Introduction

Low‑dose computed tomography (LDCT) of the chest is a widely used lung cancer screening tool in high‑risk populations.^1,2 However, LDCT also provides an opportunity to assess coronary artery calcification (CAC), emphysema, and other incidental findings.^3,4 The assessment of CAC is typically performed visually and has been simplified.³

Recent studies have confirmed a lower lung cancer mortality rate in patients undergoing thoracic LDCT screening. Therefore, since thoracic LDCT is now commonly used in clinical practice, it is crucial to consider its dual diagnostic value, which includes detecting lung cancer at an early stage and assessing cardiovascular risk in patients with a long history of smoking. In this context, we present unique prospective data investigating the value of the simple to calculate CAC parameter as a marker of coronary artery disease to predict both all‑cause and cardiovascular mortality in a group of patients undergoing LDCT screening due to long‑term smoking, within 6.8 years of follow‑up.⁵

The aim of this study was to investigate whether the visual detection of CAC in asymptomatic patients undergoing chest LDCT screening due to a history of smoking is an independent risk factor for mortality during long‑term follow‑up.

Patients and methods

The study population consisted of the participants of the Pilot Pomeranian Lung Cancer Screening Programme who underwent thoracic LDCT. The Department of Thoracic Surgery at the Medical University of Gdansk was responsible for the design, organization, and co‑ordination of the program. The program ran from February 2009 to April 2011 and included patients at a high risk for developing lung cancer. The inclusion criteria were age of 50–75 years, a history of smoking (at least 1 pack per day for 20 years), and no signs or symptoms of lung cancer. Other risk factors and comorbidities were not considered in the program.²

LDCT was performed in 1 of the 19 CT laboratories participating in the program. It involved a multi‑row CT scanner and the following exposure conditions: 120 kV, 50 mAs, 3‑second scan time, and a radiation dose of approximately 1–1.5 mSv. An ordinal scoring of calcification in the coronary arteries was applied. Four sections of the coronary arteries were identified on LDCT images: left main coronary artery, left anterior descending coronary artery, left circumflex artery, and right coronary artery. Each of the listed arteries was assigned a calcification severity category expressed as a number: 0 meant no calcification, 1 was slight calcification (less than one‑third of the vessel length), 2 denoted moderate calcification (more than one‑third but less than two‑thirds of the vessel length), and 3 meant heavy calcification (more than two‑thirds of the vessel length). Based on the ordinal analysis of the calcifications, the degree of CAC was expressed with the Coronary Artery Calcification score (CACs) ranging from 0 to 12 points. Four categories were defined after the initial analysis, encompassing 0, 1–3, 4–6, and 7–12 points, which were further investigated.

LDCT scans were assessed by experienced radiologists (with >5 years of training) following the same study protocol and detailed training. The agreement on the assessment of the coronary calcification degree carried out by 2 independent observers across the 4 CACs categories was high with unweighted κ of 0.86 (0.78, 0.95) and weighted κ of 0.92 (0.92, 0.92; P <⁠0.001). In the patients who, due to the assumptions of the Pilot Pomeranian Lung Cancer Screening, underwent thoracic LDCT more than once, only the CACs value from the first assessment was included in the analysis. The Central Statistical Office of Poland provided information on the causes of death in the study population in 3 categories: death due to malignant neoplasms (C00–C97 according to the International Classification of Diseases 10th Revision [ICD‑10]), death due to cardiovascular diseases (CVDs; I00–I99 according to ICD‑10), and death due to other causes. October 4, 2016, was assumed the follow‑up completion date.

The study was approved by the Institutional Review Board of the Medical University of Gdansk (NKEBN/109/2009). All patients provided their written informed consent to participate in the study.

Results

The study population consisted of all participants of the program (with an unknown medical history of other cardiovascular risk factors and CVDs) comprising 8535 individuals, of whom 4393 were men (51.5%). The mean (SD) age was 61.9 (6) years. Of all patients, 5403 were active smokers (63.3%). The median (IQR) smoking duration was 30 (25–40) years, the median number of pack‑years was 30 (22.5–40), and the median number of cigarettes smoked per day was 20 (20–25). The mean (SD) follow‑up duration was 6.8 (1) year. A total of 572 deaths (6.7%) were registered during the follow‑up period, including 159 cardiovascular deaths (1.9%). An analysis of the relationship between all‑cause mortality and CACs category showed that the death rate increased significantly with each CACs category, being 4.5% for the 0 category, 6.3% for the 1–3 category, 10% for the 4–6, and 17.2% for the 7–12 category (P <⁠0.001).

The incidence of cardiovascular death in each CACs category was 1%, 0.9%, 3.7%, and 8.4%, respectively (P <⁠0.001). The Cox regression HR analysis, using CACs category 0 as the reference group, showed that for the CACs 1–3 group, mortality from all causes was 1.4 times higher; for the CACs 4–6 group, it was 2.26 times higher; and for the CACs 7–12 group, it was 3.93 times higher than for the CACs category 0. Regarding cardiovascular deaths, the risk of death for CACs 4–6 was 3.7 times higher, and for CACs 7–12, it was 8.5 times higher than in the reference group CACs 0. There was no significant difference in the incidence of cardiovascular deaths between the CACs 0 and CACs 1–3 groups (Table 1). The multivariable logistic regression analysis showed that male sex, age, active smoking, total smoking time, and CACs are independent risk factors for 6‑year all‑cause mortality, while male sex, age, and CACs are independent risk factors for 6‑year cardiovascular mortality (n = 7835; Supplementary material, Tables S1–S3).

Discussion

While CACs is a widely used tool for cardiovascular risk assessment and has clear indications for evaluation during thoracic LDCT, as recommended by several scientific committees, its predictive value in asymptomatic patients undergoing LDCT for lung cancer screening has not been well established.³ Our study focused on this parameter, and its novel aspect is demonstrating that CACs, measured during a routine thoracic LDCT screening, can be a valuable predictor of both all‑cause and cardiovascular mortality, independent of other factors, such as sex, age, or smoking history.

The ability to effectively assess the progression of the atherosclerotic process in patients with CVD risk factors, and to determine cardiovascular risk based on this, enables a consistent implementation of primary prevention strategies and a healthy lifestyle in individuals at a high risk of death.⁵

It should be noted that the results of our study are not applicable to the general population. The study was conducted in an age‑restricted cohort with specific risk factors for lung cancer. Therefore, generalizing these findings to broader populations may not be entirely justified.