A joint model study for dynamically predicting the survival of patients after radical gastrectomy based on longitudinal data of
serum total protein
1Zhang Cheng,2Jin Lian,2Wang Jiayi,2Ma Tai
1
Anhui Cancer Institute Hefei 230022 Anhui China
2Department of Oncology The First Affiliated Hospital of Anhui Medical
University Hefei 230022 Anhui China
Abstract：Objective To investigate the association between the longitudinal dynamics of serum total protein TP and postoperative
survival outcomes in patients undergoing radical gastrectomy for gastric cancer. Method Patients with adenocarcinoma of the stomach or
gastroesophageal junction who underwent D2 radical surgery between 2010 and 2022 were enrolled. Longitudinally measured TP levels
during follow-up were recorded and patient survival status and duration were ascertained via cause-of-death verification or telephone
follow-up. Cluster analysis of oncological characteristics was performed using the K-prototype algorithm. In the joint model a linear
mixed model was adopted for the longitudinal sub - model and a Cox proportional hazards model was used for the survival sub -
model. This joint model aimed to explore the interaction between TP and oncological characteristic subgroups analyze the impact of
temporal trends in TP changes during follow-up on survival and dynamically predict individual survival probabilities. Result A total of
601 patients were enrolled with a median follow-up duration of 5. 74 years. Among these 438 patients were followed up until death
with a median survival time of 2. 18 years. Cluster analysis revealed two oncological characteristic clusters cluster A was featured by
more advanced tumor stages poorer differentiation and an increased risk of neural and vascular invasion. The joint model
demonstrated that the interaction between oncological characteristic subgroups and time - varying TP levels represents a critical
determinant of postoperative survival. Notably a negative association was observed between TP levels and mortality in cluster A.
Conclusion Serum TP levels following radical gastrectomy for gastric cancer exhibit a subgroup - specific negative association with
mortality. Regular monitoring of TP may offer valuable insights for optimizing nutritional support and prognostic management.
1张 澄,2 金 莲,2 王嘉怡,2 马 泰. 基于血清总蛋白纵向数据动态预测胃癌根治术后患者
生存的联合模型研究[J]. 肿瘤代谢与营养电子杂志, 2025, 12(5): 587-593.
1Zhang Cheng,2Jin Lian,2Wang Jiayi,2Ma Tai. A joint model study for dynamically predicting the survival of patients after radical gastrectomy based on longitudinal data of
serum total protein. Electron J Metab Nutr Cancer, 2025, 12(5): 587-593.
