Competitive blocking of salivary gland [18F]DCFPyL uptake via localized, retrograde ductal injection of non-radioactive DCFPyL: a preclinical study

Abstract

Background: PSMA-targeted radionuclide therapy (TRT) is a promising treatment for prostate cancer (PCa), but dose-limiting xerostomia can severely limit its clinical adaptation, especially when using alpha-emitting radionuclides. With [¹⁸F]DCFPyL as a surrogate for PSMA-TRT, we report a novel method to selectively reduce salivary gland (SG) uptake of systemically administered [¹⁸F]DCFPyL by immediate prior infusion of non-radioactive standard of [¹⁸F]DCFPyL (DCFPyL) directly into the SG via retrograde cannulation.

Methods: A dose-finding cohort using athymic nude mice demonstrated proof of principle that SG uptake can be selectively blocked by DCFPyL administered either locally via cannulation (CAN group) or systemically (SYS group). The experiments were repeated in a validation cohort of 22RV1 tumor-bearing mice. Submandibular glands (SMG) of CAN mice were locally blocked with either saline or DCFPyL (dose range: 0.01× to 1000× molar equivalent of the radioactive [18F]DCFPyL dose). The radioactive dose of [18F]DCFPyL was administered systemically 10 min later and the mice euthanized after 1 h for biodistribution studies. Toxicity studies were done at up to 1000× dose.

Results: In the dose-finding cohort, the SYS group showed a dose-dependent 12-40% decrease in both the SMG T/B and the kidney (tumor surrogate). Mild blocking was observed at 0.01× , with maximal blocking reached at 1× with no additional blocking up to 1000× . In the CAN group, blocking at the 0.1× and 1× dose levels resulted in a similar 42-53% decrease, but without the corresponding decrease in kidney uptake as seen in the SYS group. Some evidence of "leakage" of DCFPyL from the salivary gland into the systemic circulation was observed. However, experiments in 22RV1 tumor-bearing mice at the 0.1× and 1× dose levels confirm that, at the appropriate blocking dose, SG uptake of [18F]DCFPyL can be selectively reduced without affecting tumor uptake and with no toxicity.

Conclusion: Our results suggest that direct retrograde instillation of DCFPyL into the SG could predictably and selectively decrease salivary uptake of systemically administered [¹⁸F]DCFPyL without altering tumor uptake, if given at the appropriate dose. This novel approach is easily translatable to clinical practice and has the potential to mitigate xerostomia, without compromising the therapeutic efficacy of the PSMA-TRT.

Keywords: Cannulation; Competitive inhibition; PSMA; Prostate cancer; Radionuclide therapy; Salivary glands; Xerostomia.

Fig. 1

a PET/CT scan of a patient administered with [¹⁸F]DCFPyL demonstrating uptake in the bladder, kidney, salivary (parotid, sublingual, submandibular, and minor/seromucous glands), and lacrimal glands. b Immunofluorescence staining of PSMA in human salivary glands (magenta: anti-PSMA staining on the apical lumen of acinar cells; teal: DAPI nuclei staining). c Cannulated submandibular salivary glands (SMG) of mouse and human illustrating similarity in delivery of preventive therapy in our model system. Black arrowhead indicates the inserted cannula. d and e Diagram of the cannulation procedure showing how delivered cold DCFPyL can access and block PSMA sites on the acinar cell surface

Fig. 2

Experimental design for systemic study (SYS group) and intraglandular blocking study (CAN group) for the dose-finding in non-tumor-bearing cohort (a, b). The tissue/blood ratios of [¹⁸F]DCFPyL in submandibular salivary glands (SMG) and kidney at 1 h post-injection are shown for the SYS group (panels c, d) and for the CAN group (panels e, f), respectively. Mice were injected either with 50 µl of saline control or with DCFPyL at 1× to 1000× molar equivalent of the systemically injected [¹⁸F]DCFPyL dose. Each value in the graphs represents mean tissue/blood ratios ± SD, n = 5–6 for each group, with statistically significant results (P < 0.05) indicated by asterisk. The CAN 1 × group shows the expected local blocking at the SMG and lack of blocking at the kidneys (tumor surrogate). However, unexpected blocking of the kidneys is seen in the CAN group at the 10–1000× level, perhaps due to “leakage” of DCFPyL from intraglandular administration into the systemic circulation

Fig. 3

Experimental design for systemic study (SYS group) and intraglandular blocking study (CAN group) in the validation 22RV1 tumor-bearing cohort (a, b). The tissue/blood ratios of [¹⁸F]DCFPyL in submandibular salivary glands (SMG) and the tumor at 1 h post-injection are shown for the SYS group (panels c, d) and for the CAN group (panels e, f), respectively. Mice were injected either with 50 µl of saline control or with “cold” DCFPyL at 0.1 × or 1 × molar equivalent of the systemically injected “hot” [¹⁸F]DCFPyL dose. Each value in the graphs represents mean tissue/blood ratios ± SD, n = 5–7 for each group, with statistically significant results (P < 0.05) indicated by asterisk. Mice in the CAN 0.1 × and 1 × groups demonstrate statistically significant local blocking at the SMG with no reduction in [¹⁸F]DCFPyL uptake at the tumor

Fig. 4

a Experiment design for intraglandular blocking study. b, c Tissue/blood ratios of [¹⁸F]DCFPyL in salivary glands (submandibular: SMG, sublingual: SLG, parotid: PRG, right: R, left: L; b and kidneys (c) at 1 h after injection. SMG glands were cannulated infused either with saline (CAN saline control) or with 25 µl or 50 µl of DCFPyL at 0.01, 0.1, 1 × molar excess of the systemically injected “hot” [¹⁸F]DCFPyL dose. Ten minutes after SMG infusion, mice were injected with [¹⁸F]DCFPyL. Each value in the graph represents mean tissue/blood ratios ± SD, n = 4–6 for each group

Fig. 5

Effect of instillation of 10 nmoles (CAN-10) and 1 nmoles (CAN-1) of DCFPyL in submandibular glands via cannulation at different times (1 or 2 months) on saliva secretion (a, c) and body weight (b, d). a and c Each bar represents mean saliva volume (µL) normalized to body weight (g) ± SD. b and d Each bar represents mean body weight (g) ± SD, n = 4–6 each group. e: There were no microscopic differences (H&E) between the kidney, liver, and submandibular gland of various groups