Arterial hypertension in patients with takotsubo syndrome: prevalence, long-term outcome, and secondary preventive strategies: a report from the Takotsubo Italian Network register

Abstract

Aims

The aim of this study was to investigate the long-term outcome of takotsubo syndrome (TTS) patients with and without hypertension (HT) and to evaluate the effectiveness of treatment with beta-blockers (BBs) and/or renin–angiotensin–aldosterone system inhibitors (RAASi).

Methods and results

The study population includes a register-based, multicentre cohort of consecutive patients with TTS, divided into two groups according to the history of HT. Further stratification was performed for BB/RAASi prescription at discharge. The primary outcome was the composite of all-cause death and TTS recurrence at the longest available follow-up. The propensity score weighting technique was used to account for potential confounding. In the overall population (903 patients, mean age 70 ± 11 years), HT was reported in 66% of cases. At a median 2-year follow-up, there was no difference in the risk of the primary composite outcome between patients with and without HT. The adjusted Cox regression analysis showed a significantly lower risk for the primary outcome [adjusted hazard ratio (aHR): 0.69; 95% confidence interval (CI): 0.49–0.99] in patients who received BB vs. those who did not. Renin–angiotensin–aldosterone system inhibitors treatment was not associated with the primary study outcome. The lower risk for the primary outcome with BB treatment was confirmed in patients with HT (aHR: 0.37; 95% CI: 0.24–0.56) but not in patients without (aHR: 1.83; 95% CI: 0.92–3.64; P_interaction < 0.001).

Conclusion

In this TTS study, HT did not affect the long-term risk of adverse events but increased the probability of benefit from BB treatment after discharge. Owing to the favourable outcome impact of BB prescription in TTS patients with HT, a tailored pharmacological therapy should be considered in this cohort.

Introduction

Although takotsubo syndrome (TTS) has long been considered a benign disease, recent evidence from large-scale registries has demonstrated a substantial morbidity and mortality during follow-up, with a rate of adverse events comparable to that observed in patients with coronary syndromes.^1–5

The outcome of TTS is influenced by advanced age and comorbidity burden.^6–8 Hypertension (HT) is the most frequently reported comorbidity, affecting ∼65–70% of cases,^9–11 but its clinical and prognostic implications are not well understood in this patient setting.

Although the exact mechanisms of TTS are still unclear, elevated circulating catecholamines levels together with the increased sympathetic tone have been demonstrated to play a central role.^12,13 Unexpected stress or concomitant conditions may activate the cognitive centres of the brain and the hypothalamic–pituitary–adrenal axis resulting in ‘pathological’ epinephrine and norepinephrine release. The response of the cardiovascular system to this catecholaminergic surge hesitates in myocardial stunning with left ventricular wall motion abnormalities typical of TTS.

This pathophysiological framework suggests a potential role for long-term treatment with beta-blockers (BBs) and renin–angiotensin–aldosterone system inhibitors (RAASi), albeit in the absence of evidence from randomized studies.¹⁴ Observational evidence supporting the effectiveness of these medications is also controversial, probably due to the heterogeneity of the TTS population and the difficulty in identifying the profile of responder vs. non-responder subjects.^15,16 Conversely, the use of BB and RAASi as antihypertensive agents has been a cornerstone of real-world clinical practice for decades.^17,18 The aim of this study was to investigate the long-term clinical outcome of TTS patients with and without HT and to evaluate the effectiveness of pharmacological treatment with BB and/or RAASi in this particular patient cohort.

