Prediction of Targeted Therapy Efficacy in EGFR-mutant Lung Cancer Patients Using AI-based Multimodal Data

The main purpose of this study is to explore the value of multimodal imaging information and models in predicting the prognosis of EGFR-positive non-small cell lung cancer patients undergoing targeted therapy, providing a basis for selecting suitable populations for precise tumor treatment and corresponding therapy. We retrospectively analyzed patient case data, extracted preoperative CT images, H\&E-stained whole-slide digital pathology images, and pre- or postoperative genetic testing reports to extract radiomic features of tumor and peritumoral regions. These features were combined with multidimensional pathological features and gene expression distribution characteristics to construct a multimodal radiopathogenomic model, offering more precise prognostic evaluation for lung cancer patients receiving targeted therapy.

Sex

All

Minimum age

18 Years

Maximum age

80 Years

Inclusion Criteria:

1. Age 18-80 years, undergoing radical surgery for lung cancer (R0 resection);
2. Postoperative pathological stage IB-IIIA, pathology confirmed as adenocarcinoma;
3. EGFR gene testing positive, EGFR 19del/L858R mutation;
4. Receiving postoperative EGFR-TKI targeted adjuvant therapy;
5. Complete and clear preoperative imaging data, genetic testing report, and pathology report available.

Exclusion Criteria:

1. Patients negative for EGFR;
2. Incomplete surgical resection (R1, R2);
3. Did not receive EGFR-TKI targeted therapy after surgery;
4. Recurrent or advanced stage patients;
5. Incomplete preoperative or postoperative data;
6. Patients who died within 30 days post-surgery.

• RESULTVaidya P, Bera K, Gupta A, Wang X, Corredor G, Fu P, Beig N, Prasanna P, Patil PD, Velu PD, Rajiah P, Gilkeson R, Feldman MD, Choi H, Velcheti V, Madabhushi A. CT derived radiomic score for predicting the added benefit of adjuvant chemotherapy following surgery in stage I, II resectable non-small cell lung cancer: a retrospective multicohort study for outcome prediction. Lancet Digit Health. 2020 Mar;2(3):e116-e128. doi: 10.1016/S2589-7500(20)30002-9. Epub 2020 Feb 13. PMID 33334576
• RESULTChen M, Lu H, Copley SJ, Han Y, Logan A, Viola P, Cortellini A, Pinato DJ, Power D, Aboagye EO. A Novel Radiogenomics Biomarker for Predicting Treatment Response and Pneumotoxicity From Programmed Cell Death Protein or Ligand-1 Inhibition Immunotherapy in NSCLC. J Thorac Oncol. 2023 Jun;18(6):718-730. doi: 10.1016/j.jtho.2023.01.089. Epub 2023 Feb 10. PMID 36773776
• RESULTLin H, Hua J, Gong Z, Chen M, Qiu B, Wu Y, He W, Wang Y, Feng Z, Liang Y, Long W, Li R, Kuang Q, Chen Y, Lu J, Luo S, Zhao W, Yan L, Chen X, Shi Z, Xu Z, Mo Z, Liu E, Han C, Cui Y, Yang X, Chen X, Liu J, Pan X, Madabhushi A, Lu C, Liu Z. Multimodal radiopathological integration for prognosis and prediction of adjuvant chemotherapy benefit in resectable lung adenocarcinoma: A multicentre study. Cancer Lett. 2025 Apr 28;616:217557. doi: 10.1016/j.canlet.2025.217557. Epub 2025 Feb 13. PMID 39954935