Background: AZD9567 is a novel oral selective glucocorticoid receptor modulator (SGRM) developed to have an improved safety profile versus prednisolone while maintaining efficacy.

Objectives: To show clinical evidence of safety differentiation for AZD9567 compared to prednisolone with respect to glucose homeostasis and to investigate the underlying effects on the glucose metabolism in human cell systems in vitro.

Methods: In a dose escalation study (NCT02760316), healthy volunteers were randomized to a 5-day once-daily treatment with AZD9567 (20, 40, 80, 125 mg) or prednisolone (5, 20, 40 mg). We monitored the anti-inflammatory effect by TNFα release from ex vivo LPS stimulated whole blood and modelled the relationship with pharmacokinetics (PKPD). Plasma glucose was measured (AUC_0–4h) during an oral glucose tolerance test (OGTT) before and after four days of treatment.

We also studied the effects of AZD9567 and prednisolone on mRNA expression of key gluconeogenesis enzymes (tyrosine aminotransferease (TAT), phosphoenolpyruvate carboxykinase (PEPCK) and glucose 6-phosphatase (G6Pase)) in primary human hepatocytes as well as on insulin secretion in human islets microtissues.

Results: The inhibition of TNFα release increased with the plasma concentrations of AZD9567. PKPD modeling of TNFα inhibition data showed that 40 mg AZD9567 results in an anti-inflammatory activity similar to that predicted for 20 mg prednisolone.

A significantly smaller increase of the glucose AUC was observed following treatment with 20 to 125 mg AZD9567 compared to 20 mg prednisolone (p<0.05). Mean (95% CI) changes from baseline for 20, 40, 80 and 125 mg AZD9567 were 16% (6.5–26, n=7), 19% (9.3–30, n=7), 19% (9.8–30, n=7) and 34% (24–46, n=7), respectively. The corresponding values for 5, 20 and 40 mg prednisolone were 17% (10–24, n=13), 53% (43–62, n=13), and 68% (58–79, n=12), respectively.

We found significant induction of TAT, PEPCK, and G6Pase mRNA in hepatocytes after prednisolone treatment. In contrast, AZD9567 did not change the expression of these enzymes and inhibited the effect of prednisolone when co-administered. Furthermore, AZD9567 showed less suppression of insulin secretion in human islets microtissues compared to prednisolone at concentrations with comparable inhibition of TNFα release, resulting in a 12-fold better therapeutic ratio.

Conclusions: In healthy individuals, AZD9567 showed significantly reduced effect on glucose homeostasis versus prednisolone at doses with similar anti-inflammatory activity. Thus, AZD9567 shows potential as an anti-inflammatory treatment with an improved metabolic safety profile compared to prednisolone.

Disclosure of Interest: T. Hegelund Myrbäck Employee of: AstraZeneca, S. Prothon Employee of: AstraZeneca, M. Dearman Employee of: AstraZeneca, G. Edenro Employee of: AstraZeneca (former emplyee), J. Leander Employee of: AstraZeneca, M. Hashemi Employee of: AstraZeneca, G. de Miquel Employee of: AstraZeneca (former emplyee), R. Fuhr Employee of: Parexel International GmbH, T. Körnicke Employee of: Parexel International GmbH, Z. Tahib Employee of: AstraZeneca, E.-M. Andersson Employee of: AstraZeneca, P. Svanberg Employee of: AstraZeneca, K. Edman Employee of: AstraZeneca, R. Hendrickx Employee of: AstraZeneca, C. Keen Employee of: AstraZeneca, M. Kraan Employee of: AstraZeneca (former emplyee), K. Johansson Employee of: AstraZeneca (former emplyee), U. Eriksson Employee of: AstraZeneca, B. Carlsson Employee of: AstraZeneca, O. Vaarala Employee of: AstraZeneca