A GIPR antagonist conjugated to GLP-1 analogues promotes weight loss with improved metabolic parameters in preclinical and phase 1 settings

Abstract

Obesity is a major public health crisis. Multi-specific peptides have emerged as promising therapeutic strategies for clinical weight loss. Glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) are endogenous incretins that regulate weight through their receptors (R). AMG 133 (maridebart cafraglutide) is a bispecific molecule engineered by conjugating a fully human monoclonal anti-human GIPR antagonist antibody to two GLP-1 analogue agonist peptides using amino acid linkers. Here, we confirm the GIPR antagonist and GLP-1R agonist activities in cell-based systems and report the ability of AMG 133 to reduce body weight and improve metabolic markers in male obese mice and cynomolgus monkeys. In a phase 1, randomized, double-blind, placebo-controlled clinical study in participants with obesity ( NCT04478708 ), AMG 133 had an acceptable safety and tolerability profile along with pronounced dose-dependent weight loss. In the multiple ascending dose cohorts, weight loss was maintained for up to 150 days after the last dose. These findings support continued clinical evaluation of AMG 133.

Fig. 1. AMG 133 exhibited GIPR antagonist and GLP-1R agonist activities in vitro and extended PK profiles.

a, Structure of AMG 133. The GLP-1 analogue peptide sequence is H[Aib]EGTFTSDYSSYLEEQAAKEFIAWLVKGGG(GGGGS)3 K(BrAc){CONH2}. b, In vitro potency of AMG 133. c–e, PK properties after a single dose of 5 mg kg⁻¹ i.v. administration of AMG 133 to CD-1 mice (c), a single dose of 3 mg kg⁻¹ s.c. administration of AMG 133 to female cynomolgus monkeys (d) and s.c. administration of AMG 133 to obese male cynomolgus monkeys (e). AUC, area under the curve; AUC_inf, AUC zero to infinity; V_ss, volume in steady state; CL, clearance; F, fraction absorbed; EC₅₀, half maximal effective concentration; GIPR, glucose-dependent insulinotropic polypeptide receptor; GLP-1R, glucagon-like peptide-1 receptor; IC₅₀, half maximal inhibitory concentration; IV, intravenous; PK, pharmacokinetic; SC, subcutaneous; t_max, time of maximum observed concentration.

Fig. 2. AMG 133 murine surrogate reduced body weight in mouse models of obesity.

a,b, Changes in BW (a) and blood glucose (b) in male db/db mice following a single i.p. injection of the AMG 133 murine surrogate; n = 8 per group. c–h, Dose-dependent reductions in BW (c) and food intake (d) with the AMG 133 murine surrogate in DIO mice. Blood glucose (e), fasting insulin (f), triglycerides (g) and total cholesterol (h) decreased in DIO mice treated with the AMG 133 murine surrogate. **P < 0.01, ***P < 0.001 and ****P < 0.0001 versus vehicle (a–c) or versus vehicle and between doses (d–h). NS, not significant. ⁺P < 0.05, ⁺⁺P < 0.01 and ⁺⁺⁺P < 0.001 between doses (c). Data represent group mean ± s.e.m. n = 5–8 per group depending on group and dataset. Statistical tests were all adjusted for multiple comparisons and included two-way analysis of variance (ANOVA) (a–c), mixed-effects analysis (d) and ordinary one-way ANOVA (e–h). Source data

Fig. 3. AMG 133 reduced body weight in obese cynomolgus monkeys.

a–e, AMG 133 led to a reduction in BW (a), total energy intake (TEI) (b), triglycerides (c), fasting insulin (d), total cholesterol (e) and LDL-C (f). *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 versus vehicle. ⁺P < 0.05 for AMG 133 0.25 mg kg⁻¹ versus AMG 133 0.75 mg kg⁻¹. Data show group mean ± s.e.m.; vehicle n = 10, AMG 133 0.25 mg kg⁻¹ n = 9 and AMG 133 0.75 mg kg⁻¹ n = 8. Statistical test used was a two-way ANOVA adjusted for multiple comparisons (a,c–f; statistical analysis not performed on TEI). Source data

Fig. 4. Clinical PK and PD profiles of AMG 133 in humans.

