Invention Lacks Inventive Step: Delhi HC Rejects Esteve's Tramadol-Celecoxib Co-Crystal Patent Appeal

At the heart of the dispute was a pharmaceutical co-crystal involving tramadol and celecoxib. The claimed invention sought protection for a co-crystal comprising tramadol, either as a free base or physiologically acceptable salt, together with at least one coxib selected from celecoxib or its salts. The claims also covered specific forms, including a 1:1 molecular ratio of rac-tramadol hydrochloride and celecoxib, specified crystalline characteristics, a production process and pharmaceutical compositions containing the co-crystal.

Esteve filed the national-phase application before the Indian Patent Office on February 15, 2012, based on PCT Application No. PCT/EP2010/002385 dated April 19, 2010. The Request for Examination was filed on September 18, 2013. Subsequently, on October 30, 2017, the Patent Office issued a First Examination Report raising objections relating to novelty and inventive step under Section 2(1)(ja), besides objections under Sections 3(d) and 3(e).

Esteve responded to the examination objections and amended its claims. Later, hearing notices issued in November and December 2019 introduced additional prior-art documents, D3 to D7, as well as an objection concerning sufficiency of disclosure under Section 10(4). Esteve’s representative appeared at the hearing on January 17, 2020, following which written submissions and another set of amended claims were filed on January 29, 2020.

The Controller ultimately rejected the application on February 12, 2020. Esteve then challenged that decision before the Delhi High Court. Notably, although lack of novelty had initially been raised during prosecution, Esteve argued that the final rejection did not sustain the novelty objection after the claims had been amended. The company therefore maintained that the remaining objections, particularly under Section 3(d), were legally unsustainable because the claimed subject matter had effectively been accepted as novel.

On Section 3(d), Esteve argued that the Patent Office had failed to identify the relevant “known substance” against which the claimed co-crystal was required to demonstrate enhanced therapeutic efficacy. According to the appellant, the Controller was required to specify the known substance, explain how the claimed invention constituted a new form or derivative of that substance and establish the basis for comparing therapeutic efficacy. Esteve argued that the applicant could not be left to guess which known substance was being relied upon for the Section 3(d) objection.

The company further relied on experimental data in the patent specification. It pointed to Figure 6, concerning mechanical allodynia in rats, and Figure 8, concerning thermal hyperalgesia, to argue that the 1:1 co-crystal of rac-tramadol-HCl and celecoxib demonstrated greater activity than the individual drugs. The appellant highlighted an ED50 value of 2.04 for the co-crystal, compared with 5.41 for tramadol and 3.03 for celecoxib, arguing that the lower ED50 demonstrated superior efficacy and synergy.

On Section 3(e), Esteve strongly disputed the characterization of the invention as a mere admixture. It argued that a co-crystal is fundamentally different from a physical mixture because it constitutes a single crystalline phase formed through specific intermolecular interactions between the API and the co-crystal former. The appellant relied upon Powder X-Ray Diffraction data, contending that a co-crystal produces a new and unique diffraction pattern rather than merely displaying the characteristic peaks of its individual components.

The company also challenged the finding of lack of inventive step. It argued that prior art D5 disclosed combinations of tramadol with NSAIDs, but did not specifically teach or suggest the formation of a tramadol–celecoxib co-crystal. According to Esteve, D5 merely listed celecoxib among numerous NSAIDs, while D7 generally disclosed API–API co-crystals without specifically directing a skilled person towards tramadol and celecoxib.

Esteve argued that selecting tramadol and celecoxib from the large number of possible APIs disclosed in the prior art would involve an enormous number of possible combinations. It relied on evidence that approximately 500 API–API candidates had been tested by the appellant, of which only 13 formed co-crystals. The company therefore argued that the specific co-crystal was a rare and unpredictable result rather than an obvious modification.

The appellant also pointed out that D7 had a priority date dating back to 2004, yet the specific tramadol–celecoxib co-crystal was not developed for several years. It argued that this prolonged absence of the claimed combination supported the conclusion that the invention was not obvious. Esteve additionally informed the Court that D5 and D7 had been considered during prosecution of corresponding applications in jurisdictions including Israel, Brazil and the United States, where patents had nevertheless been granted.

