When do clinical trial consent disclosures render a patent invalid?

By Jane Owen, Patrick Brown

07-2019

In the recent case of InterPharma Pty Ltd v Hospira Inc (No 5) [2019] FCA 960 the Australian Federal Court has given patentees guidance as to novelty preserving practice in the conduct of clinical trials.

As part of Good Clinical Practice (GCP) and local regulatory requirements, participants in clinical trials must provide informed consent.  To ensure participants can provide informed consent, trial organisers must provide them with a participant information sheet and consent form.  These documents must disclose enough detail about the objective of the study, type of treatments to be administered and possible effects to allow prospective participants to make an informed decision about whether to participate or not. 

Although detailed and accurate disclosure is an essential requirement for the conduct of clinical trials, there is a risk that publication of this information may adversely affect the novelty of subsequent patents.  A recent decision of the Federal Court of Australia in the case of InterPharma Pty Ltd v Hospira Inc (No 5) [2019] FCA 960 provides an interesting insight into the extent to which clinical trial participant information sheets and consent forms could damage patent novelty.

Background to the case

The case concerned a challenge to the validity of Australian Patent 754484, which was held by Pfizer and Hospira and was directed to the "use of dexmedetomidine for ICU sedation" (Pfizer patent).  Dexmedetomidine is a sedative belonging to the class of α2-receptor agonists.  The compound was not novel, having been disclosed in US patent 4910214 in 1987 as a general sedative. The Pfizer patent claimed new uses and methods of treatment for dexmedetomidine in the specific area of intensive care unit (ICU) sedation. 

InterPharma intended to supply a generic version of dexmedetomidine for sale in Australia, including for use in ICU sedation, and challenged the validity of the Pfizer patent on a variety of grounds.  These included that the patent was not a patentable manner of manufacture, was not novel, lacked an inventive step and the claims were not described with sufficient clarity.  Hospira and Pfizer cross-claimed that InterPharma threatened to infringe by supplying the generic product and were granted an interlocutory injunction preventing the sale of the generic product until the Pfizer patent expired in March 2019.

The court ultimately found in Pfizer's favour, rejecting all the grounds of invalidity argued and holding that InterPharma had threatened to infringe Pfizer's patent by the sale of the generic product.

Clinical trial consent forms as prior art

In arguing that the Pfizer patent lacked novelty, InterPharma, in part, relied on patient information consent (PIC) forms from two clinical trials conducted by Abbott Laboratories, as Pfizer's predecessor in the development of the Pfizer patent.  InterPharma argued that the PIC forms amounted to prior art information which anticipated Pfizer's patent, making it not novel under s 7(1) of the Patents Act 1990 (Cth) (Patents Act).

The test for anticipation requires the information to have been "made available to at least one member of the public, who was, in that capacity, free in law and equity, to make use of it" (Insta Image Pty Ltd v KD Kanopy Australasia Pty Ltd [2008] FCAFC 139) and that it discloses "explicitly or implicitly all of the essential features of the invention, as claimed" (AstraZeneca AB v Apotex Pty Ltd [2014] FCAFC 99).

The two PIC forms discussed in this case related to clinical trials of dexmedetomidine conducted at Duke University in North Carolina (Duke study) and in the Netherlands (249 study).  Both studies were completed before the priority date of the Pfizer patent on 1 April 1998. 

Justice Kenny found the PIC forms had been made publically available as required by s 7(1) of the Patents Act, because participants were free to discuss the forms with anyone they choose prior to giving their consent.  Evidence was given that this represented standard practice in relation to GCP clinical trials.  The question then, was whether the forms disclosed all the essential features of the patent.  In determining this, Justice Kenny considered what was disclosed on the face of each PIC form and expert evidence regarding what this would teach the skilled addressee. 

The Duke study aimed to assess the effects of dexmedetomidine as an adjunct to anaesthesia during surgery.  Sedation was only referenced in the PIC form as a potential side-effect, not a likely or desired outcome, so it did not teach anything about the use of dexmedetomidine in ICU sedation.  In addition, dexmedetomidine was only to be assessed in combination with other drugs, not as "essentially the sole active agent" as claimed in the Pfizer patent. 

In contrast, the 249 study aimed to evaluate the use of dexmedetomidine in ICU sedation.  Despite this there was some disagreement between expert witnesses about what the PIC form revealed compared to the patent claims.  Although the study, as outlined in the PIC form, addressed most of the features claimed in the patent, this was found to represent a hypothesis the researchers hoped to test.  This was particularly apparent because this was a small scale phase II trial, which suggested a greater level of uncertainty in the outcomes.  As a result, the skilled addressee reading the form would not know that dexmedetomidine could be safely and effectively used in ICU sedation in the manner described, only that the researchers considered this a likely outcome.

In this case, neither the Duke, nor the 249 study was held to anticipate the Pfizer patent.  The disclosure in the PIC forms was compared to reports of a series of phase I clinical trials in Bristol-Myers Squibb Co v F H Faulding & Co Ltd [2000] FCA 316 (Bristol-Myers Squibb).  Despite covering some of the features claimed in a patent, each trial did no more than describe an investigation leading towards a resolution of the problems to which the patent was directed, so could not amount to anticipation. 

In contrast, a published abstract, considered in Bristol-Myers Squibb, describing another ongoing trial did anticipate the patent.  The abstract described the drug's activity against certain cancers, maximum tolerated dose and investigations into the use of a shorter infusion time at a lower dose.  These features, combined with indications of success from the new treatment regime were found to suggest to a skilled addressee the same method of treatment claimed by the patent with a reasonable expectation of success.

Conclusions

Although in the present case Pfizer's patent was found to be valid, the discussions in Justice Kenny's judgment and the judgment of Black CJ and Lehane J in Bristol-Myers Squibb, provide a note of caution.  PIC forms are an essential part of clinical trial practice and by their nature require that participants can freely discuss their contents before giving consent.  As a result, they are likely to be made available to the public and can form part of the prior art for the purposes of assessing novelty under s 7(1) of the Patents Act.  The discussion and outcomes in these two cases highlight that whether a PIC form affects the novelty of a patent will depend greatly on the precise details of the information disclosed and what the skilled addressee could learn from reading it.  Care must be taken when drafting participant information documents to balance the need to fully disclose the nature of the study to participants, while ensuring the patentability of resulting discoveries is not adversely affected.