Section 3(i) of the Patents Act has long left the boundaries of diagnostic method exclusions uncertain, despite repeated judicial engagement. Two recent Delhi High Court decisions now offer the clearest articulation yet, bringing much-needed coherence to this evolving area of law. Shubham Thakare and Arpanjot Kaur explain these orders, highlighting how they not only clarify the legal position but also sharpen the policy tensions underlying the provision. Shubham is a third-year B.A., LL.B. (Hons.) student at the National Law School of India University, Bengaluru, with an interest in copyright and trademark law. Arpanjot is a third-year B.A., LL.B. (Hons.) student at NLSIU. She has a keen interest in patent law and competition law, and has explored these areas through her coursework as well as mooting during law school.
Diagnosing the Diagnostic Exclusion: Have Geron and Hirotsu Resolved the Delhi High Court’s Troubles?
By Shubham Thakare and Arpanjot Kaur
Section 3(i) of the Patents Act excludes medical processes, including diagnostic methods, from patent protection. Yet, the scope of this exclusion remained uncertain for a long time. It was not always clear how far it extends, particularly in relation to diagnostic methods operating at different stages of disease detection and treatment. The Madras High Court’s decision in Chinese University (2023) clarified parts of this, but three subsequent Delhi High Court decisions, namely Natera, Sequenom, and EMD Millipore (all decided in 2025), reintroduced some uncertainties.
However, two recent decisions of the Delhi High Court, delivered within a span of two months, now provide the clearest account so far of what Section 3(i) covers and address these uncertainties. (Thanks to this Linkedin post by Mr. Sandeep Rathod and Dr. Lunalisa Potsangbam’s comment on it for highlighting the nuanced take on Section 3(i) in these cases.) In Hirotsu Bio Science Inc., decided by Justice Karia in January, the Court refused a patent for a method that detects cancer at a pre-clinical stage based on the response of a microscopic organism to urine samples. In Geron Corporation, decided by Justice Arora in March, the Court declined a patent at the other end of the spectrum, where cancer has already been diagnosed, and the claimed method measures a biological marker to guide treatment.
Against this backdrop, this post argues that these decisions bring much-needed clarity after the uncertainty that followed the Chinese University. It also suggests, at a broader policy level, that the continued exclusion of precision medicine from patent protection in India is not an issue that courts can satisfactorily resolve through interpretation alone. Any meaningful change will have to come from the legislature.
The Two Inventions
The scientific character of the inventions at stake could hardly be more different, which is precisely what makes reading them together instructive.
Geron Corporation’s application covered an in vitro method for selecting cancer patients who would benefit from imetelstat, a telomerase inhibitor. Here, the method works where the patient is already diagnosed with, or suspected of having, cancer. It measures the relative length of telomeric nucleic acids in cancer cells from a biological sample, compares the result against a known standard range, and selects patients at or below the 50th percentile for therapy. Its clinical position is post-diagnosis since it does not really identify cancer but stratifies already-identified cancer patients for targeted treatment. On the other hand, Hirotsu Bio Science’s application was for an in vitro cancer detection method using the olfactory behaviour of nematodes. The nematode C. elegans, when exposed to urine samples from cancer patients, exhibits an attraction response rather than the avoidance behaviour it displays towards healthy subjects. The method is fully automated, requires no physician involvement at any stage, and, according to the complete specification, detects early-stage cancer with sensitivity and specificity approaching 100%. Its clinical position is pre-diagnosis in that it operates upstream of any clinical encounter and produces a binary cancer-or-no-cancer signal on a biological sample from an asymptomatic or undiagnosed subject.
From Uncertainty to Uncertainty Again
In 2022, Aparajita Lath wrote on this blog (See here and here) that when it came to diagnostic method patents under Section 3(i), “the only thing that is clear is that there are no clear rules.” That fairly captured the state of the law at the time. The Patent Office had granted patents to some in vitro diagnostic processes (tests carried out outside the human body, such as analysing blood, urine, or tissue samples) and refused others without any consistent principle to explain the difference. Moreover, the 2019 Manual had dropped the earlier in vivo limitation (under which only methods performed on the body were excluded) without explanation or replacement.
The Madras High Court’s 2023 decision in Chinese University was the first serious attempt at clarity. The Court began with the statutory text, noting that the European Patent Convention limits its exclusion to diagnostic methods “practised on the human or animal body,” while Section 3(i) contains no such language. India had, in fact, proposed this limitation during the TRIPS negotiations, and the proposal was rejected. The Court treated the absence of the phrase in the draft text of the TRIPS as deliberate and held that in vitro diagnostic methods could not be carved out of the exclusion. The Court also declined to follow the EPC’s four-step diagnostic framework, under which a method is only “diagnostic” if it includes the final step of arriving at a clinical conclusion. That approach, it found, could be too easily sidestepped by drafting claims that omit the final step while retaining the substance of the method. It similarly rejected the idea that screening and diagnosis can be neatly separated, since it believed that a process described as a screening method still falls within Section 3(i) if it performs the same function as a diagnostic method.
The Court framed the operative question as whether the claimed process identifies a disease or condition in a manner that can guide treatment, assessed from the perspective of a person skilled in the art, including a medical practitioner. It clarified that the process need not yield a definitive diagnosis. It is sufficient if its result can be used to make or shape a treatment decision, in which case it falls within Section 3(i).
