Does Retatrutide Plus Cagrilintide Beat CagriSema for Weight Loss?

On 15 September 2026 a group at the University of Copenhagen and Novo Nordisk published a rat study in Nature Metabolism that has been ricocheting around peptide forums ever since. Petersen and colleagues showed that cagrilintide combined with retatrutide produced more weight loss than either drug alone, and more than matched-dose combinations that used semaglutide or tirzepatide as the incretin partner. In other words, it beat CagriSema in a head-to-head. Grey-market vendors immediately started selling the pre-blend at a 5:1 ratio. This article covers what the paper actually showed, what the human data says about each drug alone, and what to watch on bloodwork if you are already running the pair.

What the Petersen 2026 paper actually showed

The study used diet-induced obese male Sprague Dawley rats fed a high-fat, high-sucrose diet for at least 12 weeks before treatment. Four separate rat experiments, most running for 13 to 14 days of daily subcutaneous injections. The main head-to-head used 5 nmol/kg of each compound, monotherapy or in combination, against vehicle controls.

Here is what the data from the paper's own Source Data tables actually looks like:

Retatrutide alone beat semaglutide and tirzepatide monotherapy by roughly 2.5x at the same molar dose, which lines up with the human dose-response and is the reason the community treats it as the top of the incretin ladder. Cagrilintide alone was middling for weight but produced the deepest single-drug reduction in food intake. The combination doubled retatrutide's effect and cut food intake by more than two thirds.

The pair-feeding methodology is what makes this paper unusual

Most rat studies stop at "the drug caused weight loss." Petersen went further with two control designs that are worth understanding, because they are the source of every honest claim you can make from this paper.

With pair-feeding, a control group of rats gets the exact food ration the drug-treated group ate, but no drug. If they lose the same weight, the drug did nothing beyond suppressing appetite. If the drug group lost more, the surplus is drug-specific: energy expenditure, thermogenesis, altered substrate handling.

Weight-matching does something different. A separate control group is calorie-restricted independently until it reaches the same weight loss the drug caused. Any marker that improves in the drug group but not the weight-matched group is doing something the weight loss alone cannot explain.

Petersen's combination group lost 18.1% of body weight by day 14. The pair-fed group lost 10.6%. The weight-matched control was engineered to reach 17.4%, as designed. So roughly 60% of the drug's weight loss was explained by eating less, and roughly 40% was drug-specific. That drug-specific fraction rose over the study, climbing from zero on day 2 to about 64% by day 14. That extra fraction is what a "five-receptor" strategy is trying to maximise, and it is a real finding.

Standard bloodwork was dominated by weight loss, not by the drugs

This is the part that gets misquoted. Yes, cagrilintide plus retatrutide dropped triglycerides by about 55% and insulin by about 50%, and those look like drug effects if you only compare the combination to vehicle. But the weight-matched control, calorie-restricted to the same weight loss without any drug, dropped triglycerides even further, to 44.5% of vehicle versus 84.6% for the drug group.

In other words, the calorie restriction was doing the heavy lifting on standard clinical chemistry. Where the drug added value beyond calorie restriction turned up in plasma proteomics and brain transcriptomics, not in a lipid panel. The abstract says the combination "highlights enrichment of bioenergetic processes" in plasma proteomics, and drives "convergent central neuronal programmes linked to energy balance regulation" in the hypothalamus and dorsal vagal complex. Specific pathway lists sit behind the paywall.

What "five-receptor polypharmacology" actually means

Retatrutide agonises three receptors: GLP-1R for satiety and insulin secretion, GIPR for satiety and adipose metabolism, and GCGR for hepatic glucose output plus, in obese patients, energy expenditure and lipolysis. Cagrilintide adds two more: the amylin receptor (a calcitonin receptor complex with RAMP1 or RAMP3, which drives satiety through the area postrema) and the calcitonin receptor itself.

Combined, the injection hits GLP-1R, GIPR, GCGR, AMLNR and CALCR at the same time. Petersen's argument is that the additive weight loss and the drug-specific proteomic and transcriptomic signatures justify hitting all five receptors together, and the paper's discussion frames the finding as guidance for future single-molecule pentagonists. Novo Nordisk has published research programs pointing in exactly this direction.

