The Sunday Brief · Issue №29 · July 19, 2026
A phase 2 liver trial where the add-on drug didn't help — and the base drug quietly did
4 min read · Week of July 13–19, 2026
This week's clearest signal came from a phase 2 randomised trial in metabolic steatohepatitis: adding zalfermin to semaglutide did not improve liver fibrosis over placebo, while semaglutide alone showed a nominally significant improvement. It is a useful reminder that combination logic does not always beat the simpler regimen, and that the strongest human evidence this week points to the base therapy, not the novel add-on. Two large observational studies and a pharmacovigilance analysis fill in the safety and outcomes picture around GLP-1 and dual agonists.
Research spotlight: human trials lead, observational data fills in
The week's strongest evidence is human, not preclinical, but it is mixed. Two randomised trials returned qualified results — a fibrosis add-on that did not separate from placebo, and a stroke combination that missed its primary endpoint but showed a subgroup signal. Two large observational cohorts add real-world associations, which suggest patterns rather than prove causation.
Phase 2 RCT: zalfermin added to semaglutide did not improve liver fibrosis
In a phase 2, dose-ranging, double-blind randomised controlled trial in people with fibrosis and cirrhosis from metabolic dysfunction-associated steatohepatitis, adding zalfermin to semaglutide did not improve fibrosis over placebo. Semaglutide alone showed a nominally significant improvement. The result argues against the combination benefit that was hypothesised, and leaves the base therapy as the more supported arm.
In a phase III open-label randomised trial in acute ischemic stroke, combined epidermal growth factor and GHRP-6 did not improve overall disability at six months. A pre-specified subgroup of severe strokes showed reduced disability and mortality. Subgroup findings are hypothesis-generating, not confirmatory, and open-label designs carry a higher risk of bias — this needs a blinded, adequately powered confirmatory trial before the subgroup result can be relied upon.
In a matched real-world cohort of over 16,000 patients with type 2 diabetes and established atherosclerotic cardiovascular disease, tirzepatide was associated with a lower one-year risk of major cardiovascular events and death than other GLP-1 receptor agonists. As an observational study, this is an association, not causal proof; residual confounding and channelling between drug choices cannot be ruled out.
In a large real-world target trial emulation, people prescribed semaglutide for weight management had a lower one-year risk of developing diabetes than those prescribed liraglutide. No difference in cardiovascular events was detected. The emulation design strengthens the comparison, but this remains observational — an association drawn from prescribing data rather than a randomised head-to-head.
Expert consensus: semaglutide guidance for patients with cardiovascular disease and obesity
An Italian expert panel published consensus guidance on semaglutide use in patients who have both cardiovascular disease and overweight or obesity, drawing on SELECT trial data. It also outlines practical barriers to wider use, including reimbursement and clinical inertia. This is expert opinion synthesising prior trial evidence, not new primary data.
Community watch: BPC-157 and the anecdote-versus-evidence gap
This week the community conversation converged almost entirely on BPC-157, and it is a clean illustration of the publication's core tension: a peptide with very little human trial data and a large, contradictory volume of anecdote. These are community reports, not clinical evidence — what makes them worth watching is the mix, with cautionary and negative accounts sitting alongside the enthusiastic ones.
The BPC-157 evidence question, asked plainly
A community thread frames the central tension directly: BPC-157 has almost no human trial data, yet the volume of anecdotal reports is large. The post asks at what point anecdotal signal should carry interpretive weight. It is a useful articulation of an evidence-quality question, though the thread itself offers discussion rather than data.
A cautionary post on BPC-157 risks
A user posted a caution to others about downsides they associate with BPC-157. This is a single anecdotal account, not a documented adverse-event report, but negative community reports are worth logging alongside the more positive ones to keep the picture balanced.
Mixed anecdotes on BPC-157 for injury recovery
A thread asking whether BPC-157 actually helped with injuries gathered mixed responses, with some reporting perceived benefit and others none. A separate post reported no results at all. Taken together these are conflicting anecdotes, which is exactly why they cannot substitute for controlled data.
A product-handling question: reconstituted BPC-157 looking chunky
A user asked whether their BPC-157 appearing chunky after reconstitution is normal, raising product-quality and handling concerns. It is a reminder that for research peptides sourced outside a clinical setting, appearance and handling questions are a recurring practical worry within the community.
One thing to think about
Two of this week's threads point the same way from opposite directions: the strongest human trial favoured the simpler regimen over the fashionable add-on, while the loudest community conversation was about a peptide with almost no human trial data at all. The distance between what is studied and what is discussed is, this week, unusually easy to see.
Get the next issue
Subscribe to The Sunday Brief
Every Sunday: what the studies say, what the community is finding, and which vendors are credible — straight to your inbox.
Free. Unsubscribe anytime. No affiliate links. No spam.