The next-generation obesity pipeline: from “how much you lose” to “at what cost”

Evidence-first notes on bioscience and deep tech, at the edge of the lab and the market. Information only — not investment advice, not medical advice.

The 30-second version

• What. The next-generation obesity drugs push weight-loss efficacy along three vectors: triple agonists (retatrutide), oral small molecules (orforglipron), and amylin combinations (CagriSema). At the top end, retatrutide’s Phase 3 topline reported ~28% weight loss (approaching the bariatric-surgery range).
• So what — mind the phase gap. Much of the top-end number is Phase 3 / company topline (not yet peer-reviewed) or a tolerated-completer subgroup, not ITT. The real bottleneck is not the maximum percent lost but the efficacy–safety trade-off: lean/muscle-mass loss, tolerability ceilings, durability, and oral manufacturing scalability.
• Now what. Do not treat Phase 2 / company data as if it were approval-grade Phase 3 evidence. Watch the full peer review of retatrutide’s TRIUMPH-1, orforglipron’s efficacy gap versus injectables, and whether CagriSema actually demonstrates muscle preservation.

The five-minute read

The frontier question has changed

Beyond approved semaglutide (~14% class) and tirzepatide (~20% class), the next-generation pipeline extends efficacy in three directions: adding glucagon in a triple agonist, moving to pill form as an oral incretin, and combining with amylin on a satiety/muscle rationale. But at this frontier the question that actually matters is not “what is the maximum percent lost?” but “at what cost is it lost?” The bottlenecks are the lean-mass loss that accompanies rapid weight reduction, the tolerability ceiling at high doses, the durability that reverses after discontinuation, and the bioavailability and manufacturing scalability of oral agents.

Read the top-end numbers by phase

The strongest signal is Lilly’s triple agonist retatrutide. In Phase 2 (NEJM, N=338, peer-reviewed), 12 mg over 48 weeks produced −24.2% (placebo −2.1%); in the Phase 3 TRIUMPH-1 topline (company data, not yet peer-reviewed), 12 mg over 80 weeks reached 28.3%, approaching the bariatric-surgery range. Note, though, that the extension figure of 30.3% at 104 weeks is not ITT — it is a tolerated-completer subgroup (n=532). The oral small molecule orforglipron sits below the injectables in Phase 3 ATTAIN-1 (NEJM, N=3,127), at 12.4% for 36 mg over 72 weeks (versus tirzepatide ~20%, semaglutide ~15%) — so the strength of the oral route is access, not efficacy. By contrast the oral peptide high-dose oral semaglutide (OASIS 1, Phase 3, 50 mg −15.1%) is injectable-grade — the same word “oral” spans different classes.

The bottleneck lies outside the weight-loss percentage

Novo’s CagriSema (amylin + GLP-1) showed ~20.4% (treatment-policy) to 22.7% (adherence) in Phase 3 REDEFINE 1 and was submitted to the FDA on 2025-12-18 (the first GLP-1 + amylin filing) — tirzepatide-grade, but short of the triple. The amylin muscle-preservation rationale remains a hypothesis, not yet demonstrated. Amgen’s MariTide shows the mechanistic paradox of producing weight loss despite antagonizing GIP; its once-monthly (or less) dosing is a differentiator, but it is still at the Phase 2 / company-data stage. The shared limit is durability: in the STEP 1 extension, 17.3% weight loss was followed by 11.6 percentage points of regain one year after discontinuation (about two-thirds reversed) — a class limit the next generation is expected to share.

Deep dive

1. Background — extension vectors beyond the two approved agents

Obesity incretin therapy standardized weight loss in the −14 to −20% range with the approvals of semaglutide (a GLP-1 agonist) and tirzepatide (a GIP/GLP-1 dual agonist) (see Part 0 of this series). The next-generation pipeline branches from there along three axes: (1) triple agonists that add the glucagon receptor, (2) oral incretins in pill form, and (3) amylin-class combinations. Each axis carries a different trade-off, so reducing them to a single “weight-loss %” ranking is misleading.

2. What is newly shown — the ceiling of efficacy and its conditions

Triple agonist — retatrutide (Lilly; GIP/GLP-1/glucagon). In Phase 2 (NEJM, N=338, peer-reviewed), 12 mg over 48 weeks gave −24.2% (placebo −2.1%), the dose-response did not plateau, and 83% achieved ≥15% loss. The Phase 3 TRIUMPH-1 topline (company data, not yet peer-reviewed) reported 28.3% for 12 mg over 80 weeks, with 45.3% achieving ≥30% and 27.2% achieving ≥35%; the BMI≥35 extension was 30.3% at 104 weeks. But that 104-week 30.3% is not ITT — it is a tolerated-completer subgroup (n=532), and the Phase 2 (peer-reviewed) and Phase 3 (company topline) results must be kept separate.

