Rezolute Presents Positive CGM Sub-Analysis from Failed Phase 3 sunRIZE Study
Read source articleWhat happened
Rezolute announced at the Pediatric Endocrine Society meeting additional continuous glucose monitoring (CGM)-based outcomes from the Phase 3 sunRIZE study of ersodetug in congenital hyperinsulinism. Despite failing the primary endpoint of weekly hypoglycemia events and key secondary endpoint in December 2025, these new analyses show statistically significant and consistent improvements in glycemic control for ersodetug versus placebo across multiple pre-specified and post-hoc CGM endpoints. The company is using these data to support a 'totality of evidence' argument in ongoing FDA discussions for a potential salvage path in congenital HI. However, investors should remain skeptical: post-hoc analyses from a failed trial are rarely sufficient to change regulatory decisions, and the stock remains deeply discounted due to the primary failure and litigation overhang. The positive CGM signals may modestly increase the probability of a limited congenital salvage, but the binary outcome still hinges on upcoming FDA feedback and the tumor HI upLIFT trial.
Implication
The CGM sub-analysis offers a marginal improvement in the congenital salvage thesis, but the stock still trades as a binary option on FDA feedback and upLIFT results. Without clear FDA endorsement of a salvage path, the bear case remains dominant. Entry at $2.00 or below is still preferred, and the WAIT rating is appropriate until more clarity emerges.
Thesis delta
The positive CGM sub-analysis from sunRIZE suggests that ersodetug has some glycemic benefit, increasing the chance that FDA may accept a 'totality of evidence' approach for a limited congenital HI indication from maybe 10% to 20%. However, the primary endpoint failure remains the dominant fact, and the thesis remains centered on the upcoming FDA meeting and upLIFT readout. The probability of the bear scenario ($1.25) decreases slightly from 40% to 35%, but the base case ($4.00) remains the most likely.
Confidence
Low