Guest Post by Richard Bone. Mr. Bone is a partner at the VLP Law Group.
The popular view of the PTO’s new AIA reviews, or “patent trials”, is that they have been disproportionately unfavorable to patentees in their outcomes, provoking characterizations such as “patent death squad” for the Patent Trial and Appeal Board (PTAB), the body that administers the trials.
Yet the PTAB’s own statistics paint a different picture: that, in fact, as few as 25% of patent claims challenged are actually held “invalid” by the PTAB, and that figure actually represents only 11% of all of the claims in all of the patents which received at least one AIA challenge. That being the case, patent owners ought to be less fearful than most commentators suggest.
To understand the discrepancy between published statistics and public perception, it’s necessary to dissect the data. Fortunately the PTAB has made various layers of data available on its website. The data is cumulative over all petitions filed from the inception of AIA trials, in September 2012. For all data, the PTAB counts proceedings that have received a final adjudication, whether that be as a result of a settlement, or a request for adverse judgment by the patent owner, or because the PTAB issued a final written decision. Between September 2014 and April 2015, the data was presented on three occasions, as a sequence of bar-graphs condensed into a single graphic. From April 2015, the PTAB has presented the data monthly in a more informative manner, with multiple graphic representations and an accompanying narrative. From both formats, it’s possible to see what is going on, though the numbers mask a number of subtleties.
Just as patents are asserted on a claim by claim basis, so invalidity is adjudicated claim by claim. The overwhelming majority of patent trials to date have been Inter Partes Reviews (IPRs), with almost all of the remainder being “Covered Business Method Reviews” (CBMs). Although the PTAB’s data through January 2015 was presented in aggregate form for the two types of proceeding, the more recent data shows that, statistically, the outcomes of the two types of proceeding are very similar. In both an IPR and a CBM, the petitioner (party challenging the patent) must identify the claims challenged, and advance a ground of invalidity of each. The fee structure for filing an IPR or a CBM includes a challenge to up to 20 claims of a patent, but exacts an additional fee of $200 per claim challenged in excess of 20. Subsequently, if the petition is accepted and trial proceeds, the PTAB levies a further fee of $400 per claim in excess of 15 on which trial is instituted. With this type of fee structure, and accepting that – beyond the filing fees – there is an effort and therefore a cost to prepare a separate argument for each claim challenged, and that there is also an overall page limit for the petition, many petitioners may have been choosy over which claims to challenge. In some cases that choice will have been based on the patent claims asserted in litigation, which need not have been all of the claims in the patent. In others, it will have been based on an assessment that, after demolishing certain well-chosen claims, what remains of the patent is essentially toothless.
So, according to the PTO, through July 2015, in patents for which AIA petitions were filed, 47% of the claims were challenged. This number is little changed from the 45% figure from September 2014. However, the July 2015 numbers, when broken down into the different types of proceedings, show that across all CBMs 58% of patent claims were challenged. This could be because of the longer page limits for CBM petitions (80 pages vs. 60 pages for IPRs) but could equally be because Section 101 challenges are permissible in CBM’s: when arguing invalidity of an independent claim under section 101, it is not significant additional effort to argue invalidity of claims depending from it.
It has to be assumed from the way that the data is characterized that trials recently instituted that have not yet received a final adjudication are not included in the data. Because of the fact that a patent trial typically takes 18 months from filing a petition to issuance of a final written decision, any large scale trend that has happened recently will take some time to be reflected in the PTAB’s cumulative averages.
In any case, just based on claims challenged to date, patent owners could see – on average – patents surviving with more than half of their claims intact, not adjudicated as patentable per se but simply not challenged. Furthermore, the PTAB’s latest charts suggest that the numbers do not include the petitions that were wholly denied, and thus did not result in a trial. Thus, the reported data is based on petitions that resulted in a trial on at least one challenged claim. The statutes governing inter partes review state that trial should only be instituted if “there is a reasonable likelihood that the petitioner would prevail with respect to at least 1 of the claims challenged in the petition”, and this takes into account any preliminary response filed by the patent owner. If the petition was judged not to have met that standard for any claim challenged then trial does not proceed (the petition is “denied”): those claims are not included in the PTAB’s data and hence the overall numbers of surviving claims are actually slightly more favorable to patent owners because trial is not always instituted. It is difficult to calculate this effect precisely because in many instances multiple petitions are filed against a single patent and those several petitions are consolidated into a single proceeding for trial. It is not clear from the PTAB’s data how such petitions that formally do not proceed to trial on their own because they are subsumed into another proceeding are differentiated from those petitions that fall short on all challenged claims and thus lead to no trial on the challenged patent. Nevertheless, comparing fiscal years 2013, 2014, and the first 10 months of 2015, we are seeing a growing number of petitions denied: rising from 14% in 2013 to 32% in 2015, a trend that can only be welcome to patent owners.