a, AMG 133 PK parameters; t_max is presented as the median and t_1/2 is presented as the geometric mean. AUC_0–28 is presented as the geometric mean (CV%) after the last dose of AMG 133 on day 57 for MAD cohorts. b–e, Mean (s.e.m.) percent change from baseline in BW after single doses, n = 6–7 for AMG 133 and n = 12 for placebo at day 1 (b) and multiple doses, n = 6–8 for AMG 133 and n = 6 for placebo at day 1 (c). Mean (s.e.m.) change from baseline in BMI, n = 6–8 for AMG 133 and n = 6 for placebo at day 1 (d) and waist circumference, n = 6–8 for AMG 133 and n = 6 for placebo at day 1 (e) after multiple doses of AMG 133. Arrows indicate when the investigational product was administered: at day 1 in the SAD cohorts (b) and at day 1, 29 and 57 in the MAD cohorts (c–e). Source data

Fig. 5. Changes in metabolic and inflammatory parameters in humans with multiple ascending doses of AMG 133.

a–f, Mean (s.e.m.) change from baseline in fasting glucose, n = 5–8 for AMG 133 and n = 6 for placebo at day 1 (a), fasting insulin, n = 4–8 for AMG 133 and n = 5 for placebo at day 1 (b), fasting C-peptide n = 3–8 for AMG 133 and n = 5 for placebo at day 1 (c), fasting glucagon, n = 3–8 for AMG 133 and n = 5 for placebo at day 1 (d), HbA1c, n = 6–8 for AMG 133 and n = 6 for placebo at day 1 (e) and hs-CRP, n = 6–8 for AMG 133 and n = 6 for placebo at day 1 (f) after MADs of AMG 133. Arrows indicate when the investigational product was administered at day 1, 29 and 57. Source data

Extended Data Fig. 1. Conjugation of GLP-1 peptide analog to GIPR antibody.

AMG 133 chemical structure.

Extended Data Fig. 2. In vitro functional activity of AMG 133 or AMG 133 murine surrogate.

(a–d) Representative dose-response curves of cAMP assays in cells expressing human GIPR (A), human GLP-1R (B), mouse GIPR (C), or mouse GLP-1R (D). 50 pM or 90 pM GIP was used in the antagonist mode (A&C) of the assay. Data are represented as mean ± SEM of n = 2 replicates per treatment. Source data

Extended Data Fig. 3. Phase 1 trial patient disposition.

(a) SAD participant flow chart (CONSORT diagram). (b) MAD participant flow chart (CONSORT diagram).

Extended Data Fig. 4. Changes in free fatty acids and lipid parameters in humans with multiple ascending doses of AMG 133.

(a–d) Mean (SE) change from baseline in fasting free fatty acids, n = 6–8 for AMG 133 and n = 6 for placebo at Day 1 (A), fasting triglycerides, n = 6 for AMG 133 and n = 6 for placebo at Day 1 (B), fasting total cholesterol, n = 6–8 for AMG 133 and n = 6 for placebo at Day 1 (C), fasting LDL, n = 6 for AMG 133 and n = 6 for placebo at Day 1 (D) after multiple ascending doses of AMG 133 or placebo. Arrows indicate when the investigational product was administrated: at day 1, 29, and 57. Source data

Extended Data Fig. 5. In vitro functional activity comparison between AMG 133 and hGIPR-Ab/P1.

(a) Representative dose-response curves of cAMP assays with GIP (agonist mode) and AMG 133 or hGIPR-Ab/P1 + 50 pM GIP (antagonist mode) in cells expressing human GIPR. Data are represented as mean ± SEM of n = 2 replicates per treatment. (b) Representative dose-response curves of cAMP assays with GLP-1, AMG 133 or hGIPR-Ab/P1 in cells expressing human GLP-1R. Data are represented as mean ± SEM of n = 2 replicates per treatment. Source data

Extended Data Fig. 6. The HPLC-TOF-MS spectrum and SEC chromatograph of AMG 133.

(a) HPLC-TOF chromatogram, UV 214 nm. (b)TOF-MS, full spectrum. (C) SEC chromatogram, UV 280 nm.