E190｜Semaglutide’s New Rivals: Big Pharma’s Battle for the Second Half of the Weight-Loss Drug Market
Summary
Weight-loss drugs have moved from Novo Nordisk’s single-target lead into a scramble where nearly every company now has weapons from everyone else’s arsenal. Semaglutide pushed one-year weight loss to roughly 15%, above what diet and exercise alone can typically achieve; Lilly’s GLP-1/GIP dual agonist tirzepatide took that to roughly 20%–22%, with slightly better gastrointestinal tolerability. New patients are therefore tilting toward Lilly, but payer agreements, brand recognition and the cost of retitration mean the installed base will migrate only gradually.
Amylin is the most closely watched new mechanism not because it has already been proven muscle-sparing, but because it could become a second backbone alongside GLP-1. Unlike GIP, whose single-agent efficacy is weak and whose value mainly comes from combination synergy, early animal and phase 1/2 human data suggest Amylin monotherapy could also deliver more than 15% weight loss, offering another path for people who do not respond to or cannot tolerate GLP-1. Muscle preservation and lower rates of nausea, vomiting and diarrhea remain only potential benefits. 郭霆 uses an aircraft carrier as the metaphor: if the first weapon fails to take out the target, having a second one ready has enormous strategic value.
Novo Nordisk’s CagriSema showed that a stronger mechanism does not automatically create a generational gap in phase 3. The company had communicated weight loss of up to 25%, but REDEFINE 1 read out at roughly 22%, broadly in line with tirzepatide, sending the stock down 20%–30% at one point premarket; REDEFINE 2 also failed to open a significant lead. Only 57.3% of patients remained on the highest dose at the end of the trial, and the flexible-dose design clouded the result, prompting Novo to rerun the program in REDEFINE 11 with tighter dose management.
The trading frenzy around Amylin reflects a pipeline land grab, not a result already decided. Roche and Zealand struck a deal worth up to $5.3B, including a $1.65B upfront payment—the highest upfront payment in a multinational pharma deal in 2025, according to the show; AbbVie then acquired a phase 1 molecule from Gubra. Roche is pursuing a lower-risk commercial strategy of acquiring a batch of competitive molecules, while Zealand had only completed phase 1 and was moving into phase 2; a meaningful comparison still requires full phase 3 data.
The next generation may not belong to Amylin: GLP-1, GIP, glucagon and even triple agonists are still competing over the balance between efficacy and organ-specific fat loss. Lilly’s retatrutide delivered 8.7%–24.2% weight loss at 48 weeks in phase 2, but showed mild heart-rate-related signals; BI is focusing on fatty liver, while the show expected Innovent’s mazdutide to become the first GLP-1/glucagon combination approved for weight loss. Novo also bought a triple-target molecule from 联邦制药, underscoring that pharma companies are still running molecular combinations; it will be difficult to declare a winner before phase 3 results are in.
Oral drugs could be the equivalent of moving from the PC internet to the mobile internet, making market expansion more important than simply chasing the weight-loss numbers of injectables. They bypass needle aversion, cold-chain requirements, clinician monitoring and complex dose titration, opening access in countries with inadequate infrastructure and among less-severe patients. The show expected Lilly’s oral obesity drug to deliver phase 3 results in Q2 that year, with the first product potentially approved in the second half of the following year. The injectable market is expected to reach $100B–$150B by 2030–2035, with oral drugs adding another roughly $50B.
Restored manufacturing capacity removes a growth bottleneck, but also takes away Novo and Lilly’s most convenient explanation for demand. Both companies say all formulations are now adequately supplied, which could push the FDA to clear out compounding pharmacies while forcing investors to examine true uptake. Tirzepatide’s slightly better efficacy and tolerability, combined with Lilly’s more diversified pipeline, have earned it a higher valuation premium than Novo. “Once the capacity problem was solved, they had effectively set a small trap for themselves.”
Generics, compounding pharmacies and patent expiries will turn weight-loss drugs from a high-margin duopoly into a layered market where price and function compete in parallel. Compounding pharmacies, using the shortage exemption, were once estimated to hold 5%–20% of the U.S. weight-loss drug market, with Chinese peptide companies serving as important API suppliers; originators are fighting back with direct sales of branded drugs at roughly $400–$500 per month. China could see semaglutide generics first in 2–3 years, Europe around 2030 and the U.S. in 2033–2034; low-cost drugs will absorb basic weight-loss demand, while new products will have to target muscle preservation, fatty liver, kidney disease and other niches.