Why the pair beat CagriSema in rats

The mechanistically interesting part of the head-to-head is not that cagrilintide plus retatrutide beat cagrilintide plus semaglutide. It is what that says about CagriSema's own design.

CagriSema is cagrilintide 2.4 mg plus semaglutide 2.4 mg. Semaglutide is GLP-1 alone. Retatrutide is GLP-1 plus GIP plus glucagon. If you swap semaglutide for retatrutide in the combination with cagrilintide, you are not adding a new mechanism, you are upgrading the incretin backbone from one receptor to three. So when Petersen's data shows cagri plus reta beating cagri plus sema, it is telling you that the semaglutide component of CagriSema was doing less work than the equivalent slot in a triple agonist would do.

That has a practical implication for the CagriSema plus retatrutide stack some people run: the semaglutide arm of CagriSema is largely redundant with retatrutide's own GLP-1 agonism, and you are paying an extra dose of GLP-1 tone (plus its extra GI burden) for no unique receptor engagement.

What the human data actually says

Retatrutide and CagriSema have serious human trials behind them. The combination in this paper does not.

Retatrutide alone in humans

The Phase 2 obesity trial (Jastreboff et al., 2023, NEJM) enrolled 338 adults across placebo and retatrutide doses from 1 mg to 12 mg weekly, over 48 weeks. Weight loss climbed dose-dependently:

At 12 mg, 83% of participants lost at least 15% of body weight (Jastreboff et al., 2023).

The Phase 2 MASLD substudy (Sanyal et al., 2024, Nature Medicine) reported an 82% reduction in liver fat at 12 mg over 48 weeks, with 86% of that arm reaching a normal liver fat percentage (Sanyal et al., 2024). The Phase 2 T2D substudy (Rosenstock et al., 2023, Lancet) showed HbA1c reductions up to 2.02% at 12 mg (Rosenstock et al., 2023).

TRIUMPH-1, the Phase 3 obesity trial, reported topline results by Lilly press release in May 2026: 28.3% weight loss at 12 mg over 80 weeks in 2,339 adults, and 30.3% weight loss at 104 weeks in a BMI ≥35 extension. The peer-reviewed manuscript is not yet indexed on PubMed, so those numbers should be treated as topline rather than as literature. Adverse events at 12 mg included nausea in 42.4% and vomiting in 25.3%, dysesthesia in 12.5% versus 0.9% on placebo, and roughly 18% discontinuation, some of which Lilly attributed to perceived excessive weight loss rather than adverse events. A network meta-analysis (Zhang et al., 2026) puts the pooled retatrutide resting heart rate rise at +3.46 bpm (Zhang et al., 2026).

CagriSema in humans

REDEFINE 1 (Garvey et al., 2025, NEJM) randomised 3,417 adults without diabetes across CagriSema, semaglutide 2.4 mg, cagrilintide 2.4 mg, and placebo over 68 weeks (Garvey et al., 2025):

Cagrilintide added about 5.5 percentage points on top of maximal-dose semaglutide, at the cost of gastrointestinal adverse events in 79.6% of the combination arm versus 39.9% on placebo. Novo Nordisk missed its own 25% target with the combination, which cooled the commercial reaction to the trial.

REDEFINE 2 (Davies et al., 2025, NEJM) covered CagriSema in type 2 diabetes with 68-week weight loss of 13.7% versus 3.4% on placebo (Davies et al., 2025). REDEFINE 5 (Yamauchi et al., 2026, Lancet Diabetes Endocrinol) covered a Japanese and Taiwanese cohort with 18.4% loss on CagriSema versus 11.9% on semaglutide alone (Yamauchi et al., 2026). Novo Nordisk filed the CagriSema NDA with the FDA in December 2025. A decision is expected in Q4 2026.

There is no human trial of cagrilintide plus retatrutide

Zero. That is the single most important line in this section. Every published claim about the combination in humans is extrapolation from the two-drug monotherapy datasets above and from a 14-day rat study. Anyone running the stack is running it on that basis.

