When cheaper GLP-1 drugs arrive, country by country

Semaglutide's 2006 Patent Starts a Twenty-Year Clock

On March 20, 2006, a team of researchers at a Danish pharmaceutical company filed a patent application. The paperwork described a new molecule — a 31-amino-acid peptide that mimicked a hormone your gut naturally produces after you eat. Nobody outside the lab cared. Patent filings happen every day. This one started a clock.

Twenty years later, that clock ran out. On March 20, 2026, the core compound patent on semaglutide — the active ingredient in Ozempic, Wegovy, and Rybelsus — expired in every market that never granted an extension. What happened next depends entirely on which country you were standing in.

Semaglutide is the world's second-best-selling prescription medicine. Its three brand names pulled in $26 billion in combined global sales in 2024, and the number was still climbing. Ozempic is the weekly injection for type 2 diabetes. Wegovy is the weekly injection for obesity. Rybelsus is the daily pill for type 2 diabetes. They are the same molecule doing the same job at different doses, sold under different names, for reasons that have everything to do with pricing and nothing at all to do with chemistry.

The global GLP-1 drug market was valued at $64 billion in 2025 and is projected to hit $212 billion by 2035. Novo Nordisk, the Danish company that owns semaglutide, posted total semaglutide sales of DKK 228 billion in fiscal year 2025 — across Ozempic, Wegovy, and Rybelsus combined. Ozempic alone generated DKK 127 billion of that. These are numbers that do not suggest a company with anything to worry about, which is exactly what makes the geography of what happened next so brutal.

On March 20, 2026, the world split in three. Countries where no patent extension was granted lost exclusivity that day. Countries where extensions were granted kept the door shut — for years, sometimes a decade. And in both cases, the molecule in question was identical. Only the paperwork differed.

The countries that got hit first include India, Brazil, Canada, South Africa, and Saudi Arabia. In India, over 40 generic semaglutide brands hit pharmacy shelves within days of March 20. Saudi Arabia's first generic launched on September 16, 2026. South Africa got its first generic in August 2026, with 13 more applications already queued behind it. Canada became the first G7 nation to approve a generic version, with Health Canada clearing Dr. Reddy's product on April 28, 2026. Dr. Reddy's had already announced plans to launch in 87 countries during 2026, starting with Canada, India, Brazil, Turkey, and other emerging markets. The first wave landed.

The second wave is scheduled for around 2031. That is when the European Union, the United Kingdom, and Japan lose their patent protection — those markets granted supplementary protection certificates that extended the core patent's life well past the 2026 baseline.

The third wave is the United States, where the relevant compound patent listed in the FDA's Orange Book expires December 5, 2031, with additional exclusivities pushing effective generic market entry to 2032.

Generic in Mumbai, Still Protected in Manhattan

The result is a paradox Novo Nordisk is navigating in real time: it faces the most consequential patent cliff in pharmaceutical history, and yet its most lucrative markets are sealed for another five to six years. The drug is going generic in Mumbai while the patents hold in Manhattan. The same Tuesday in March 2026 was simultaneously a catastrophe and a non-event, depending on your longitude.

That single filing on March 20, 2006 — PCT application WO2006/097537 — produced one compound, one formal expiry date, and a dozen wildly different outcomes, each determined by which country's patent office was asked and what that country's rules say about extensions. The molecule is identical everywhere. The legal calendar is not.

This is a story about what happens when a $26 billion drug meets its expiry date — and why that date means something completely different depending on your zip code, your postal code, or your PIN code, and frankly whether your government's patent office had a particularly generous afternoon back in 2006.

Canada's story begins with a form that someone forgot to fill out.

Ozempic was Canada's best-selling drug in 2025 — not best-selling diabetes drug, not best-selling injectable — best-selling drug, full stop, at CAD $2.9 billion in annual sales, more than three times the next drug on the list. Over one million Canadians were on it at the time. So whatever happened next in Canada mattered enormously. It also happened because of a clerical error so small it is almost physically painful to contemplate.

Canadian patent CA 2,601,784 covers the compound semaglutide. Novo Nordisk filed it on March 20, 2006. If all maintenance fees had been paid, it would have run until March 20, 2026 — the standard 20-year term. On top of that, the company held a Certificate of Supplementary Protection linked to that patent, which is Canada's mechanism for extending effective patent protection, and that would have pushed the deadline to March 20, 2028. Together, these two instruments were the entire legal foundation for Ozempic in Canada.