Deep dive
1. Semaglutide’s real breakthrough was crossing the efficacy threshold, not merely reducing injection frequency
郭霆 first separates two ways of defining a “generation”: GLP-1s evolved from short-acting products requiring one or multiple injections a day, to once-weekly diabetes drugs, and then to long-acting, highly potent versions; in the weight-loss market, semaglutide is considered first generation and tirzepatide second generation.
Native GLP-1 has a half-life so short that it is difficult to turn into a drug “unless you keep pumping it in nonstop.” Pharma companies had to modify the amino acids and molecular structure to slow metabolism and clearance before they could extend dosing to once a week.
Liraglutide’s roughly 5% weight loss beat the 1%–2% delivered by earlier products and was also the first formally FDA-approved weight-loss drug, in 郭霆’s account. But for someone weighing 100 kg, losing 5 kg in a year could still be surpassed by diet and exercise. Semaglutide pushed the figure to roughly 15%, the point at which it “really became a drug.”
郭霆 compares the progression with ChatGPT: 1.0 and 2.0 had the technical outlines, while 3.0 became a phenomenon. Novo Nordisk’s lead was driven by years of R&D, but also by chance—the company initially did not know that intensifying glucose lowering would trigger a simultaneous leap in weight loss.
2. GLP-1 acts like a satiety instruction; GIP pushes single-target efficacy one step further
Understanding of where GLP-1 acts has also shifted. The focus used to be on metabolic organs such as the liver and gut; increasingly, explanations point to the brain, where it tells the body: “You’re full. You don’t need to eat anymore.”
郭霆 stresses that these gut-pancreatic hormones are not switches whose behavior can be fully derived from first principles. They are more like neural networks: often, researchers have to administer them to animals, and even to humans, before they know how appetite, gastric emptying, blood sugar and adverse events will interact.
Lilly did not simply copy semaglutide. It put GLP-1 and GIP into the same dual-target molecule, tirzepatide, sold under brand names including Zepbound and Mounjaro. Its roughly 20%–22% one-year weight loss represents a clear but not overwhelming advantage over semaglutide’s roughly 15%.
The combination logic is asymmetric: GIP has no particularly strong weight-loss effect on its own, but it can increase GLP-1 efficacy while making nausea and vomiting “a little less bad.” 郭霆’s candid answer is: “Lilly doesn’t know either—nobody does.” Many mechanistic explanations remain post-hoc attribution.
3. Better tirzepatide will capture incremental patients, but will not erase semaglutide overnight
泓君’s question is direct: if doctors know tirzepatide works better and has milder side effects, why do they still prescribe semaglutide? 郭霆 splits the market into the installed base and new demand. Switching the former means retitration and restarting the insurance process, so patients generally do not switch mid-course.
New-patient migration is also constrained by insurers’ manufacturer-share agreements, physician habits and brand recognition. Doctors will try a product on 1 or 2 patients and expand gradually, rather than replace every prescription overnight after seeing a single set of clinical results.
Consumers have their own inertia. Both companies advertise on U.S. television, semaglutide has been on the market longer and is better known, and some patients will simply say, “I want this semaglutide.” Lilly is taking share, but the process is gradual.
4. Amylin is attracting attention because it could become a second backbone therapy
Amylin is secreted mainly by the pancreatic islets, circulates throughout the body and engages in complex cross-talk with the brain, liver, gut and other organs. 郭霆 compares it to giving an AI a prompt and then confronting a black box: “It’s hard to know what it is actually doing,” so researchers can only infer the mechanism from experimental results.
The most important signal in the early data is not muscle preservation, but that Amylin monotherapy may deliver roughly 15% or more weight loss, close to semaglutide. That could create an independent market among people with weak average responses, no response at all or an inability to tolerate GLP-1.
Muscle preservation still needs to be viewed cautiously. Animal studies show potential, but the signal in phase 1 and phase 2 human data is “not particularly clear.” Gastrointestinal tolerability also appears only slightly better and does not eliminate nausea, vomiting or diarrhea. Yushan’s warning about the field’s ongoing evolution and uncertainty is worth keeping: this is not yet an established product selling point.
