2025 年 89 巻 10 号 p. 1627-1636
Background: Initially regarded as a benign acute cardiomyopathy, recent insights have shown that takotsubo syndrome (TTS) carries a prognosis comparable to that of acute coronary syndrome, with a notable impact of inflammatory burden. Given the seasonal variation seen in air pollution, inflammation, and coronary events, we investigated whether chronobiology and inflammation contribute to adverse outcomes.
Methods and Results: Between 2008 and 2020, all consecutive TTS patients were retrospectively included in a multicenter registry. We analyzed the impact of seasonal variation and inflammation on in-hospital events, including acute cardiac failure, cardiogenic shock, and death, as well as 30-day mortality. In-hospital events were identified in 238 (42.6%) patients. Higher rates of in-hospital events and 30-day mortality were observed during winter and spring than in summer and autumn. Multivariate analysis identified the presence of dyspnea on admission (odds ratio [OR] 4.02; 95% confidence interval [CI] 2.61–6.17; P<0.001), a neurological trigger (OR 2.58; 95% CI 1.21–5.50; P=0.014), hyperleukocytosis (OR 1.04; 95% CI 1.02–1.17; P=0.002), and left ventricular ejection fraction at admission (OR 0.98; 95% CI 0.96–1.00; P=0.011) as independent predictors of adverse outcomes.
Conclusions: In TTS, higher rates of in-hospital events and 30-day mortality were observed during winter and spring. Inflammatory burden and neurological disorders emerged as independent predictors of poor prognosis.
Originally viewed as a benign acute cardiomyopathy, recent insights have revealed that takotsubo syndrome (TTS) shares a similarly poor prognosis to that observed in acute coronary syndrome,1 with a particular emphasis on the detrimental impact of inflammatory burden. In addition to the well-established role of catecholamine surges in TTS pathophysiology,2–5 which trigger a switch in β2-adrenergic signaling from Gs- to Gi-protein, leading to further impaired cardiac function, recent studies have underscored the significance of both systemic and local inflammation, including monocyte-mediated myocardial inflammation during the acute phase and residual, long-lasting systemic inflammation.6–9
Challenging the previous notion of inflammation as merely a consequence of catecholamine release with little impact in the pathophysiology of TTS, recent evidence highlights the strong prognostic impact of inflammation on early acute heart failure, supraventricular arrhythmias, in-hospital mortality, and mid- to long-term outcomes, including impaired left ventricular ejection fraction (LVEF) recovery and increased cardiovascular mortality.7,8,10–14 Intramyocardial inflammatory activation during the acute phase of TTS has been demonstrated in different ways, including endomyocardial biopsies, autopsies, and cardiac magnetic resonance imaging with ultra-small iron oxide particles that are phagocytosed by macrophages.7,15
As a critical component of wound healing, immune cell infiltration into the damaged myocardium can also initiate a process known as sterile inflammation, where the immune system is activated in the absence of an infectious agent. This phenomenon may contribute to prolonged inflammation, as evidenced by elevated plasma concentrations of interleukin-6 (IL-6) and monocyte chemoattractant protein-1 , detectable up to 5 months after the onset of TTS.7 Other studies have highlighted the seasonality of inflammation and coronary events, potentially driven by air pollution or bronchopulmonary infections.16,17
Each season is characterized by multiple changes in environmental and physiological factors, such as low-temperature-induced vasoconstriction in winter, cyclical changes in blood pressure and cholesterol levels, changes in habitus, a peak of nitrogen oxides or “coarse” particles, such as PM10 and PM2.5 (fine particulate matter). These changes are likely to contribute to the atherothrombotic burden, with substantial increases in stroke, myocardial infarction, and aortic dissection being reported in winter.16,18–20 Although the chronobiological impact on the incidence of TTS has been explored previously, showing a higher prevalence in summer and peaks on Mondays and Tuesdays,19 the influence of seasonal variation on TTS severity remains unclear.
The aim of the present study was to investigate whether chronobiology and inflammation contribute to in-hospital events and adverse outcomes among TTS patients.
We conducted an observational multicenter retrospective cohort study including all consecutive TTS patients admitted to 3 hospitals in France, namely Strasbourg, Colmar, and Haguenau, from September 2008 to July 2020. Patients with suspected TTS were identified using the key words “stress,” “catecholamine,” “catecholaminergic,” and “Takotsubo” within the hospital databases. The diagnosis of TTS was confirmed according to the European Society of Cardiology Criteria.21,22 Patients with acute coronary occlusion, percutaneous coronary intervention, myocarditis, or cardiac arrest at first medical contact were excluded from the study. Two cardiologists reviewed all cases, blinded to admission date, and the data were recorded in the Alsace Takotsubo (ATAK) registry.
All participants provided written informed consent before study enrollment. The study protocol was approved by Strasbourg University institutional review board (CE 2016-9). The study complies with the principles outlined in the Declaration of Helsinki.
