Abstract / Summary
The metabolic characteristics of untreated testicular germ cell tumours (TGCTs) on ¹⁸F-fluorodeoxyglucose (FDG) PET/CT remain incompletely characterised. This study evaluated differences in PET-derived metabolic parameters between seminoma and non-seminomatous germ cell tumour (NSGCT) and explored their association with histological phenotype. This retrospective single-centre study included 30 treatment-naïve patients with histologically confirmed TGCT who underwent ¹⁸F-FDG PET/CT before antitumour treatment, comprising 12 patients with seminoma and 18 with NSGCT. PET-derived metabolic parameters and serum tumour markers were compared between groups. Receiver operating characteristic analysis, analyses restricted by disease extent, and exploratory multivariable modelling were performed. NSGCT showed higher primary-tumour SUVmax than seminoma (median, 16.66 vs. 8.41; P = 0.007) and higher patient-level highest-lesion SUVmax (21.91 vs. 8.41; P < 0.001). M1 disease was more frequent in NSGCT than in seminoma (55.6% vs. 8.3%). Among individual biomarkers, patient-level highest-lesion SUVmax showed the highest apparent discrimination (AUC, 0.944), followed by AFP (AUC, 0.910) and primary-tumour SUVmax (AUC, 0.796). In M0 disease, primary-tumour SUVmax remained higher in NSGCT (16.96 vs. 8.00; P = 0.001), whereas the apparent AUC difference between patient-level highest-lesion SUVmax and primary-tumour SUVmax decreased from 0.148 to 0.023. In an exploratory log-linear model adjusted for primary-tumour size and M status, the higher primary-tumour SUVmax in NSGCT persisted (adjusted geometric mean ratio [NSGCT/seminoma], 1.67; 95% CI, 1.07–2.60; P = 0.024). NSGCT demonstrated higher primary-tumour ¹⁸F-FDG uptake than seminoma, and this difference remained in exploratory sensitivity analyses accounting for tumour size and disease extent. Although patient-level highest-lesion SUVmax showed greater apparent discrimination at the patient level, its performance was partly influenced by metastatic disease extent. These findings warrant validation in larger independent cohorts.