2023 World Conference on Lung Cancer (Posters)-P2.05. An Artificial Neural Network System to Predict the Immune-related Adverse Events Based on the RNA Data of ORIENT-3 Study

P2.05. An Artificial Neural Network System to Predict the Immune-related Adverse Events Based on the RNA Data of ORIENT-3 Study - PDF(Slides)

Back to course

Pdf Summary

Researchers from the Department of Medical Oncology at the National Cancer Center in China have developed an artificial neural network (ANN) system to predict immune-related adverse events (irAEs) in patients with squamous-cell non-small-cell lung cancer (sqNSCLC) who are treated with sintilimab. Sintilimab is a treatment option for patients who have failed first-line chemotherapy. The researchers used RNA data from the ORIENT-3 study, which collected detailed information on irAEs and RNA data. The RNA data was normalized and converted logarithmically before being used to construct the ANN system. Patients treated with sintilimab were randomly divided into a training cohort (70%) and a test cohort (30%). Univariate logistic regression and receiver operating characteristic curve analysis were used to select genes for constructing the ANN models. The top 15 significant genes were used to build 900 ANN models, which were tested for accuracy using three-fold cross validation. The accuracy of the final ANN system was then tested in the test cohort. The results showed that the ANN system had a potential value in predicting irAEs in sqNSCLC patients treated with sintilimab. The frequency of ANN models with a weight over 0.40 (indicating a smaller AUC over 0.70) was 95.22%. The AUC in the training cohort was 1.00, while in the test cohort, it was 0.84. This research provides a potential tool for predicting irAEs in patients receiving sintilimab treatment, which could help improve patient management and treatment outcomes.

Asset Subtitle

Tongji Xie

Meta Tag

Speaker Tongji Xie

Topic Metastatic NSCLC: Immunotherapy - Biomarker

Keywords

artificial neural network

irAEs

sqNSCLC

sintilimab

RNA data

training cohort

test cohort

genes

ANN models

patient management