2024 World Conference on Lung Cancer (WCLC) - Posters-P1.08A.01 Identifying Candidates for Postoperative Radiotherapy in Patients with Non-Small Cell Lung Cancer: A Multicenter Study

P1.08A.01 Identifying Candidates for Postoperative Radiotherapy in Patients with Non-Small Cell Lung Cancer: A Multicenter Study

Back to course

Pdf Summary

The study, "Identifying Candidates for Postoperative Radiotherapy in Patients with Non-Small Cell Lung Cancer: A Multicenter Study," focuses on evaluating the efficacy of a deep learning model in predicting disease-free survival (DFS) for non-small cell lung cancer (NSCLC) patients, aiming to identify those who could benefit from postoperative radiotherapy (PORT). Traditional methods like the Cox model provide population-level risk assessments, whereas the deep learning model, specifically DeepSurv, utilized in this study offers individualized predictions based on each patient's unique profile.<br /><br />Participants with histologically proven pN2 NSCLC who had complete resections were used to train the model at one institution, while participants from the PORT-C randomized controlled trial formed the test set. Additional patients from four independent medical centers served as external validation sets. The model's performance was measured using the concordance index (C-index) to ensure accurate predictions.<br /><br />The DeepSurv model categorized patients into two subgroups: those recommended for PORT and those not. The study compared DFS across these subgroups to determine the model's clinical impact on treatment recommendations. <br /><br />The conclusions drawn from the study emphasized the potential application of this deep learning model for predicting DFS and identifying NSCLC patients who might benefit from PORT. It highlighted the need for further prospective validation in clinical trials, efforts to improve the model’s interpretability, and strategies to integrate it into clinical workflows effectively.<br /><br />This multicentric collaboration included institutions such as the National Cancer Center in Beijing, Zhejiang Cancer Hospital, and West China Hospital, among others, reflecting a robust and comprehensive research effort.

Asset Subtitle

Zeliang Ma

Meta Tag

Speaker Zeliang Ma

Topic Local-Regional NSCLC

Keywords

deep learning

non-small cell lung cancer

postoperative radiotherapy

disease-free survival

DeepSurv model

Cox model

concordance index

pN2 NSCLC

clinical validation

multicenter study