Data CollectionPatient data at admission were collected through a detailed review of electronic medical records. The day of coronary angiography was designated as the “diagnosis day.” Seasons were defined as follows: spring, March 20 to June 20; summer, June 21 to September 21; autumn, September 22 to December 20; and winter, December 21 to March 19 of the following year. A secondary analysis was performed by combining diagnoses made in spring and winter, and those made in summer and autumn.
Ejection fraction measurements were obtained via transthoracic echocardiography using the Simpson biplane method or ventriculography, when available. Due to evolving medical practices, a transition from troponin I to high-sensitivity troponin occurred in 2019. Therefore, an intermediate criterion, “N-troponin”, defined as the multiplier of the upper limit values set at 0.04 μg/L for troponin I and 57 ng/L (men) and 37 ng/L (women) for high-sensitivity troponin, was used.
Information on in-hospital complications, including heart failure, cardiogenic shock, and death, was collected through careful review of electronic medical records. Follow-up was conducted by telephone interviews with patients or, if unavailable, with their attending physicians or cardiologists, or through a review of electronic hospital records.
EndpointsThe primary endpoint in this study was a composite of in-hospital acute heart failure, cardiogenic shock, and in-hospital death. Patients who experienced any of these events were classified as “high-risk TTS,” whereas those who did not were classified as “low-risk TTS.” Mortality was also assessed at 30 days. Acute heart failure was defined according to 2021 European Society of Cardiology guidelines as an onset of symptoms and/or signs of heart failure occurring during hospitalization.23 Cardiogenic shock was defined according to the Society for Cardiovascular Angiography and Interventions clinical expert consensus as a patient with hypoperfusion requiring intervention beyond volume resuscitation (inotrope, pressor, mechanical support, including extracorporeal membrane oxygenation) to restore perfusion.24
Statistical AnalysisCategorical variables are presented as frequencies and percentages. Normally distributed continuous variables are presented as the mean±SD, whereas continuous variables with a skewed distribution are presented as the median with interquartile range (IQR). The significance of differences in categorical variables was evaluated using the χ2 test or Fisher’s exact test. Normally distributed continuous variables were compared using unpaired Student’s t-tests, whereas continuous variables with a skewed distribution were compared using the Wilcoxon test.
Seasonal variation was assessed using one-way analysis of variance (ANOVA) for normally distributed continuous variables, with post hoc pairwise comparisons adjusted by the Tukey-Kramer method. For non continuous variables with a skewed distribution, the Kruskal-Wallis test was used, followed by post hoc pairwise analysis using the Steel-Dwass test.
Associations between baseline patient characteristics and the occurrence of in-hospital events were evaluated using univariate and multivariate Cox regression analyses. Variables with P<0.05 in the univariate analysis were included in 2 multivariate models. Due to the collinearity between C-reactive protein (CRP) levels, leukocyte count and season, we created 2 distinct models: Model 1 included leukocyte counts and seasons; Model 2 included CRP levels. Kaplan-Meier curves and the log-rank test were used to analyze mortality during follow-up.
Given that previous studies have reported seasonal variations in CRP levels among healthy individuals, with peaks in winter and spring,25 and that air pollution has been observed to peak in winter in our region in France,16 additional subgroup analyses were conducted by combining winter and spring in one group and summer and autumn in another.
Two-tailed P<0.05 was considered statistically significant. Odds ratios (ORs) and hazard ratios (HRs) are reported with 95% confidence intervals (CIs). All analyses were performed using JMP 13 software® (SAS Institute, Cary, NC, USA).
Patient enrolment in this study is shown in Figure 1. In all, 559 TTS patients were identified between 2008 and 2020 and enrolled in the ATAK Registry. The baseline characteristics of the patients are presented in Table 1. Most patients were women (85%) and aged ≥65 years, with a mean age of 70±14 years. Cardiovascular risk factors were prevalent, including hypertension (58%), dyslipidemia (42%), diabetes (19%), and active smoking (18%). In addition, there were high rates of psychiatric disorders (31%), chronic respiratory disease (23%), cancer (23%), vascular disease (20%), and a history of stroke or transient ischemic attack (12%). Most patients presented with chest pain (49%) or dyspnea (41%) upon hospital admission. Physical triggers were identified in 44% of cases. The median length of hospitalization was 7 days (IQR 4–11 days).
Flowchart showing patient enrolment into the study between 2008 and 2020 from 3 hospitals in France (Strasbourg, Colmar, and Haguenau). ACS, acute coronary syndrome; TTS, takotsubo syndrome.
