By Michael Borella and Ashley Hatzenbihler --
There is ample evidence that patent examiner allowance rates vary dramatically from examiner to examiner and art unit to art unit.[1] This has resulted in the general understanding that there are "easy" examiners and "tough" examiners.
Naturally, there is little complaining about "easy" examiners (until you are defending against a patent with very broad claims, at least). But when patent attorneys stand around the coffee machine at work (or, more accurately, are on Zoom or Teams calls), there is a not small amount of eye rolling and wringing of hands over "tough" examiners.
Over the last several years, a particular type of "tough" examiner has emerged -- one that will reject just about any claim as ineligible under 35 U.S.C. § 101 . . . and never withdraw the rejection.[2] In other words, if you get a 101 rejection from one of these examiners, prosecution is basically over. Interviews don't help. Arguments and amendments go nowhere, with the examiner engaging in vigorous goalpost-shifting.[3] More often than not, the natural habitats of such examiners are the verdant plains of technology centers 1600 and 3600.
The USPTO's Patent Trial and Appeal Board (PTAB) is often a last resort for patent applicants who cannot afford further appeals to the federal judiciary. Therefore, should not the PTAB lay the judicial smack down on these errant plains-dwellers? If only.
Last year, we studied all substantive 101 decisions from the PTAB that came down in 2021. The results were puzzling, striking, and, not in small part, abysmal. The PTAB affirmed examiners' 101 rejections 87.1% of the time, whereas the overall affirmance rate across all grounds of rejection was 55.6%. Why the discrepancy? The data does not provide that information, though it could be due to the notorious vagueness of the test set forth in Alice Corp. v. CLS Bank Int'l, the Federal Circuit's conflicting case law, and/or the PTAB not following the USPTO's 101 guidance.
Regardless of cause, the numbers do not lie. The PTAB is a brutal tribunal for applicants attempting to argue that an examiner's 101 rejection is in error.
But was 2021 an outlier or a blip on the radar? Does the data from 2022 exhibit a similar affirmance rate or has there been a "regression to the mean" of sorts?
To answer this question, we once again reviewed every substantive 101 appeal decided by the PTAB in 2022. As was the case for the 2021 data, this required a particular search strategy as well as manually combing the text of each decision.
From the PTAB's search interface, we specified the following criteria: decision dates between January 1, 2022 and December 31, 2022, a proceeding type of "appeal", a decision type of "decision", and an issue type of "101". Even so, the results were over-inclusive, returning decisions that just mentioned 101 in passing. Thus, we further filtered these decisions to focus only on those in which the applicant appealed an examiner's Alice-based 101 rejection and the PTAB ruled on this grounds of appeal. This took the 634 decisions returned by the search engine down to 482 -- the substantive 101 decisions.[4]
Of these, 426 were examiner affirmances, for an affirmance rate of 88.4%. Yep, the PTAB got a little worse for applicants in 2022. The month-by-month and total statistics are provided in the table below.
Not unlike 2021, the examiner affirmance rate fluctuated with no clear month-over-month trend. For example, the lowest 101 affirmance rate was in May (77%) while the highest was in October (100%).[5] Also like 2021, there is a slight downward trend in the number of appeals from Q1 to Q4. The monthly average of appeals on 101 grounds was 49.5 for the first four months of the year and 34.25 for the last four months of the year (as compared to 78.25 and 49.25, respectively, for 2021).
But one data point that sticks out is that the total number of appeals has dropped significantly year-over-year. In 2021, 708 substantive 101 decisions were handed down, while in 2022 this number was -- as noted above -- just 482. That is better than a 30% decline. From our (admittedly anecdotal) experience, this is not because examiners are getting easier. The opposite appears to be the case. It seems more likely that applicants have determined that appealing 101 rejections to the PTAB has betting odds akin to those of the Cleveland Browns making it to the Super Bowl.
Given that the PTAB does not provide much of a recourse for applicants stuck with a 101 rejection from an unrelenting examiner, we next wondered what grounds of rejection are popular with the PTAB. For the abstract idea exception to patentable subject matter, the three main categories are mathematics, mental processes, and methods of organizing human activity. In other words, a claim is deemed ineligible for patenting if it is directed to mathematics, a mental process,[6] or a method of organizing human activity[7] without significantly more.
Of all substantive affirmances of 101 rejections by the PTAB, 14.1% were based on mathematics, 58.4% on mental processes, and 64.6% on methods of organizing human activity. Use of the latter two categories is quite widespread among PTAB judges, with many decisions incorporating new grounds of rejection to accentuate the examiner's mental process rejection with a method of organizing human activity rejection or vice-versa. Indeed, grounds of both mental processes and methods of organizing human activity were found in 31.9% of all affirmances.