What to watch on bloodwork if you are running this pair

This section extrapolates from single-agent monitoring data. The retatrutide monitoring canon is covered in more depth in the retatrutide bloodwork guide, and the class baseline sits in the GLP-1 bloodwork guide for semaglutide and tirzepatide.

Draw everything at trough, immediately before your weekly injection, and hold every follow-up draw to the same point relative to the injection. Otherwise you are measuring your dosing schedule rather than your physiology.

Baseline panel, before the first dose of either compound

Pull the glucose axis first: fasting glucose, HbA1c, fasting insulin, HOMA-IR. Both compounds move all four. On the liver, run ALT, AST, GGT and total bilirubin, because retatrutide's glucagon arm can transiently raise ALT and AST as hepatic fat clears.

Pancreatic enzymes lipase and amylase belong on the panel because pancreatitis is a GLP-1 class signal (Wen et al., 2025, RR 1.44 across 62 RCTs) (Wen et al., 2025). Amylin agonists alone do not carry that signal. For lipids, get total cholesterol, LDL, HDL, triglycerides and ApoB. Retatrutide lowers triglycerides and LDL through an ANGPTL3/8 pathway that goes beyond calorie restriction.

Thyroid needs TSH and free T4, given the class labelling around medullary thyroid carcinoma and the fact that rapid weight loss itself shifts thyroid binding. On the kidney, run creatinine, eGFR and cystatin C. Cystatin C is the muscle-independent measure and matters for muscular users, whose creatinine misreports kidney function in both directions.

Haematology means haemoglobin, haematocrit and ferritin, and this is critical if TRT is on board. Add CRP for inflammation, since semaglutide reduced CRP 39 to 48% in STEP (Verma et al., 2022) (Verma et al., 2022) and retatrutide should do the same. If you are on TRT, add total testosterone, free testosterone, SHBG and estradiol. Weight loss raises SHBG and lowers estradiol, so free T will drift down even if your protocol has not changed.

Week 4 check

The early markers most likely to move. Fasting glucose drops 0.3 to 0.6 mmol/L; do not read that as hypoglycaemia unless you also inject exogenous insulin. ALT may rise transiently as hepatic fat clears (Newsome et al., 2021 semaglutide NASH data supports the pattern) (Newsome et al., 2021). If GGT stays flat while ALT and AST rise, the picture is hepatic remodelling, not hepatotoxicity.

Triglycerides typically drop 15 to 25% by week 4 on retatrutide alone. Creatinine rises transiently if GI side effects have caused dehydration. Lipase rose 28 to 31% on liraglutide in LEADER (Steinberg et al., 2017) (Steinberg et al., 2017), and the pattern is broadly consistent across the class. Positive predictive value for actual pancreatitis is under 1%, but the marker still deserves the check. Resting heart rate begins climbing on retatrutide.

Week 12 check, full panel plus body composition

Repeat the full lipid panel including ApoB. Combined class effect expected: triglycerides down 25 to 40%, LDL down 10 to 20%. HbA1c should be down 0.5 to 1.2%, even in non-diabetics. HOMA-IR below 2.0 is the target, computed as fasting glucose (mmol/L) × fasting insulin (mIU/L) ÷ 22.5.

The liver panel is now in its peak deviation window. Use the GGT split to distinguish remodelling from insult. Cystatin C at this point matters more than creatinine, because falling creatinine is likely lean mass loss rather than better kidney function. Albumin below 38 g/L means your protein intake has cratered. Watch HCT and ferritin to catch either hemoconcentration or protein-driven iron drop.

If you are on TRT

Two things change. First, haematocrit now has two competing pressures: androgen-driven erythrocytosis pushing it up, and rapid weight loss causing plasma volume shifts that can push it up transiently and then normalise. The standard 54% ceiling still applies; the check frequency should go up. Second, aromatase falls with fat loss (Cohen, 2001) (Cohen, 2001), so estradiol will drift down on a stable TRT dose. Expect to reduce any aromatase inhibitor dose by 25 to 50%. SHBG will rise (Niskanen et al., 2004) (Niskanen et al., 2004), which pushes free T down even if total T is stable, and you may need a 10 to 20% testosterone dose bump to hold free T in target.

