肝泡型包虫病根治性切除术后教科书式结局的影响因素:回顾性队列研究

Predictive study of textbook outcomes after radical resection of hepatic alveolar echinococcosis:retrospective cohort study

目的探索肝泡型包虫病(hepatic alveolar echinococcosis,HAE)根治性切除术后教科书式结局(textbook outcomes,TO)的影响因素。方法回顾性收集2015年至2022年期间在四川大学华西医院甘孜分院确诊的427例HAE患者的临床资料,首先采用最小绝对收缩和选择算子回归(least absolute shrinkage and selection operator regression,LASSO)筛选TO达成的潜在影响因素(预测模型A),再采用单因素和多因素logistic回归分析探索TO的影响因素(预测模型B),进一步构建列线图并进行验证与比较。结果共纳入427例成功接受根治性肝切除术的HAE患者,其中实现TO患者174例(40.7%)。相较于非TO组,TO组患者的P 1+2期[62.6%(109/174)vs 42.3%(140/253)]、N0期[77.0%(134/174)vs 63.6%(161/253)]、病灶直径≤10 cm[77.0%(134/174)vs 64.8%(164/253)]、白蛋白-胆红素指数(albumin Bilirubin index,ALBI)≤–2.6[70.1%(122/174)vs 59.3%(150/253)]及小肝切除[71.8%(125/174)vs 50.2%(127/253)]比例高,前白蛋白(181 mg/L vs 169 mg/L)和白蛋白(39 g/L vs 38 g/L)水平高,术后住院时间短(15 d vs 19 d),住院费用低(51727元vs 62715元),P<0.05。LASSO回归分析结果(预测模型A)表明,P分期、病灶直径、ALBI和切肝方式为HAE术后达成TO的潜在影响因素(P<0.05),P 1+2期者的TO率高于P 3+4期者,病灶直径≤10 cm者的TO率高于病灶直径>10 cm者,ALBI≤–2.6者的TO率高于ALBI>–2.6者,小肝切除者的TO率高于大肝切除者。多因素logistic回归分析结果表明(预测模型B),P 1+2期者的TO率高于P 3+4期者(OR=1.800,P=0.025),小肝切除者的TO率高于大肝切除者(OR=1.974,P<0.001)。预测模型A的准确性及稳定性高于预测模型B(AUC:0.754 vs 0.712;C-index:0.756 vs0.707)。结论在接受根治性切除的HAE患者中,实现TO的比例未达到一半,致力于实现TO不仅能显著缩短患者的住院时间,还能有效降低医疗费用。基于P分期、病灶直径、ALBI、切肝方式建立的TO预测模型优于仅基于P分期和切肝方式的预测模型。

Objective To explore the influencing factors of textbook outcomes(TO)after radical resection for hepatic alveolar echinococcosis(HAE).Methods Clinical data from 427 patients diagnosed with HAE at the Ganzi Branch of West China Hospital,Sichuan University,between 2015 and 2022,were retrospectively collected.The least absolute shrinkage and selection operator regression(LASSO)was first used to screen potential influencing factors for achieving TO(predictive model A).Univariate and multivariate logistic regression analysis were then used to explore the influencing factors of TO(predictive model B).A Nomogram was further constructed and validated.Results A total of 427 patients who successfully underwent radical liver resection were included,of which 174 patients(40.7%)achieved TO.Compared to the non-TO group,patients in the TO group had higher proportions of WHO PNM stage of P 1+2[62.6%(109/174)vs 42.3%(140/253)],N0[77.0%(134/174)vs 63.6%(161/253)],lesion diameter≤10 cm[77.0%(134/174)vs 64.8%(164/253)],albumin-bilirubin index(ALBI)≤–2.6[70.1%(122/174)vs 59.3%(150/253)],and minor liver resection[71.8%(125/174)vs 50.2%(127/253)],higher prealbumin(181 mg/L vs 169 mg/L)and albumin(39 g/L vs 38 g/L)levels,shorter postoperative hospital stay(15 d vs 19 d),and lower hospitalization costs(51727 RMB vs 62715 RMB),P<0.05.LASSO regression analysis indicated that P stage,lesion diameter,ALBI,and liver resection method were potential influencing factors for achieving TO after HAE surgery(model A),P<0.05.The TO rate for P 1+2 stage was higher than that of P 3+4 stage,the TO rate for lesion diameter≤10 cm was higher than that of lesion diameter>10 cm,the TO rate for ALBI≤–2.6 was higher than that of ALBI>–2.6,and the TO rate for minor liver resection was higher than of major liver resection.Multivariate logistic regression analysis showed that P stage(OR=1.800,P=0.025)and liver resection method(OR=1.974,P<0.001)were influencing factors for achieving TO(model B).The TO rates for P1+2 stage and minor liver resection were higher.Predictive model A demonstrated higher accuracy and stability compared to predictive model B(AUC:0.754 vs 0.712,C-index:0.756 vs 0.707).Conclusions Among patients undergoing radical resection for HAE,less than half achieved TO.Striving to achieve TO can significantly shorten the length of hospital stay and effectively reduce medical costs.The TO predictive model based on P stage,lesion diameter,ALBI,and liver resection method is superior to the model based solely on P stage and liver resection method.