AstraZeneca’s Dapagliflozin Patents: UK Revocation and Reflections on Indian Litigation

INTRODUCTION

AstraZeneca's anti-diabetic medication dapagliflozin which has been sold as Forxiga/Farxiga has been the subject of high-stakes patent litigation across the world. The UK High Court in April 2025 revoked AstraZeneca's patent (EP 1 506 211)[1] and its associated Supplementary Protection Certificates for dapagliflozin which resulted as a setback to the company's monopoly on this medication. While in India, AstraZeneca has fought a concurrent battle to enforce its patents for the same drug against a series of generic manufacturers, and this has given rise to controversies regarding double patenting, inventive step, disclosure requirements, and anti-evergreening provisions. This blog will analyze the UK judgment and Indian litigations, evaluating the insights from UK jurisdiction that could facilitate the advancement of the Indian patent system.

BACKGROUND

India

Dapagliflozin is an oral type 2 diabetes drug that decreases blood glucose by inhibiting glucose reabsorption in the kidney. It has been a commercially successful drug and has other indications for heart failure and chronic kidney disease. Dapagliflozin was developed by AstraZeneca in collaboration with Bristol Myers Squibb and they obtained patents globally. In India, two patents were granted for this drug, being, IN 205147 (IN'147)[2], a genus patent disclosing a general Markush structure of all SGLT2 inhibitor compounds, and IN 235625 (IN'625)[3], a species patent specifically covering dapagliflozin. The patent genus IN'147 had an earlier filing priority which resulted in expiry on October 2020, whereas the species patent IN'625 was filed subsequently and had an expiry date in May 2023. This twin-patent set the stage for double patenting controversies. AstraZeneca purchased rights to these patents from BMS in 2014 and, when IN'147 was close to expiration, generics in India were poised to release low-cost versions of dapagliflozin. AstraZeneca preemptively sued several Indian generic firms (including Intas, Alkem, Zydus, Torrent, MSN, and others) for patent infringement, claiming both IN'147 and IN'625[4]. The fundamental argument was whether one invention could be patented twice and whether the subsequent species patent was valid in the face of prior disclosure by the previous genus patent[5]. AstraZeneca's attempts to uphold its dapagliflozin patents were met with strong opposition from generic manufacturers and an ultimately adverse judicial position. The lawsuit culminated in late 2020, shortly after the genus patent IN'147 lapsed and generics proceeded to launch. AstraZeneca instituted several suits claiming both IN'147 and IN'625, and requested interim injunctions to enjoin generics from producing or selling dapagliflozin. These matters produced two landmark orders by solo judges of the Delhi High Court in November 2020 and a detailed appellate ruling by a Division Bench of the High Court in July 2021, all of which were contrary to AstraZeneca's contention.

[Image Sources: Shutterstock]

One Invention, One Patent: The Delhi High Court adopted a straightforward principle: a single invention (in this case, the dapagliflozin compound) cannot be patentized by two different patents. The courts noted that AstraZeneca's two patents fundamentally addressed the same subject matter of dapagliflozin one in broad terms (genus) and one in specific terms (species). Most importantly, the "Field of Invention" in both IN'147 and IN'625 was a word-to-word repeat, showing that the later patent did not define a clearly different invention. The species patent IN'625 did not reveal any technical improvement or improved efficacy over what is revealed in IN'147. Actually, the specification of IN'625 was silent on improvement of dapagliflozin compared to the common class of compounds; it just asserted the compound per se and its employment. The Division Bench's rationale was that granting two patents over the same drug would actually prolong the period of monopoly in an unjustified way, which was against the patent law principle of only a single patent for an invention[6]. If such double patenting was allowed, a patentee would be able to obtain sequential 20-year terms by initially patenting a genus and subsequently a species, which goes against the legislative policy of capping the patent term and promoting evergreening which is referred to under section 3(d) Indian patent's Act[7].