Red-line thresholds where you stop or reduce

ALT or AST above 3× ULN, or an ALT/AST rise with concurrent GGT rise more than 1.5× baseline: hold, retest in two weeks. If it does not resolve, stop.

Lipase above 5× ULN, or any lipase elevation with epigastric pain radiating to the back: stop retatrutide immediately and image the pancreas.

Resting HR consistently above 100 bpm at rest, or an increase above 10 bpm sustained past week 16: reduce retatrutide.

Fasting glucose below 3.5 mmol/L on two occasions. Only realistic if exogenous insulin is on board; cut mealtime insulin 50% per the pramlintide precedent (Ratner et al., 2004, boxed warning for pramlintide plus insulin) (Ratner et al., 2004).

Haematocrit above 54% on TRT co-administration: therapeutic phlebotomy. The standard TRT threshold does not relax.

Albumin below 35 g/L, or ongoing loss of more than 1% body weight per week for four straight weeks: reduce dose and push protein intake up.

The muscle-sparing claim, revisited

This is where the story turns against amylin. The most useful human data on body composition in this class is the REDEFINE 1 DXA substudy, presented as conference material at ObesityWeek 2025 and re-presented at ECO 2026, currently unpublished in peer-reviewed form. The relevant fat-to-lean split of total weight lost was:

Cagrilintide alone had the worst fat-to-lean split of the three active arms. Adding cagrilintide to semaglutide moved the split in the wrong direction, from 69.7% to 66.9% fat share. The amylin monotherapy result specifically inverts the marketing.

Mechanistically that makes sense. Amylin is a satiety peptide that acts on the area postrema and hypothalamus. It does not act on skeletal muscle protein synthesis, and it does not modulate IGF-1, mTOR, or androgen receptors. It is a hunger suppressant with a different mechanism from GLP-1, which is why stacking them adds efficacy, but the extra tissue coming off is not preferentially fat.

There is no human DXA data on cagrilintide plus retatrutide. Retatrutide's own Phase 2 body composition (Coskun et al., 2025) sits in roughly the same range as tirzepatide's DXA breakdown from SURMOUNT-1 (Look et al., 2025) (Look et al., 2025), commonly reported as around 75% fat / 25% lean of total weight lost, which is stronger than either amylin arm above. If you were adding cagrilintide to retatrutide to protect muscle, the case for that has just failed its own confirmatory trial. This is the same argument the eloralintide article made about amylin monotherapy generally.

Practical implications for people already running the pair

If you have already decided to run this stack, here is what actually matters.

The pre-blend vials are a vendor problem, not a pharmacology problem

Retatrutide is investigational, sold under Lilly's TRIUMPH programme. Cagrilintide exists only in Novo's CagriSema clinical stock. Neither is legal to compound in the US. Grey-market vendors work around this by selling both as "research chemicals not for human use." The dominant pre-blend on the market is a 5:1 mass ratio, typically 12.5 mg retatrutide plus 2.5 mg cagrilintide in a single vial, marketed as "Reta/Cagri Blend", "RETACAG", "GLP-3 Cagrilintide Blend" and similar names. A higher-dose 20 mg / 4 mg vial exists at the same ratio.

Two problems with the pre-blend. First, the 5:1 ratio was picked to loosely mimic CagriSema's molar dosing, not because any published trial tested that ratio. Second, a legitimate COA on a pre-blend requires two HPLC assays plus mass spectrometry confirmation, one per peptide. Most vendor COAs verify only the total peptide mass or the retatrutide fraction. Running the stack as two separate vials at least lets you titrate each independently, adjust the ratio if you tolerate one better than the other, and audit each COA.

Rat mg/kg is not a dose

The paper's headline efficacy sat at 5 nmol/kg per compound. That does not translate to a human dose. Standard allometric scaling (FDA 2005 guidance, Nair and Jacob, 2016) (Nair and Jacob, 2016) divides rat mg/kg by 6.2 for a mid-sized adult rat to give a first-in-human starting dose in mg/kg. That is a scaling equivalence for a Phase 1 starting dose, not a promise the pharmacodynamics translate.

