儿童暴发性心肌炎早期识别与重症化预警研究进展

Research Progress on Early Recognition and Risk Stratification for Progression to Critical Illness in Pediatric Fulminant Myocarditis

摘要: 儿童暴发性心肌炎(FM)起病隐匿，入院24 h“黄金窗口期”内误诊率与病死率居高不下，其早期识别与重症化风险分层系改善预后的核心前提。本文系统整合了近年来关于FM底层病理机制(泛凋亡与免疫风暴)、多组学分子标志物、多模态影像参数及人工智能预警模型的前沿循证数据。病理生理层面，泛凋亡级联引发的线粒体崩塌与失控性免疫风暴被确认为FM进行性恶化的底层驱动力。预警维度，可溶性生长刺激表达基因2蛋白(sST2, AUC 0.887)联合心型脂肪酸结合蛋白(H-FABP)可有效填补传统肌钙蛋白的窗口期假阴性；系统性免疫炎症指数(SII > 1050)及6 h乳酸清除率(<10%)系重症化的强独立预测因子。影像维度，整体纵向应变(GLS)的机械力学衰减显著先于宏观射血分数急降。干预维度，基于血管活性药物评分(VIS > 40)的体外膜肺氧合(ECMO)早期介入，并联IL-1受体靶向阻断，可实质性逆转血流动力学崩溃。鉴于此，深度融合sST2等高维分子标志物、GLS形变力学与AI算法，构建涵盖初筛、确诊至机制的“三级阶梯式”预警体系，系改善FM预后的核心策略。现代重症管理已实质性迈入“精准免疫调节并联机械辅助”的双轨模式。

Abstract: Pediatric fulminant myocarditis (FM) has an insidious onset, with persistently high rates of misdiagnosis and mortality during the 24hour “golden window” after admission. Early recognition and risk stratification for progression to critical illness are therefore the core prerequisites for improving outcomes. This article systematically integrates recent evidence on the underlying pathophysiological mechanisms of FM (PANoptosis and immune storm), multiomics biomarkers, multimodal imaging parameters, and artificial intelligencebased early warning models. At the pathophysiological level, mitochondrial collapse triggered by the PANoptosis cascade and uncontrolled immune storm have been identified as the driving forces behind the progressive deterioration of FM. For early warning, soluble suppression of tumorigenicity 2 (sST2, AUC 0.887) combined with hearttype fatty acidbinding protein (HFABP) effectively fills the windowperiod falsenegative gap of conventional troponin; a systemic immuneinflammation index (SII > 1050) and 6hour lactate clearance rate (<10%) are strong independent predictors of critical illness. On the imaging front, mechanical attenuation of global longitudinal strain (GLS) occurs significantly earlier than the acute decline in macroscopic ejection fraction. Regarding intervention, early initiation of extracorporeal membrane oxygenation (ECMO) guided by a vasoactiveinotropic score (VIS > 40), together with IL1 receptortargeted blockade, can substantially reverse hemodynamic collapse. Given this, the core strategy to improve FM outcomes is to deeply integrate highdimensional biomarkers such as sST2, GLS deformation mechanics, and AI algorithms into a threetiered early warning system covering screening, definitive diagnosis, and mechanistic evaluation. Modern critical care management of FM has essentially entered a dualtrack model of “precise immunomodulation combined with mechanical support”.