Fully automated calcium scoring predicts all-cause mortality at 12 years in the MILD lung cancer screening trial

• DOI: 10.1371/journal.pone.0285593
• Publication Date: 2023-05-16T18:08:21Z
• Authors: Federica Sabia; Maurizio Balbi; Roberta E. Ledda; Gianluca Milanese; Margherita Ruggir...; Camilla Valsecchi; Alfonso Marchianò; Nicola Sverzellati; Ugo Pastorino
• Keywords: Lung Neoplasms; Science; Q; R; 610; Coronary Artery Disease; Risk Assessment; 3. Good health; Artificial Intelligence; Risk Factors; Medicine; Humans; Calcium; Vascular Calcification; Early Detection of Cancer; Research Article

Coronary artery calcium (CAC) is a known risk factor for cardiovascular (CV) events and mortality but not yet routinely evaluated in low-dose computed tomography (LDCT)-based lung cancer screening (LCS). The present analysis explored the capacity of fully automated CAC scoring to predict 12-year Multicentric Italian Lung Detection (MILD) LCS trial. study included 2239 volunteers MILD trial who underwent baseline LDCT from September 2005 January 2011, with median follow-up 190 months. score was measured by commercially available artificial intelligence (AI) software stratified into five strata: 0, 1-10, 11-100, 101-400, > 400. Twelve-year all-cause 8.5% (191/2239) overall, 3.2% = 4.9% 8.0% 11.5% 17% In Cox proportional hazards regression analysis, 400 associated higher both univariate model (hazard ratio, HR, 5.75 [95% confidence interval, CI, 2.08-15.92] compared 0) after adjustment confounders (HR, 3.80 [95%CI, 1.35-10.74] 0). All-cause significantly increased increasing (7% ≤ vs. 400, Log-Rank p-value <0.001). Non-cancer at 12 years 3% (67/2239) 0.8% 1.0% 2.9% 3.6% 8.2% (Grey's test p < 0.001). Fine Gray's competing model, predicted non-cancer (sub-distribution hazard SHR, 10.62 1.43-78.98] 0), association no longer significant confounders. conclusion, effective predicting setting.