Background Body CT scans are frequently done for a wide range of clinical indications, but potentially valuable biometric information typically goes unused. We aimed to compare the prognostic ability of automated CT-based body composition biomarkers derived from previously developed deep-learning and feature-based algorithms with that of clinical parameters (Framingham risk score [FRS] and body-mass index [BMI]) for predicting major cardiovascular events and overall survival in an adult screening cohort.Methods In this retrospective cohort study, mature and fully automated CT-based algorithms with predefined metrics for quantifying aortic calcification, muscle density, ratio of visceral to subcutaneous fat, liver fat, and bone mineral density were applied to a generally healthy asymptomatic outpatient cohort of adults aged 18 years or older undergoing abdominal CT for routine colorectal cancer screening. To assess the association between the predictive measures (CT-based vs FRS and BMI) and downstream adverse events (death or myocardial infarction, cerebrovascular accident, or congestive heart failure subsequent to CT scanning), we used both an event-free survival analysis and logistic regression to compute receiver operating characteristic curves (ROCs) . Findings 9223 people (mean age 57•1 years [SD 7•8]; 5152 [56%] women and 4071 [44%] men) who underwent CT scans between April, 2004, and December, 2016, were included in this analysis. In the longitudinal clinical follow-up (median 8•8 years [IQR 5•1-11•6]), subsequent major cardiovascular events or death occurred in 1831 (20%) patients. Significant differences were observed for all five automated CT-based body composition measures according to adverse events (p<0•001). Univariate 5-year area under the ROC (AUROC) values for predicting death were 0•743 (95% CI 0•705-0•780) for aortic calcification, 0•721 (0•683-0•759) for muscle density, 0•661 (0•625-0•697) for ratio of visceral to subcutaneous fat, 0•619 (0•582-0•656) for liver density, and 0•646 (0•603-0•688) for vertebral density, compared with 0•499 (0•454-0•544) for BMI and 0•688 (0•650-0•727) for FRS.Univariate hazard ratios for highestrisk quartile versus others for these same CT measures were 4•53 (95% CI 3•82-5•37) for aortic calcification, 3•58 (3•02-4•23) for muscle density, 2•28 (1•92-2•71) for the ratio of visceral to subcutaneous fat, 1•82 (1•52-2•17) for liver density, and 2•73 (2•31-3•23) for vertebral density, compared with 1•36 (1•13-1•64) for BMI and 2•82 (2•36-3•37) for FRS. Multivariate combinations of CT biomarkers further improved prediction over clinical parameters (p<0•05 for AUROCs). For example, the 2-year AUROC from combining aortic calcification, muscle density, and liver density for predicting death was 0•811 (95% CI 0•761-0•860).Interpretation Fully automated quantitative tissue biomarkers derived from CT scans can outperform established clinical parameters for presymptomatic risk stratification for future serious adverse events and add opportunistic value to CT scans performed fo...

We sought to determine if vertebral trabecular attenuation values measured on routine body computed tomography (CT) scans obtained for a variety of unrelated indications can predict future osteoporotic fractures at multiple skeletal sites. For this Health Insurance Portability and Accountability Act (HIPAA)-compliant and Institutional Review Board (IRB)-approved retrospective cohort study, trabecular attenuation of the first lumbar vertebra was measured in 1966 consecutive older adults who underwent chest and/or abdominal CT at a single institution over the course of 1 year. New pathologic fragility fractures that occurred after a patient's CT study date were identified through an electronic health record database query using International Classification of Diseases (ICD)-9 codes for vertebral, hip, and extremity fractures. Univariate and multivariate Cox proportional hazards regression were performed to determine the effect of L trabecular attenuation on fracture-free survival. Age at CT, sex, and presence of a prior fragility fracture were included as confounders in multivariate survival analysis. Model discriminative capability was assessed through calculation of an optimism-corrected concordance index. A total of 507 patients (mean age 73.4 ± 6.3 years; 277 women, 230 men) were included in the final analysis. The median post-CT follow-up interval was 5.8 years (interquartile range 2.1-11.0 years). Univariate analysis showed that L attenuation values ≤90 Hounsfield units (HU) are significantly associated with decreased fracture-free survival (p < 0.001 by log-rank test). After adjusting for age, sex, prior fracture, glucocorticoid use, bisphosphonate use, chronic kidney disease, tobacco use, ethanol abuse, cancer history, and rheumatoid arthritis history, multivariate analysis demonstrated a persistent modest effect of L attenuation on fracture-free survival (hazard ratio [HR] = 0.63 per 10-unit increase; 95% confidence interval [CI] 0.47-0.85). The model concordance index was 0.700. Ten-year probabilities for major osteoporosis-related fractures straddled the treatment threshold for most subcohorts over the observed L HU range. In conclusion, for patients undergoing body CT scanning for any indication, L vertebral trabecular attenuation is a simple measure that, when ≤90 HU, identifies patients with a significant decrease in fracture-free survival. © 2018 American Society for Bone and Mineral Research.