Methods

Study population

This observational, multicentre, cohort study included consecutive patients with TTS diagnosis admitted at 18 Italian hospitals and prospectively enrolled in the Takotsubo Italian Network (TIN) register. Patients were recruited according to the TIN diagnostic criteria, subsequently revised and incorporated by the Heart Failure Association and by the InterTAK Diagnostic Criteria (see Supplementary material online, Table S1).^19–21

All patients underwent coronary angiography and left ventriculography within 24 h of symptom onset. In the TIN register, patient data are collected through a standardized electronic data collection form that includes information on patient demographics, signs and symptoms at presentation, medical history, trigger events (primary or secondary TTS), ST-segment changes on admission electrocardiogram, and echocardiographic and coronary angiography parameters.²²

Left ventricular ejection fraction (LVEF) was calculated using the biplane Simpson’s rule from the apical four- and two-chamber views at admission and before discharge.⁷ Right ventricular (RV) wall motion was evaluated by visual assessment for the detection of RV involvement as previously described. The echo transducer was adjusted to the level of the RV chamber to achieve the optimal visualization of RV size and endocardial borders.^6,23

Left ventricular outflow tract obstruction was detected by continuous wave Doppler. Using the modified Bernoulli equation, a cut-off value of 25 mmHg for dynamic intraventricular pressure gradient was considered to indicate significant left ventricular outflow tract obstruction.²⁴ Mitral regurgitation was quantified from colour Doppler imaging and semi-quantitatively graded as absent, mild, moderate, or severe, using the standardized criteria.²⁴

The glomerular filtration rate was calculated from the creatinine value at admission by using the Chronic Kidney Disease Epidemiology Collaboration equation. In-hospital events, including ventricular malignant arrhythmias, acute heart failure, cardiogenic shock, need for an intra-aortic balloon pump, and/or inotropic agents, were also collected.

For the present analysis, patients were divided into two groups according to the history of HT, defined documented HT history in patients’ medical records and/or current antihypertensive treatment at the time of the hospitalization. Blood pressure–lowering drugs considered were α-blockers and other centrally acting agents, calcium blockers with mainly vascular effects, RAASi, BB, and thiazides. Patients with unknown status regarding a history of HT or dead during the hospitalization were excluded from this analysis. Further stratification was performed in accordance with the prescription of BB or RAASi at hospital discharge. The study selection process is summarized in Supplementary material online, Figure S1.

All participants provided informed written consent, and the study was approved by the local ethics committee at each participating site.

Follow-up and study outcomes

After hospital discharge, follow-up was performed through outpatient clinic visits, medical charts, or structured telephone interview by local investigators. The occurrence of the pre-specified adverse clinical events was recorded. Takotsubo syndrome recurrence was considered if a documented new episode met the diagnostic criteria for TTS.^19–21 Patients who did not experience the pre-specified outcomes of interest and those lost to follow-up were censored.

The study outcome measures were assessed at the longest available follow-up. The primary outcome of the present study was the composite of death for any cause and TTS recurrence. The secondary outcomes were the single components of the primary outcome.

Statistical analysis

The normal distribution of continuous parameters was visually assessed through histograms. Normally distributed variables were expressed as mean ± standard deviation and compared using Student’s t-test; variables with a skewed distribution were reported as median and interquartile range (IQR) and were compared with the Mann–Whitney U test. Categorical variables were reported as numbers and percentages and compared using the χ² test, or Fisher exact test, when appropriate. Survival free from the study outcomes was estimated by the Kaplan–Meier method. The unadjusted and adjusted hazard ratios (HRs) for the outcomes of interest were calculated using the Cox proportional hazard regression model and presented as HR with their 95% confidence intervals (CIs). First, we compared the clinical outcome of patients with vs. those without HT. Then, we compared the effectiveness of BB and RAASi in the overall population and patients with and without HT.

We used the propensity score weighting technique to account for potential selection bias in treatment assignment between groups (average treatment effect weights). Propensity score models were developed using a non-parsimonious approach and by incorporating a large number of baseline covariates potentially related to the outcome and/or treatment decision regardless of their statistical significance or collinearity with other variables included in the model.²⁵ The list of variables included in the models is reported in Supplementary material online, Table S2. After weighting, a standardized mean difference below 0.10, which reflects an optimal balance for all covariates included in the propensity score models, was achieved (see Supplementary material online, Figures S2–S4).

