Tumor-targeting for biopsy and radiotherapy is increasingly considered in the management of patients with localized prostate cancer. Dose painting of radiotherapy to an identifiable tumor site within the prostate may improve outcome without increasing treatment-related toxicity. (1) We sought to determine the role of MRI-guided biopsy to confirm visible tumors in the treatment planning process of focal radiotherapy boost..

Thirty-two patients with localized prostate cancer and a visible dominant tumor nodule on MRI were prospectively enrolled between 2012 and 2015. Patients underwent a confirmation MRI-guided tumor biopsy using an integrated diagnostic and interventional MRI technique in a 3T scanner (Verio, Siemens) using an endorectal coil system (Sentinelle, InVivo).(2) Axial images were acquired at 3 T (Verio, Siemens) using an endorectal coil and body matrix coil. The protocol included T2w FSE images (TR/TE 115/5410 ms; TF 12, R/L phase encode, FOV 14cm, 100% phase oversampling, 2 mm slices, 256x256x20, 0.5 mm in-plane, 203Hz/pixel readout, GRAPPA 2, 3 nex, 6 min 4 s), DWI (2D TE/TR= 83/6000 ms, 3 b-values 100, 600, 1000 s/mm2, FOV 18cm, 3mm slices, 128x128x17, FatSat, PE A/P, GRAPPA 2, 6 nex, 6 min 20 sec) and DCE (3D FLASH, TE/TR/alpha = 1.71/4.3 ms/20 deg, FOV 18 cm, 3 mm slices, 128 x 128 x 20, PE A/P, GRAPPA 2, tres 4.8s, 3 min 14 s).

Tumors were scored according to PIRADS v2 classification and all PIRADS=3-5 lesions were targeted for biopsy. Biopsies were acquired through a custom transperineal template under online stereotactic navigation (Aegis, Hologic). Axial FSE images were acquired with needles in situ to evaluate tumor-targeting accuracy (2D TE/TR = 11/1250 ms, R/L phase encode, 80% oversampling, TF 8, FOV 16cm, 256 x 256 x 20, 70% phase resolution for 0.6 x 0.9 mm in-plane resolution, 3 mm slice thickness, GRAPPA 2, 30 s).

Median age at biopsy was 67 years (range 51-75), and median PSA was 7 (2-29). Eight patients with Gleason (3+3), 17 with (3+4), 3 with (4+3) and 5 with (4+4) disease on prior TRUS-guided biopsy were enrolled. A total of 52 tumors were targeted for biopsy, mean per patient 1.7 (1 - 4), with 1.9 (1-5) biopsies acquired per target. Intervention time was a mean of 51 minutes (22-112).

Malignancy was confirmed in 13/13 PIRADS=5 tumors, 27/35 PIRADS=4 tumors, and 1/4 PIRADS=3 tumors (Figure 2). The majority of biopsy needles (70%) were deflected at the margin of tumors on confirmatory MRI. (Figure 1) Overall, 68% (66) of biopsies were positive for disease. The mean percentage of positive cores was 62% for central biopsies and 43% for marginal deflections. Few patients (n=3) were upgraded following MRI-guided and tumor-targeted biopsy to high-risk Gleason 8 disease. Four of 35 PIRADS=4 lesions had false negative biopsies confirmed with subsequent sampling, and the remainder were suspected marginal miss. All false negative samples occurred with dense tumors ≤1.2cm whereby needles were all deflected at the margin of the tumor. The negative biopsy rate fell from 40% for targets with single sampling to 13% for those with multiple sampling. Cores visualized at the centre of a tumor yielded a negative rate of 11%, while cores visualized at the deflected margin of the tumor yielded a negative rate of 42%.

Current standard of care prior to definitive local therapy includes 12-core TRUS-guided biopsy to establish the diagnosis of prostate cancer, and characterize the underlying histopathologic profile. MRI in increasing performed prior to local therapy to guide focal dose-painting strategies. We find that the value of supplementing with and MRI-guided and tumor-targeted biopsy is limited when boosting PIRADS 4-5 disease. The probability of a false-positive PIRADS 4 lesion on MRI is lower than the probability of sampling error on directed biopsy due to targeting inaccuracies incurred through needle deflection by tumors. The results of this study suggest biopsy confirmation of PIRADS=5 lesions is not be necessary in the treatment planning process for focal therapy, and of limited impact for PIRADS=4 lesions. Our observation of needle deflection by tumours when performing trans-perineal biopsy highlights errors in limited sampling, and potential challenges if using an MRI-directed approach through TRUS/MRI fusion. Strategies to mitigate this problem in small (<1cm) dense tumors include online MRI guidance with imaging confirmation of successful central tumor targeting, and/or multiple sampling.