Baseline Clinical Characteristics of Patients With TTS and According to Season of Occurrence
| All TTS patients (n=559) |
Season of occurrence | P value | ||||
|---|---|---|---|---|---|---|
| Spring (n=157) | Summer (n=126) | Autumn (n=123) | Winter (n=153) | |||
| Age (years) | 70±14 | 69±14 | 72±13 | 69±14 | 72±14 | 0.18 |
| Female sex | 476 (85) | 133 (85) | 105 (83) | 105 (85) | 133 (87) | 0.86 |
| Height (cm) | 163±8 | 163±7 | 163±8 | 163±8 | 163±8 | 0.98 |
| Weight, kg | 66±15 | 65±15 | 66±16 | 66±14 | 67±16 | 0.86 |
| Hypertension | 326 (58) | 94 (60) | 78 (62) | 67 (54) | 87 (57) | 0.64 |
| Dyslipidemia | 233/558 (42) | 70 (45) | 52/125 (42) | 55 (45) | 56 (37) | 0.45 |
| Diabetes | 107 (19) | 34 (22) | 21 (17) | 28 (23) | 24 (16) | 0.34 |
| Active smoker | 103 (18) | 24 (15) | 29 (23) | 26 (21) | 24 (16) | 0.24 |
| Prior TTS | 16/558 (3) | 5/156 (3) | 2 (2) | 1 (1) | 8 (5) | 0.13 |
| Prior pacemaker | 23/558 (4) | 4/156 (3) | 9 (7) | 4 (3) | 6 (4) | 0.25 |
| Prior AF | 95 (17) | 27 (17) | 28 (22) | 13 (11) | 27 (18) | 0.11 |
| Vascular diseasesA | 112/557 (20) | 27/155 (17) | 38 (30)* | 17 (14)† | 30 (20)† | 0.009 |
| Prior stroke/TIA | 68 (12) | 20 (13) | 15 (12) | 15 (12) | 18 (12) | 0.99 |
| Dementia | 43/557 (8) | 11 (7) | 10/125 (8) | 12 (10) | 10 (7) | 0.75 |
| Psychiatric disorder | 172 (31) | 40 (25) | 39 (31) | 48 (39) | 45 (29) | 0.11 |
| Respiratory disease | 109/469 (23) | 26/131 (20) | 17/102 (17) | 24/102 (24) | 42/134 (31)*,† | 0.04 |
| Cancer history | 131 (23) | 43 (27) | 26 (21) | 26 (21) | 36 (24) | 0.52 |
| Pheochromocytoma | 4 (1) | 3 (2) | 0/125 (0) | 1/123 (1) | 0 (0) | 0.16 |
| Trigger of TTS (n=557) | 0.34 | |||||
| Unknown | 118 (21) | 35 (22) | 24 (19) | 31/122 (25) | 28/152 (18) | |
| Emotional | 150 (27) | 68 (43) | 51 (40) | 48/122 (39) | 79/152 (52) | |
| Physical | 246 (44) | 42 (27) | 40 (32) | 30/122 (25) | 38/152 (25) | |
| Neurological | 43 (8) | 12 (8) | 11 (9) | 13/122 (11) | 7/152 (5) | |
| Symptoms on admission | ||||||
| Chest pain | 274 (49) | 72 (46) | 53 (42) | 72 (59) | 77 (50) | 0.05 |
| Dyspnea | 231 (41) | 60 (38) | 51 (40) | 45 (37) | 75 (49) | 0.14 |
| Syncope | 28 (5) | 6 (4) | 8 (6) | 9 (7) | 5 (3) | 0.35 |
| QTc (s) | 470±53 | 470±52 | 471±52 | 462±54 | 476±55 | 0.35 |
| TTS pattern (n=554) | ||||||
| Apical | 402 (73) | 116/155 (75) | 92/124 (74) | 85/122 (70) | 109 (71) | 0.75 |
| Midventricular | 125 (23) | 35/155 (23) | 27/124 (22) | 32/122 (26) | 31 (20) | 0.69 |
| Basal | 18 (3) | 4/155 (3) | 4/124 (3) | 4/122 (3) | 6 (4) | 0.93 |
Unless indicated otherwise, data are given as n (%), n/N (%), mean±SD. *P<0.05 compared with spring. †P<0.05 compared with summer. AVascular diseases includes any kind of artery disease excluding ischemic cerebral event (stroke, transient ischemic attack [TIA]). AF, atrial fibrillation; QTc, corrected QT interval; TTS, Takotsubo syndrome.
The characteristics of TTS patients, stratified by season of occurrence, are presented in Tables 1 and 2.
