Science pulse: management of cardiogenic shock and trial design: time for a paradigm shift! Insights from the Critical Care Clinical Trialists meeting

Cardiogenic shock (CS) is still an under-recognized disease with high mortality, up to 40–50% at 30 days, depending on several risk factors and phenotype and aetiology of CS as well as on centres and countries. Despite all attempts to perform high-quality randomized controlled trials (RCTs) in CS, most of them are neutral. As a result, CS recommendations primarily rely on expert opinion and international standardized procedures to treat CS patients are still lacking. Between 13 and 15 June 2024, the 6th Critical Care Clinical Trialists (3CT) workshop took place in Washington DC (USA). A multidisciplinary, international group of experts (clinical trialists, clinicians of different specialties, patient representatives, nurses, regulators, and industry representatives) met to discuss recent advances in CS research and to explore future approaches to optimize trial design (Figure 1). Together with three participants, we summarized three main sessions of this meeting. 3CT Meeting 2024, Washington, DC. In this session, major insights were drawn from the Extracorporeal Life Support (ECLS)-Shock trial, which was pragmatically designed to have a large generalizability therefore including post-cardiac arrest patients. Recruitment was completed within 3.5 years. This RCT indicated that early routine implementation of ECLS for patients with CS due to acute myocardial infarction (AMI) may not be beneficial. Instead, early initiation of ECLS should be reserved for patients where the benefits clearly outweigh potential harms, necessitating further studies to identify this subgroup of patients. The DAN-GER Shock RCT studied the use of percutaneous left ventricular assist device (LVAD) (Impella©) in patients with predominantly left ventricular failure and ST-elevation myocardial infarction–related CS. Despite slow enrolment over 10 years, the strict inclusion and exclusion criteria, particularly the exclusion of comatose patients following out-of-hospital cardiac arrest, may have contributed to its success. The trial showed an improvement in mortality at 180 days, illustrating the impact of precise endpoint selection in CS trials. High prevalence of acute kidney injury requiring renal replacement therapy was discussed. Importantly, the trial revealed significant sex differences, showing that women experienced worse outcomes and greater delays in the initiation of care. This underscores the importance of considering disparities, especially sex, race, and ethnicity differences in CS management and research. In addition to RCTs, observational studies and registries are valuable to provide information about the heterogeneity of care practices and their impact on patient outcomes. Addressing this variability is crucial for the development of more effective treatment protocols. Finally, shifting from CS as a singular syndrome to a condition with distinct phenotypes and endotypes represents a major change in understanding and treatment. This approach will allow for prognostic and predictive enrichment in clinical trials, tailoring therapies based on individual patient characteristics and thereby enhancing the precision and effectiveness of interventions in the future. Inflammation was highlighted as a driver of outcome in CS, moving away from the ‘cardiocentric’ approach in CS. The variability in the pathobiology of CS contributing to the clinical manifestation of shock and organ dysfunction, which include metabolic dysfunction, alteration of the renin–angiotensin–aldosterone system pathway, endothelial dysfunction, and hyper-inflammation, was emphasized. Omics-based biomarkers reflecting underlying mechanistic signatures [e.g. bio-adrenomedullin, DPP3, and interleukin (IL)-6] as well as genotyping may lead to a new approach in characterizing and classifying CS patients. The session included evidence from the re-analysis of completed observational studies and RCT to identify inflammatory/host response biomarker–driven sub-phenotypes using unsupervised machine learning [Shock CO-OP research programme (NCT06376318)]. The main aim of this international research programme is to allow prognostic enrichment and support future RCTs of personalized interventions (i.e. predictive enrichment: identifying a molecular subclass of CS patients with the highest probability of responding to a specific treatment). Insights were also provided about the use of DPP3 and IL-6 not only as biomarkers to improve risk stratification but also as biological targets of therapy (e.g. antibodies) in medical and post-cardiac surgery CS. Similarly, Dr Helle Søholm (Copenhagen DK) presented the design of the DOBERMANN trial (NCT05350592), with AMI patients at intermediate/high risk of CS development (CS SCAI B) to be randomized (2 × 2) to receive dobutamine infusion and/or a single dose of tocilizumab (IL-6 receptor antagonist) vs. placebo. In this trial, NT-pro-BNP will be used as a surrogate endpoint for development of CS. In this session, several methodological aspects that may prevent neutral results in CS trials were discussed. The selection of the optimal study population should maximize the proportion of patients likely to benefit from the investigated therapeutic intervention. The choice of clinically meaningful endpoints and the optimal length of follow-up to capture the efficacy of an intervention are crucial. The advantages and downsides of novel statistical methods such as the ‘win ratio’ were elaborated. This method uses hierarchical—preferably continuous—endpoints as alternative to the traditional studies and gives appropriate priority to the more clinically important events while also increasing statistical power of the trial. Study participants are compared in a pairwise fashion, which result in wins, losses, or ties. By dividing the total number of wins by the total numbers of loss, the win ratio describes the estimated probability that an individual will have more favourable outcomes. Additional hurdles inherent to strategy trials, especially when devices are studied, were discussed. Double-blinded or placebo-controlled designs are often impossible, and randomization by patients is unlikely to unravel relevant differences in management. Alternative approaches have been highlighted, such as clustering by centres. In such strategy studies, the control arm should be standardized as much as possible. The study protocol should describe the measures in the control arm in detail, including possible escalation strategies in case of treatment failure. This should minimize bias derived from heterogeneity in the ancillary treatments and reduce crossovers. The importance of adequate post-intervention management was also emphasized. Finally, special considerations for dealing with CS patients presenting after out-of-hospital cardiac arrest were mentioned, since this condition confers a doubled risk of death, and many deaths are caused by anoxic brain injury, complicating the interpretation of CS trials. During the meeting, all parties agreed on the unmet need for unequivocable—and optimally positive—results from high-quality studies in CS. To reach this ambitious goal, a thorough and comprehensive re-evaluation of current CS trial methodologies and clinical practices is needed, advocating for a more personalized, collaborative, and standardized approach to improve patient outcomes in CS. Global networking and collaboration in CS research can facilitate the sharing of knowledge and resources, leading to more robust and generalizable findings. Shifting from CS as a singular syndrome to a condition with distinct phenotypes and endotypes represents a major change in understanding and will have major impact on the treatment. Furthermore, this approach will allow for prognostic and predictive enrichment in clinical trials, tailoring therapies based on individual patient characteristics and thereby enhancing the precision and effectiveness of interventions in the future. Therefore, it is indeed time for a paradigm shift, and we can only achieve this together. Janine Pöss Hannah Schaubroeck Vanessa Blumer Sabri Soussi Mattia Arrigo Alexandre Mebazaa

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.