High speed automated intraoperative microscopy of the prostate circumference to ensure tumor-free margins in radical prostatectomy

点击翻译按钮获取中文摘要

Project Summary: In prostate cancer (PCa), the presence and amount of residual tumor at the surface of the excised prostate is the only prognostic factor that is affected by surgical technique, vs. other factors, which are fixed and non-modifiable. Yet, positive surgical margins (PSM), defined as the presence of tumor cells at the inked surface of the removed specimen, are common, especially in advanced stage cancers, and are a strong independent risk factor for clinical progression and secondary treatment. In addition, elevated post-operative PSA, which is understood to be related to tumor left behind in the patient, is also associated with high risk of progression and secondary treatment. Thus, complete tumor removal is important to achieve to improve patient outcomes and reduce overtreatment, yet there is a delicate balance between resection radicality and minimizing damage to the neurovascular bundles to preserve post-operative function. The NeuroSAFE trial demonstrated that real-time detection and correction of PSMs by comprehensive histology of the prostate circumference results in improved patient outcomes, yet there are no widely adoptable methods to achieve this. In an effort to address this technical gap, we have developed video-rate structured illumination microscopy (VR-SIM) and demonstrated that it enables accurate diagnosis of PCa in the biopsy setting, and that it can rapidly deliver gigapixel microscopic images of the entire prostate surface for detection of PSMs. VR-SIM delivers the surface area coverage and resolution needed to detect PSMs, yet it is also fast and relatively simple and inexpensive, opening the possibility for widespread adoption. In this project, our interdisciplinary team of engineers and clinicians with significant prostate expertise will further advance this technology towards clinical translation, by completing critical technology development steps and by prospectively validating it in a large patient series. Specifically, we will increase the mechanical automation speed of the device, and will develop a fully automated system for handling of the removed prostate, to enable gigapixel panoramas of the entire prostate circumferential surface to be delivered within 10 minutes of removal and with minimal tissue processing and user intervention. We will then leverage our recently developed dual-color fluorescent stain that replicates standard H&E with high specificity, to develop expert-validated clinical image atlases using biopsies and cadaveric specimens to enhance image interpretation of VR-SIM prostate panoramas. These improvements will be combined to test the system in a prospective 250-patient clinical study, to determine the accuracy of the device for intra-operative detection of PSMs and prediction of post-operative PSA based on measurement of PSM extent. Finally, we will measure and model expert reviewer behavior using a novel web-enabled visual observer tracking method, which could be used in future computer-assisted search algorithms to expedite intra-operative image review. Successful completion will set the stage for multi-center clinical trials to validate the clinical utility of VR-SIM for real-time detection and correction of PSMs, and improvement of patient outcomes.