Echocardiographic and Biological Parameters of Patients With TTS According to Season of Occurrence
| Spring (n=157) |
Summer (n=126) |
Autumn (n=123) |
Winter (n=153) |
P value | |
|---|---|---|---|---|---|
| Echocardiography | |||||
| LVEF (%) | |||||
| On admission | 38±12* | 37±12* | 42±11 | 38±11* | 0.008 |
| At discharge | 51±13 | 51±12 | 52±11 | 50±13 | 0.35 |
| LVEF recovery (%) | 12±15 | 14±12 | 11±12 | 11±10 | 0.30 |
| At baseline | |||||
| WBC (×109/L) | 11.4±5.6 | 11.3±5.2 | 11.5±5.3 | 11.5±5.4 | 0.99 |
| Hemoglobin (g/dL) | 12.7±1.9 | 12.8±2.1 | 13.2±1.8 | 12.6±1.8 | 0.097 |
| Platelet (×109/L) | 273±101 | 263±104 | 270±96 | 259±116 | 0.67 |
| Sodium (mEq/L) | 137±5 | 137±5 | 137±5 | 137±5 | 0.60 |
| Potassium (mEq/L) | 4.0±0.7 | 3.9±0.6 | 3.9±0.5 | 4.0±0.6 | 0.60 |
| Creatinine (μmol/L) | 66 (53–89) | 66 (52–87) | 65 (53–81) | 64 (53–81) | 0.72 |
| eGFR (mL/min/1.73 m2) | 84 (59–95) | 81 (60–91) | 82 (61–91) | 81 (61–91) | 0.66 |
| CRP (mg/L) | 8 (3–48) | 7 (3–34) | 11 (3–45) | 16 (4–60)† | 0.04 |
| BNP (ng/L) | 350 (117–1,057) | 388 (185–959) | 333 (105–949) | 458 (153–1,181) | 0.48 |
| N-troponin | 28 (6–85) | 23 (7–92) | 26 (4–73) | 41 (9–100) | 0.34 |
| Peak | |||||
| WBC (×109/L) | 13.4±6.9 | 13.7±7.8 | 13.2±6.3 | 13.5±7.1 | 0.95 |
| CRP (mg/L) | 37 (7–94) | 19 (4–105) | 23 (4–99) | 37 (6–103) | 0.35 |
| BNP (ng/L) | 579 (228–1,463) | 628 (282–1,210) | 482 (244–1,212) | 643 (259–1,441) | 0.75 |
| N-troponin | 67 (18–143) | 58 (23–143) | 71 (25–158) | 64 (19–165) | 0.91 |
| At discharge | |||||
| WBC (×109/L) | 8.3±5.2 | 8.5±3.9 | 8.4±3.6 | 9.0±5.6 | 0.57 |
| CRP (mg/L) | 9 (3–29) | 8 (3–19) | 9 (3–27) | 14 (4–31) | 0.29 |
| BNP (ng/L) | 297 (131–670) | 365 (174–694) | 320 (152–583) | 272 (144–562) | 0.40 |
| N-troponin | 10 (3–40) | 10 (3–32) | 12 (3–48) | 10 (3–34) | 0.82 |
Unless indicated otherwise, data are given as the mean±SD or median (interquartile range). *P<0.05 compared with autumn; †P<0.05 compared with summer. BNP, B-type natriuretic peptide; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction; N-troponin, the multiplier of the upper limit values of the troponins (see Methods); TTS, Takotsubo syndrome; WBC, white blood cell.
The proportion of patients with lung disease differed significantly between winter and spring and summer (31% vs. 20% and 17%, respectively; P=0.04). In contrast, there was a significant peak in patients with vascular disease (excluding stroke/transient ischemic attack) during summer (30%; P=0.009). Emotional stress was more frequently identified as a trigger during winter (52%), although this did not reach statistical significance. Elevated CRP levels at admission were more frequently observed in winter, suggesting increased inflammatory burden during this season. However, no seasonal variation was found in markers of myocardial injury (as assessed by troponin levels), neurohormonal activation (B-type natriuretic peptide [BNP] levels), or TTS patterns (apical vs. other types).
Events occurring during hospitalization, by season of diagnosis, are presented in Table 3. Among the 39 patients who died during hospitalization, 16 (41.0%) deaths were attributed to cardiovascular causes. It appeared that season had no effect on the primary endpoint. However, patients hospitalized in spring experienced a higher incidence of in-hospital cardiac arrests than those hospitalized in summer or autumn (10% vs. 2% and 2%, respectively; P=0.02). In addition, patients hospitalized in spring were more likely to develop heart conduction problems requiring temporary pacemaker implantation (vs. summer and autumn groups; P=0.02) or permanent pacemaker implantation (vs. summer group; P=0.047). At hospital discharge, inflammatory status, neurohormonal activation, and myocardial damage were similar across all groups.
