Microaxial Flow Pump Use in Different Phenotypes of Cardiogenic Shock—a Secondary Analysis of the DanGer Shock Trial

BACKGROUND: Three cardiogenic shock (CS) phenotypes have been proposed and validated in various datasets: noncongested (phenotype I), cardiorenal (phenotype II), and cardiometabolic CS (phenotype III). The DanGer Shock trial demonstrated a mortality benefit of microaxial flow pump (mAFP) use in myocardial infarction-related CS. In this post-hoc analysis, we aimed to assess the trajectories and outcomes of these phenotypes in the DanGer Shock population. METHODS: Patients randomized to the DanGer Shock trial were retrospectively assigned to 1 of 3 CS phenotypes at admission. Missing values for phenotyping were imputed using multiple random forest imputation. Outcomes were 180-day mortality and the trajectories of key clinical, laboratory and hemodynamic parameters first 72 hours within phenotypes, stratified by allocation to mAFP or standard of care. RESULTS: Of 355 adult patients in the trial, 145 (41%), 38 (11%), and 172 (48%) patients were in the noncongested, cardiorenal, and cardiometabolic phenotypes, respectively. The 180-day mortality was greater in cardiometabolic (69%) compared with the noncongested (33%) and cardiorenal CS (47%) groups. Clinical metabolic and hemodynamic trajectories and their treatment response differed between phenotypes. mAFP use was associated with lower mortality in noncongested CS (odds ratio, 0.51 [0.28-0.91], P = .02). The odds of mortality were 0.81 [0.57-1.16] (P = .25) in cardiometabolic and 0.91 [0.35-2.34] (P = .84) in cardiorenal CS (P for interaction: .43). CONCLUSION: In this post-hoc analysis of the DanGer Shock trial, predefined CS phenotypes showed distinct outcomes, with the noncongested phenotype faring best and the cardiometabolic the worst. The greatest apparent benefit of mAFP was observed in noncongested CS. These findings are hypothesis-generating and warrant confirmation in prospective studies. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov unique identifier: NCT01633502. Condensed Abstract In this DanGer Shock post-hoc analysis, study participants were retrospectively assigned to 1 of 3 cardiogenic shock (CS) phenotypes: noncongested, cardiorenal, or cardiometabolic CS. In total, 41%, 11%, and 48% of all 355 patients were in these phenotypes, respectively. The 180-day mortality was greater in the cardiometabolic (69%) compared with the noncongested (33%) and cardiorenal CS (47%) groups. mAFP use was associated with lower mortality in the noncongested and numerically in the cardiometabolic, but not in cardiorenal CS group. In summary, CS phenotypes showed distinct outcomes, with the noncongested phenotype faring best and the cardiometabolic faring the worst. The greatest apparent benefit of mAFP was observed in noncongested CS.