The respondent’s principal argument was that tramadol and celecoxib were already well-known therapeutic agents, particularly for pain management. According to the Controller, prior-art documents already taught the use of tramadol with NSAIDs, while D5 specifically identified celecoxib as one of the NSAIDs that could be used with tramadol.

The respondent placed particular reliance on D5, which disclosed pharmaceutical combinations involving slow-release tramadol and NSAIDs. Although naproxen was used as an example, D5 expressly listed celecoxib among the NSAIDs that could be substituted or used in such combinations. The Controller argued that a person skilled in the art would therefore have been motivated to consider tramadol and celecoxib together.

D7 was relied upon for the co-crystallization aspect. According to the Controller, D7 disclosed API–API co-crystals, hydrogen bonding between APIs and methods for preparing such co-crystals. It also described improvements that could result from co-crystallization, including enhanced solubility, dissolution, bioavailability, stability and dose response.

The respondent therefore presented D5 and D7 as complementary prior art: D5 supplied the motivation and disclosure for combining tramadol with celecoxib, while D7 supplied the teaching for forming API–API co-crystals and the methods for preparing them. On this basis, the Controller argued that the claimed invention represented no more than an obvious application of known teachings.

The Controller also disputed Esteve’s claim of enhanced therapeutic efficacy. The respondent argued that the data did not demonstrate sufficient improvement across the broad therapeutic conditions covered by the claims. It particularly questioned whether the reported results established genuine synergy rather than merely comparable activity.

With respect to Section 3(e), the Controller maintained that the combination of two active pharmaceutical agents could amount to a mere admixture where the required synergistic effect was not established. The respondent consequently sought dismissal of the appeal.

However, the Court focused its detailed analysis on Section 2(1)(ja) and the relevant prior art, particularly D5 and D7.

The Court found that D5 expressly disclosed celecoxib as an NSAID that could be used in combination with tramadol. It further noted that D5 stated that its examples using naproxen were illustrative and that a person skilled in the art would know how the combination could be modified using other NSAIDs, including celecoxib.

The Court then examined D7. It found that D7 disclosed API–API co-crystals and described improvements such as increased solubility, dose response, dissolution, bioavailability and stability. The Court observed that these were substantially aligned with the improvements claimed by Esteve for its co-crystal.

The Court also identified a structural similarity between the claimed invention and D7. D7 disclosed hydrogen bonding between APIs, while Esteve’s own specification defined a co-crystal as a crystalline material in which components were held together through weak interactions, including hydrogen bonds.

Most importantly, the Court considered D5 and D7 together rather than as isolated disclosures. It concluded that D5 supplied the teaching concerning the tramadol–celecoxib combination, while D7 supplied the teaching concerning API–API co-crystallization and the associated benefits and preparation methods.

The Court further observed that D7’s disclosure of improvements associated with co-crystals provided the necessary teaching, motivation and suggestion to apply the co-crystal technology to the tramadol–celecoxib combination disclosed in D5. D5’s disclosure that certain combinations could produce synergistic analgesic effects further supported this motivation.

The Court also found an overlap between the claimed manufacturing process and D7. The claimed process involved dissolving the components, combining them, cooling and evaporating solvent to obtain the co-crystal. D7 likewise disclosed dissolving separate components in a solvent, combining them and allowing the co-crystal to precipitate or crystallize as the solvent evaporated.

The Court held that the disclosure of the tramadol–celecoxib combination in D5, when read with the co-crystal technology and associated improvements disclosed in D7, was sufficient to make the claimed invention obvious to a person skilled in the art.This was the decisive issue in the case.

The Court also dismissed the pending applications and directed that there would be no order as to costs.

In practical terms, the judgment means that Esteve’s Indian patent application for the claimed tramadol–celecoxib co-crystal remains rejected. The High Court did not order the Patent Office to grant the patent; instead, it upheld the rejection on the ground of lack of inventive step.