Clinical features, complications, and outcomes of exogenous and endogenous catecholamine-triggered Takotsubo syndrome: A systematic review and meta-analysis of 156 published cases.

Abstract Innumerable physical stress factors including externally administered catecholamines, and pheochromocytomas and paragangliomas (PPGLs) have been reported to trigger Takotsubo syndrome (TS). A systematic search of PubMed/MEDLINE identified 156 patients with catecholamine‐induced TS up to December 2017. Data were compared within the catecholamine‐induced TS cohort, but some comparisons were also done to a previously published large all‐TS cohort (n = 1750). The mean age was 46.4 ± 16.4 years (72.3% women). The clinical presentation was dramatic with high complication rates in (68.2%, n = 103; multiple complications 34.6%, n = 54). The most common TS ballooning pattern was apical or mid‐apical (45.2%, n = 69), followed by basal pattern (28.8%, n = 45), global pattern (16.0%, n = 25), mid‐ventricular (8.3%, n = 13), focal (0.6%, n = 1), and unidentified pattern (1.9%, n = 3). There was an increase in the prevalence of apical sparing ballooning pattern compared to all‐TS population (37.7% vs 18.3%, P < .00001). Higher complication rates were observed in TS with global ballooning pattern compared to apical ballooning pattern (23/25, 92% vs 38/65, 58.5%; P = .0022). Higher complication rates were observed in patients with age < 50 years than patients >50 years (73/92, 79.3% vs 29/56, 51.8%, P = 0.0009). Recurrence occurred exclusively in patients with PPGL‐induced TS (18/107 patients, 16.8%). PPGL‐induced TS was characterized by more global ballooning's pattern (22/104, 21.2% vs 3/49, 6.1%, P = 0.02), and lower left ventricular ejection fraction (25.54 ± 11.3 vs 31.82 ± 9.93, P = 0.0072) compared to exogenous catecholamine‐induced TS. In conclusion, catecholamine‐induced TS was characterized by a dramatic clinical presentation with extensive left ventricular dysfunction, and high complication rate.


| BACKGROUND
Takotsubo syndrome (TS) is an acute cardiac disease entity with a clinical presentation resembling that of an acute coronary syndrome. 1,2 The syndrome is characterized by a striking regional left ventricular wall motion abnormality (LVWMA) with a circumferential pattern extending beyond the coronary artery supply territory and resulting in a conspicuous ballooning of the left ventricle during systole. 3,4 Innumerable physical stress factors have been reported to trigger the disease. 5 Among the physical stressors are external administration of epinephrine 6 and norepinephrine 7 and the disease conditions causing increased catecholamine elevations, sometimes massively, such as pheochromocytomas and paragangliomas (PPGLs). [8][9][10] PPGLs are catecholamine-secreting tumors that arise from chromaffin tissue of the sympathetic nervous system. 11,12 PPGLs presentation may be vague and the interpretation of the symptoms and signs may be difficult. 13 The classic triad (hypertension, hyperhidrosis, and palpitation) and paroxysmal hypertension were the symptoms that usually lead to the suspicion of PPGLs previously. 13 PPGLs are more frequent in certain groups, for example, in patients with adrenal incidentalomas with 0.6% to 4.2% being affected but is otherwise generally rare. [14][15][16] Most PPGLs are nowadays diagnosed due to an incidentaloma, then due to catecholamine excess symptoms and finally because of screening in a previously known familial syndrome (eg, multiple endocrine neoplasia type 2, von Hippel Lindau syndrome, neurofibromatosis type 1, and mutations in succinate dehydrogenase B, C, and D) 12,13,[17][18][19] Cushing's syndrome due to ectopic ACTH-production from a PPGL can occasionally occur, 20,21 and thus all adrenal tumors should have a 1 mg overnight dexamethasone suppression test to exclude cortisol excess. 22,23 Sometimes an adrenal medullary hyperplasia may be the culprit of catecholamine excess. 24 The data on catecholamine-induced TS are limited and mostly consist of case reports and case series. Thus, in this systematic review and meta-analysis, the clinical features, complications and outcomes of 156 published cases of externally administered epinephrine-and norepinephrine-, and PPGL-induced TS are described. Furthermore, a comparison was made between the whole group and a study including all-TS population. The externally administered catecholamine-induced TS is also compared to the PPGL-induced TS.

