Association Between Automated Coronary Artery Calcium From Routine Chest Computed Tomography Scans and Cardiovascular...

Area of Science:

• Cardiology
• Oncology
• Radiology
• Medical Imaging Analysis

Background:

• Cardiovascular disease (CVD) is a leading cause of mortality in colorectal or gastric cancer patients.
• Identifying patients at increased CVD risk is crucial for this population.
• Coronary artery calcium (CAC) scoring is an established CVD predictor, but its utility in cancer patients is under-evaluated.

Purpose of the Study:

• To evaluate the association between automated CAC scoring using chest CT and atherosclerotic CVD risk.
• To assess the prognostic value of CAC in colorectal or gastric cancer patients.

Main Methods:

• Retrospective cohort study of colorectal or gastric cancer patients (2011-2019).
• Automated CAC quantification using deep learning software on non-ECG-gated chest CT scans.
• Classification into four CAC groups (0, 0-100, 100-400, >400) and assessment of major adverse cardiovascular events (MACE) using Fine and Gray models.

Main Results:

• 3153 patients included; mean follow-up 4.1 years.
• MACE incidence rates increased with higher CAC scores.
• CAC > 400 was associated with a 2.33-fold higher risk of MACE compared to CAC=0 (aOR, 2.33 [95% CI, 1.24-4.39]).

Conclusions:

• Elevated CAC scores (CAC>400) on routine chest CT scans are linked to increased MACE risk in cancer patients.
• Automated CAC scoring offers valuable prognostic information for atherosclerotic CVD risk stratification in this cohort.