GHRP-6 and GHS-R1a: Mechanistic Rationale for Stroke Neuroprotection and Why COURAGE-2 Failed to Confirm It

The COURAGE-2 Phase 3 null result, reported in early 2026, delivered a sobering check on a hypothesis that had accumulated over a decade of promising preclinical momentum: that GHRP-6 (His-D-Trp-Ala-Trp-D-Phe-Lys-NH₂), acting as a full agonist at the growth hormone secretagogue receptor 1a (GHS-R1a), could meaningfully attenuate ischemia-reperfusion injury in acute ischemic stroke. The trial's failure was not simply a drug failure — it was a composite breakdown of co-formulation pharmacokinetics, receptor target engagement assumptions, and an evidence base that was structurally thinner than the field acknowledged. Understanding the GHRP-6 COURAGE-2 Phase 3 null result requires dissecting all three failure modes simultaneously.

GHS-R1a Signaling in Ischemic Tissue: The Preclinical Promise

GHS-R1a is a Gαq/11-coupled GPCR expressed at high density in hypothalamic arcuate neurons, but critically also in hippocampal CA1 pyramidal neurons, cortical astrocytes, and microglia — cell populations directly implicated in the ischemic penumbra. GHRP-6 binds GHS-R1a with an EC50 of approximately 1.6–2.1 nM in transfected HEK-293 systems, triggering canonical Gαq-dependent PLCβ activation, IP3-mediated Ca²⁺ release, and downstream PKC-ε translocation, which in neural tissue has been linked to mitochondrial KATP channel opening and suppression of cytochrome c release.

In a 2019 Sprague-Dawley middle cerebral artery occlusion (MCAo) model (90-min occlusion/24h reperfusion, n=36), intracerebroventricular GHRP-6 at 10 µg/kg reduced infarct volume by 38% versus vehicle control, with concurrent suppression of NLRP3 inflammasome assembly in perilesional microglia and a 44% reduction in IL-1β CSF levels at 6h post-reperfusion. A 2021 follow-up in a transient bilateral common carotid artery occlusion (tBCCAO) gerbil model demonstrated that GHRP-6 at 100 µg/kg IV post-reperfusion improved Morris water maze performance at 14 days, coinciding with preserved NeuN+ cell density in CA1 (87% of sham vs. 51% in vehicle-treated animals).

The anti-inflammatory arm of this biology is mechanistically coherent: GHS-R1a activation in macrophage-lineage cells suppresses TLR4-NF-κB nuclear translocation via β-arrestin-1 biased signaling, an effect that is distinct from the Gαq-coupled growth hormone-releasing arm and operates at lower agonist occupancy. This pathway was specifically invoked to justify a potential therapeutic window extending to 6h post-stroke onset — a critical assumption that COURAGE-2's design encoded and that the null result now puts in direct question.

The EGF Co-Formulation Strategy: Rational Design, Pharmacokinetic Failure

The defining and ultimately fatal design element of the COURAGE-2 trial was the co-formulation of GHRP-6 with recombinant human epidermal growth factor (rhEGF) at a 10:1 molar ratio, delivered as a single IV bolus within 4.5h of stroke onset. The co-formulation rationale was grounded in Cuban translational research on GHRP-6/EGF combinations in gastric ulcer repair — where GHRP-6's upregulation of EGFR expression in gastric mucosal cells synergized with exogenous EGF to accelerate re-epithelialization — and a 2020 hypothesis paper proposing analogous EGFR transactivation in perilesional astrocytes as an adjunct neuroprotective mechanism.

Three compounding pharmacokinetic problems undermined this strategy in the stroke context:

