Williams & Connolly LLP
Patents on processes have been a mainstay of the US patent system since its earliest days; the very first patent issued in the United States in 1790, signed by George Washington, was for “making Pot ash and Pearl ash by a new Apparatus and Process”. In the biomedical field, process patents – particularly those directed at methods of treatment and diagnostic methods – play a particularly important role that differs in important respects from patents on chemical compounds or industrial processes. Because these patents often involve discoveries made during a drug’s clinical development, they may have terms that extend well beyond that of the initial patent on a molecule and in many cases become the only patent protecting a drug.
Process patents in the life sciences, however, present certain unique issues that are reflected in recent case law regarding the patentability requirements of 35 USC Sections 101 and 112, and which may be of interest to life sciences stakeholders and attorneys.
A perennial ‘hot topic’, and a major source of confusion and controversy in US patent law for many years, is 35 USC Section 101, the statute specifying the subject matter eligible for patent protection. On the surface, the statute is expansive, reciting that “any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof” may be patented. But like many generally worded and venerable statutes, it has developed layers of decisional law that have taken it in directions that are not apparent from its text. The Supreme Court has held that “laws of nature, natural phenomena, and abstract ideas” may not be patented, even if they superficially fall within one of the categories of statutory subject matter.
The prohibition on patenting “laws of nature”, in particular, has given rise to a number of important decisions in the life sciences arena that affect method-of-treatment and diagnostic-method patents – and that frequently hinge on the distinction between the two. The difficulty arises because, on some level, virtually every life sciences patent involves laws of nature; pharmaceuticals and diagnostics work because science works. A landmark case addressing this issue is Mayo v Prometheus, a 2011 Supreme Court decision invalidating claims to a “method of optimizing therapeutic efficacy for treatment of an immune-mediated gastrointestinal order”. The method involved measuring the concentration of a particular metabolite in a patient’s blood and determining that the dose should be decreased if the metabolite’s level is above a given threshold. The court determined that this method was effectively directed to a law of nature – the relationship between the metabolite concentration and the likelihood that the drug would be ineffective or harmful.
Troublingly for patent owners, the method in Mayo included an explicit step of administering the drug, immediately raising questions:
- How far does Mayo go?
- Are methods of treatment, which necessarily rely on the natural relationship between a drug and its pharmacological target, subject to challenge?
Since Mayo, however, the US Court of Appeals for the Federal Circuit has answered in the negative, interpreting Mayo to be ineligible because it was not truly a method of treatment. The Federal Circuit explored the line between patent-eligible treatments and ineligible laws of nature in its 2018 Vanda v West-Ward decision. Vanda involved a method that was superficially similar to Mayo’s, involving treatment with the drug iloperidone. The claimed method involved genotyping the patient to determine whether he or she is a poor metaboliser of the CYP2D6 enzyme, then administering a different dose of iloperidone depending on this result so as to avoid certain risky cardiac side effects. Crucially, unlike in Mayo, the claims recited expressly the administration of particular amounts of iloperidone, and the end result of the process was treatment by effectuating this administration – not merely an observation as in Mayo.
The Federal Circuit has relied on this same distinction – and overturned lower courts that did not do so – multiple times since. Its 2019 decision in Natural Alternatives International v Creative Compounds considered a method of treatment in which a naturally occurring compound, beta-alanine, was provided in an amount above natural levels that was “effective to increase beta-alanylhistidine dipeptide synthesis”. Overturning the district court, the Federal Circuit concluded that even though beta-alanine is naturally occurring, and the claim involved the natural consequences of administering it, the claim was concretely directed to “using a natural product in unnatural quantities to alter a patient’s natural state, to treat a patient with specific dosages outlined in the patents”. And in 2020, the court applied similar principles in Boehringer Ingelheim v Mylan Pharmaceuticals, again reversing a lower court’s ineligibility determination. In Boehringer Ingelheim, the claim was regarding administering a particular class of drugs to patients with kidney problems; the challenger argued that the claim was directed to the natural law that the drugs in question are metabolised by the liver, not the kidney. But again, the Federal Circuit disagreed on the basis that even though the natural law may have been a crucial aspect of the invention, the claim was a concrete method of treatment involving administering a drug to a patient.
The apparent thrust of recent Federal Circuit case law, therefore, is that method-of-treatment claims are generally likely to survive eligibility challenges – but only if they can really be characterised as methods of treatment, and particularly if the end result of the entire method is administration of a drug to a patient. Patent applicants drafting claims would be well advised to spell out a treatment process expressly, even where the focus of their innovation is on a diagnostic test rather than on a novel drug or regimen.
What about diagnostic methods that do not involve administration of a drug? In the Federal Circuit’s own words, it has “consistently held diagnostic claims unpatentable as directed to ineligible subject matter”. But in the very case where it said that – the 2020 decision Illumina v Ariosa Diagnostics – it upheld a claim to diagnostic technology that culminated in analysing DNA and was entirely decoupled from any method of administration of any drug. The district court had struck down the claims on the basis that they were “directed to a testable quantity of genetic information found in nature”, but the Federal Circuit disagreed. In fact, the Federal Circuit stated outright that “[t]his is not a diagnostic case”. A key to this result was that the claims recited, and could be characterised as being directed to, preparing a sample of “cell-free DNA that is enriched in fetal DNA”. And as part of that preparation, the claims involved “specific process steps”. This focus of the claim was enough to overcome any argument about the end result of the claim, as well as the challenger’s contention on appeal that the claims centred around a discovery about naturally occurring differences between maternal and foetal DNA.
