Split Decision at PTAB
[UPDATE: January 18, 2019] Mike and others on the Yahoo Boards raise an alternative argument why the PTAB got it wrong: "Instead of trying to show NR in milk was somehow disabled by some mystic bacteria, they should have focused on the fact that NR exists in milk in trace amount. One gallon of milk contains only 0.8mg of B3, of which 60% is Nicotinamide and 40% is NR. On the contrary milk is rich in Tryptophan which is a NAD+ precursor. The concentration of Tryptophan in one gallon of milk is 1600mg. So it is quite obvious that Tryptophan was the active ingredient [in the dog milk] and NR was just trace contaminant.
The Patent Trial and Appeal Board (PTAB) yesterday issued a split decision on Elysium's attempt to invalidate the '086 patent, covering pharmaceutical NR compositions. [Previously, Elysium failed to invalidate ChromaDex's other major NR patent, '807, covering health supplements.]
The PTAB ruled that Claims 1, 3, 4, and 5 of the '086 Patent were invalid as having been anticipated by prior art, because skim milk and buttermilk were known to cure black tongue in dogs in the 1920s, even though NR had not been discovered at the time, and it was not known what caused the cure.
However, Claim 2 of the patent covered isolated NR, and the NR in the dog-curing-milk was not isolated, so Claim 2 remains valid.
You can read the PTAB's decision here:
We'll get into the excruciatingly fine legal analysis shortly, and see whether the PTAB's reasoning makes sense, but first let's just look at the big picture issues, like what does it mean, and what happens next.
What Does It Mean?
There are a lot of reasons why nothing changes as a result of this ruling.
First, there is a decent chance that both sides will want to appeal to the Federal Circuit Court of Appeals, because neither side is likely to be happy with the PTAB's decision, so the ruling could be either overturned or broadened, with significant impact.
Second, there is a federal patent infringement case in Delaware that will consider a broader array of invalidity and non-infringement defenses by Elysium, as well as damage claims by ChromaDex, so the PTAB's decision, even if it were not appealed, does not finally resolve the dispute either way.
Third, the legal impact of the split decision, even if it were the final word, doesn't disturb the parties' positions very greatly. ChromaDex might like to have five claims to assert, or 25, but Claim 2 by itself, plus the entirely of the '807 patent, seem like they are enough to support all the patent infringement actions ChromaDex might want to assert right now against Elysium.
In the long term, if this ruling turns out to be the law, then it creates a potential loophole for pharmaceutical compositions in which NR is present in less than 25% concentration, but Basis is more like an 80% concentration, so that won't help Elysium.
I think of Elysium as having gotten a runner to third base when the inning ended, which is impressive, but worth zero points. Or alternately, Elysium dodged four of the five bullets that ChromaDex shot, but it only takes one bullet to fatally injure your business, and Elysium came up just shy of what they needed.
But even that analysis pays too much attention to the '086 patent and ignores the '807 patent, which is not currently in doubt, and more obviously applies to health supplements, which is what Elysium's only product currently is.
So that's why the result of this is a big Who Cares, and why ChromaDex's share price has not risen or fallen much as a result of the decision -- if anything, the share price has been more stable than usual.
What Happens Next?
The PTAB did a great job explaining its reasoning, but the reasoning it explained is not entirely compelling, so I imagine that both parties will want to appeal, and it appears that they have 63 days to file their notice of appeal, which is March 20. [Sixty-three days sounds like an arbitrary number, but it is exactly nine weeks, which is the smallest number of weeks that gives the parties at least two months, so maybe there is some lawyerly logic in the number 63.]
Elysium probably has some pharmaceuticals dreams, and it will not want to be limited to insipid 25% doses, and it will crave a new audience to argue that the 25% rule isn't rational. So I would expect Elysium to appeal.
ChromaDex doesn't need to resuscitate the four invalidated claims to win against Elysium's Basis, but it might be helpful to have those claims for future battles against future infringers -- maybe even against future versions of Elysium's products, and ChromaDex has already laid out compelling arguments, so it wouldn't be surprising to see them appeal (or cross-appeal, depending on who gets there first).
Indeed, if ChromaDex is is committed to vigorously enforcing its IP rights, then it would be surprising if it did not appeal, unless the PTAB's decision was a preclusive and silencing slam-dunk. Judge McMahon's decision in New York was exactly that sort of a hard blow to Elysium's legal position, so we do not expect to see an appeal in New York, but I don't think the PTAB has done anything comparable to either side.
