Machine Learning Approaches for Phenotyping in Cardiogenic Shock and Critical Illness

Jentzer, Jacob C.; Rayfield, Corbin; Soussi, Sabri; Berg, David D.; Kennedy, Jason; Sinha, Shashank S.; Baran, David A.; Brant, Emily; Mebazaa, Alexandre; Billia, Filio; Kapur, Navin K.; Henry, Timothy D.; Lawler, Patrick R.

doi:10.1016/j.jacadv.2022.100126

Cited by 25 publications

(24 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In CS, the selection of patients for trial enrollment could be based on readily available clinical and biological data [20,45 ▪▪ ] or host response omics-based biomarkers (e.g., inflammation, endothelial dysfunction) using a phenotyping approach [46]. The reanalysis of major ‘neutral’ clinical trials in CS from the perspective of biological signatures to identify distinct ‘hidden’ subphenotypes using unsupervised machine learning (i.e., agnostic to outcome) might help to inform targeted new therapies in future clinical trials [47 ▪ ,48 ▪▪ ]. This may not only allow an improved signal-to-noise ratio, but also the discovery of novel targets of pharmacological therapy (i.e., actionable biomarkers).…”

Section: Future Direction For Clinical Trials Optimization: Predictiv...mentioning

confidence: 99%

Cardiogenic shock: a major challenge for the clinical trialist

Sarma

Jentzer

Soussi

2023

Current Opinion in Critical Care

View full text Add to dashboard Cite

Purpose of review Cardiogenic shock (CS) results in persistently high short-term mortality and a lack of evidence-based therapies. Several trials of novel interventions have failed to show an improvement in clinical outcomes despite promising preclinical and physiologic principles. In this review, we highlight the challenges of CS trials and provide suggestions for the optimization and harmonization of their design. Recent findings CS clinical trials have been plagued by slow or incomplete enrolment, heterogeneous or nonrepresentative patient cohorts, and neutral results. To achieve meaningful, practice-changing results in CS clinical trials, an accurate CS definition, a pragmatic staging of its severity for appropriate patient selection, an improvement in informed consent process, and the use of patient-centered outcomes are required. Future optimizations include the use of predictive enrichment using host response biomarkers to unravel the biological heterogeneity of the CS syndrome and identify subphenotypes most likely to benefit from individualized treatment to allow a personalized medicine approach. Summary Accurate characterization of CS severity and its pathophysiology are crucial to unravel heterogeneity and identify the patients most likely to benefit from a tested treatment. Implementation of biomarker-stratified adaptive clinical trial designs (i.e., biomarker or subphenotype-based therapy) might provide important insights into treatment effects.

show abstract

Section: Future Direction For Clinical Trials Optimization: Predictiv...mentioning

confidence: 99%

Cardiogenic shock: a major challenge for the clinical trialist

Sarma

Jentzer

Soussi

2023

Current Opinion in Critical Care

View full text Add to dashboard Cite

show abstract

“…Numerous RCTs in patients with CS are currently ongoing and will expand the evidence base ( Supplemental Table 2 , http://links.lww.com/CCM/H343), including several that aim to compare advanced temporary MCS devices (e.g., Impella and VA-ECMO) versus optimal medical therapy whose proposed sample sizes exceed the sum total of patients enrolled in published RCTs utilizing these devices (74, 75). Integration of phenotyping and shock severity classification is essential for future research to encapsulate the heterogeneity of CS populations and understand pathophysiologic mechanisms which could yield novel treatment strategies in specific subgroups (21, 33). Implementing a more consistent standard of care including a multidisciplinary shock team with an evidence-based CS protocol has the potential to improve the generalizability of RCTs into clinical practice but requires additional rigorous investigation to establish the efficacy and safety of each proposed element (70).…”

Section: Future Cs Researchmentioning

confidence: 99%

“…Components to be considered when evaluating the severity and phenotype of patients with cardiogenic shock. Markers of shock severity ( bottom ) are distinct from the assessment of shock phenotype ( top ) (14, 21, 33). MCS = mechanical circulatory support, SCAI = Society for Cardiovascular Angiography and Intervention.…”

Section: Prognostication In Csmentioning

confidence: 99%

“…As many as 30-50% of patients with CS have concomitant cardiac arrest (CA), which is associated with myocardial dysfunction, global ischemia-reperfusion injury, systemic inflammation, and higher mortality (5,22,(28)(29)(30). The end stage of CS characterized by severe lactic acidosis and multiple organ failure has been recently termed "hemometabolic" or "cardiometabolic" shock and is associated with RV congestion, greater shock severity, and worse outcomes (4,19,(31)(32)(33).…”