In 2019, Novo Nordisk missed a maintenance fee payment. The base fee was CAD $250. With the late charge added, the total bill would have been CAD $450. Nobody paid it. The last payment on record arrived on October 23, 2018. After that, nothing.

Here is where the bureaucratic calendar becomes relevant. Before October 30, 2019, Canada's Intellectual Property Office was not required to notify patent holders when a maintenance fee went unpaid. You were simply expected to know. The Patent Law Treaty amendments that took effect on October 30, 2019 changed this: from that date forward, CIPO was required to issue notice before rights were extinguished, giving patentees a window to remedy the situation. Novo Nordisk's missed payment fell in the gap — the fee was due in 2019, the notice requirement did not yet exist, and by the time anyone might have reviewed the file, the window to reverse the lapse had already shut. The deadline to undo the damage was August 31, 2020. Nobody showed up.

Canada's Six-Year Silent Patent Lapse

When the patent lapsed, the Certificate of Supplementary Protection built on top of it became void. The extended protection to 2028 vanished along with the patent it depended on. Six years of additional Canadian exclusivity, gone.

The lapse sat quietly in CIPO's records until June 2025, when Science magazine columnist Derek Lowe reported on it. At that point, the patent had been dead for six years, and Canada's data exclusivity period for semaglutide — a separate regulatory barrier under Canada's Food and Drug Regulations — was still running but expiring fast. It lapsed on January 4, 2026.

Novo Nordisk's public response was pointed. A spokesperson told Fortune magazine that "all intellectual property decisions are carefully considered at a global level." Companies do not say that about accidents. Industry observers took the statement as an invitation to speculate. One theory: the lapse was a calculated pivot toward secondary patents covering injection devices and dosing regimens, which run to 2033. Another: it reflected a strategic shift in focus toward CagriSema, Novo Nordisk's next-generation combination therapy. Neither theory explains why the company would have voluntarily surrendered Canadian protection three years before the core patent's natural expiry, in what became its second-largest global semaglutide market. The simpler reading is that a renewal was missed, the window closed before anyone noticed, and the official line afterward was aggressive reframing. Novo Nordisk has not confirmed this.

The outcome was clean and fast. No semaglutide patents appear on the Canadian Patent Register, which meant the regulatory linkage regime — the mechanism that would have triggered a 24-month automatic stay upon a generic challenger filing — was simply unavailable. There was no procedural tripwire left. Health Canada approved Dr. Reddy's Laboratories' generic semaglutide injection on April 28, 2026, making Canada the first G7 country to authorize a generic version of Ozempic. Three days later, on May 1, 2026, Apotex received the second approval. On June 29, 2026, Health Canada authorized Sevmia — Apotex's generic version of Wegovy, indicated for chronic weight management in patients aged twelve and over. As of that date, six additional generic semaglutide submissions were still under review, with Sandoz projecting launch pricing up to 70 percent below Ozempic's branded list price.

The drug that was Canada's best-seller by a factor of three met generic competition because of a $250 administrative fee and six years of nobody checking the file.

Canada's story had a villain you could put in a spreadsheet. India's story was just chaos, and mostly the good kind.

The Indian compound patent on semaglutide, IN 262697, expired on March 20, 2026. Novo Nordisk would have known this was coming. The entire Indian pharmaceutical industry knew it was coming. What happened next was less a market transition and more a starting pistol fired into a crowd of people who had already been sprinting for two years.

Twenty-Six Indian Generics Within Weeks of Expiry

By the morning of March 21, the day after patent expiry, more than a dozen Indian manufacturers were selling generic semaglutide. Dr. Reddy's, Sun Pharma, Zydus Lifesciences, Torrent, Alkem, Glenmark, Natco, Eris, and USV had all filed launch notices, priced packs, and briefed their field forces. By the end of March, Pharmarack was counting 26 brands from 13 companies. By May, PharmaTrac had 35 brands from 17 companies. The industry had been preparing so methodically that the actual patent expiry date was almost an afterthought. The race started; the starting line caught up to the runners.

