[Thanks to Swaraj Barooah sharing an in-progress draft from a project he is working on examining cancer litigation in India. The questions in there led me to write this post.]
Can a court say a company has no right to confidentiality in its regulatory data, and then hand that data to a rival’s expert in a sealed cover? The Delhi High Court decided it can. On 3 November 2026, the Supreme Court is scheduled to take up two Special Leave Petitions (SLPs) for final disposal filed by Zydus (formerly Cadila Healthcare) and Hetero, against a Delhi High Court order of 18 February 2025. The Feb 2025 order was pronounced with regards to two suits:
- In the first one, (CS (COMM) 540/2016), Roche challenged DCGI’s approval of Hetero’s version of bevacizumab, called Cizumab/Bevaas approved on 13 May 2016 and launched in June 2016.
- In the second, (CS (COMM) 1119/2016), Roche challenged DCGI’s approval of Cadila’s version of trastuzumab, called Vivitra, approved on 28 October 2015 and launched in December 2015.
Roche says the approvals were wrongly granted because biosimilarity was not shown through the required tests, and seeks declarations of invalidity and injunctions against calling the drugs biosimilar. The defendants’ contention is that the plaintiff is seeking production of the documents which are not relevant to the claim of passing off made out in the plaints. The High Court directed both companies (Hetero and Cadila) to file their regulatory files in sealed cover, for examination by Roche’s lawyers and an expert before a confidentiality club. The Supreme Court stayed the order in Zydus’s petition on 7 April 2025. Hetero obtained no stay.
I must note here that the reasoning followed in the order is scattered throughout and is built on contradictory terms. An order mandating discovery makes sense when a party shows that it has a case which needs determination via a corpus of documents that rests with the opposing party. Roche claims passing off, specifically extended passing off actions against Cadila and Hetero, but does so on the basis of DCGI’s apparent wrongful approval. It is not clear how the basis for any of these causes of actions are satisfied. Regardless, the said order assumes that Roche showed existence of a legitimate cause of action sufficiently and directs discovery.
Over the last 10+ years, different bloggers have followed Roche’s litigation over trastuzumab biosimilars since it first sued Biocon and Mylan in 2014, when Madhulika wrote on Justice Manmohan Singh’s interim injunction. Much discussion has questioned whether such suits should exist at all, argued by Prashant in 2016 and in 2019. Adarsh tracked the Supreme Court’s December 2019 order in the Reliance matter and the later orders on Rule 122DC. Tejaswini analysed the 2023 order that allowed the present suits to proceed, and anticipated that a disclosure application and a confidentiality club might follow.
Interestingly, the Feb 2025 order, carrying impact, has had no significant analytical coverage (including in mainstream news) which I could find! With the matter now before the SC at the final disposal stage, this post asks a narrower question: once the suits were allowed to proceed, how did the DHC reason its way to this discovery order? The order rests on precedent and a strong case on relevance, but it leaves four things unexplained: (1) why it denies the defendants any confidentiality and then builds a confidentiality club that arms only Roche with an expert; (2) why the regulator is functionally absent in the order; (3) why the discovery test applied is narrower than the one the Court itself quotes; and (4) why every class of document was ordered. It also never asks a policy question that runs through similar suits by Roche.
Understanding the Feb 2025 Order
Roche filed interim applications seeking production/discovery (under Order XI Rules 12 and 14 Code of Civil Procedure) of the regulatory dossiers, covering cell-line development, product characterisation, pre-clinical and clinical data, committee reviews, package inserts, marketing material and sales accounts. The single judge bench of Justice Amit Bansal allowed Roche’s applications. The Court held that the real controversy is whether the defendants’ trials and biosimilarity claims were valid and whether the Drug Controller General of India (DCGI) approvals conformed to prescribed law and procedure. Yet although the controversy is about approvals granted by the DCGI, the discovery order runs against the drug makers, Hetero and Cadila, and not against the DCGI.
The Court, while emphasizing on how discovery and interrogatories are designed to narrow the issues, shorten trial duration, and save litigation costs, took a liberal approach. Relying on the Supreme Court’s ruling in M.L. Sethi, alongside Delhi High Court’s judgement in M. Sivasamy; Koninklijke Philips v. Vivo; Largan Precision v. Motorola, held that discovery follows where documents are relevant, in the other side’s possession, necessary at that stage and needed for fair disposal. The Court also explicitly rejected defendants’ defense that discovery requires a preliminary prima facie case or that the prayers constituted a roving and fishing expedition.
The Court’s analysis relied heavily on the line of precedent established in the similar Biocon and Reliance biosimilar litigations. Tracing the history of the 2016 Biocon judgment, the Court highlighted that because regulatory approvals were granted on the basis of data benchmarked against the plaintiffs’ reference products, the defendants cannot claim confidentiality after publicizing their drugs as biosimilars. Recognizing that courts lack technical expertise to evaluate biological data, the Court affirmed that plaintiffs require access to clinical trial data to assist the Court. That sits oddly with what follows, because the Court then set up a Confidentiality Club in both suits, with two advocates and one expert for Roche and two advocates for the defendants.
So to briefly summarise what I’ve been able to make out from the court documents: Roche approached the Court questioning DCGI approvals granted to defendants claiming that this led to ‘passing off’ (or ‘extended passing off’), and the Court, instead of focusing on DCGI’s reasons for giving approval , asked the defendants’ to hand over data to Roche behind an unevenly formed confidentiality club, while observing the data cannot be claimed to be confidential by the defendants!