The rate of missing baseline values, if any, is shown in Supplementary material online, Table S3. Missing data were handled using multiple imputations with the method of chained equations. Twenty imputed data sets were generated.

Further sensitivity analyses were conducted to evaluate the consistency of the main results for the primary outcome in subgroups of clinical interest: age ≥ or < the overall population mean age, males or females, and LVEF ≥ or <50% at discharge.

A mediation analysis was also conducted to explore the causal pathways in case of a significant association between BB/RAASi with the primary composite outcome. For this purpose, age was assessed as a potential mediator related to the outcome effectiveness of pharmacological treatment. The proportion mediated was calculated as the ratio of the indirect effect to the total effect and was interpreted as mediation when positive or as suppressor when negative.²⁶

For all test, a P-value of <0.05 was considered statistically significant. Analysis was performed by using SPSS software version 25.0 (SPSS Inc., Chicago, IL, USA) and R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Study population

The study included 903 patients (mean age 70 ± 11 years, 91.6% females) and HT was reported in 594 (65.7%). The baseline characteristics and clinical course of TTS patients during the index event both in the overall population and according to the presence or not of HT are summarized in Table 1. Hypertension patients were older (73 vs. 65, P < 0.001) and showed a higher prevalence of cardiovascular risk factors including diabetes (P < 0.001), hypercholesterolaemia (P < 0.001), and tobacco use (P = 0.028) compared with subjects without HT.

Hypertension patients showed a significantly higher proportion of dyspnoea at clinical presentation (18.2 vs. 12.3%, P = 0.023) and lower glomerular filtration rate value (P < 0.001) than the group without HT. No differences in terms of LVEF, both at admission and at discharge, were observed; however, HT patients showed a higher proportion of acute heart failure during the hospitalization (16.4 vs. 9.5%, P = 0.005).

Renin–angiotensin–aldosterone system inhibitors were prescribed in 41.1% patients at discharge, BB in 60.7%. Renin–angiotensin–aldosterone system inhibitors treatment was higher in HT patients (49.7 vs. 24.8%, P < 0.001); conversely, there was no difference in the prescription of BBs between groups (60.6 vs. 60.9%, P = 0.924). The characteristics of patients who were prescribed BB and/or RAASi according to the presence or not of HT are summarized in Supplementary material online, Tables S4 and S5.

Long-term outcome

The median follow-up time was 24.0 months (IQR: 11.0–38.0 months). During follow-up, the composite outcome was reported in 138 (15.3%) patients; all-cause death occurred in 101 (11.2%) patients, and TTS recurrence in 45 (5.0%). There was no difference in the risk of the study outcomes between patients with and without HT both at unadjusted and adjusted Cox regression analysis (Figure 1 and Table 2).

Overall, the unadjusted Cox regression model showed a significantly lower risk for the primary composite outcome in the BB compared with the no-BB group (HR: 0.61; 95% CI: 0.44–0.86; P = 0.005). The risk of all-cause death was also lower in the BB than in the no-BB group (HR: 0.48; 95% CI: 0.32–0.72; P < 0.001), whereas no difference was detected for the risk of TTS recurrence (HR: 1.22; 95% CI: 0.65–2.30; P = 0.530; Figure 2). The adjusted regression model confirmed the significantly lower risk for the composite outcome (adjusted HR: 0.69; 95% CI: 0.49–0.99; P = 0.041) and for all-cause death (adjusted HR: 0.59; 95% CI: 0.39–0.89; P = 0.012) in patients who received BB vs. those who did not. Overall, RAASi treatment was not associated with the primary and secondary study outcomes both at unadjusted and adjusted analysis (Figure 2).

