PSMA-mediated fluorescence-guided robot-assisted prostatectomy at tracer dose in a procine model

Introduction: Fluorescence-guidance is an upcoming methodology for improvement of surgical accuracy. However, the ability to provide intraoperative fluorescence-guidance is both dependent on the amount/concentration of the fluorescent dye accumulated in the lesion(s) of interest and hampered by tissue-dependent signal attenuation. This generally results in application of relatively high doses of a fluorescent tracer, which can have negative impact on toxicity as well as costs. While tracer dose application (100 ug) is the mainstay in nuclear medicine, to date there is still controversy about the feasibility of receptor-mediated fluorescence imaging in vivo at tracer dose levels. Following validation of the performance of new bimodal tracer designs for PSMA in mouse tumor models we have set-out to validate the performance of fluorescence guidance in a large animal (porcine) model with increased translational character. Reasoning that basal PSMA levels in the prostate should already allow evaluation of the feasibility of PSMA-mediated fluorescence imaging in vivo, we evaluated the ability to identify the prostate in vivo after administration of a tracer dose of a PSMA-targeted tracer. Materials and Methods: Tracer kinetics and in vivo imaging of the prostate were assessed in a porcine model (N=15, 40kg/animal) after intravenous administration of a high affinity Cy5-functionalized PSMA targeting tracer (EuK-(SO₃)Cy5-MAS₃; 100 ug, K_D: 19.2 ± 5.8 nM [1]). Blood and urine samples were obtained over the course of 240 minutes. In vivo imaging and robot assisted prostatectomy were performed at 4-6 hours post injection, in combination with Cy5 imaging using a clinical Cy5-modified laparoscopic system. Following in vivo imaging, tracer uptake in excised tissues (prostate, kidney, ureter, liver, abdominal fat, muscle, splenic tissue and salivary gland) was evaluated ex vivo using fluorescence confocal microscopy. Results: Fluorescence was detected in the blood at early time points after tracer administration (t=5 - 20 minutes) with decreasing intensity over time. Fast urinary clearance was observed, with low levels of fluorescence still present in the urine at the time of imaging and subsequent prostatectomy (t1/2 = 105 min). In vivo fluorescence imaging allowed visualization of the prostate and permitted robot-assisted prostatectomy based on tracer uptake in the prostate. Confocal microscopy of excised tissues confirmed tracer uptake in the kidney and prostate and revealed a faint tracer-related fluorescence signal in the ureter. No fluorescence was detected in the liver, abdominal fat, muscle, splenic tissue or the salivary gland. Conclusion: Our findings indicate that intraoperative PSMA-mediated fluorescence imaging using a tracer dose regime is feasible. The applied approach in (non-tumor bearing) pigs widens translation opportunities from molecule to man and decreases the chance of applying non-successful image-guidance approaches in patients.