Development and Validation of a Nomogram Prognostic Model for Resected Limited-Stage Small Cell Lung Cancer Patients
Qingpeng Zeng1, Jiagen Li1, Fengwei Tan1
1Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, No. 17, Panjiayuan Nanli, Beijing, 100021, China.
Background:
In this study, we developed and validated nomograms for predicting the survival in surgically resected limited-stage small cell lung cancer (SCLC) patients.
Methods:
The SCLC patients extracted from the Surveillance, Epidemiology, and End Results database between 2000 and 2014 were reviewed. Significant prognostic factors were identified and integrated to develop the nomogram using multivariable Cox regression. The model was then validated internally by bootstrap resampling, and externally using an independent SCLC cohort diagnosed between 2000 and 2015 at our institution. The prognostic performance was measured by the concordance index (C-index) and calibration curve.
Results:
A total of 1006 resected limited-stage SCLC patients were included in the training cohort. Overall, 444 cases from our institution constituted the validation cohort. Seven prognostic factors were identified and entered into the nomogram construction. The C-indexes of this model in the training cohort were 0.723, 0.722, and 0.746 for predicting 1-, 3-, and 5-year overall survival (OS), respectively, and 0.816, 0.710, and 0.693, respectively, in the validation cohort. The calibration curve showed optimal agreement between nomogram-predicted survival and actual observed survival. Additionally, significant distinctions in survival curves between different risk groups stratified by prognostic scores were also observed. The proposed nomogram was then deployed into a website server for convenient application.
Conclusions:
We developed and validated novel nomograms for individual prediction of survival for resected limited-stage SCLC patients. These models perform better than the previously widely used staging system and may offer clinicians instructions for strategy making and the design of clinical trials.