郭霆 explains the excitement through pharma’s concept of a backbone. Semaglutide is strong enough as a monotherapy to be paired with mechanisms targeting fatty liver and other indications; if Amylin can also stand on its own, it competes from another dimension rather than merely adding a marginal benefit. GIP and glucagon have not yet shown comparable evidence as monotherapies.
5. CagriSema’s phase 3 shortfall weakened Novo Nordisk’s generational-upgrade thesis
Novo combined semaglutide and an Amylin drug in a 1:1 ratio to create CagriSema. In phase 2 among patients with type 2 diabetes, semaglutide produced roughly 5% weight loss after 4–6 months, while the combination delivered more than 10%. That doubling of efficacy led the market to expect it to overtake tirzepatide.
But REDEFINE 1 read out at roughly 22%, falling short of Novo’s repeatedly communicated 25% target and landing broadly in line with tirzepatide; the stock fell 20%–30% at one point premarket that day. The subsequent diabetes study, REDEFINE 2, also failed to open a significant lead, leaving Novo’s shares under pressure.
The tolerability data cited by 泓君 were even more striking: only 57.3% of patients were on the highest dose at the end of the trial, versus 70.2% for semaglutide and 82.5% in the earlier monotherapy group. 郭霆’s explanation is that Novo allowed flexible dose reductions; some patients had reached their target weight and were not unable to tolerate the drug—they simply “didn’t want to keep enduring it.”
The design could not restore market expectations, so Novo plans to launch REDEFINE 11 with stricter dose management and take on tirzepatide in a direct comparison. Based on the show’s assessment at the time, CagriSema could still be approved relatively soon, but putting 2 drugs into 1 injection means convenience and potential safety are not settled questions.
6. Roche and AbbVie are buying Amylin for the size of the market, not an exclusive mechanism
On the day the show was first recorded, Roche and Zealand signed an exclusive collaboration worth up to $5.3B, including a $1.65B upfront payment. 泓君 called it the largest upfront payment in a multinational pharma deal in 2025. The core objective was to obtain an Amylin-based weight-loss pipeline.
郭霆 sees relatively limited molecular differentiation in Roche’s assets. The strategy looks more like a low-risk effort to accumulate competitive assets after confirming that the weight-loss market is large, rather than an attempt to create a mechanism no one else has. Zealand had only completed phase 1 and was still advancing phase 2, so its assets could not be compared directly with mature phase 3 programs.
AbbVie subsequently acquired a phase 1 Amylin molecule from the European company Gubra, whose early monotherapy weight-loss data were also “quite impressive.” The transaction and Gubra’s data strengthened the signal that Amylin monotherapy might stand on its own, while Roche’s deal with Zealand is betting on a GLP-1/Amylin combination. Long-term safety, muscle preservation and phase 3 reproducibility remain unanswered.
Novo still holds the next-generation Amycretin, a single molecule targeting both GLP-1 and Amylin, with injectable and oral versions whose early data were described as “also performing very well.” Another path is a small-molecule oral Amylin drug: the program designed in Shanghai by Structure Therapeutics (硕迪生物) was still preclinical at the time, with human trials planned within the year.
7. Glucagon and triple agonists could still rewrite the “Amylin successor” narrative
郭霆 expands the battlefield to the entire gut-pancreatic hormone library: GLP-1, GIP, Amylin, glucagon and PYY may affect appetite, gastrointestinal motility, blood sugar and different organs. Pharma companies can keep “mixing and matching,” but cannot select the winning combination from mechanism alone.
Glucagon is not a standout weight-loss drug on its own, but in combination with GLP-1 it may intensify fat loss in the liver, kidneys, heart and other organs. BI is therefore focusing its combination on fatty liver. Innovent’s mazdutide showed standout weight loss in China, and the show expected it to launch around midyear as China’s first homegrown weight-loss drug.
Lilly’s GLP-1/GIP/glucagon triple agonist retatrutide delivered 8.7%–24.2% weight loss at 48 weeks in phase 2, but also produced mild heart-rate and cardiac signals. Whether the incremental efficacy is worth the new risks must be answered by phase 3, patient by patient and indication by indication.
8. Cross-trial phase 3 comparisons are tempting—and most likely to manufacture false certainty
郭霆 cautions that starting weights in obesity trials may be 80, 100 or 120 kg, while the country’s diet, the male-female mix, diabetes status and trial duration may also differ. Putting retatrutide’s 48-week phase 2 results directly beside semaglutide’s 68-week phase 3 results requires “an enormous amount of effort.”