Illumina illustrates a longstanding principle of Section 101 case law: characterisation of the claim can decide the case. In litigation, both patentees and accused infringers are well advised to consider whether a claim can be fruitfully characterised as an unpatentable diagnostic or as a patent-eligible method of treatment. And when drafting or evaluating claims, patent applicants and patentees should be aware that the more a claim focuses on specific, concrete, physical steps, the more likely it is to be upheld. In general terms, things you do to a patient or in a test tube are likely to be held to be patent-eligible; things you observe, determine or decide are less likely to be so. That remains true even if the inventive step in giving a drug, or the reason for mixing substances together, involves a diagnostic observation.
Another set of issues to be aware of when asserting or evaluating life sciences method claims arises from the requirements of 35 USC Section 112. Section 112 encompasses three separate patentability requirements:
- the ‘definiteness’ requirement that the claims point out with reasonable certainty what they are claiming;
- the ‘enablement’ requirement that the specification of the patent contain sufficient information for the hypothetical person of ordinary skill in the art to make and use the invention (or to perform the method); and
- the ‘written description’ requirement that the specification describe the invention in sufficient detail that a person of ordinary skill can reasonably conclude that the inventor ‘possessed’ the invention.
All three of these requirements are frequently litigated and all three have a substantial body of case law associated with them that may be relevant to a case involving a method-of-treatment or diagnostic-method claim – or to any other patent case. Such method claims, however, pose a few particular issues with respect to the written description requirement that are worth some special consideration.
One practical reality of method-of-treatment claims is that they frequently arise from discoveries made after the invention of the active pharmaceutical compound. How much is known about the compound, its activity and dosing by the time a method-of-treatment application is filed can have profound effects on what claims the application ultimately can support. When a molecule is discovered to have new and potentially beneficial biological properties that may enable its use as a treatment, the natural inclination is to patent that use. Waiting too long to seek patent protection can jeopardise the ability to seek patent protection at all or can lead to an award of priority to a competitor. On the other hand, a patent application filed before the clinical use of the drug is well understood can fail to provide sufficient written description support for a method-of-treatment claim.
The recent patent dispute over the multiple sclerosis drug Tecfidera illustrates the challenges. The drug in question, known as ‘DMF’, was a compound that had been known for decades; it had been used as a topical treatment for psoriasis since the late 1950s. In the 2000s, however, Biogen (and certain collaborators) developed it as a multiple sclerosis drug. In 2007, a neurologist working for Biogen hypothesised that its efficacy might be driven by maximising the peak concentration of the drug in circulation, as opposed to by a particular level of continuous exposure. Biogen scientists who were not clinicians also did extensive pre-clinical work regarding the mechanism of action of DMF and related compounds against various neurological diseases including multiple sclerosis. Biogen filed a patent application on all this work that also included various exemplary doses at which DMF might be administered, including a dose of 480 milligrams per day (mg/day). That same dose of 480mg/day yielded positive clinical trial results, and Biogen subsequently refocused its application on methods of treatment of multiple sclerosis by administering orally 480mg/day of DMF. The patent claims issued and Tecfidera became a multibillion-dollar drug administered according to the patented method.
However, following a challenge by Mylan, the US District Court for the Northern District of West Virginia invalidated Biogen’s patent for lack of written description support. The problem, in a nutshell, was that the patent specification described a “slew of competing possibilities” for different conditions and regimens to treat them. While the specification did include treatment methods that encompassed multiple sclerosis, and did mention 480mg/day as a dosage, the court concluded that it contained “no blaze marks” that would lead from its generalised disclosure of possible treatments to the treatment of multiple sclerosis using DMF, nor to link a dose of 480mg/day of DMF to that treatment. The court also attached importance to the fact that during prosecution, Biogen contended that the ability of a 480mg/day dose of DMF to treat multiple sclerosis would have been unexpected. Mylan successfully argued that in the absence of a specific disclosure of that dose to treat that condition, nothing in the specification would have changed this expectation.
While the particulars of the Biogen v Mylan case depended on the complicated history of Biogen’s research and patent prosecution, it illustrates that method-of-treatment claims may be less susceptible to written description challenges the more expressly they recite a particular method of treatment. Where a specific claimed method of treatment is recited in the patent specification (or the patent claims set out in the original patent application), it is frequently difficult to mount a written description challenge; in contrast, method-of-treatment patents that list particular conditions and dosages among a large range of options, without adequately linking the two, may be especially susceptible to such challenges. In terms of drug development, it can be disadvantageous to rely solely on pre-clinical work on a drug’s mechanism of action to try to support method-of-treatment claims, rather than to file applications more specifically directed to a regimen that will be used in a clinical trial for a particular disease. And in cases (unlike Biogen v Mylan) where the focus of a method-of-treatment claim is on a novel combination of drugs, claims will be strengthened if particular doses for each are recited in the specification (even where they might have been present, outside the combination, in the prior art).
A strength but also a quirk of US patent law is that it reflects more than 200 years of statutory and case-law development that has been applied to myriad different technical disciplines. Patent practitioners and life sciences decision makers alike are wise to give careful consideration to the ways in which the rich body of precedents applies to life sciences method claims. With that consideration, they can survive patentability challenges and form an important part of pharmaceutical patent portfolios.