Objective To analyze the level of serum neuron-specific enolase (NSE) in patients with small cell lung cancer, and establish a dynamic prognostic model for SCLC.

Methods A total of 404 patients with small cell lung cancer were collected. The longitudinal analysis submodel was used to analyze the correlation between the longitudinal changes of serum NSE level and covariates after treatment. The submodel of survival analysis (COX regression model) was used to predict the potential risk factors affecting the survival of patients. Two submodels were combined to analyze the relationship between longitudinal NSE changes and patient survival, and a personalized dynamic prognosis prediction model was established.

Results The results of longitudinal submodel analysis showed that the short follow-up time and TNM stage 3–4 were correlated with the increase of serum NSE level (P ＜ 0.05 to P ＜ 0.01). The results of COX regression model analysis showed that the high baseline (pre-treatment) NSE level, older age, high TNM stage and CgaSyn (+) were associated with the poor prognosis (P ＜ 0.05 to P ＜ 0.01). The results of combined model analysis showed that longitudinal NSE changes were associated with patient survival, and the higher the NSE, the worse the patient prognosis (P ＜ 0.05). A personalized dynamic prognosis prediction model was established, the patients were randomly selected for validation, and the results proved that the model could effectively predict the survival outcome of patients.

Conclusions The change of longitudinal NSE is related to the survival of patients. The personalized dynamic prediction model based on random effects can effectively predict the survival of patients.