Another factor we looked into was whether certain PTAB panels or judges were making formulaic rejections -- in other words cutting and pasting large sections of their reasoning for the 101 rejections from case to case and only changing the discussion of the facts. It was not too difficult to identify a few instances of this.
For example, Appeal 2021-002509 (decided January 31, 2022) and Appeal 2021-002913 (also decided January 31, 2022) involve two different applicants claiming two different technologies, but the decisions rely on reasoning that is largely word-for-word the same. The deciding PTAB panels had two judges in common and both decisions were written by the same judge. Another example of liberal cutting-and-pasting can be found in Appeals 2021-002840 and 2021-002807, also both decided on the same day by a panel with two judges in common.
It is important to understand that this does not imply laziness or malfeasance on the part of anyone involved in these decisions. Instead, this is more evidence that it is very easy to find virtually any invention ineligible by robotically deconstructing the claims into small enough parts and ignoring the advantage or improvements provided the claims as a whole. Not convinced? Try our rationale for invalidating the eligible claims of Diamond v. Diehr that is based on observed USPTO reasoning.
There is a false narrative that has been floating around for the better part of two decades. It implies that broad claims on obvious technological variants can be easily obtained from the USPTO. There may have been some small truth to this notion in the 1990s, but today the pendulum has swung so far in the other direction that narrowly-scoped, complex, innovative technologies are often denied patentability just because they involve software.
As a consequence, the irrational arbitrariness of Section 101 provides limited access to patenting for individual inventors as well as small and mid-sized companies. Maybe someday the PTAB will lay the judicial smack down in a less one-sided fashion, but for now it appears that 101 appellants need to be wary against flying elbows of ineligiblity.
[1] See the histogram of examiner grant rates compiled by Patent Bots, for example.
[2] Okay, the word "never" is admittedly somewhat hyperbolic, but not far from the truth.
[3] Again, we want to emphasize that most examiners are reasonable, even with their 101 rejections. Some are quite clear in terms of what they deem to be sufficient to overcome a 101 rejection. But there is a notable contingent who are not, and anecdotally it seems as if the size of this contingent has grown since early 2021.
[4] Many of these decisions also reviewed rejections on other grounds (e.g., Sections 102 or 103). We did not consider anything but the 101 determinations. We also omitted rejections based on laws of nature or natural phenomena, which accounted for only 10 of the decisions and did not impact the results in any significant fashion.
[5] Ouch. Just ouch. At least relatively few appeals were decided in October.
[6] We have previously detailed how illogically and broadly that this category is being construed.
[7] This is another overly-broad category. What isn't a method of organizing human activity at some level?
The New York Times Is at It Again Regarding Patents
By Kevin E. Noonan --
The problem seems to be that Humira has made a pharma company a lot of money (purportedly $116 billion), that the drug is expensive (said to cost upwards of $50,000/year) and that the drug company has amassed a large number of patents to protect its intellectual property. The bigger problem is that the article fails to recognize several important facts relating to the circumstances under which Humira's makers made this money and amassed its patent estate (or "thicket" as the anti-patent crowd likes to call it).
The first of these is that until 2010 there was no pathway for "generic" (accurately, "biosimilar") competition for innovator biologic drugs (the class of drugs including Humira). This has nothing to do with patents; the FDA could not approve a biosimilar competitor by law until the Biologic Price Competition and Innovation Act (BPCIA) was passed as part of Obamacare (see "Follow-on Biologics News Briefs - No. 11"; "House Passes Health Care Reform Bill -- Biosimilar Regulatory Pathway Makes Cut, Pay-for-Delay Ban Does Not"). Thereafter, the FDA needed to develop guidelines (see "FDA Looks to Multiple Sources, Including EMA Guidelines, in Developing Biosimilar Approval Standards") and Guidances (see "FDA Publishes Draft Guidelines for Biosimilar Product Development"; "FDA Releases "Final" Guidances for Industry regarding the Biosimilar Approval Pathway") establishing the standards under which biosimilar drugs could be approved; while still on-going for certain types of biosimilars these efforts took about 5 years to be promulgated (the FDA engaging with stakeholders to ensure the efforts were fair and robust enough to minimize the possibility of approving drugs that were not "similar enough" to avoid safety, potency, or efficacy issues).
