Prostate tumour visualisation with PET: is image fusion with MRI the answer?

摘要

Prostate-specific membrane antigen positron emission tomography (PSMA PET)/CT plays an important role in prostate cancer due to superior accuracy for staging than conventional imaging (CT, bone scan). Intraprostatic PSMA PET/CT activity has additive benefit for detection of significant tumour foci [1] and prognostic properties [2]. Although combined PSMA PET/CT and MRI improve sensitivity due to enhanced localisation of prostatic lesions, especially in multifocal disease [1], these scans are often acquired and viewed (by urologists) separately. The CT component of PET/CT lacks detailed soft tissue delineation when compared to MRI, and the prostate volume tends to be larger on CT than MRI, which in turn may affect accurate localisation and targeting of prostatic lesions. Incorporation of all imaging with PET/CT and MRI image fusion may enhance prostate target lesion visualisation and improve prostate cancer management through more accurate biopsy and surgical planning. Image fusion can combine PSMA PET/CT and MRI images into a single image ‘stack’ for easier visualisation. While PSMA PET and MRI fusion can also be performed using a hybrid PSMA PET/MRI scanner, limited scanner access and costs mean this form of fusion is uncommon. Image fusion of separate PSMA PET/CT and MRI scans using software with semi-automated methods represents a feasible alternative to hybrid PET/MRI and warrants investigation due to potentially enhanced prostate target lesion visualisation and, thus, improved diagnosis (biopsy) and surgical (robot-assisted radical prostatectomy [RARP]) planning (Fig. 1). In the diagnostic setting, the current pathway of MRI then prostate biopsy has limitations, including MRI-invisible tumours, inter-reader interpretation, access to MRI, and biopsy targeting accuracy. In particular, patients with a negative/equivocal MRI but rising PSA density are recommended to undergo a blind, systematic prostate biopsy with no targeting. Use of PSMA PET prior to initial biopsy is likely to improve sensitivity overall (90% vs 83%) when compared to MRI, and in lesion detection in patients with negative/equivocal MRI (90%), according to the PRIMARY study (Australian New Zealand Clinical trials registry: ANZCTRN 12618001640291) [1]. Conversely, a negative PSMA PET is likely to indicate an absence of significant disease, and image fusion can serve to better determine prostate anatomy when considering possible benign central and transition zone avidity. One can expect that the combined visualisation of PSMA PET and MRI will be useful for accurate diagnosis of clinically significant disease through both higher yield in target biopsy and avoidance of biopsy for negative/equivocal and low-risk patients, as was observed when MRI was first incorporated into biopsy planning. In the surgical planning setting, surgeons are currently limited to separate consideration of imaging (MRI, PSMA PET) and biopsy information. The result is often a wide neurovascular bundle, bladder neck and/or urethral resection where cancer is suspected, potentially affecting functional (continence, potency) recovery, or positive margin if preoperative information did not detect a peripheral tumour. Improved visualisation via image fusion may capitalise on increased sensitivity of multifocal prostatic disease from PSMA PET for more precise prostatectomy and decision on nerve sparing. Enhanced visualisation through three-dimensional (3D) virtual and printed models, based on MRI alone, have shown reduced positive surgical margin rates during RARP [3, 4]. By creating 3D models based on PET and MRI fused images (Fig. 1), dual benefits of 3D visualisation and combined imaging information may assist surgical planning especially in patients with discordant PET and MRI findings. Many methods exist for image fusion. Practically for urologists, semi-automated methods for image fusion that register PET/CT to MRI by overlaying of information are already available on clinical software in