Accurate measurement of coronary artery calcium in cancer patients using the CT component of PET/CT scans.

Objective: The aim of this study is to establish the correlation between a CAC score derived from the CT component of PET/CT scan (CAC-PET) using in-house software as compared to the conventional technique (CAC-Standard). In addition, the incidence of high CAC scores in asymptomatic cancer patients with low-to-intermediate cardiovascular risk will be determined.

Methods: 100 patients referred for oncologic PET/CT were prospectively recruited to have a conventional CAC score after their PET/CT. Patients with a history of cardiac disease were excluded. The nongated CT images from the PET/CT (CAC-PET) were analysed using validated in-house software with the results compared to those from gated CT analysed using the standard technique (CAC-Standard).

Results: The correlation of CAC scores between the two scan types was moderate [slope, 0.95; R2 = 0.91; limits of agreement (LOA) = 0.29-5.65]. Using a conventional categorical analysis, there was complete agreement in 73% of patients with one category difference in the remainder. [interclass correlation (ICC) = 0.90; Cohen's kappa = 0.63]. In total 28% of these asymptomatic low-to-intermediate-risk cancer patients had CAC scores over 300.

Conclusion: Estimation of CAC from the CT component of PET/CT scans is a reliable method for the detection of significant CAC in cancer patients and correlates well with the standard method. This technique should permit the calculation of cardiovascular risk in cancer patients undergoing PET/CT without any additional radiation exposure. A significant number of asymptomatic low-to-intermediate-risk cancer patients were found to have a high risk of cardiovascular disease.