Beware Clinical Trial Protocols as Prior Art

Beware Clinical Trial Protocols as Prior Art

Bruce DeKock

Introduction

          Developing new medicines requires innovation and testing.  Protecting innovation is essential for pharmaceutical companies.  Most scientists and clinicians know that patent applications should be filed to protect new innovations, and that patent applications should be filed before publication of the data confirming the innovation worked.  But surprisingly, publications that do not disclose the innovation per se, such as a clinical trial protocol, can be used to invalidate patents.  This article explores how publications that do not expressly disclose the results of a clinical trial can be used to invalidate patents on methods for administering pharmaceutical products and provides suggestions to innovators on steps to take to protect their future patent filings.

Two recent court cases invalidated patent claims based on disclosures of the upcoming clinical trial.  In the recent case of Bayer Pharma AG v. Mylan Pharma. Inc, et al,[1] the Federal Circuit held that a journal article describing a yet to be completed clinical trial anticipated several claims in a patent, rendering those claims invalid.  Similarly, in Salix Pharm. v. Norwich Pharm.,[2] the Federal Circuit upheld the use of a clinical trial protocol published on ClinicalTrials.gov to invalidate the inventions claimed in a patent as obvious.  Neither the disclosure of the clinical trial nor the protocol contained any efficacy data (by definition!).  Yet in both cases the publication of the upcoming clinical trial was one of the key references used to invalidate the patent claims.  How did that happen?

          New inventions may be protected in the United States and other countries by means of a patent.  In the United States, to obtain a patent, an invention must be both new (or “not anticipated” in the language of the patent statute) and not obvious at the time of filing the patent application.  To be “new” is rather straightforward, only requiring a difference, however trivial, from what was known before the patent application was filed.

          Whether an invention is obvious is a different analysis.  Typically, obviousness is evaluated by comparing two or more prior publications and asking whether a person of ordinary skill in the art would have combined the publications to come up with the innovation and would have had a reasonable expectation of success in doing so.

          The publications available for an obviousness analysis are varied and encompass a broader range of materials than one might expect.  A publication is available as prior art if it has been disseminated or otherwise made available to the extent that a person of ordinary skill, exercising reasonable diligence, can locate it.  .[3]  The question is not whether particular members of the public actually received the information.[4]  Instead, the question is whether a person of ordinary skill in the art could, after reasonable diligence, access a publication. [5]  Indexing or searchability bolster the status of a document as a printed publication, but are not absolutely necessary,[6] Thus a doctoral thesis cataloged in a library[7], abstracts circulated at a scientific conference[8], and yes, clinical trial protocols can all be used as publications to evaluate patentability of an invention. 

          In Bayer Pharma, Bayer’s patent[9] referenced a Phase III clinical trial (the “COMPASS trial”) evaluating the efficacy of administering rivaroxaban with and without aspirin to treat coronary artery disease (CAD) and peripheral artery disease (PAD).  Interestingly, the claims contained a limitation that the amounts of rivaroxaban and aspirin were administered in amounts that are “clinically proven effective.”  During prosecution of the patent application, the publication at ClincialTrials.gov of the clinical protocol for the COMPASS trial was asserted by the examiner as prior art.  Bayer was able to overcome the rejection by showing the patent application was filed before the protocol published.  No doubt though that Bayer was concerned that prior disclosures of the clinical trial could jeopardize the patent. 

          Unfortunately for Bayer though, a journal article published by Foley et al. before the filing of the patent application disclosed the COMPASS trial and stated that the trial studies the effects of rivaroxaban in patients with CAD or PAD.[10]  Foley stated that the patients were to receive different doses of rivaroxaban (an anticoagulant) and aspirin or a placebo.  Foley noted however that (1) there was no prospective data available on the effects of direct oral anticoagulant (DOAC) agents in patients with PAD, and (2) the COMPASS trial is ongoing and not expected to be completed until after the filing date of Bayer’s patent application.

          Despite the prospective nature of the disclosure in the Foley journal article, the Federal Circuit affirmed that the journal article anticipated several claims of Bayer’s patent.  Bayer argued that the claim limitation “clinically proven effective” was a meaningful limitation that excluded the Foley journal article, since the journal article published before the results of the COMPASS trial were in.  According to Bayer, “clinically proven effective” required proof of efficacy as shown by results from a clinical trial.  But the Federal Circuit disagreed.  According to the Federal Circuit, the limitation “clinically proven effective” was a “functionally unrelated limitation” that failed to make the challenged claims patentable.  The limitation “clinically proven effective” had no functional relationship with the claimed method.  Such proof of effectiveness did not transform the process of taking the drugs, and irrespective of whether the method was proven to be effective, did not change the method.  Accordingly, the Federal Circuit ignored the limitation “clinically proven effective” and upheld the PTAB’s finding that the Foley journal article anticipated several of the method claims.[11] 

