Nomogram Prognostic Model

Nomogram Prognostic Model predicts overall survival for advanced non-small-cell lung cancer (NSCLC) patients by applying a multivariable nomogram that integrates clinical and routine laboratory variables to generate individualized risk assessments.

Key Features:

• Nomogram-based prognostic model: Integrates multiple clinical and laboratory variables to provide individualized survival risk estimates for advanced NSCLC patients.
• Development dataset: Built using data from 1,817 advanced NSCLC patients derived from the control arms of four Phase III randomized clinical trials.
• Validation strategy: A subset of 524 patients was used to validate an existing nomogram model, while the remaining patients served as independent validation cohorts for the newly developed model.
• Included variables: Sex, histology, performance status, liver metastasis, hemoglobin level, white blood cell counts, peritoneal metastasis, skin metastasis, and lymphocyte percentage.
• Statistical evaluation metrics: Prognostic performance assessed by hazard ratios, integrated area under the curve (AUC), concordance index, calibration plots, and log-rank P-values.
• Performance in SUN1087 cohort: Integrated AUC 0.83, log-rank P = 3.87e-11, concordance index 0.717.
• Performance in SAVEONCO cohort: Integrated AUC 0.755, log-rank P = 4.94e-6, concordance index 0.678.
• Performance in VITAL cohort: Integrated AUC 0.723, log-rank P = 1.36e-11, concordance index 0.654.

Scientific Applications:

• Overall survival prediction: Provides individualized overall survival probability estimates for patients with advanced NSCLC.
• Risk stratification for treatment planning: Enables stratification of patients by prognostic risk to inform clinical decision-making and personalized treatment strategies.
• Clinical trial enrichment: Facilitates patient selection and enrollment by identifying prognostic risk groups for clinical trial stratification.

Methodology:

Analysis of 1,817 advanced NSCLC patients from control arms of four Phase III randomized trials with a subset of 524 used to validate an existing nomogram and remaining patients used as independent validation cohorts; the model integrates specified clinical and laboratory variables and its performance was evaluated using hazard ratios, integrated AUC, concordance index, calibration plots, and log-rank tests.