Visual ordinal grading of aortic valve calcification on routine non-gated chest CT predicts prognosis and alters management

Objective BSCI/BSTI guidelines recommend reporting aortic valve calcification (AVC) on all chest CTs regardless of indication. We assessed AVC frequency, severity, and association with aortic stenosis (AS) on echocardiography and its prognostic implications. Methods Retrospective, single-centre analysis of consecutive chest CTs (January-December 2015) for 200 patients per age group (< 40, 40-49, 50-59, 60-69, 70-79, 80-89, >= 90) performed for medical, surgical, and oncological indications. CTs were re-reviewed for the presence and graded severity of AVC and coronary artery calcification (CAC). Corresponding echocardiography reports (within 5 years) reviewed for AS. Comorbidities and clinical outcomes were recorded. Results One thousand three hundred seventy-seven patients were included (mean age 64 +/- 20 years, 55% female). AVC was present in 25% (350/1377) and was more prevalent in males (p < 0.001). Frequency and severity increased with age (p < 0.001). 38% (524/1377) had an echocardiogram (median inter-test interval 4.3 months [IQR 0.4-17.5]). Sixteen per cent (29/178) with AVC had AS of any severity (8% [15/178] mild; 8% [14/178] moderate; 0% [0/178] severe). Sensitivity and specificity for AVC predicting AS were 91% and 70%, respectively. Extrapolating findings, 8% of individuals with AVC and without an echocardiogram may have undiagnosed AS. All-cause mortality occurred in 53% (734/1377), which AVC predicted independently of CAC and age (p < 0.001). Adjusting for confounders, severe AVC predicted all-cause mortality (HR 1.56 [1.10-2.22], p = 0.013). Conclusions AVC identified AS in 16% of patients. Additionally, severe AVC is an independent predictor of all-cause mortality in multivariable analysis. Validation in a prospective cohort is required to inform clinical practice guidelines.