Beware the “on-sale bar”

killed by the on-sale bar
Victim of the on-sale bar

A patent is invalid if the invention was sold more than a year before filing (on-sale bar), and the recent “Helsinn v Teva” case shows that the courts can be harsh.  

A patent applicant can accidentally ruin their own patent in various ways.  In the US, one error is to first sell an invention (presumably in the form of a product with the same claims), and then wait more than a year to file the patent. This error violates the 35 USC 102 on-sale bar of classic (pre-2012) patent law. This error also violates the latest 2012 AIA version of patent law, which phrases this as “on sale, or otherwise available to the public”.

However, the relationship between an “invention” and a “product” isn’t always clear, and it also isn’t always clear if a “sale” has taken place.  Is the product identical to the invention’s claims?  If I privately show you a cardboard box and say “I have an invention inside, want to buy it sometime if it works?”, and you say “Maybe”, is this a sale that invalidates a later patent?  In the event of doubt, how will the courts rule?  Will they err on the side of protecting the patent, or invalidating the patent?

In their May 1, 2017 “Helsinn Healthcare via Teva Pharmaceuticals” decision, the Federal Circuit took a harsh and patent unfriendly approach. Indeed, this ruling was so harsh that Lamar Smith, the Congressional sponsor of the 2012 AIA law, stated that the court was ignoring the intent of Congress.

To greatly simplify the Helsinn case: back in 2001, Helsinn was doing FDA clinical trials on the efficacy of various palonosetron drug formulations to reduce nausea during chemotherapy. During these trials, they signed a supply agreement with MGI (another company) stating that if the FDA approved some of Helsinn’s various drug formulations, and if MGI subsequently made purchase orders for these drug formulations, and if Helsinn subsequently accepted these purchase orders, then Helsinn would sell the drug to MGI. In 2003, after the clinical trials were successful, but before FDA approval, and before any actual product changed hands, Helsinn began filing for various patents.

Teva, a competitor, decided to challenge these patents as being invalid due to the “on-sale bar”. But was there really a sale? Was there an invention yet? Was it disclosed to the public? A lower court ruled in favor of Helsinn, but the Federal Circuit reversed.

The Federal Circuit used the Uniform Commercial Code (UCC – a set of laws intended to “save” ambiguous contracts by automatically supplying missing terms) to argue that there was a “sale” despite all the “ifs” and ambiguity. They then argued that the invention was in the (undelivered) product and that the invention existed before clinical proof that it actually worked.  Strangely, they even argued that the AIA “on sale, or otherwise available to the public” language didn’t apply because… this would change past practice. Apparently Congress, (despite good reasons and clear intent) somehow doesn’t have enough authority to make these changes?

The case is presently being appealed, but the moral is: be careful.

Thales patent eligibility ruling

Thales Visionix head mounted display
Not the Thales Head Mounted Display!

A recent Federal Circuit case, Thales Visionix Inc. v United States, continues the process of restoring sanity to the ongoing “Alice” patent eligibility mess.

As previously discussed, since the Supreme Court’s 2014 “Alice” decision, patent law has been burdened with an unworkable “is it abstract?” test for patent eligibility. This test is similar to a medieval test to determine witches:  Step 1:determine if the test subject is a witch abstract”;  Step 2:if so, does the witch float? is there something more?”

The Federal Circuit, charged with cleaning up patent law, sometimes chips away at this nonsense, and sometimes adds to the nonsense. In two earlier cases (Enfish and Rapid Litigation), the Federal Circuit established at least a few reasonable step 1 rules (such as read the entire claim), and now we have another.

Thales Visionix had a patent on a motion-tracking Head Mounted Display (HMD).  This patent claimed a HMD arrangement of inertial sensors and signal processing elements, used in the F-35 fighter jet. Any enemy pilot blown up by this HMD system might not consider this to be “abstract”, but a lower US court was not so easily impressed. They used their own “Alice: it’s abstract” weapon to shot down the HMD patent.

