Predicting Tolerance to Anthracycline Chemotherapy Using Electrocardiogram-Based Artificial Intelligence in Sarcoma.

Area of Science:

• Cardiology
• Oncology
• Artificial Intelligence

Background:

• Anthracycline chemotherapy can cause cardiotoxicity, necessitating cardiac monitoring.
• Assessing chemotherapy tolerance and cardiac function is crucial for patient management.

Purpose of the Study:

• To evaluate the utility of artificial intelligence-enabled electrocardiogram (AI-ECG) in assessing anthracycline chemotherapy tolerance.
• To determine if AI-ECG can predict cardiac function and monitor changes during chemotherapy.

Main Methods:

• Adult sarcoma patients receiving anthracycline chemotherapy with pre-treatment ECGs were analyzed.
• AI-ECG nomograms predicted age and ejection fraction (EF); changes correlated with chemotherapy tolerance.
• Sensitivity and specificity of AI-ECG for predicting low EF (<50% or <35%) were measured.

Main Results:

• AI-ECG accurately identified patients with low EF (100% sensitivity, 94% specificity for EF <50%).
• A trend suggested higher dose reductions with increased AI-ECG aging (OR 5.13, P=.32).
• Two patients with decreased EF post-treatment showed significant increases in AI-ECG low EF prediction.

Conclusions:

• AI-ECG predictions on ECGs may monitor EF decreases during anthracycline chemotherapy.
• Further research is recommended to validate AI-ECG aging as a marker for chemotherapy tolerance.