Prices fell immediately and sharply. Before March 20, Ozempic cost between ₹8,800 and ₹11,175 per month in India, and Wegovy, which had only launched in June 2025 at prices up to ₹17,345 per month before being cut, was going for up to ₹16,400. Generics launched at ₹1,800 to ₹4,200 per month for pen devices, ₹325 to ₹440 for weekly packs, and about ₹1,290 per month for vials. That vial price is the number that clarifies everything about this market: Natco and Eris charged approximately three times more for an autoinjector pen version than for a vial containing the exact same molecule. The molecule does not know which container it is in. The patent on the container, however, very much does.

Novo Nordisk's response took ten days. On April 1, 2026, the company cut Wegovy prices in India by 48% and Ozempic by 36%. That speed tells you everything about how seriously it took the competitive threat. The pricing war Novo Nordisk had been fighting since November 2025, when it had already slashed Wegovy's starting pack from ₹16,261 to ₹10,850, accelerated into something close to a rout.

By April 2026, generics accounted for 137,000 of 170,000 injectable semaglutide units sold in India. Four in five units, in two months, against an originator that had just cut its prices by nearly half. In value terms, generics had taken Rs 44 crore of an Rs 88 crore total semaglutide market, 50% by value, against a product that Novo Nordisk had been actively discounting to compete.

None of this was a surprise, and Novo Nordisk had tried to stop it. In December 2025, the company took Dr. Reddy's to the Delhi High Court seeking an injunction. Justice Manmeet Pritam Singh Arora declined. The court found Dr. Reddy's had established a credible prima facie challenge against patent IN'697, and the interim injunction was refused. What the court did provide was a temporary compromise: Dr. Reddy's could manufacture semaglutide for export to markets where the patent had already lapsed, but agreed on record not to sell domestically in India until March 2026. On March 9, 2026, eleven days before the patent expired, a division bench of the Delhi High Court dismissed Novo Nordisk's appeal of that ruling. The runway was clear.

The tablet side was more complicated. A separate court order directed Sun Pharmaceutical Industries to seek prior court approval before commercially launching generic semaglutide tablets. Sun Pharma had received subject expert committee approval for tablet strengths of 3 mg, 7 mg, and 14 mg, and then found itself arguing jurisdiction, with Sun Pharma contending it was a Maharashtra company with no relevant Delhi operations, and Novo Nordisk countering that the company sells products in Delhi and would sell semaglutide there too. The tablets stayed in legal limbo while the injectable market went generic around them.

Ozempic's Four-Month Branded Window in India

Ozempic had launched in India on December 12, 2025, priced at ₹8,800 per month at its lowest strength. That gave the branded product less than four months on Indian shelves before it was competing against 26 versions of itself at half the price or less. By October 2026, 12 generic manufacturers were selling 20 semaglutide brands in India, and the lowest available monthly price for vials had settled at ₹1,290. The brand that had taken twenty years and billions of dollars to develop had three months of Indian exclusivity. Novo Nordisk's patent filing clock ran from 2006 to 2026 in every country simultaneously. In India, it just happened to matter on day one.

Brazil runs on a regulatory system with its own logic, and that logic turned out to be very convenient for anyone who wanted cheaper semaglutide. ANVISA — Brazil's drug regulator, whose name sounds like a cleaning product but functions as the gatekeeper for everything pharmaceutical in Latin America's largest market — refused Novo Nordisk's application for a patent term extension on semaglutide. No extension means the core patent expired March 20, 2026, on schedule, same as India. The first generic approval followed in May 2026. That two-month gap between patent death and first generic clearance is the part that should get your attention.

Brazil's historical average for the gap between patent expiration and first generic approval runs two to four years. Regulators have forms to process, studies to review, and somewhere to be. The semaglutide approval in May 2026 demolished that average completely. Brazilian regulators had apparently decided that access to GLP-1 drugs was not a situation requiring the usual pace of deliberation.

The legal mechanism that made this possible is ANVISA's patent linkage system, which blocks generic approval only if a valid patent exists in Brazil's official database. Once ANVISA declined to extend semaglutide's patent term, there was no valid patent entry to block anything. The door was not merely unlocked — it had been removed from its hinges by a regulatory agency that made a deliberate policy choice. That choice was this: access matters more than originator protection here. Brazil looked at semaglutide's pricing, looked at its obesity burden, and made a decision. The paperwork followed.

