Quantification of 18F-FDG Delivery Rate for Liver Inflammation Using Shortened Dynamic PET Imaging Protocols.

Liver inflammation is a diagnostic hallmark of metabolic dysfunction-associated steatohepatitis (MASH), a severe form of chronic metabolic dysfunction-associated steatotic liver disease. 18F-FDG delivery rate (K 1) in the liver, measured by dynamic PET with tracer kinetic modeling, has the potential to assess liver inflammation and diagnose MASH noninvasively. However, a 1-h scan protocol is typically used to generate the liver [Formula: see text], which is less practical for clinical use. In this study, we aimed to investigate shortened scan protocols for quantification of liver 18F-FDG [Formula: see text] in patients with MASH.

Methods

Eighty-two subjects, including 68 patients with metabolic dysfunction-associated steatotic liver disease and 14 healthy volunteers, were scanned on either a short-axial-field-of-view PET/CT scanner or a total-body PET/CT system following a full 1-h dynamic scan protocol. Two shortened scan protocols were proposed with appropriate tracer kinetic models: a 15-min dynamic scan with a 2-tissue reversible model and a 10-min dynamic scan with a 2-tissue irreversible model. Liver [Formula: see text] values generated from the shortened scan protocols were compared with those generated from the full 1-h scan and used to assess biopsy-determined liver inflammation and MASH using receiver-operating-characteristic analysis.

Results

Liver 18F-FDG [Formula: see text] values generated from the shortened scan protocols were approximately equal to [Formula: see text] values generated from the full 1-h scan. Such [Formula: see text] showed equivalent capability to differentiate between healthy, low-inflammation, and high-inflammation groups (P < 0.0005), similar to the [Formula: see text] obtained from a full 1-h scan. When combined with liver CT that assesses steatosis, the 2 shortened scan protocols achieved an area under the curve of 0.95 for receiver-operating-characteristic analysis when differentiating MASH and non-MASH with both scanners.

Conclusion

It is feasible to develop a liver PET/CT parametric imaging approach using a dynamic scan of 15 min or less with an appropriate tracer kinetic model for evaluating liver inflammation and diagnosing MASH on both total-body and conventional PET scanners. The 2 proposed shortened scan protocols are more practical than the 1-h protocol for measuring liver 18F-FDG [Formula: see text].