For the best part of 10 years, since the judgment of Lord Hoffmann in Kirin-Amgen v Hoescht Marion Roussel[1], it has been widely assumed that there is no file wrapper estoppel in the UK and no doctrine of equivalents either.  Both these assumptions are thrown into doubt by the Patent Court’s decision in Actavis v Eli Lilly [2].  The judge, Mr Justice Arnold, held that the court should not shut its eyes to the prosecution file but should ensure a patentee did not abuse the system by accepting narrow claims during prosecution and then argue for a broader construction in a subsequent infringement action.  The judge also applied the more or less defunct Improver or Protocol Questions[3] in a manner that potentially opens the door to a doctrine of equivalence in all but name.

Use of the File Wrapper

In Kirin-Amgen, Lord Hoffmann said that there were good reasons why the German and English courts discouraged, if not actually prohibited, the use of patent office files in aid of the construction of claims.  The reasons included that the meaning of a patent does not change depending on whether a party has access to the office file or not and anyway the file may reflect no more than a patentee’s desire to get the patent granted quickly without further argument with the examiner.  In short, it was a difficult and likely unproductive exercise to use the file wrapper to try to work out what the patentee meant by a claim’s terms.  

Taking the hint, parties and the courts made little use of file wrappers in subsequent cases with Sir Robin Jacob observing in Eli Lilly v Human Genome Sciences[4] that he was not confident it was legitimate to do so even though the UK courts had yet to decide the matter conclusively.  Going further, Mr Justice Floyd in Qualcomm v Nokia [5] warned that it was positively counterproductive to be referred to prosecution documents as it alerted the tribunal to the fact that there might be something to be said for the alternative construction. 

Doctrine of Equivalents

While Lord Hoffmann considered that Article 69 of the European Patent Convention firmly shut the door on any doctrine which extended protection outside the claims, he acknowledged that equivalence could be an important part of the background that would affect the skilled person’s understanding of a claim.  This was consistent with Article 2 of the Protocol on the Interpretation of Article 69, which states that, in balancing the literal and purposive construction of a patent’s claims, due account should be taken of “any element which is equivalent to an element specified in the claims”.  However, he doubted the usefulness of the Improver Questions in reaching a purposive construction saying that there was only one compulsory question: what would a person skilled in the art have understood the patentee to have used the language of the claim to mean?

Following this decision, the Improver Questions have seldom been used in UK patent actions.  In Smith & Nephew v Convatec Technologies[6] it was even questioned whether a 1992 decision on infringement of a numerical limitation would have been decided the same way following Kirin-Amgen as it had been using the Improver Questions.

The Actavis v Eli Lilly Decision

The Actavis case concerned Eli Lilly’s European patent protecting its cancer treatment, Alimta.  The patent claims the active ingredient, pemetrexed disodium, in combination with vitamin B12 and, optionally, a folic protein binding agent.  Actavis sought declarations that its proposed generic version of Alimta, which would contain one of pemetrexed diacid, pemetrexed dipotasium or pemetrexed dimethamine, did not infringe.  As Actavis was not challenging validity, it was allowed to seek declarations of non-infringement under the French, Italian and Spanish designations of the patent as well as the UK designation.

In his judgment, Mr Justice Arnold carefully reviewed the case law on both use of file wrappers and the doctrine of equivalence.  On the latter, he identified three main classes of case in which patentees resort to arguments on equivalence, namely where:

  1. the patent has been poorly drafted (as in the Improver case);
  2. technology has moved on significantly since the patent was filed (as in Kirin-Amgen); and
  3. the patent regrets a decision taken during the course of prosecution.

The Actavis case was a clear example of the third category. Eli Lilly had attempted to broaden its claims in prosecution to cover “pemetrexed”, failed and accepted the narrower claims to pemetrexed disodium.  It had reserved the right to file a divisional application, but did not do so.  It clearly now wished it had broader claims in order to catch Actavis on infringement.  But there was “no reason why the law should be sympathetic to the patentee”, which not only had the benefit of skilled professional advice but also had the opportunity to appeal against adverse decisions by patent examiners.  Further, allowing examiners’ decisions effectively to be overturned by the courts on claim construction would undermine their important role.  Against this background, Mr Justice Arnold made use of the prosecution history as one of several reasons why Eli Lilly’s broad construction, that pemetrexed disodium encompassed other salts because the important pemetrexed anion was present irrespective of the cation used, should be rejected.

He also used the Improver Questions, partly because there was no dispute that the alleged infringement was not within the primary, literal meaning of the claim and partly because the parties used them for consistency in their submissions on the non-UK designations.


Mr Justice Arnold observed that the US has a doctrine of equivalence and a doctrine of file wrapper estoppel to counterbalance it.  While following recent case law the UK had neither, file wrapper estoppel is now back albeit only in the case of the regretful patentee.  Even then, as Arnold J warned, the courts should be cautious and use the prosecution history only when it is short, simple and shows clearly why the claims are in their granted form and not broader. 

As regards equivalence, using the Improver Questions having held that there is plainly no literal infringement comes very close to Lord Hoffmann’s description of the US approach in Kirin-Amgen as being “to adhere to literalism in construing the claims and evolve a doctrine which supplements the claims by extending protection to equivalents”.  Even if not explicitly recognised as a doctrine of equivalence, the decision may therefore lead to a resurrection of the Improver Questions by patentees effectively arguing just that.