Using [F-18]Dcfpyl in Prostate Cancer Patient-Derived Xenograft (lucap PDX) Organoids to Define Parameters Influencing PSMA-targeted PET Imaging

12 Objectives: Prostate Specific Membrane Antigen (PSMA) is a transmembrane protein that is found at low levels on normal prostate cells but is increasingly expressed in prostate cancer (PCa) during disease progression. Hence, development of PSMA-targeted PET imaging agents has proven successful in identifying PCa and monitoring PSMA expression levels during treatment. Standard treatment for PCa patients is androgen deprivation therapy (ADT), which inhibits the ability of Androgen Receptor (AR) to induce PCa growth. However, ADT has shown variable effects in PET imaging studies on PSMA expression in cancerous lesions. Thus, there is need for defining the genomic and phenotypic properties that impact PSMA detection, especially relative to ADT. ADT has been found to increase PSMA in vitro in LNCaP cells, but cell lines do not model the intratumoral heterogeneity associated with PCa in patients. With the development of the LuCaP PDX platform and in vitro organoid culture methods designed to preserve the genomic integrity of the PDX, the role of AR regulation in PSMA expression can be explored as it relates to disease diversity in a variety of patient-derived models. Therefore, LuCaP PDX organoid cultures provide a clinically relevant ADT model in vitro where PSMA expression could be determined with a PSMA-targeted probe. Towards this end, PSMA PET imaging agent [18F]DCFPyL was used initially to characterize PSMA in vitro in LuCaP PDXs and organoid model systems. Following this, an in vitro ADT protocol was established using LuCaP141 organoids (AR+) in which changes in PSMA were studied with [18F]DCFPyL over the treatment time course. Methods: [18F]DCFPyL was synthesized following our recently published method1. Representative LuCaP PDXs were chosen for study based on PSMA gene expression, ADT response, and AR mutational status, using PC3 cells that express PSMA as a positive control. For the organoid models, PDX or PC3 tissues were processed into a single cell suspension, cultured and harvested at 7 days. In vitro binding studies using [18F]DCFPyL were done to assess the binding affinity (Kd) and PSMA protein levels (Bmax) with organoid, LuCAP PDX, and PC3 membrane preparations. In LuCaP141 (ADT-responsive), organoids were compared in androgen-replete and androgen-deprived (no dihydrotestosterone [DHT], plus enzalutamide [MDV3100]) states. Results: Levels of PSMA detected by [18F]DCFPyL in PDX tumor and PDX-derived organoids were remarkably similar, ranging from 0.01-4.66 femtomoles PSMA per microgram of protein. LuCaP77 demonstrated the highest amount in both the PDX and organoid models and LuCaP167 demonstrated the lowest amount detected in the PDX and undetectable PSMA in organoids. For the most part, the PSMA level of each in vivo LuCaP PDX tumor was comparable to its cultured organoid counterpart. PSMA levels were slightly higher in PDX vs. organoids for those models expressing low levels of PSMA, probably due to the contributions of other cell types (tumor stroma) present in the PDX but not the organoids (Figure 1A). PSMA in LuCaP141 organoids increased gradually following ADT, becoming significant at day 7 (Figure 1B). These data demonstrate the fidelity and utility of PDX-derived organoids, which unlike PDX tumors, are facile for biochemical and genetic manipulations. Conclusions: [18F]DCFPyL was used to characterize PSMA in the LuCaP series, showing similarity between in vivo PDX and in vitro organoid models. Furthermore, AR was shown to negatively regulate PSMA expression in LuCaP141 organoids, validating in this model that PSMA changes in treatment response can be exploited in therapeutic monitoring. The use of multiple patient-derived genomically-heterogeneous models coupled with molecular correlates of PSMA response to ADT is anticipated to provide insights into the heterogeneity of PSMA detection in patients. Reference: 1 Basuli, F.et al. J Label Compd Radiopharm 60, 168-175, (2017).