Introduction: The case of F. Hoffmann-La Roche Ltd. & Anr. vs. Cipla Ltd. represents a pivotal moment in Indian patent law, particularly in the pharmaceutical sector. This high-profile dispute centered on the alleged infringement of a patent for Erlotinib, a groundbreaking cancer treatment drug marketed as Tarceva. The case raised critical questions about patent validity, infringement, and the scope of protection for polymorphic forms of pharmaceutical compounds. It also highlighted the tension between intellectual property rights and public interest in access to affordable life-saving drugs. Decided by the Delhi High Court in 2012, the judgment offers profound insights into the application of India’s Patents Act, 1970, particularly Section 3(d), which governs the patentability of new forms of known substances.
Detailed Factual Background: F. Hoffmann-La Roche Ltd. (Roche), a Swiss pharmaceutical giant, and OSI Pharmaceuticals Inc. (OSI), a New York-based company, were the plaintiffs in this case. Roche, known for its extensive research in pharmaceuticals and diagnostics, invests heavily in collaborative research, spending approximately 7 billion Swiss Francs annually. OSI, along with Pfizer Products Inc., jointly owned a patent for Erlotinib, a Human Epidermal Growth Factor Receptor (HER/EGFR) inhibitor used to treat advanced or metastatic non-small cell lung cancer (NSCLC). Erlotinib, sold under the Trademark Tarceva by Roche, was a tablet formulation approved by the U.S. FDA in 2004 and the European Union in 2005. In India, Roche introduced Tarceva in April 2006, following its registration by the Central Drug Standard Control Organization in December 2005.
The patent in question, Indian Patent No. 196774 (IN’774), was granted on February 23, 2007, to OSI and Pfizer for Erlotinib Hydrochloride, chemically named NOVEL [6,7-BIS(2-METHOXYETHOXY) QUINAZOLIN-4-YL]-(3-ETHYNYLPHENYL) AMINE HYDROCHLORIDE. The patent covered both the drug and its manufacturing process. Roche, under a licensing agreement with OSI dated January 8, 2001, was authorized to manufacture, market, and enforce intellectual property rights for Tarceva in India and other countries.
Cipla Ltd., a prominent Indian pharmaceutical company, was the defendant. Cipla announced plans to launch a generic version of Erlotinib, named Erlocip, in India and for export, as reported in the English daily Mint on January 11, 2008. This prompted Roche and OSI to file a suit against Cipla, alleging infringement of IN’774. Cipla, in turn, challenged the validity of the patent and argued that its product, Erlocip, was a polymorphic form (Polymorph B) of Erlotinib Hydrochloride, distinct from the patented compound, which was a mixture of Polymorphs A and B.
The dispute also involved a subsequent U.S. patent, US 6900221 (US’221), which described Polymorph B of Erlotinib Hydrochloride as a more stable form suitable for tablet formulation. In India, Roche’s application for a patent on Polymorph B (IN/PCT/2002/00507/DEL) was rejected under Section 3(d) of the Patents Act, which requires new forms of known substances to demonstrate enhanced therapeutic efficacy. This rejection became a focal point in the case, as Cipla argued that its product aligned with the unpatented Polymorph B, thus not infringing IN’774.
Detailed Procedural Background:The suit, CS(OS) No. 89/2008, was filed by Roche and OSI on January 15, 2008, in the Delhi High Court, seeking a permanent injunction to restrain Cipla from manufacturing, selling, or marketing Erlocip, along with claims for damages, rendition of accounts, and delivery up. Concurrently, the plaintiffs filed an interim injunction application under Order XXXIX, Rules 1 and 2 of the Code of Civil Procedure (CPC). On January 16, 2008, the court issued a notice on the interim application, and Cipla admitted to marketing Erlocip for three weeks. The interim application was argued over several hearings, but on March 19, 2008, the court dismissed the injunction request, directing Cipla to maintain accounts and provide an undertaking to pay damages if the suit was decreed.