The PTAB's Decision
I will give the PTAB credit for a thoughtful, thorough, and well-written decision. It seems apparent that they in fact heard and considered the parties' arguments, and maybe I even agree with more than half of what they wrote.
But it's a pretty tough task to completely silence opposing doubts when you argue that a patent on a pharmaceutical fed to humans containing a specific ingredient to achieve a specific affect was anticipated by milk fed to dogs at a time when neither the specific ingredient nor its specific effect were known to exist. How does that teach one the art, which is the whole point of patent disclosure? So here goes.
There were three contested terms at issue to be defined -- "Pharmaceutical Composition," Active Agent," and "Isolated." On "Pharmaceutical Composition," the PTAB sort of went both ways, but ChromaDex won the other two:
The PTAB seems to have agreed with both sides, accepting Elysium's contention that food can be a pharmaceutical composition, and accepting ChromaDex's contention that not every food is a pharmaceutical composition.
Patent Owner’s contention that foods are not pharmaceutical compositions is unpersuasive. Although we agree with Patent Owner that not all foods are necessarily pharmaceutical compositions, the Specification of the ’086 patent expressly teaches that the pharmaceutical composition of the present invention can include food.
That begs the key issue, which is WHEN food is or is not a pharmaceutical composition, and thus that issue was not resolved as a part of the claim construction, but later on in the analysis.
ChromaDex should have won the entire dispute when it convinced the PTAB that the patent applied to pharmaceutical compositions in which NR is an active agent:
We have considered the parties’ arguments, and find that the term “pharmaceutical agent” as used in the present claims calls for the presence of at least one active agent...We agree with Patent Owner that, based on these teachings of the Specification, one skilled in the art would have been able to identify an effective amount of NR for use in a composition. Thus, based on the full trial record, we do not find that Patent Owner’s proposed construction [requiring that NR be an "active agent"] renders the claim indefinite. Based on the foregoing analysis, we construe the term “pharmaceutical composition” to be a composition, including a food composition, which contains NR as an active agent in an amount effective for the treatment or prevention of a disease or condition associated with the nicotinamide riboside kinase pathway of NAD+ biosynthesis.
That's ChromaDex 1, Elysium 0. I credit the PTAB for seeing past Elysium's weird arguments.
I also would have expected that that determination would have been decisive of a favorable outcome for ChromaDex -- how could NR have been an "active agent" in buttermilk? -- so the issue of whether the PTAB in fact complied with its own construction will be central to any appeal.
Elysium lost again on "Isolated." The PTAB said,
Claim 2 recites the limitation that the NR “is isolated from a natural or synthetic source.” In our Institution Decision, we construed the term to mean “that the nicotinamide riboside is separated or substantially free from at least some of the other components associated with the source of the molecule such that it constitutes at least 25% (w/w) of the composition.”
Elysium wanted that to say, isolated means "separated from at least one other component," which ChromaDex mockingly pointed out would include separating the milk from the cow.
PTAB agreed with ChromaDex,
As we noted in our Institution Decision, construing the term “is isolated” as suggested by Petitioner would render the term unreasonably broad in that it would encompass separation of even an insignificant amount of other components. The teachings in the Specification of the ’086 patent counsel against such a broad construction when defining the term “isolated” with respect to NR.
And based on that construction, Claim 2, which applied to isolated NR, survived, because the NR in Goldberger's dog-milk wasn't isolated.
Claims 1 and 3-5
So how could the prior art of feeding dogs some milk in which NR had to be an active agent have constituted prior art anticipating Niagen if no one said -- or even knew -- what the active agent was in the dog milk? That comes down to the concept of inherency.
I'll pause here to note that one of our Elysium Precognition Trolls was close enough to Elysium's legal team that they were able to forecast ten days in advance that the PTAB's decision would turn on the issue of inherency:
That's something short of rocket science, since inherency was discussed several times at Trial, but so were a lot of other things, including the several issues that Elysium lost on.
So here is a paragraph that we will be spending a lot of time on if there is an appeal:
Given that a food, such as the skim milk used by Goldberger et al. can be a pharmaceutical agent, there remains the questions of whether NR was necessarily present in Goldberger’s skim milk and whether such NR was necessarily active in the manner required by our claim construction. (emphasis added)
ChromaDex answers the question, "No," for two reasons.