At the end of March, Novo also bought from 联邦制药 a GLP-1/GIP/glucagon molecule that had completed phase 1 in China. 联邦制药 did not publish the specific phase 1 data, so 郭霆 could only infer that, after seeing Lilly’s phase 2 results, Novo still made the deal because it believed the molecule was “at least no worse than Lilly’s.”
The first-generation market has become a two-horse race between Novo and Lilly. The next generation has at least 4 or 5 major pharma companies—Roche, Pfizer, AstraZeneca, AbbVie and others—entering aggressively. 郭霆’s view is not “a three-way battle,” but that once you reach the battlefield, “you realize this is a free-for-all.”
9. Oral drugs are not substitutes for injectables; they are a new market entry point
郭霆’s analogy is that “oral versus injectable is a little like moving from the PC internet to the mobile internet.” Many people do not want injections, and many countries lack cold chains, physician monitoring and dose-titration capacity. Oral drugs can therefore reach less-severe patients and regions that previously could not support injectable treatment.
Injectable drugs require patients to climb through doses step by step, with dose or product changes based on vomiting and other reactions. A shortage of any low-dose SKU can prevent a new patient from starting. Oral formulations reduce healthcare-system friction first; they do not necessarily deliver stronger efficacy.
Lilly is ahead. Its oral molecule originally came from Japan’s Chugai, with Lilly acquiring rights outside Japan. The show expected its phase 3 results in Q2 that year, with the first oral weight-loss drug potentially approved in the second half of the following year.
Roche acquired Carmot to obtain an oral GLP-1 and an injectable GLP-1, then added Amylin through Zealand. The market expectation cited by 郭霆 was roughly $100B–$150B for injectables by 2030–2035, with oral drugs representing another roughly $50B.
10. The difficulty of investing in weight-loss drugs is that many mechanisms can only be judged when the cards are turned over
郭霆 sorts drugs along a certainty gradient. Oncogenic driver mutations are clear switches: turn them off and the tumor stops growing, making the outcome easier to reason about. Other drugs have only preclinical mechanistic support, while metabolic drugs contain a large amount of “alchemy” whose workings are nearly impossible to assess before phase 2 or phase 3.
His investment method is not to pretend there is a standard answer, but to combine mechanism, clinical data, commercial prospects, competition and valuation while acknowledging that he “also gets things wrong quite often.” That is why even beautiful early Amylin data cannot skip complete phase 3 testing.
11. Restored capacity shifts the growth narrative from supply constraints to real demand
Over the past 2 years, semaglutide and tirzepatide often could not scale because particular doses were in short supply. The lowest dose is especially important: obesity does not require immediate treatment, so patients may prefer to wait rather than skip titration and start directly at a high dose.
The contradiction is that, during a shortage, slow growth can be attributed to excess demand over supply, allowing the market to maintain high expectations. Now that both companies say every formulation is adequately supplied, analysts can instead ask: “If growth is not fast enough, does that mean demand is not as large as everyone thought?”
Tirzepatide is still growing faster than semaglutide, both exposing unmet demand and potentially taking share. Lilly’s more diversified pipeline has also encouraged investors to pay a higher premium. Novo is simultaneously absorbing setbacks in programs such as CagriSema.
At the time of recording, pharmaceutical products had not yet been affected by Trump’s tariff war. Public information suggested that a substantial portion of Lilly’s tirzepatide API might come from China, while 郭霆 described Novo as having no Chinese supply-chain component. Even if tariffs are imposed later, high margins and public pressure over patients losing access make a complete supply stoppage unlikely.
12. “Normal-weight people lose weight too” does not mean the benefits have exceeded the risks
郭霆 gives the approved-use boundary as a BMI of 30 or higher, or 27 or higher with an obesity-related disease. For these patients, the FDA considers the benefits to outweigh the risks. For people without a relevant disease, weight loss may still occur, but current evidence is insufficient to treat the drugs as consumer products.
The importance of these drugs is that they break through the ceiling imposed by self-discipline: exercise and diet alone rarely deliver 20% weight loss within a year, while drugs can. But normal-weight people still bear small-probability risks such as gastrointestinal reactions and bowel obstruction, and weight can rebound after discontinuation.
Improper use magnifies the risk. 郭霆 has seen people who were already in good shape inject a high dose without low-dose titration in pursuit of rapid weight loss within 1 month, resulting in severe vomiting and diarrhea and even hospitalization. “Can normal-weight people use it?” cannot be answered by looking only at whether they can lose weight.
