Initial haemodynamic assessment of cardiogenic shock: back to basics?

This editorial refers to ‘Prognostic Significance of Hemodynamic Parameters in Patients with Cardiogenic Shock’, by D. D. Berg et al. https://doi.org/10.1093/ehjacc/zuad095. Despite tremendous advancements in vascular medicine and critical care over the last decades, cardiogenic shock remains a dreadful medical condition with a dire prognosis, killing up to 40–50% of patients affected.1 After a string of trials over the last decades investigating various pharmacologic and device treatments yielded neutral results, risk stratification and individualized treatment using a team-based approach have regained focus in the community. Several prediction models based on routine and experimental clinical and laboratory findings have been proposed.2 While some of these prediction models have been validated in various cardiogenic shock populations, they lack sufficient robustness to guide therapeutic decisions and may be too complex or time-consuming for daily use, resulting in low application rates in clinical practice. The Swan-Ganz catheter, from here on referred to as pulmonary artery catheter (PAC), allows direct measurement of both pulmonary and cardiac pressures as well as oxygen saturation, which is used to calculate several haemodynamic parameters, including cardiac output and vascular resistances. It therefore allows differential diagnosis in cardiogenic shock and early recognition of patients in potential need of escalation to mechanical circulatory support. In the late 1990s and early 2000s however, several randomized controlled trials suggested that PAC use does not result in a mortality benefit and may even increase complication rates.3 In return, PAC application rates declined. In recent years, these trends reversed4 and some expert groups have proposed routine use of PACs in cardiogenic shock patients. Indeed, several contemporary, retrospective registry-based analyses suggested improved mortality rates associated with PAC use in this patient group.5 Based on existing evidence, current clinical practice guidelines and consensus documents on the management of cardiogenic shock on both sides of the Atlantic therefore recommend a selected use of PAC in patients not responding to initial pharmacological interventions, in cases of diagnostic uncertainty or planned escalation to mechanical circulatory support systems.6–9 In this issue of the European Heart Journal: Acute Cardiovascular Care, David Berg and colleagues present a detailed, contemporary investigation into the association between invasively measured haemodynamic parameters at cardiovascular intensive care unit (CICU) admission and in-hospital mortality in nearly 1500 patients with cardiogenic shock within the renowned Critical Care Cardiology Trial’s Network (CCCTN) registry.10 Amongst the predictors of in-hospital mortality in the unadjusted analysis were a low mean arterial pressure (MAP), a low systolic blood pressure, a low systemic vascular resistance, an elevated right atrial pressure (RAP), elevated RAP/pulmonary capillary wedge pressure (PCWP), and a low pulmonary artery pulsatility index. Ultimately, only MAP as a measure of systemic arterial tone and RAP as an index of right ventricular function remained statistically significant predictors of mortality in the multiparameter model. Conversely, and somewhat surprising, neither cardiac index nor cardiac power output (CPO) or PCWP was associated with in-hospital mortality in unadjusted or adjusted analyses. Findings were consistent when adjusting for background catecholamine support, across causes of cardiogenic shock as well as in those with and without mechanical circulatory support at the time of invasive haemodynamic assessment. The authors need to be congratulated for providing such a detailed and large-scale analysis of the prognostic significance of invasively measured haemodynamic parameters using a PAC in a contemporary cohort of cardiogenic shock patients across various geographic regions in North America. Their analysis is amongst the largest systematic analyses of invasively measured haemodynamic parameters ever conducted. They provide nearly complete haemodynamics and co-therapies in close to 1500 cardiogenic shock patients across the disease spectrum. On average, invasive haemodynamics were recorded 1 h after CICU admission, underscoring the high quality of participating centres. The findings of Berg et al. are remarkable and somewhat surprising and contradicting prior literature. Analyses from the SHOCK trial registry and the CardShock study suggested that indices associated with cardiac output (CO), including CPO and cardiac power index (CPI), are the strongest predictors of mortality in cardiogenic shock patients.11,12 Why were indices of cardiac output not predictive of mortality in the CCCTN analysis? It is important to note that such a lack of prognostic value of CPO is not new, data from another contemporary analysis recently yielded similar results.5 Herein, the majority of patients suffered from cardiogenic shock not due to acute myocardial infarction, where historical data on CPO are scarce. Furthermore, in the CardShock study, haemodynamic parameters assessed for outcome prediction were median values from the first 4 days as opposed to very early values obtained in the CCCTN population. In the SHOCK analysis, PAC measurements were undertaken within the first 12 h but the overall population with CPO measurements available was small (n = 189). The authors state that in a vast majority of patients in the CCCTN registry, cardiac output was estimated using the Fick principle with the remaining ones measured by thermodilution, two measurements that have been shown to only modestly agree.13 Additional aspects include the inclusion of nearly 25% of patients after cardiac arrest, a condition associated with decreased systemic oxygen demand making CO estimates based on Fick less reliable.14 Another potential aspect is that in contemporary practice, patients with signs of hypoperfusion and low output may rapidly be treated with catecholamines and/or mechanical circulatory support systems, rendering indices based on cardiac output not prognostic. In this regard, it is of great interest that MAP was the strongest predictor of mortality and hypoperfusion, given that most clinicians will target a MAP of at least 65 mmHg in the early phase after CICU admission thus confounding MAP values. As discussed by the authors, the results suggest that not only does the intensity of vasopressor support needed to achieve a satisfying MAP matter but also the achieved MAP. Whether achievement of a higher MAP based on vasopressor use results in improved outcomes is not known but is currently being tested by a dedicated randomized clinical trial (NCT05168462). One interpretation of the current analysis is that conventional haemodynamic monitoring using an arterial line and a central venous line is sufficient for initial risk stratification and therapeutic guidance, at least in the first hours after CICU admission as also recommended by recent guidelines and position statements. Such monitoring provides the treating clinician with MAP as parameter of systemic arterial tone as well as RAP as a measure of right ventricular function. A limitation of the herein reported CCCTN registry data is that it does not include information on changes of haemodynamic parameters over time thus leaving the value of longitudinally collected invasive haemodynamic parameters up for speculation. The CCCTN comprises a network of advanced CICU’s in North America that provides granular data on consecutive medical admissions. As previously reported and in line with the results from other larger registry-based analyses, a wide variation in PAC use patterns has been described even amongst CCCTN participating hospitals, at least in part driven by institutional tendency.15 This underscores the need for high quality evidence regarding the use of PAC from adequately powered randomized controlled trials in patients with cardiogenic shock according to modern diagnostic criteria. The Pulmonary Artery Catheter in Cardiogenic Shock Trial (PACCS) is an ongoing randomized controlled trial to assess the effects of a routine early invasive haemodynamic assessment on in-hospital mortality in ∼400 patients with cardiogenic shock due to acutely decompensated heart failure (NCT05485376; clinicaltrials.gov). Patients will be randomized to either PAC placement within 6 h of randomization and 24 h of presentation or no or delayed PAC 48 h after randomization. The PACCS trial will provide much anticipated randomized trial data on the benefits and harms of routine early PAC use in a contemporary cardiogenic shock population. Until publication of these results, patients should continue to be included in trials and registries and a conservative approach upon admission using conventional haemodynamic monitoring providing information on systemic arterial tone and impaired right ventricular function may be sufficient for initial risk stratification. None declared. No new data were generated or analysed in support of this research.