Abstract
1466
Objectives: Coeliac ganglia (CG) has been reported to cause pitfalls on 68Ga-PSMA on Positron Emission Tomography/Computer Tomography (PET/CT) scans. The aim of this study was to investigate the morphologic features and 68Ga-PSMA avidity of coeliac ganglia on hybrid PET/Magnetic Resonance (MR) imaging. METHODS 68Ga-PSMA whole body PET/MR examinations in 104 patients (mean age 64years; mean dose 166MBq), referred for primary staging or follow-up of prostate cancer, were retrospectively reviewed to investigate the radiotracer uptake (maximum standardized uptake value - SUVmax) and morphologic features (size, shape and location on T1-weighted and T2-weighted MR images) of the coeliac sympathetic ganglia. The background 68Ga-PSMA activity was measured in gluteal muscles and descending aorta. Oval, nodular (with 2 or more nodules) and longitudinal shapes of CG were regarded as mistakable with lymph nodes, whereas linear (regular, linear, crescent, irregular or dashed) as non-mistakable. RESULTS On MR scans coeliac ganglia were visible in the case of 101 patients out of 104 on both sides and in the case of one patient only on the left side. In the case of 2 patients reliable identification of both CG, and, in the case of another patient, of the right CG was not possible neither on MR nor on PET scans, probably due to their extreme thinness. Mistakable with lymph nodes shape presented 80% (82/102) of the left CG and 20% (20/101) of the right CG. The left CG were thicker (2-10, mean 4±1.4mm) than the right ones (0.5-7, mean 3±1.3mm). Mean SUVmax was 2.68±1.10 (range 0.02-5.48) in the left CG and 2.38±1.17 (range 0.02-5.91) in the right CG. Increased 68Ga-PSMA uptake, SUVmax above 2, and SUVmax above 2.5, was detected in 77% (79/102) and 55% (56/2012) of the left CG and 61% (62/101) and 44% (44/101) of the right CG, respectively. Uptake below background (in gluteal muscles, in descending aorta or both) was noted only in 20% of the left and 17% of the right CG, and was probably caused either by their thin shape or ‘halo’ artifact around the kidneys. Both - erroneous shape and elevated 68Ga-PSMA uptake - was observed in 65% (66/102) of the left and 16% (16/101) of the right CG. The majority of CG (54% of right and 41% of left) lied between coeliac trunk (CT) and superior mesenteric artery (SMA) not exceeding their levels. Only 19% of the right CG and 16% of the left ones were protruding (markedly or slightly) above the level of CT (reaching SMA with the lower pole or finishing above). Five percent of the right CG and 12% of the left CG were long, extending as well above CT as below SMA (distinctly of slightly). Large amount of CG were protruding below SMA (23% of the right and 31% of the left), sometimes reaching the level of the left renal vein. CONCLUSIONS The oval, nodular or longitudinal shape of coeliac ganglia itself, especially of the thicker left CG, on MR scans may cause mistaking them for metastatic lymph nodes. On whole body PET/MR imaging evident and sometimes high 68Ga-PSMA uptake in coeliac ganglia increases the risk of a misinterpretation of them as prostate cancer metastases. Coeliac ganglia on in-vivo PET/MR imaging are located lower than it has been presented in most anatomical atlases so far.