Clinical Events Among Patients With TTS According to Season of Occurrence
| Season | P value | ||||
|---|---|---|---|---|---|
| Spring (n=157) |
Summer (n=126) |
Autumn (n=123) |
Winter (n=153) |
||
| In-hospital events | |||||
| Death | 13 (8) | 8 (6) | 3 (2) | 15 (10) | 0.099 |
| Cardiovascular death | 6 (4) | 2 (2) | 1 (1) | 7 (5) | 0.193 |
| Acute heart failure | 56 (36) | 50 (40) | 34 (28) | 51 (33) | 0.24 |
| Cardiogenic shock | 26 (17) | 15 (12) | 9 (7) | 22 (14) | 0.13 |
| Composite endpointA | 73 (46) | 60 (48) | 40 (33) | 65 (42) | 0.06 |
| Supraventricular arrhythmia | 34 (22) | 27 (22) | 18 (15) | 35 (23) | 0.35 |
| Cardiorespiratory arrest | 15 (10) | 3 (2)* | 3 (2)* | 11 (7) | 0.02 |
| Ventricular tachycardia | 2 (1) | 4 (3) | 1 (1) | 6 (4) | 0.25 |
| Ventricular fibrillation | 5 (3) | 2 (2) | 1 (1) | 3 (2) | 0.54 |
| Sick sinus syndrome | 1 (1) | 1 (1) | 4 (3) | 3 (2) | 0.30 |
| Complete AV block | 4 (3) | 2 (2) | 0 (0) | 0 (0) | 0.09 |
| Temporary pacemaker | 6 (4) | 0 (0)* | 0 (0)* | 2 (1) | 0.02 |
| Permanent pacemaker | 7 (4) | 0 (0)* | 1 (1) | 4 (3) | 0.047 |
| Implantable defibrillator | 1 (1) | 0 (0) | 0 (0) | 1 (1) | 0.66 |
| Infection | 60 (38) | 43/124 (35) | 38 (31) | 59 (39) | 0.52 |
| Death within 30 days | 11/151 (7) | 3/121 (2)† | 3/118 (3)† | 13/148 (9) | 0.046 |
| Cardiovascular death | 6 (4) | 2 (2) | 1 (1)† | 8 (5) | 0.123 |
Unless indicated otherwise, data are n (%) or n/N (%). *P<0.05 compared with spring; †P<0.05 compared with winter. AComposite of in-hospital acute heart failure, cardiogenic shock, and in-hospital death. AV, atrioventricular; TTS, Takotsubo syndrome.
Thirty-day mortality was significantly higher among patients hospitalized in winter than among those hospitalized in summer or autumn (9% vs. 2% and 3%, respectively; Table 3). Among the 30 patients who died during the first 30 days, 17 (56.7%) deaths were attributed to cardiovascular causes. It is important to note that only a few patients died in the hospital beyond the initial 30-day period and were therefore excluded from the 30-day mortality count.
In subgroup analysis, the spring+winter group had a higher incidence of cardiorespiratory arrest (8% vs. 2%; P=0.003), in-hospital deaths (9% vs. 4%; P=0.03), cardiogenic shock (15% vs. 10%; P=0.04), and 30-day mortality (8% vs. 2%; P=0.002) than the summer+autumn group (Figure 2). Patient characteristics stratified by type of TTS (high-risk vs. low-risk TTS) are presented in Supplementary Table 1. Patients who experienced in-hospital events and 30-day mortality had higher CRP levels than those who did not (Figure 3).
Intrahospital deaths, acute heart failure, cardiogenic shock, cardiorespiratory arrest, mortality at 30 days, and survival curves to Day 30 in patients grouped according to season of takotsubo syndrome (summer+autumn vs. winter+spring).
Box plots illustrating the prognostic impact of C-reactive protein levels at admission for in-hospital events (Left) and 30-day mortality (Right). “In-hospital events” was a composite of in-hospital acute heart failure, cardiogenic shock, and in-hospital death. The boxes show the interquartile range, with the median value indicated by the horizontal line; whiskers show the range. Values above each box are the median and interquartile range CRP concentration.
Results of univariate analyses for predictors of adverse intrahospital events are presented in the Supplementary Table 2. Multivariate analysis revealed that the independent predictors of high risk TTS (Table 4) were the presence of dyspnea on admission (OR 4.02; 95% CI 2.61–6.17; P<0.001), a neurological trigger (OR 2.58; 95% CI 1.21–5.50; P=0.014), hyperleukocytosis (OR 1.04; 95% CI 1.02–1.17; P=0.002), and LVEF at admission (OR 0.98; 95% CI 0.96–1.00; P=0.011).
