Novartis AG v. Union of India: the enhanced-therapeutic-efficacy gloss on Section 3(d)

Novartis AG v. Union of India is the case that fixed the meaning of Section 3(d) of the Patents Act 1970. The provision had been inserted by the Patents (Amendment) Act 2005, the legislation by which India brought its patent regime into compliance with the Agreement on Trade-Related Aspects of Intellectual Property Rights and re-opened product patents on pharmaceuticals after a quarter-century of process-only patenting under the 1970 Act. The new Section 3(d), however, did more than transpose TRIPS minima. It carved out a category of claim that, in the legislative judgment, called for a heightened patentability filter: incremental modifications of known pharmaceutical substances. The Bench of Aftab Alam, J. and Ranjana Prakash Desai, J., on 1 April 2013, supplied the gloss that has since defined how that filter operates.

The judgment is reported at (2013) 6 SCC 1. It runs to more than two hundred pages in the SCC print volume. The reasoning operates on three levels: a legislative-history level in which the Bench traces the journey of Indian patent law from the Ayyangar Committee Report of 1959 through the 1970 Act and the TRIPS amendments of 1999, 2002 and 2005; a textual level in which the Bench works through Section 3(d), its Explanation, and its interaction with Section 2(1)(j) (invention) and Section 2(1)(ja) (inventive step); and a substantive level in which the Bench applies the construction to the beta-crystalline form of Imatinib Mesylate.

A drafting note before the body. The molecule in suit is marketed in India and most of the world as Glivec; in the United States the trade name is Gleevec. Novartis used the Indian spelling in its Indian patent filings and in the litigation. The judgment uses "Glivec" throughout; this article follows the same convention.

The statutory architecture

Three provisions hold the weight of the decision.

Section 2(1)(j) defines "invention" as "a new product or process involving an inventive step and capable of industrial application." Section 2(1)(ja) defines "inventive step" as a feature of the invention that involves technical advance as compared to existing knowledge or has economic significance or both, and that makes the invention not obvious to a person skilled in the art. These are the threshold conditions of patentability under the 1970 Act.

Section 3(d), as recast by the 2005 amendment, sits in the chapter on inventions not patentable. The text reads:

the mere discovery of a new form of a known substance which does not result in the enhancement of the known efficacy of that substance or the mere discovery of any new property or new use for a known substance or of the mere use of a known process, machine or apparatus unless such known process results in a new product or employs at least one new reactant.

The Explanation, which the Bench treated as the operative interpretive key, reads:

for the purposes of this clause, salts, esters, ethers, polymorphs, metabolites, pure form, particle size, isomers, mixtures of isomers, complexes, combinations and other derivatives of known substance shall be considered to be the same substance, unless they differ significantly in properties with regard to efficacy.

The architecture is layered. Section 3(d) operates in addition to Sections 2(1)(j) and 2(1)(ja), not in substitution for them. A claim must first cross the inventive-step threshold; if it concerns a new form of a known substance, it must then independently cross the Section 3(d) threshold. The Bench's central interpretive move is the gloss on "efficacy" — the word the Explanation makes the criterion for treating a derivative as a separate substance.

The factual matrix

The product in suit is the beta-crystalline form of Imatinib Mesylate. The base molecule, Imatinib, had been disclosed in Novartis's 1993 patent (the "Zimmermann patent"), filed when India was not granting pharmaceutical product patents but disclosed in many jurisdictions that were. Imatinib is a tyrosine-kinase inhibitor used in the treatment of chronic myeloid leukaemia. The mesylate salt of Imatinib was disclosed in the same family; the beta-crystalline form — a specific polymorph of that salt with reported improvements in flow, hygroscopicity and bioavailability — was the subject of a separate patent application filed in 1998 in the United States and lodged in India in 1998 in the mailbox arrangement under Section 5(2) and the TRIPS-transitional provisions.