Content codes:Purpose: To determine if abdominal aortic calcification (AAC) at CT predicts cardiovascular events independent of Framingham risk score (FRS). Materials and Methods:For this retrospective study, electronic health records for 829 asymptomatic patients (mean age, 57.9 years; 451 women, 378 men) who underwent nonenhanced CT colonography screening between April 2004 and March 2005 were reviewed for subsequent cardiovascular events; mean follow-up interval was 11.2 years 6 2.8 (standard deviation). Institutional review board approval was obtained. CT-based AAC was retrospectively quantified as a modified Agatston score by using a semiautomated tool. Kaplan-Meier curves and Cox proportional hazards models were used for time-to-event analysis; receiver operating characteristic curves and net reclassification improvement compared predictive abilities of AAC and FRS.Results: An index cardiovascular event occurred after CT in 156 (19%) of 829 patients (6.7 years 6 3.5, including heart attack in 39 [5%] and death in 79 [10%]). AAC was higher in the cardiovascular event cohort (mean AAC, 3478 vs 664; P , .001). AAC was a strong predictor of cardiovascular events at both univariable and multivariable Cox modeling, independent of FRS (P , .001). Kaplan-Meier plots showed better separation with AAC over FRS. The area under the receiver operating characteristic curve (AUC) was higher for AAC than FRS at all evaluated time points (eg, AUC of 0.82 vs 0.64 at 2 years; P = .014). By using a cutoff point of 200, AAC improved FRS risk categorization with net reclassification improvement of 35.4%. Conclusion:CT-based abdominal aortic calcification was a strong predictor of future cardiovascular events, outperforming the Framingham risk score. This finding suggests a potential opportunistic role in abdominal nonenhanced CT scans performed for other clinical indications.

To assess the feasibility of four-dimensional (4D) flow MRI as a noninvasive imaging marker for stratifying the risk of variceal bleeding in patients with liver cirrhosis. Materials and Methods: This study recruited participants scheduled for both liver MRI and gastroesophageal endoscopy. Risk of variceal bleeding was assessed at endoscopy by using a three-point scale: no varices, low risk, and high risk requiring treatment. Fourdimensional flow MRI was used to create angiograms for evaluating visibility of varices and to measure flow volumes in main portal vein (PV), superior mesenteric vein, splenic vein (SV), and azygos vein. Fractional flow changes in PV and SV were calculated to quantify shunting (outflow) from PV and SV into varices. Logistic analysis was used to identify the independent indicator of highrisk varices. Results: There were 23 participants (mean age, 52.3 years; age range, 25-75 years), including 14 men (mean age, 51.7 years; age range, 25-75 years) and nine women (mean age, 53.2 years; age range, 31-72 years) with no varices (n = 8), low-risk varices (n = 8), and high-risk varices (n = 7) determined at endoscopy. Four-dimensional flow MRI-based angiography helped radiologists to view varices in four of 15 participants with varices. Independent indicators of high-risk varices were flow volume in the azygos vein greater than 0.1 L/min (P = .034; 100% sensitivity [seven of seven] and 62% specificity [10 of 16]) and fractional flow change in PV of less than 0 (P , .001; 100% sensitivity [seven of seven] and 94% specificity [15 of 16]). Conclusion: Azygos flow greater than 0.1 L/min and portal venous flow less than the sum of splenic and superior mesenteric vein flow are useful markers to stratify the risk of gastroesophageal varices bleeding in patients with liver cirrhosis.

No abstract