How I Do It - MRI-ultrasound fusion prostate biopsy using the Fusion MR and Fusion Bx systems

There is increasing evidence to support the use of multiparametric magnetic resonance imaging (MRI) in men at risk for clinically significant prostate cancer to help identify lesions and inform biopsy. Randomized, level 1 evidence demonstrates that men who are managed with MRI and MRI-ultrasound fusion targeted biopsy (MRF-TB) have more clinically significant prostate cancer and less clinically insignificant prostate cancer detected and avoid biopsy altogether more often than men who undergo systematic, whole-gland prostate biopsy (SPB). Furthermore, strategies that incorporate MRF-TB have lower rates of upgrading on radical prostatectomy compared to SPB. However, generalizing this data to wider practice is challenging because there is a learning curve for interpreting MRI and performing MRF-TB, and some of the fusion technologies are better than others. We describe our group's early experience with the Fusion MR and Fusion Bx systems (Focal Healthcare, Toronto, ON, Canada). These products are designed with elastic fusion technology that is user-friendly, intuitive and accurate. The Fusion MR contouring system is straightforward and allows for contouring with several MRI sequences simultaneously. The Fusion Bx biopsy system has a semi-robotic arm that accounts for prostate deformation and patient movement and allows for freehand-like access, which is a seamless transition from SPB for clinicians. There were 68 lesions targeted in the first 51 patients. The overall cancer detection rate was 22%/61%/83% for PI-RADS 3/4/5, respectively. The Gleason grade group 2 prostate cancer or higher rate was 6%/47%/75% for PI-RADS 3/4/5, respectively. There were no major complications in this cohort of patients. Limitations of this study include small number of patients and lack of formal follow up to rule out sepsis. Overall, the Fusion MR and Fusion Bx systems are accurate, straightforward and safe to use for MRF-TB. Early experience does not show any significant learning curve.