The Indian application emerged from the mailbox after the 2005 amendments and was examined by the Patent Office at Chennai. The Controller refused the application on 25 January 2006 on three grounds — anticipation by the Zimmermann disclosure, lack of inventive step, and failure of Section 3(d). The five Indian generic manufacturers (Cipla, Hetero, Ranbaxy, Natco, Sun Pharma) and the patient-cum-civil-society opponent (the Cancer Patients Aid Association) had filed pre-grant oppositions; the Controller's order incorporated their submissions.

Novartis filed a writ petition in the Madras High Court challenging the constitutional validity of Section 3(d) — arguing vagueness under Article 14 and TRIPS-inconsistency. A Division Bench of the Madras High Court rejected both challenges on 6 August 2007. The vagueness argument was held untenable; the TRIPS-compatibility question was held not within the writ court's jurisdiction.

The merits appeal of the Controller's refusal moved to the newly-constituted Intellectual Property Appellate Board. The IPAB, by an order of 26 June 2009 authored by Prabha Sridevan as Chairperson and S. Chandrasekaran as Technical Member, set aside the anticipation and inventive-step findings but affirmed refusal under Section 3(d). The IPAB held the beta-crystalline form was a new form of the known substance Imatinib Mesylate and had not been shown to differ significantly in properties with regard to efficacy.

The IPAB order was carried directly to the Supreme Court under a special-leave petition. The procedural shortcut — leapfrogging the Madras High Court on the merits — was approved at the leave stage in view of the public-importance dimension of the Section 3(d) construction.

The Court's reasoning

The judgment is structured as a sustained interpretive exercise in three parts. The Bench first reads Section 3(d) in the legislative-history context of the 2005 amendments. It then constructs "efficacy" as "therapeutic efficacy". It then applies that construction to the evidentiary record on the beta-crystalline form.

Legislative history — the anti-evergreening rationale

Alam, J. devoted the opening third of the judgment to the legislative history of Indian patent law. The Ayyangar Committee Report of 1959 had recommended a process-only patent regime for pharmaceuticals on the ground that product patents in India would impede access to medicines. The 1970 Act gave effect to that policy choice. The TRIPS Agreement of 1994 obliged India to re-open product patents by 1 January 2005, the end of the transitional period under Article 65.

The 2005 amendments accordingly re-opened product patents but added Section 3(d) as a screen specifically targeted at incremental pharmaceutical claims. The Bench drew on the Parliamentary debate, on the Mashelkar Committee Report and on the contemporaneous Standing Committee proceedings to establish a single legislative purpose: to prevent the practice known as "evergreening", in which a patent holder secures successive patents on minor modifications of a known molecule and thereby extends the effective term of market exclusivity beyond the original twenty-year period.

The Bench was explicit that Section 3(d) is not a TRIPS-implementing provision in the narrow sense — it is an India-specific filter that operates within the space that TRIPS leaves open. Article 27 of TRIPS lays down minimum patentability criteria — novelty, inventive step, industrial application. TRIPS, the Bench held, does not prohibit higher national standards; Article 1 expressly contemplates that "Members may, but shall not be obliged to, implement in their law more extensive protection than is required by this Agreement." The corollary — and this is the Bench's most-cited dictum on TRIPS — is that TRIPS operates as a floor, not a ceiling.

The gloss — "efficacy" means "therapeutic efficacy"

Section 3(d)'s text uses the word "efficacy"; the Explanation requires that a polymorph or other derivative differ "significantly in properties with regard to efficacy" to be treated as a separate substance. The decisive interpretive question was whether "efficacy" carries any specific clinical meaning, or whether it is open-textured enough to capture any property improvement that ultimately bears on the drug's performance.

Novartis's argument was the latter. Improved bioavailability — the proportion of an administered dose that reaches systemic circulation in unchanged form — translates, ultimately, to improved therapeutic effect. Improved flow and reduced hygroscopicity facilitate manufacturing and formulation, and thereby support clinical reliability. On Novartis's reading, the Explanation captured any property that, downstream, supported the therapeutic objective.