Wong, Siah and Lo (2019) put the probability that a metabolic drug at Phase 1 makes it to approval at about one in five, and preclinical attrition adds another 40 to 50% loss layer that Wong does not measure (Wong et al., 2019). The plain read: rat weight loss does not predict human weight loss, and four out of five metabolic candidates that clear a first-in-human trial still fail. Rat mg/kg is not a dose, it is a rumour.

The paper also tested higher rat doses that hit real ceilings. At 30 nmol/kg per compound, food intake dropped by roughly 89%, one animal was terminated for adverse behaviour, and at 10 nmol/kg per compound blood glucose fell into hypoglycaemia range. Human tolerability will inherit that ceiling.

The decision tree: add cagrilintide, dose up retatrutide, or add tirzepatide

If you are at 8 mg retatrutide and want more weight loss, the three options are not equivalent.

Dose up retatrutide 8 to 12 mg. This has the data. Jastreboff 2023 showed an incremental 1.4 percentage points from 8 to 12 mg over 48 weeks; TRIUMPH-1 showed roughly 2.3 percentage points from 9 to 12 mg over 80 weeks. Discontinuation for adverse events rose from about 7% at 9 mg to 11% at 12 mg, and nausea hit 42% at the top dose. The marginal weight per marginal milligram is falling, but the direction is at least documented in humans.

Add cagrilintide. No published human trial has tested cagrilintide with retatrutide. The closest analogue, semaglutide plus cagrilintide in REDEFINE 1, added about 7 percentage points on top of semaglutide alone but moved the fat-to-lean split slightly worse. Whether that additivity holds when the base is already a triple agonist is unknown. Amylin adds an independent satiety mechanism through the area postrema that retatrutide does not directly engage, which is the pharmacological argument for stacking. The counter-argument is that adding it after retatrutide is already suppressing appetite hard may hit the same ceiling with additional GI cost.

Add tirzepatide. Mechanistically redundant. Retatrutide already agonises GLP-1 and GIP, so adding tirzepatide doubles that tone without engaging a new receptor. GI adverse events stack additively. Incremental weight loss is not supported by any trial. This is the least defensible of the three.

Honest ordering for tolerable GI at 8 mg reta: titrate to 10 and then 12 mg first, because that path has data. Consider cagrilintide only if 12 mg is well tolerated and still short of goal. Do not add tirzepatide.

Key takeaways

• Petersen 2026 is a 13 to 14 day rat study, not a human trial. It showed cagrilintide plus retatrutide at 5 nmol/kg beat CagriSema by roughly 4.5 percentage points and tirzepatide combinations by roughly 7, at matched molar doses.
• Pair-feeding controls show roughly 40% of the combination's weight loss is drug-specific and 60% is from reduced food intake.
• The five-receptor argument (GLP-1R, GIPR, GCGR, AMLNR, CALCR) is a hypothesis for a future single-molecule agonist, not a claim about the current pair in humans.
• Human data for the pair does not exist. Retatrutide alone reaches 28.3% at 80 weeks in TRIUMPH-1 topline. CagriSema alone reaches 20.4% at 68 weeks in REDEFINE 1.
• Cagrilintide monotherapy had the worst fat-to-lean split of the three active arms in the REDEFINE 1 DXA substudy. The muscle-sparing argument for amylin agonism does not survive human data.
• Grey-market pre-blend vials at a 5:1 mass ratio solve a vendor logistics problem, not a pharmacology one. Run the stack as two vials if you must run it.
• If you are on TRT, expect estradiol to fall, SHBG to rise, and free testosterone to drift down as you lose weight. Adjust accordingly.
• Rat mg/kg is not a dose. Roughly four in five metabolic drugs that clear Phase 1 still fail.

