68Ga-PSMA HBED-CC PET/MRI in intermediate and high-risk prostate cancer improves the intraprostatic tumor localization compared to multiparametric MR

Objectives Prostate-specific membrane antigen (PSMA)-ligand PET shows promising results in recurrent prostate cancer (PC) and preliminary studies state its usefulness also for primary PC. Multiparametric MR (mpMR) is currently the gold standard for evaluation of the intraprostatic spread of primary PC. Aim of the study was to assess the diagnostic performance of combining the molecular signal from ⁶⁸Ga-PSMA HBED-CC PET with dedicated mpMR by using integrated PET/MRI for localization of primary PC within the prostate and comparing it to mpMR and PET alone.

Methods ⁶⁸Ga-PSMA HBED-CC PET/MRI was performed in 66 men with biopsy-proven PC with 53 (n=25 intermediate, n=28 high-risk patients) included in the final analysis. The prostate was divided into sextants for histopathology and co-registered with imaging. Three independent readers blinded to histopathology assessed mpMR using PIRADS v1, PET and combined ⁶⁸Ga-PSMA HBED-CC PET/MRI using a 5-point Lickert scale. SUVmax and SUVmean was measured in each sextant. On patient base sensitivity of ⁶⁸Ga-PSMA HBED-CC PET/MRI, PET-imaging and mpMR was calculated after dichotomization of data. On sextant base diagnostic performance was calculated based on receiver-operating-characteristics (ROC) analysis for each modality for all patients and separated for intermediate and high-risk patients. Histopathology was performed on sextant base with determination of the largest tumor and the highest Gleason Score per sextant.

Results 202 of 318 (63.5%) sextants contained cancer at pathologic examination. On patient base sensitivities of mpMR, PET-imaging and ⁶⁸Ga-PSMA HBED-CC PET/MRI on a patient basis were 66.0%, 92.5% and 98.1%, respectively. PET-imaging and PET/MRI statistically significantly outperformed mpMR (p<0.001, respectively). On sextant base ⁶⁸Ga-PSMA HBED-CC PET/MRI statistically outperformed mpMR (AUC 0.88 vs. 0.73, p<0.001) and PET-imaging (AUC 0.88 vs. 0.83, p=0.002) for intraprostatic tumor localization of PC. Sensitivities and specificities at maximum Youden-selected treshold were 75.7%, 64.4%, 57.9% and 97.4%, 94.0%, 81.9% for ⁶⁸Ga-PSMA HBED-CC PET/MRI, PET-imaging alone and mpMR, respectively. Subanalysis for intermediate and high risk patient also showed superiority of PET/MR vs. mpMR (AUC 0.84 vs 0.62, p5mm tumors (odds ratios 4.47 and 4.50, p<0.001, respectively). PET/MR detected significantly more high (蠅7) vs. low (&#8804;6) Gleason Score tumors (odds ratio 1.86, p=0.041). PET showed high uptake ratio between malignant vs. non-malignant tissue (5.02 (range 0.89-29.8)), but no significant correlation was observed between quantitative PET-parameters and Gleason-Score or PSA-value.

Conclusions ⁶⁸Ga-PSMA HBED-CC PET/MRI improves diagnostic accuracy for PC localization compared to mpMR. Prospective studies are warranted and promising to evaluate e.g. its potential for biopsy guidance.