AstraZeneca AB v Pharmacor Pty Ltd (No 4)

AstraZeneca AB v Pharmacor Pty Ltd (No 4) [2026] FCA 663 is a Federal Court decision about discovery in a patent case. It is not the final merits decision. The Court had to decide whether Pharmacor should get broad discovery of documents relating to work performed by AstraZeneca's inventors for use in an inventive step challenge.

The Court refused the broader discovery Pharmacor wanted. It held that the inventors' actual thought processes, selection steps and comparisons were only of secondary significance to the inventive step issue raised on the application. The Court also found the request was not targeted or proportionate and could put the trial timetable at risk. Instead, it ordered a narrower category of discovery limited to documents already discovered in UK proceedings.

The parties were AstraZeneca AB and AstraZeneca Pty Ltd as applicants, and Pharmacor Pty Limited as respondent and cross-claimant. The reasons say broader background appears in earlier 2026 decisions in the same proceeding. This judgment itself deals with a procedural fight about documents, not the final patent outcome.

The timing mattered. The trial had originally been listed to start on 31 August 2026, but by order dated 12 May 2026 it was moved so that the hearing on issues other than quantum would begin on 19 October 2026. Timetabling orders had already been made. That meant any discovery order had to be assessed against a relatively tight path to trial.

Two interlocutory applications were heard on 26 May 2026. The first was AstraZeneca's application to set aside a Notice to Produce served by Pharmacor on 8 May 2026, except for certain identified documents. That issue was resolved between the parties, and costs were reserved.

The second application was Pharmacor's application for discovery. Pharmacor changed the form of the discovery it sought more than once. It first proposed one form in its submissions, then amended it after receiving AstraZeneca's evidence, and then amended it again during argument. The Court noted that the alternative form eventually sought was wider than the initial request.

Pharmacor said the documents were relevant to a pleaded ground of invalidity. In broad terms, it alleged that the asserted claims lacked an inventive step in light of common general knowledge together with WO 128, because the claims did not satisfy the selection principles summarised in Ranbaxy. Pharmacor's submission, as recorded by the Court, was that the relevant inquiry involved comparing dapagliflozin with other compounds in the broader prior art class disclosed in WO 128 to see whether there was some special advantage or quality peculiar to dapagliflozin and whether that was the basis for its selection.

AstraZeneca did not run the application on the basis that the documents were wholly irrelevant to the facts in issue. Instead, it argued that the inventive step case was legally tenuous and that the discovery sought was too broad. The Court also recorded AstraZeneca's position that the request was, in substance, a fishing expedition.

The immediate question was whether the Court should order the broader discovery sought by Pharmacor. That required more than a simple relevance check. In Federal Court practice, discovery is also shaped by proportionality, burden, timing and the real significance of the documents to the issues in dispute.

The legal complication was that Pharmacor's request was tied to an inventive step case said to depend on selection patent principles. The reasons refer back to earlier decisions in the same proceeding where the Court had not been persuaded, for the purposes of those applications, that those principles formed part of Australian law in the way argued, including in the context of obviousness. The Court repeated that the legal foundation of the point remained uncertain on this interlocutory application, while acknowledging that a different view could be taken at trial after fuller submissions.

The Court also noted an important structural point about the argument. It said selection principles are typically invoked by a patentee seeking to overcome a prior publication, using the Full Court's language in Ranbaxy, rather than by a party seeking revocation of claims. AstraZeneca submitted, and Pharmacor accepted, that Pharmacor therefore had to prove the negative of those principles. The Court set out what that appeared to mean on Pharmacor's case, including showing there was no substantial advantage secured by the use of the selected members and that the selection was not in respect of a quality of a special character peculiar to the group.

That legal uncertainty mattered because the strength of the case underpinning a discovery request can affect whether discovery should be ordered at all. The Court expressly said the strength of the case told against the discovery sought by Pharmacor for reasons explained in an earlier judgment.

Downes J refused to make either of the broader discovery orders sought by Pharmacor. The central reason was that the steps actually taken by the inventors to select dapagliflozin, and to compare it with other compounds, were of only secondary relevance or significance to the inventive step issue raised on the application.

In reaching that conclusion, the Court relied on authority stating that inventive step under section 7 of the Patents Act is assessed objectively by reference to whether the invention would have been obvious to a person skilled in the relevant art. It is not assessed by direct reference to what the inventor actually thought or did. The reasons quote Beach J in Bluescope Steel Ltd v Dongkuk Steel Mill Co, Ltd to that effect, and also refer to Teva and Pfizer. The Court said it was not persuaded that invoking the selection principles from Ranbaxy made the inventors' actual thought processes or comparison work critical to Pharmacor's case or took the application outside that line of authority.