---

References

1. Cohen, P. G. (2001). Aromatase, adiposity, aging and disease. Medical Hypotheses, 56(6), 702-708. PubMed
2. Coskun, T., Wu, Q., Schloot, N. C., et al. (2025). Effects of retatrutide on body composition in people with type 2 diabetes: a substudy of a phase 2 trial. Lancet Diabetes & Endocrinology. PubMed
3. Davies, M. J., Bajaj, H. S., Broholm, C., et al. (2025). Cagrilintide-semaglutide in adults with overweight or obesity and type 2 diabetes. New England Journal of Medicine, 393(7), 648-659. PubMed
4. Garvey, W. T., Blüher, M., Osorto Contreras, C. K., et al. (2025). Coadministered cagrilintide and semaglutide in adults with overweight or obesity (REDEFINE 1). New England Journal of Medicine, 393(7), 635-647. PubMed
5. Jastreboff, A. M., Kaplan, L. M., Frías, J. P., et al. (2023). Triple-hormone-receptor agonist retatrutide for obesity: A phase 2 trial. New England Journal of Medicine, 389(6), 514-526. PubMed
6. Look, M., Dunn, J. P., Kushner, R. F., et al. (2025). Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study. Diabetes, Obesity and Metabolism, 27(5), 2720-2729. PubMed
7. Nair, A. B., & Jacob, S. (2016). A simple practice guide for dose conversion between animals and human. Journal of Basic and Clinical Pharmacy, 7(2), 27-31. PubMed
8. Newsome, P. N., Buchholtz, K., Cusi, K., et al. (2021). A placebo-controlled trial of subcutaneous semaglutide in nonalcoholic steatohepatitis. New England Journal of Medicine, 384(12), 1113-1124. PubMed
9. Niskanen, L., Laaksonen, D. E., Punnonen, K., et al. (2004). Changes in SHBG and testosterone during weight loss in abdominally obese men. Diabetes, Obesity and Metabolism, 6(3), 208-215. PubMed
10. Petersen, J., Merrild, C., Holm, S. K., Kurgan, N., Hassan, S., Mathiesen, C. V., et al. (2026). Cagrilintide and retatrutide combination therapy enhances weight loss and metabolic outcomes in obese male rats. Nature Metabolism. DOI: 10.1038/s42255-026-01603-y
11. Ratner, R. E., Dickey, R., Fineman, M., et al. (2004). Amylin replacement with pramlintide as an adjunct to insulin therapy improves long-term glycaemic and weight control in Type 1 diabetes mellitus. Diabetic Medicine, 21(11), 1204-1212. PubMed (pivotal trial behind the FDA boxed warning for pramlintide plus insulin)
12. Rosenstock, J., Frías, J., Jastreboff, A. M., et al. (2023). Retatrutide, a GIP, GLP-1 and glucagon receptor agonist, for people with type 2 diabetes: A phase 2 trial. Lancet, 402(10401), 529-544. PubMed
13. Sanyal, A. J., Kaplan, L. M., Frías, J. P., et al. (2024). Triple hormone receptor agonist retatrutide for metabolic dysfunction-associated steatotic liver disease. Nature Medicine, 30(7), 2037-2048. PubMed
14. Steinberg, W. M., Buse, J. B., Ghorbani, M. L. M., et al. (2017). Amylase, lipase, and acute pancreatitis in people with type 2 diabetes treated with liraglutide (LEADER). Diabetes Care, 40(7), 966-972. PubMed
15. Verma, S., Bhatta, M., Davies, M., et al. (2022). Effects of once-weekly semaglutide 2.4 mg on C-reactive protein in adults with overweight or obesity. EClinicalMedicine, 55, 101737. PubMed
16. Wen, J., Nadora, D., Bernstein, E., et al. (2025). Evaluating the rates of pancreatitis and pancreatic cancer among GLP-1 receptor agonists. Endocrinology, Diabetes and Metabolism, 8(5), e70113. PubMed
17. Wong, C. H., Siah, K. W., & Lo, A. W. (2019). Estimation of clinical trial success rates and related parameters. Biostatistics, 20(2), 273-286. DOI
18. Yamauchi, T., Becker, N. P., Hagemann, C. A., et al. (2026). Co-administered cagrilintide and semaglutide versus semaglutide alone in Japan and Taiwan (REDEFINE 5). Lancet Diabetes & Endocrinology, 14(6), 450-462. PubMed
19. Zhang, Y., Zhang, C., & Gong, X. (2026). Effect of GLP-1 receptor agonists on heart rate in non-diabetic individuals with overweight or obesity: A systematic review and network meta-analysis. European Journal of Medical Research, 31(1), 108. PubMed