Section: Key Pointsmentioning

confidence: 99%

See 1 more Smart Citation

Advances in the Management of Cardiogenic Shock

Jentzer

Pöss

Schaubroeck

et al. 2023

Critical Care Medicine

Self Cite

View full text Add to dashboard Cite

OBJECTIVES: To review a contemporary approach to the management of patients with cardiogenic shock (CS). DATA SOURCES: We reviewed salient medical literature regarding CS. STUDY SELECTION: We included professional society scientific statements and clinical studies examining outcomes in patients with CS, with a focus on randomized clinical trials. DATA EXTRACTION: We extracted salient study results and scientific statement recommendations regarding the management of CS. DATA SYNTHESIS: Professional society recommendations were integrated with evaluated studies. CONCLUSIONS: CS results in short-term mortality exceeding 30% despite standard therapy. While acute myocardial infarction (AMI) has been the focus of most CS research, heart failure-related CS now predominates at many centers. CS can present with a wide spectrum of shock severity, including patients who are normotensive despite ongoing hypoperfusion. The Society for Cardiovascular Angiography and Intervention Shock Classification categorizes patients with or at risk of CS according to shock severity, which predicts mortality. The CS population includes a heterogeneous mix of phenotypes defined by ventricular function, hemodynamic profile, biomarkers, and other clinical variables. Integrating the shock severity and CS phenotype with nonmodifiable risk factors for mortality can guide clinical decision-making and prognostication. Identifying and treating the cause of CS is crucial for success, including early culprit vessel revascularization for AMI. Vasopressors and inotropes titrated to restore arterial pressure and perfusion are the cornerstone of initial medical therapy for CS. Temporary mechanical circulatory support (MCS) is indicated for appropriately selected patients as a bridge to recovery, decision, durable MCS, or heart transplant. Randomized controlled trials have not demonstrated better survival with the routine use of temporary MCS in patients with CS. Accordingly, a multidisciplinary team-based approach should be used to tailor the type of hemodynamic support to each individual CS patient’s needs based on shock severity, phenotype, and exit strategy.

show abstract

“…The current study aims were to characterize the CS survivors underlying heterogeneity using a clustering approach (i.e. phenotyping) 4–6 . Accordingly, we applied latent class analysis (LCA) in a population of CS survivors [French and European Outcome Registry in Intensive Care Units (FROG‐ICU) cohort] 7 to identify distinct clinical phenotypes at intensive care unit (ICU) discharge, used host‐response biomarkers to characterize them (i.e.…”

Section: Aimsmentioning

confidence: 99%

Clinical phenotypes of cardiogenic shock survivors: insights into late host responses and long‐term outcomes

Soussi,

Ahmadiankalati,

Jentzer

et al. 2023

ESC Heart Failure

Self Cite

View full text Add to dashboard Cite

AimsAn elevated risk of adverse events persists for years in cardiogenic shock (CS) survivors with high mortality rate and physical/mental disability. This study aims to link clinical CS‐survivor phenotypes with distinct late host‐response patterns at intensive care unit (ICU) discharge and long‐term outcomes using model‐based clustering.Methods and resultsIn the original prospective, observational, international French and European Outcome Registry in Intensive Care Units (FROG‐ICU) study, ICU patients with CS on admission were identified (N = 228). Among them, 173 were discharged alive from the ICU and included in the current study. Latent class analysis was applied to identify distinct CS‐survivor phenotypes at ICU discharge using 15 readily available clinical and laboratory variables. The primary endpoint was 1 year of mortality after ICU discharge. Secondary endpoints were readmission and physical/mental disability [short form‐36 questionnaire (SF‐36) score] within 1 year after ICU discharge. Two distinct phenotypes at ICU discharge were identified (A and B). Patients in Phenotype B (38%) were more anaemic and had higher circulating levels of lactate, sustained kidney injury, and persistent elevation in plasma markers of inflammation, myocardial fibrosis, and endothelial dysfunction compared with Phenotype A. They had also a higher rate of non‐ischaemic origin of CS and right ventricular dysfunction on admission. CS survivors in Phenotype B had higher 1 year of mortality compared with Phenotype A (P = 0.045, Kaplan–Meier analysis). When adjusted for traditional risk factors (i.e. age, severity of illness, and duration of ICU stay), Phenotype B was independently associated with 1 year of mortality [adjusted hazard ratio = 2.83 (95% confidence interval 1.21–6.60); P = 0.016]. There was a significantly lower physical quality of life in Phenotype B patients at 3 months (i.e. SF‐36 physical component score).ConclusionsA phenotype with sustained inflammation, myocardial fibrosis, and endothelial dysfunction at ICU discharge was identified from readily available data and was independently associated with poor long‐term outcomes in CS survivors.

show abstract

Machine Learning Approaches for Phenotyping in Cardiogenic Shock and Critical Illness

Cited by 25 publications

References 64 publications

Cardiogenic shock: a major challenge for the clinical trialist

Cardiogenic shock: a major challenge for the clinical trialist

Advances in the Management of Cardiogenic Shock

Clinical phenotypes of cardiogenic shock survivors: insights into late host responses and long‐term outcomes

Contact Info

Product

Resources

About