Predicting Efficacy of Neoadjuvant Immunotherapy in Lung Cancer based on Tertiary Lymphoid Structure and...

Area of Science:

• Oncology
• Immunology
• Computational Pathology

Background:

• Neoadjuvant immunotherapy is a key lung cancer treatment, but predicting patient response remains challenging due to immune microenvironment complexity.
• Current methods lack the ability to accurately predict individualized efficacy for neoadjuvant immunotherapy in lung cancer patients.

Purpose of the Study:

• To develop a novel multi-instance learning model (GLAM) for predicting individualized short-term and long-term efficacy of neoadjuvant immunotherapy in lung cancer.
• To integrate fine-grained tertiary lymphoid structure (TLS) features and global immune microenvironment features from H&E-stained whole-slide images (WSI) for enhanced predictive power.

Main Methods:

• A semi-supervised learning approach was used to train a network for predicting TLS maturity, a prognostic indicator.
• A multi-instance learning model with self-attention was constructed, combining TLS and global immune features via cross-attention to predict clinical endpoints.
• The study analyzed 194 lung cancer patients who received neoadjuvant immunotherapy, utilizing post-operative WSI.

Main Results:

• The GLAM model demonstrated strong predictive performance for both short-term and long-term efficacy endpoints.
• For short-term efficacy, AUCs of 0.951 for major pathological response and 0.864 for pathological complete response were achieved.
• For long-term efficacy, AUC of 0.911 for 2.5-year recurrence status and a C-Index of 0.805 for predicting recurrence time were obtained.

Conclusions:

• The GLAM model offers a novel computational method for predicting individualized efficacy of neoadjuvant immunotherapy in lung cancer.
• This approach can significantly aid in personalized treatment planning for lung cancer patients undergoing neoadjuvant immunotherapy, improving clinical decision-making.