Brazilian law requires that generics be priced at least 35 percent below the reference product's price. That is the floor. When multiple manufacturers enter, historical price reductions in Brazil reach 50 to 80 percent below originator levels. The 35 percent is the legal minimum that stops mattering the moment competition arrives. Novo Nordisk's lawyers presumably understand this. The 35 percent figure exists to protect patients from lazy monopoly pricing on day one; the competitive dynamics handle everything after that.

As of September 2026, Brazil's approved generics covered diabetes pen formats only. Obesity-indication generics remained in the queue. The distinction matters for patients because the weight-management indication commands a higher price and reaches a different, larger population of potential users. Brazil has classified semaglutide as a synthetic analogue of a biologic peptide, which triggers a higher analytical bar than standard small-molecule testing: manufacturers must document impurity profiles, aggregates, and immunogenicity data. That technical bar partly explains why the obesity-indication approvals are lagging. You are testing the same molecule. Brazil just wants more paperwork about exactly which version of the same molecule has arrived at the door.

Biocon and Biomm Target Brazil's Off-Patent Market

Biocon Limited, the Indian biopharmaceutical company, partnered with Brazilian firm Biomm to file and commercialize an off-patent semaglutide candidate in Brazil. Biocon had supply chain history with GLP-1 molecules and the Biomm partnership followed the same structure playing out across emerging markets: Indian manufacturing capacity, local regulatory relationships, and a market that had just legally opened.

Brazil also had a compounding situation that preceded all of this. Brazilian law permits compounding when no registered equivalent exists in the required dose or form. Compounded semaglutide from Brazilian and Paraguayan pharmacies was already circulating as of mid-2024, before the patent expired, before ANVISA approved anything. Compounded products do not undergo ANVISA bioequivalence testing. Patients were buying a product made by a pharmacy that had read the same scientific literature as Novo Nordisk but had not been required to prove equivalence to anyone. Whether that is an access solution or a public health gamble is a question Brazilian regulators were still working through while the approved generics arrived behind it.

The ANVISA refusal of a patent term extension is where this story gets its cleaner edge. India's patent expired by the calendar. Brazil's expired because a regulator made a choice. Both outcomes look identical from the pharmacy counter — the originator's price drops, the generics appear, the supply expands. The mechanism that got there was different. One was arithmetic; the other was policy. When the history of GLP-1 access gets written, the ANVISA refusal will sit alongside the Canadian maintenance fee in the footnotes: two different paths to the same result, neither of them the dramatic courtroom showdown the industry expected.

China has nearly 148 million adults with diabetes. If that sounds like a number that should have more commas in it, you are reading it correctly. It is roughly the population of Russia, except this particular Russia has type 2 diabetes. This is the market Novo Nordisk is defending with data exclusivity rather than patents, because the patents ran out.

The core semaglutide patent in China expired March 20, 2026, the same day as India. As of October 2026, zero generic semaglutide products had been approved in mainland China. India had 35 brands from 17 companies by May. China had none. The molecule was identical, the filing date was identical, and the outcome was not. The barrier is data exclusivity, not patent protection, and the distinction matters.

A patent protects the molecule itself. Data exclusivity protects the clinical trial data the originator submitted to prove the drug works, blocking generic companies from relying on that evidence for a set number of years even after the patent has expired. Under China's older framework, semaglutide's data protection runs to approximately April 2027, one year after the patent died. That gap — legally unprotected molecule, legally protected clinical data — matters enormously to regulators and not at all to patients who cannot afford the branded price. China's new Regulatory Data Protection framework took effect in May 2026. Under that system, innovative drugs properly filed under the new rules could have protection extended to approximately 2032. The exact endpoint is unsettled: data protection blocks generic entry at least until 2027, and under the new framework potentially much longer.

Seventeen Chinese Manufacturers Queue for Approval

None of this stopped seventeen Chinese manufacturers from queuing up anyway. At least seventeen domestic companies had progressed to Phase III clinical trials or the pre-market application stage by the time the patent expired, with nine specifically targeting obesity. As of March 20, 2026, twelve domestically produced semaglutide generics from ten companies had already submitted registration applications to the NMPA. They are waiting.

