Why does Roche’s oncology blockbuster Herceptin (trastuzumab) have 1,536 patents in a comprehensive search of international patents but only 393 when searching for patent titles? This four-fold difference is a window into pharmaceutical strategy.
For complex biologics, especially monoclonal antibodies, the platform product strategy is gaining ground. Popularized in part by Humira (adalimumab), which is now fending off biosimilar competition, this platform-based approach involves continuously expanding the indications for the molecule in question over time. In addition to treating new diseases, this strategy also focuses on patenting the drug’s use in combination with other therapies, making it a foundational part of modern treatment protocols. The patent data for Trastuzumab (Herceptin) is a key example: its patent portfolio is dominated by patents for new indications (452 patents) and combination therapies (354 patents) in our review of patent filings from USPTO (United States), EPO (European Patent Office). WIPO (PCT international applications), JPO (Japan), CNIPA (China) and KIPO (Korea).
The temporal dimension of these biologics patent portfolios also pokes holes in the traditional notion of a “patent cliff,” where a drug loses protection and faces generic competition. For instance, Herceptin, first approved in 1998, shows strong patent activity decades later: 832 patents filed in the 2010s, followed by 590 patents in just the first five years of the 2020s.
Meanwhile, the patenting strategy for small molecules is more like building a fortress. Bristol Myers Squibb and Pfizer’s Eliquis (apixaban) exemplifies this approach with 226 manufacturing patents and 213 formulation patents. Similarly, Bayer and Johnson & Johnson’s Xarelto (rivaroxaban) leads with 282 manufacturing patents. These companies are building what patent lawyers call a “thicket.” This involves patenting every crystalline form, every manufacturing step, every coating and release mechanism to create impenetrable barriers against generic competition.
| Rank | Drug Name | Methodology A: IFI Claims (Comprehensive w/ pharma filter) | Methodology B: IFI Claims (Title-only search) |
| 1 | Trastuzumab (Herceptin) | 1,536 | 393 |
| 2 | Bevacizumab (Avastin) | 919 | 348 |
| 3 | Lenalidomide (Revlimid) | 824 | 481 |
| 4 | Ibrutinib (Imbruvica) | 792 | 452 |
| 5 | Biktarvy (HIV combo) | 744 | 57 |
| 6 | Apixaban (Eliquis) | 681 | 686 |
| 7 | Rivaroxaban (Xarelto) | 679 | 621 |
| 8 | Adalimumab (Humira) | 585 | 248 |
| 9 | Etanercept (Enbrel) | 517 | 310 |
| 10 | Pembrolizumab (Keytruda) | 497 | 88 |
The data for this analysis was sourced from the IFI CLAIMS patent database, accessed via Google’s BigQuery platform on June 25, 2025. To provide a transparent and insightful view, we adopted two distinct methodologies to count patents for a list of 15 top-selling drugs.
The “comprehensive search” option` represents a quality-filtered total. This method counts any patent that mentions a drug’s brand or generic name within its title or abstract, but only if the patent is also classified under a relevant pharmaceutical category using Cooperative Patent Classification (CPC) codes (e.g., A61K, A61P). This step ensures that the results are focused on medically relevant inventions and not merely incidental mentions.
To show the sensitivity of patent counts to search parameters, we also performed a title-only search. This restrictive method only counts patents that feature the drug name in the title. By comparing these two results, we can demonstrate how a narrow search scope can dramatically underrepresent the full patent landscape.
For instance, for Herceptin, here’s the breakdown of the difference:
-
Title-Only Search (393 patents): These patents are likely directly about Herceptin. The main invention is a new formulation, a new use, or a new manufacturing process for Trastuzumab itself. The title is the “headline,” so if Herceptin is the star of the show, its name will be there.
-
Comprehensive Search (1,536 patents): This includes the 393 “title-only” patents plus an additional approximately 1,100 patents where Herceptin is mentioned in the abstract but not the title. In general, this type of additional patents are often about combination therapies, comparative studies or broader methods.
biologics maintain intense patenting activity decades after approval, while small molecules show a more traditional lifecycle:
| Drug Name | First Patent | Years Active | 2000s Patents | 2010s Patents | 2020s Patents* |
| Trastuzumab (Herceptin) | 2000 | 24.5 | 243 | 832 | 590 |
| Bevacizumab (Avastin) | 2005 | 19.9 | 72 | 600 | 311 |
| Pembrolizumab (Keytruda) | 2016 | 8.9 | 0 | 115 | 395 |
| Apixaban (Eliquis) | 2007 | 18.0 | 15 | 377 | 290 |
| Rivaroxaban (Xarelto) | 2006 | 18.6 | 16 | 493 | 233 |
*2020s data through February 2025