To Be Or Not To Be Confidential
Roche’s case for the disclosure of defendants’ data was based on the public interest. In para 51, the Court then held that the defendants cannot claim confidentiality because they used plaintiff’s data to establish biosimilarity. Yet in paragraph 60, the Court, pointing to the confidentiality of the data, directed the formation of a confidentiality club of two advocates and an expert for Roche and two advocates for the defendants. It’s hard to unravel the Court’s reasoning here. If the data is not confidential and its disclosure serves the public interest, the logical conclusion is open disclosure, not a sealed cover for one litigant. In fact, extending this, one could also argue that a public-interest claim would apply evenly to all regulator approved clinical trial data (See Swaraj and Md Sabeeh’s article here for a discussion on the dangers of not having transparent regulatory practices) .
Fairness between the parties breaks down in the club’s composition. It is hard to see why only Roche gets an expert as the member of the Confidentiality Club, and not the defendants. It stands in contrast to the Biocon ruling, where each side had an expert in the Club. The order also sets no test for the expert’s independence, although the concern about an in-house person acting as expert had surfaced in the Reliance proceedings. (See here) Moreover, weeks prior to this order, the same judge in Syngenta Limited and Anr. vs. GSP Crop Science Private Limited clarified the significance of the role of Confidentiality Club members, their access to confidential information and the process for their nomination. (See here)
The Court didn’t limit discovery to a certain class of documents; rather the order directs the filing of every class of document on Roche’s list, from cell-line development to licence renewals. The court’s reason for relevance supports the disclosure of the core dossier, but it does not obviously reach sales accounts, which concern profits and not the validity of approvals. The list of documents sought from Hetero further asks for any other document on development, testing and approval, which comes close to the roving inquiry the defendants complained of.
Consequently, the confidentiality club lets Roche investigate and examine the approval granted by the DCGI to Cadil and Hetero, even though Roche didn’t grant the approvals for the defendants’ drugs. This point becomes concerning in the light of the fact that the order does not record the regulator’s position.
The Regulator is Central to the Issue, but Missing from the Order
Roche’s suits challenge DCGI approvals, and the DCGI is a defendant in both. Roche claimed that the approvals were granted without the comparative testing that the Drugs and Cosmetics Act, the 1945 Rules and the 2012 Similar Biologics Guidelines require to establish biosimilarity. This means that either the regulator was not given proper data for approval, or that it did not insist on required data for approval. Either way, the most obvious question for the Court to have inquired would have been: What data did the DCGI receive, and why did it approve it? Yet the order does not record any stand taken from the regulator, whether it produced any of its documents, or whether the court or any of the parties asked it to. The burden fell instead on the defendants, who must open their confidential files to a competitor. Though the Court itself acknowledged the data’s confidentiality, it did not first ask the DCGI to explain how and why it granted approval, starting instead with the defendants’ files and a rival’s expert’s opinion on the matter.
This leaves a larger question unanswered that others on the blog have long asked: whether it is for the rival to challenge a regulator’s decision in a civil suit? The 2023 order in the matter tried to settle it by accepting Roche’s stand that it challenges not the DCGI’s competence or discretion but the manner in which the approvals were granted. However, the fallout of that shows through the Feb 2025 order that Roche did not grant these approvals and is not the regulator, and still it gets to examine, whether the DCGI did its job properly.
A Test with a Missing Limb
The test the Court applies for discovery is also incomplete. When it quotes Sivasamy in paragraph 24, the conditions include that production of documents is necessary at that stage of the suit and that the court considers it just. However, when the order later restates the test in paragraph 31, it drops the criterion of justness. It adopts the judge’s own view in Largan which noted that the only factor is whether the document bears on the dispute. If that is the whole test, then stage, necessity and justness have no work left to do, which sits uneasily with the rules the court quotes. The order also moves from relevance, to the finding that production of documents will save time and costs but doesn’t explain how. The reason to raise this is that similar suits brought by Roche are still pending for over more than a decade. The Biocon suit, filed in February 2014, had not reached admission and denial of documents by September 2026, and the Reliance suit was still being renumbered in March 2023, over three years after the Supreme Court asked for it to be decided within about twelve months.
The order also leans on precedent whose reach it assumes more than it examines. It treats the interim discovery ruling in Biocon and Reliance as final because the Supreme Court restored the single judge’s order in December 2019. However, the passage quoted in the present order pertained merely to making an interim direction operational. Crucially, the Supreme Court explicitly caveated that its observations were confined to that specific appeal and should not influence the High Court proceedings. Thus, relying on it as precedent for sweeping discovery stretches the Supreme Court’s limited intervention too far.
The Question the Order didn’t Ask
The order treats the matter as one concerned with disclosure of proprietary data, but it belongs to a larger question which is common to the long series of suits brought by Roche against Biocon, Mylan, Reliance, Hetero and Cadila. Roche’s Trastuzumab patent was no longer in force from 2013, and Bevacizumab seemed to have never had any patent protection. However, these suits brought by Roche present an avenue for the innovator drug companies to potentially restrict competition vis-a-vis biosimilars by claiming regulatory non-compliance, and extended passing off and goodwill dilution. It should be upon the courts to deal with these issues expeditiously and appropriately, to ensure justice is served in a timely manner. Earlier posts on the blog have also asked whether such suits work to keep competitors out once patents are gone. This question regains importance in the light of fact that these litigations involve drugs which are essential for cancer patients. Sadly, in the said order, public interest appears only as a ground for Roche’s case for disclosure to protect themselves from dilution of the goodwill, and not on the point that the biosimilars in question provide affordable cancer medication to a large number of Indian patients.
The Way Ahead
The grounds of the SLPs before the Supreme Court, and the specific questions the Court will decide them on, are not public; nor does any interim order record them. However, while the judgment and significant developments are awaited, the questions left open by the DHC order will arise again in biosimilar litigation, and they matter to the companies, the regulator and the patients who depend on these drugs.
Note: Any clarifications, corrections, or alternate perspectives on this are welcome!
