Construction and validation of a nomogram model for the prediction of the prognosis of pulmonary large cell neuroendocrine carcinoma
10.3760/cma.j.cn115355-20240428-00213
- VernacularTitle:肺大细胞神经内分泌癌预后预测列线图模型的构建与验证
- Author:
Yi HAN
1
;
Fei QI
;
Hongmei ZHANG
;
Hongbo WU
;
Yong ZHANG
;
Tongmei ZHANG
Author Information
1. 首都医科大学附属北京胸科医院综合科，北京 101149
- Publication Type:Journal Article
- Keywords:
Lung neoplasms;
Carcinoma, neuroendocrine;
Nomograms;
Prognosis
- From:
Cancer Research and Clinic
2025;37(8):569-576
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the prognostic influencing factors of patients with pulmonary large cell neuroendocrine carcinoma (LCNEC), to develop a nomogram-based predictive model for the overall survival (OS) of LCNEC patients and to make validation.Methods:The clinical data of 2 947 patients with LCNEC in the Surveillance, Epidemiology, and End Results (SEER) database (the modeling group) and 147 patients with LCNEC in Beijing Chest Hospital Affiliated to Capital Medical University from 2010 to 2023 (the validation group). The data of patients in the both groups were compared. Cox proportional hazards model was used to screen out the factors influencing the OS of patients with LCNEC. A nomogram model was constructed to predict the OS based on the multivariate analysis result. Internal validation of the predictive model's performance was conducted through 500 repeated samplings based on the Bootstrap method. The predictive performance of the nomogram model was evaluated by using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve. The consistency index (CI) was used to analyze the discrimination of the nomogram model in predicting the survival of LCNEC patients; calibration curves were used to analyze the consistency between the survival predicted by the nomogram model and the actual survival outcomes; and the decision curve analysis (DCA) was used to assess the net benefit of the model for actual clinical decision-making.Results:The differences in the proportions of patients with different age, gender, race, tumor staging, N stage, M stage, hepatic metastasis or not, pulmonary metastasis or not, chemotherapy and radiotherapy or not between the modeling group and the validation group were statistically significant (all P < 0.05). The median OS time of LCNEC patients in the modeling group was 14.0 months, with the 1-year OS rate of 53.3% and the 5-year OS rate of 21.2%; the median OS time of LCNEC patients in the validation group was 17.5 months, with the 1-year OS rate of 58.7%; there was no statistically significant difference in OS between the 2 groups ( P = 0.280). In the modeling group, the median OS time of female and male LCNEC patients was 18.0 and 12.0 months, respectively, and the difference in OS between the 2 groups was statistically significant ( P < 0.05); for patients with stage Ⅰ-Ⅱ, Ⅲ, and Ⅳ LCNEC, the median OS time was 48.0, 16.0, and 6.0 months, respectively, and the difference in OS among the 3 groups was statistically significant ( P < 0.05); the median OS time of patients receiving surgery and not receiving surgery was 28.0 and 8.0 months, respectively, and the difference in OS between the 2 groups was statistically significant ( P < 0.05). The differences in OS among female and male, patients in stages Ⅰ-Ⅱ, Ⅲ and Ⅳ, patients who underwent surgery or not were statistically significant (all P < 0.05). The results of multivariate Cox regression analysis in the modeling group showed that patients aged >60 years old (>60 years old vs. ≤60 years old: HR = 1.234, 95% CI: 1.114-1.367, P < 0.01), M 1 stage (M 1 stage vs. M 0 stage, HR = 2.646,95% CI: 2.385-2.935, P < 0.001), T 2-4 stage (T 2-4 stage vs. T 1 stage: HR = 1.199, 95% CI: 1.147-1.252, P < 0.001), N 1-3 stage (N 1-3 stage vs. N 0 stage: HR = 1.281, 95% CI: 1.225-1.340, P < 0.001) were independent risk factors of the OS in patients with LCNEC; female (female vs. male: HR = 0.877, 95% CI: 0.805-0.956, P = 0.003), surgery (yes vs. no: HR = 0.612, 95% CI: 0.554-0.676, P < 0.001), chemotherapy (yes vs. no: HR = 0.520, 95% CI: 0.470-0.575, P < 0.001) were independent protective factors of the OS in patients with LCNEC. A nomogram model for predicting 1, 3, and 5-year OS rates of LCNEC patients was constructed based on age, gender, T stage, N stage, M stage, surgery and chemotherapy. The result of ROC curve analysis indicated that the AUC of the nomogram model for predicting 1, 3, and 5-year OS rates in the modeling group was 0.822, 0.821 and 0.821, respectively, while the AUC of 1-year OS rate predicted by the validation group was 0.660. The CI of the modeling group and the validation group was 0.756 and 0.660, respectively. The calibration curve showed that 1, 3, and 5-year OS rates predicted by the modeling group were highly consistent with the actual OS rates. The DCA showed that the nomogram model for predicting OS in the modeling group and the validation group both had good clinical net benefits. Conclusions:The constructed nomogram model for predicting the prognosis of LCNEC patients is proved to be reliable and has good clinical values.
