Introduction: Cardiogenic shock represents one of the cardiovascular emergencies with the highest in-hospital mortality, despite advances in therapeutic management and mechanical circulatory support. The pathophysiological heterogeneity characterizing this syndrome has highlighted the inadequacy of a unitary characterization, making it necessary to adopt phenotypic systems capable of capturing its complexity. Among these, the model proposed by Zweck et al (2021) identifies three distinct phenotypes based on hemodynamic and metabolic patterns derived from cluster analysis. However, the model’s dependence on the calculation of Euclidean distances from cluster centroids limits its applicability at the bedside, necessitating the development of more accessible phenotyping tools. Aim of the study: Building on this operational limitation, the present study aims to: 1) perform a multidimensional characterization of the study population; 2) apply Zweck et al.'s phenotyping system to an Italian cohort of patients with cardiogenic shock and 3) compare the clinical, hemodynamic, and biochemical characteristics of the phenotypes thus identified; 4) derive a bedside phenotypic classification based on readily accessible clinical criteria, and develop, on the basis of this classification, an early predictive score for identifying patients at higher risk of unfavourable clinical evolution. Materials and methods: A prospective, single-centre study was conducted on 25 patients with cardiogenic shock admitted to the Cardiac Intensive Care Unit of the Ca’ Foncello Hospital in Treviso. Patients underwent daily multimodal monitoring encompassing clinical, laboratory, echocardiographic and hemodynamic parameters until the point of recovery. The cohort was subsequently classified according to the system of Zweck et al., distinguishing a non-congested, a cardiorenal and a cardiometabolic phenotypes, which were then subjected to systematic clinical comparison. An operational bedside classification was then derived, based on indicators of clinical course (such as weaning from vasoactive drugs, use of intra-aortic balloon pump, and SOFA score) from which an early predictive score was subsequently developed using variables measurable at admission. Results: Application of the Zweck et al. system distributed the 25 patients across the three phenotypes as follows: 20% in the non-congested, 52% in the cardiorenal and 28% in the cardiometabolic phenotypes. Comparison between adjacent phenotypes revealed significant differences along the metabolic, renal, hepatic and hemodynamic axes, with progressive deterioration of tissue perfusion parameters and vasopressor requirements. The operational bedside classification confirmed the stratification into three levels of clinical severity. In-hospital mortality showed an increasing gradient: 0% in the low support phenotype, 12,5% in the effective-support phenotype and 83,3% in the support-refractory phenotype. The early predictive score demonstrated excellent discriminative characteristics for the early identification of the highest-complexity phenotype (AUC 0,974), and no patient with a score below the threshold progressed to the support refractory phenotype. Conclusions: The study confirmed the replicability of the Zweck phenotypic system in a small Italian cohort, documenting a gradient of clinical and prognostic severity consistent with the existing literature. The predictive score showed excellent discriminative characteristics for the early identification of the highest-risk phenotype at admission. Although derived from a small, single-centre sample, the results suggest the potential utility of integrative bedside scoring tools as a complement to the phenotypic classification currently in use. The findings lay the groundwork for external validation studies in larger populations which are indispensable for defining the role of the tools developed in this study in early risk stratification.
Introduzione: Lo shock cardiogeno è una delle emergenze cardiovascolari con più elevata mortalità ospedaliera, nonostante i progressi in ambito terapeutico e di supporto circolatorio meccanico. L’eterogeneità fisiopatologica che caratterizza questa sindrome ha reso evidente l’inadeguatezza di una sua caratterizzazione unitaria, rendendo necessaria l’adozione di sistemi fenotipici capaci di coglierne la complessità. Tra questi, il modello proposto da Zweck et al. (2021) identifica tre fenotipi distinti, sulla base di pattern emodinamici e metabolici. Tuttavia, la limitata applicabilità al letto del paziente di tale modello, rende necessario lo sviluppo di strumenti di fenotipizzazione più accessibili. Scopo dello studio: Partendo da questa limitazione operativa, il presente studio si propone di: 1) effettuare una caratterizzazione multidimensionale della popolazione in studio 2) applicare il sistema di fenotipizzazione di Zweck et al. a una coorte italiana di pazienti in shock cardiogeno e 3) confrontare le caratteristiche cliniche, emodinamiche e bioumorali dei fenotipi così identificati; 4) di derivare una classificazione fenotipica bedside fondata su criteri clinici facilmente accessibili e di sviluppare, a partire da tale classificazione, uno score predittivo precoce per l'identificazione di pazienti a maggiore rischio di evoluzione sfavorevole. Materiali e metodi: È stato condotto uno studio prospettico monocentrico su 25 pazienti con shock cardiogeno ricoverati presso l’UCIC dell’Ospedale Ca’ Foncello di Treviso. I pazienti sono stati sottoposti a monitoraggio multimodale giornaliero, comprendente parametri clinici, laboratoristici, ecocardiografici ed emodinamici, fino al momento di recovery o dell’exitus. La coorte è stata classificata secondo il sistema tripartito di Zweck et al., distinguendo un fenotipo non-congesto, uno cardiorenale e uno cardiometabolico, sottoposti a confronto clinico sistematico. È stata poi derivata una classificazione fenotipica bedside, fondata su indicatori di decorso clinico (come lo svezzamento dai farmaci, il Vasoactive-inotropic score, l’utilizzo di contropulsatore aortico