Failure to Demonstrate Inventive Step: Dapagliflozin is a member of the Markush formula of IN'147, and the inventive idea was generally taught by the genus patent. AstraZeneca's contention was that IN'147 merely generically encompassed DAPA in principle, while IN'625 gave the actual disclosure and a working invention of the compound. The judges were unconvinced: without any showing of "technical improvement or increased effectiveness" in the particular patent, it seemed to be merely an effort to encroach upon what had already been revealed, in a bid to secure more exclusivity. The November 2, 2020 single-judge ruling extending to suits against Alkem and Intas categorically held that IN'625 "had no technical advancement and hence lacked an inventive step mentioned under Section 2(1)(ja)[8]," considering that IN'147 wholly disclosed dapagliflozin. These results reflect the UK court's "arbitrary selection" consideration in both jurisdictions, dapagliflozin selected from a known class with an unproven special benefit was held not patentable. The Indian Division Bench took one step further to explain that where the same applicant or inventor is the mind behind both the genus and species patents, the inventive step bar is judged not from the eyes of a normal skilled person reading another's prior patent, but indeed from the eyes of the inventor who already possessed knowledge of the genus. In such a case, the inventor is assumed to be "a person in the know," and will need to lucidly explain the precise enhancements or surprising properties that make a second patent necessary. AstraZeneca's IN'625 had failed this test by not recognizing any such enhancement over IN'147.

Disclosure Obligations: Section 8[9] mandates disclosure by patent applicants regarding corresponding foreign patent applications and developments thereon. A contravention can be a basis for revocation which is drawn under Section 64(1)(m)[10]. In the case of dapagliflozin, the defendants emphasized that AstraZeneca violated Section 8 by failing to report to the Indian Patent Office that, in the U.S., a terminal disclaimer had been submitted to surmount double-patenting objections, such that the U.S. species patent lapsed simultaneously with the genus patent[11]. This fact was material as it reflected that AstraZeneca recognized the two patents were not inventively different. The Delhi High Court noted this lacuna seriously. The single judge and the Division Bench both held prima facie non-compliance with Section 8(2), which contributed to the injunction denial.

Markush Claim vs. Specific Disclosure: One of the theme issues in this litigation was the so-called "coverage-disclosure dichotomy". AstraZeneca argued that dapagliflozin was not delineated clearly in the genus patent, suggesting that the species patent disclosed a new invention which is the individual molecule and that was worthy of its own protection. The generics, on the other hand, argued that simply not referring to the compound in a list of examples doesn't remove it from the scope of the previous patent, as the Markush structure was sufficiently broad to encompass it. The Delhi High Court concurred with the latter interpretation: IN'147's protection was sweeping and facilitated to the degree that a person skilled in the art could have reached dapagliflozin even if the molecule wasn't expressly pinpointed. AstraZeneca's applications for interim injunctions were rejected at all levels. The November 2020 Single Judge orders denied relief, and upon appeal, the Division Bench on July 20, 2021 rejected AstraZeneca's appeals. AstraZeneca subsequently approached the Supreme Court of India, which in July 2022 also turned down AstraZeneca's request to enjoin the generics.

The refusal by the Supreme Court effectively sealed the High Court's conclusions, opening the way for Indian Pharma companies to keep producing low cost dapagliflozin. By then, several generic formulations were on sale in India at a tiny fraction of AstraZeneca's cost, showing the public interest aspect of the case. Significantly, the issue before the Supreme Court was at an interlocutory stage, and the revocation proceedings on substance could have gone ahead, nonetheless, considering that IN'625 was to expire in May 2023, any additional challenge became of limited academic interest. The Delhi High Court rulings are a strong precedent on the non-permissibility of double patenting and stringent examination of marginally inventive pharmaceutical innovations in India.

