Strong Correlation Between SUV_max on PSMA PET/CT and Numeric Drop-In γ-Probe Signal for Intraoperative Identification of Prostate Cancer Lesions.

Abstract

Prostate-specific membrane antigen (PSMA) PET is used to select patients with recurrent prostate cancer for metastasis-directed therapy. A surgical approach can be achieved through radioguided surgery (RGS), using a Drop-In γ-probe that traces lesions that accumulate the radioactive signal. With the aim of guiding patient selection for salvage surgery, we studied the correlation between the SUV_max of lesions on preoperative PSMA PET/CT and their intraoperative counts/s measured using the Drop-In γ-probe. <b>Methods:</b> A secondary analysis based on the prospective, single-arm, and single-center feasibility study was conducted (NCT03857113). Patients (<i>n</i> = 29) with biochemical recurrence after previous curative-intent therapy and a maximum of 3 suggestive lesions within the pelvis on preoperative PSMA PET/CT were included. Patients treated with androgen deprivation therapy within 6 mo before surgery were excluded. All patients received an intravenous injection of ^99mTc-PSMA-I&S 1 d before surgery. Radioguidance was achieved using a Drop-In γ-probe. Correlation was determined using the Spearman rank correlation coefficient (ρs). Subgroup analysis was based on the median SUV_max <b>Results:</b> In total, 33 lesions were visible on the PSMA PET/CT images, with a median overall SUV_max of 6.2 (interquartile range [IQR], 4.2-9.7). RGS facilitated removal of 31 lesions. The median Drop-In counts/s were 134 (IQR, 81-220) in vivo and 109 (IQR, 72-219) ex vivo. The intensity of the values correlated with SUV_max (ρs = 0.728 and 0.763, respectively; <i>P</i> < 0.001). Subgroup analysis based on median SUV_max in the group with an SUV_max of less than 6 showed no statistically significant correlation with the numeric signal in vivo (ρs = 0.382; <i>P</i> = 0.221) or the signal-to-background-ratio (ρs = 0.245; <i>P</i> = 0.442), whereas the group with an SUV_max of 6 or more showed respective statistically significant positive correlations (ρs = 0.774 [<i>P</i> < 0.001] and ρs = 0.647 [<i>P</i> = 0.007]). <b>Conclusion:</b> Our findings indicate that there is a direct relation between SUV_max on PSMA PET/CT and the readout recorded by the surgical Drop-In probe, thereby indicating that SUV_max can be used to select patients for PSMA RGS. For more definitive subgroup definitions for treatment recommendations, further studies are necessary to validate the present findings.