Multivariate Analysis for Predictors of Adverse Intrahospital Events (Death, Heart Failure, Cardiogenic Shock)
| Model 1A | Model 2A | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | 1.01 | 1.00–1.03 | 0.124 | 1.01 | 1.00–1.03 | 0.074 |
| Female sex | 0.56 | 0.35–0.89 | 0.015 | 0.65 | 0.40–1.04 | 0.072 |
| Weight | 0.98 | 0.97–1.00 | 0.012 | 0.99 | 0.98–1.00 | 0.049 |
| Prior AF | 1.16 | 0.77–1.75 | 0.482 | 1.17 | 0.77–1.77 | 0.458 |
| Vascular diseaseB | 1.31 | 0.87–1.98 | 0.192 | 1.21 | 0.81–1.80 | 0.706 |
| Respiratory disease | 0.75 | 0.49–1.15 | 0.191 | 0.91 | 0.59–1.39 | 0.651 |
| Trigger of TTS | ||||||
| Unknown | Ref. | Ref. | Ref. | Ref. | Ref. | Ref. |
| Physical | 1.45 | 0.87–2.42 | 0.163 | 1.74 | 1.05–2.89 | 0.031 |
| Emotional | 1.35 | 0.76–2.38 | 0.302 | 1.46 | 0.82–2.57 | 0.197 |
| Neurological | 2.58 | 1.21–5.50 | 0.014 | 2.61 | 1.22–5.58 | 0.013 |
| Symptoms on admission | ||||||
| Chest pain | 0.67 | 0.45–1.00 | 0.047 | 0.64 | 0.43–0.96 | 0.03 |
| Dyspnea | 4.02 | 2.61–6.17 | <0.001 | 4.04 | 2.59–6.30 | <0.001 |
| Season | ||||||
| Summer | Ref. | Ref. | Ref. | |||
| Autumn | 1.14 | 0.67–1.93 | 0.632 | |||
| Winter | 1.51 | 0.90–2.53 | 0.117 | |||
| Spring | 1.26 | 0.76–2.10 | 0.377 | |||
| LVEF at admission | 0.98 | 0.96–1.00 | 0.011 | 0.97 | 0.96–0.99 | 0.001 |
| WBC | 1.04 | 1.02–1.07 | 0.002 | |||
| Hemoglobin | 0.97 | 0.89–1.05 | 0.439 | 1.01 | 0.92–1.10 | 0.896 |
| eGFR | 1 | 0.99–1.00 | 0.194 | 1 | 0.99–1.00 | 0.22 |
| BNP (per 100-ng/L increase) | 0.99 | 0.97–1.00 | 0.178 | 0.99 | 0.97–1.00 | 0.095 |
| CRP (per 10-mg/L increase) | 1.02 | 0.99–1.04 | 0.133 | |||
ADue to the collinearity between CRP levels, leucocyte count, and season, 2 distinct models were created: Model 1, which included all variables with P<0.05 in the univariate analysis, including leucocyte count and season; and Model 2, which included all variables with P<0.05 in the univariate analysis, including CRP levels. Due to the collinearity between creatinine, eGFR, sodium and potassium, only eGFR was included in the models. BVascular diseases includes any kind of artery disease excluding ischemic cerebral event (stroke, TIA). CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.
Independent predictors of 30-day mortality included the occurrence of TTS in winter (HR 4.46; 95% CI 1.18–29.12; P=0.04) and a neurological trigger (HR 5.39; 95% CI 1.08–26.87; P=0.04; Table 5).
Cox Regression Model of Predictors of 30-Day Mortality
| Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Season | ||||
| Spring | 4.04 (1.06–26.27) | 0.04 | 3.37 (0.85–22.36) | 0.09 |
| Summer | 1.53 (0.25–11.63) | 0.64 | 1.31 (0.21–10.14) | 0.77 |
| Autumn | Ref. | – | Ref. | – |
| Winter | 5.33 (1.47–34.09) | 0.008 | 4.46 (1.18–29.12) | 0.03 |
| Age | 1.05 (1.01–1.08) | 0.004 | ||
| Female sex | 0.48 (0.21–1.21) | 0.11 | ||
| Height | 1.01 (0.97–1.06) | 0.53 | ||
| Weight | 1.01 (0.99–1.03) | 0.29 | ||
| Hypertension | 1.01 (0.48–2.18) | 0.98 | ||
| Dyslipidemia | 0.48 (0.19–1.08) | 0.08 | ||
| Diabetes | 1.63 (0.67–3.56) | 0.26 | ||
| Active smoker | 0.75 (0.22–1.95) | 0.59 | ||
| Prior TTS | 5.357e–9 | 0.21 | ||
| Prior pacemaker | 1.9295e–9 | 0.13 | ||
| Prior AF | 1.10 (0.37–2.68) | 0.84 | ||
| Vascular diseasesA | 2.29 (1.02–4.87) | 0.046 | 2.10 (0.89–4.75) | 0.09 |
| Prior stroke/TIA | 1.95 (0.72–4.52) | 0.18 | ||
| Dementia | 2.10 (0.62–5.43) | 0.21 | ||
| Psychological disorder | 0.59 (0.22–1.38) | 0.24 | ||
| Respiratory disease | 1.61 (0.65–3.67) | 0.28 | ||
| Cancer history | 2.87 (1.35–6.05) | 0.007 | 1.86 (0.80–4.25) | 0.15 |
| Pheochromocytoma | 1.5087e–8 | 0.57 | ||
| Trigger of TTS | ||||
| Unknown | 1.374e–9 | <0.001 | 4.684e–10 | 1.00 |
| Emotional | Ref. | – | Ref. | – |
| Physical | 4.22 (1.45–17.85) | 0.006 | 2.86 (0.78–10.46) | 0.11 |
| Neurological disorder | 4.46 (1.00–19.93) | 0.05 | 5.39 (1.08–26.87) | 0.04 |
| Symptoms on admission | ||||
| Chest pain | 0.46 (0.20–1.00) | 0.0495 | 0.84 (0.36–1.96) | 0.68 |
| Dyspnea | 2.47 (1.16–5.55) | 0.02 | 1.50 (0.65–3.47) | 0.34 |
| Syncope | 0.81 (0.05–3.79) | 0.83 | ||
| QTc | 1.00 (0.99–1.01) | 0.94 | ||
| LVEF on admission | 0.95 (0.92–0.98) | 0.004 | 0.97 (0.94–1.01) | 0.099 |
| Type of TTS | ||||
| Apical | 1.68 (0.69–5.00) | 0.27 | ||
| Midventricular | 0.58 (0.17–1.50) | 0.28 | ||