UK and Europe

AstraZeneca's EP 1 506 211 patent which is similar to the dapagliflozin species patent in India expired in May 2023, yet AstraZeneca had received SPCs adding protection until May 2028 (SPC/GB13/021 for dapagliflozin and a corresponding SPC/GB14/050 for a dapagliflozin/metformin combination). These SPCs, a type of patent term extension found in the EU/UK for pharmaceuticals, effectively precluded generics from being introduced on the market after the 20-year period of the patent. A number of generics (Teva, Viatris, Glenmark) replied by contesting the validity of the dapagliflozin patent/SPC in UK courts. This resulted in a March 2025 High Court trial. In a ruling handed down by Deputy Judge Michael Tappin KC on April 28, 2025, the High Court of England & Wales held AstraZeneca's patent for dapagliflozin invalid on several grounds. Accordingly, the related SPCs that rely on the validity of the patent were also invalidated[12]. The ruling was in response to two principal arguments presented by the generic challengers, bein

Inventive Step/"Arbitrary Selection" from Prior Art: The generics referred to an earlier published patent application (WO 01/27128, published in 2001)[13] which revealed a wide genus of SGLT2 inhibiting compounds, a disclosure including dapagliflozin amongst several potential analogues. AstraZeneca's patent for the particular compound was claimed to contribute nothing technically over this earlier disclosure of the genus which was found by the high court. That is, the choice of dapagliflozin from among the many suspects in the familiar Markush pool was considered an obvious selection, devoid of inventive step under Section 3 of the UK Patents Act 1977[14] and Article 56 EPC[15].

Lack of Plausibility of Therapeutic Efficacy: The High Court also questioned whether the specification of the patent made it reasonable to believe that dapagliflozin would indeed be effective as an anti-diabetic medicine[16]. In accordance with UK precedents Warner-Lambert v. Generics[17] and Akebia v. Fibrogen[18], a patent should disclose sufficient to make the claimed therapeutic effect plausible at filing. In the present case, it was held that the patent did not adequately disclose data or facts to make it plausible for its intended purpose under Section 14(3) of the UK Patents Act 1977[19] and Article 83 EPC[20].

Commercial Success Irrelevant: AstraZeneca had contended that the significant commercial success of dapagliflozin and favorable clinical results highlighted its inventive merit. The High Court, however, asserted unequivocally that ex post commercial success is not a relevant factor in the legal determination of validity. The focus must remain on what the skilled person would have understood from the prior art and the patent disclosure at the relevant time, not on real-world market performance after the fact.

Overall, the UK High Court’s decision unambiguously revoked the dapagliflozin patent and its SPC extension, opening the door for immediate generic entry.

RECOMMENDATIONS AND CONCLUSION

The saga of AstraZeneca’s dapagliflozin patents is seen to be a victory for generics in both the UK and India and offers important lessons and prompts reflection on possible patent law reforms, especially for India.

The Indian courts in this case firmly applied the one-invention-one-patent principle. To avoid uncertainty, India could consider explicitly codifying a double patenting prohibition in the Patents Act or Patent Office guidelines. This would guide examiners to refuse a second patent on substantially the same invention. For instance, if an applicant already has a patent on a genus of compounds, any later application claiming a species falling wholly within that genus should trigger an automatic inquiry of “whether there exists a new technical effect or significant improvement?” If not, the application should be rejected upfront for double patenting citing obviousness. Clear rules would disincentivize evergreening attempts and reduce litigation. While Section 13 of the Act directs examiners to cite earlier patents, an explicit bar akin to the U.S. “obviousness-type double patenting” doctrine could be introduced.

The dapagliflozin case highlights the challenges broad Markush claims pose. The Patent Office should ensure that Markush claims are not allowed to overreach without sufficient support. Requiring applicants to identify representative compounds and disclose specific activity data for at least a reasonable subset can prevent situations where a later selected compound seems like a fresh invention simply because the original patent was too generic. If a Markush claim is extremely broad, examiners might restrict the claim or demand divisionals so that undisclosed species are not effectively “blocked” by an overly broad patent. This could mitigate later disputes by ensuring genus patents are granted at a scope commensurate with their disclosure.

Indian law could take a cue from the UK/EPO’s evolving plausibility doctrine. Although not explicitly required by statute, adopting a practice where patent applications for new drugs or new uses include experimental data or a sound scientific rationale would improve patent quality. The Indian Patent Office can issue guidelines that mere assertions of efficacy in the specification are not enough but some proof or at least a well-reasoned biological mechanism should be disclosed to make the claim credible.

