br Framingham risk or CAC
4.2. Framingham risk or CAC which is preferred?
In the absence of a universal clinical risk tool that predicts Jasplakinolide failure, atrial fibrillation/arrhythmias and atherosclerosis, risk
stratifiers like the FRS will continue to be used to determine cardiac risk in clinical settings including cardiac oncology. Clinical risk sys-tems such as FRS are the guideline preferred method of stratifying patients for atherosclerotic disease rather than using a cardiac CT derived CAC score [5,20,21]. According to primary prevention guide-lines there are however certain restricted circumstances when CAC assessment could be considered and these are defined by the initial clinical risk assessment [5,20,21]. Thus to be concordant with cur-rent primary prevention guidelines, CAC assessment can be used in conjunction with FRS (or alternative clinical stratification system) but not as a replacement . Co-incidental CAC assessment at CT chest probably should be considered in the same manner. In the presence of a high clinical risk score absence of CAC does not normal-ize cardiovascular risk and primary prevention should be continued. Coincidental detection of CAC in low risk or intermediate cardiovas-cular risk patients might however prompt statin therapy depending on the extent of CAC [2,5]. Our sample size did not allow calculation of a net reclassification index to evaluate the impact of CAC detection on statin eligibility. Nevertheless 7% of low FRS patients and 45% of intermediate to high FRS patients were found to have CAC and were not on a statin which based on the results of our study might be considered.
Fig. 2. Framingham risk score and cancer stage adjusted Cox proportional regression analysis for cardiac events (cardiac mortality, coronary revascularization, heart failure hospitalization, new atrial fibrillation). The presence of coronary artery calcium (CAC) red line was associated with reduced event free survival (77% at ten years) in comparison to patients without CAC (blue line) (93% at ten years) (p = 002).
The majority of the cardiac events in the current study were coronary revascularization and heart failure events (there were very few episodes of new atrial fibrillation). The utility of FRS to predict heart failure events in breast cancer patients may be limited . Furthermore it is possible that FRS may underestimate coronary athero-sclerosis in breast cancer patients due to confounding influences such as the impact of cancer therapy or due to the influence of cancer on athero-sclerosis. In the initial Framingham cohort cancer diagnoses were present in 0.6% of subjects before study entry and the inadvertent under subscription of cancer patients may have limited the utility of FRS in this population . More recently Gernaat et al. demonstrated in 1103 breast cancer patients that FRS underestimates cardiovascular events in comparison to age and gender matched controls . In view of these potential limitations of FRS, our data do not argue for an aban-donment of traditional clinical risk scoring systems, rather that CAC detection on chest CT could be used alongside clinical risk scores to augment cardiac risk prediction.
4.3. Impact of CAC on survival analysis
Cardiac event free survival at 10 years was 93% in patients with no CAC versus 77% in those with CAC (Fig. 2). Coronary revascularization and heart failure hospitalizations accounted for the majority of cardiac events. There were only 2 cases of cardiac mortality, both cases occurred in patients with CAC. Non-cardiac death occurred in 83 patients (32.4%) during follow up. CAC was not associated with non-cardiac death which, for the majority of cases, was due to a cancer related cause (86%). Advanced cancer stage (Stage III–IV versus I–II) was associated with the composite end point that included all-cause mortality. Inclusion of all-cause mortality in the survival analysis reduced the event free survival time for both patients with CAC and no CAC (47% versus 64% respectively at 10 years) (Fig. 1). Despite the reduced event free survival with inclusion of all-cause mortality, CAC was predictive of survival whereas FRS was not. CAC therefore remains a significant risk factor to predict event free survival even in the context of a diagnosis of breast cancer. This is consistent with prior data demonstrating the importance of cardiovas-cular disease to determine prognosis in breast cancer patients .
This was a single center, retrospective study of consecutive breast cancer patients attending a cardiac oncology clinic and the applicability of the findings to an unselected breast cancer population may be lim-ited. To detect coronary calcification we utilized reconstructed sections from varying widths as per clinical protocols. In original descriptions, CAC scoring was performed using 3 mm section widths, section widths greater than this may be less sensitive, but specificity is maintained at 90% across the section widths used in this study .