Roche appealed the interim order in FAO(OS) No. 188/2008, but the Division Bench dismissed the appeal on April 24, 2009. A subsequent Special Leave Petition (SLP) before the Supreme Court (Civil No. 20111/2009) was dismissed on September 28, 2009, with a direction to expedite the trial without being influenced by the Division Bench’s observations.
Cipla filed a written statement and a counterclaim (C.C. No. 52/2008) on January 21, 2008, seeking revocation of IN’774 on grounds including lack of inventive step, obviousness, and non-patentability under Section 3(d). The plaintiffs filed a replication and a written statement to the counterclaim. On September 18, 2008, the court framed five issues, covering infringement, patent validity, the effect of US’221, and the plaintiffs’ entitlement to relief.
Evidence was recorded by a court-appointed commissioner, with both parties submitting affidavits from witnesses, including experts in pharmaceutical sciences and patent law. The plaintiffs presented affidavits from Mr. Shivprasad Laud (PW-1), Prof. Roger Griffin (PW-2), and Prof. Nick Thatcher (PW-3). Cipla’s witnesses included Mr. R. Gopalakrishnan (DW-1), Ms. Shashikala Kanathala (DW-2), Prof. Ashwini Nangia (DW-3), and Dr. Rajender Kumar Lohiya (DW-4). Extensive documentary evidence, including patent specifications, X-ray diffraction (XRD) data, and prior art references, was exhibited.
Cross-examination occurred between April 2009 and November 2010, with final arguments concluding on June 1, 2012. The court reserved judgment on June 1, 2012, and pronounced it on September 7, 2012.
Issues Involved in the Case:The court framed the following issues for adjudication:
Whether the manufacture, marketing, and sale of Erlocip by Cipla infringes the plaintiffs’ Indian Patent No. 196774? (Onus on Plaintiffs)? Whether Indian Patent No. 196774 is liable to be revoked on the grounds raised in Cipla’s written statement and counterclaim? Whether the plaintiffs are entitled to a permanent injunction as prayed for?Whether Cipla proves that the plaintiffs’ US Patent 6900221 indicates that the compound of claim No. 1 of IN’774 is a mixture of Polymorphs A and B, requiring separation for acceptable efficacy, and its effect on IN’774?
Plaintiffs’ Submissions:Roche and OSI argued that IN’774 was a valid patent covering Erlotinib Hydrochloride, and Cipla’s Erlocip infringed it by replicating the patented compound. They emphasized Tarceva’s significance as a life-saving drug, developed after substantial research, and its global approvals. The plaintiffs contended that Section 48 of the Patents Act grants exclusive rights to prevent unauthorized use, and their licensing agreement authorized Roche to enforce these rights.
On infringement, the plaintiffs asserted that Erlocip contained Erlotinib Hydrochloride, identical to the patented compound, and Cipla’s marketing violated their rights. They argued that IN’774’s claims were not limited to a specific polymorphic form, covering both Polymorphs A and B. The plaintiffs relied on the Catnic approach for claim construction, urging a purposive interpretation to include variants like Polymorph B. They cited clinical trial data from 1997 (Ex. PW1/X2) showing Erlotinib’s administration in tablet form, predating Polymorph B’s invention, to argue that the suit patent encompassed solid forms.
Regarding Polymorph B, the plaintiffs argued that it was irrelevant to therapeutic efficacy, as per Dr. Nick Thatcher’s testimony (PW-3), who stated that polymorphism does not affect patient outcomes. They contended that Section 3(d)’s explanation, which considers polymorphs as the same substance unless differing significantly in efficacy, supported their claim that Polymorph B fell within IN’774. The plaintiffs also referenced the Controller’s order in pre-grant opposition (December 15, 2008), which treated Polymorph B as the same substance as IN’774’s compound.