In the absence of rules (the Supreme Court thoughtfully declined to provide any), lower courts have often used a type of “guilt by association” logic, where if a given patent claim has some elements in common with another claim previously ruled to be abstract, then that claim is also abstract. The lower court argued that the Thales patent claims were abstract because the claims allegedly “used mathematical equations (previously determined to be abstract) for determining the relative position of a moving object to a moving reference frame”.

However. the Federal Circuit disagreed.  They determined that just because a claim contains a patent-ineligible (abstract) concept (e.g. mathematical equations) does not mean that the entire claim is (step 1) “abstract.” Rather, the question is if the patent-ineligible concept (math, natural law) is being used to improve some other technique. If so, then the claim as a whole is not abstract. This is similar to their earlier, more biotech-focused, ruling in the Rapid Litigation case.

So as a medieval logic analogy, just because a woman has a cat does not automatically mean that the woman is a witch, if the cat is unusually good at catching mice. If improved mouse catching can be shown, the step 1 conclusion is that the woman is not a witch. There is no need to go on and subject the poor woman to a step 2 “witch float” test.  This is good because there is a high casualty rate at step 2.

Sending Alice to Planet Blue

Planet Blue -- Neptune perhaps?
Planet Blue — Neptune perhaps?

Forget “patent invalid because abstract”, the Planet Blue decision finds the real issue is preemption, and non-preemptive software claims are not abstract.

As previously discussed, in recent years the Supreme Court (SCOTUS) issued a series of confused rulings such as the “Alice” and “Mayo” decisions.  SCOTUS ruled that “abstract” inventions were not patentable, but didn’t define “abstract”, thus making the issue very subjective and bringing much confusion to software and biotech patents.

Fortunately, perhaps realizing that they did more harm than good, SCOTUS has recently started to decline further cases of this type.  They are apparently now leaving it to the Federal Circuit (the court right below SCOTUS) to clean things up.

The Federal Circuit has never known what to do with this, and for every step forward there is at least one step backward. Some damage control efforts, include their “Enfish”, “Bascom”, and “Rapid Litigation” decisions.   Another case, “Mcro, Inc. v. Bandi Namco Games, provides a bit more damage control. (Mcro calls itself “Planet Blue”, so we will call this the “Planet Blue” decision.)

The “Planet Blue” patent involved some improved methods of lip synchronization for animated cartoons, and their claims covered some novel but general rules to do this.  A lower court had originally invalidated these as being abstract, but the Federal Circuit reversed and said that the claims were OK.

As per their earlier Enfish decision, the Federal Circuit did not simply assume that software is “abstract” (under Alice step 1).  Instead, they asked the broader question, why is “abstract” a problem anyway?  As they interpreted it, “The concern underlying the exceptions to § 101 is not tangibility, but preemption.”  [Emphasis added]

Some history about patents and “preemption”: Back in1853 Samuel Morse, inventor of the telegraph, tried to get a claim for the use of electromagnetism for any method of printing characters or signs. However, this particular claim omitted any other details. The 1853 SCOTUS said “no”, because claims where “it matters not by what process or machinery the result is accomplished” (e.g. are preemptive) are not going to be allowed.

In Planet Blue, the Federal Circuit (possibly giving up on the more recent rulings as being hopelessly confused), has essentially gone back to the 1853 Morse case for some coherent guidance.  The Federal Circuit ruled that with regards to the Planet Blue claims, “The claim uses the limited rules in a process specifically designed to achieve an improved technological result in conventional industry practice… Claim 1 of the ’576 patent, therefore, is not directed to an abstract idea.” In other words, this passes Alice “step 1”, and you get a free “get out of abstract patent eligibility rejections” card.

The advantage of this ruling is that “preemption” is a more objective issue, and by changing the analysis from “abstract” to “preemption”, it would be possible to remove a lot of randomness from the patent process. If sustained (and this is a big “if”), it would be good news for software patents, biotech patents, and indeed all types of patents.

Rapid Litigation: Biotech patent win

Rapid litigation and biotech patents
Rapid Litigation v Cellzdirect: Federal Circuit cleaning up a patent mess

In Rapid Litigation v Cellzdirect, the Federal Circuit further clarified a biotech patent eligibility mess caused by SCOTUS’s Alice and Mayo decisions.

