Baraclude patent invalidation does not spell doom for lead optimization.
The recent invalidation of Bristol-Myers Squibb’s composition-of-matter patent for hepatitis B drug Baraclude (entecavir) by the District Court of Delaware (Burke, 2013) has caught the attention of not only those following the fortunes of the company, but also those with an interest in the patentability of small molecule drugs discovered by lead optimization processes.
The most immediate question raised by the ruling against US Patent No. 5,206,244 – namely the likely impact on Bristol-Myers Squibb’s revenue – is straightforward to answer. Baraclude’s US sales contributed only $240m of the drug’s $1.4bn total global sales in 2012, so from an earnings perspective the loss of US patent protection is not a severe blow, especially considering that Bristol-Myers Squibb’s opponent in court, Teva, is yet to gain FDA approval for generic entecavir.
It is the broader question of how this ruling could influence future decisions of patent validity that is more complex and interesting. Sanford Bernstein analyst Ronny Gal has commented in a research note to investors that “[t]he case has the potential to reverse well-accepted drug dogma” (Silverman, 2013). This refers to the general assumption that the expiry date of a drug’s composition-of-matter patent represents the very earliest date at which a generic version can be expected to enter the market.
A composition-of-matter patent protects a drug’s chemical structure, preventing others from selling the substance regardless of purpose, and such patents are widely viewed as constituting secure protection until expiry. In contrast, it is the more vulnerable “formulation” and “use” patents (deployed by brand companies to extend market exclusivity beyond the term of the composition-of-matter patent) that generic companies typically attack in an attempt to achieve early market entry.
Thus, the realization that composition-of-matter patents may not be sacrosanct has caused a stir. Of course, it is possible that the Baraclude ruling could be overturned on appeal, but if it were upheld would it spell doom for the patenting of traditional small molecule drugs? And would it open up the possibility of a wave of attacks by generics companies years before the expiry of primary composition-of-matter patents? Datamonitor’s view is that such extrapolation from the Baraclude case is misguided, although no less so than the hitherto undue faith that all composition-of-matter patents protecting marketed drugs are inherently invulnerable to attack.
For a patent to be valid in the US, the invention must be novel, nonobvious, and useful. Neither the novelty nor usefulness of Baraclude was in doubt: Bristol-Myers Squibb was the first to describe the chemical structure, and the drug’s antiviral activity is well documented. Instead, the case contested by Bristol-Myers Squibb and Teva hinged on the question of whether Baraclude was an obvious drug to make from the perspective of someone with “ordinary skill in the art.” In patent law, this hypothetical person has a thorough knowledge of the “prior art” (the relevant scientific knowledge publically available at the time at which the invention was made) but is unable to make innovative and imaginative leaps.
Teva’s argument centered on its identification of an antiviral compound, referred to in the trial as 2’-CDG and closely related to the naturally occurring 2’-deoxyguanosine, that Teva claimed to be prior art of relevance to the validity of the ‘244 patent. Baraclude differs from 2’-CDG by one of the smallest possible differences – the addition of a methylene unit (equating to insertion of a single carbon atom into the structure of 2’-CDG).
Teva argued that 2’-CDG constituted an obvious starting point for someone skilled in the art seeking to design a new antiviral drug. Furthermore, according to Teva, the addition of a methylene unit was an obvious change to make since not only was it a very conservative structural alteration, but also there was precedent for such modifications within other series of antiviral drugs.
The judge found these arguments convincing. In contrast, Bristol-Myers Squibb failed to persuade the court that the person with ordinary skill in the art would have been deterred from selecting 2’-CDG as a starting point. Bristol-Myers Squibb’s argument was largely based on the fact that 2’-CDG has toxicity that precludes it from being a useful drug. However, crucially, Bristol-Myers Squibb could not adequately demonstrate that knowledge of the toxicity existed prior to the date of invention of Baraclude, and thus this line of defense proved fruitless. Similarly, Bristol-Myers Squibb was found lacking in its attempts to suggest that addition of a methylene would have been outside the scope of potential modifications contemplated by someone skilled in the art.
To put it in the simplest terms, the ‘244 patent was fundamentally weak from the outset. Indeed, had the United States Patent and Trademark Office (USPTO) been made aware of the existence of 2’-CDG at the time the patent was filed, it is possible that the patent would never have been granted. Recognizing this, Teva had also submitted a claim of inequitable conduct on the part of Bristol-Myers Squibb, since inventors are under obligation to disclose to the USPTO all relevant prior art at the time of filing a patent application. The court found in favor of Bristol-Myers Squibb in this regard – even though there was evidence that the inventors were aware of 2’-CDG at the time – concluding that the inventors had shown oversight but no deliberate attempt to mislead. Nonetheless, had the Bristol-Myers Squibb inventors included in their patent application a more thorough assessment of all potentially relevant prior art, the Baraclude story may have been a considerably shorter one.
Further illustrating the weakness of the patent, the judge highlighted numerous instances in which Bristol-Myers Squibb’s own expert witness was unable to contradict Teva’s argument under cross-examination. As a general rule of thumb, the smaller the difference between the prior art and the invention, the greater the demand on the inventor to demonstrate why the invention is nonobvious. For Baraclude the bar was high, and Bristol-Myers Squibb had little to stand on.
All of this therefore means that the invalidation should cause minimal concern for those involved in small molecule drug discovery. Contrary to the suggestion of others, the court’s decision does not threaten the standard practices of drug discovery by lead optimization. It does not even imply that small structural changes will no longer be admissible. If at the time of invention there had been strong reason to expect that the small change versus 2’-CDG might compromise antiviral activity, or if there had been data in the literature clearly dissuading one from selecting 2’-CDG as a starting point, or if the invention of Baraclude had involved greater structural modification versus the prior art, then the ‘244 patent would have been on very much firmer ground.
Rather than undermining the principles of lead optimization, the court’s decision merely reminds us that, in return for the market exclusivity granted to a patent holder, it is expected that the invention represents an advance that could not have been made straightforwardly by a person of ordinary skill in the art following clearly signposted clues in the literature. This creates something of an uncomfortable conflict for many scientists, since it is in their training to show how their work relates to and is derived from what has gone before, and to show how they applied a logical sequence of steps to arrive at their new findings. Indeed, some of the evidence quoted by the judge against Bristol-Myers Squibb’s case was drawn from the inventors’ own later publications. Unfortunately for Baraclude, the patent system seeks unexpected discoveries, not inventions that could be engendered by minor stepwise extensions of existing knowledge.
Burke C (2013) Bristol-Myers Squibb Company v. Teva Pharmaceuticals USA, Inc. Civil Action No. 1 0-805-CJB, Memorandum Opinion. Available in two parts at http://freepdfhosting.com/983cc1cdae.pdf and http://freepdfhosting.com/a962026227.pdf [Accessed February 15, 2013].
Silverman E (2013) Not So Obvious: A Bristol-Myers Patent Defeat May Change Wall Street Views. Forbes, February 12. http://www.forbes.com/sites/edsilverman/2013/02/12/not-so-obvious-a-bristol-myers-patent-defeat-may-change-wall-street-views/ [Accessed February 15, 2013].