On patent validity, the plaintiffs argued that Cipla failed to prove obviousness or lack of inventive step. They disputed Cipla’s reliance on European Patent 0566226 (EP’226) as prior art, asserting that EP’226’s Markush structure covered millions of compounds, and Example 51 (a methyl-substituted quinazoline) was not the closest prior art. Instead, they highlighted EP’851’s compound (6,7-dimethoxy-4-(5-indolylamino)-quinazoline) with a superior IC50 value (1 nM) as the starting point for a skilled person. The plaintiffs argued that substituting methyl with ethynyl required inventive ingenuity, not mere workshop improvement, and Cipla provided no evidence of motivation to make this substitution.
The plaintiffs cited cases like Daiichi Sankyo v. Matrix Laboratories (670 F. Supp. 2d 359, Fed. Cir. 2010) to explain IC50 values and Glaverbel v. British Coal Corporation (1995 RPC 255) for claim construction, emphasizing that external documents like US’221 should not limit IN’774’s scope. They also relied on Pfizer Inc. v. Ranbaxy (457 F.3d 1284, Fed. Cir. 2006) and Abbott v. Dey (287 F.3d 1097, Fed. Cir. 2002) to argue that subsequent patents (US’221) do not negate earlier claims.
Defendant’s Submissions
Cipla challenged IN’774’s validity, seeking revocation under Section 64 of the Patents Act on grounds of obviousness, lack of inventive step, and non-patentability under Section 3(d). They argued that IN’774 was a derivative of known quinazoline compounds, particularly those disclosed in EP’226, filed by Zeneca Ltd. in 1993. Cipla contended that Example 51 of EP’226, a methyl-substituted quinazoline, was the closest prior art, and replacing methyl with ethynyl to arrive at Erlotinib was obvious to a person skilled in the art. They supported this with Prof. Ashwini Nangia’s affidavit (DW-3), which cited five patents (EP 0477700, US 4138590, US 5427766, US 5736534, WO 93/04047) demonstrating the interchangeability of methyl and ethynyl substituents.
Cipla argued that IN’774 lacked inventive step, as quinazoline derivatives were known for anti-cancer properties, and EP’226 provided sufficient motivation to experiment with substitutions. They referenced C.W. Thornber’s article on bio-isosterism (Ex. D-4) to support the predictability of such substitutions. Cipla also alleged that IN’774 was a derivative of Gefitinib (AstraZeneca’s drug), which was denied a patent in India for being in the public domain, accusing Roche of “evergreening” to extend patent monopolies.
On infringement, Cipla asserted that Erlocip was Polymorph B of Erlotinib Hydrochloride, distinct from IN’774’s mixture of Polymorphs A and B. They relied on US’221’s specification, which described Polymorph B as more stable and suitable for tablets, and Ms. Shashikala Kanathala’s XRD analysis (DW-2) confirming that Tarceva and Erlocip were Polymorph B. Since Roche’s application for Polymorph B (IN’507) was rejected, Cipla argued that manufacturing Polymorph B did not infringe IN’774. They cited the rejection order under Section 3(d) to argue that Polymorph B was not patentable without enhanced efficacy.
Cipla also raised public interest concerns, noting that Tarceva cost ₹4,800 per tablet, while Erlocip was priced at ₹1,600, making it more accessible for cancer patients. They argued that granting an injunction would harm public access to affordable drugs, citing the need to balance patent rights with public welfare.
Cipla referenced cases like Merck (applying Catnic principles to chemical compounds) and EPO Board of Appeal decisions (Ex. DW4/40-42) to support their interpretation of prior art and obviousness. They also relied on Sudhir Engineering Company vs. Nitco Roadways Ltd. (1995 (34) DRJ 86) to argue that public documents could be considered despite marking objections.
Detailed Discussion on Judgments Cited by Parties
Both parties relied on several precedents to support their arguments, each applied in specific contexts:
Daiichi Sankyo v. Matrix Laboratories (670 F. Supp. 2d 359, Fed. Cir. 2010): Cited by the plaintiffs to explain IC50 values as a measure of drug potency. The case clarified that lower IC50 values indicate higher potency, supporting the plaintiffs’ argument that EP’851’s compound (IC50 of 1 nM) was the most potent prior art, not EP’226’s Example 51.