| METHODS
All cases of epinephrine-induced TS, norepinephrine-induced TS, and PPGL-induced TS from 1990, the year where the Japanese term Takotsubo was introduced, to December 2017 were critically reviewed.
The cases were retrieved by systematical searches in PubMed/Medline using the search terms "Takotsubo," "apical ballooning," "stress cardiomyopathy," and "broken heart syndrome," and linking them with the terms "pheochromocytoma," "paraganglioma," "catecholamines," "epinephrine," "adrenaline," "norepinephrine," and "noradrenaline." Cases with PPGL-induced transient left ventricular dysfunction where the clinical features and course were consistent with TS were also included. Articles were initially screened by title for relevance and then by abstract, with full-text articles of potentially relevant reports reviewed. The reference lists of the retrieved full-text studies were scanned to identify additional relevant reports. Only case-reports or reports on a series of cases where enough information was available on every case were included ( Figure 1).
The following information was extracted from the publications: the age and gender of the patients, the clinical presentation, the type of ECG changes and the cardiac biomarkers in all patients.

| Statistical analysis
Continuous variables are presented as means ± standard deviations and categorical data as absolute values and percentages. Fisher's exact test or Chi-square test was used as appropriate to compare categorical data, and two-tailed unpaired student's t test was used for continuous variables. A P < .05 was considered significant.

| RESULTS
The systematic search identified 391 potentially relevant records, with an additional 10 records (including one case of PPGL-induced TS from 1987 25 with features typical for mid-apical TS) identified through review of the reference lists. After excluding all report not reporting on original cases, 148 articles were screened for eligibility. Further 12 non-English case reports were then excluded since information could not be extracted. Consequently, 136 publications were included; 8 publications reported on more than one case report (from 2 to 7 cases; Figure 1). Because of the extreme similarities between two reported pheochromocytoma-induced TS cases, 26,27 one case which was reported later was excluded. 27 In total, 156 cases reports (107 cases with PPGL-induced TS, 9,28-43 42 cases epinephrineinduced TS, 6,44-52 and 7 cases norepinephrine-induced TS [53][54][55][56][57][58][59] ) constitute the patient cohort for the meta-analysis. Among the cases included were two non-English case reports (one case in Swedish 60 and one in German 61 ) where enough information could be obtained.
Ten of 156 of the cases (9 PPGLs and 1 epinephrine), were deemed to be catecholamine-induced TS but were reported before the TS-era, that is, cases reported before 1999 when the first reports on TS were published in English. The remaining 146 cases were reported during the TS-era.

| Catecholamine-induced TS vs all-TS population
Compared to all-TS population, 5 the patients in catecholamineinduced TS were 20 years younger (P < .0001; Table 1). The TS prevalence in men was increased to 27.7% in catecholamine-induced TS compared to 10.2% in all-TS population; however, women were still predominating. The disease was more severe in catecholamineinduced TS with significantly higher heart rate and lower left ventricular ejection fraction; 16.3% of patients had global left ventricular dysfunction compared to 0% in all-TS population. 5 There was also increase in the prevalence of apical sparing ballooning pattern compared to all-TS population (37.7% vs 18.3%, P < .00001). The disease severity was reflected in significantly higher complication rate (68.2% vs 21.8%), more cardiogenic shock (37.7% vs 9.9%) and higher use of inotropic medications (32.7% vs 12.2%). Significantly higher recurrence rate was reported in catecholamine-induced TS (all in patients with PPGL-induced TS) than all-TS population.

| Exogenous catecholamine-induced TS vs PPGL-induced TS
The disease in PPGL-induced TS was more severe with significantly increased prevalence of global TS-pattern and lower left ventricular ejection fraction (Table 1). Complication rates were higher in PPGLinduced TS, but this did not reach significant levels. with PPGL-induced TS may deteriorate rapidly and the TS localization may transform from regional to global. 10  The more severe disease in PPGL-induced TS than in exogenous catecholamine-induced TS may be attributed to higher catecholamine surge in the former. 9 Unfortunately, it was not possible to evaluate the degree of catecholamine elevations in PPGL-induced TS in comparison to that of externally administered catecholamines. However, it has been reported that in external epinephrine-induced TS, the administration of >1 mg epinephrine was associated with significantly higher complication rate than administration of ≤1 mg epinephrine (92% vs 42.9%). been reported in other studies. 6,8,9,78 Furthermore, almost all TS studies have, apart from PPGL-induced TS, shown either normal or mild to moderate plasma epinephrine elevation. 79 Consequently, it is justified to conclude that there is no direct causal relation between epinephrine and TS but epinephrine may, as any other physical trigger factor, induce TS through sympathetic nervous system hyper-activation including cardiac sympathetic nerve terminals with norepinephrine seethe and spill over. 1,10,80

| Limitations
The analysis of catecholamine-induced TS was based on retrospective studies of case reports or series with the inherent limitations of all retrospective studies, in particularly that of ascertainment bias. The absolute levels of the cardiac biomarkers and catecholamine levels could not be utilized for the estimation of comparable mean values because of lack of standardization and uniformity across the case reports. The TS localization pattern was not always accurate, and this point has been discussed previously. 81 The possibility of additional trigger factors, which could have triggered TS, especially in patients with exogenous catecholamine-induced TS, cannot be ruled out. 82