The Court also found the discovery sought by Pharmacor was neither targeted nor proportionate. Importantly, the Court said the alternative form of order sought during argument had become wider than the initial request. That was a serious problem given the procedural posture of the case. The proceeding was listed for trial in less than five months, previous trial dates had already been vacated, and Pharmacor's evidence in chief was due by 30 June 2026. The Court said the form of order sought by Pharmacor placed both the trial timetable and the trial dates in peril.

That timetable risk was not a side issue. It was a key reason the Court declined broad discovery. The Court weighed the legal uncertainty of the inventive step case, the secondary significance of the inventor documents, and the practical danger of derailing preparation for trial. On that combination of factors, the broader orders were refused.

Instead, the Court accepted a narrower form of discovery proposed by AstraZeneca. That order was confined to documents already discovered in proceedings in the United Kingdom, with AstraZeneca allowed 14 days to retrieve and review them. Although AstraZeneca's primary position had been that no discovery should be ordered, the Court considered this limited approach appropriate in all the circumstances, including because it would allow the proceeding to remain on track for the October trial.

Costs on the interlocutory matters were reserved. The parties were ordered to provide a form of order reflecting both their agreement on the Notice to Produce issue and the Court's position on discovery and costs as reflected in the hearing transcript.

For businesses, this case is a reminder that discovery strategy has to be disciplined. If you are asking for documents, broad categories aimed at internal R&D work, inventor deliberations or comparison exercises may not be enough just because they sound potentially useful. You need to show why those documents matter to the legal test the Court must actually apply.

This is especially important in patent cases involving inventive step. The Court's reasoning aligns with established authority that the inventive step inquiry is objective. That means the Court is generally concerned with what a hypothetical person skilled in the art would have found obvious, not with reconstructing the inventor's subjective path. A request focused heavily on what the inventors actually thought or did may therefore be treated as secondary unless there is a stronger legal basis for it.

The case also shows that legal uncertainty can affect procedural outcomes. Here, the Court noted that it had previously not been persuaded, for the purposes of earlier applications, that the selection patent principles relied on by Pharmacor formed part of Australian law in the way argued. The Court did not finally determine that issue in this judgment, and it expressly said a different view could be taken at trial after fuller submissions. But for the purpose of deciding discovery, that uncertainty counted against ordering a broad and burdensome document production exercise.

In practical terms, if your business wants discovery, you should expect the Court to ask: is the issue legally solid enough, are the categories narrow enough, and can the case still get to trial on time? If the answer to any of those questions is weak, the Court may cut the request back sharply.

For businesses responding to discovery, this decision is useful authority for resisting overreach. The Court accepted that concerns about oppression, wording, timetable disruption and the fishing-expedition character of a request can all matter. Even where some relevance exists, that does not guarantee broad discovery.

There is also a records-management point. AstraZeneca still had to provide a narrower category of documents, namely documents already discovered in UK proceedings. Businesses involved in cross-border disputes should assume that documents produced overseas may become the practical starting point for discovery in Australia. Keeping technical and litigation records organised can reduce cost and delay when a narrower order is made.

Is this a final patent decision? No. It is a procedural ruling about discovery only.

Did the Court say inventor documents are always irrelevant? No. The Court said that in this application they were of only secondary significance to the inventive step issue being advanced.

Why did the Court focus so much on timing? Because the case was already on a path to an October 2026 trial, previous trial dates had been vacated, and Pharmacor's evidence in chief was due by 30 June 2026. The Court considered the broader request capable of putting both the timetable and the trial dates in peril.

What kind of discovery did the Court allow? A narrower category limited to documents already discovered in UK proceedings, with 14 days for retrieval and review.

The judgment was delivered on 26 May 2026 and the reasons were published on 27 May 2026. The hearing concerned two interlocutory applications listed on 26 May 2026. The trial had originally been listed for 31 August 2026, but by order dated 12 May 2026 it was moved so that the hearing on issues other than quantum would commence on 19 October 2026.

This page should be read as an explanation of the discovery ruling only. It does not state the final result of the patent proceeding, and it does not resolve whether the selection patent principles argument ultimately succeeded or failed at trial.

This page is based on the published Federal Court reasons in AstraZeneca AB v Pharmacor Pty Ltd (No 4) [2026] FCA 663. The reasons identify the case as a patents and discovery decision and explain why limited discovery was ordered instead of the broader categories sought.

The judgment refers to earlier 2026 decisions in the same proceeding for the broader patent background and for earlier discussion of the selection principles issue. This page therefore focuses on the procedural ruling made here and avoids broader claims about the final merits of the patent dispute.