Related Experiment Video
Updated: Jun 25, 2025
08:50
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
6.9K
Artificial intelligence-based prognostic model accurately predicts the survival of patients with diffuse large B-cell
Huilin Peng1, Mengmeng Su2, Xiang Guo3
1Department of Lymphatic Oncology, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of Sciences, Hangzhou, Zhejiang, 310022, China.
BMC Cancer
|May 21, 2024
Summary
A new molecular-contained prognostic model (McPM) and its simplified version (sMcPM) offer superior prognostic stratification for diffuse large B-cell lymphoma (DLBCL) patients compared to the International Prognostic Index (IPI). These AI-driven tools improve risk assessment and guide personalized DLBCL treatments.
Area of Science:
- Oncology
- Hematology
- Bioinformatics
Background:
- Diffuse large B-cell lymphoma (DLBCL) exhibits significant molecular heterogeneity.
- The current International Prognostic Index (IPI) lacks molecular data, limiting its prognostic accuracy.
- Artificial intelligence (AI) can identify novel molecular indicators for improved DLBCL risk stratification.
Purpose of the Study:
- To develop and validate a novel prognostic scoring system for DLBCL incorporating molecular data.
- To compare the performance of the new model against the established IPI.
- To create a simplified, clinically applicable version of the prognostic model.
Main Methods:
- Retrospective analysis of 401 DLBCL patients (2015-2019).
- Utilized random survival forest for variable weighting and a combination of bidirectional long-short term memory (Bi-LSTM) and logistic hazard techniques to build the molecular-contained prognostic model (McPM).
- Developed a simplified McPM (sMcPM) and compared predictive performance against IPI using C-index, integrated Brier score (IBS), and receiver operating characteristic (ROC) curve analysis.
Main Results:
- The McPM demonstrated superior predictive accuracy for overall survival (OS) and progression-free survival (PFS) compared to IPI.
- The simplified sMcPM, using key indicators (extranodal involvement, LDH, MYC status, AMC, PLT), showed comparable OS and significantly better PFS stratification than IPI.
- AI-driven model development identified crucial molecular and clinical predictors for DLBCL prognosis.
Conclusions:
- The novel McPM, integrating clinical and molecular factors, offers enhanced prognostic stratification for DLBCL, outperforming the IPI.
- The sMcPM provides a practical and effective tool for clinical risk stratification in the molecular era.
- These advanced models can guide individualized precision treatments and support new drug development in DLBCL.
