Are polymorph patents necessarily obvious? A recent CAFC decision may read-through to Revlimid’s polymorph patents.
The Federal Circuit has issued a precedential decision addressing whether a patent covering a given polymorph was invalid as obvious, Grunenthal GmbH v. Alkem Laboratories Ltd. Though the Court explained that it was not establishing a categorical rule that polymorph patents can never be obvious, the case nonetheless provides important guidelines for when a polymorph patents are likely to be invalid. For those following Revlimid®’s patent cases, the immediate question is—does the Grunenthal case have read-through to Celgene’s polymorph patents?
The Grunenthal decision.
Assertio ($ASRT), formerly Depomed, sells Nucynta® ER for severe pain. The drug is covered by U.S. Patent No. 7,994,364 and listed in the Orange Book. The ‘364 patent is directed to a Form A polymorph of the chemical compound tapentadol hydrochloride and a method of treating pain and urinary incontinence.
There are two known polymorphs for tapentadol hydrochloride—Form A and Form B. The Form B polymorph was previously disclosed in an earlier patent, U.S. Patent No. 6,248,737. In addition, the prior art generally disclosed methods of polymorph screening. For instance, the Byrn article disclosed a conceptual approach for investigating polymorphs for a given compound. Byrn discusses several different solvents that can be used to recrystallize the substance, as well as other variables—temperature, concentration, agitation, pH, etc.—that could affect the resulting recrystallization.
The Federal Circuit affirmed the lower court’s ruling that the ‘364 patent was not obvious. Importantly, the Court found there was no reasonable expectation of success in producing Form A. That was another way of saying that, based upon the available scientific knowledge at the time, the likelihood of successfully synthesizing a stable Form A polymorph of tapentadol hydrochloride was too unpredictable, and could thus not be a matter of routine experimentation.
More specifically, the Byrn article outlines a number of variables that could be adjusted to attempt solution recrystallization. However, Byrn did not necessarily disclose when it would be appropriate to use which variables. In other words, Byrn disclosed several solvents that can be used, such as water, methanol, ethanol, among many others. Byrn also disclosed that variables such as temperature, concentration, agitation, and so forth, could also be varied. But the Court found that Byrn “does not provide guidelines regarding which temperature, concentration, agitation, or pH levels are likely to result in polymorphs of particular compounds.” (Slip op. at 14).
Given this, the Court found that though Byrn provided guidance on how to go about investigating polymorphs, it was insufficient to render specific polymorphs invalid, such as the Form A polymorph in the ‘364 patent. The Court relied upon precedent holding that to have a reasonable expectation of success, the prior art must do more than simply motivate a scientist to “vary all parameters or try each of numerous possible choices until one possibly arrived at a successful result”. (Slip op. at 15, citing Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1365 (Fed. Cir. 2007)).
The Court also rejected the argument that, even if Byrn did not provide a specific roadmap to discovering the Form A polymorph, doing so was nonetheless “obvious to try.” In other words, the argument in favor of finding the patent obvious was that scientists were clearly motivated to try finding a polymorph for tapentadol hydrochloride, and Byrn provided the general guidance for doing so.
The problem was that Byrn disclosed too many variables to adjust, and not enough of specifically enabling methodology. Since Byrn provided no guidance on which variables were critical, and which directions were likely to be successful, the Court held that even if there was a commercial demand to finding Polymorph A, the resulting discover of it was still patentable.
Does the Grunenthal case have readthrough to Revlimid’s polymorph patents?
The crux of the Grunenthal case is that for a polymorph patent to be invalid as obvious, the prior art must do more than provide general guidance on a plethora of factors that can be explored. The Grunenthal decision did not categorically hold that polymorph patents can never be obvious. Indeed, the Court stated, “[o]ur decision today does not rule out the possibility that polymorph patents could be found obvious.” (Slip op. at 19).
The Court’s decision that the ‘364 patent in Grunenthal was not obvious was based upon the available prior art as of the prior date of that patent. Likewise, for other, later patents, prior art may exist that more precisely instructs a pathway for discovering polymorphs for a given compound. Thus, whether the Grunenthal decision reads through to other polymorph patents will depend upon the particular compound and the date the patents.
One obvious case involving polymorph patents that has attracted considerable attention is Celgene’s Revlimid® patent cases currently pending. The priority date for Celgene’s two polymorph patents currently being asserted against the generics is roughly 2003 for the ‘800 patent and the ‘217 patent. The priority date for the ‘364 patent in the Grunenthal case is roughly 2004.
This suggests that, since the Revlimid® polymorph patents were filed earlier than Depomed’s ‘364 polymorph patent, and the ‘364 patent survived an obviousness challenge, then it may be unlikely that sufficient prior art exists that could invalidate the Revlimid® polymorph patents.
Is this conclusive? No. There could still be prior art lurking out there uncovered by Dr. Reddy’s that is more apropos of what is disclosed in Revlimid®’s polymorph patents. But it is something. The Grunenthal case is one of few precedential opinions from the Federal Circuit addressing the obviousness of polymorph patents head-on, and therefore it is an important ruling for all future polymorph patent cases.
If anything, the Grunenthal decision may increase the importance of Dr. Reddy’s non-infringement position—namely, that its formulation is amorphous.
 Stephen Byrn et al., Pharmaceutical Solids: A Strategic Approach to Regulatory Considerations, 12 Pharmaceutical Res. 945 (1995).