Intracerebral hemorrhage, a type of stroke where blood leaks directly into brain tissue, remains one of the most damaging and least treatable medical emergencies. Within hours of the initial bleed, secondary processes kick in: the clot expands, surrounding tissue swells, and neurons begin to die. Researchers have long searched for agents that could interrupt those secondary processes and limit the total damage.
A phase 2 randomized trial registered under NCT07613437 and known as the GLICH trial is now positioned to test whether a GLP-1 receptor agonist peptide, specifically semaglutide administered by subcutaneous injection, can influence hematoma expansion, perihematomal edema, and neurological outcomes in patients who arrive at hospital within 24 hours of a brain bleed. The trial has not yet begun recruiting, but its design offers a detailed window into how researchers are thinking about GLP-1 biology in an acute neurological context.
What GLP-1 receptor agonists are
GLP-1 stands for glucagon-like peptide-1. It is a naturally occurring signaling molecule released in the gut after eating. Synthetic versions that mimic or extend its activity have been studied extensively for metabolic conditions, but researchers have increasingly noted that GLP-1 receptors are also expressed in the brain and on immune cells throughout the body.
That distribution has sparked interest in whether GLP-1 receptor agonist peptides might do more than regulate blood sugar. Early preclinical and observational data have suggested anti-inflammatory and neuroprotective signals, though large prospective human trials in acute brain injury settings remain sparse. The GLICH trial is designed to generate exactly that kind of controlled human data.
Trial design and patient population
The trial is an investigator-initiated, open-label, phase 2 study with blinded endpoint assessment. That design, sometimes called PROBE, allows clinicians to know which treatment a patient receives while keeping the people measuring outcomes unaware of group assignment. It is a common compromise when blinding all parties to an injectable therapy is logistically difficult.
Researchers aim to enroll 200 adults with primary intracerebral hemorrhage located in one of two specific brain regions: the putamen, with a blood volume between 10 and 30 milliliters, or the thalamus, with a blood volume between 5 and 15 milliliters. Those size thresholds matter because very small bleeds may resolve on their own without room to demonstrate a treatment effect, while very large bleeds carry a high early mortality that can obscure any benefit from a therapeutic agent.
Patients must present within 24 hours of the last time they were known to be well. That 24-hour window reflects the belief that the secondary injury processes most likely to be modifiable, particularly edema formation and inflammatory cascades, are still actively developing in that timeframe. Sites span three centers: Wenzhou and Hefei in China, and hospitals in the Hong Kong Special Administrative Region, with recruitment targets proportional to each center's catchment population.
Dosing schedule
Participants randomly assigned to the intervention group receive subcutaneous semaglutide at a concentration of 0.5 mg per milliliter at four time points: the day of enrollment (day 0), then again on day 7, day 14, and day 21. The comparison group receives standard care alone. The four-injection schedule across three weeks was likely chosen to maintain receptor engagement through the acute and subacute phases of hemorrhagic injury without extending treatment into the longer recovery period, though the trial record does not explicitly state that rationale.
Imaging endpoints
The primary window for measuring brain changes relies on serial non-contrast computed tomography scans. Scans are scheduled at enrollment, at three days plus or minus one day, and at seven days plus or minus one day. Researchers will use these images to track three related but distinct phenomena: overall hematoma volume, hematoma expansion over time, and perihematomal edema, the swelling that develops in tissue immediately surrounding the clot.
Hematoma expansion is considered a particularly important target because it is an independent predictor of poor outcomes and occurs in a meaningful proportion of patients in the hours to days after onset. Perihematomal edema, meanwhile, can persist and worsen for days even after the initial bleed stabilizes, representing a potentially separate therapeutic window.
A brain MRI is also planned around days 5 through 9 after the hemorrhage. MRI offers sensitivity to neuronal injury patterns that CT cannot detect, giving researchers a complementary picture of how much brain tissue has been affected beyond the visible clot.
Clinical and biological measurements
The trial collects functional and cognitive scores at three time points: day 0 (baseline), day 90, and day 180. The tools chosen cover a range of domains. The National Institutes of Health Stroke Scale measures overall neurological deficit severity. The modified Rankin Scale captures how much disability a person has in daily life. The Montreal Cognitive Assessment and the Mini Mental State Examination both probe cognitive function, which can be subtly or severely affected by hemorrhage in the putamen or thalamus.
Alongside imaging and clinical scores, blood samples are drawn at baseline, day 3, day 7, day 14, and day 30 for what the trial record calls omics analysis. Omics is a broad term covering genomics, proteomics, metabolomics, and related large-scale molecular profiling techniques. Collecting samples at multiple time points allows researchers to map how biological markers change across the acute and early recovery phases, and potentially to identify signatures that predict who responds to treatment.
What the trial can and cannot tell us
Phase 2 trials are primarily designed to assess whether a treatment produces a measurable biological signal and to gather safety data in a specific population, rather than to deliver a definitive answer about effectiveness. With 200 participants split equally between two groups, the GLICH trial is sized to detect meaningful differences in imaging endpoints, but it would not be large enough on its own to confirm clinical benefit or establish the approach as a standard of care.
Still, the trial record represents a meaningful step. GLP-1 receptor biology in the central nervous system is an active area of preclinical research, and the anti-inflammatory properties that have been hypothesized in animal models need prospective human data to evaluate properly. If the phase 2 results show a reduction in edema or hematoma expansion, that would generate a hypothesis worth testing in a larger phase 3 program.
The omics data collected alongside imaging and clinical scores could also yield insights that go beyond a simple yes or no on semaglutide. Identifying which patients show the strongest biological response, and what molecular features predict that response, could help future researchers design more targeted trials or understand the mechanisms driving secondary injury after intracerebral hemorrhage more broadly.
