GLP-1 Drug Patent Landscape

What this research found

An attempt to find prior-art vulnerabilities in three GLP-1 receptor agonist drugs — liraglutide, dulaglutide and tirzepatide — by comparing when each first appeared in the scientific literature against when the FDA approved it. All three showed exactly a four-year gap, and no publication fell inside the ten-year pre-approval window that would flag a potential early disclosure. The report's own conclusion is that this null result mainly reflects a six-year hole in literature coverage for liraglutide rather than genuinely unchallengeable patents.

• All three drugs show an identical four-year lag from first publication to FDA approval: liraglutide 2005 to 2009, dulaglutide 2010 to 2014, and tirzepatide 2018 to 2022.
• No early disclosures were detected. Using publication more than 10 years before approval as the test, zero of the 1,003 indexed publications qualified for any of the three drugs.
• That null result is only trustworthy for the newest drug. The literature set begins in 2005, which leaves a six-year gap for liraglutide (1999 to 2004) and a one-year gap for dulaglutide over precisely the period when core patents would have been filed, while tirzepatide's window is fully covered.
• Co-citation patterns separate tirzepatide from its predecessors. Liraglutide and dulaglutide share a Jaccard index of 0.190 across 70 co-cited papers, whereas tirzepatide reaches only 0.065 with dulaglutide and 0.023 with liraglutide, on 8 co-citations each — consistent with its distinct dual GIP/GLP-1 mechanism.
• Academic attention peaks at very different times relative to launch: liraglutide at 29 papers in 2016, seven years after approval; dulaglutide at 9 papers in 2015, one year after; and tirzepatide at 5 papers in 2020, two years before approval.
• Back-calculating filing dates from approval puts estimated core patent expiry at roughly 2019 for liraglutide and 2024 for dulaglutide, both already passed, against roughly 2032 for tirzepatide.

How it was done

PubMed supplied a timeline of 1,003 publications from 2005 to 2025 mentioning the three drugs, while OpenAlex contributed 465 of those papers with richer metadata for the network work, and FDA Orange Book records anchored the approval dates. With no direct filing dates available from patent office interfaces, core patent filing was approximated as approval minus ten years, and any publication predating that estimate was flagged as a candidate early disclosure. Drug co-occurrence across the OpenAlex papers was scored with the Jaccard index to build a three-node network, and the output was a 15-page report with figures on publication velocity, the co-occurrence network, and research lag. An initial run using OpenAlex alone produced impossible negative research lags, which is why the two-source split was adopted.

Data sources

• PubMed — 1,003 publications mentioning the three drugs, 2005 to 2025
• OpenAlex — 465 linked papers with metadata (liraglutide 334, dulaglutide 105, tirzepatide 26)
• FDA Orange Book — approval dates and patent information
• ChEMBL — compound structure data

Figures from this analysis

Outputs produced

• uspto_patents.json

  USPTO patent data - API limitation documentation

• openalex_literature.json

  OpenAlex scientific literature (2005-2025)

• chembl_compounds.json

  ChEMBL compound data for known GLP-1 agonist drugs

• timeline_data.csv

  Unified timeline of publications and drug approvals

• paper_drug_links.json

  Drug-to-paper linkage mapping from text mining

• analysis_summary.json

  Comprehensive temporal and network analysis summary

• drug_cooccurrence.csv

  Co-occurrence matrix for drug pairs

• early_disclosures.csv

  Early disclosure papers (>10 years before approval)