MBX Biosciences released initial Phase 1 data for MBX 4291 on May 11, 2026 — a once-monthly GLP-1/GIP dual receptor agonist prodrug that produced mean weight loss of 7 percent in 8 weeks with no reported nausea or vomiting in the first multiple-ascending-dose cohort. The pharmacokinetic profile — a delayed time-to-peak around 13-14 days and an effective half-life around 26 days — supports the company's once-monthly dosing pitch. With Amgen's MariTide and Pfizer's PF-3944 already in or heading toward Phase 3, the monthly obesity category is filling out faster than any other GLP-1 segment.
Research-context information only. Peptides and investigational drugs discussed below are research compounds. Doses and outcomes reported come from published clinical trial data. This article reports what has been documented, not what should be done. Consult a licensed physician for personal medical decisions.
What MBX 4291 Actually Is
MBX 4291 is a GLP-1/GIP dual receptor agonist peptide prodrug engineered on MBX Biosciences' PEP (Precision Endocrine Peptide) platform. Two structural details define the molecule.
Receptor profile matches tirzepatide. MBX 4291 activates the same two receptors — GLP-1 and GIP — as Lilly's weekly tirzepatide. The mechanistic case for dual GLP-1/GIP agonism is established: tirzepatide produced ~21 percent placebo-adjusted weight loss at 72 weeks in SURMOUNT-1, meaningfully better than weekly semaglutide's ~15 percent at 68 weeks in STEP 1. The MBX 4291 differentiation pitch is not a new mechanism — it is the same mechanism at one-quarter the injection frequency.
The prodrug chemistry delivers slow release. PEP-platform peptides are engineered to maintain "consistent drug concentrations with low peak-to-trough ratios" — the company's framing for how it solves the half-life problem. Standard weekly GLP-1s like semaglutide and tirzepatide carry half-lives of 5-7 days, requiring weekly dosing to maintain therapeutic concentrations. MBX 4291's prodrug structure releases active peptide gradually, producing a time-to-peak around 13-14 days and a half-life at maximum concentration of approximately 26 days. The mathematical implication is that one subcutaneous dose maintains active drug in circulation across a full 28-day cycle.
The result is a once-monthly investigational injection with a familiar receptor profile but unfamiliar PK engineering.
The Phase 1 Numbers
MBX Biosciences disclosed data from two ongoing Phase 1 cohorts in the May 11 portfolio update.
| Cohort | Design | Doses | Outcome |
|---|---|---|---|
| SAD Part A | Single ascending dose | 15, 60, 90, 180 mg (4 of 5 complete) | PK supports monthly dosing, no nausea/vomiting |
| MAD Part B (Cohort 1) | 4 weekly + 1 monthly | 30 mg weekly × 4 + 120 mg | 7% mean weight loss at 8 weeks (range 0-16%) |
| MAD Part C | 12-week dosing | 30 active + 10 placebo per cohort, 2 cohorts | Data Q4 2026 |
Three structural details matter beyond the headline weight-loss number.
Tolerability was unusually clean for a GLP-1/GIP. "Only one subject experienced an event of diarrhea, nausea or vomiting through eight weeks" across the MAD Part B Cohort 1, per MBX. By comparison, tirzepatide's Phase 1 multiple-dose studies and Phase 2 obesity trials both reported nausea rates in the 30-50 percent range across active doses. The prodrug slow-release profile may flatten peak plasma concentrations enough to reduce GI receptor activation spikes — the same biological mechanism MariTide and PF-3944 are betting on with different chemistry.
The 7 percent figure is blinded preliminary data. MBX did not break out placebo-arm weight loss separately. With 6 active and 2 placebo subjects in Cohort 1, the placebo-adjusted number cannot be calculated from this readout. Phase 1 weight-loss data from any single small cohort should be read as a tolerability and PK signal rather than a definitive efficacy estimate.
