Multimodal radiopathomics for predicting immunotherapy response and prognosis in advanced non-small cell lung cancer

Objectives To develop and validate an integrative radiopathomics framework combining pretreatment multiphase contrast-enhanced computed tomography (CT) radiomics and hematoxylin-and-eosin (H&E)-stained whole-slide image (WSI)-derived pathomics for assessment of immunotherapy benefit and overall survival (OS) risk stratification in advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs). Methods This multicenter retrospective study included 344 patients with advanced NSCLC treated with ICI-based therapy between June 2019 and June 2024. Pretreatment non-contrast, arterial-phase and venous-phase chest CT images and H&E-stained WSIs were analyzed. Radiomics, pathomics and integrated radiopathomics models were developed in the training cohort using feature selection and support vector machine to distinguish durable clinical benefit (DCB) from non-durable clinical benefit (NDB) and were evaluated in one internal and two external validation cohorts. The final radiopathomics score (RPS) was further assessed for OS stratification and incorporated with clinically relevant covariates into an interpretable survival nomogram. Results Among the 344 patients included (mean age 66.5 ± 8.5 years; 309 men), the radiopathomics model achieved an area under the receiver operating characteristic curve (AUC) of 0.972 in the training cohort, 0.882 in the internal validation cohort, and 0.908 and 0.898 in the two external validation cohorts, respectively. RPS stratified OS risk in the pooled cohort (hazard ratio 4.65, 95% confidence interval 3.41–6.34; P < 0.001) and provided additional prognostic separation within programmed death-ligand 1 (PD-L1) expression strata. A nomogram combining RPS, Eastern Cooperative Oncology Group performance status, liver metastasis and PD-L1 expression achieved concordance indices of 0.765–0.899 across validation cohorts. Conclusions The radiopathomics model integrating pretreatment multiphase contrast-enhanced CT radiomics with pathomics derived from H&E-stained WSIs consistently distinguished DCB from NDB in patients with advanced NSCLC treated with ICIs. The resulting RPS enabled OS risk stratification and provided additional prognostic separation within PD-L1 expression strata. Together, these findings offer preliminary insight into immunotherapy response and may inform individualized treatment planning.

Authors

Institutions

Publication Details

Journal
Frontiers in Immunology
Published
2026-11-18
DOI
https://doi.org/10.3389/fimmu.2026.1928594
Primary Topic
Radiomics and Machine Learning in Medical Imaging
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Multimodal radiopathomics for predicting immunotherapy response and prognosis in advanced non-small cell lung cancer

Weiming Xu, Junru Peng, Lingling Zhou, Jianfei Tu et al.
3 links
Frontiers in Immunology
Radiomics and Machine Learning in Medical Imaging
article

Multimodal radiopathomics for predicting immunotherapy response and prognosis in advanced non-small cell lung cancer

Weiming Xu, Junru Peng, Lingling Zhou, Jianfei Tu, Xiaojie Zhang, Weibo Mao, Feng Ye, Hai Yang, Bojun Liu, Mengyu Xu, Yinuo Xu, Jinying Wu, Haokai Hua, Linqiang Lai, Wenjie Wang
article en

Abstract

Objectives To develop and validate an integrative radiopathomics framework combining pretreatment multiphase contrast-enhanced computed tomography (CT) radiomics and hematoxylin-and-eosin (H&E)-stained whole-slide image (WSI)-derived pathomics for assessment of immunotherapy benefit and overall survival (OS) risk stratification in advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs). Methods This multicenter retrospective study included 344 patients with advanced NSCLC treated with ICI-based therapy between June 2019 and June 2024. Pretreatment non-contrast, arterial-phase and venous-phase chest CT images and H&E-stained WSIs were analyzed. Radiomics, pathomics and integrated radiopathomics models were developed in the training cohort using feature selection and support vector machine to distinguish durable clinical benefit (DCB) from non-durable clinical benefit (NDB) and were evaluated in one internal and two external validation cohorts. The final radiopathomics score (RPS) was further assessed for OS stratification and incorporated with clinically relevant covariates into an interpretable survival nomogram. Results Among the 344 patients included (mean age 66.5 ± 8.5 years; 309 men), the radiopathomics model achieved an area under the receiver operating characteristic curve (AUC) of 0.972 in the training cohort, 0.882 in the internal validation cohort, and 0.908 and 0.898 in the two external validation cohorts, respectively. RPS stratified OS risk in the pooled cohort (hazard ratio 4.65, 95% confidence interval 3.41–6.34; P < 0.001) and provided additional prognostic separation within programmed death-ligand 1 (PD-L1) expression strata. A nomogram combining RPS, Eastern Cooperative Oncology Group performance status, liver metastasis and PD-L1 expression achieved concordance indices of 0.765–0.899 across validation cohorts. Conclusions The radiopathomics model integrating pretreatment multiphase contrast-enhanced CT radiomics with pathomics derived from H&E-stained WSIs consistently distinguished DCB from NDB in patients with advanced NSCLC treated with ICIs. The resulting RPS enabled OS risk stratification and provided additional prognostic separation within PD-L1 expression strata. Together, these findings offer preliminary insight into immunotherapy response and may inform individualized treatment planning.

Frontiers in ImmunologyVol. 17
Wenzhou Medical University (CN), Lishui Central Hospital (CN), Lishui City People's Hospital (CN), Zhejiang Taizhou Hospital (CN), Taizhou University (CN)
Zhejiang Medical Association
Good health and well-being
Radiomics and Machine Learning in Medical Imaging
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon