¹⁶—and may reflect differences in baseline cardiovascular risk. In a recent analysis of the GEIST registry,⁵ the prevalence of hypertension in patients with TC varied according to the age at presentation of symptoms, ranging from 22% in younger patients (<45 years) to 78% in those over 75 years.

ST-segment elevation on initial ECG also had a great variability, ranging from 12.5%¹⁷ to 65%¹⁶, potentially influencing both diagnosis and treatment decisions. Notably, studies with higher rates of ST-elevation^{16 18} also reported greater use of β-blockers, which may reflect a more aggressive in-hospital management strategy based on perceived myocardial injury. In the study of Citro et al⁹, elderly patients (>65 years) had a significantly greater prevalence of ST-elevation on hospital admission with TC.

In the past, TC was thought to be an acute reversible and transient disease with an overall good prognosis.² However, some recent studies have shown that in-hospital mortality is comparable with acute myocardial infarction.³ Cardiac dysfunction can cause life-threatening complications such as fulminant heart failure, cardiogenic shock, heart rupture and ventricular fibrillation.²³ The physiological rationale for β-blocker use in Takotsubo is based on mitigating excessive adrenergic stimulation, thought to underlie transient myocardial dysfunction.¹¹ Even with LV dysfunction in the acute phase, myocardial viability is preserved, and because of this, patients with TC have more complications during hospitalisation and the early use of β-blockers appears to reduce the incidence of such events.²⁴ Especially in patients who presented with hypertension or cardiogenic shock during hospitalisation, the use of β-blockers seems to be associated with improved long-term survival.²⁵ Notably, studies reporting higher mortality attributed most in-hospital deaths to non-cardiovascular causes such as hypoxic-ischaemic encephalopathy and sepsis¹⁸, highlighting the role of comorbid conditions in influencing prognosis.

Although the CI suggests a potential trend towards benefit, the lack of statistical significance limits firm conclusions. In-hospital use of β-blockers varied substantially, ranging from 15.7%¹⁴ to 89%.¹⁶ Overall, β-blockers were administered to approximately half of the patients (48%), despite wide variation in follow-up duration and study design. The cohort by Petursson et al,¹⁵ although retrospective, had the second-longest follow-up (877 days) and a high β-blocker usage rate (86%), while Isogai et al’s study,¹⁴ with the shortest follow-up (30 days), reported the lowest rate of β-blocker use (15.7%). This variability suggests that institutional protocols, study periods and patient selection may have influenced β-blocker prescription practices. Sometimes, the presence of haemodynamic instability represents a limitation to the early initiation of β-blocker therapy.²⁶

Our study has limitations. Our findings from five observational studies underscore the need for prospective, randomised trials with clearly defined inclusion criteria, stratification by clinical phenotype and evaluation of outcomes beyond mortality, such as recurrence, arrhythmias and functional recovery. Specific subgroups—such as male patients, those with significantly reduced LV ejection fraction or ventricular arrhythmias—may derive differential benefit and warrant targeted investigation. Additionally, all studies had at least moderate overall risk of bias, and two had serious risk of bias. Randomised, double-blind studies with a larger number of patients could provide more robust results.

Conclusion

In this meta-analysis, our findings suggest that early β-blocker therapy is not associated with a significantly reduced in-hospital mortality in patients with TC. Until further evidence from controlled trials is available, this therapy should be guided by individualised clinical judgement rather than routine practice.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.