The Predictive Accuracy of Aortic Valve Calcium Score in Mediastinal Radiation Therapy Survivors

Mediastinal radiation is frequently used to treat patients with malignant diseases, including breast, esophageal, and lung cancers and lymphoma. However, radiation therapy can lead to long-term adverse cardiovascular events, including premature coronary artery disease, myocardial restriction, pericardial constriction, and valvular heart disease, with the aortic valve (AV) most commonly affected.¹

According to U.S. and European guidelines for the diagnosis and management of aortic stenosis (AS), AV calcium score, measured by Agatston units (AU) on cardiac computed tomography (CT), is recommended for risk stratification of patients with low-flow, low-gradient AS.^2,3 In patients with concordant echocardiographic findings of AS severity (severe, aortic valve area [AVA] <1.0 cm² and maximal velocity [Vmax] ≥4.0 m/s; or nonsevere, AVA ≥1.0 cm² and Vmax <4.0 m/s), an abnormal AV calcium score demonstrates excellent discrimination for assessing disease severity (C-statistic, women, 0.92; men, 0.89).⁴ Whether established calcium score thresholds apply to patients with radiation therapy–associated AS is unknown. This question arises from recent evidence demonstrating that in cancer survivors treated with mediastinal radiation, coronary artery calcium scores do not reliably predict acute myocardial ischemic events,⁵ and these scores have limited specificity for major adverse cardiovascular events.⁶ Therefore, we aimed to evaluate the accuracy of AV calcium scoring in survivors of mediastinal radiation.

This was a retrospective cohort study of adult (aged ≥18 years) cancer survivors who were previously treated with mediastinal radiation therapy and who had undergone transcatheter aortic valve replacement (TAVR) protocol CT for concordant severe AS (AVA <1.0 cm² and Vmax ≥4.0 m/s) at 2 university-affiliated hospitals from 2015 to 2024. A control group with concordant severe AS detected by echocardiography but without prior mediastinal radiation exposure was included for comparison. Patients with prior AV interventions were excluded. Electronic medical records were manually reviewed for demographic and imaging data. Because of the prolonged time interval between cancer treatment and AV intervention, radiation doses were often unavailable. Echocardiograms and TAVR CT images were reviewed for relevant measurements by experienced cardiovascular imagers, following established guidelines. The study was approved by the Institutional Review Board at each institution, and the requirement for informed consent was waived given its retrospective nature.

The study cohort was comprised of 97 patients with concordant severe AS determined by echocardiographic criteria; 63 of these patients had previous mediastinal radiation, and 34 did not. For radiation survivors, the median age was 73 years (Q1-Q3: 62–81 years), and most patients (n = 45, 71%) were female. Four patients (6.3%) had bicuspid AV anatomy. The most common underlying malignant diseases were Hodgkin disease (n = 28, 45%) and breast cancer (n = 24, 39%), and the median time interval between radiation therapy and TAVR CT was 25 years (Q1-Q3: 8–39 years). The median echocardiographic parameters were as follows: AVA, 0.7 cm² (Q1-Q3: 0.6–0.8 cm²); AV Vmax, 4.4 m/s (Q1-Q3: 4.1–4.6 m/s); peak AV gradient, 75 mm Hg (Q1-Q3: 68–83 mm Hg); mean AV gradient, 45 mm Hg (Q1-Q3: 41–53 mm Hg); left ventricular ejection fraction, 60% (Q1-Q3: 58%65%); and left ventricular stroke volume index, 39 mL/m² (Q1-Q3: 33–46 mL/m²). The demographic and echocardiographic characteristics of patients with and without prior mediastinal radiation were similar, except those without radiation had a wider AVA (AVA: 0.8 cm² [Q1-Q3: 0.7–0.9 cm²]; P = 0.025). The median AV calcium score was lower in patients with (1,715 AU [Q1-Q3: 1,313–2,317 AU]) vs without (2,827 AU [Q1-Q3: 1,683–3,423 AU]) previous mediastinal radiation (P < 0.001).

To evaluate the discriminatory ability of published AV calcium score cutoffs in this cohort of patients with echocardiographically concordant severe AS, we stratified scores using the established sex-specific thresholds for severe AS (1,300 AU in women and 2,000 AU in men).^3,4 We found that whereas most (70%) radiation survivors had abnormal AV calcium scores, 19 (30%) had scores below the defined cutoffs for severe AS (Figure 1). In contrast, only 1 (3%) patient without previous radiation had a score below the defined threshold (P < 0.001). These findings persisted when the European AV calcium score thresholds were applied (1,200 AU in women and 2,000 AU in men).² Notably, among radiation survivors, a similar proportion of male patients (6/18; 33%) and female patients (13/45; 29%) scores below the defined cutoffs for severe AS.

FIGURE 1. CT AV Calcium Scores in Patients With Severe Aortic Stenosis, Stratified by Previous Mediastinal Radiation Exposure.

Violin plots displaying AV calcium scores in patients with echocardiographically concordant severe AS, stratified by prior mediastinal radiation exposure. The y-axis shows AV calcium scores divided by the respective sex-specific thresholds for severe AS in the general population (1,300 Agatston units [AU] in women and 2,000 AU in men³). Values >1 represent AV calcium scores above the sex-specific threshold, and values <1 represent AV calcium scores below the sex-specific threshold. AS = aortic stenosis; AV = aortic valve.

Our findings suggest that AV calcium scores may be lower in mediastinal radiation survivors with severe AS compared with the general population.⁴ In the general population, severe AS results from progressive calcification of the AV leaflets, driven by age-related degenerative changes and atherosclerotic risk factors. In contrast, after mediastinal radiation, the AV cusps and leaflets undergo thickening and fibrosis, with or without calcification.¹ Therefore, radiation survivors may develop severe AS despite a lower calcification burden, and consequently, AV calcium scores may be less sensitive for the diagnosis of severe AS in these patients.

Despite the limited sample size of our cohort, these findings suggest a distinct pathophysiology in mediastinal radiation survivors that may lead to underestimation of AS severity by standard calcium scoring, thus highlighting the need for careful interpretation of imaging results in these patients.