A Phase 3 stroke trial testing a two-peptide cocktail missed the target it was built to hit. The interesting part is where it did not.
The COURAGE-2 trial, published July 16 in the Journal of Clinical Neuroscience ↗, gave 188 people with an acute ischemic stroke either standard care or standard care plus an intravenous drug pair: GHRP-6 ↗, a small growth-hormone-releasing peptide, combined with epidermal growth factor (EGF). An ischemic stroke is when a clot cuts off blood to part of the brain, and the damage spreads outward from the dead core in the hours that follow. The idea here was not to break the clot but to protect the brain tissue still at risk around it.
Patients got the treatment within 12 hours of symptoms, twice a day for a week. Ninety-five received the peptide pair, 93 got standard care alone. The main question was disability at three and six months, scored on the modified Rankin Scale, a zero-to-six ladder where zero is no symptoms and six is death.
The primary endpoint missed
Across everyone in the trial, the peptide pair did nothing measurable. No difference in disability, no difference in daily function on a second scale, no difference in survival. By the standard the trial set for itself, COURAGE-2 is a negative study.
Then the authors looked at the sickest patients. Among the 27 people who came in with the most severe strokes, those given EGF and GHRP-6 ended up less disabled at six months (a Rankin score of 2.6 versus 4.7 in the control group) and far less likely to die (hazard ratio 0.18, which works out to roughly an 82 percent lower risk of death). In the group whose stroke sat in the middle cerebral artery, the largest brain-supply vessel, the treated patients also showed more shrinkage of the dead tissue at 30 days.
Why the subgroup is a lead, not a verdict
A drug that appears to help only the sickest quarter of a trial is one of two things. It could be a real effect that the full, mixed population was too diluted to reveal. Or it could be noise that a small subgroup and a p-value squeaking under 0.05 dressed up as a finding. COURAGE-2 cannot tell you which. The severe-stroke group held 27 people, the survival benefit's confidence interval ran from 0.03 all the way to 0.96, and the analysis was carved out after the main result came back flat. The authors are careful about it, concluding only that the trial "suggested potential benefit in moderate-to-severe stroke" and warrants a study built around that subpopulation.
On safety the pair looked clean. Serious adverse events landed in 48 of 188 patients across both arms, and none were judged treatment-related.
The peptide angle
GHRP-6 is better known as a growth-hormone secretagogue ↗, a research peptide that spikes growth hormone and appetite by acting on the ghrelin receptor. Its use here is a repurposing bet: pair it with EGF and aim the combination at the surviving neurons after a stroke rather than at metabolism. That places it among the neuroprotective ↗ peptide candidates, a corner of the field that has produced far more failed trials than approvals. COURAGE-2 does not break that pattern. What it does is point at a narrower question, the severe-stroke patient, that an adequately powered trial could actually answer.