肝切除术后肝功能衰竭风险预测模型研究进展

Research progress of risk prediction models for post-hepatectomy liver failure

肝细胞癌(HCC)是最常见的原发性肝癌,是全球范围内最常见和高致死率的癌症之一。目前,根治性肝切除术依然是早期和部分中晚期HCC最有效的治疗手段之一。肝切除围手术期并发症是影响HCC患者长短期预后的关键影响因素,其中肝切除术后肝功能衰竭(PHLF)是肝切除术后常见的并发症之一。PHLF是肝切除患者围手术期死亡的主要病因,术前及时发现高危PHLF患者是亟待解决的临床实际问题和研究热点。传统肝功能评估方法应用广泛,能够区分出PHLF高危患者,但其预测准确性相对不高。近些年以来,随着人工智能技术的发展,越来越多的高级算法纳入更全面危险因素的模型被应用于PHLF预测领域。国内外学者通过各类统计学方法构建了新式PHLF相关预测模型,并证实了模型的准确性得到明显提升。经过前期大量文献检索,笔者通过归纳总结PHLF风险预测模型的相关文献,对其研究进展进行综述,以方便临床医生和研究者更全面地了解各类型的PHLF预测模型。

Hepatocellular carcinoma(HCC)is the most common primary liver cancer and ranks among the most prevalent and highly lethal cancers worldwide.Radical hepatectomy remains one of the most effective treatment methods for early and some intermediate to advanced stages of HCC.Complications during the perioperative period of liver resection are critical factors affecting the long-term and short-term prognosis of HCC patients,with post-hepatectomy liver failure(PHLF)being a common complication after liver resection.PHLF is a major cause of perioperative death in liver resection patients,and the timely identification of patients with high-risk PHLF before surgery is a pressing clinical issue and research focus.Traditional methods of liver function assessment are widely used and can distinguish high-risk PHLF patients,but their predictive accuracy is relatively low.In recent years,with the development of artificial intelligence technology,an increasing number of advanced algorithms and models incorporating more comprehensive risk factors have been applied in the field of PHLF prediction.Scholars both in China and abroad have constructed new PHLF-related prediction models through various statistical methods,confirming a significant improvement in the accuracy of these models.After an extensive literature review,the authors summarize the relevant literature on PHLF risk prediction models,providing a comprehensive overview of the research progress,so as to facilitate clinicians and researchers in gaining a more comprehensive understanding of various types of PHLF prediction models.