The Bench rejected that reading. Alam, J. held that "efficacy" in Section 3(d) means therapeutic efficacy — the capacity of the substance to produce its intended healing effect in the human body. The construction is anchored in three considerations: the pharmacological vocabulary in which "efficacy" carries the clinical-effect meaning; the structural placement of Section 3(d) as an anti-evergreening filter, which would be defeated by an open-textured reading; and the Explanation's enumeration — salts, esters, ethers, polymorphs, metabolites, pure form, particle size, isomers — which catalogues precisely the kinds of variations that pharmaceutical chemistry routinely generates without therapeutic gain.

The corollary is the physico-chemical / therapeutic distinction. Properties such as bioavailability, hygroscopicity and flow are physico-chemical attributes of the substance; they may, in some cases, translate to a clinical advantage, but the translation is not assumed. The patent applicant must adduce evidence of the therapeutic consequence — and that evidence must support a "significant" difference. A reported improvement in bioavailability, standing alone, does not cross the threshold; what is required is evidence that the bioavailability improvement produces a clinically meaningful enhancement of the substance's healing effect.

The application — beta-crystalline Imatinib Mesylate fails

Applied to the record, the construction was dispositive. Novartis had tendered animal-study data showing a 30 per cent improvement in oral bioavailability of the beta-crystalline form over Imatinib base. It had tendered data on improved hygroscopicity and flow that supported manufacturing reliability. It had not tendered clinical data demonstrating a therapeutic advantage of the beta-crystalline form over the salt form already known.

The Bench held the record insufficient. Bioavailability improvements are physico-chemical; the patent applicant had not bridged the inferential gap between bioavailability and clinical effect. Improvements in flow and hygroscopicity are even further from the therapeutic locus — they support manufacturing, not clinical performance. The beta-crystalline form was, accordingly, the same substance as Imatinib Mesylate under Section 3(d)'s Explanation, and the claim failed the heightened patentability filter.

The Bench paused on the anticipation and inventive-step findings only briefly. The IPAB had found the beta-crystalline form crossed the Section 2(1)(j)/2(1)(ja) thresholds — novelty and inventive step. The Supreme Court did not need to revisit those findings to decide the appeal; Section 3(d) operated as an independent and sufficient ground of refusal. The judgment leaves open whether, on a fuller evidentiary record, the beta-crystalline form would have crossed the Section 3(d) threshold — the holding is that, on the record before the Bench, it did not.

The doctrinal contribution

Novartis contributes to Indian patent law on four axes. The "enhanced therapeutic efficacy" gloss fixes the meaning of "efficacy" and supplies a workable physico-chemical versus therapeutic distinction that the Patent Office and the IPAB have since applied to a steady flow of incremental-pharma applications. The anti-evergreening rationale locates Section 3(d) in the public-health architecture of the 2005 amendments — it is a substantive filter with a specifically articulated purpose, applied through the Madras High Court's Bayer Pharma v. Union of India (2014), the Delhi High Court's AstraZeneca AB v. Intas Pharmaceuticals (2021) and Boehringer Ingelheim v. Vee Excel Drugs (2023) lines. The strict reading of the Explanation treats its enumeration as substantive — the default is identity; the burden is on the applicant to displace it. And TRIPS as floor, not ceiling — anchored in Article 1 — supplies the standard rejoinder to the TRIPS-inconsistency argument and has informed every subsequent Indian engagement with the TRIPS compatibility of Section 3(d) and adjacent provisions.

What the judgment did not decide

The Bench did not lay down the quantum of therapeutic improvement required by Section 3(d) — the text speaks of a "significant" difference; subsequent practice has worked out the quantum case-by-case on expert evidence. It did not address the interaction with divisional or selection patents where a later patent claims a narrow embodiment from an earlier broader claim. It did not address biological substances — the Explanation's enumeration is chemistry-coded, and the application to biologics has emerged as a live question in subsequent practice. And it did not engage with compulsory licensing as a complementary instrument — the Sections 84–92 dimension was developed a year later in Bayer Corporation v. Union of India, to which we cross-refer.

