br Sepsis induced cardiomyopathy SIC the clinical
Sepsis-induced cardiomyopathy (SIC): the clinical picture The occurrence of cardiovascular abnormalities during sepsis has been recognised for over 50 years [4,5]. However, an intrinsic myocardial dysfunction in patients with septic shock was first described in 1984 by Parker and colleagues, who reported an increase in ventricular volumes and a decreased ejection fraction . SIC has generally been defined as an intrinsic and reversible systolic and diastolic dysfunction of both the left and right sides of the 3X FLAG Peptide induced by sepsis . However, there is no consensus on specific diagnostic criteria , and different studies have used markedly different criteria to identify SIC. This variability has resulted in an apparently shifting prevalence of SIC, ranging from 24 to 72% [, , , , ]. It is however clear that the prevalence of SIC rises with increasing disease severity . These early studies suggested a protective effect of ventricular dilation . However, more recent studies using similar loading-dependent variables have failed to identify a positive effect on outcome [10,15,16]. On the contrary, utilising more sensitive markers of cardiac dysfunction such as tissue Doppler, mitral flow Doppler and speckle tracking that are less dependent on loading conditions, cardiac dysfunction appears to be strongly associated with a worse outcome [9,, , , , ]. These differences underline the complexity of the disease, the heterogeneity of patient cohorts, and the sometimes conflicting methodologies used for SIC diagnosis. Part of this complexity stems from the difficulty in differentiating an intrinsic cardiac dysfunction from abnormalities in vascular and autonomic status. An important characteristic of SIC is its potential reversibility observed in numerous studies [6,11,22,23]. However, this concept of reversibility has not been tested in large outcome studies with robust methodologies, so the evidence is currently inconclusive . While cardiac function has been extensively studied in human sepsis, much less data are available on concurrent structural cardiac damage . Both macroscopic and microscopic findings of myocarditis have been noted at post-mortem [, , , ] while evidence of non-ischemic cardiac injury compatible with inflammation or tissue acidosis was observed in vivo using cardiac magnetic resonance . The cardiomyocytes of septic patients showed scattered foci of disruption of the contractile apparatus and translocation of connexin-43, an indication of cell injury [28,31]. Of note, only minimal signs of cardiomyocyte apoptosis or necrosis were seen, suggesting that cell death does not account for the severity of SIC in clinical patients [31,32]. Low levels of cardiomyocyte apoptosis and necrosis were also confirmed in large animal experimental studies [33,34]. Cardiomyocyte death has been noted in some rodent models of sepsis [, , ] but this may reflect the severity and acuity of the model, variations in resuscitation, and species differences. Despite the frequent finding of minimal cardiomyocyte death, a correlation has been found in septic patients between a rise in cardiac troponins, a circulating biomarker of cell injury, and both mortality [, , ] and the degree of myocardial dysfunction . This apparent paradox can be explained by the non-necrotic release of troponins [41,42], further supporting the concept of reversible intrinsic myocardial damage.
Preclinical studies Given these limited clinical studies, most evidence for the pathophysiology underlying SIC has originated from preclinical models. Ex vivo models include the study of the isolated whole heart (i.e., Langendorff model), papillary muscles, permeabilised muscle fibres and isolated cardiomyocytes or mitochondria . In vivo models have utilised a variety of species, insult types and severity, study duration, and degree of supportive care provided. Insult types used to model sepsis in animals include injection of pathogen components (e.g. endotoxins or zymosan), administration of live bacteria (e.g., intravenous, intraperitoneal or intratracheal injection of bacteria, intraperitoneal inoculation with fecal slurry, implantation of bacterial and fibrin clots), or disruption of the host barriers resulting in polymicrobial sepsis (e.g., caecal ligation and puncture, CLP, or colon ascendens stent peritonitis, CASP). Each model presents unique features that attempt to recapitulate specific aspects of clinical sepsis in humans. The details, and limitations, of the different approaches have been described in numerous reviews [, , ] and minimum quality thresholds to move preclinical research forward have been recently proposed . Of note, a large proportion of the pre-clinical sepsis literature is based on endotoxin models. The administration of endotoxin produces a reproducible, rapid and robust activation of the innate immune system. However, endotoxin fails to recapitulate the complexity of sepsis pathophysiology and numerous differences are present between the sepsis and the endotoxin phenotypes. For these reasons, current guidelines discourage extensive use of models based on endotoxin administration .