Characterization of priMary And sEcondary STress Related takOtsubo

Background and rationale Takotsubo syndrome (TTS) is an acute and reversible form of myocardial injury often preceded by a physical or emotional trigger. Although TTS was generally considered a benign disease for its reversible nature, it is now clear that hemodynamic and electrical instability during the acute phase exposes patients to frequent serious adverse in-hospital complications. However, the pathophysiology of TTS is far from being completely understood. Consistent evidence demonstrated that the environmental events experienced by most of these patients and perceived as stressful (both physical or emotional) induce a brain activation and a stress-related response, with increasing bioavailability of local and circulating stress mediators, such as catecholamine and cortisol, which showed to play a major role in the etiology of to the "neurogenic stunning myocardium" responsible for this clinical condition. Recent studies strengthened the hypothesis of an outstanding link between the brain stress response system and heart in TTS patients using neuroimaging approach. The fundamental anatomic structures involved in the stress response are the neocortex, limbic system, reticular formation, brainstem, and spinal cord along with the hypothalamic-pituitary-adrenal axis which finally leads to cortisol secretion. In this regard, substantial structural differences in the neocortex and the limbic network (insula, amygdala, cingulate cortex, and hippocampus), have been shown among TTS patients compared to healthy controls using brain functional magnetic resonance imaging (fMRI), along with a hypoconnectivity of the central brain regions holding a regulatory function of the autonomic and limbic system. Moreover, a recent PET/TC study demonstrated that heightened limbic activity precedes the development of TTS and that patients with the highest activity of the amygdala develop the syndrome earliest, supporting the hypothesis that a neurobiological substrate may predispose them to this clinical syndrome.

Of interest, since TTS has been identified in an increasing number of hospitalized patients, an important clinical heterogeneity emerged among those affected, and clinical characteristics such as physical triggers along with acute neurologic or psychiatric disease, high troponin levels, and low ejection fraction showed to identify a specific category of individuals with TTS at higher risk of mortality and in-hospital complications. Therefore, it is recently emerging that so far two different categories of TTS patients were described, with different clinical features and risk profiles: primary TTS, which mainly affects patients after an emotional stressor, in the absence of epicardial coronary disease, with minor troponin release, slightly reduced and rapidly reversed left ventricular (LV) dysfunction and benign prognosis; the second one of secondary TTS, occurring after a physical stressor, in the presence of epicardial coronary artery disease (CAD), with major troponin release, with more severe or persistent LV dysfunction and associated with worse prognosis. An interesting hypothesis, that remains to be tested, is that only primary TTS might result from reversible left ventricular dysfunction of neurogenic origin through activation of neurons originating in the limbic system which may cause reversible vasoconstriction of coronary microvasculature, while secondary TTS may be due to a direct catecholamine-induced myocardial damage provoked by a physical trigger in patients with concomitant CAD. However, the pathophysiological determinants of primary and secondary TTS respectively have never been investigated so far as well as a different brain-heart axis activation in these two categories of TTS patients has never been demonstrated.

Myocardial dysfunction following a physical stressor has been also described in sepsis-induced cardiomyopathy, a well-known systemic complication of the cytokine storm following the host immune response to infection, which can significantly affect the prognosis of these patients. Both left ventricular systolic dysfunction (LVSD) and LV diastolic dysfunction (LVDD) have been described in the first period of severe septic shock as in secondary TTS, but except for the reversible nature of the myocardial dysfunction, clinical and echocardiographic characteristics (e.g. typical left ventricular kinetic abnormalities, electrocardiographic features, and possible detrimental effects of vasopressors use) seemed to distinguish secondary TTS from sepsis-induced cardiomyopathy, thus suggesting a different host systemic response to the same "stressors" and different pathophysiological mechanisms underlying these clinical conditions.