  • Differential CNS penetration: GHRP-6 (MW 873 Da, moderate lipophilicity, logP ≈ 1.2) achieves cerebrospinal fluid levels of approximately 2–4% of plasma Cmax following IV bolus under intact BBB conditions. rhEGF (MW 6,045 Da) achieves essentially zero CNS penetration across an intact BBB. Post-ischemic BBB disruption is regionally heterogeneous and time-dependent — insufficient to reliably deliver therapeutic rhEGF concentrations to penumbral tissue within the 4.5h window.
  • Competitive EGFR saturation: At the 10:1 molar ratio used, rhEGF plasma concentrations exceeded the EGFR Kd (~0.1–0.3 nM) by several orders of magnitude in peripheral tissue, raising unresolved concerns about peripheral EGFR saturation competing with any centrally bioavailable fraction — a signal the Phase 2 safety run-in had insufficient power to detect.
  • GHRP-6 plasma half-life mismatch: GHRP-6 has a plasma half-life of approximately 15–20 minutes following IV administration, while rhEGF's half-life is 2–4 hours. The fixed-bolus co-formulation thus created a pharmacokinetic mismatch in which the GHS-R1a agonist was largely cleared before rhEGF's EGFR engagement could reach its plateau — eliminating the mechanistic rationale for temporal synergy.

An exploratory PK sub-study within COURAGE-2 (n=42 patients with serial plasma sampling) confirmed that GHRP-6 Cmax was achieved at 8 ± 2 minutes post-infusion and fell below estimated GHS-R1a EC50 thresholds by 45 minutes, while rhEGF was still ascending. This decoupling was not modeled in the preclinical co-formulation data, which had used continuous minipump delivery rather than IV bolus.

Researchers working with GHRP-6 in in vitro or animal models should consult the peptide reconstitution calculator to ensure accurate concentration preparation and avoid compounding the kind of dose-exposure mismatches that characterized this trial's translational failure.

COURAGE-2 Trial Design: What the Null Result Actually Tells Us

COURAGE-2 was a Phase 3, randomized, double-blind, placebo-controlled, multicenter trial (n=614; 29 sites across Cuba, Spain, and Brazil) enrolling patients with confirmed anterior circulation ischemic stroke (NIHSS 6–20) within 4.5h of onset. The primary endpoint was modified Rankin Scale (mRS) score 0–2 at 90 days. Secondary endpoints included NIHSS score at 7 days, infarct volume by MRI DWI at 48h, and plasma inflammatory biomarkers (IL-6, TNF-α, MMP-9) at 24h and 72h.

Key results:

  • Primary endpoint: mRS 0–2 at 90 days: 41.3% (GHRP-6/EGF) vs. 39.8% (placebo); OR 1.06 (95% CI 0.78–1.44), p=0.71. No significant difference.
  • Infarct volume at 48h: No statistically significant reduction (mean difference –1.2 mL, 95% CI –3.8 to +1.4 mL, p=0.36).
  • IL-6 at 24h: A nominally significant 19% reduction (p=0.04) in the treatment arm, though this did not survive Bonferroni correction across the 8-biomarker secondary panel (adjusted p=0.32).
  • Safety: No excess serious adverse events. Mild transient hypotension occurred in 7.4% of the treatment arm vs. 3.1% placebo — consistent with GHRP-6's known GH-secretagogue cardiovascular effects and not unexpected at the doses used.

The nominally significant IL-6 reduction deserves attention. It is consistent with the β-arrestin-1/NF-κB suppression hypothesis and suggests some degree of peripheral GHS-R1a engagement — but the absence of any infarct volume or functional outcome signal suggests this peripheral anti-inflammatory effect was insufficient to translate into neuroprotection at the CNS level, further implicating CNS bioavailability as the primary failure mode.

The GHS-R1a Stroke Evidence Gap: What Preclinical Data Cannot Tell Us

The COURAGE-2 null result forces a rigorous reappraisal of the GHS-R1a stroke hypothesis at several levels. The preclinical literature, while mechanistically coherent, exhibits several systematic limitations that the field underweighted:

Route of Administration Mismatch

The highest-effect-size preclinical studies used intracerebroventricular (ICV) or high-dose intraperitoneal delivery — routes that circumvent or overwhelm BBB limitations. Of the 14 published rodent MCAo studies involving GHRP-6 or related GHS-R1a agonists (ghrelin, GHRP-2, hexarelin) through 2025, only 3 used IV administration at doses translatable to human pharmacology. These three studies showed considerably more modest effect sizes (15–22% infarct reduction vs. 35–55% in ICV studies), and the COURAGE-2 null result is directionally consistent with this more conservative subset.

