Peptide Research 2024–2026: What the Studies Actually Show

For the approved GLP-1-based agents, the body of evidence became unusually robust in 2024–2026 – large, randomised, placebo-controlled trials in journals such as the New England Journal of Medicine (NEJM). The first direct head-to-head comparison, SURMOUNT-5 (NEJM 2025), showed markedly greater weight loss for Tirzepatide than for Semaglutide (around 20% versus around 14%), with overall comparable tolerability.

In parallel, the indications have expanded beyond pure weight reduction. The ESSENCE trial (NEJM 2025) demonstrated for the first time that Semaglutide improves the advanced fatty liver disease MASH – prompting the US agency FDA to grant a corresponding approval in 2025. The large cardiovascular trial SELECT (NEJM 2023, over 17,000 participants) had already shown a roughly 20% reduction in major cardiovascular events that can only partly be explained by weight loss. And with Orforglipron (ATTAIN-1, NEJM 2025), a first substance of this class that can be taken as a tablet is moving closer to approval.

• SURMOUNT-5 (NEJM 2025): Tirzepatide ~20% vs. Semaglutide ~14% weight loss – first direct comparison.
• ESSENCE (NEJM 2025): Semaglutide improves the fatty liver disease MASH → 2025 FDA approval for it.
• SELECT (NEJM 2023): ~20% fewer major cardiovascular events, partly independent of weight loss.
• Orforglipron (ATTAIN-1, NEJM 2025): oral GLP-1 tablet, approval application submitted to the FDA.

Behind the approved agents, a new generation of multi-agonists is pushing forward that is still in clinical testing. Retatrutide acts on three receptors at once (GIP, GLP-1 and glucagon). In the peer-reviewed Phase 2 trial (Jastreboff et al., NEJM 2023) it achieved a weight loss of a good 24% after just under a year; a separate Phase 2a study (Nature Medicine 2024) also showed a very pronounced reduction in liver fat. The headline figures of around 30% circulating in the media, however, so far come from the Phase 3 programme TRIUMPH and for now exist only as company press releases (topline) – not as a complete peer-reviewed publication. There is no approval yet.

Survodutide (a glucagon/GLP-1 dual agonist) showed a clear improvement in liver disease in a Phase 2 trial on MASH (Sanyal et al., NEJM 2024) and received expedited review procedures (FDA Breakthrough Therapy, EMA PRIME); the Phase 3 trials are ongoing. That more drug does not automatically mean more success is shown by CagriSema (Cagrilintide plus Semaglutide): in REDEFINE 1 (NEJM 2025) the weight loss, at around 23%, fell short of the high expectations, and in a direct comparison the combination missed non-inferiority against Tirzepatide.

• Retatrutide: Phase 2 (NEJM 2023) ~24% weight loss; ~30% figures from Phase 3 (TRIUMPH) so far only as company topline.
• Survodutide: Phase 2 success in MASH (NEJM 2024), expedited review procedures, Phase 3 ongoing – not yet approved.
• CagriSema (REDEFINE 1, NEJM 2025): ~23% weight loss – below expectations; missed non-inferiority against Tirzepatide.
• Assessment: promising, but "investigational" – always read specific figures together with trial status.

Serious research also means naming failures. The hypothesis that GLP-1 agents could protect the brain beyond their metabolic effects (the "type 3 diabetes" idea of Alzheimer's disease) was considered one of the most exciting bets of recent years. The large EVOKE/EVOKE+ trial with oral Semaglutide tested it in several thousand people with early Alzheimer's disease.

The result (topline at the end of 2025, full publication in The Lancet 2026): the primary endpoint was missed – there was no slowing of cognitive decline, despite a few favourable biomarker signals. The planned extension was discontinued. This is an important damper on the expectation that the incretin class is a cure-all far beyond metabolism – and a good example of why one must present the success and failure of the same substance in a balanced way.

• EVOKE/EVOKE+ (oral Semaglutide in early Alzheimer's disease): primary endpoint missed.
• No slowing of cognitive decline; extension phase discontinued (The Lancet 2026).
• A damper on the "type 3 diabetes" hypothesis – a substance can shine in one indication and fail in another.

While the approved class delivers study after study, little has moved with the peptides traded in the "biohacking" scene – except for the regulatory pressure. For BPC-157, first studies in humans did appear, but these are small, uncontrolled case series (predominantly from a single research group in a niche journal). A systematic review (HSS Journal 2025) screened 544 publications and found 36 relevant ones – of which 35 were from the animal/laboratory domain and only a single one clinical. The conclusion: preclinically plausible, practically unproven in humans. The FDA has listed BPC-157 on the 503A "Category 2" list since 2023 (significant safety concerns: immunogenicity, impurities, lack of human data); WADA bans it in sport.

Similarly with TB-500: it is important to distinguish between TB-500 (a synthetic fragment) and the body's own full protein Thymosin-β4. Genuine clinical trials (e.g. SEER-1, Int J Mol Sci 2023) are ongoing for the full protein in eye diseases – not for "TB-500" as a wellness agent. For Epitalon, a first independent replication of telomerase activation did succeed in 2025 (Biogerontology 2025), but exclusively in cell culture – this is expressly not evidence of an anti-ageing effect in living humans. And CJC-1295/Ipamorelin went through several FDA advisory procedures (PCAC) in 2024 that came out against a compounding clearance – among other things because of cardiac concerns and insufficient evidence.

• BPC-157: only small, uncontrolled case series; systematic review 2025 → 35 preclinical vs. 1 clinical study.
• FDA 503A "Category 2" (safety concerns) and WADA ban apply to BPC-157, TB-500 and CJC-1295.
• TB-500 ≠ Thymosin-β4: genuine trials are ongoing for the full protein, not for wellness "TB-500".
• Epitalon: telomerase effect replicated in 2025 only in cell culture – no evidence in living humans.
• CJC-1295/Ipamorelin: FDA advisory panels came out against a compounding clearance in 2024.

Perhaps the most consequential shift is not a single substance but a method. AI-assisted peptide design – based on structure prediction (AlphaFold), generative models and de novo design tools – is changing how new peptides come into being in the first place. Instead of modifying natural compounds, molecules can increasingly be designed on the computer towards a target, for example against attack points previously considered "undruggable".

This is flanked by chemical tricks (cyclisation, stapling, lipid conjugation) that make peptides more stable and longer-acting. The market for peptide therapeutics is growing accordingly (according to market researchers such as Precedence Research, roughly 49 billion US dollars in 2024, with moderate annual growth). Important for context: this progress concerns serious drug development – it does not ennoble any of the freely traded "wellness" peptides.

• AI design (AlphaFold, generative models, de novo design) is the central methodological breakthrough.
• Chemical stabilisation (cyclisation, stapling, lipidation) extends duration of action and bioavailability.
• Market growth demonstrates the interest – but says nothing about the safety of individual grey-market peptides.