Biochemical profiles of primary vs. secondary TTS patients are still largely unidentified, and whether a different profile exists associated with these two categories of patients remains unknown. The known mechanisms associated with the development of TTS include elevated levels of circulating plasma catecholamines and their metabolites. TTS has been reported to be characterized by a myocardial macrophage inflammatory infiltrate and an increase in systemic proinflammatory cytokines. Indeed, recent studies reported that patients with acute TTS had elevated levels of the pro-inflammatory cytokines IL-6, IL-8 and CXCL1 in the blood, however, to the best of our knowledge no study before ever investigated the inflammatory burden on primary and secondary TTS.

A pro-inflammatory response is also known as a key pathogenetic mechanism of sepsis-induced cardiomyopathy. Indeed, pro-inflammatory cytokines, such as tumor necrosis factor-alpha (TNF-α), interleukin-1-beta (IL-1β) and chemokines activated by pathogen-associated molecular patterns (PAMPs) have been implicated in the pathogenesis of the myocardial dysfunction following sepsis, as well as endothelial dysfunction and impaired endothelium-derived NO release which can alter the physiological regulation of blood flow distribution. Of note, in sepsis-induced cardiomyopathy, the altered immune response to pathogens leads to a down-regulation of β-adrenergic receptors, and to the attenuation of the adrenergic response at the cardiomyocyte level, as the opposite of what has been demonstrated in TTS. However, if different biological responses to the same stressors may underly the different clinical features of secondary TTS, sepsis-induced cardiomyopathy, and patients with sepsis/septic shock without myocardial dysfunction have never been investigated so far.

Moreover, since the importance of acute stress mediators is clear, the role of exposure to chronic stress in the pathogenesis of TTS and whether exists a difference in primary vs. secondary TTS patients remain unexplored. In this regard, hair cortisol has recently emerged as a new biomarker of long-term HPA activity, providing an alternative method for assessing the degree of psychosocial stress exposure months prior to an acute stressful event. Indeed, it has been shown that cortisol, like other serum lipophilic components, can cross the capillaries that nourish the follicle by diffusion and be incorporated into the growing hair proportionally to its circulating concentration. Indeed, since the hair grows at a speed of about 1 cm/month, the hair sampling at 6 cm from the scalp may allow to estimating the plasmatic cortisol levels up to six months earlier. Therefore, the invastigators aim at investigating if a different chronic exposure to stress, evaluated through hair cortisol analysis, may play a role in the pathogenesis of primary vs. secondary TTS.

Rationale Primary and secondary TTS showed an important clinical heterogeneity identifying two different subtypes of patients with different outcomes and risk profiles. the invastigators hypothesize that a different activation of the brain structures involved in acute stress response, as well as a different exposure to chronic stress, may subtend the different clinical and risk profiles observed in primary vs. secondary TTS patients. Moreover, the invastigators hypothesize that distinct signatures of circulating biomarkers may be associated with these two categories of TTS patients. Therefore, identifying these specific signatures may help in the diagnosis of these patients and pave the way for the identification of specific pathophysiologic pathways and the development of future therapies. In addition, as primary vs. secondary TTS showed different clinical outcomes, the invastigators hypothesize that the identification of such specific signatures may help in the prognostic stratification of TTS patients.

As well, the invastigators want to assess the different clinical features of patients with secondary TTS, sepsis-induced cardiomyopathy, and sepsis/septic shock without myocardial dysfunction, to see whether a different biological response to the same stressful event (infection) may underly different pathophysiological mechanisms.

Thus, our hypotheses are the following:

• A different activation of brain structures involved in acute stress response may subtend the different clinical and outcomes profiles observed in primary vs. secondary TTS.
• A different exposure to chronic stress, evaluated through hair cortisol assay, may predispose to either primary or secondary TTS.
• Specific plasma circulating biomarkers could discriminate between primary vs. secondary TTS, helping to identify the underlying pathophysiological mechanisms.
• Specific plasma circulating biomarkers could discriminate between sepsis-induced cardiomyopathy, secondary TTS, and patients with sepsis/septic shock without myocardial dysfunction helping to identify the underlying pathophysiological mechanisms.

