MELD亚型间意向治疗生存率的差异:为终末期肝病创建的所有模型并非等同
该研究纳入2016至2020年登记的成人肝移植候选者,按MELD评分主要驱动变量分为MELD-Br、MELD-INR和MELD-Cr三个亚型。结果显示MELD-Cr亚型1年意向治疗生存率最低(65%),且女性低于男性;校正MELD-Na后,MELD-Cr亚型等待期死亡风险更高、移植率更低。MELD亚型可能有助于动态评估死亡风险并指导器官分配。
要点
- MELD-Cr亚型1年意向治疗生存率为65%,低于MELD-Br的78%和MELD-INR的75%。
- MELD-Cr亚型女性1年意向治疗生存率63%,低于男性的67%。
- 校正MELD-Na后,MELD-Cr亚型等待期死亡风险比为1.339,移植率风险比为0.688。
核心结论
- 在等效登记实践下,MELD-Cr亚型登记者的意向治疗生存率更低。
- MELD亚型可能作为更精细的变量用于动态评估死亡风险并指导器官分配。
研究设计与样本
- 研究评估了2016至2020年间登记进行肝移植的所有成人患者,排除了接受MELD例外或进行多器官移植的患者。
- 使用K-Means聚类分析,根据MELD评分的主导变量将每位患者分类为MELD-Br、MELD-INR或MELD-Cr。
- 主要结局为意向治疗生存率,定义为从登记起1年内无论是否接受肝移植的生存。
主要发现
- MELD-Br组n=13,658,MELD-INR组n=13,809,MELD-Cr组n=12,412,三组1年意向治疗生存率分别为78%、75%和65%,p<0.01。
- 女性在各MELD亚型的意向治疗生存率均低于男性,例如MELD-Cr亚型女性63% vs 男性67%,p<0.0001。
- MELD-Cr亚型登记时MELD最高(MELD-Cr 23.4 vs MELD-Br 19.2 vs MELD-INR 21.0),且3个月内MELD下降幅度最大(23% vs 12% vs 21%)。
- 在包括MELD-Na的校正分析中,与其他亚型相比,MELD-Cr与更高的等待期死亡率相关(风险比1.339,95% CI 1.279-1.402),与更低的肝移植率相关(风险比0.688,95% CI 0.664-0.713)。
临床怎么用
- 作者指出,MELD评分的驱动因素很重要,尤其肌酐升高驱动的MELD与更低的1年意向治疗生存率相关。
- 作者指出,在肌酐主导亚型中,女性的1年意向治疗生存率也低于男性。
- 作者认为,肌酐可能作为预后标志物,即使肌酐水平改善预后仍然较差,需讨论替代移植途径。
- 作者指出,肾损伤类型很重要,在肌酐主导亚型中,急性肾损伤患者比慢性肾病患者更可能死亡或仍留在等待名单上。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND & AIMS: Kidney dysfunction is a major determinant of prognosis in patients with decompensated cirrhosis awaiting transplantation. We hypothesized that for identical model for end-stage liver disease (MELD) scores at listing, outcomes before and after liver transplantation may vary if the predominant driver of the MELD score is serum creatinine (Cr) vs. serum bilirubin (Br) or international normalized ratio (INR). METHODS: We evaluated all adult patients registered for liver transplantation (LT) between 2016-2020 and excluded patients receiving MELD exceptions or undergoing dual organ transplantation. Using K-Means clustering analysis, we classified each patient as MELD-Br, MELD-INR or MELD-Cr depending on the dominant variable for their MELD score. The primary outcome was intent-to-treat (ITT) survival, defined as survival within 1 year from listing with or without LT. RESULTS: MELD scores of LT waitlist registrants were clustered into three subtypes: MELD-Br (n = 13,658), MELD-INR (n = 13,809), and MELD-Cr (n = 12,412). One-year ITT survival rates were 78% (MELD-Br), 75% (MELD-INR), and 65% (MELD-Cr), p <0.01. ITT survival was lower for each MELD subtype for females compared to males (e.g. MELD-Cr: 63% females vs. 67% males, p <0.0001). The MELD-Cr subtype had the highest MELD at listing (MELD-Cr 23.4 vs. MELD-Br 19.2 vs. MELD-INR 21.0) and the largest decline in MELD over 3 months (23% vs. 12% vs. 21%). In adjusted analyses including MELD-Na, MELD-Cr was associated with higher waitlist mortality (hazard ratio 1.339, 95% CI 1.279-1.402) and lower LT rates (hazard ratio 0.688, 95% CI 0.664-0.713) compaed to the other subtypes. CONCLUSIONS: For equivalent listing practices, registrants with the MELD-Cr subtype have lower ITT survival. MELD subtype may serve as a more sophisticated variable for dynamic assessment of mortality risk and to guide organ allocation. IMPACT AND IMPLICATIONS: The model for end-stage liver disease (MELD) score is an excellent predictor of waitlist mortality; however, our work highlights that the driver of a patient's MELD score matters and particularly those driven by elevated creatinine are associated with lower 1-year intent-to-treat survival. The 1-year intent-to-treat survival is also lower for women compared to men within the Cr-dominant subtype. These results are important for physicians and patients undergoing LT evaluation as creatinine may serve as a marker of prognosis and even if creatinine levels improve the prognosis remains poor, necessitating discussion about alternative pathways for transplant. Our work also highlights that the type of kidney injury matters, in that those with acute kidney injury were more likely to die or remain on the waitlist than those with chronic kidney disease within the creatinine-dominant subtype.
引用
- Rosenstengle C, Serper M, Asrani SK, et al. Variation in intention-to-treat survival by MELD subtypes: All models created for end-stage liver disease are not equal. J Hepatol. 2024. PMID: 39181212. doi:10.1016/j.jhep.2024.08.006
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。