Dual-Energy CT muscle fat fraction as a new imaging biomarker of body composition and survival predictor in critically ill patients.

OBJECTIVE: To analyze changes in the muscular fat fraction (FF) during immobilization at the intensive care unit (ICU) using dual-energy CT (DECT) and evaluate the predictive value of the DECT FF as a new imaging biomarker for morbidity and survival. METHODS: Immobilized ICU patients (n = 81, 43.2% female, 60.3 ± 12.7 years) were included, who received two dual-source DECT scans (CT1, CT2) within a minimum interval of 10 days between 11/2019 and 09/2022. The DECT FF was quantified for the posterior paraspinal muscle by two radiologists using material decomposition. The skeletal muscle index (SMI), muscle radiodensity attenuation (MRA), subcutaneous-/ visceral adipose tissue area (SAT, VAT), and waist circumference (WC) were assessed. Reasons for ICU admission, clinical scoring systems, therapeutic regimes, and in-hospital mortality were noted. Linear mixed models, Cox regression, and intraclass correlation coefficients were employed. RESULTS: Between CT1 and CT2 (median 21 days), the DECT FF increased (from 20.9% ± 12.0 to 27.0% ± 12.0, p = 0.001). The SMI decreased (35.7 cm2/m2 ± 8.8 to 31.1 cm2/m2 ± 7.6, p < 0.001) as did the MRA (29 HU ± 10 to 26 HU ± 11, p = 0.009). WC, SAT, and VAT did not change. In-hospital mortality was 61.5%. In multivariable analyses, only the change in DECT FF was associated with in-hospital mortality (hazard ratio (HR) 9.20 [1.78-47.71], p = 0.008), renal replacement therapy (HR 48.67 [9.18-258.09], p < 0.001), and tracheotomy at ICU (HR 37.22 [5.66-245.02], p < 0.001). Inter-observer reproducibility of DECT FF measurements was excellent (CT1: 0.98 [0.97; 0.99], CT2: 0.99 [0.96-0.99]). CONCLUSION: The DECT FF appears to be suitable for detecting increasing myosteatosis. It seems to have predictive value as a new imaging biomarker for ICU patients. CLINICAL RELEVANCE STATEMENT: The dual-energy CT muscular fat fraction appears to be a robust imaging biomarker to detect and monitor myosteatosis. It has potential for prognosticating, risk stratifying, and thereby guiding therapeutic nutritional regimes and physiotherapy in critically ill patients. KEY POINTS: The dual-energy CT muscular fat fraction detects increasing myosteatosis caused by immobilization. Change in dual-energy CT muscular fat fraction was a predictor of  in-hospital morbidity and mortality. Dual-energy CT muscular fat fraction had a predictive value superior to established CT body composition parameters.

Fat quantification in dual-layer detector spectral CT: How to handle iron overload, varying tube voltage and radiation dose Indices.

OBJECTIVES: Opposed to other spectral CT techniques, fat quantification in dual-layer detector CT (dlCT) has only recently been developed. The impact of concomitant iron overload and dlCT-specific protocol settings such as the dose right index (DRI), a measure of image noise and tube current, on dlCT fat quantification was unclear. Further, spectral information became newly available <120 kV. Therefore, this study's objective was to evaluate the impact of iron, changing tube voltage, and DRI on dlCT fat quantification. MATERIAL AND METHODS: Phantoms with 0 and 8mg/cm3 iron; 0 and 5mg/cm3 iodine; 0, 10, 20, 35, 50, and 100% fat and liver equivalent, respectively, were scanned with a dlCT (CT7500, Philips, the Netherlands) at 100kV/20DRI, 120kV/20DRI, 140kV/20DRI, and at 120kV/16DRI, 120kV/24DRI. Material decomposition was done for fat, liver, and iodine (A1); for fat, liver, and iron (A2); and for fat, liver, and combined reference values of iodine and iron (A3). All scans were analyzed with reference values from 120kV/20DRI. For statistics, the intraclass correlation coefficient (ICC) and Bland-Altman analyses were used. RESULTS: In phantoms with iron and iodine, results were best for A3 with a mean deviation to phantom fat of 1.3±2.6% (ICC 0.999 [95%-confidence interval 0.996-1]). The standard approach A1 yielded a deviation of -2.5±3.0% (0.998[0.994-0.999]), A2 of 6.1±4.8% (0.991[0.974-0.997]). With A3 and changing tube voltage, the maximal difference between quantified fat and the phantom ground truth occurred at 100kV with 4.6±2.1%. Differences between scans were largest between 100kV and 140kV (2.0%[-7.1-11.2]). The maximal difference of changing DRI occurred between 16 and 24 DRI with 0.4%[-2.2-3.0]. CONCLUSION: For dlCT fat quantification in the presence of iron, material decomposition with combined reference values for iodine and iron delivers the most accurate results. Tube voltage-specific calibration of reference values is advisable while the impact of the DRI on dlCT fat quantification is neglectable.

Limb Occlusion Rate after EVAR With Individualized Graft Limb Selection and a Liberal Protocol of Primary Relining.

BACKGROUND: This study analyses limb occlusion rates after endovascular aneurysm repair (EVAR) with a strategy including stent-graft limb selection and liberal primary stenting (relining) in anatomies at higher risk for occlusion with uncovered self-expandable or balloon-expandable stents. METHODS: All patients undergoing elective EVAR with a bifurcated stent-graft between January 2010-August 2018 were included. A protocol involving personalized stent-graft selection and liberal primary relining based on preoperative imaging was followed during the whole period. Primary endpoints were technical success and primary limb patency during follow-up. Secondary endpoints included mortality and limb reintervention rates. Risk factors associated to limb patency and reintervention rates were analyzed. RESULTS: Six hundred and fifteen patients (548 males; mean age 72.9 ± 9 years) were included. Overall technical success was 98.5% (606/615). One (0.16%) patient died during the first 30 days. Of the 1230 limbs, 96 (8%) were deemed at risk for occlusion in view of the anatomy, and primarily relined in 62 patients. Estimated primary limb patency at 6 months, 1 year and 3 years was 99.5 ± 0.2%, 99.2 ± 0.3% and 98.5 ± 0.5%, respectively. Freedom from limb-related reintervention at 6 months, 1 year and 3 years was 98.1 ± 0.4%, 97.4 ± 0.5% and 95.6 ± 0.7%, respectively. Only one (1%) of the 96 relined limbs occluded during follow-up. No differences were found in terms of patency or freedom from reintervention between limbs at risk that were primary stented and limbs without adjunctive stents. Gore Excluder stent-grafts presented better patency (Breslow P = 0.005) and lower reintervention rates (Breslow P = 0.001) than other devices during follow-up. Peripheral artery disease was also a risk factor for reintervention (Breslow P = 0.015). CONCLUSION: Liberal use of primary limb relining in patients with iliac anatomy at higher risk for occlusion appears to be a safe and effective strategy to preserve limb patency after EVAR. Gore Excluder graft limbs present better patency and lower reintervention rates than other stent-graft types.