The glucagon liver target (the MASH angle) appeared in a Phase 2a MASLD study (Nat Med, n=98): liver fat −81.4/−82.4% at 8/12 mg with 79–86% steatosis resolution (placebo 0%). But this is steatosis measured by MRI-PDFF, not MASH resolution or fibrosis improvement (no biopsy, n=98). The CKM liver-axis link is evidence-based but limited to steatosis.

Oral incretins — orforglipron / high-dose oral semaglutide. The oral non-peptide small molecule orforglipron (Lilly) gave 12.4% for 36 mg over 72 weeks in Phase 3 ATTAIN-1 (NEJM, N=3,127; placebo 0.9%), below tirzepatide ~20% and semaglutide ~15%. The ≥20% loss rate was a modest 20.1%, while glycemic control was injectable-grade (A1C −1.8%). A tolerability ceiling was also seen: AE-driven discontinuation was dose-dependent (5.3/7.9/10.3% at 6/12/36 mg vs placebo 2.7%), with nausea ~36% and vomiting ~24% at the top dose. By contrast the oral peptide high-dose oral semaglutide (OASIS 1, Phase 3 Lancet, n=667) was injectable-grade at 50 mg −15.1% (placebo −2.4%) — the same word “oral” spans peptide and non-peptide classes.

Amylin & combination — CagriSema (Novo). Fixed-dose cagrilintide 2.4 + semaglutide 2.4 once-weekly SC showed ~20.4% (treatment-policy) to 22.7% (adherence) in Phase 3 REDEFINE 1 and was submitted to the FDA on 2025-12-18 (the first GLP-1 + amylin filing). The amylin satiety/muscle-preservation rationale remains a hypothesis, and muscle preservation is undemonstrated (see the muscle/lean-mass note). Amycretin (oral/SC) is at an early pipeline stage.

MariTide (Amgen). Whereas tirzepatide is a GIP agonist, MariTide is a GIP antagonist (a GLP-1 agonist + GIP-antagonist antibody-peptide conjugate, ~21-day half-life) that still produces weight loss — a mechanistic paradox and an unresolved debate in GIP biology. Phase 2 (non-diabetic obesity) showed ~20% (company data); Phase 3 MARITIME is ongoing, with once-monthly (or less) dosing as the differentiator.

3. Strengths and limits of the methods (phase gap)

The strength is that many of these are large randomized controlled trials (orforglipron ATTAIN-1 N=3,127; retatrutide P2 NEJM N=338, both peer-reviewed). The decisive limit, however, is the phase gap: the top-end number (retatrutide 28.3%) is Phase 3 company topline (not yet peer-reviewed), and the 30.3% extension is a tolerated-completer subgroup. MariTide ~20% is Phase 2 / company data. Most comparisons are also cross-trial rather than head-to-head, so direct rankings between agents do not hold. The glucagon liver data, too, is a steatosis surface marker, not a MASH hard outcome. Erasing these distinctions would violate the analysis-standards demo-gap (the bio-domain version = phase gap).

4. Connections to neighbouring domains (CKM)

• Glucagon → liver (MASH). The liver-fat target of triple and glucagon arms connects to the CKM liver axis. Boehringer Ingelheim’s survodutide (GLP-1/glucagon) met its co-primary liver endpoint in Phase 3 SYNCHRONIZE-MASLD (Nat Med, n=216) at −12.2% for 6.0 mg over 48 weeks (placebo −1.0%), re-confirming the glucagon–liver thesis independently of retatrutide. But MASH hard outcomes (fibrosis progression, cirrhosis) are a separate task.
• Muscle preservation → sarcopenia/CKM. The lean-mass loss of rapid weight reduction has clinical implications for frail and CKM-vulnerable groups. Mitigation approaches such as amylin or bimagrumab combinations remain hypotheses with insufficient demonstration.