The first approved biosimilar was Zarxio from Sandoz, which competes with Amgen's Neupogen, an adjuvant for cancer patients for treating side effects of chemotherapy (see "FDA Approves Sandoz Filgrastim Biosimilar"). Humira biosimilars have been pursued by several companies and FDA has approved seven (see, e.g., "FDA Approves Amjevita -- Amgen's HUMIRA® Biosimilar") and this table:
(With one of them, Cyltezo, being designated as an interchangeable biosimilar, see "FDA Approves Another Interchangeable Biosimilar Drug", which is highly sought after because it has advantages, see "FDA Issues Final Guidance Regarding Biosimilar Interchangeability", similar to what can be achieved for a small molecule drug.)
Immediately it will be clear that a great deal of the $116 billion cited to raise the temperature of the debate (and the purported perfidy of Humira's producer) was made during the time that it was impossible to compete, and that once a pathway had been opened several companies took the steps to do so. But although there was this large patent estate accumulated when litigation ensued, the number of patents asserted and claimed was but a minuscule portion of the estate. As for types of patents accumulated, much is made in the article about the patent on the active pharmaceutical product (the drug) expired in 2016, as if that was the only patent upon which Humira was entitled to rely. The first source of error in this assertion is that most of the current biologic drugs were approved so long ago that they similarly no longer have patent protection on the drug molecule itself. But due to the complexity of producing these drugs commercially, they all have protection on those methods (without which the drugs could not be produced and regarding which each sponsor company invested money, time, and effort to develop). Humira is not alone nor an outlier on such protection and these patents protecting how the drugs are made are no less worthy than the drug patent itself.
The irony of ironies in this story printed this Sunday is that this Tuesday, January 31st, those seven Humira biosimilar-approved companies will be able to sell their biosimilar Humira free of all the patents in the patent estate, pursuant to a settlement agreement (see, e.g., "HUMIRA® Biosimilar Update -- Settlement in AbbVie v. Amgen Case Announced and AbbVie v. Boehringer Ingelheim Litigation Begins"; "AbbVie Announces Global Resolution of HUMIRA® (adalimumab) Patent Disputes with Sandoz"), with others as yet not approved being licensed to go on the market on July 1st and September 30th of this year. As mentioned above, the overwhelming number of the Humira patents were not asserted but could have been asserted to prevent these Humira biosimilars from being marketed for many more years. Inconsistent with the bad guy caricature promulgated by the Times, the cases were settled, benefiting the public at the real-world expense of Humira's drug maker.
Two other points bear mentioning. First, the 7th Circuit Court of Appeals decided last year that the Humira patent estate was not an antitrust violation and thus fear of antitrust liability is not a factor in these settlements (see "Mayor and City Council of Baltimore v. AbbVie Inc. (7th Cir. 2022)"). The second point is that the FTC in a 2009 white paper (see "No One Seems Happy with Follow-on Biologics According to the FTC") predicted that the price reduction benefit of biosimilars would be about 30%, that is that the cost of a biosimilar equivalent to a patented biologic drug would be about 70% of the reference biologic drug price. With admittedly few data points, that prediction has been borne out so far, meaning that instead of $116 billion the cost for an equivalent period of time an amount of Humira biosimilar sales can be expected to be $82 billion. As Sen. Dirksen would say that is real money but hardly the type of windfall public benefit that small molecule drug generics represent (which sell for about 10% of the innovator price).
The real issue is that the development cost of biologic drugs is much higher than traditional small molecule drugs as is the cost of producing them. Everyone thinks "drugs cost too much" and want them to be cheaper but the reality (in a capitalist society) is that there needs to be sufficient prospect of return on investment to justify development. The entire economic argument is TLDR (which is why articles like the one in Sunday's Times is both easy and incomplete); for a good and accurate explication of the patent side of this issue, Professor Adam Mossoff at George Mason University has published a report for the Hudson Institute (see "Unreliable Data Have Infected the Policy Debates Over Drug Patents"), which, while generating less heat than the Times article, does shed enormously more light (see "Faux-Populist Patent Fantasies from The New York Times"). Suffice it to say that while it may make the medically self-righteous feel better, it does little to advance the real debate about how to ensure that people who need drugs (and medical care generally) can get them. That's too complicated for a lazy winter Sunday afternoon reading the Times, but the issue deserves more than this shallow level of analysis and rhetoric.
* Where you are likely to be talking to yourself in under five minutes if you bring up the subject.
Posted at 11:47 PM in Media Commentary | Permalink | Comments (8)