The statistics for those claims on which trial is instituted become more complicated, however. As of September last year, the PTO reported that trial was instituted on 66% of challenged claims. By July of this year, that number had dropped slightly to 63% of all challenged claims in all types of petitions: however, CBM’s had an institution rate that was slightly above the average, at 65%.
At this point, looking at the situation from the point of view most favorable to patent owners, trial is only being instituted on 31% of the claims in the patents that have been challenged in an AIA review (and this may be a smaller proportion overall, if it is accepted that a growing number of petitions are being denied outright). It is also fair to say that any claim whose validity was challenged in an IPR or CBM petition and on which trial was not instituted has emerged from that process as a stronger claim, unless the claim escaped that evaluation because the parties entered a settlement even before the PTAB had issued a decision on whether to institute a trial.
Once trial has been instituted, however, patent owners appreciate that the claims proceeding to trial are in considerable jeopardy. For the claims that proceed to trial, the PTAB identifies 4 distinct categories of outcome: held unpatentable (in a final written decision); held patentable (in a final written decision); cancelled (e.g., by patent owner’s amendment), or disclaimed (by patent owner); and “remaining patentable”.
The last of these categories is perhaps the most controversial in the way that the numbers are presented, and one that has been least well understood. It arises from a settlement that precludes the need for a final written decision by the Board. According to the PTAB, by July 2015, 45% of all IPR’s in which trial was instituted were terminated by settlement. One has to assume that a common form of settlement permits the patent owner to retain their patent without an adjudication of invalidity of any of the challenged claims in return for dropping a contemporaneous infringement suit against the petitioner. Claims subject to a settlement in this way have been undoubtedly weakened by the fact of there being a determination of probable invalidity (an institution of a patent trial) in the public record, but they have nevertheless “survived” a formal written judgment from the PTAB.
Claims cancelled by the patent owner are a casualty of the proceeding and arise – most likely – from a desire by the patent owner to minimize the ongoing costs of defending the patent before the PTAB. In any view of the proceedings, an instituted IPR or a CBM is, to any patent owner, an expensive way to lose a patent.
The most important two categories, those claims that are found invalid, or whose patentability is upheld, by the PTAB in a final written decision are those that grab the headlines. Respectively, 42% and 9% of claims proceeding to trial are found invalid and valid. In the grand scheme of things, these are a small fraction of the total number of claims challenged (26% and 6%), and a still smaller fraction of the total number of claims in patents receiving challenges under the AIA. However, what is clear is that patent owners who have decided to fight to the very end are on the whole being severely bruised in the process. The PTO reinforces this point with a break out of 447 IPR trials to date that have “reached final written decisions”: in 2/3rds of them (295 trials), all claims on which trial had been instituted were found unpatentable; by contrast, only in 16% of final written decisions were all claims adjudicated to be patentable. In CBM’s these numbers are even further skewed in favor of petitioners: 79% of final written decisions in CBM’s involved wholesale invalidation of the claims taken to trial, whereas in only 4% were all of the claims upheld. When considered claim by claim, in CBM’s the validity of just 2% of claims are upheld in a final written decision of the Board.
There are a number of questions that the data do not answer: for instance, how many patents sustained a challenge to all of their claims and were ultimately found invalid in their entireties.
But it is clear how the gap between perception and the actual numbers can be bridged: final written decisions are on balance unfavorable to patentees, more so in CBM’s than in IPR’s. It is likely to be the most valuable claims that are contested in this way. Even though such claims represent a tiny proportion of all patent claims that came under jeopardy in an AIA review, it may be little consolation to patent owners that the claims they are left with – if any – are those they would have least interest in enforcing.