Abstract
BACKGROUND: Magnetic resonance imaging (MRI) guided, minimally invasive transperineal interventions have been shown to produce positive clinical outcomes, specifically regarding prostate biopsy and cryoablation of prostate cancer. Many vendors, however, do not provide guidance to clinicians as to how their devices, biopsy needles and cryoprobes should be inserted during procedures. As a result, many leading research institutions have developed in-house solutions for needle insertion guidance. Institutions with less research and engineering support may find the development of their own in-house solutions infeasible, leading to a gap in clinical practice. Our purpose was to develop a replicable toolset using minimal hardware and software to facilitate MRI-guided prostate biopsy and cryoablation. The toolset was validated first in a series of phantom trials, then in a clinical cohort of 24 patients from a single institution. METHODS: This study utilized a grid-like rigid trajectory guide coupled with custom software to simulate in phantoms, and then perform on patients, prostate-focused interventional procedures using a closed bore General Electric 450W scanner. Only standard imaging sequences were employed, allowing this work to be generalizable between different scanners with no additional setup. Interventionalists imported the coordinates of the targets in the prostate and the 3 saline-filled fiducial markers on the grid into custom software, which then identified the grid hole and depth of insertion necessary to drive needles to the targeted points. In phantom trials, the targeting error of each insertion was logged. All patients enrolled in the study were men over the age of 50 years with either biopsy-confirmed prostate cancer with magnetic resonance (MR)-visible lesions or prostatic fluid collection. Each patient was reviewed by a multidisciplinary focal prostate therapy team consisting of one interventional radiologist and two urologists. Patients were seen in interventional radiology and urology clinics. During each procedure the number of needle insertions and adjustments was tracked to better evaluate accuracy in clinical practice. RESULTS: The targeting grid and software allowed rapid, repeatable needle insertions with a mean error of 1.05 mm and a standard deviation 0.38 mm. Across 24 prostate-focused procedures (9 biopsies, 14 cryoablations, 1 fluid aspiration), the tools guided interventionalists on their initial insertion of biopsy needles or cryoprobes. In an average procedure, 81.8% of the inserted needles required no adjustment from their first insertion. All procedures achieved technical success during the intervention, satisfying the leading clinician's standards for biopsy sample collection and/or ablation coverage. CONCLUSIONS: Phantom and patient trials found that the proposed tools and techniques enabled clinicians to quickly and accurately place their needles in simulated tissue and clinical procedures. This minimal-investment adaptation of a diagnostic scanner for interventional purposes lowers barriers to entry for image-guided prostate interventions, while the trajectory guide's simplicity requires no specialized on-site engineering support and can be readily fabricated, promoting greater ease-of-use.