PSMA-PET/MRI-Ultrasound Multimodal Fusion Navigation for Da Vinci Robot-Assisted Radical Prostatectomy: A Randomized Controlled Trial

Overview

Radical prostatectomy faces the core dilemma of balancing functional preservation with tumor eradication. While nerve-sparing techniques improve urinary control, intraoperative tumor localization remains imprecise, resulting in positive surgical margin (PSM) rates of 11%-38% and elevated recurrence risk. Traditional preoperative 2D imaging fails to dynamically guide surgical boundaries. Although multimodal fusion studies (e.g., MRI or PSMA-PET/CT) attempt to address this, they struggle to achieve simultaneous precision in lesion identification and real-time spatial tracking. This study pioneers a PSMA-PET/MRI-ultrasound multimodal fusion navigation system for the Da Vinci surgical robot, leveraging three innovations: PSMA-PET/MRI dual-modality synergy for subclinical lesion detection at millimeter resolution; Non-rigid point-cloud registration algorithms to dynamically compensate for intraoperative prostate deformation, enabling 3D ultrasound-PET/MRI elastic fusion; Utilizing the telipro port of the Da Vinci surgical robot to achieve intraoperative picture-in-picture navigation, real-time localization of the tumor boundary, and precise resection as well as precise protection.This study aims to verify the safety and effectiveness of the world's first PSMA-PET/MRI-ultrasound multimodal fusion navigation system adapted for the Da Vinci surgical robot. This system is expected to reduce the positive margin rate to less than 10%, increase the rate of nerve preservation by 30%, shorten the postoperative urinary control recovery time to within 2 weeks, and establish a standard process for robotic surgery navigation. This will provide a new paradigm for precise surgical treatment of prostate cancer.

Prostate cancer, the second most prevalent malignancy in men globally, has long grappled with a core dilemma in radical surgery: balancing functional preservation against oncological efficacy. Although nerve-sparing techniques significantly improve postoperative urinary control and sexual function (with robotic surgery achieving \>80% continence recovery rates), conventional approaches relying on intraoperative visual tumor boundary assessment result in positive surgical margin (PSM) rates of 11%- 38%, increasing biochemical recurrence risk exceeding 40% \[1,2\]. For locally advanced cases, sacrificing functional structures to ensure oncological radicality leads to postoperative erectile dysfunction rates up to 95% and urinary incontinence exceeding 50% \[3\].The essence of this conflict lies in: Extended resection reduces PSM rates but damages neurovascular bundles (NVBs) governing micturition and erectile function; Limited resection preserves function yet increases PSM risk due to residual microlesions-particularly in anatomically complex zones like the prostatic apex and anterior wall, where visual localization errors typically exceed 3 mm. Preoperative imaging limitations exacerbate this: MRI offers high anatomical resolution (0.5 mm³) but cannot track intraoperative organ deformation; PSMA-PET/CT detects micrometastases with 98% sensitivity, yet spatial registration errors between metabolic/anatomical data exceed 2 mm \[4\]. Current multimodal fusion approaches are inadequate: MRI-based fusion misses early-stage lesions due to limited tumor contrast; PSMA-PET/CT fusion suffers from metabolic-anatomical misalignment. Thus, a navigation system enabling simultaneous subclinical lesion detection and dynamic deformation compensation is imperative to resolve the function-versus-curability dilemma. We have adopted the following approaches to complete the construction of the intraoperative navigation system: (1) On the PET/MRI before the operation, the prostate and the lesion were delineated: at least two nuclear medicine physicians independently reviewed the images and then provided a unified report; the external contour of the prostate and the three-dimensional lesion schematic diagram of the lesion were then delineated by a urologist; (2) On the intraoperative ultrasound, the prostate was delineated: the prostate image was captured in real time by BK ultrasound and then the external contour of the prostate was delineated by a urologist; (3) The multimodal fusion of the three-dimensional lesion delineated by BK ultrasound and PSMAPET/MRI was achieved through the MIM software built into the BK ultrasound; (4) The intraoperative resection was guided by the Da Vinci Tilepro functional module. So far, 6 cases have been successfully completed and compared with 6 T3a patients randomly selected from previous conventional surgeries. Currently, due to the small sample size, although the differences in the surgical margins have not reached a statistically significant difference, a trend of difference has been demonstrated. Due to the short follow-up period, the postoperative PSA and urination conditionshave not been included in the statistical cohort.