5. Commercialization and investment view (TRL, companies)

• retatrutide (Lilly, LLY) — Phase 3 (TRIUMPH), upper TRL. Topline announced; awaiting full peer review and regulatory filing.
• orforglipron (Lilly, LLY) — Phase 3 (ATTAIN) peer review complete. The access/manufacturing-scalability advantage of an oral small molecule is a thesis, but a separate hypothesis rather than a trial endpoint.
• CagriSema (Novo, NVO) — Phase 3 (REDEFINE) complete; FDA submission 2025-12-18.
• MariTide (Amgen, AMGN) — Phase 2 / company data; Phase 3 (MARITIME) ongoing.
• survodutide (Boehringer Ingelheim, private) — Phase 3 (SYNCHRONIZE), MASH direction.

Because clinical topline is share-price sensitive, any narrative of pipeline superiority or success/failure can be misread as a stock implication. This post describes phase attribution neutrally and offers no ranking or investment judgment.

6. The skeptic’s bottom line

Carrying forward the original knowledge-asset skeptic gate (§7) and per-item skeptic notes:

• [Required caveat] Do not treat Phase 2 / company topline as approval-grade or Phase 3 evidence (the bio-domain version of the analysis-standards demo-gap = phase gap). Efficacy must be attributed with trial name and phase, and manufacturer sponsorship must be stated alongside.
• Efficacy–safety frontier. Behind the maximum-weight-loss headline, the real bottlenecks are muscle loss, tolerability, durability, and access (oral manufacturing).
• retatrutide: glycemic/heart-rate signals of the glucagon arm; P2→P3 reproducibility (only topline exists); unresolved muscle loss.
• orforglipron: the efficacy–tolerability ceiling of high-dose oral (nausea ~36%, vomiting ~24%, discontinuation 10.3%).
• CagriSema: ~20% is tirzepatide-grade but short of the triple (retatrutide ~28%), so it is not “beyond expectations,” and muscle preservation is undemonstrated.
• durability (class limit): in the STEP 1 extension, 17.3% loss was followed by 11.6 pp regain one year after discontinuation (about two-thirds reversed) — expected to be shared by the next generation.

7. What to watch

1. Full peer review of retatrutide TRIUMPH-1 (durable efficacy on an ITT basis, glucagon safety) — if a significant additional weight loss versus tirzepatide is reproduced, the “triple frontier” is confirmed; if offset by plateau or safety, it weakens.
2. Whether orforglipron narrows the efficacy gap versus injectables (the access thesis).
3. Whether amylin combinations such as CagriSema demonstrate lean-mass preservation (muscle-preservation differentiation).
4. Whether the glucagon class converts steatosis into a MASH hard outcome (fibrosis).

References

The original knowledge asset (sources: []) did not specify DOIs/URLs, so the items below are attributed by trial name, journal, and phase. DOIs/URLs must be finalized before publication.

• retatrutide Phase 2 (obesity). New England Journal of Medicine. N=338, 12 mg 48 weeks −24.2%.
• retatrutide Phase 2a MASLD. Nature Medicine. n=98, liver fat −81.4/−82.4%.
• retatrutide Phase 3 TRIUMPH-1. Company topline (not yet peer-reviewed), 12 mg 80 weeks 28.3%.
• orforglipron Phase 3 ATTAIN-1. New England Journal of Medicine. N=3,127, 36 mg 72 weeks 12.4%.
• high-dose oral semaglutide, OASIS 1 Phase 3. The Lancet. n=667, 50 mg −15.1%.
• CagriSema Phase 3 REDEFINE 1 (Novo). ~20.4–22.7%, FDA submission 2025-12-18.
• MariTide Phase 2 (Amgen). Company data, ~20%.
• survodutide Phase 3 SYNCHRONIZE-MASLD (Boehringer Ingelheim). Nature Medicine. n=216, 6.0 mg −12.2%.
• semaglutide STEP 1 extension (durability). 17.3% loss, 11.6 pp regain one year after discontinuation.

Disclosure

This post is for information only and is not investment advice or medical advice.

COI note. The trials referenced are sponsored by listed pharmaceutical companies (Lilly LLY, Novo NVO, Amgen AMGN) and Boehringer Ingelheim (private). Quantitative claims are attributed exactly as in each trial’s phase and “company topline / peer-reviewed” status, stated as neutral fact with no softening. In particular, retatrutide’s 28.3% (company topline, not yet peer-reviewed) and the 104-week 30.3% (tolerated-completer subgroup, not ITT) are not approval-grade evidence. The author holds no position in, and has no financial interest in, the listed companies mentioned. Pipeline rankings are cross-trial comparisons, not head-to-head; this post takes no view on pipeline superiority.

Original knowledge-asset skeptic verdict: proceed-with-caveats (gate §7 phase-gap required caveat inserted).