The lower risk for the primary composite outcome with BB treatment was confirmed in patients with HT (adjusted HR: 0.37; 95% CI: 0.24–0.56; P < 0.001) but not in patients without (adjusted HR: 1.83; 95% CI: 0.92–3.64; P = 0.086), with a significant sub-group interaction (P_interaction <0.001; Figure 3). In patients with HT, BB treatment was associated with a lower risk for all-cause death (adjusted HR: 0.28; 95% CI: 0.17–0.46; P < 0.001), not observed in those without HT (P_interaction <0.001). The absence of a significant association between RAASi and the study outcomes was confirmed both at unadjusted and adjusted Cox regression models, independently from the coexistence of HT (Figure 3).

There was no significant interaction between the subgroups of clinical interest on the effect of BB treatment with the primary composite outcome (see Supplementary material online, Figure S5). Also, the absence of a significant association between RAASi and the primary outcome was confirmed in the sub-group analysis.

The association between BB treatment and the primary composite outcome was not mediated by age (proportion mediated: 0.0%).

Discussion

This is the first registry-based study evaluating the effect of HT on the clinical outcome of TTS patients and assessing the effectiveness of treatment with BB or RAASi in TTS patients with and without HT.

The main findings of this can be summarized as follows:

1. Hypertension was a frequently reported condition but did not affect TTS patients’ long-term outcome.

2. Beta-blocker prescription at hospital discharge was associated with a lower risk of the primary composite outcome and all-cause death at long-term follow-up.

3. Hypertension played as a modifier of the effect of BB on long-term outcome, as BB reduced the risk of the primary outcome only in patients with HT.

4. Renin–angiotensin–aldosterone system inhibitors did not affect significantly patient outcome, regardless of the coexistence of HT.

Our study confirms that HT was a high-prevalence condition, being reported in 64% of cases, which is consistent with previous observational studies.^9–11 Since TTS patients are generally elderly and present multiple comorbidities, one may assume that the prevalence of HT merely reflects the demographics and clinical profile of this particular population.^6–8 In fact, HT patients were more likely to be older and showed a higher proportion of coexistent cardiovascular risk factors, comorbidities, and worse renal function, compared with those without HT. These conditions are all associated with a reduced survival in the general population and may justify the higher mortality risk of HT patients, which was not confirmed after adjustment for potential confounders. This result is also consistent with a previous study from the GEIST register, which reported no significant association of HT with the risk of TTS recurrence in the long term.⁹

In the present study, BB treatment was associated with a lower risk of the primary outcome and all-cause death at long-term follow-up. The pathological response of the myocardium to an abnormal sudden release of catecholamines provides the rationale for BB treatment that, by controlling both the basal sympathetic level and the hypothalamic–pituitary–adrenal axis activation, might reduce the risk of TTS recurrence and other adverse cardiovascular events, including mortality. However, there is no clear evidence supporting the long-term use of BB after clinical and LVEF recovery in TTS.^16,27,28 In a sub-analysis of 1115 subjects with available medication information from the InterTAK registry, there was no statistical association between BB treatment and all-cause mortality at 1-year follow-up.³ Conversely, in a recent analysis from the TIN register, including patients older than in the InterTAK register and followed for a longer period (median 2 years), we already observed a positive independent association between BB treatment and mortality.¹⁵ It could be hypothesized that the beneficial effect of BB treatment on clinical outcome becomes significant after longer treatment exposure. Indeed, our results were consistent with a retrospective population-based cohort study including 519 TTS patients, where BB treatment was independently associated with a lower risk for the composite of TTS recurrence and death at 5-year follow-up.²⁹

Beta-blocker may have pleiotropic effects and may be more effective in patients with the highest catecholamine levels. In this study, BBs were associated with higher overall survival in patients with HT, but not in those without, suggesting that HT may play as a modifier of the effect of BB on long-term outcome. This result could be explained by an increased catecholaminergic activity in TTS patients with HT who may constitute a particular patient phenotype characterized by higher sympathetic activation in response to triggering events, greater susceptibility to catecholamine-mediated myocardial damage, and low cardiac reserve. Therefore, this patient population might have a higher probability of benefit from BB treatment in the long term.

