Can Coherus invalidate Amgen's Enbrel® patents? Part 2

Zachary Silbersher

In an earlier post, we discussed the pending IPRs filed by Coherus Biosciences against Amgen’s two patents covering its Enbrel® protein (entanercept).  Whereas our earlier post summarized Coherus’s argument for why the fusion protein claimed in Amgen’s ‘182 and ‘522 patents should be found obvious by the PTAB, this post will summarize Amgen’s preliminary response. 

At its most basic, Amgen’s protein is a fusion protein between IgG and TNFR.  One of the key aspects of the protein is that the fusion occurs at the hinge region, and includes all of the constant heavy domain regions of IgG (CH2 and CH3).  Coherus relied upon prior art (Smith, Watson and Zettlmeissl) that, together, purportedly taught fusion proteins at the hinge region of IgG. 

Overall, Amgen’s response is a strong one that highlights that there are numerous highly factual and scientific disputes between the parties.  This is not a case where Amgen argues that Coherus’ prior art simply fails to teach one or more elements of the patented claims.  Instead, Coherus picks and chooses from different pieces of prior art to show that the claimed fusion of IgG and TNFR at the hinge region was obvious.  While Coherus has theoretically found all the elements of the claimed fusion protein in the prior art, Amgen is arguing that Coherus is driven by hindsight, and it would not have been obvious for any scientist to combine the prior art in the way that Coherus has done.  Amgen previously defeated Kyle Bass’ IPR at the institution stage by essentially making the same argument.  Now, however, Coherus has brought forward new prior art that, it claims, cures the gaps in Bass’ argument.  Because Amgen’s response raises many highly factual and scientific disputes, it is possible that the PTAB may grant institution of the IPRs.  Yet, in this case, given the intricacies of the scientific disputes, that will not necessarily be a strong signal that Coherus will ultimately prevail.  Amgen has several responses, and we briefly summarize the main ones.

First, Amgen argues that neither Watson nor Zettlmeissl teach what Coherus claims that they do.  According to Amgen, the patents require that the entire portion of the hinge region of IgG is included in the fused protein.  But neither Watson nor Zettlmeissl supposedly teaches that.  Instead, the fusion protein in Watson omits five residues of the hinge domain, and Zettlmeissl adds an artificial linker sequence between IgG1 and CD4.  This is illustrated in the image below, which was taken from Amgen’s preliminary response in IPR2017-02066, Paper 7, p. 14.

This is one of Amgen’s strongest arguments because it comes very close to showing that Coherus’ prior art may be missing certain elements required by the patent.  That is typically the type of argument that warrants denial of institution.  Amgen backs up its argument with evidence that the patents should be interpreted to require all of the IgG constant regions except the first domain (CH1).  Amgen and Coherus also appear to quibble over what Watson actually teaches – does it say that fusion near the hinge region was important or was not important?  While Amgen has identified a clear distinction between its protein and the prior art, it may not have done enough to show that these distinctions over Watson and Zettlmeissl necessitate defeating a prima facie showing of obviousness.  In other words, why these distinctions matter.

Second, Amgen also argues that Coherus relies upon prior art proteins that were already considered by the Board and the Patent Office in prior proceedings.  That included original examination of the patents and the previous, unsuccessful IPR filed by Kyle Bass.  The PTAB does, from time to time, reject IPRs solely on the basis that it previously considered the same prior art or the same arguments.  (There is a statutory basis for this under 35 U.S.C. §§ 325(d) and 314(a).)  Even though Coherus relies upon references not previously considered by the Patent Office (Watson and Zettlmeissl), Amgen argues that the fusion proteins disclosed in those references were previously considered.

Prevailing on an equitable argument under §§ 325(d) or 314 that the PTAB previously considered the same or substantially the same arguments is difficult.  And it is fairly discretionary, which means that prior precedent is only so helpful.  In this case, Amgen does a better than average job of showing that the substance of Coherus’ prior art was previously considered by the PTAB, either during examination of the patents or Bass’ IPR.  Yet, Coherus appears to rely upon Watson and Zettlmeissl not just for the proteins they taught, but also for teachings that fusing at the hinge region was optimal or beneficial.  To the extent those precise teachings were not previously considered, that waters down Amgen’s argument that, on equitable grounds, Coherus’ IPR petitions should be tossed.

Third, Amgen argues that Watson does not teach a fusion protein for therapeutically treating inflammation conditions.  Instead, Watson teaches a fusion protein for use as a research tool or probe.  This is a worthy point, and it undoubtedly weakens the force of Watson as prior art.  It suggests that because Watson was really directed to a different scientific problem than the one allegedly solved by Amgen’s patents, then an earlier scientist would not be motivated to combine Watson with any other art to discover the proteins patented by Amgen.  On the other hand, this is a highly factual issue that the PTAB may feel requires further evidence.   In that case, the PTAB may institute the IPRs, but signal that further stress-testing of this argument, through cross-examination of the experts, is needed to adequately resolve it. 

Fourth, Amgen argues that a scientist would not have combined the teachings of Smith and Zettlmeissl.  Even though Zettlmeissl may have taught that it was optimal to fuse IgG to other proteins at the hinge region, according to Amgen, that only applied to proteins without the light chain regions.  But the protein taught in Smith do contain the light chain regions.  So, according to Amgen, a scientist would have had no reason to combine Smith with Zettlmeissl.  This is another motivation to combine argument that may warrant institution, but not guranantee ultimate success by Coherus.

Finally, Amgen faults Coherus for not considering all possible secondary considerations.  Even if an invention is deemed obvious in light of prior art, it can nevertheless be patentable in light of certain secondary considerations.  For example, under certain circumstances, unexpected results or certain types of commercial success can save a patent even in the face of damning prior art. 

Amgen enumerates several secondary considerations that Coherus allegedly failed to address.  Yet, secondary considerations are really the patent owner’s burden to raise.  Unless they were raised earlier, such as during prosecution of the patents, the PTAB will typically not deny institution of an IPR based on secondary considerations raised by the patent owner for the first time in its preliminary response.  Many of the secondary considerations identified by Amgen fall into this category.

Unexpected results is the exception.  Amgen relied upon alleged unexpected results of its protein to get the patents in the first place, i.e., certain unexpected beneficial properties.  Coherus challenges those properties as not really unexpected.  Amgen argues that Coherus’ science is faulty.  This is another highly factual question that will likely come down to a battle of the experts.  Because Amgen’s claim that its protein exhibited properties that were “unexpected” was never fully vetted previously by Patent Office, either during examination of the patent applications or during Bass’ IPR, then Coherus has likely done enough to gain institution of the IPRs, but whether it can ultimately prevail on this argument remains to be seen.