Similarly, the experience of Salix Pharmaceuticals stands as a cautionary example of how seemingly innocuous documents can be trouble.  Salix initiated a clinical study to evaluate the efficacy of rifaximin for diarrhea associated IBS (or dIBS).  According to Salix, while IBS is one of the most common chronic medical conditions, the etiology of IBS was unknown.  Salix filed a protocol for a Phase II trial to administer rifaximin to patients aged 18 and over to receive daily doses of placebo BID, rifaximin 275 mg BID, rifaximin 550 mg BID, or 1100 mg BID for 14 days.  The protocol included the outcome measures of providing adequate relief of symptoms and evaluating a durability of response over a 12-week post-treatment period.[12]

          A little more than one year after the protocol published, Salix filed a patent application claiming an invention involving a method for treating dIBS with rifaximin.  The application included the results from the Phase II study showing the results of treatment and improvement over the placebo.  The patent claimed an invention involving a method of treating dIBS by administering a 1650 mg/day dose of rifaximin.  Salix disclosed to the Patent Office the protocol and other publications showing use of rifaximin to treat IBS.  The patent granted with a claim to a method of treatment for dIBS by administering 1650 mg/day of rifaximin for 14 days, wherein removing the subject from treatment after the 14 days results in a durability of response of about 12 weeks of adequate relief of symptoms.[13]  The patent included only one other claim, the above method in which rifaximin was administered as 550 mg three times per day.

Then came the ANDA filings from generics, including Norwich, challenging the validity of the patent.  Norwich argued that the invention was obvious – that the protocol combined with a publication by Pimental[14] disclosed all the features of the invention claimed by Salix.  Pimentel taught rifaximin resulted in statistically greater global improvement in IBS symptoms than placebo, and that the improvements were “seemingly” sustained through 10 weeks of follow-up after 10 days of treatment.  Pimental further noted that recent studies suggested that the “optimal dosage of rifaximin may, in fact, be higher than that used in our study.”[15]

Salix gamely argued against the references and argued that a person of ordinary skill would have had no expectation of success in using the claimed dose.  According to Salix, Pimentel did not find an improvement in the symptoms of abdominal pain and diarrhea.  Salix further argued that the protocol did not disclose results, and that a person of ordinary skill in the art would not have reasonably expected the trial to be successful simply because the trial had begun.

Salix’s arguments were for naught.  The Federal Circuit upheld that the patent’s claimed invention was invalid as obvious.  The Federal Circuit’s reasoning illuminates how courts approach the question of obviousness.  The court stated

[W]e are hesitant to conclude as a general matter that the disclosure of a Phase II clinical trial plan, standing alone, provides an expectation of success sufficient to render obvious a dosage that was not included within the planned clinical trial.  . . .  But the Protocol was not asserted alone; it was asserted in combination with Pimentel.[16]

According to the court, Pimentel supplied all that was missing from the protocol:  administering 400mg rifaximin TID resulted in relief from symptoms, and the suggestion that an optimal dose may in fact be higher than used in the study.  So while neither the protocol nor Pimentel together individually disclosed all of the features of the claim, together they did, and supplied the required reasonable expectation of success.[17]

This notion – that publications can be combined to invalidate patents - is a key take home message for scientists and clinicians.  Publications that by themselves would appear to not disclose an invention, can be combined to render obvious inventions that still require experimentation to prove.  And since parts of publications can be combined, any publication prior to the filing of a patent application is potentially dangerous. 

The results in Bayer and Salix feel in some sense deeply unfair.  The FDA never would approve a drug based on a clinical trial protocol and a paper that suggests the trial may be successful.  The FDA would want to see the trial results!  So how could this outcome have been avoided?  Here are some suggestions.

File patent applications early.  Of course, it is required to file patent applications before publications disclosing the invention,[18] but also best practice to file before any publications merely relating to the invention.  With respect to clinical trial protocols, request a delay in publication of the protocol if possible, to prevent disclosure and maintain secrecy before the patent application is filed. 

If the protocol publication cannot be delayed, file a patent application disclosing the invention as if the trial were successful, and include prophetic examples.  If the application is filed as a US provisional application, update the examples to include study results if possible, prior to the filing of a US non-provisional application or international application.[19]

Alternatively, if the patent application must be filed after a publication such as a clinical trial protocol, then limit the amount of disclosure contained in the publication and do not speculate about future results.  Use broad or general statements to describe the anticipated trial.  Do not use language that may express future success, but instead use language that more directly tracks the hypothesis to be tested.  Perhaps even reference uncertainty in the literature about causes of the disease, or with the proposed treatment.[20]

So be careful with publications relating to clinical trials before filing your patent application.  Who knows how a future court may take your publication, perhaps combined with other publications, to find your innovative method of treatment anticipated or obvious, even before it is tested.