Dr David Eyre

October 2026

This decision overturned a $107.5 million infringement award. The CAFC held that Wyeth’s method of treatment claims were invalid for lack of enablement because the patent did not disclose to the skilled person how to identify a therapeutically effective daily dose without undue experimentation.  

As alluded to above, an amendment during prosecution that was introduced to gain novelty caused a novelty–enablement “squeeze”. The claimed subject matter was considered novel, but not now enabled by the specification as originally filed. Thus this decision highlights an important lesson: amendments introduced to overcome prior art may also increase the disclosure burden that the patent must satisfy. 

Background 

The patents at issue, US 10,603,314 and US 10,596,162, concerned the treatment of gefitinib- and erlotinib-resistant non-small cell lung cancer (NSCLC). The claimed invention involved administering an irreversible EGFR inhibitor capable of covalently binding to a specified cysteine residue in the receptor. 

An essential feature of the claims was the requirement to administer a daily “unit dosage” of the inhibitor to a patient. Although the patents contained in vitro data demonstrating the activity of several compounds, the specification did not include any examples of dosages in a clinical setting or instructions as to how to find an effective therapeutic dose in a patient. 

The importance of this distinction between in vitro data and clinical practice was central to this case.  

As noted above, during prosecution, Wyeth amended the claims to require that a “unit dosage” be “administered daily to the patient”. This amendment was introduced to distinguish the claims from the prior art and secure grant. However, the specification also contained a definition of “unit dosage” as a predetermined quantity “calculated to produce the desired therapeutic effect”. 

The CAFC considered that definition as being very significant. Once the patentee had chosen to define “unit dosage” by reference to therapeutic efficacy, the CAFC viewed the claims as necessarily requiring administration of an effective therapeutic dose to a patient. This is in contrast to disclosures for compounds capable of inhibiting EGFR activity in vitro. 

Thus the question for the CAFC regarding enablement was not simply whether the disclosed compounds possessed the claimed biological activity. Rather, the question of enablement became whether the specification disclosed to the skilled person how to determine a therapeutically effective daily dose of the claimed compounds without undue experimentation. 

The CAFC concluded that the specification did not do this. 

Although the specification disclosed broad dose ranges, those ranges were described only as “general” and “projected”. The specification did not describe how the ranges had been derived, how an appropriate dose should be selected within those ranges, or how the disclosed dosages related to the therapeutic effect that was an essential feature of the claim. Also, AstraZeneca presented evidence that therapeutically effective doses for at least two of the three specifically disclosed compounds would exceed the maximum tolerated human dose. Further, this evidence was not rebutted by Wyeth. Indeed, the inventors’ own evidence indicated that concentrations shown to be effective in laboratory testing could not safely be administered to patients. 

Therefore the CAFC concluded that the information in the patent specification did not disclose sufficient information to identify a workable therapeutic daily dose Rather, this critical task was left for a skilled clinician to work out. The CAFC considered this to require undue experimentation and thus the claims were not enabled. 

Importantly, this decision does not deny that method of treatment claims can satisfy enablement without human trial results. The CAFC emphasised that this decision does not mean that clinical data is required for every pharmaceutical patent.  

It appears that had the specification avoided linking “unit dosage” explicitly to therapeutic efficacy, or used different language, the enablement analysis might have proceeded very differently. 

That is, by defining “unit dosage” as being the amount calculated to achieve a therapeutic effect, the patentee effectively introduced a functional therapeutic limitation into the claims. That definition then became central to claim construction and significantly increased the enablement burden. Accordingly, this case illustrates the risks associated with definitions for claim features that are given in the patent specification. 

Conclusion 

In conclusion, this case confirms that amendments to address novelty or inventive step objections may also alter the technical contribution to the art that must be supported by the patent specification. Care must therefore be taken that consider each added feature not only from a patentability perspective but also in terms of whether the application provides sufficient technical disclosure across the scope of the amended claim. 

Also, in drafting patents for prosecution in the US patentees (and their attorneys) should be cautious when including definitions as part of a “glossary” of terms in the specification. This case illustrates that definitions intended simply to clarify terminology might later be used to place functional constraints on the claims and such constraints, might potentially raise the bar for the sufficiency and enablement of the claims. 

Furthermore, where claims relate to therapeutic administration, it is prudent to include as much information as possible regarding dose selection, therapeutic rationale, and any basis for proposed dosage ranges. While clinical data may not always be available at filing, applicants should ensure that the specification explains how the skilled person can find effective therapeutic regimens without having to engage in substantial research. 

Finally, Wyeth serves as a reminder that patent drafting often involves a delicate balancing exercise. For European patent attorneys, the decision reinforces the importance of considering patentability and disclosure requirements together. A claim amendment may solve one problem during prosecution, but if the supporting disclosure has not kept pace, it may simply create another.