Synchronous vs independent reading of prostate-specific membrane antigen positron emission tomography (PSMA-PET) and magnetic resonance imaging (MRI) to improve diagnosis of prostate cancer.

OBJECTIVES

To identify whether synchronous reading of multiparametric magnetic resonance imaging (mpMRI) and 68 Ga-PSMA-11 positron emission tomography (PET)/computed tomography (prostate-specific membrane antigen [PSMA-PET]) images can improve diagnostic performance and certainty compared with mpMRI/PSMA-PET reported independently and synthesized, while also assessing concordance between imaging modalities and agreement with histopathology.

METHODS

This was a retrospective analysis of 100 patients randomly selected from the PRIMARY trial, a prospective Phase II multicentre imaging trial. Three dual-trained radiologist/nuclear medicine physicians re-reported the mpMRI and PSMA-PET both independently and synchronously for the same patients in random order, blinded to previous results. Diagnostic performance was assessed for mpMRI/PSMA-PET images read synchronously or independently and then synthesized. Agreement between imaging results and histopathology was examined. 'Concordance' between imaging modalities was defined as overlapping lesions. Reporting certainty was evaluated by the individual reporters for each modality.

RESULTS

International Society of Urological Pathology Grade Group ≥2 cancer was present in 60% of patients on biopsy. Synchronous reading of mpMRI/PSMA-PET increased sensitivity compared to mpMRI or PSMA-PET alone (93% vs 80% vs 88%, respectively), although specificity was not improved (63% vs 58% vs 78%, respectively). No significant difference in diagnostic performance was noted between mpMRI/PSMA-PET read synchronously and mpMRI or PSMA-PET reported independently and then synthesized. Most patients had concordant imaging (60%), while others had discordant lesions only (28%) or a mixture (concordant and discordant lesions; 12%). When mpMRI/PSMA-PET findings were concordant and positive, 95% of patients had clinically significant prostate cancer (csPCa). When PSMA-PET alone was compared to synchronous PSMA-PET/MRI reads, there was an improvement in reader certainty in 20% of scans.

CONCLUSION

Synchronous mpMRI/PSMA-PET reading improves reader certainty and sensitivity for csPCa compared to mpMRI or PSMA-PET alone. However, synthesizing the results of independently read PSMA-PET and mpMRI reports provided similar diagnostic performance to synchronous PSMA-PET/MRI reads. This may provide greater flexibility for urologists in terms of referral patterns, reducing healthcare system costs and improving efficiencies in prostate cancer diagnosis.