In 2012 and 2014, the Supreme Court (SCOTUS), which wanted to invalidate a few pesky financial and medical diagnostics patents, decided to use legal “nuclear weapons” rather than legal “fly swatters“.  The SCOTUS legal nuclear weapons were the now-infamous two-step “abstract material” patent eligibility test.  Step 1 of this test is summarized below:

Step 1) Determine if the patent is directed to “abstract” subject material.  If not then congratulations, the patent passes this test and is patent-eligible.  

SCOTUS refused to define “abstract material”. This allowed the USPTO and the lower courts to run amok and start randomly invalidating patents, including biotech patents. This, in turn, started to damage US biotech and software startups, which need patents to get funding, as well as for protection from larger competitors.

The Federal Circuit, one step below SCOTUS, and charged with cleaning-up patent law, waited several years for SCOTUS to correct itself, to no avail.  The Federal Circuit is itself confused about these issues, and their rulings largely depend on which judges hear a particular case. However, in mid-2016, they issued a series of patent eligibility decisions, first Enfish, then Bascom, now Rapid Litigation (Celsis) v Cellzdirect that suggest that at least some of the judges want to do some damage control.

“Abstract” is just as undefined for biotech as it is for software, but for biotech, it seems have more of a “natural law-ish” flavor.  Arguably a technologically illiterate approach since everything involves natural laws, but alas SCOTUS has no STEM majors.

Fortunately, a few Federal Circuit judges are STEM majors, and they have the power to do at least some damage control by providing official interpretations/clarifications of SCOTUS decisions. Here they clarified that in step 1, “directed to” is not the same thing as “involving”.

The patent in this case, 7,604,929, was about an improved method of freeze storing liver cells (cryopreservation of hepatocytes).  Before the patent, everyone in the field believed that freezing damages living cells, and that multiple freeze-thaw cycles should be avoided.  The inventors discovered that some hepatocytes were resistant to this problem, and used this discovery to invent an improved hepatocyte cryopreservation method.  This method first freeze-thawed the cells, then used a density gradient to select for the freezing resistant cells, and then refroze these resistant cells again, producing very freeze resistant hepatocytes.

An earlier court had ruled this patent invalid by arguing that discovering that some hepatocytes could survive multiple freeze-thaw cycles involved a “law of nature”. However, the Federal Circuit pointed out that “involved” isn’t enough grounds to invalidate a patent, because the ‘929 methods also claimed other steps, such as using density gradients.  So “directed to” is more than just “involves“.

This Federal Circuit decision makes it harder to invalidate a biotech patent, at least in the most common cases where the claims also have other steps in addition to the “natural law” steps.  The fact that the ‘929 claims were relatively simple helps to further clarify the legal issues.

Bascom: another Alice software win

Bascom - 35 USC 101 Alice abstract test step 2 is like a 35 USC 103 obviousness test
Bascom – fish delivering mail

The Bascom court decision helps software patents by suggesting that step 2 of the Alice test should follow established obviousness rules.

US software patents got another win from the Federal Circuit Court this week.  This court, which has a Congressional mandate to clarify patent law (actual results may vary), made an important clarification to the Alice (software patent killer) “abstract subject matter” test in the Bascom v AT&T case (Bascom).

As you may recall from our last “Enfish” episode, the Supreme Court (SCOTUS) “Alice” decision created a judge-ordered incoherent two-step test that any given patent must pass in order to be patent-eligible under 35 USC 101 rules.  To simplify:

Step 1) Determine if the patent is directed to “abstract” subject material.  If not then congratulations, the patent passes this test.  

Step 2) But if the patent is found to be “abstract” in step 1, then determine if the patent contains “something extra” beyond just “abstract” subject material.  If there is nothing extra, then reject the patent as being “abstract”.   

The Federal Circuit is one step down from SCOTUS.  The Federal Circuit doesn’t have enough clout to overrule SCOTUS, but does have enough clout to overrule the USPTO and the lower courts.  More importantly, the Federal Circuit has both the clout and mandate to “clarify” SCOTUS decisions. Sometimes they do so, sometimes they make it worse. But I think they got it right in this case.

