基于聚类分析的原发性肝癌患者预后预测

Prediction of prognosis of primary liver cancer patients based on cluster analysis

目的通过探索原发性肝癌患者术前的临床信息,进而评估患者的临床表型特点,对患者进行根治性肝癌切除术的预后状况进行预测,为制定个体化诊治方案和治疗策略提供临床依据。方法对386名原发性肝癌患者的34个基线临床资料进行主成分分析和聚类分析,再使用Kaplan-Meier方法和Cox比例风险模型比较各亚组的临床信息特点和预后差异。结果通过聚类分析确定了4组临床表型亚组,4组患者的一些临床表型和预后效果存在统计学差异:第1亚组的患者预后效果最好,第3亚组的预后最差。与亚组1相比,第2亚组的复发风险比为1.32(95%CI:1.03～1.70),第3亚组的复发风险比为3.60(95%CI:1.97～6.58);在考虑死亡风险时,第2亚组的死亡风险提高了1.43倍(95%CI:1.10～1.86),第3亚组的死亡风险是第1亚组的4.11倍(95%CI:2.11～8.00)。结论基于临床表型信息的聚类分析可以确定原发性肝癌患者的临床表型亚组,可为术后个体化辅助治疗提供一定的参考。

Objective To explore the preoperative clinical information of patients with hepatocellular carcinoma,and then to assess the clinical phenotype characteristics of patients.To predict the prognosis of patients undergoing radical hepatectomy,and to provide clinical evidence for the development of individualized treatment and treatment strategies.Methods Principal component analysis and cluster analysis were performed on34baseline clinical data of386patients with primary hepatocellular carcinoma,and the clinical information characteristics and prognosis differences of each subgroup were compared using Kaplan-Meier method and Cox proportional risk model.Four subgroups of clinical phenotype were identified by cluster analysis.Result There were statistically significant differences in the clinical phenotype and prognosis of the four groups after radical hepatectomy.Patients in the first subgroup had the best prognosis and those in the third subgroup had the worst prognosis.The recurrence risk ratio in the second subgroup was1.32(95%CI:1.03-1.70)compared with subgroup1,and the recurrence risk ratio in the third subgroup was3.60(95%CI:1.97-6.58).When considering the risk of death,the risk of death in the second subgroup was increased by1.43times(95%CI:1.10-1.86),and the risk of death in the third subgroup was4.11times(95%CI:2.11-8.00).Conclusion Cluster analysis can be used to determine the subgroup of clinical phenotype in patients with hepatocellular carcinoma,which can provide certain reference for postoperative individualized adjuvant therapy.