This clarity was muddied by the Delhi High Court’s trilogy in Natera, Sequenom, and EMD Millipore, decided in 2025. As Kartik Sharma has argued previously on this blog, the outcomes were broadly defensible, but the reasoning raised concerns that matter for future application. In Sequenom, the Court’s engagement with the Chinese University case was confined to a narrow point about the distinction between screening and diagnosis. The Madras High Court’s more fundamental analysis, particularly its rejection of the EPO framework as a guide to Indian law, was left unaddressed. In EMD Millipore, the Court drew on concepts such as invasiveness, health risk, and professional judgment from EPO guidance on surgical methods rather than diagnostic methods. These concepts find no footing in the text of Section 3(i) and originate in a statutory framework that uses language absent from the Indian Act. The Court also observed, correctly, that Section 3(i) does not distinguish between in vitro and in vivo processes. Yet, in the same discussion, it appeared to treat in vitro methods as more likely candidates for protection. The four-step European diagnostic framework was mentioned but ultimately set aside without explanation, leaving the question of its applicability to future cases uncertain.
How Hirotsu and Geron Settle the Law
There was a concern that, in light of the aforementioned trilogy, the Delhi High Court might take a similar path and further complicate the application of Section 3(i). That concern is, at least in part, addressed by Hirotsu and Geron. Both decisions largely move away from the trilogy’s analytical vocabulary and return to the capability-based approach articulated in the Chinese University case. In doing so, they respond, directly or indirectly, to the problems Sharma had identified.
Justice Arora’s judgment in Geron is the more structurally significant of the two. At paragraph 25, she sets out twelve principles (reproduced below) that together form the most consolidated statement of Section 3(i) doctrine in Indian patent law to date. A future court could well treat this as a working restatement of the law on diagnostic exclusions, without needing to revisit the earlier cases in detail.
What is notable is that the framework of invasiveness and professional judgment, which had complicated the trilogy, finds no place here. Those concepts were borrowed from EPO guidance on surgical methods, which is a body of law that operates under different statutory language and has no grounding in the text of Section 3(i). By contrast, the capability test, reflected in Principles II and V of Geron, flows directly from what the first prong of Section 3(i) actually asks: whether a claimed process is diagnostic for the treatment of human beings. A method satisfies this if it is inherently capable of contributing to a treatment decision, assessed from the perspective of a person skilled in the art, including a medical practitioner. The method need not, on its own, conclusively establish the existence of a medical condition. It is enough that its output feeds into clinical decision-making. That is the inquiry Chinese University had already pointed towards. Geron now makes it explicit and consolidates it into a workable set of principles.
Hirotsu confirms the same conclusions on different facts and adds one clarification that the twelve principles do not expressly address. Justice Karia held that it is immaterial who performs the method. Consequently, a diagnostic process does not escape Section 3(i) simply because no physician is involved in operating it. This closes the circumvention route that the EMD’s professional judgment language had inadvertently left open (namely, that an applicant could argue an automated or technician-operated method falls outside the exclusion precisely because no professional judgement is exercised in running it). Moreover, Hirotsu expressly affirms that Section 3(i) draws no distinction between in vivo and in vitro processes, and does not exhibit the instability identified in EMD Millipore. This position is also reflected in Principle III of Geron.
Hirotsu and Geron thus together clarify that under Section 3(i), a diagnostic method is patent-ineligible across the full clinical spectrum, from pre-diagnosis detection of a condition to post-diagnosis selection of treatment, where:
- The claimed process is inherently capable of contributing to a treatment decision, even if it does not establish a definitive medical condition;
- This is assessed from the perspective of a person skilled in the art, including a medical practitioner, reading the claims in the context of the complete specification;
- It does not matter whether the method is performed in vivo or in vitro;
- It does not matter whether the method is labelled a screening test or a diagnostic test; and
- It does not matter who performs the method, meaning an automated or technician-operated process is not exempt simply because no physician is involved in running it.
The Problem still Open
While Hirotsu and Geron go a considerable distance in settling the uncertainties under Section 3(i), they also bring forth a broader policy concern around the exclusion of precision medicines. These depend on the ability to identify which patients will respond to which therapies. The processes that enable that identification, whether through telomere length measurement, nematode chemotaxis, or other biomarker-based methods, are, under the current framework, likely to fall within the exclusion from patent protection.
The Madras High Court in the Chinese University case had already taken notice of this problem. It suggested that there may be a case for limiting “diagnostic” in Section 3(i) to in vivo processes, balanced by a regime of compulsory licensing, but declined to pursue this as it fell within Parliament’s domain. Mani and Sharma observe in their piece that the Court recognised the need to balance innovation incentives against public health without attempting to legislate that balance itself. That observation has remained unaddressed for over two years. Geron and Hirotsu show us clearly that the issue extends across the entire diagnostic continuum, from upstream pre-clinical detection to downstream treatment-selection methods.
The courts, for their part, have brought a degree of coherence to the section. The question of whether the present balance between encouraging innovation and preserving access to medical practice is appropriate for a jurisdiction that is increasingly significant in pharmaceutical and diagnostic research is something that courts now cannot resolve. There is a compelling case for legislative engagement on this front.
