Prognostic value of coronary artery calcium screening in asymptomatic smokers and non-smokers

Abstract

Aims: To determine the extent and prognostic significance of coronary artery calcium in asymptomatic smokers and non-smokers. Population data are available on the prognostic impact of smoking on atherosclerotic imaging measurements of the carotid and peripheral arteries. Limited data are available on the impact of cigarette smoking on the prognostic value of coronary calcium.

Methods and results: A referred patient registry of 10,377 asymptomatic individuals (40% were current smokers) was followed for death from all-causes at 5 years. Univariable and multivariable Cox proportional hazard models were calculated to estimate time to all-cause mortality. Cumulative 5-year survival was 96.9 and 98.4% for smokers when compared with non-smokers (P < 0.0001). Using a stratified Cox proportional hazards survival analysis, survival for non-smokers ranged from 99.7 to 89.6% with calcium score of 0-10 and >1000 (P < 0.0001). In comparison, smokers had survival rates ranging from 99.5 to 81.4% for calcium score of 0-10 to >1000 (P < 0.0001). When further evaluating the effect of age on prognosis by coronary calcium, there was an additive relationship between age and calcium that was exacerbated with smoking, resulting in higher relative risk ratios for older smokers with coronary calcium (P < 0.0001). For smokers <50 years of age, a calcium score >1000 was associated with a relative risk ratio that was elevated 8.9-fold (P = 0.029). Thus, resulting in an expected reduction in life expectancy of 4.8 years for smokers <50 years of age with a calcium score >400 (P < 0.0001).

Conclusion: The prognostic value of coronary artery calcium scoring was accurate in identifying a high-risk cohort of asymptomatic smokers and non-smokers. Young smokers with high-risk calcium scores have a four- to nine-fold increased risk of dying when compared with similarly aged non-smokers. When prospectively applied, evidence of a high-risk calcium score may be useful in educating patients as to their expected risk of dying over the next 5 years.