| Basal | 5.3389e–9 | 0.20 | ||
| Laboratory test on admission | ||||
| WBC | 1.04 (0.98–1.10) | 0.16 | ||
| Hemoglobin | 0.78 (0.65–0.94) | 0.009 | 1.00 (0.82–1.23) | 0.98 |
| Platelets | 1.00 (1.00–1.00) | 0.96 | ||
| Sodium | 0.94 (0.88–1.01) | 0.07 | ||
| Potassium | 1.11 (0.59–2.01) | 0.75 | ||
| Creatinine | 1.00 (1.00–1.01) | 0.34 | ||
| eGFR | 0.99 (0.98–1.01) | 0.24 | ||
| CRP (per 10-mg/L increase) | 1.05 (1.00–1.09) | 0.07 | ||
| BNP (per 100ng/L increase) | 1.03 (0.99–1.06) | 0.16 | ||
| N-troponin | 1.00 (1.00–1.00) | 0.94 | ||
AVascular diseases include any kind of artery disease excluding ischemic cerebral event (stroke, TIA). CI, confidence interval; HR, hazard ratio. Other abbreviations as in Tables 1,2.
The major findings of this multicenter study indicate that TTS occurring in winter and spring is associated with increased severity and 30-day mortality. The high incidence of heart failure, unaffected by season, masks the real impact of season on the worst outcomes of TTS, as evidenced by a higher burden of cardiogenic shock, cardiac arrest, and in-hospital death among patients with TTS occurring during winter and spring. Neurological disorders and inflammatory burden emerged as independent predictors of early mortality.
This study reinforces the prognostic significance of neurological triggers, consistent with the InterTAK prognostic score, which highlights neurological events as a major factor in TTS outcomes.26 The involvement of the hypothalamic-pituitary-adrenal axis, particularly through catecholamine surges, likely accounts for the severity of TTS triggered by neurological factors, independent of inflammatory mechanisms.
Recent studies have emphasized the role of monocyte diapedesis following catecholamine surges as a potential consequence of endothelial dysfunction, increased cytoadhesion, and increased permeability, all of which contribute to inflammatory cell infiltration into the myocardium.2,5,27,28 Resident cardiac macrophages, which are generally anti-inflammatory and cardioprotective, whereas infiltrating monocyte-derived macrophages adopt a proinflammatory and harmul role in TTS.9,29 Liao et al.29 demonstrated that adrenergic surges shift monocytes towards a proinflammatory phenotype, this response peaking the day after the adrenergic surge and returning to baseline by 7 days later. Using pharmacologic and genetic models, Liao et al.29 highlighted macrophages as key regulators of TTS pathogenesis. From a clinical viewpoint, high inflammatory burden has already been correlated with mortality and incomplete recovery following TTS.11,30 Recent analysis by our group demonstrated that the incorporation of inflammatory markers at discharge enhances the predictive capacity of the conventional InterTAK score.30 Our findings in the present study align with these data, showing that a higher inflammatory burden (as evidenced by the leukocyte count) was a marker of poor prognosis in high-risk TTS patients. Although CRP was not an independent predictor of in-hospital events, likely due to its kinetics and collinearity with the leukocyte count, elevated leukocyte counts proved to be a more sensitive marker of inflammation. In addition, our finding that a higher inflammatory burden was a marker of poor prognosis could underscore a specific deleterious role of leukocytes infiltrating into the myocardium in TTS.29
Considering seasonal variations in air pollution and cardiovascular events, we explored whether season influenced TTS severity. Literature reviews have previously characterized a chronobiology of TTS,19 noting a higher prevalence in the morning, on certain days of the week (i.e., Monday, Tuesday), and during summer, possibly related to increased catecholamine release. Although our study did not specifically focus on the seasonal prevalence of TTS, we observed increased severity during winter and spring, with no summer peak in incidence, contrary to previous findings.19 Given that the major previous studies were conducted in US and Japanese populations, in contrast to our French cohort, the effects of environmental factors (e.g., temperature fluctuations, air pollution, and seasonal variations in catecholamine secretion) and genetics warrant further investigation.
