Clinical Report: Enhanced Multimodal Radiomics Utilizing Foundation Models for Anticipating Chemo-Immunotherapy Outcomes in Advanced Lung Squamous Cell Carcinoma

Overview

This study presents a foundation model-driven multimodal fusion framework for predicting treatment response to first-line chemo-immunotherapy in advanced lung squamous cell carcinoma (LUSC). The model demonstrated an AUC of 0.903 in the training cohort and 0.863 in the validation cohort.

Background

Advanced lung squamous cell carcinoma (LUSC) is a major subtype of non-small cell lung cancer (NSCLC) with limited treatment options and heterogeneous responses to therapy. The integration of immune checkpoint inhibitors with chemotherapy has improved outcomes, yet reliable predictive biomarkers for treatment response remain scarce. This study addresses the need for non-invasive methods to stratify patients based on their likelihood of benefiting from chemo-immunotherapy.

Data Highlights

Key Findings

• The FusionModel achieved an AUC of 0.903 and accuracy of 0.885 in the training cohort.
• In the validation cohort, the FusionModel maintained an AUC of 0.863 and accuracy of 0.836.
• FusionModel outperformed single-modality feature models.
• SHAP analysis revealed that deep learning features were significant contributors to the model's predictions.

Clinical Implications

The foundation model-driven multimodal fusion framework offers a non-invasive tool for predicting treatment responses in advanced LUSC.

Conclusion

The study demonstrates that a multimodal fusion model can predict chemo-immunotherapy responses in advanced LUSC.

Related Resources & Content

1. ASCO AI in Oncology, 2026 -- Improved Immunotherapy Response Prediction in NSCLC With Deep-Learning Radiomic Biomarker
2. ASCO AI in Oncology, 2026 -- Multimodal Model Uses Pathology Data to Predict Immunotherapy Response in NSCLC
3. European Radiology, 2025 -- CT-Based Radiogenomics Evaluation of Metastatic Lung Adenocarcinoma: A Study of Single and Multi-Site Analysis and Its Impact on Patient Outcomes
4. The ASCO Post, 2026 -- Deep-Learning CT Biomarker Predicts Survival Better Than Traditional Measures in Immunotherapy-Treated Advanced NSCLC
5. Journal of Clinical Oncology, 2026 -- Therapy for Stage IV Non–Small Cell Lung Cancer Without Driver Alterations: ASCO Living Guideline, 2026.3.0
6. Journal of Clinical Oncology, 2022 -- Pembrolizumab Plus Chemotherapy in Squamous Non–Small-Cell Lung Cancer: 5-Year Update of the Phase III KEYNOTE-407 Study
7. asco ai in oncology — Improved Immunotherapy Response Prediction in NSCLC With Deep-Learning Radiomic Biomarker
8. asco ai in oncology — Multimodal Model Uses Pathology Data to Predict Immunotherapy Response in NSCLC
9. RECIST 1.1 - RECIST
10. Therapy for Stage IV Non–Small Cell Lung Cancer Without Driver Alterations: ASCO Living Guideline, 2026.3.0 | Journal of Clinical Oncology
11. Pembrolizumab Plus Chemotherapy in Squamous Non–Small-Cell Lung Cancer: 5-Year Update of the Phase III KEYNOTE-407 Study | Journal of Clinical Oncology