Characterization of the sigma-1 receptors status with (S)-(-)-[18F]fluspidine PET of an orthotopic mouse model of glioblastoma to assess its potential in glioblastoma management

Introduction
The sigma-1 receptor (S1R) is a chaperone protein of the mitochondrion-associated endoplasmic reticulum membrane and regarded as potential therapeutic target for a variety of malignant tumors including glioblastoma. Noninvasive assessment of changes in the availability of S1R could help to better understand the pathophysiology of glioblastoma and to improve diagnosis or treatment follow-up. We aim to evaluate the potential of (S)-(−)-[18F]fluspidine, a highly specific S1R radioligand, to characterize the expression of S1R in an orthotopic glioblastoma model in mouse with small-animal PET/MRI.

Methods
U87 human glioblastoma cells were stereotactically implanted into the striatum of three female nude mice (8 weeks old). At a median tumor size of 27 mm3 (determined with MR) 60 min dynamic PET scans were performed after i.v. injection of (S)-(-)-[18F]fluspidine (9.1 ± 2.1 MBq; Am: 140 ± 50 GBq/µmol, EOS). Time-activity curves (TACs) from the tumor and the contralateral regions were analyzed (PMOD v3.9). Peak-to-end ratios (P/E; Peak: SUV mean from 2-9 min, end: SUV mean from 45-60 min) were used to compare regions. Statistics: unpaired two-tailed Student’s T-test (P < 0.05).
To determine the S1R affinity (KD) and density (Bmax) in the tumor and the contralateral region, in vitro autoradiography was performed with (S)-(-)-[18F]fluspidine using cryosections of tumor bearing mice.

Results & discussion
Autoradiography showed an equal affinity of [18F]fluspidine for S1R in the contralateral and tumor region (KD: 17.5 ± 1.3 vs. 18.0 ± 4.9 nM), but a higher Bmax in the tumor (490 ± 43 vs. 1756 ± 40 fmol/mg prot). PET TACs reflected significantly different kinetic profiles in the tumor and the contralateral side (P/E: 3.10 ± 0.48 vs. 1.81 ± 0.16, p < 0.05) (Figure 1). Interestingly, the tumor regions were characterized not only by a lower initial uptake compared to the contralateral side (SUV2-9 min p.i.: 0.89 vs. 1.17, p < 0.05) but also by a slower washout resulting in equivalent SUVs in both regions at 45 min p.i. (SUV45-60 min p.i.: 0.38 vs. 0.34 respectively). This slower radiotracer washout from the tumor compared to the contralateral side is assumed to be caused by the higher S1R site density found in the tumor, along with the specific pathophysiology of the tumor itself (neovascularization, oncotic pressure).

Conclusions
The PET investigation revealed a significant difference in the pharmacokinetics of (S)-(-)-[18F]fluspidine between the brain tumor and the contralateral region, probably related to different S1R availabilities. These results show the suitability of (S)-(-)-[18F]fluspidine for the non-invasive determination of the S1R status in an orthotopic glioblastoma model.