The front-runner is Hangzhou Jiuyuan Gene Engineering, owned by Huadong Medicine. Jiuyuan filed for NMPA marketing authorization in April 2024 after completing its Phase III clinical trial. The NMPA accepted the application for review in February 2026, and Jiuyuan Genetic Biopharmaceutical announced the acceptance via a Hong Kong Stock Exchange filing on February 25, 2026, for a product branded Jikeqin. Jiuyuan registered as Class 3.3, the biologic pathway. Qilu Pharmaceutical went the opposite direction, filing in approximately September 2024 as a Class 2.2 modified new chemical entity, a classification that carries lower regulatory requirements. Both strategies have their adherents: Xibeijiang Pharmaceutical joined Jiuyuan on the Class 3.3 track, while Changchun Bcht Biotechnology filed alongside Qilu on Class 2.2.

The classification debate reflects a structural question China has not resolved. The FDA settled it by rule: any peptide under forty amino acids is a generic drug rather than a biologic. Semaglutide has thirty-one amino acids and falls cleanly into the generic pathway in the United States. China's NMPA has made no equivalent declaration. Both pathways remain open. The agency has said nothing definitive about which one it prefers, which is either a deliberate regulatory strategy or a very long committee meeting. Both are consistent with what is known.

One protection Novo Nordisk cannot deploy in China is a patent term extension. China's fourth patent law amendment introduced extension provisions and took effect June 1, 2021. Ozempic was approved in China in April 2021, two months before the amendment. Extensions do not apply retrospectively. The compound patent ran twenty years, reached March 2026, and stopped there.

The strategic irony is that Chinese companies were already deeply embedded in the global semaglutide supply chain before any of this. Novo Nordisk submitted a letter to the US Department of Commerce estimating that Chinese companies had shipped enough semaglutide into the US in just over six months to manufacture 1.5 billion starter doses of Wegovy. Eight Chinese pharmaceutical manufacturers were actively supplying raw semaglutide and tirzepatide API to US compounding pharmacies in 2024. Chinese ingredient manufacturers were selling semaglutide powder at up to seven times production cost, with a monthly supply's worth of raw material priced at approximately forty-nine cents. By the second quarter of 2025, those US-bound shipments had plunged ninety percent year-on-year after regulatory crackdowns on compounding eliminated the demand. The manufacturing infrastructure does not disappear when a customer does. It redirects.

EPO Revokes Novo's Cardiovascular Outcomes Patent

Then on October 8, 2026, the EPO Board of Appeal published decision T 0187/25, revoking Novo Nordisk's semaglutide cardiovascular outcomes patent — specifically the one covering semaglutide's use in reducing major adverse cardiovascular events in patients with type 2 diabetes and high cardiovascular risk, which had been due to run until 2037. The decision appeared online on October 26, 2026. It does not open the Chinese generic market directly, but it removes one instrument from Novo Nordisk's secondary patent portfolio globally, and every generic manufacturer monitoring that portfolio noticed the gap where a 2037 patent used to be.

China is not going generic tomorrow. The data exclusivity wall is real, the classification dispute is unresolved, and seventeen manufacturers in a regulatory queue is not the same as seventeen manufacturers on pharmacy shelves. But the queue exists, the supply chain exists, the patient population is the largest in the world for this indication, and the cardiovascular outcomes patent just got revoked in Europe with global implications. The window will open. Whether it opens in 2027 or somewhere closer to 2032 is the only remaining question, and China's pharmaceutical industry is not waiting around to find out.

While Mumbai pharmacies were running low on generic semaglutide stock and Toronto was processing its fourth generic application, the United States and the European Union were doing something entirely different: absolutely nothing. Not because they were unaware. Because the legal machinery protecting these markets was specifically designed to ensure that no amount of awareness would matter.

The US compound patent for semaglutide, US 8,129,343, did not expire on March 20, 2026 like India's did. The USPTO granted Novo Nordisk a 1,040-day Patent Term Extension under 35 U.S.C. 156, the mechanism created by the Hatch-Waxman Act to compensate drug companies for the time the FDA spends reviewing a drug before it can be sold. Those 1,040 days moved the expiry date to December 5, 2031. December 5, 2031 is exactly 14 years after the FDA approved Ozempic on December 5, 2017, which is also the maximum post-approval exclusivity the statute permits. The law says you can extend a patent, but the combined term cannot give you more than 14 years of post-approval exclusivity. Novo Nordisk hit the ceiling exactly. The statute also says only one patent per active ingredient can receive this extension, so Wegovy and Rybelsus cannot queue up for their own individual extensions on the same molecule. They ride the same December 2031 date.

