Background: Androgen receptor (AR) signaling is a key driver of prostate cancer and inhibiting it remains as the main therapeutic approach for treating the disease.^1,2 The presence of AR and its different isoforms evolves as a response to treatments and as such exhibits significant inter- and intratumoral heterogeneity.  Therefore, knowledge of AR expression profile and its occupancy in tumors can play vital role in understanding response and resistance to treatments. The only AR tracer, [¹⁸F]-FDHT, investigated in humans has its challenges with poor in vivo stability, lower affinity to AR than its endogenous AR ligand, dihydrotestosterone (DHT), and its challenging radiosynthesis.^3,4 Additionally for efficient tumor uptake, [¹⁸F]-FDHT needs to be bound to sex hormone blinding globulin (SHBG) which is not expressed in rodents, making it unsuitable for preclinical studies.⁵ To address these challenges, our laboratory has explored selective androgen receptor modulator (SARM) backbone⁶ for developing novel non-steroidal AR tracer for PET imaging. Here we describe the synthesis and evaluation of [¹⁸F]-FSARM3, a fluorine-18 radiolabeled agent for imaging AR. 

Methods: The [¹⁸F]-FSARM3 was synthesized with a copper catalyzed radiolabeling of a boronic acid⁷SARM3 derivative (B(OH)₂-SARM3). The tracer’s specificity and selectivity to AR was investigated in vitro by monitoring its relative binding to prostate cancer cell lines with varying levels of AR expression with and without non-radioactive FSARM3 and DHT as blocking agents. The affinity of FSARM3 to AR was quantified with a cell competitive binding assay with AR-positive LNCaP prostate cancer cell lines using [¹⁸F]-FSARM3 and [¹⁸F]-FDHT. The tracer’s in vitro stability and its binding to human serum protein was studied over a 120 min period. The in vivo profile of [¹⁸F]-FSARM3 was investigated in castrated and non-castrated subcutaneous LNCaP tumor bearing male nude mice with PET/CT imaging and biodistribution studies.  

Results: We have developed a 60-minute one-step radiosynthesis for [¹⁸F]-FSARM3 tracer resulting in high radiochemical purity of 97.9±2.3 % and radiochemical yield of 2.7±1.4 %. Using multiple AR positive prostate cancer cell lines we demonstrated that the tracer uptake is specific, that is blockable (LNCaP, MYC-CaP & 22Rv1 p < 0.002). Selectivity was established using AR negative PC3 prostate cancer cell line that did not have any specific uptake (PC3: p=0.915). Additionally, FSARM3 has extremely high affinity to AR based on our competitive binding assays in LNCaP cells with [¹⁸F]-FSARM3 and [¹⁸F]-FDHT along with non-radioactive FSARM3 (IC₅₀=4.7 nM and IC₅₀=2.1 nM respectively) or DHT (IC₅₀=4.7 nM and IC₅₀=2.4 nM) as the competitors (Figure 1A). Based on the in vitro stability assay, the tracer remained intact in human serum up to 120 min (99.7 %) and in practice no tracer was bound to human serum proteins (protein bound fraction: 0.5±1.7 %). The in vivo studies in castrated LNCaP tumor bearing mice revealed higher tumor uptake of [¹⁸F]-FSARM3 compared to blocking and experimental non-castrated mice cohorts (1.0±0.7 vs. 0.2±0.1 vs. 0.1±0.0 %ID/g) (Figure 1B). The PET imaging revealed great tumor delineation of the castrated experimental cohort tumors. (Figure 1C) The tracer possessed great pharmacokinetics with minimal uptake in the blood pool (0.2±0.1 %ID/g) and muscle (0.2±0.1 %ID/g) and fast clearance through the gastrointestinal system (liver: 0.8±0.6 %ID/g; l. intestine: 2.7±2.2 %ID/g) by 2 h post injection. 

Conclusions: We have developed a fluorine-18 labeled PET probe [¹⁸F]-FSARM3 for imaging AR based on non-steroidal backbone and demonstrated its high affinity to AR using in vitro and in vivo models.  To our knowledge, this is the first PET tracer that can clearly image AR in castrated mice models without supplementing with SHBG.  Based on superior pharmacokinetic and stability profile, we believe [¹⁸F]-FSARM3 will outperform currently clinically used PET tracer, [¹⁸F]-FDHT.

Image/Figure:

Click to view full size

Image/Figure Caption:

Figure 1. Competitive androgen receptor binding assay of [¹⁸F]-FSARM3 and [¹⁸F]-FDHT in LNCaP prostate cancer cells using non-radioactive FSARM3 or dihydrotestosterone (DHT) as a competitor. (A) In vivo biodistribution of [¹⁸F]-FSARM3 in LNCaP tumor bearing nude male mice 2 h post injection. (B) Axial PET/CT images of castrated LNCaP tumor bearing mice 60 and 120 min post injection red arrow indicating the location of the subcutaneous tumor. (C)

Author