Hypertension pathophysiology is complex and multifactorial and, in the context of TTS, has never been investigated. In HT subjects, the sympathetic nervous system increases cardiac output, peripheral vascular resistance, and hydrosaline retention.^30–32 Furthermore, high catecholamine levels may induce α−1 adrenergic receptor-mediated endothelial dysfunction, vascular smooth muscle proliferation, and increased arterial stiffness, thus contributing to the maintenance and worsening of HT.³³ The renin–angiotensin–aldosterone system also acts as a blood pressure regulator through multiple mechanisms including pressure natriuresis and sodium retention, salt sensitivity, vasoconstriction, and endothelial dysfunction.^34,35

In the individual HT patient, blood pressure values may depend more on one mechanism than the other, and respond heterogeneously to different classes of antihypertensive drugs. Since catecholamine levels are supraphysiological in TTS patients, we hypothesized that HT may be largely dependent on the catecholaminergic overdrive in this patient population. Also, TTS patients with HT may have higher catecholamine levels than those without and benefit more from beta-adrenergic receptor blockade.

Conversely, we did not find an association between RAASi treatment and long-term outcome after TTS recovery, irrespective of the presence or absence of HT. Albeit the effectiveness of RAASi in TTS has long been debated, current evidence is limited to few studies with controversial results.^3,29,36 Our results suggest that RAASi are not associated with a more favourable long-term outcome in TTS and, conversely, the use of BB seems to be more useful for the effective prevention of adverse events in TTS patients with a history of HT.

Study limitations

Owing to the nature of this real-world multicentre observational study, our results should be considered hypothesis-generating and interpreted accordingly. To account for potential selection bias in treatment assignment, in this study, we employed the propensity score technique to balance several baseline patient-related characteristics. Although we included many variables in the model (non-parsimonious approach), we cannot exclude a residual selection bias secondary to other concealed confounders.

Since medication persistence and adherence during follow-up was not reported in the TIN register, the association of pharmacological treatment with the study outcomes should be interpreted with caution.

This analysis aimed at assessing the class effects of BB and RAASi rather than individual agents due to the sample size of the study. Dosage information was not complete in TIN and therefore could not be analysed in this study.

The follow-up of the TIN register was designed to evaluate adverse clinical events related to TTS, but not to monitor the response to pharmacological treatment. Therefore, we do not have information on blood pressure and heart rate, which precludes inference about the effect of drug treatment on the control of patients’ haemodynamic parameters. Also, data on frailty or patient’s functional status (e.g. sedentary time, physical activity, etc.) were not prospectively collected in this registry. The low number of males in this study, which reflect the general gender proportion of TTS, might limit the generalizability of the results in male patients.^3,37

Sub-group analyses usually do not provide definitive conclusions, but rather should be viewed as an important approach to generating hypotheses, which need confirmation by dedicated and adequately powered prospective studies.

Conclusions

In this study, HT did not influence the clinical outcome of TTS patients but was a modifier of the effect of pharmacological treatment after discharge. This study suggests that BB should be considered as the first-line therapy in TTS patients with HT, while in patients without HT, there is no clear benefit of BB or RAASi treatment.

Albeit needing confirmation by randomized studies, this evidence suggests the importance of a tailored pharmacological therapy in TTS, especially in patients with HT.

Author contributions

A.S. and R.C. contributed to the conception and design of the work. A.S. and M.D.M. performed the data analysis. E.B., G.P., F.S., O.V., C.Z., G.P., M.Cr., G.N., C.M., F.R., P.I., J.S-U., M.Ca. and F.A.-C. contributed to the acquisition of the data for the work. A.S., M.B., G.I., and R.C. drafted the manuscript. All authors critically revised the paper. All authors read and approved the final version of the paper.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology.

Funding

No financial support to declare.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References

Author notes