Preoperative Cardiac Risk Stratification in Dogs with Mammary Tumors Using Two-Dimensional Speckle Tracking...

Area of Science:

• Veterinary Cardiology
• Cardio-oncology
• Comparative Medicine

Background:

• Subclinical left ventricular dysfunction is linked to breast cancer in humans.
• Cardiac effects of chemotherapy-induced myocardial toxicity (CMTs) in dogs are not well understood.
• Early detection of cardiac changes is crucial for managing cancer treatment side effects.

Purpose of the Study:

• To compare echocardiographic parameters in dogs with CMTs versus healthy controls.
• To assess the feasibility of combining myocardial deformation imaging with data-driven analysis for preoperative cardiac assessment.
• To explore the potential of machine learning for identifying subtle cardiac alterations.

Main Methods:

• Pilot study involving clinical and echocardiographic assessments, including 2D speckle-tracking echocardiography (2D-STE).
• Exploratory machine learning (XGBoost), ROC analysis, calibration, and decision curve analysis were used.
• Limited sample size necessitated hypothesis-generating analyses without external validation.

Main Results:

• Dogs with CMTs showed reduced global longitudinal strain (GLS) and mitral annular plane systolic excursion (MAPSE) despite normal conventional systolic indices.
• Deformation imaging parameters were more sensitive in detecting subtle myocardial changes.
• The exploratory machine learning model achieved moderate discrimination (AUC-ROC = 0.75), but findings are preliminary.

Conclusions:

• Conventional systolic indices may underestimate early myocardial changes in dogs with CMTs.
• Myocardial deformation imaging integrated with exploratory analysis is feasible for veterinary cardio-oncology.
• Further large-scale, validated studies are warranted to confirm these preliminary findings.