Response to “Consideration of Statin Therapy in the Interpretation of Artificial Intelligence-Derived Coronary Artery Calcium Density”

We thank Dr. Liu [1] for thoughtful comments on our study [2] and for highlighting the importance of statin therapy in interpreting coronary artery calcium (CAC) density. We agree that the lack of detailed information on lipid-lowering therapies, particularly statin use, limits our study and may have influenced the interpretation of the CAC density.

However, the relationship between statin therapy, CAC density, and cardiovascular outcomes is complex, and the direction of potential bias is not straightforward. Although detailed information regarding statin type, dose, duration, and treatment intensity was unavailable, dyslipidemia was included in our multivariable models and was defined as a prior diagnosis of dyslipidemia or the current use of lipid-lowering medication. Therefore, lipid-related risks, and treatment exposure were included in our analyses.

We agree that statin therapy may promote plaque densification and increase the CAC density. However, if the inverse association between CAC density and cardiovascular events is primarily driven by statin-related confounding factors, this inverse association may be strongest among statin-treated individuals. Conversely, in a CAC Consortium study by Osei et al. [3], CAC density was associated with cardiovascular and coronary heart disease mortality among statin non-users, whereas this association was not evident among statin users. Similarly, in a recent systematic review and meta-analysis, the inverse association between CAC density and cardiovascular events was most evident among statin-naïve individuals [4]. Bhatia et al. [5] demonstrated that the inverse association between CAC density and coronary heart disease risk persisted among statin-naïve participants in the multi-ethnic study of atherosclerosis, and no significant interaction was observed between statin use and CAC density. Collectively, these findings suggest that CAC density is unlikely to function simply as a surrogate marker of statin exposure but may instead reflect the natural evolution and biological stability of calcified plaques. Indeed, treatment-related plaque densification may attenuate or obscure the biological significance of CAC density, rather than exaggerate its inverse association with cardiovascular risk.

We noted that the relationship between CAC and statin therapy may be bidirectional in contemporary clinical practice. As Dr. Liu noted [1], statin therapy may influence the characteristics of CAC. Conversely, the NOTIFY-1 trial demonstrated that systematic notification of incidental CAC identified on nongated chest CT significantly increased statin initiation and improved low-density lipoprotein cholesterol reduction [6]. Therefore, CAC findings may influence preventive treatment decisions. In this context, statin exposure may represent a potential confounder and a consequence of CAC detection and cardiovascular risk assessment.

Therefore, although unmeasured statin therapy remains an important limitation of our study, the current evidence does not suggest that statin therapy explains or exaggerates the inverse association between CAC density and major adverse cardiovascular events observed in our cohort. Future studies linking chest CT-derived CAC phenotypes with longitudinal prescription data, such as those available through Korea’s National Health Insurance Service database, would allow a more accurate characterization of statin exposure and help determine whether the inverse association between CAC density and cardiovascular events persists after accounting for lipid-lowering therapy. Furthermore, longitudinal studies incorporating serial CT examinations may provide insight into the effects of statin therapy on CAC volume and density and help distinguish treatment-related plaque densification from the natural evolution of calcified plaques.

We thank Dr. Liu for highlighting this important point [1]. We agree that statin therapy should be carefully considered in future investigations on CAC density. As artificial intelligence-based quantification of CAC volume and density becomes increasingly available in routine chest CT practice, the integration of imaging biomarkers with longitudinal treatment data is essential to distinguish the effects of plaque biology from those of preventive therapy and optimize the clinical application of CAC density for cardiovascular risk stratification.