Related Experiment Video
Updated: Jun 27, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Machine-learning and mechanistic modeling of metastatic breast cancer after neoadjuvant treatment
Sebastien Benzekry1, Michalis Mastri2, Chiara Nicolò3
1Computational Pharmacology and Clinical Oncology (COMPO), Inria Sophia Antipolis-Méditerranée, Cancer Research Center of Marseille, Inserm UMR1068, CNRS UMR7258, Aix Marseille University UM105, Marseille, France.
Abstract:
Clinical trials involving systemic neoadjuvant treatments in breast cancer aim to shrink tumors before surgery while simultaneously allowing for controlled evaluation of biomarkers, toxicity, and suppression of distant (occult) metastatic disease. Yet neoadjuvant clinical trials are rarely preceded by preclinical testing involving neoadjuvant treatment, surgery, and post-surgery monitoring of the disease. Here we used a mouse model of spontaneous metastasis occurring after surgical removal of orthotopically implanted primary tumors to develop a predictive mathematical model of neoadjuvant treatment response to sunitinib, a receptor tyrosine kinase inhibitor (RTKI). Treatment outcomes were used to validate a novel mathematical kinetics-pharmacodynamics model predictive of perioperative disease progression. Longitudinal measurements of presurgical primary tumor size and postsurgical metastatic burden were compiled using 128 mice receiving variable neoadjuvant treatment doses and schedules (released publicly at https://zenodo.org/records/10607753). A non-linear mixed-effects modeling approach quantified inter-animal variabilities in metastatic dynamics and survival, and machine-learning algorithms were applied to investigate the significance of several biomarkers at resection as predictors of individual kinetics. Biomarkers included circulating tumor- and immune-based cells (circulating tumor cells and myeloid-derived suppressor cells) as well as immunohistochemical tumor proteins (CD31 and Ki67). Our computational simulations show that neoadjuvant RTKI treatment inhibits primary tumor growth but has little efficacy in preventing (micro)-metastatic disease progression after surgery and treatment cessation. Machine learning algorithms that included support vector machines, random forests, and artificial neural networks, confirmed a lack of definitive biomarkers, which shows the value of preclinical modeling studies to identify potential failures that should be avoided clinically.
Insights
Neoadjuvant receptor tyrosine kinase inhibitor (RTKI) treatment in breast cancer models shrinks tumors but fails to prevent postsurgical metastasis. Preclinical modeling and machine learning found no reliable biomarkers for predicting treatment response or metastatic progression.
Area of Science:
- Oncology
- Mathematical Biology
- Translational Research
Background:
- Neoadjuvant treatments in breast cancer aim to reduce tumor size before surgery and assess treatment response.
- Preclinical models rarely simulate the full neoadjuvant-surgery-monitoring sequence.
- Understanding perioperative disease progression is crucial for optimizing breast cancer therapy.
Purpose of the Study:
- To develop and validate a predictive mathematical model for neoadjuvant treatment response in breast cancer.
- To evaluate the efficacy of sunitinib, a receptor tyrosine kinase inhibitor (RTKI), in a preclinical model of spontaneous metastasis.
- To identify potential biomarkers for predicting individual patient kinetics and treatment outcomes.
Main Methods:
- Utilized a mouse model with spontaneous metastasis after primary tumor resection.
- Developed a novel mathematical kinetics-pharmacodynamics model validated with longitudinal data from 128 mice.
- Applied non-linear mixed-effects modeling and machine learning (SVM, random forests, ANN) to analyze tumor size, metastatic burden, and biomarker data.
Main Results:
- Computational simulations indicated neoadjuvant RTKI treatment inhibited primary tumor growth but showed limited efficacy against postsurgical (micro)-metastatic disease.
- Machine learning analyses failed to identify definitive biomarkers (circulating tumor cells, MDSCs, CD31, Ki67) for predicting individual kinetics or metastatic progression.
- Significant inter-animal variability in metastatic dynamics and survival was quantified.
Conclusions:
- Neoadjuvant RTKI therapy may not be sufficient to prevent metastatic spread after surgery in breast cancer.
- Preclinical mathematical modeling is valuable for identifying potential clinical trial failures early.
- The lack of predictive biomarkers highlights the complexity of perioperative breast cancer management and the need for improved therapeutic strategies.
