敗血症性心筋症における乳酸推移に基づく表現型と28日死亡率との関連：多施設共同後ろ向きコホート研究

To investigate phenotypic classification based on plasma lactate trajectories and its association with 28-day mortality in patients with septic cardiomyopathy (SCM).

A multicenter retrospective cohort study was conducted. Patients with SCM admitted to the intensive care units (ICUs) of three tertiary hospitals (Affiliated Hospital of Guizhou Medical University, Guizhou Provincial People's Hospital, and GuiQian International Hospital) from October 2019 to December 2024 were enrolled. Baseline characteristics, intervention strategies, and clinical outcomes were collected, including the first blood lactate level upon ICU admission and daily lactate values within 7 days after SCM diagnosis. The primary outcome was 28-day mortality, and secondary outcomes included 28-day vasoactive drug-free days, duration of mechanical ventilation, and ICU length of stay. Latent class growth model (LCGM) was used to identify lactate trajectory-based phenotypes. Differences in the above indicators among phenotypes were compared, and univariate and multivariate Cox regression analyses were performed to identify independent risk factors for 28-day mortality in SCM patients.

A total of 216 SCM patients were included, among whom 41 died and 175 survived within 28 days. LCGM identified four lactate trajectory phenotypes: phenotype 1 (persistent low level, n = 142), phenotype 2 (moderate level with rapid decline, n = 32), phenotype 3 (moderate level with slow increase, n = 20), and phenotype 4 (high level with slow decline, n = 22). The 28-day mortality rates of patients in phenotypes 1-4 showed a significant increasing trend (9.2%, 21.9%, 30.0%, and 68.2%, respectively, P < 0.05). Phenotype 1, which exhibited the lowest mortality and lowest acute physiology and chronic health evaluation (APACHE), was designated as the baseline group. Univariate Cox regression analysis showed that phenotypes 3 and 4 were associated with 28-day mortality in SCM patients (both P < 0.05). Multivariate Cox regression analysis revealed that phenotype 3 [hazard ratio (HR) = 2.831, 95% confidence interval (95%CI) was 1.243-6.447] and phenotype 4 (HR = 2.389, 95%CI was 1.223-4.663) were independent risk factors for 28-day mortality (both P < 0.05). Kaplan-Meier survival curves showed that the survival probability gradually flattened after 3 weeks of hospitalization. Using phenotype 1 as the baseline, patients with phenotypes 3 and 4 had a higher risk of death, and the risk in phenotype 4 was significantly higher than that in phenotype 3 (P < 0.05). Accordingly, the four patient groups were stratified into high-risk (high level with slow decline), intermediate-risk (moderate level with slow increase), and low-risk (moderate level with rapid decline and persistent low level) categories.

Four phenotypes of SCM were successfully identified based on lactate trajectories, which were closely associated with 28-day mortality and could be used for risk stratification, providing new insights for prognosis assessment and individualized treatment of SCM.