Range matters. Individual responses ran from 0 percent to 16 percent at 8 weeks in Cohort 1. The high-responder ceiling of 16 percent is the data point that suggests further dose escalation could push mean weight loss higher in subsequent cohorts. Phase 1 MAD Part C uses 30 subjects per cohort across two cohorts for 12 weeks — large enough to characterize both the mean response and the responder distribution at higher doses.
For context: weekly semaglutide produced 6.0 percent weight loss at 12 weeks in STEP 1, weekly tirzepatide produced approximately 8.5 percent at 12 weeks in SURMOUNT-1, and weekly retatrutide hit roughly 9 percent at 12 weeks in the Phase 2 dose-ranging study. MBX 4291's 7 percent at 8 weeks tracks the lower half of that range — at one-quarter the dosing frequency.
The Wider MBX Pipeline
The May 11 update was an obesity portfolio update — not a single-asset release. Two additional MBX programs warrant attention.
MBX 5765 is a four-receptor amycretin. MBX 5765 was nominated as the company's lead amycretin candidate alongside the MBX 4291 readout. The molecule combines GLP-1, GIP, glucagon (GCG), and dual amylin and calcitonin receptor agonist (DACRA) activity in a single peptide, also engineered for once-monthly dosing. IND-enabling studies are scheduled to begin in Q2 2026, putting first-in-human dosing roughly 12-18 months out. Amycretin chemistry (named for combining amylin agonism with GLP-1 / GIP / glucagon agonism) is the same biological logic Novo Nordisk's amycretin and Lilly's eloralintide are pursuing — adding amylin and calcitonin signaling on top of incretin receptor activation to recruit additional weight-loss pathways.
A separate GLP-1/GIP/GCG triple agonist nomination is queued for Q3 2026. MBX is also developing a non-amycretin triple agonist molecule competing more directly with retatrutide's receptor profile. No clinical data are available — the molecule has not yet been nominated as a development candidate.
The cumulative effect: by Q4 2026, MBX has the potential to have three obesity assets in or approaching clinical development (MBX 4291 in Phase 1 MAD, MBX 5765 in IND-enabling, plus the triple agonist nomination), all engineered for monthly dosing.
How MBX 4291 Reshapes the Monthly GLP-1 Map
The monthly GLP-1 category now has at least four serious investigational programs.
| Candidate | Sponsor | Mechanism | Stage | Best Phase Data | Timeline |
|---|---|---|---|---|---|
| MariTide | Amgen | GLP-1 agonist + GIP antagonist (peptide-antibody conjugate) | Phase 3 | Up to 20% at 52w (Ph2) | 2027 readout, 2028+ launch |
| PF-3944 | Pfizer | Biased GLP-1 monthly | Phase 3 (starting 2026) | ~13% at 26w (Ph2b) | 2027-2028 readout |
| MBX 4291 | MBX Biosciences | GLP-1/GIP prodrug | Phase 1 MAD | 7% at 8w (Cohort 1) | Q4 2026 next readout |
| MBX 5765 | MBX Biosciences | GLP-1/GIP/glucagon + amylin/calcitonin amycretin | IND-enabling | None yet | FIH likely 2027 |
Three implications follow from a field this crowded.
Pricing pressure builds even before launch. Four sponsors all racing toward the same once-monthly category means at least two of them have to price aggressively to win share at launch. The convenience premium that monthly dosing was supposed to command erodes when convenience becomes a category feature rather than a single-drug differentiator.
MariTide's first-mover position is no longer monopolistic. Amgen's projected 2028 launch will face PF-3944 within roughly 6-12 months, and MBX 4291 within roughly 24 months. The monthly category will arrive as a competitive market, not a monopoly.
Compounded weekly retatrutide retains the efficacy lead through at least 2030. Retatrutide's 28.7 percent ceiling at 68 weeks dwarfs any monthly Phase 2 number reported so far. Even if MBX 4291 and MBX 5765 hit their best-case Phase 2 numbers, neither is likely to close that gap. Patients who prioritize raw kilograms of weight loss per dollar will continue choosing weekly compounded retatrutide regardless of which monthly drug clears the FDA first.
