Validation and further development of a spectral CT-based imaging biomarker for sarcopenia diagnostics

• Applicants: Hans Meine, Ph.D.; Dr. Isabel Molwitz; Professor Dr. René Werner
• Subject Area: Nuclear Medicine, Radiotherapy, Radiobiology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 536311164

Background: Sarcopenia is a skeletal muscle disorder that increases adverse events like falls, disability, and mortality. It is of high prevalence in oncologic patients and associated with complications under chemotherapy and shorter survival. Muscle mass or muscle quality need to be determined to confirm the diagnosis of sarcopenia and allow early interventions with nutritional counseling and physical therapy. Existing radiological or clinical techniques are biased, e.g., by contrast agent, water retention, or the need for prospective sequence selection.Muscle quality depends on the muscle fat content, its composition, and architecture. A new approach that is less dependent on the above-mentioned limitations is to determine muscle quality by quantifying the muscle fat fraction using spectral CT material decomposition. However, fat distribution varies within a muscle, which requires 3D-based analyses. Further, for objective assessment of the muscle composition and architecture, radiomics analyses appear suited but have not yet been employed for spectral CT data. The suitability of the spectral CT fat fraction, distribution, and muscle composition to serve as an imaging biomarker for sarcopenia remains to be determined.Aim: The aim of this project is to further develop and validate a novel spectral CT-based biomarker for sarcopenia diagnostics, including automated 3D segmentation and radiomics-based pattern analysis.Methods: For this project, patients with gastrointestinal carcinoma under palliative chemotherapy receiving dual-layer detector spectral CT (dlCT) staging examinations, will be included. The work program consists of 5 work packages (WPs). WP 1 and 3 are about patient recruitment, development and calibration of the material decomposition software, and the confirmatory proof of the accuracy of dlCT muscle fat quantification based on a sample size calculation of n=127 patients. MRI chemical-shift relaxometry and mDIXONquant will serve as reference techniques. In WP 2, automated 3D segmentation algorithms for the abdominal muscles will be developed, dlCT and MRI co-registered, and improvement of dlCT fat quantification robustness by 3D volume- versus 2D slice-based analyses will be evaluated. In WP 4 we will assess the value of the dlCT fat fraction and other dlCT parameters to predict clinical muscle strength and muscle function compared to standard CT muscle mass and density. Moreover, the suitability of dlCT values to identify patients who suffer from sarcopenia, i.a. according to bioelectrical impedance analysis, will be evaluated. In WP 5, radiomic signatures for spectral data of the skeletal muscle will be developed and their predictive value for clinical muscle function and sarcopenia detection assessed.In conclusion, this work program will serve to validate and further develop spectral CT muscle quality assessment derived from regular staging scans and its suitability to serve as an imaging biomarker for sarcopenia diagnostics.

• DFG Programme: Research Grants