Glaverbel v. British Coal Corporation (1995 RPC 255): Cited by the plaintiffs for claim construction principles, emphasizing that courts should interpret patent claims purposively without relying on external documents like US’221 to limit IN’774’s scope. The court in Glaverbel held that specifications should be read in context, which the plaintiffs argued supported a broad interpretation of IN’774.
Pfizer Inc. v. Ranbaxy (457 F.3d 1284, Fed. Cir. 2006): Cited by the plaintiffs to argue that subsequent patents (US’221) do not negate the validity or scope of earlier patents (IN’774). The case involved a challenge to a patent’s validity, where the court upheld the earlier patent’s claims, supporting Roche’s position.
Abbott v. Dey (287 F.3d 1097, Fed. Cir. 2002): Cited by the plaintiffs to reinforce that subsequent developments or patents do not limit the scope of an earlier patent. The case dealt with claim construction, aligning with the plaintiffs’ reliance on Catnic principles.
Catnic Components Ltd. v. Hill & Smith Ltd. ([1982] RPC 183): Referenced by both parties, particularly by the plaintiffs, for purposive claim construction. The House of Lords in Catnic established that patent claims should be interpreted to cover variants that achieve the same result in substantially the same way. The plaintiffs argued that Polymorph B was a variant within IN’774’s scope, while the court applied Catnic to conclude that IN’774 did not intend to cover Polymorph B.
Sudhir Engineering Company vs. Nitco Roadways Ltd. (1995 (34) DRJ 86): Cited by Cipla to address objections to document marking. The Delhi High Court held that public documents could be considered despite procedural objections, supporting Cipla’s reliance on patent specifications and Controller’s orders.
EPO Board of Appeal Decisions (T 424/86, T 133/84, T 145/84, dated 1984-1986): Cited by Cipla (Ex. DW4/40-42) to support their obviousness argument. These decisions discussed prior art and the predictability of chemical substitutions, aligning with Cipla’s claim that methyl-to-ethynyl substitution was obvious based on EP’226.
Patent Validity:The court examined Cipla’s counterclaim for revocation, which alleged obviousness, lack of inventive step, and non-patentability under Section 3(d). Cipla argued that IN’774 was obvious based on EP’226’s Example 51, supported by Prof. Nangia’s testimony and prior art patents demonstrating methyl-ethynyl interchangeability. The plaintiffs countered that EP’851’s compound was the closest prior art due to its superior IC50 value, and the substitution required inventive ingenuity.
The court found that Cipla failed to discharge its onus to prove obviousness. It noted that EP’226’s Markush structure encompassed millions of compounds, and selecting Example 51 as the starting point was not sufficiently motivated. The court accepted the plaintiffs’ argument that a skilled person would prioritize EP’851’s compound, and Cipla’s evidence did not establish why ethynyl substitution was an obvious “workshop result.” The court also rejected Cipla’s bio-isosterism argument, finding insufficient evidence of a predictable outcome. Thus, IN’774 was upheld as valid.
Infringement:The court’s analysis of infringement was central to the judgment. The plaintiffs argued that Erlocip contained Erlotinib Hydrochloride, infringing IN’774, and that Polymorph B was covered by the patent’s claims. Cipla contended that Erlocip was Polymorph B, distinct from IN’774’s mixture of Polymorphs A and B, and that the rejection of IN’507 permitted its manufacture.
Applying the Catnic approach, the court construed IN’774’s claims purposively to determine whether Polymorph B was intended to be covered. The court found that Roche’s separate patent applications for Polymorph B (US’221 and IN’507) indicated an intent to treat it as a distinct invention. US’221’s specification revealed that IN’774’s compound was a mixture of Polymorphs A and B, while Polymorph B was more stable and suitable for tablets. The court noted Roche’s failure to disclose US’221 or IN’507 in the suit, suggesting non-disclosure of Polymorph B’s relevance.