OBJECTIVES Primary Objective The primary objective of the study is to assess the activation of brain structures involved in acute stress response in primary vs. secondary TTS through Positron Emission Tomography (PET) analysis.

Secondary Objectives

1. To assess chronic stress exposure through hair cortisol levels in primary vs. secondary TTS.
2. To identify plasma circulating biomarkers in primary vs. secondary TTS both in the acute setup and at 3 months follow-up.
3. To identify if different plasma circulating biomarkers may be identified in sepsis/septic shock without cardiac dysfunction vs sepsis-induced cardiomyopathy;
4. To identify if different plasma circulating biomarkers may be identified in sepsis/septic shock without cardiac dysfunction vs. secondary TTS;
5. To identify if different plasma circulating biomarkers may be identified in sepsis-induced cardiomyopathy vs secondary TTS, underlying different pathophysiological mechanisms.

Study Design Single center prospective interventional study due to procedure (without drug nor device)

Study duration The study will last 42 months after the approval by the local Ethics Committe, of which the first 24 will be advocated to the enrolment phase, 12 to follow-up, and the following 6 to data analysis and scientific drafting.

Procedures The invastigators will evaluate demographic data (i.e., age, sex, race), classical cardiovascular risk factors, history of previous acute coronary syndromes, malignancy history, psychiatric history, and medical treatments.

At the time of coronary angiography for patients with primary or secondary TTS and within the first 48 hours for patients with sepsis or septic shock, arterial blood samples will be collected in 1 Vacuette® 9 mL CAT Serum Clot Activator tube and 4 Vacuette® 6 mL EDTA tubes. Furthermore, a hair sample of 6 cm will be collected for hair cortisol assay.

For patients with primary or secondary TTS blood samples will be collected by venipuncture with 1 Vacuette® 9 mL CAT Serum Clot Activator tube and 4 Vacuette® 6 mL EDTA tubes also at 3 months follow-up.

Blood samples will be immediately centrifuged to obtain whole blood, serum, and plasma samples, and then aliquoted into Eppendorf-type tubes. All samples will then be stored at -80°C until the analysis.

Circulating biomarkers levels of pro and anti-inflammatory pathways (IL-6, IL-1beta, IL-10, IL-18) will be measured using commercially available quantitative Enzyme-Linked ImmunoSorbent Assay (ELISA). The invastigators will also evaluate Endothelin-1 and BDNF levels and finally, the invastigators will assess catecholamine and relative receptors expression through the ELISA test.

Follow up visit Participation in this study requires for patients with primary vs secondary TTS, a follow-up visit at 3 months (90 +/- 5 gg) after enrollment to assess a Positron Emission Tomography (PET) analysis, and during the same visit, venous blood sampling as described above will be further performed.

Positron Emission Tomography (PET) analysis 18F-FDG-PET/CT imaging will be performed 3 months after the acute event at the Department of Nuclear Medicine of Fondazione Policlinico Universitario A. Gemelli IRCCS using an integrated scanner (e.g Biograph 64 Siemens Healthcare, Erlangen, Germany, or similar). Intravenous 18F-FDG (370 MBq) will be given following an overnight fast. Three-dimensional PET imaging will be performed after 1 h of quiet waiting. A non-gated, non-contrast CT will be acquired for attenuation correction. Brain structures analyzed will include the neocortex, limbic system (insula, amygdala, cingulate cortex, and hippocampus), reticular formation, brainstem, and spinal cord.

Analysis of stress-associated neural activity will be performed in a retrospective analysis using validated research methods by investigators blinded to clinical data. Amygdalar 18F-FDG uptake will be measured by placing circular regions of interest (ROIs) over the right and left amygdalae and measuring tracer accumulation as a standardized uptake value (SUV).