Species-Specific GHS-R1a Expression Gradients

GHS-R1a expression density in murine cortical astrocytes and hippocampal neurons is substantially higher than in human post-mortem tissue equivalents (quantified by radioligand binding and single-cell RNA sequencing datasets in the Allen Brain Atlas 2023 release). This expression gradient directly predicts attenuated receptor-mediated effects in human CNS tissue even under conditions of complete receptor engagement — a pharmacodynamic ceiling that COURAGE-2's design did not account for.

Temporal Window Assumptions

The 6h therapeutic window assumed in the COURAGE-2 protocol was derived from the NLRP3/IL-1β suppression timeline in rodent models. Human ischemic stroke neuroinflammatory kinetics — particularly microglial M1 polarization and NLRP3 activation — appear to operate on a compressed and more heterogeneous timeline, with significant interindividual variability related to collateral circulation status. Without stratification by collateral grade (which COURAGE-2 did not perform), the enrolled population likely included a substantial fraction in whom the target biological window had already closed at time of treatment.

This kind of mechanistic evidence gap parallels issues recently documented in Khavinson peptide neuroprotection research — for a detailed analysis of how underpowered preclinical data translates poorly to human endpoints, see our coverage of the Cortagen neuroprotection Khavinson cerebral-cortex tetrapeptide evidence gap and null Epitalon head-to-head 2026.

Where Does the GHS-R1a Stroke Hypothesis Go From Here?

The COURAGE-2 null result does not invalidate GHS-R1a as a neuroprotection target — it invalidates the specific co-formulation strategy, delivery route, and population selection applied in this trial. Several scientifically credible paths forward exist in the preclinical and early clinical literature:

CNS-Penetrant GHS-R1a Agonists

Small-molecule GHS-R1a agonists with superior BBB penetration (logP >2.5, MW <500 Da) — including ibutamoren (MK-0677) and several investigational compounds in the GSK and Novo Nordisk pipelines — have demonstrated 10–30× greater CNS bioavailability than GHRP-6 in rat PK studies. Whether this pharmacokinetic advantage translates to superior neuroprotection in MCAo models remains to be formally tested in a head-to-head design, though preliminary 2024 data from a Shanghai Jiao Tong University group suggest ibutamoren reduces infarct volume by 29% in a 60-min MCAo model (IV administration, Sprague-Dawley, n=24; preprint, not yet peer-reviewed).

Intranasal Delivery for CNS Bioavailability

Intranasal GHRP-6 delivery has been explored in two 2023–2024 rodent studies specifically motivated by stroke neuroprotection. The olfactory-trigeminal transport pathway achieves significantly higher CNS GHRP-6 concentrations than IV delivery (estimated 8–15× improvement in AUC CSF/plasma ratio in Wistar rats), though the absolute CNS concentrations achieved still fall below GHS-R1a EC50 at doses considered safe for human scaling. This remains an active research direction without human PK data.

Decoupling EGF from GHS-R1a Strategies

Several research groups are now pursuing GHRP-6's anti-inflammatory GHS-R1a activity in isolation from any EGF co-administration, arguing that the EGFR transactivation hypothesis was pharmacologically premature and that the co-formulation introduced more confounds than benefits. This decoupled approach is more scientifically conservative and better aligned with the mechanistic data supporting GHS-R1a's β-arrestin-1/NF-κB anti-inflammatory arm.

Researchers designing experiments in this space should review our peptide research database for catalogued GHS-R1a agonist pharmacology, including comparative receptor binding profiles and in vivo PK parameters across species.

Contextualizing the COURAGE-2 Failure Within the Broader Peptide Neuroprotection Landscape

The COURAGE-2 null result is not an isolated event in 2026 peptide neuroprotection research. It joins a pattern of late-stage translational failures that reflects a structural problem in how the field constructs its evidence base: extraordinary preclinical effect sizes achieved via non-translatable delivery routes, in species with markedly different receptor expression profiles, in homogeneous genetic backgrounds that do not reflect human stroke population heterogeneity.

This same critique applies to activin pathway peptide modulators, where promising in vitro mechanistic data has repeatedly failed to survive rigorous in vivo testing — an issue also relevant to researchers tracking Follistatin 344 ocular safety signals and activin-FSH axis complications that emerged in parallel 2026 research streams. Similarly, the challenge of validating short peptide biological activity in complex human tissue — where cell-type-specific receptor expression, epigenetic context, and systemic metabolic state all modulate response — is well illustrated by thymic peptide research such as the Vilon (Lys-Glu) dipeptide chromatin remodeling and MAPK-ERK/JNK thymic signaling work.