A key structural issue exposed by the AstraZeneca Dapagliflozin litigation in India is the failure to resolve revocation proceedings before the patent’s expiry, despite strong prima facie findings of invalidity. While the Delhi High Court and Supreme Court swiftly dealt with interim injunction applications in 2020–21, the core revocation petitions filed under Section 64 were not adjudicated to finality, and the patent expired in May 2023 without a formal decision on validity. This stands in sharp contrast to the UK, where the High Court conducted a full revocation trial promptly and invalidated the patent within two years post-expiry, providing clarity on the legal status of the molecule. The Indian experience reflects a systemic problem of delayed resolution in pharmaceutical patent trials, allowing potentially weak patents to remain technically in force and impeding generic entry. To address this, India must implement procedural reforms to foster faster adjudication of patent challenges—such as mandating strict adherence to timelines, enabling scientific advisors for complex technical issues, and strengthening the functioning of specialized IP benches.

In conclusion, the dapagliflozin case reinforces the importance of balancing innovation incentives with public interest, which is basically strong patents to reward true innovation, but no tolerance for patent layering or unwarranted extensions. By learning from the UK jurisdiction, India can continue to develop a patent regime that encourages real drug discovery while ensuring affordable access to life-saving medicines once the legitimate patent term has elapsed.

Author:  Shakthi Bharathee, in case of any queries please contact/write back to us via email to chhavi@khuranaandkhurana.com or at  

[1] European Patent No. EP 1 506 211 B1.

[2] Indian Patent No. IN 205147.

[3] Indian Patent No. IN 235625.

[4] FE Bureau, Supreme Court Rejects AstraZeneca Plea Against Manufacturing of Low-Cost Versions of Its Diabetes Drug, Fin. Express (July 18, 2022), https://www.financialexpress.com/life/supreme-court-rejects-astrazeneca-plea-against-manufacturing-of-low-cost-versions-of-its-diabetes-drug-2599439.

[5] Business Today Bureau, AstraZeneca in Legal Fight with 12 Drug Makers over Diabetic Drug Dapagliflozin, Bus. Today (Oct. 13, 2020), https://www.businesstoday.in/industry/pharma/story/astrazeneca-in-legal-fight-with-12-indian-drug-makers-over-rights-to-diabetic-drug-dapagliflozin-275589-2020-10-13.

[6] Times News Network, Court Rejects AstraZeneca’s Plea on Diabetes Drug Patent, Times of India (Nov. 18, 2020), https://timesofindia.indiatimes.com/business/india-business/court-rejects-astrazenecas-plea-on-diabetes-drug-patent/articleshow/79291466.cms.

[7] The Patents Act, No. 39 of 1970, § 3(d) (India).

[8] The Patents Act, No. 39 of 1970, § 2(1)(ja) (India).

[9] The Patents Act, No. 39 of 1970, § 8 (India).

[10] The Patents Act, No. 39 of 1970, § 64(1)(m) (India).

[11] U.S. Patent No. 6,514,117

[12] Mathieu Klos, UK High Court Revokes AstraZeneca’s SPC and Patent for Diabetes Drug Dapagliflozin, JUVE Patent (Apr. 28, 2025), https://www.juve-patent.com/cases/uk-high-court-revokes-astrazenecas-spc-and-patent-for-diabetes-drug-dapagliflozin.

[13] World Intellectual Property Organization, International Publication No. WO 01/27128 A1.

[14] Patents Act 1977, c. 37, § 3 (UK).

[15] European Patent Convention, art. 56.

[16] EIP, Plausibility at the Forefront of the UK High Court’s Decision in Finding AstraZeneca’s Patent Invalid, EIP Legal Update (Apr. 30, 2025), https://eip.com/us/latest/article/plausibility_at_the_forefront_of_the_uk_high_court_s_decision_in_finding_astrazeneca_s_patent_covering_blockbuster_diabetes_drug_invalid.

[17] Warner-Lambert Co. Ltd. v. Generics (UK) Ltd., [2018] UKSC 56.

[18] Akebia Therapeutics Inc. v. FibroGen Inc., [2021] EWHC 866 (Pat).

[19] Patents Act 1977, c. 37, § 14(3) (UK).

[20] European Patent Convention, art. 83.