A Scoring System for Predicting Neoadjuvant Chemotherapy Response in Primary High-Grade Bone Sarcomas: A Multicenter...

Area of Science:

• Oncology
• Radiology
• Medical Informatics

Background:

• Clinical tools for assessing neoadjuvant chemotherapy response in high-grade bone sarcomas are lacking.
• Accurate prediction of chemotherapy response is crucial for treatment planning and patient prognosis.

Purpose of the Study:

• To investigate the predictive value of clinical findings for chemotherapy response in primary high-grade bone sarcomas.
• To develop and validate a scoring system for predicting chemotherapy response.

Main Methods:

• Retrospective multicenter cohort study of 322 patients with primary high-grade bone sarcomas.
• Utilized machine learning models (logistic regression, decision trees, SVM, neural networks) to analyze clinical data including imaging (X-ray, CT, MR, PET-CT) and laboratory values.
• Developed a scoring system based on identified significant predictors of good histological response.

Main Results:

• A scoring system incorporating longest diameter reduction, bone boundary changes, tumor necrosis on MR, SUVmax decrease, and alkaline phosphatase decrease was developed.
• A score ≥4 accurately predicts a good response to chemotherapy, achieving an Area Under the Curve (AUC) of 0.893.
• The scoring system demonstrated high predictive performance, with an AUC of 0.901 for nonmeasurable lesions.

Conclusions:

• A novel, comprehensive scoring system has been developed to predict neoadjuvant chemotherapy response in primary high-grade bone sarcomas.
• This scoring system offers a valuable clinical tool for evaluating treatment efficacy.
• The findings support improved patient management through personalized chemotherapy response prediction.