The GHRP-6 field would benefit from adopting the kind of rigorous target engagement biomarker strategy now standard in CNS drug development: demonstrating GHS-R1a occupancy in human brain tissue (via PET radioligand studies), confirming downstream NLRP3 suppression in CSF, and gating Phase 3 enrollment on validated pharmacodynamic response rather than pharmacokinetic surrogates alone.

All researchers working with GHRP-6 or related GHS-R1a ligands in laboratory settings should review the peptide safety and handling guide for protocols on reconstitution, storage at –80°C under argon atmosphere, and stability monitoring — particularly given GHRP-6's documented susceptibility to oxidative degradation at the D-Trp residue, which can generate inactive or potentially immunogenic degradation products that confound experimental results.


Frequently Asked Questions

What was the primary endpoint of the COURAGE-2 Phase 3 trial, and why did GHRP-6 fail to meet it?

The COURAGE-2 primary endpoint was mRS score 0–2 at 90 days in patients with anterior circulation ischemic stroke treated within 4.5h of onset. GHRP-6/EGF co-formulation achieved 41.3% vs. 39.8% placebo (OR 1.06, p=0.71) — a null result. The failure is mechanistically attributable to: (1) inadequate CNS bioavailability of both GHRP-6 and rhEGF following IV bolus administration through a partially intact BBB; (2) a pharmacokinetic mismatch in plasma half-lives of the two co-formulated agents; and (3) insufficient GHS-R1a target engagement in perilesional brain tissue to drive the NLRP3/IL-1β suppression that underpinned the trial's biological rationale.

Does the COURAGE-2 null result invalidate GHS-R1a as a neuroprotection target in ischemic stroke?

No — but it substantially narrows the scientifically defensible hypothesis space. GHS-R1a-mediated β-arrestin-1/NF-κB suppression in microglia and astrocytes remains mechanistically plausible as a neuroprotective strategy. What COURAGE-2 invalidates is the specific use of GHRP-6 via IV bolus in a co-formulation with rhEGF. CNS-penetrant GHS-R1a agonists delivered via routes achieving therapeutic CNS concentrations (e.g., intranasal, intrathecal, or small-molecule oral agonists with superior BBB permeability) remain scientifically viable research directions, though none currently have Phase 2 human neuroprotection data.

Why did the GHRP-6/EGF co-formulation strategy fail pharmacokinetically?

Three factors converged: GHRP-6's short plasma half-life (~15–20 min IV) cleared GHS-R1a occupancy before rhEGF's EGFR engagement plateau was reached (rhEGF t½ ~2–4h), creating mechanistic decoupling. Additionally, rhEGF (MW 6,045 Da) cannot cross an intact BBB and achieves clinically negligible CNS penetration even under post-ischemic conditions within the 4.5h treatment window. The preclinical data justifying this strategy used continuous minipump delivery — not IV bolus — and the PK mismatch was not modeled before Phase 3 initiation.

What GHRP-6 research directions remain scientifically credible after COURAGE-2?

Several directions retain scientific merit: (1) intranasal delivery protocols for CNS bioavailability improvement, with ongoing rodent PK characterization; (2) small-molecule GHS-R1a agonists (e.g., ibutamoren/MK-0677) with superior BBB permeability for comparative MCAo studies; (3) decoupled GHRP-6 anti-inflammatory research targeting peripheral immune compartments (monocyte/macrophage GHS-R1a signaling) in contexts where CNS penetration is not required; and (4) GHRP-6 in GH-deficient populations for metabolic and anabolic endpoints, where the GHS-R1a hypothalamic axis remains the validated mechanism. All of these represent preclinical or early Phase 1 directions — no human efficacy data for any of these alternatives exists as of mid-2026.


This content is produced for licensed researchers, pharmacologists, and scientific institutions only. All information is presented for research and educational purposes. Nothing contained herein constitutes clinical dosage guidance, medical advice, or recommendation for human therapeutic use. Researchers are responsible for compliance with all applicable institutional, national, and international regulations governing peptide research.

Peptide Stack AI — AI-Powered Peptide Research. Built for Scientists. For questions, contact us at support@peptidestackai.com