Diagnostic performance of PSMA PET/CT, multiparametric MRI, and combined imaging for local prostate cancer recurrence after radical prostatectomy.

OBJECTIVES

Approximately 20-50% of patients develop biochemical recurrence (BCR) of prostate cancer within 10 years following radical prostatectomy (RP). The accurate identification of recurrent disease is crucial for guiding salvage treatment decisions. While multiparametric MRI (mpMRI) and prostate-specific membrane antigen positron emission tomography/computed tomography (PSMA PET/CT) are both utilized for detecting local recurrence, their combined diagnostic benefits remain unclear. This study seeks to evaluate the diagnostic performance of both modalities alone and in conjunction for detecting local recurrence following RP in patients with BCR.

METHODS

A retrospective single-institution analysis included 37 post-RP patients with BCR who received mpMRI and PSMA PET/CT. Five board-certified radiologists reviewed images in three phases: mpMRI only, PSMA PET/CT only, and both modalities combined. Multidisciplinary tumor board consensus served as the reference standard. Diagnostic performance, inter-reader agreement, and radiologist confidence with each modality was examined.

RESULTS

MpMRI outperformed PSMA PET/CT, yielding a higher sensitivity (73.0% vs. 65.2%) and specificity (77.1% vs. 75.7%). Interpretation of mpMRI and PSMA PET/CT together achieved the highest diagnostic accuracy (77.8%), representing a statistically-significant increase over PSMA PET/CT (p = 0.026) but a non-statistically-significant increase over mpMRI (p = 0.441). Combined imaging also resulted in greater specificity (90.0%) and inter-rater reliability (κ = 0.622). However, in some cases performance decreased with both modalities due to interpretive pitfalls.

CONCLUSION

While mpMRI remains the preferred imaging modality for post-RP local recurrence surveillance, the integration of PSMA PET/CT may lead to improved specificity and inter-rater reliability. However, radiologists must understand each modality's limitations to avoid interpretive pitfalls.