The court emphasized that the plaintiffs bore the onus to prove infringement by showing that Polymorph B was subsumed within IN’774’s claims. However, Roche provided no positive evidence to clarify the role of reactants in US’221 or demonstrate that Polymorph B’s properties were inconsequential. The court rejected the plaintiffs’ reliance on Dr. Thatcher’s testimony, finding it insufficient to address how reactants affected the compound’s stability or dosage.
Cipla’s evidence, particularly Ms. Shashikala’s XRD analysis (Ex. DW2/3-4), confirmed that both Tarceva and Erlocip were Polymorph B, aligning with US’221’s specifications. The court found that Roche’s non-denial of marketing Polymorph B and failure to clinically examine Erlocip undermined their case. The rejection of IN’507 under Section 3(d) further supported Cipla’s position, as Polymorph B was not patented in India.
The court also addressed the plaintiffs’ “fiction” argument under Section 3(d), which considers polymorphs as the same substance unless differing in efficacy. The court held that Section 3(d) creates a limited consideration for patentability, not a deeming fiction for infringement purposes. Extending the fiction beyond Section 3(d)’s scope was impermissible, and the plaintiffs’ failure to prove infringement independently was fatal.
Effect of US’221: The court found that Cipla discharged its onus to prove that US’221 indicated IN’774’s compound was a mixture of Polymorphs A and B, requiring conversion to Polymorph B for stability. The court relied on US’221’s specification and Cipla’s XRD evidence, noting that Roche’s separate patent applications for Polymorph B contradicted their claim that it was covered by IN’774. The court rejected the plaintiffs’ argument that US’221 only taught conversion, not separation, finding that the need for additional steps to achieve Polymorph B distinguished it from IN’774.
Injunction and Relief:Given the finding of non-infringement, the court held that Roche was not entitled to a permanent injunction or damages. The plaintiffs’ prayer focused on Tarceva, which was Polymorph B, not strictly IN’774’s compound, further weakening their case.
Public Interest: The court briefly addressed public interest, noting Cipla’s argument about Erlocip’s affordability (₹1,600 vs. ₹4,800 for Tarceva). However, since non-infringement was established, public interest did not influence the outcome.
Final Decision:The Delhi High Court dismissed both the suit (CS(OS) No. 89/2008) and Cipla’s counterclaim (C.C. No. 52/2008). The court held that Cipla’s Erlocip did not infringe IN’774, as it was Polymorph B, not covered by the patent’s claims. The court upheld IN’774’s validity, finding that Cipla failed to prove obviousness or grounds for revocation. No costs were awarded.
Law Settled in This Case:
This case clarified several aspects of Indian patent law: Purposive Claim Construction: The Catnic approach applies to chemical patents, requiring courts to interpret claims to determine the inventor’s intent, especially when variants like polymorphs are involved.
Section 3(d) Interpretation: The explanation to Section 3(d) considers polymorphs as the same substance for patentability unless they significantly differ in efficacy, but this does not create a deeming fiction for infringement purposes.
Burden of Proof in Infringement: Plaintiffs must provide positive evidence to prove that a defendant’s product falls within the patent’s claims, particularly when variants are involved. Non-disclosure of relevant patents (e.g., US’221) can weaken the plaintiff’s case.
Polymorphic Forms: Separate patent applications for polymorphic forms suggest an intent to treat them as distinct inventions, impacting infringement analysis.
Public Documents: Courts can consider public documents like patent specifications despite procedural objections, as per Sudhir Engineering.
Case Title: F. Hoffmann-La Roche Ltd. & Anr. Vs. Cipla Ltd.
Date of Order: September 7, 2012
Case No.: CS(OS) No. 89/2008
Name of Court: High Court of Delhi, New Delhi
Name of Judge: Hon’ble Mr. Justice Manmohan Singh
Disclaimer: The information shared here is intended to serve the public interest by offering insights and perspectives. However, readers are advised to exercise their own discretion when interpreting and applying this information. The content herein is subjective and may contain errors in perception, interpretation, and presentation.
Written By: Advocate Ajay Amitabh Suman, IP Adjutor [Patent and Trademark Attorney], High Court of Delhi