Although the NR was might have been present and active in Goldberger's dog-milk, it was not necessarily present and active, because:
(1) It might have been degraded, and thus not present, and
(2) It might have been bound, and thus not active
No one tested the milk at the time, and there are no studies that show that NR is necessarily present (not degraded) and necessarily active (not bound or otherwise made non-bio-available) in every milk sample.
Inherency requires that a condition necessarily have been true, not just probably true, and it was Elysium's job to show by a preponderance of evidence that the NR in the milk (1) necessarily was an active agent, and (2) necessarily was present and active in the dog-milk.
The PTAB was convinced:
We find that the references’ teachings support the conclusion that not only is NR a constituent of milk, but that it is active in the production of NAD+.
How did they get to that?
Based on the evidence of record we find Patent Owner’s contention that the NR in milk is not active because Trammell I teaches that NR in milk is bound to other components in milk to be unpersuasive in light of this clear demonstration in the record that NR is an active agent in milk.
Say again? What was the "clear demonstration?"
* The researchers in Trammell I reported that both conventional milk and organic milk contained NR.
* Trammell II reports a study of the bioavailability of NR taken orally. In the background discussion, the researchers report that milk is a source of such NR.
* Canto reports...that “NR is found in milk . . . constituting a dietary source for NAD+ production.”
* Bogan...teaches that “NR is a newly discovered salvageable precursor of NAD+ that occurs in cow’s milk.”
* A filing with the Food and Drug Administration relating to the safety of an NR supplement Niagen teaches that “[h]umans are exposed to NR via dietary sources such as milk.”
* Dr. Baur states in his declaration: As it is now known that blacktongue in dogs is a disease caused by NAD+ deficiency, it follows that the resolution or prevention of blacktongue by milk supplementation, as shown in Goldberger et al., is direct evidence that the milk stimulated greater NAD+ biosynthesis upon oral administration. (Ex. 1005, Goldberger et al., at 1404.) This conclusion is confirmed by later studies, discussed above in paragraphs 13-14, directly demonstrating that oral intake of NR increases NAD+ concentration in multiple tissues. Thus, Goldberger et al. teaches the oral administration of a composition containing NR that necessarily increases NAD+ biosynthesis upon oral administration
Each of these references clearly demonstrates that not only that NR is present in milk, but that it is a source or precursor for NAD+ production.
I disagree. Those references all show that NR is typically present in milk, and is considered an NAD precursor, but none of them says that NR is necessarily present in milk -- that is, it must not be degraded or destroyed -- or that it necessarily is an active ingredient -- that is, it is must not be deactivated or unavailable in some circumstances. Thus, they are not responsive to the issues raised by Dr. Zhou, that NR is not always present and might not be active, and nobody tested the milk back in the 1920's, so we can't be sure.
I would also point out -- although for some reason it doesn't get discussed in the parties' papers -- that there are NAD precursors in milk, such as nicotinamide and tryptophan, so the fact that the dog-milk had AN active agent doesn't mean that NR was an active agent.
So what does the PTAB say about these counter-arguments?
Patent Owner also contends that Petitioner has not shown that the milk used in Goldberger et al. was not degraded by natural bacteria. Resp. 23. We are not persuaded. As Patent Owner points out, Goldberger et al. expressly states that the milk used was fresh skim milk that has been allowed to stand in an ice box for not more than 24 hours. Reply 22–23; Ex. 1005, 1403. The Inventor, Dr. Brenner, in his Declaration filed during prosecution of the ’086 patent stated that NR is “stable for at least 24 hours at room temperature in milk.” Ex. 1003, 132, see also Ex. 1023, 22 (NR levels do not change significantly when milk is stored at room temperature for 24 hours). Based on this admission, we find one of ordinary skill in the art would have expected the level of NR to remain substantially the same when the milk is stored in an ice box. While there may be unusual circumstances in which bacteria in the milk renders the NR inactive, nothing of record suggests that the skim milk of Goldberger at al. is unusual in any way. As such, there is no basis for us to speculate that the NR in such skim milk was degraded or otherwise inactive. Inherency cannot be avoided by taking “extraordinary measures.”
So the question comes down to who is speculating? Is ChromaDex speculating that the NR might have been absent or inactive, or is the PTAB speculating that it probably was present and active?
The key word in that sentence is "probably" because the test for inherency is "necessarily," not probably.
Does the PTAB's evidence show that the NR was necessarily present and active?
