DSM IP Assets BV & Anor v Algal Omega 3 Limited (In Administration) & Anor

This Court of Appeal decision is a substantial patents appeal about microbial oil extraction technology. DSM said Mara infringed three patents. Mara denied infringement and sought revocation. The trial judge had reached a mixed result, and both sides appealed different parts of that outcome.

For ordinary businesses, the useful lesson is straightforward: patent value depends on exact claim wording, earlier published material and what the skilled team in the field would already know. Courts will not rescue a weakly drafted or obvious claim just because the commercial objective behind it seems clear.

DSM owned patents relating to methods of producing and extracting microbial oils containing polyunsaturated fatty acids, including DHA. These oils had commercial importance because DHA is used in products such as infant formula. Mara was accused of using processes that fell within DSM’s patent rights.

The dispute was not just about whether Mara’s processes matched DSM’s patents. Mara also attacked the patents themselves. That is common in patent litigation: an alleged infringer often argues both that it does not infringe and that the patent should never have been granted, or should now be revoked, because it lacks novelty or inventiveness.

The trial judge found that EP155 survived and EP801 did not. On appeal, Mara tried to overturn the validity finding on EP155. DSM tried to revive EP801 by arguing that the judge had interpreted one claim feature wrongly and had been wrong to find the patent obvious over an earlier publication called Kobzeff. DSM also argued that, if EP801 were valid, more of Mara’s processes should have counted as infringing.

The Court of Appeal therefore had to work through technical background, common general knowledge, prior publications and claim wording. Even though the science was specialised, the legal pattern is familiar: what exactly does the contract-like wording of the patent claim mean, and was the claimed idea really inventive when measured against what was already known?

The appeal involved two main strands. First, EP155. The only claim in issue was a method for obtaining a polyunsaturated fatty acid-containing lipid from Schizochytrium biomass by contacting the biomass with an enzyme, specifically a protease. Mara argued that this claim was obvious in light of an earlier publication known as Bijl.

Secondly, EP801. DSM had proposed an amended claim for a process involving lysing a microbial cell, adding base to raise pH to 8 or above to demulsify the composition, using one or more further demulsification techniques such as salt, heat, agitation or a second base, and then separating a lipid. The claim also said the lipid contained less than 5% by weight of an organic solvent and that the lysing comprised enzymatic treatment.

A central fight on EP801 was what the phrase about less than 5% organic solvent actually meant. DSM said it referred to the lipid during the process, so the process itself had to stay effectively solvent-free. Mara said, and the judge had held, that it referred to the lipid resulting from the process. That difference mattered because earlier technical material might make the claim obvious on one reading even if DSM’s broader commercial narrative was that the invention was about avoiding solvent-based extraction.

The Court of Appeal dismissed both appeals. That meant Mara failed to overturn the finding that EP155 was valid, and DSM failed to overturn the finding that EP801 was invalid for obviousness.

On EP801, the court agreed with the trial judge’s interpretation of the claim wording. It held that the phrase requiring the lipid to contain less than 5% by weight of an organic solvent referred to the lipid resulting from the claimed process, not to the lipid at every stage during the process. The court said that reading fitted the structure of the claim and made practical sense. It also stressed that a process claim can legitimately be defined in part by its result.

The court rejected DSM’s argument that the broader purpose of the invention should drive a different reading. Purpose matters in patent interpretation, but it is not everything. The court said the language used still matters, and if the drafter chose a particular limitation, the court must give it meaning in context.

The court also held that, even on DSM’s preferred interpretation, the judge had been entitled to find EP801 obvious over Kobzeff. It accepted that the skilled team would see salt and heat as obvious techniques to try for demulsification and that the judge had not fallen into hindsight reasoning.

If you are not in biotech, the science can be ignored. The business lesson is about diligence and drafting. Many SMEs treat patents as simple badges of exclusivity: either a business has one or it does not. This case shows that the real question is narrower. What exactly is claimed, how vulnerable is that claim to earlier publications, and does the wording match the commercial story your business is telling customers, investors or partners?

If you are licensing in technology, a surviving patent may justify exclusivity, minimum purchase commitments or premium pricing. But if a key patent is later held obvious, the economics can change quickly. The same applies if you are buying assets from a distressed business, entering a manufacturing partnership, or relying on a supplier’s assurance that a process is protected. You need to know whether the protection is broad, narrow, contested or already under attack.

This case also shows why technical teams and legal teams need to work together. The court paid close attention to common general knowledge, routine testing and what a skilled team would find obvious to try. A commercial team may see a process improvement as valuable, but that does not automatically make it patentable.

Before you sign a licence, supply agreement, collaboration deal or investment round that depends on process IP, run a practical review. The aim is not to re-litigate the patent. It is to understand whether the legal right actually supports the commercial assumptions in the deal.

In a dispute, do not focus only on whether your process looks similar at a high level. Patent cases often turn on a specific step, threshold, ingredient or result. Equally, if you are enforcing your own IP, be realistic about validity risk. A claim that sounds commercially attractive may still fail if an earlier publication and common general knowledge make it obvious.