13. GLP-1’s incremental indications may be larger than weight loss alone—and require more restraint
Obesity is increasingly being redefined from a health risk into a disease because it is linked to diabetes, cardiovascular disease, sleep apnea, kidney disease and fatty liver. Lilly and Novo are using independent clinical trials to prove the core drugs’ effects in these conditions, and some results have already materially changed market perceptions.
A more ambitious theory holds that obesity causes chronic inflammation and that GLP-1 may help normalize the immune system by improving metabolism. Mice that eat 20% less and become healthier and longer-lived are then used to speculate about the link between caloric restriction and longevity. 郭霆 explicitly calls these ideas “opening up the imagination,” not medical advice.
Alzheimer’s disease has produced more specific but still limited signals. After aggregating data from thousands to tens of thousands of people, Novo observed that the risk might be cut roughly in half, but the actual number of new cases was only a few dozen. Long-term Danish medical-record data also showed that, for patients with type 2 diabetes, each year of GLP-1 use was associated with roughly a 10% reduction in risk.
These observations prompted Novo to launch phase 3 trials of semaglutide in Alzheimer’s disease, with the show expecting results within a year. If 1 weekly injection truly reduced risk substantially, it would open a vast market; until low event counts and observational bias are ruled out, it cannot be presented as established efficacy.
14. China is both a source of innovative pipelines and the center of the generic and compounding-pharmacy supply chain
郭霆 estimates that Chinese and Asian companies account for “at least half” of early-stage pipelines. Pfizer’s second-generation oral molecule came from a Japanese company; Merck bought a preclinical oral small molecule from 豪森 for an upfront payment of more than $100M; and 恒瑞’s tirzepatide-like molecule, after strong phase 2 data in China, was packaged by a group of U.S. venture investors for development in the United States.
U.S. compounding pharmacies used the shortage exemption to buy active ingredients from the supply chain and formulate their own injections. Market analysts once estimated that they held 5%–20% of the U.S. weight-loss drug market, while Hims & Hers spent heavily on advertising. After the FDA removed the drugs from the shortage list, the regulatory and political battle between pharmacies and originators remained unresolved.
郭霆 says the top 5–10 Chinese peptide companies supplied the vast majority of the semaglutide and tirzepatide APIs required by U.S. compounding pharmacies, including 药明康德, 蓝帆 and 翰宇. His estimate is that 1 kg of API may cost only several thousand RMB, while the end customer pays a few hundred dollars per month.
The risk is not that the formula is completely mysterious, but that generics lack equivalent evaluations, audits and end-to-end traceability. Roughly 20% of U.S. doctors have introduced patients to compounding pharmacies, often because insurance is inadequate and branded drugs are too expensive. If the raw materials, temperature control and operations are reliable, the felt efficacy may be similar, but there is no guarantee that every batch is problem-free.
15. Patent expiry will push down the price of basic weight loss and force new drugs into functional niches
Formal generics must use analytical, preclinical and human pharmacokinetic data to demonstrate equivalence with the originator; compounding pharmacies are only an emergency channel during shortages. Semaglutide contains more than 30 amino acids and requires fermentation and chemical modification, so patent protection covers not only the active ingredient but also the complex manufacturing process.
The show’s assessment was that Chinese patents could expire first in 2–3 years, Europe around 2030 and the United States around 2033–2034. But a single product may be covered by multiple patent families whose expiry dates are separated by as much as 10 years, with the final date depending on patent litigation between originators and generic manufacturers.
Branded semaglutide in China cost roughly RMB1,000 per month at the time, while generics could bring that down to the low hundreds or even RMB100–200. Branded drugs sold directly in the United States cost roughly $400–$500 per month, and even with insurance, out-of-pocket costs could still run into the low hundreds. Price matters especially because Wegovy is used for an average of only 6–8 months, while diabetes-treated semaglutide can continue for more than 4 years.
Efficacy is already approaching the 25%–30% weight-loss ceiling associated with bariatric surgery, making it increasingly difficult to keep competing for the title of strongest drug on earth. 郭霆’s view of the next 5–10 years is that semaglutide and its generics will absorb basic demand, while next-generation products move into oral dosing, muscle preservation, fatty liver, kidney disease, respiratory disorders and other functional niches. “The second half may be even more exciting than the first.”