Januvia: What a Real Patent Cliff Looks Like

The drug

Januvia (sitagliptin) was approved by the FDA in October 2006 as the first DPP-4 inhibitor - a class of oral diabetes drugs that work by blocking the enzyme dipeptidyl peptidase-4, which breaks down incretin hormones. By preserving incretins, sitagliptin helps the body produce more insulin after meals and reduces glucose production.

Janumet followed in 2007 - a single pill combining sitagliptin with metformin, the most widely prescribed diabetes drug in the world. Metformin reduces glucose production in the liver; sitagliptin helps the body produce more insulin after meals. Two mechanisms in one tablet, one prescription, one copay. Together they became Merck's largest franchise outside of vaccines - a $6 billion-a-year business at peak.

Sitagliptin is a small molecule. Its molecular weight is 523 daltons. It is made through chemical synthesis, not in living cells. This distinction - which we explored in Why Biosimilars Aren't Generics - determines everything about what happens when the patent expires. Januvia's key US patent expires in November 2026, with generics eligible to launch as early as May 2026 under settlement agreements.

The chart tells the story in three phases. Growth from 2009 to 2012, as sitagliptin became the standard oral add-on for type 2 diabetes. A long plateau from 2012 to 2021, holding steady above $5 billion. Then a steep decline - from $5.3 billion in 2021 to $2.2 billion in 2024 - before a single generic has launched.

The triple squeeze

Januvia's revenue is being compressed from two directions simultaneously.

From outside the US: generic competition is already here. Januvia lost exclusivity in Europe at the end of 2021. Generic sitagliptin launched across the EU in 2022. In India, multiple manufacturers (Zydus, Glenmark, Cadila) launched generic versions in mid-2022 at a fraction of the branded price. The US is the last major market where Januvia still has patent protection - and that protection ends in May 2026.

From above: GLP-1 receptor agonists. Drugs like Ozempic (semaglutide) and Mounjaro (tirzepatide) have reshaped the type 2 diabetes market. They produce better glucose control, significant weight loss, and cardiovascular benefits that DPP-4 inhibitors cannot match. Physicians are increasingly starting patients on GLP-1s instead of DPP-4s, and switching existing patients over. The clinical case for DPP-4 inhibitors has weakened - not because sitagliptin stopped working, but because better options arrived.

From below: generic anticipation. Payers and pharmacy benefit managers have been positioning formularies for the May 2026 generic launch for years. Prior authorization requirements have tightened. Preferred status has shifted. The economic pressure of an approaching cliff starts well before the cliff itself - insurers have no incentive to lock in patients on a branded drug that is about to become available at 90% off.

The result is a 64% revenue decline from peak - and the generics have not shipped yet.

The generic flood

Januvia's key US patent (US 7,326,708, covering the sitagliptin phosphate salt) expires in November 2026, with pediatric exclusivity extending to May 2027. But under settlement agreements between Merck and multiple generic challengers, generics can launch seven months early - in May 2026. Generic sitagliptin/metformin (Janumet equivalent) follows the same timeline, with the extended-release version (Janumet XR) by July 2026.

This is how small-molecule patent cliffs work. The generic is chemically identical to the brand. The FDA approval pathway (ANDA) requires only bioequivalence, not clinical trials. Development costs roughly $1-3 million per product. Dozens of companies can file simultaneously. And pharmacists can substitute automatically - no physician decision required, no state-by-state variation.

Want the next brief? Subscribe — free, every week.

Within a year of generic launch, prices typically drop 80-90%. Within two years, the brand retains single-digit market share. This is the Lipitor model, the Plavix model, the Singulair model. It is fast, steep, and irreversible.

Merck has seen this before

Singulair (montelukast), Merck's blockbuster asthma drug, lost US patent protection in August 2012. The revenue collapse was textbook:

$5.5 billion in 2011. $1.2 billion in 2013. A 78% decline in two years. That is the speed of a small-molecule generic cliff. Januvia's trajectory is heading for the same destination - the only question is how fast it gets there.

Merck's bigger problem

Januvia's cliff is significant, but it is not Merck's largest exposure. That distinction belongs to Keytruda (pembrolizumab), the PD-1 checkpoint inhibitor that generated $29.5B in 2024 revenue - making it the world's best-selling drug. Keytruda's key patents begin expiring around 2028.

But Keytruda is a biologic, not a small molecule. Its patent cliff will follow the Humira model, not the Januvia model - biosimilar competition, not generic competition. The decline will be slower. Merck will have more time.

The irony is that Januvia - the smaller franchise - will experience the faster, more complete cliff. Merck is losing $6 billion the fast way and $30 billion the slow way, simultaneously.

A $6 billion franchise, built over a decade, dismantled in under four years - and the generics have not even launched yet. That is what a real patent cliff looks like. No patent settlements to delay entry. No complex manufacturing to limit competition. No physician switching decisions to slow adoption. Just a chemical formula that anyone can copy, and a price that drops to a fraction of what it was. The molecule dictates the market.