The USPTO and the lower courts were often ignoring Alice step 1 and just assuming that a patent had “abstract” subject matter. The previous Federal Circuit “Enfish” decision pointed out that this was improper.  By contrast, the newer Federal Circut “Bascom” case is now clarifying that another common practice, just asserting that the patent fails Alice step 2 because it lacks an undefined “something extra”, isn’t right either.

More specifically, in Bascom, the Federal Circuit pointed out that based on earlier SCOTUS decisions (e.g. Mayo v. Prometheus, which SCOTUS used for “Alice”), Alice step 2 tests if the claim is “well-understood, routine or conventional”.  According to the Federal Circuit’s interpretation, SCOTUS was probably thinking about something similar to an obviousness test.   The Federal Circuit also pointed out that there are well-established rules for establishing obviousness, which the USPTO and the lower courts were (also) ignoring.

Specifically, the Bascom case was an appeal of a lower court decision that had earlier found the Bascom patent claims to be “abstract” and therefore invalid.  The lower court’s arguments (in simplified form) were that the Bascom patent claim language words described conventional computer pieces, and therefore the Bascom claims failed Alice step 2 due to lack of “something more”.  

In Bascom, the Federal Circuit Court, after “clarifying” Alice by pointing out that SCOTUS’s Alice step 2 resembled an obviousness test, then pointed out that the lower court had failed to follow established rules to determine obviousness (35 USC 103 rules).  These 35 USC 103 rules require that the combination of the pieces and the motive for combining the pieces also be considered.  Here conventional pieces, arranged in a non-conventional way, are often not obvious.

The Federal Circuit then looked at the Bascom claims, determined that they were not obvious, and (again somewhat simplifying) therefore had “something more” that satisfied step 2 of the Alice test.  They then overruled the lower court and found the Bascom patent to be “not abstract” and therefore valid under 35 USC 101.

If there were more cases like the Bascom decision, it might almost introduce some sanity to the Alice test. Sadly, for every coherent Federal Circuit ruling, there is another incoherent ruling.  Stay tuned…

Federal Circuit rules against PTAB “Chewbacca defenses”

European patent oppositions: litigation by unfair surprise?
Chewbacca defense?

Litigation by unfair surprise is out, at least for US patent PTAB cases — for European patent oppositions, not so much. 

One thing that I will never forget about my experience with European patent oppositions is that to American eyes, the European process appears to be rather “due process of law” impaired.

Europe has a system by which issued patents can be challenged. This system is called a patent opposition. In the US, the counterpart to this is the Patent Trial and Appeal Board (PTAB), which can conduct Inter Partes Review (IPR) hearings.

The two systems have various differences. For example, although in theory, issues should be argued in advance by written briefs, the European opposition process also allows parties to introduce new issues during the last-minute oral arguments.

This allows for litigation by “unfair surprise”, rather than by reasoned arguments.  In this sort of setting, I have seen that illogical but last-minute “Chewbacca defenses” can work quite well.  The clock is ticking, proceedings are going to finish in an hour, and suddenly you have to discuss entirely new and logically irrelevant issues. It is as if you suddenly have to shift gears and focus on if the Star Wars character Chewbacca lives on the planet Endor or not.

In this regard, it is refreshing to see that the Federal Circuit in Dell Inc., v. Acceleron, LLC (March 15, 2016) has confirmed that this sort of litigation by unfair surprise is unacceptable for US PTAB patent reviews. Due process wins!  That is, arguments made during oral argument must be restricted to only those arguments previously discussed in writing beforehand.  Thank you, Federal Circuit.

Part of why the European patent opposition process permits unfair surprise may be due to the difference between European civil law practice, and American (and UK) common law practice. The European civil law system, which generally does not use juries, assumes that the judges are all-wise, and therefore need little protection against bad input (irrelevant or improper evidence).  By contrast, the common law system, which is jury oriented, has greater protections (rules of evidence) against bad input.

Since all judges are human, a key difference between the two systems is that, at least for patent opposition matters, the European system is running with fewer quality control measures.

I prefer the American system, thanks. In any event, the take-home lesson is that for international patents, sometimes there can be unexpected differences between countries.