Inflammatory responses appeared more pronounced during winter,31 which may explain the increased severity of TTS during winter. Prior studies have emphasized seasonal variation in CRP levels among healthy individuals,31 with peaks in winter and spring. Similarly, analyses of RNA from deceased human donors and analysis of the transcriptome profile of leukocytes and adipocytes revealed a winter-associated proinflammatory profile, marked by increased IL-6 receptor, CRP, and monocyte levels.25,32 In addition, air pollution, particularly fine particles linked to IL-6 secretion, exhibits a winter peak that parallels increased coronary events in northeast France.16 This suggests that the seasonality observed in TTS severity may reflect an underlying proinflammatory state, creating conditions favorable for severe TTS presentations. However, the collinearity between CRP and winter prevents a clear delineation of their independent effects on TTS severity.
Further, seasonality may also affect the cerebral cellular architecture, with studies showing increased neuronal gene expression in the hypothalamus during winter.32 Although a connection between these cerebral changes and the hypothalamic-pituitary-adrenal axis in TTS is plausible, the functional impact of these changes remains speculative and requires further exploration.
The effects of inflammatory infiltrates in TTS are linked to oxidative stress and local cytokine production, causing direct myocardial damage. Nitric oxide release during TTS, likely secondary to M1 macrophage activation, leads to inhibition of glycolysis and contractile dysfunction due to energy depletion.33–36 Scally et al. found elevated IL-6 levels during both the acute and late TTS phase, with persistent IL-6 elevation correlating with incomplete recovery of left ventricular function.7,8 Similar findings of Tarantino et al. reinforced the role of IL-6 in determining TTS prognosis.14
Neurohormonal activation, specifically peak BNP levels, could play a crucial role in assessing TTS prognosis. Prior studies have linked catecholamine surges and inflammation to BNP levels, which not only reflect cardiac dysfunction but also act as markers of systemic inflammation.11,37 BNP appears to mitigate oxidative stress, promote myocardial relaxation, and modulate vascular permeability, possibly as a protective response to inflammatory damage.38,39 In patients with longstanding inflammation, we have recently established the noxious role of proinflammatory cytokines in vascular damage through the NADP/oxidase/sodium-glucose cotransporter 2 (SGLT2) pathway,40 an effect prevented by empagliflozin. Moreover, a correlation between low-grade inflammation and SGLT2 expression in the human vasculature and heart was recently demonstrated.41 Investigating anti-inflammatory strategies, such as the use of SGLT2 inhibitors, holds promise for TTS treatment. Our recent preclinical studies in a rat model of TTS induced by isoproterenol infusion demonstrated that empagliflozin, an SGLT2 inhibitor, not only prevented TTS onset but also mitigated left ventricular dysfunction, inflammatory infiltration, and fibrotic remodeling.42 Because this therapeutic class has already demonstrated its efficacy in acute heart failure,43 further clinical studies are required to evaluate its impact on TTS.
Study LimitationsThis study has several limitations. Its retrospective design may have introduced selection bias, because patients who did not undergo coronary angiography were not included. TTS in severely ill patients may have been missed. The retrospective design of the study resulted in inconsistent timing in the collection of intrahospital data. Although echocardiographic and biological parameters at admission were measured over a standardized timeframe, discharge parameters were collected at varying time points, depending on the length of hospitalization. In addition, the study population was limited to the Alsatian region, spanning only 3 centers within close proximity, potentially affecting the generalizability of our findings. However, the proximity of the study centers had the advantage that all patients were simultaneously exposed to the same pollution variations and seasons. Future studies should aim to assess TTS chronobiology on a larger, international scale. Furthermore, key biochemical measurements, such as catecholamine and proinflammatory cytokine levels, were not assessed due to the retrospective nature of the study, limiting our ability to fully elucidate their role in TTS.
This retrospective multicenter study suggests the possible chronobiological influence of inflammation on the severity of TTS. An increased inflammatory response (leukocyte counts) and neurological factors were associated with worse cardiovascular outcomes. The findings underscore the potential benefit of exploring anti-inflammatory therapies, such as SGLT2 inhibitors, in the management of TTS.
This study was supported by GERCA (Groupe pour l’Enseignement et la Recherche Cardiologique en Alsace, France).
The authors declare that they have no known conflicts of interest associated with this publication.
This work was approved by the Ethics Committee of the Faculties of Medicine, Odontology, Pharmacy, Nursing, Physiotherapy, Midwifery and Hospitals of Strasbourg. Ref CE: 2016-9.
Data will be made available upon reasonable request.
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-24-0762
