Selective PET imaging of CXCR4 using the Al18F-labeled antagonist LY2510924

Abstract

Background: [⁶⁸Ga]PentixaFor detects C-X-C chemokine receptor type 4 (CXCR4) overexpression in various malignancies, such as multiple myeloma and non-Hodgkin lymphomas, as well as in endocrine and inflammatory disorders. This study aimed to develop an Al¹⁸F-labeled radiotracer derived from LY2510924 for CXCR4-targeted imaging, leveraging the physical and logistical advantages of fluorine-18.

Methods: We designed a CXCR4-specific radioprobe, [¹⁸F]AlF-NOTA-SC, based on LY2510924 by incorporating a triglutamate linker and NOTA chelator to enable Al¹⁸F-labeling. The in vitro CXCR4 affinity was assessed using cell-based binding assays. Subsequently, in vivo pharmacokinetics and tumor uptake of [¹⁸F]AlF-NOTA-SC were assessed in naïve mice and mice with xenografts derived from U87.CD4/U87.CD4.CXCR4 and MM.1 S cells. Finally, biodistribution was determined in a non-human primate using PET-MR.

Results: Compared to Ga-PentixaFor, AlF-NOTA-SC demonstrated similar in vitro affinity for human CXCR4. [¹⁸F]AlF-NOTA-SC was produced with a decay-corrected radiochemical yield of 21.0 ± 7.1% and an apparent molar activity of 16.4 ± 3.6 GBq/µmol. In [¹⁸F]AlF-NOTA-SC binding assays on U87.CD4.CXCR4 cells, the total bound fraction was 7.1 ± 0.5% (58% blocking by AMD3100). In naïve mice, the radiotracer did not accumulate in any organs; however, it showed a significant CXCR4-specific uptake in xenografted tumors (SUVmean_U87.CD4 = 0.04 ± 0.00 (n = 3); SUVmean_{U87.CD4.CXCR4} = 3.04 ± 0.65 (n = 3); SUVmean_{MM.1 S} = 1.95 ± 0.11 (n = 3)). In a non-human primate, [¹⁸F]AlF-NOTA-SC accumulated in CXCR4 expressing organs, such as the spleen and bone marrow.

Conclusion: [¹⁸F]AlF-NOTA-SC exhibited CXCR4-specific uptake in vitro and in vivo, with fast and persistent tumor accumulation, making it a strong candidate for clinical translation as an ¹⁸F-alternative to [⁶⁸Ga]PentixaFor.

Keywords: Al18F; Al18F-NOTA-SC; CXCR4; PET.

Fig. 1

Chemical structure of LY2510924 (left) and [¹⁸F]AlF-NOTA-SC (right)

Fig. 2

(A) Radioligand binding assay of [¹⁸F]AlF-NOTA-SC in U87 and MM.1 S cells. The light blue bar represents the internalized fraction and the dark blue the membrane-bound fraction. The values are expressed as mean ± SD (n = 3). (B) In vitro autoradiography of [¹⁸F]AlF-NOTA-SC on U87.CD4.CXCR4 tumor slices and (C) co-incubated with AMD3100 (75 µM) ***p ≤ 0.001

Fig. 3

Ex vivo biodistribution profile of [¹⁸F]AlF-NOTA-SC, with dark blue bars representing 10 min p.i. (n = 4) and light blue bars representing 60 min p.i. (n = 3) (A), and whole body maximum intensity projection (MIP) image 50–60 min p.i. in naïve mice (B)

Fig. 4

Ex vivo biodistribution profile of [¹⁸F]AlF-NOTA-SC (n = 3) 75 min p.i., (A) and MIP image (50–60 min p.i.) in mice xenografted with U87.CD4 and U87.CD4.CXCR4 cells (B), with the corresponding TACs (E, F). Ex vivo autoradiography was performed on U87.CD4 and U87.CD4.CXCR4 tumor slices 75 min p.i., using scaling appropriate for U87.CD4 in (C) and for U87.CD4.CXCR4 in (D)

Fig. 5

Ex vivo biodistribution profile of [¹⁸F]AlF-NOTA-SC (n = 4) in mice subcutaneously xenografted with MM.1 S cells, with dark blue bars representing 75 min p.i. (n = 3) and light blue bars representing AMD3100 blocking (n = 1) (A). MIP image 50–60 min p.i. in mice subcutaneously xenografted with MM.1 S cells (B), with AMD3100 blocking (D) and the corresponding TACs in (F). Comparison with [¹⁸F]FDG scan (C) and MIP image of the systemic model (E)

Fig. 6

Dynamic whole-body PET scan of [¹⁸F]AlF-NOTA-SC in a NHP (n = 1). The averaged maximum intensity projection (MIP) image (60 min p.i.), expressed as SUV is shown in (A). The TACs of the kidneys, spleen, liver, lung, heart, and spine, expressed as SUV or %ID, are shown in (B) and (C). The term spine has been defined as the complex structure of vertebrae (including bone marrow), intervertebral discs and spinal cord