IHGAMP : Pan‐Cancer HRD Prediction From Routine H&E Whole‐Slide Images Using Foundation Models
APMIS · 2026-04 · 134 (4)
DOI: 10.1111/apm.70211 ↗Kurz gefasst
Fachartikel in der Zeitschrift APMIS, online erschienen am 24.04.2026. Verfasst von Sanwal Ahmad Zafar und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai, China. Hauptthemen laut MEDLINE: Neoplasms, Homologous Recombination. Die Arbeit nennt 22 Literaturangaben. Gefördert von: Shanghai Jiao Tong University.
Abstract (Originalsprache)
Homologous recombination deficiency (HRD) guides PARP inhibitor and platinum‐based therapy selection but relies on costly next‐generation sequencing with limited accessibility. Here, we show that IHGAMP, a computational framework using vision transformer foundation models, predicts HRD from routine H&E whole‐slide images with AUROC 0.766 (95% CI 0.727–0.803) using OpenCLIP embeddings on the TCGA held‐out test set ( n = 8109 patients, 31 cancer types), improving to 0.812 with histopathology‐pretrained OpenSlideFM. External evaluation on 927 patients from seven independent cohorts confirmed generalization in adenocarcinoma/serous histologies (CPTAC‐LUAD AUROC 0.723; PTRC‐HGSOC platinum resistance AUROC 0.673), with attenuation in squamous histologies. True external validation using the frozen TCGA‐trained model on CPTAC‐LUAD yielded AUROC 0.671, confirming cross‐institutional generalization. Pathway specificity analysis demonstrated that the apparent TP53 correlation was mediated through HRD (partial r = 0.112 after controlling for scarHRD), and evaluation on MMR‐deficient endometrial carcinoma yielded below‐random prediction (AUC = 0.445), ruling out general genomic instability capture. Sensitivity analysis across label definitions (scarHRD ≥ 33 and ≥ 42) showed robust performance, and a BRCA1/2 mutation‐trained comparator performed near‐randomly (AUC = 0.486), confirming that learned features reflect HRD biology beyond single‐gene effects. IHGAMP supports a screening/triage role to prioritize confirmatory molecular testing where appropriate.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai, China.
- Services Institute of Medical Sciences, Lahore, Pakistan.
- Stephenson Cancer Center, OU Health, The University of Oklahoma, Oklahoma City, Oklahoma, USA.
ORCID-Profile: Sanwal Ahmad Zafar, Areeba Ali Khan
Förderung
- Shanghai Jiao Tong University
Zitieren
APA
Zafar, S. A., Chengliang, L., Khan, A. A., Faisal, M. S., & Qin, W. (2026). IHGAMP : Pan‐Cancer HRD Prediction From Routine H&E Whole‐Slide Images Using Foundation Models. APMIS, 134(4). https://doi.org/10.1111/apm.70211
Vancouver
Zafar SA, Chengliang L, Khan AA, Faisal MS, Qin W. IHGAMP : Pan‐Cancer HRD Prediction From Routine H&E Whole‐Slide Images Using Foundation Models. APMIS. 2026;134(4). doi:10.1111/apm.70211
BibTeX
@article{Zafar2026,
author = {Zafar, Sanwal Ahmad and Chengliang, Liu and Khan, Areeba Ali and Faisal, Muhammad Salman and Qin, Wei},
title = {IHGAMP : Pan‐Cancer HRD Prediction From Routine H&E Whole‐Slide Images Using Foundation Models},
journal = {APMIS},
year = {2026},
volume = {134},
number = {4},
doi = {10.1111/apm.70211}
}
EUVIMED