e il SOFA score) dalla quale è stato successivamente sviluppato uno score predittivo precoce basato su variabili misurabili all’ammissione. Risultati: L'applicazione del sistema di Zweck et al. ha permesso di distribuire i 25 pazienti nei tre fenotipi nel modo seguente: 20% nel fenotipo non-congesto, 52% nel fenotipo cardiorenale e 28% nel fenotipo cardiometabolico. Il confronto tra fenotipi adiacenti ha evidenziato differenze significative lungo gli assi metabolico, renale, epatico ed emodinamico, con un progressivo deterioramento dei parametri di perfusione tissutale e del supporto vasopressorio. La classificazione fenotipica bedside ha confermato la stratificazione in tre livelli di severità clinica. La mortalità ospedaliera ha mostrato un gradiente crescente: 0% nel fenotipo a basso supporto, 12,5% nel fenotipo a supporto efficace e 83,3% nel fenotipo refrattario al supporto. Lo score predittivo precoce ha dimostrato ottime caratteristiche discriminative per l'identificazione precoce del fenotipo a maggiore complessità (AUC 0,974) e nessun paziente con punteggio inferiore alla soglia ha presentato un’evoluzione verso il fenotipo a prognosi peggiore. Conclusioni: Lo studio ha confermato la replicabilità del sistema fenotipico di Zweck in una coorte di dimensioni limitate, documentando un gradiente di severità clinica e prognostica coerente con la letteratura. Lo score predittivo ha mostrato caratteristiche discriminative eccellenti per l'identificazione precoce del fenotipo a maggior rischio all'ammissione. I risultati suggeriscono la potenziale utilità di strumenti di scoring bedside integrativi rispetto alla classificazione fenotipica attualmente utilizzata e pongono le basi per studi di validazione esterna su popolazioni più ampie.
Nuove prospettive nella fenotipizzazione dello shock cardiogeno: caratterizzazione dei profili fenotipici e analisi dei fattori condizionanti la prognosi
GALBIATI, SOFIA
2025/2026
Abstract
Introduction: Cardiogenic shock represents one of the cardiovascular emergencies with the highest in-hospital mortality, despite advances in therapeutic management and mechanical circulatory support. The pathophysiological heterogeneity characterizing this syndrome has highlighted the inadequacy of a unitary characterization, making it necessary to adopt phenotypic systems capable of capturing its complexity. Among these, the model proposed by Zweck et al (2021) identifies three distinct phenotypes based on hemodynamic and metabolic patterns derived from cluster analysis. However, the model’s dependence on the calculation of Euclidean distances from cluster centroids limits its applicability at the bedside, necessitating the development of more accessible phenotyping tools. Aim of the study: Building on this operational limitation, the present study aims to: 1) perform a multidimensional characterization of the study population; 2) apply Zweck et al.'s phenotyping system to an Italian cohort of patients with cardiogenic shock and 3) compare the clinical, hemodynamic, and biochemical characteristics of the phenotypes thus identified; 4) derive a bedside phenotypic classification based on readily accessible clinical criteria, and develop, on the basis of this classification, an early predictive score for identifying patients at higher risk of unfavourable clinical evolution. Materials and methods: A prospective, single-centre study was conducted on 25 patients with cardiogenic shock admitted to the Cardiac Intensive Care Unit of the Ca’ Foncello Hospital in Treviso. Patients underwent daily multimodal monitoring encompassing clinical, laboratory, echocardiographic and hemodynamic parameters until the point of recovery. The cohort was subsequently classified according to the system of Zweck et al., distinguishing a non-congested, a cardiorenal and a cardiometabolic phenotypes, which were then subjected to systematic clinical comparison. An operational bedside classification was then derived, based on indicators of clinical course (such as weaning from vasoactive drugs, use of intra-aortic balloon pump, and SOFA score) from which an early predictive score was subsequently developed using variables measurable at admission. Results: Application of the Zweck et al. system distributed the 25 patients across the three phenotypes as follows: 20% in the non-congested, 52% in the cardiorenal and 28% in the cardiometabolic phenotypes. Comparison between adjacent phenotypes revealed significant differences along the metabolic, renal, hepatic and hemodynamic axes, with progressive deterioration of tissue perfusion parameters and vasopressor requirements. The operational bedside classification confirmed the stratification into three levels of clinical severity. In-hospital mortality showed an increasing gradient: 0% in the low support phenotype, 12,5% in the effective-support phenotype and 83,3% in the support-refractory phenotype. The early predictive score demonstrated excellent discriminative characteristics for the early identification of the highest-complexity phenotype (AUC 0,974), and no patient with a score below the threshold progressed to the support refractory phenotype. Conclusions: The study confirmed the replicability of the Zweck phenotypic system in a small Italian cohort, documenting a gradient of clinical and prognostic severity consistent with the existing literature. The predictive score showed excellent discriminative characteristics for the early identification of the highest-risk phenotype at admission. Although derived from a small, single-centre sample, the results suggest the potential utility of integrative bedside scoring tools as a complement to the phenotypic classification currently in use. The findings lay the groundwork for external validation studies in larger populations which are indispensable for defining the role of the tools developed in this study in early risk stratification.| File | Dimensione | Formato | |
|---|---|---|---|
|
Galbiati_Sofia.pdf
Accesso riservato
Dimensione
5.03 MB
Formato
Adobe PDF
|
5.03 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/109109