The bilateral amygdalar mean SUVs will be assessed both as an average and individually by dividing by regulatory regional neural uptake to generate a ratio of amygdalar to regulatory activity (20). The primary measure for amygdalar activity (AmygA) will be derived by averaging the SUVs of the right and left amygdalae, and dividing that value by the mean temporal lobe (TL) SUV (20). Additionally, because the ventromedial prefrontal cortex (vmPFC) serves as a regulatory centre under stressful conditions, AmygA will be also adjusted for mean vmPFC SUV in a secondary analysis.

Sample size calculation To the best of our knowledge, there is no previous literature comparing primary vs. secondary TTS. Hence, this configures as a pilot study. As such, no formal sample size calculation is needed. Based on common rules of thumb for pilot studies by Julious (2005), who refers as 12 subjects per group, and on the annual number of TTS patients observed at our Unit, the invastigators plan to enroll 60 patients in 24 months, of which 15 with primary TTS and 15 with secondary TTS and, as well, 15 with sepsis/septic shock and 15 with sepsis-induced cardiomyopathy.

Statistical analysis The sample will be described in its demographic, anthropometric, clinical, instrumental, variables through descriptive statistical techniques. Qualitative variables will be expressed as absolute and relative percentage frequency, while quantitative variables as either mean and standard deviation (SD) or median and interquartile range (IQR), as appropriate. Gaussian distribution will be previously assessed by the Shapiro-Wilk test. In the case of missing data, these will be treated by data imputation methods through the imputeR package, using multiple imputation with mean-centered Lasso Regression methods in case of quantitative data, while classification trees for imputation by function "RpartC", centered on fashion, that is the most represented class object, on qualitative data.

Comparisons between groups at baseline will be evaluated by Chi-square test or Fisher's exact test, with Freeman-Halton's extension, where appropriate, in case of qualitative data, while comparisons between quantitative data will be performed by one-way ANOVA or Kruskal-Wallis test. Pairwise comparisons will be computed by either the Student's t-test or Mann Whitney's U test. In this latter case, by using "coin" R package, Wilcoxon-Pratt signed rank test will be applied to deal with the presence of ties.

The evaluation of the primary endpoint, in terms of SUV values differences between primary and secondary TTS will be assessed by either the Student's t-test or Mann Whitney's U test.

Among the secondary endpoints, the hair cortisol and biomarkers difference at 3 months as compared to baseline will be primarily assessed in the individual groups using paired Student's t test or Wilcoxon rank sum test for paired data.

Between- and within-groups differences over time and at baseline in biomarkers and cortisol levels will be further represented by "violin plots", by using R packages "ggpubr", "ggstatsplot", "ggplot2", "ggprism" and "ggsignif".

The primary endpoint, as well as the secondary endpoints aimed to compare the variation in the aforementioned indexes between the two groups will be performed by linear mixed effects models for repeated measures (LMMRMs), with "glmmTMB" R package, including outcome data in long-format as dependent variable. Fixed effects will include subsample allocation, time as qualitative variable and their interaction, whilst random effects patients' ID. In depth, correlations between repeated measures will be modelled using either a compound symmetry structure or an unstructured covariance matrix, otherwise. The best fitting for the models will be tested by "DHARMa" R package, which uses a simulation-based approach to create readily interpretable scaled (quantile) residuals for fitted LMMs and provides test for model over/underdispersion. The resulting residuals are standardized to values between 0 and 1 and can be interpreted as intuitively as residuals from a linear regression.

In order to use these models on small samples, a Bartlett-like adjustment will be made by bootstrap resampling with the "pbkrtest" R package. In fact, taking into account that the approximation of the maximum likelihood (LR) T test statistic to the chi-square distribution in large samples is not accurate on a low scale, Bartlett's correction divides by a correction term, with the term of adjustment which is an estimate of the first moment of the null distribution of the T-statistic.

P-values <0.05 will be considered statistically significant. Suggestive p values will also be reported (0.05 ≤ p < 0.10). All analyses will be conducted with R software, version 4.2.0 (CRAN ®, R Core 2022, Vienna, Austria).