Accord Healthcare v Sanofi (Paris Judicial Court, RG 21/06416, 2024)

Dhenne Avocats · 11 October 2026

Our analysis

Summary

The Paris Judicial Court annuls, for lack of inventive step, the French part of European patent EP 2 493 466, which claims cabazitaxel, in combination with prednisone or prednisolone, for treating metastatic castration-resistant prostate cancer in patients previously treated with docetaxel. The infringement action against Accord’s hybrid generic is dismissed. Having rejected the priority, insufficiency and novelty attacks, the court holds that the Tropic phase III trial, announced before the priority date, gave a reasonable expectation of success. The judgment thus distinguishes the weight of an announced clinical trial when relied on against novelty and against inventive step.

Facts and procedure

Sanofi Mature IP’s patent, filed on 27 October 2010 and granted on 10 March 2021, claims priority from US applications, the earliest dated 29 October 2009 (§ 3). Sanofi Winthrop Industrie and Sanofi-Aventis France are its licensee and sub-licensee (§ 1). Accord Healthcare France has marketed a hybrid generic of Jevtana in France since 14 June 2021 (§ 2). Sanofi’s applications for interim injunctions were refused in 2021, the challenge to validity being regarded as serious (§ 5). At the EPO, the oppositions were rejected on 15 December 2023 and an appeal is pending (§ 3). Accord sued for nullity on 6 May 2021, Sanofi for infringement on 8 September 2021.

The applicable law

Following the Cour de cassation (commercial chamber, 22 November 2016) and G 2/98, it holds that the invention is the same where the skilled person can derive it directly and unambiguously from the earlier application as a whole (§ 37). Where the therapeutic effect is a functional feature of the claim, the application must make it credible to the skilled person, without any need for proof (§ 51, citing the Cour de cassation of 6 December 2017, G 1/03 and G 2/21). The court extends that test to novelty: a feature forms part of the state of the art only if its earlier disclosure makes it workable, that is to say credible (§ 62). It also reads Article L. 615-2 of the Intellectual Property Code as entitling licensees to sue for infringement in respect of their own loss (§§ 15 to 18).

Question

Is a second medical use patent new and inventive where the clinical trial designed to establish the claimed effect had been announced, together with its protocol, before the priority date?

Decision

Since claim 1 covers only administration in two separate preparations, the US application’s silence on a single preparation is irrelevant, and the 29 October 2009 priority is validly claimed (§§ 40 to 46). Insufficiency fails, since the patent need not identify every non-responding subgroup (§ 53). So does the novelty attack: the documents describing the Tropic trial and the NHSC document set out the protocol without making efficacy credible (§§ 65 and 66), and it is not shown that the trial end date published on ClinicalTrials.gov predates the priority date (§ 68).

Inventive step is lacking. The Tropic trial is a relevant starting point, and the invention is obvious if that person would have pursued the hypothesis it tested with a reasonable expectation of success (§ 98). Declining to lay down an absolute rule, the court holds that a clinical trial does not in itself destroy inventive step but informs the teaching of the rest of the prior art (§ 102), having noted, with the Paris Court of Appeal judgment of 22 March 2023 (fingolimod), that the announcement of a phase III trial may strengthen the expectation of success (§ 101). It relies on encouraging phase II results in taxane-pretreated breast cancer (§ 105), on cabazitaxel’s lower affinity for P-gp (§ 106) and on the rational choice of a 25 mg/m² dose (§ 112). The trial therefore had a reasonable chance of showing a favourable effect, including a modest gain in survival (§ 116). Claim 1 and the dependent claims lack inventive step (§§ 117 to 121). The French part of the patent is annulled (§ 122), the infringement claims are dismissed (§ 124) and the Sanofi companies are ordered jointly and severally to pay EUR 150,000 in costs under Article 700 of the Code of Civil Procedure (§ 126). Provisional enforcement is maintained, save for entry in the register, since annulment has absolute effect once pronounced (Article L. 613-27 of the Intellectual Property Code, §§ 129 and 130).

Key points for practice

• Announcing a clinical trial and its protocol does not destroy the novelty of a therapeutic use unless the disclosure makes the effect credible (§§ 62, 65 and 66).
• The same trial, read with the rest of the prior art, may nonetheless found a reasonable expectation of success and defeat inventive step (§§ 98, 102 and 116).
• Practical point: where the trial was announced before filing, inventive step turns on the earlier data that feed or temper the skilled person’s optimism.

Provisions applied

European Patent Convention

art. 52, art. 54, art. 56, art. 83, art. 87, art. 88, art. 138

National law

Arts L. 613-27, L. 614-12 and L. 615-2 of the French Intellectual Property Code; Arts 31, 32, 514, 514-1, 696 and 700 of the French Code of Civil Procedure

Case law cited

Cass. com., 22 November 2016, No 15-16.647; Cass. com., 6 December 2017, No 15-19.726; Cass. com., 17 May 2023, No 19-25.509; Cass. com., 20 November 2012, No 11-18.440; CA Paris, 22 March 2023, RG 22/11165; EPO, G 2/98, G 1/03, G 2/21, T 609/02, T 293/07

Related decisions

• G 2/21, 23 March 2023, cited for the credibility of the therapeutic effect (§§ 51 and 102).
• G 2/98, 31 May 2001, cited for the concept of the same invention in priority matters (§ 37).
• T 609/02, 27 October 2004, cited on credibility based on preclinical data (§ 102).

Prepared by Dhenne Avocats from the text of the decision (full text reproduced in the PIBD, pibd.inpi.fr), consulted on 11 October 2026. Only the official text is authoritative.