I don't think so. Most of the comments they quote are just general statements, like "Milk is a source of NR" and "NR is found in milk," which do not begin to address, or even attempt to address, whether NR is necessarily or always present and bioavailable in milk.
Dr. Baur's statement least of all. He chains together these links in his syllogism:
1. Dog milk cures an NAD deficiency
2. NR cures an NAD deficiency
3. Therefore, Dog milk necessarily contains NR that cures an NAD deficiency
Any college undergrad in an Intro to Logic class would tear that to shreds. You obviously don't need to take Intro to Philosophy to get admitted to grad school in the natural sciences.
But the PTAB consists of lawyers that shouldn't fall for such shoddy reasoning.
Admixture I also suspect the PTAB might have gotten it wrong on "admixture." It's a weird word, archaic and/or technical. Presumably it's different from "mixture." Possibly in the dictionary "mixture" means something that IS mixed, and "admixture" means "the action of adding an ingredient to something else."
If that were right, then the PTAB should not say, as it does,
We find nothing in the record that supports Patent Owner’s contention that the term “admixture” requires that the ingredients be purposefully mixed.
Maybe not in the record, but in the dictionary, which would be enough.
But if you don't like semantic battles, a different approach leads to the same conclusion. The word "isolated" is also present, and it's hard to imagine what isolated could mean if its opposite is also present, mixed. It would make more sense to interpret the phrase "isolated" to mean extracting the NR, and then "admixture" to mean adding it, rather than accepting Elysium's weird shortcut that that something is isolated while still in its natural state, and also admixed without anyone doing anything. It would seem that those interpretations, one of which (admixed) the PTAB accepted, disregard the claim language.
A funny thing happens near the end of the written decision. On page 36, there is a sentence fragment, which I have highlighted in yellow:
Goldberger and Tanner expressly states that the buttermilk used was “fresh, locally produced, and of fair quality.” Ex. 1006, 93. As such, there is no basis for us to speculate that the NR in such buttermilk Based on the foregoing analysis, we conclude Petitioner has established that the buttermilk used in Goldberger and Tanner was a pharmaceutical composition comprising NR as an active agent in an amount effective for the treatment or prevention of a disease or condition associated with the nicotinamide riboside kinase pathway of NAD+ biosynthesis.
That looks like a word processing error, which is a sign that a more aggressive editorial process hit that portion of the document. And so should it have.
The PTAB is trying to say that the dog milk was just like any other milk, because it was "fresh, locally produced, and of fair quality," as opposed to milk that had spoiled or for another reason lacked normal properties.
But that begs the question of whether the normal properties of milk necessarily caused the effects seen in the Goldberg experiment, which they might have, but there is no evidence that they did, and some evidence that they might not have. The presence of alternate NAD precursors, along with the evidence in the record that NR might not be active, ought to be sufficient to defeat inherency, I would think.
Which takes us back to the malformed sentence, "there is no basis for us to speculate that the NR in such buttermilk."
I would have said just the opposite, that there is no basis to speculate that the NR in the buttermilk WAS present, WAS active, or CAUSED the observed effects in the dogs. Entirely possible, but there is no evidence in the record to demonstrate that it was necessarily so, and the evidence cited by the PTAB does not even attempt to make that case. Not even Elysium's own expert plausibly made that case: Even though he used the word "necessarily," he introduced it awkwardly into his conclusion, without its being even implicit in his premises.
My guess is that that paragraph got re-worked a couple times in an attempt to preemptively demonstrate to the Federal Circuit that they applied the correct standard. But saying it does not make it so, and failing to say it because you can't wordsmith around the problem is worse.
At best, Elysium dodged four of the five arrows in ChromaDex's quiver, but was felled by the last one, unless they can convince the Federal Circuit to adopt an interpretation of "isolated" that is so nonsensical that it was even rejected by the PTAB, which is the most sympathetic audience Elysium will ever have.
With a little luck, the Delaware Court will take this ruling as cause to proceed with the patent infringement case on both patents, so we can actually get to the bottom of the dispute, although Delaware is slow right now (I hear) and the Trump-McConnell government shutdown is not helping (I heard that some clerks in some federal courts are getting told to stay home since they are not getting paid).
I predict that both sides will appeal, since both sides have a lot to be disappointed about with this decision. Elysium will be praying to Jude, the patron saint of lost causes. ChromaDex has a decent chance of joining the 25% of appellants who achieve complete reversal of the PTAB's decision, but, as with all litigation, the only sure winners are the lawyers.