Preoperative prediction of microsatellite instability in rectal cancer using multiparametric MRI-derived intratumoral heterogeneity and exploration of biological mechanisms: a multicenter study

Objective To develop a multi-parametric MRI-based intratumoral heterogeneity (ITH) scoring model combined with clinical features for preoperative prediction of microsatellite instability (MSI) in rectal cancer, and to explore MSI-related molecular features. Methods A retrospective multicenter study of 464 rectal cancer patients (training n = 198, internal validation n = 110, external validation n = 156). Three-dimensional ITH scores were derived from T2WI and DWI. Six machine learning algorithms were used to build clinical, ITH, and combined models, evaluated by AUC, calibration curves, and SHAP analysis. TCGA transcriptomic data and immunohistochemistry (n = 10) were used to explore MSI-associated molecular features. Results Age, tumor length, location, and N stage were independent MSI predictors. The combined ExtraTrees model achieved AUCs of 0.865, 0.839, and 0.785 across three cohorts, outperforming single models. PPI network analysis identified AKT1 as the hub gene. MSS tumors showed significantly higher AKT1 expression than MSI-H tumors (H-score: 143.60 ± 26.14 vs. 76.74 ± 38.43, P = 0.012). Conclusions The ITH-clinical combined model enables effective non-invasive preoperative MSI prediction, with ITH providing incremental predictive value. Differential AKT1 expression offers preliminary molecular evidence for understanding MSI in rectal cancer.

Authors

Institutions

Publication Details

Journal
Abdominal Radiology
Published
2026-11-21
DOI
https://doi.org/10.1007/s00261-026-05874-7
Primary Topic
MRI in cancer diagnosis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Preoperative prediction of microsatellite instability in rectal cancer using multiparametric MRI-derived intratumoral heterogeneity and exploration of biological mechanisms: a multicenter study

Wanjun Hu, Yuping Bai, Yuxin Yan, Lixin Li et al.
Abdominal Radiology
MRI in cancer diagnosis
article

Preoperative prediction of microsatellite instability in rectal cancer using multiparametric MRI-derived intratumoral heterogeneity and exploration of biological mechanisms: a multicenter study

Wanjun Hu, Yuping Bai, Yuxin Yan, Lixin Li, Duo Yang, Yingying Jiang, Peng Liu, Xiaotong Wang, Baohong Gu
article en

Abstract

Objective To develop a multi-parametric MRI-based intratumoral heterogeneity (ITH) scoring model combined with clinical features for preoperative prediction of microsatellite instability (MSI) in rectal cancer, and to explore MSI-related molecular features. Methods A retrospective multicenter study of 464 rectal cancer patients (training n = 198, internal validation n = 110, external validation n = 156). Three-dimensional ITH scores were derived from T2WI and DWI. Six machine learning algorithms were used to build clinical, ITH, and combined models, evaluated by AUC, calibration curves, and SHAP analysis. TCGA transcriptomic data and immunohistochemistry (n = 10) were used to explore MSI-associated molecular features. Results Age, tumor length, location, and N stage were independent MSI predictors. The combined ExtraTrees model achieved AUCs of 0.865, 0.839, and 0.785 across three cohorts, outperforming single models. PPI network analysis identified AKT1 as the hub gene. MSS tumors showed significantly higher AKT1 expression than MSI-H tumors (H-score: 143.60 ± 26.14 vs. 76.74 ± 38.43, P = 0.012). Conclusions The ITH-clinical combined model enables effective non-invasive preoperative MSI prediction, with ITH providing incremental predictive value. Differential AKT1 expression offers preliminary molecular evidence for understanding MSI in rectal cancer.

Abdominal Radiology
Gansu University of Traditional Chinese Medicine (CN), Lanzhou University Second Hospital (CN), Shanghai First People's Hospital (CN)
Good health and well-being
MRI in cancer diagnosis
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