Novo Nordisk's 2025 Form 20-F lists 2032 as the effective compound expiry across Ozempic, Rybelsus, and Wegovy. That is slightly later than December 2031, which is the FDA Orange Book date. The gap reflects secondary patents — formulation and delivery patents that run past the core compound patent, extending to 2038 for specific products. For Rybelsus, the oral tablet, the estimated US generic entry date is around May 2034, governed by a separate cluster of patents on the delivery agent that makes oral semaglutide work at all. The US market is not one cliff. It is a staircase, and each step is a different patent covering a slightly different aspect of the same drug.

All Seven US Patent Challengers Settle Confidentially

Seven generic manufacturers litigated Paragraph IV challenges against Novo Nordisk in the US: Alvogen, Rio Biopharmaceuticals, Sun Pharma, Dr. Reddy's, Mylan, Zydus, and Apotex. All seven settled. Every settlement is confidential. The confidentiality is the signal. When a generic company settles a US drug patent case on secret terms, the settlement almost always contains a permitted entry date, and industry observers read the collective silence as pointing to generic market entry around 2032. Apotex received the first US FDA Tentative Approval for a generic semaglutide injection on April 10, 2026 — tentative meaning the drug is scientifically approved but cannot be sold until the applicable patents expire. Tentative approval is the FDA's way of saying your science is fine and your wait is not over.

In Europe, the base patent EP 1863839 expired on March 19, 2026. The Supplementary Protection Certificate extended European protection by five years, the maximum SPC term, pushing commercial generic entry to approximately March 2031. The UK SPC, numbered SPC/GB18/023, expires March 19, 2031, five years to the day from the base patent's expiry. SPCs are granted by individual national patent offices across EU member states, so expiry dates can differ slightly across Germany, France, Italy, and Spain, but for semaglutide the March 2031 date is consistent across major European markets. Japan's protection has similarly been extended to approximately 2031. Three major pharmaceutical markets, independently administered, all converging on the same year. The fact that they landed on the same date is a coincidence the same way compound interest is a coincidence.

European generic manufacturers are not idle. Under the SPC manufacturing waiver introduced by EU Regulation 2019/933, they can produce semaglutide during the SPC period for export to non-EU markets where the patent has already lapsed, and they can stockpile product in the final six months before the EU SPC expires to launch on day one. The pipeline is building. The patients cannot access it yet.

The EPO has been chipping at the secondary estate. In T 0291/22, the Board of Appeal found that Novo Nordisk's GLP-1 compositions patent lacked inventive step, with the decision published November 2024. In T 0187/25, decided October 8, 2026, the Board revoked the cardiovascular outcomes patent covering semaglutide's use in reducing major adverse cardiovascular events in diabetic patients at high cardiovascular risk, a patent that had been due to run until 2037. The opponents included STADA, Teva, Sandoz, Apotex, Generics UK, and Adalvo — a roster that reads less like a legal filing and more like a guest list Novo Nordisk did not send invitations for. Neither EPO decision opens the EU market before 2031. What they do is reduce the defensive depth available when the wall eventually comes down.

So this is where the story ends, for now. Generic semaglutide is already a fact of life in India, Canada, Brazil, South Africa, and Saudi Arabia. It is a queue in China. It is a tentative approval in the United States, sitting in a file drawer at the FDA, waiting for a date approximately five years away. The same molecule that costs the equivalent of roughly a dollar fifty a month in vial form in parts of India costs around a thousand dollars a month in the United States. The price difference is not chemistry. It is paperwork. One patent, filed on a Tuesday in March 2006 by a team that almost certainly did not know the molecule would become the world's second-best-selling prescription medicine, produced one expiry date that meant completely different things in a dozen countries simultaneously. Some of that paperwork expires in 2031. The rest expires later. And somewhere in a Canadian pharmaceutical archive, there is still a record of a two-hundred-and-fifty-dollar fee that nobody paid, which turned out to be the most expensive administrative error in the history of Danish pharmaceuticals.