The doctrinal arc

The Novartis line has had three phases. The first, in the years immediately after the judgment, was a phase of consolidation at the Patent Office and the IPAB. The "enhanced therapeutic efficacy" gloss was applied to a steady stream of incremental-pharma applications; the evidentiary norm — pharmacokinetic data supplemented by clinical-effect data, with explicit comparator — became the working baseline.

The second phase, from the mid-2010s into the early 2020s, was a phase of High Court engagement. The Madras High Court in Bayer Pharma (2014), the Delhi High Court in a series of decisions involving AstraZeneca, Bristol-Myers Squibb and Boehringer Ingelheim, and the Calcutta High Court in pharmaceutical infringement actions worked through the Section 3(d) threshold in post-grant contexts. The doctrinal architecture remained stable; the application became more granular.

The third phase is one of extension into adjacent areas: agrochemical hydrates (the Delhi High Court's recent disturbance of routine crystalline-form refusal in Syngenta Crop Protection AG v. Controller of Patents, May 2026), biologics and biosimilars, and software-related claims (where Section 3(k) runs in parallel). Novartis remains the anchor across these extensions.

What practitioners take

For the patent applicant. The drafting and the evidentiary record must be organised around the therapeutic axis. Property data on bioavailability, hygroscopicity, particle size or flow — necessary as it is for the chemistry-and-manufacturing narrative — does not, by itself, carry the Section 3(d) burden. The specification should include clinical or in-vivo data demonstrating a therapeutic consequence of the property improvement, with an explicit comparator and an explicit therapeutic metric.

For the opponent. The pre-grant opposition under Section 25(1) and the post-grant opposition under Section 25(2) are the principal procedural vehicles. The opposition should be framed around the Explanation's identity default — the polymorph or salt is the same substance unless the applicant shows otherwise — and should focus on the gap between property data and therapeutic data. F. Hoffmann-La Roche v. Cipla has since clarified that Section 3(d) operates through opposition and revocation channels, not as a freestanding infringement defence.

For portfolio strategy. Polymorph and salt claims should be filed only where the therapeutic-data record can be assembled; formulation and use-of-known-substance claims should be designed around the Section 3(d) exclusion rather than against it. A portfolio strategy that assumes routine patentability of incremental modifications will not survive Indian examination.

For TRIPS-and-policy. The "floor, not ceiling" reading of Article 1 is settled. International-trade pressure points on Section 3(d) have not displaced the construction.

The doctrinal residue

Novartis is the foundational Indian authority on incremental-pharma patentability. The "enhanced therapeutic efficacy" gloss, the physico-chemical / therapeutic distinction, the strict reading of the Explanation and the TRIPS-as-floor architecture have together defined the working framework for the Patent Office, the appellate forum (first the IPAB, now the High Courts after the IPAB's abolition by the Tribunals Reforms Act 2021) and the litigating bar. The decision sits at the head of the post-TRIPS Indian pharmaceutical patent canon — completed at the post-grant end by Bayer Corporation v. Union of India on compulsory licensing and at the infringement-defence end by F. Hoffmann-La Roche v. Cipla on the patent-eligibility-versus-defence question. Read together, the three judgments supply the architecture of Indian pharmaceutical patent law in the TRIPS era.

Related editorial pieces

• Bayer Corporation v. Union of India: India's first compulsory licence
• F. Hoffmann-La Roche v. Cipla: the s.3(d) eligibility / defence distinction and the Tarceva framework
• Delhi HC in CMYK Printech: the trademark-copyright interface in printing-and-publishing trade dress
• Delhi HC in Hindware v. Google Ads: keyword-bidding, trademark infringement and intermediary liability