Are the findings of fact in Teva v. Sandoz clearly erroneous?

The  so-called "peak average molecular weight" mentioned in Teva v. Sandoz is more accurately termed the mode of the distribution.  It is not an average.  This is a major issue of fact in the case.
[As  LBE said at PatentlyO:  As to an issue identified by the CAFC, the claim talks about an “average” molecular weight, but Mp is a mode, not an average. Mp is not a species within the genus of average molecular weights. This is a clearly erroneous finding of fact. Just because the examiner was fooled does not remedy the error.]

What exactly are the findings of fact which the CAFC needs to review?  The Supreme Court dissent might suggest there aren't any.


As to the recently-decided Supreme Court case in Teva v. Sandoz, 2015 U.S. LEXIS 628, one recalls the last words of the dissent:

Although it relied on expert testimony to understand the science underlying petitioners’ claims, the District Court made no “findings of fact” as that term is used in Rule 52(a)(6). 

One might question whether crediting opinions of expert establishes "findings of fact"; assuming that opinions may be deemed facts, one readily sees that the opinions/facts in the Teva case. are false and thus are clearly erroneous.  It is foreseeable that the CAFC will find the underlying facts relied upon by SDNY to be clearly erroneous.  

Given that the majority opinion in 2015 US Lexis 628 seems to suggest that a district court's reliance on expert opinions is subject to clear error review, one expects a consequence of the majority decision will be more expense directed to expert opinions and expert opinions to challenge expert opinions.    This becomes an interesting avenue to transmogrify "purchased opinions" into something which can be treated as facts.

A separate consequence will be re-evaluation of a party stipulating to infringement to seek review at the CAFC (eg, Gevo v Butamax).  

Problems with the facts of the SDNY decision:

From Wikipedia:

The mode is the value that appears most often in a set of data. The mode of a discrete probability distribution is the value x at which its probability mass function takes its maximum value. In other words, it is the value that is most likely to be sampled. The mode of a continuous probability distribution is the value x at which its probability density functionhas its maximum value, so, informally speaking, the mode is at the peak.

(...)

The mode is not necessarily unique, since the probability mass function or probability density function may take the same maximum value at several points x₁, x₂, etc. The most extreme case occurs in uniform distributions, where all values occur equally frequently.




Thus, when SDNY interpreted "average molecular weight" to be the mode of the molecular weight distribution, it represented an average by something which is not an average:



The Court construed "average molecular weight" to mean "peak molecular weight detected using an appropriately calibrated suitable gel filtration column." (CCO at 40 and n.10; Sept. Tr. (Grant) 211:5-15; PTX 986 at 25.)


If there were a "finding of fact" that an "average molecular weight" was the mode of the molecular weight distribution, it would be false and thus clearly erroneous.

Because the mode of a distribution is not necessarily unique, such a finding of fact would also be ambiguous.

As to the text



Second, although SEC can provide Mp, this disclosure is insufficient, Mylan argues, in view of the fact that it can also provide Mw and Mn. (Ryu Decl. ¶ 41.) This argument fails because, as explained above, Mp is the only type of AMW that can be provided directly by SEC without "further data manipulation and calculation." (Grant  [] Decl. ¶ 70 (Teva v. Mylan claim construction).) Since the patents-in-suit do not disclose any additional calculations, the Court credits Dr. Grant's representation that a person of ordinary skill in the art would understand that AMW means Mp. (See id.)


First, because Mp (which is the mode of the molecular weight distribution) is not an average, the finding that it is an AMW (average molecular weight) is wrong and clearly erroneous.  Second, true measures of average molecular weight, such as Mn or Mw, could be calculated from SEC curves by methods known to one of ordinary skill at the time of filing.  That the mode can be observed directly (although not necessarily unambiguously ) does not negate that true measures of average molecular weight could be calculated at the time of filing from the data in the specification.  A finding of fact that one of ordinary skill would understand that AMW means Mp is false and thus clearly erroneous.