A cancer drug combination patent that had spent close to a decade winding through the Indian Patent Office has been sent back for a fresh look, as the Delhi High Court found that the Controller who refused it never actually explained the reasoning behind three separate grounds of rejection. In Array Biopharma Inc. v. Deputy Controller of Patents and Designs, Hon’ble Justice Tushar Rao Gedela set aside the patent refusal covering a combination of a B-Raf inhibitor, an EGFR inhibitor and optionally a PI3K-alpha inhibitor, used together in treating proliferative diseases such as colorectal cancer, and directed the Patent Office to reconsider the application from scratch within six months. The ruling is worth reading because it does not simply overturn one Controller’s assessment of the science, it sets out in some detail what a reasoned refusal on inventive step, on Section 3(d) and on Section 3(i) of the Patents Act, 1970 actually needs to look like, and Indian pharmaceutical patent applicants have been waiting a long time for that kind of clarity.
A long road through the Patent Office
Array Biopharma’s application claimed priority from a United States filing made in August 2012 and entered the Indian national phase in January 2015 after an international PCT filing. The application was published in mid-2015 and examination was formally requested the following year. A First Examination Report followed in 2018, and after several rounds of responses, hearings and adjournments under Section 15 of the Patents Act, the Deputy Controller finally refused the application in June 2023, nearly eleven years after the original priority date. The refusal rested on three grounds at once, lack of inventive step under Section 2(1)(ja), non-patentability as a mere derivative of a known substance under Section 3(d), and non-patentability as a method of medical treatment under Section 3(i), with a fourth objection on claim clarity thrown in for good measure under Section 10(4)(c) and Section 10(5).
Array’s case throughout was that the claims were straightforward product claims for a pharmaceutical combination entity, supported by clinical data showing genuine synergistic effect between the constituent agents, including measurable tumour regression in specific test groups, and that the Controller had misread both the science and the statute in refusing the application on all three substantive grounds.
The inventive step finding falls apart on inspection
The Court’s treatment of the inventive step objection followed a document by document review of all four prior art references the Controller had relied on. The first disclosed B-Raf inhibitors generally and taught combinations with MEK inhibitors, but said nothing about pairing them with EGFR or PI3K-alpha inhibitors. The second concerned diagnostic methods for detecting certain mutations and referred only in passing to EGFR signalling inhibitors used alongside unspecified RAF inhibitors, without naming the specific B-Raf inhibitor at issue in Array’s claims. The third disclosed a different B-Raf inhibitor combined with an unspecified PI3K inhibitor but made no mention of EGFR inhibitors at all. The fourth was directed at a class of PI3K inhibitors as standalone chemical entities and discussed EGFR only in the context of laboratory models, again without disclosing the specific combination Array had claimed.
Taken individually or read together, the Court found that none of these four documents disclosed or even gestured toward the specific dual or triple combination at the heart of Array’s application. More significantly, the Court held that the Controller’s order never engaged with the question a proper inventive step analysis actually requires, namely why a person skilled in the art, faced with this scattered prior art, would have been motivated to select these particular compounds out of a vast field of known options and combine them in this particular way. Without identifying a closest prior art document, articulating the specific technical problem the invention solved, or explaining the motivation to combine, the Controller’s finding on inventive step could not stand as a reasoned decision, whatever the ultimate scientific merits might turn out to be on a fresh look.
Section 3(d) needs a named starting compound
On the Section 3(d) objection, the Court’s reasoning was more compact but no less pointed. Section 3(d) bars patents for the mere discovery of a new form of a known substance, or a derivative of a known substance, unless it demonstrates significantly enhanced efficacy over what came before, and the provision only makes sense if the decision maker can point to the specific known compound the claimed invention is supposedly derived from. The Deputy Controller’s order never did this. It invoked Section 3(d) without ever naming which known compound Array’s combination was allegedly a derivative of, and the Court held that this omission was fatal on its own terms. Beyond the missing identification, the Court also reasoned as a matter of principle that a combination of two or more independent active pharmaceutical agents, each with its own distinct chemical structure and mechanism of action, cannot logically be treated as a derivative of itself or of its own constituent parts, drawing support from the same reasoning applied in the Patent Office’s own earlier Topotarget ruling. A combination claim of this kind, in other words, sits outside the conceptual reach of Section 3(d) almost by definition, and the Controller’s order gave the Court nothing to suggest otherwise.
A product claim does not become a method just because it mentions administration
The Section 3(i) objection drew the most detailed reasoning in the judgment, and it addresses a recurring point of confusion in pharmaceutical patent prosecution. Section 3(i) excludes processes for the medicinal or therapeutic treatment of human beings from patentability, but it does not touch products, and the Controller had treated Array’s claims as falling within this exclusion because the specification described simultaneous, separate or sequential administration of the active agents and set out clinical dosing protocols. The Court disagreed with this characterisation. It held that describing how the constituent parts of a combination product might be administered is simply a way of defining the scope of the product claim itself, a description of how the invented combination works in practice, and does not convert a product claim into a disguised method of treatment. Clinical examples and dosing schedules in a specification exist to demonstrate that an invention is workable and industrially applicable, a separate statutory requirement under Section 10, and they do not expand or redefine what the claim itself actually covers. Since Claim 1 was drafted as a genuine product claim directed to a pharmaceutical combination and not to a treatment protocol, Section 3(i) had no application to it at all, and the Court found the Controller’s contrary conclusion unsustainable.
What gets settled and what does not
Taken together, the three findings leave the Patent Office with a fairly precise template for handling this application on remand, and arguably for handling similar combination patent applications more generally. A refusal on inventive step needs to identify the closest prior art and explain the motivation to combine rather than simply listing documents that happen to touch on related subject matter. A refusal under Section 3(d) needs to name the known compound the claim is allegedly derived from, since a combination claim built from genuinely independent active ingredients cannot be waved away as a derivative without that anchor. And a refusal under Section 3(i) cannot rest on administration language or dosing examples appearing somewhere in the specification if the claim itself is drafted as a product. None of this means Array’s application is now guaranteed to succeed. The Court was careful to remand the matter for a genuinely fresh, de novo consideration, directing the Controller to decide the application within six months after giving Array a proper hearing, and specifying that this fresh determination should not be influenced by anything the judgment itself observed along the way. The substantive question of whether this particular combination actually clears the inventive step bar, once properly analysed, remains open.
Conclusion
Cases like this one matter more than their individual outcome suggests, because they function as an accountability check on how the Patent Office writes its refusal orders, not just what conclusions those orders reach. I think the most useful part of this judgment for practitioners is the reminder that Sections 3(d) and 3(i) are not general purpose tools a Controller can invoke whenever a pharmaceutical claim looks complicated, they have specific analytical requirements attached to them, a named base compound in one case and a genuine method of treatment in the other, and an order that skips those requirements is vulnerable on appeal regardless of how the underlying science might eventually shake out. At the same time I would caution against reading this as a ruling that favours patentees generally in the combination therapy space, since the Court went out of its way to avoid prejudging the merits and sent the file back precisely so the Patent Office could do the analysis properly the second time. For applicants who have watched combination patents get refused on thinly reasoned Section 3(d) and 3(i) objections over the years, this judgment gives them a concrete standard to hold future refusal orders against, and that alone makes it worth tracking closely as the Controller works through the remanded application.
Array Biopharma Inc. v. Deputy Controller of Patents and Designs, C.A.(COMM.IPD-PAT) 37/2023, Delhi High Court, decided 23rd July 2026.
