Medichem, S.A. v. Rolabo, S.L.

United States Court of Appeals for the Federal Circuit

05-1179, -1248

MEDICHEM, S.A.,

Plaintiff-Appellee,

v.

ROLABO, S.L.,

Defendant-Appellant.

John G. Taylor, Frommer Lawrence & Haug LLP, of New York, New York,
argued for plaintiff-appellee. With him on the brief were Barry S. White and James K.
Stronski.

Thomas P. Heneghan, Michael Best & Friedrich LLP, of Madison, Wisconsin,
argued for defendant-appellant. With him on the brief were Jeffrey S. Ward and
Charlene L. Yager.

Appealed from: United States District Court for the Southern District of New York

Judge Jed S. Rakoff
 United States Court of Appeals for the Federal Circuit

05-1179, -1248

MEDICHEM, S.A.,

Plaintiff-Appellee,

v.

ROLABO, S.L.,

Defendant-Appellant.

________________________

DECIDED: February 3, 2006
________________________

Before SCHALL, GAJARSA, DYK, Circuit Judges.

GAJARSA, Circuit Judge.

This is the second round of a protracted litigation to establish priority of invention

between Stampa et al.’s U.S. Patent No. 6,084,100 (“the ‘100 patent”), assigned to

Medichem, S.A. (“Medichem”), and Jackson’s U.S. Patent No. 6,093,827 (“the ‘827

patent”), assigned to Rolabo, S.L. (“Rolabo”). In the first round appealed to this court,

we remanded to the district court, requiring it to establish an interference-in-fact under

35 U.S.C. § 291 before determining priority. Medichem, S.A. v. Rolabo, S.L., 353 F.3d

928 (Fed. Cir. 2003) (“Medichem II”). Rolabo now appeals from the judgment on

remand, in which the United States District Court for the Southern District of New York

found the existence of an interference-in-fact and awarded priority of invention to

Medichem. See Medichem, S.A. v. Rolabo, S.L., Memorandum Order, No. 01 Civ.
3087, 2004 WL 2674632 (S.D.N.Y Nov. 22, 2004) (“Medichem III”). For the reasons

discussed below, we affirm the judgment of the district court on the proper

establishment of the interfering subject matter and on the finding of the existence of an

interference-in-fact. We reverse, however, the district court’s award of priority to

Medichem, based on the insufficiency of the evidence that Medichem introduced at trial

to corroborate the testimony of its inventors regarding reduction to practice of the

invention.

BACKGROUND

A. The Patents

Medichem and Rolabo are both pharmaceutical manufacturers based in

Barcelona, Spain. Rolabo’s ‘827 patent and Medichem’s ‘100 patent both claim a

process for making loratadine from two precursor chemicals via a chemical reaction

known as the McMurry reaction. Loratadine is the active ingredient in the allergy

medication Claritin®. McMurry reactions involve the coupling of two starting materials in

the presence of low-valent titanium. In general, McMurry reactions can lead to two

types of products, diols and alkenes; loratadine, the desired end product of this reaction,

is an alkene. McMurry reactions can be optimized for alkene production by adjusting

various reaction parameters, such as the temperature and length of the reaction in this

case, and also by adding additional reactants. The only significant

05-1179, -1248 2
difference between the processes claimed by Medichem1 and Rolabo2 is that

Medichem’s process requires the reaction to be carried out in the presence of a type of

chemical known as a tertiary amine.3 In contrast, the Rolabo process permits by not

excluding, but does not require, the presence of a tertiary amine. Conceptually,

therefore, the Medichem invention, which requires a tertiary amine, is a species within

the genus of the Rolabo invention.

B. Proceedings to Date

Medichem brought an action under 35 U.S.C. § 291, alleging an interference-in-

fact between the ‘100 and ‘827 patents, claiming priority of invention, and seeking

invalidation of Rolabo’s patent under 35 U.S.C. § 102(g). Transcript of Verdict at 653-

67, Medichem, S.A. v. Rolabo, S.L., No. 01 Civ. 03087, 2002 U.S. Dist. LEXIS 27086

(S.D.N.Y May 8, 2002) (“Medichem I”). Because Rolabo was the party with the earlier

effective filing date, Medichem sought to establish priority by proving an actual reduction

1
Claims 1 and 2 of Medichem’s ‘100 patent read:
1. A process for the preparation of loratadine consisting of
reacting, in an organic solvent and in the presence of a tertiary
amine, 8-chloro-5,6- dihyrobenzo[5,6]cyclohepta[1,2-b]pyridin-11-
one, of formula VII with a low-valent titanium species. (emphasis
added).
2. The process of claim 1, wherein the low-valent titanium
species are generated by reduction of titanium tetrachloride with
zinc dust.
2
Claims 1 and 17 of Rolabo’s ‘827 patent read:
1. A process for preparing 5,6-dihydro-11H-
dibenzo[a,d]cyclohept-11-enes comprising reacting a
dibenzosuberone or an aza derivative thereof with an aliphatic
ketone in the presence of low valent titanium wherein said low
valent titanium is generated by zinc.
17. A process as claimed in claim 1 for preparing Loratadine.
3
A tertiary amine is a compound in which nitrogen is bonded three times to
carbon. A commonly used tertiary amine is pyridine.

05-1179, -1248 3
to practice that was even earlier.4 After a bench trial, the district court found that there

was no interference-in-fact between the claimed inventions, but it nonetheless awarded

priority to Medichem. Id.

On appeal, this court vacated the priority holding, opining that because the

existence of an interference-in-fact is a jurisdictional requirement under 35 U.S.C.

§ 291, it was therefore a precondition to the district court’s consideration of the priority

issue. Medichem II, 353 F.3d at 935-36. We explained that the first step in an

interference analysis is for the court to determine whether an interference exists under

35 U.S.C. § 291 by asking whether the “patents . . . have the same or substantially the

same subject matter in similar form as that required by the PTO pursuant to 35 U.S.C.

§ 135.” Id. at 934 (internal quotations omitted). In order to make this determination, we

use the “two-way” test which states that two patents interfere only if (1) invention A

either anticipates or renders obvious invention B, where Party A’s claimed invention is

presumed to be prior art vis-à-vis Party B and (2) vice versa. Id. (citing Eli Lilly & Co. v.

Bd. of Regents of the Univ. of Wash., 334 F.3d 1264, 1268 (Fed. Cir. 2003)).

In Medichem II, we held that Medichem’s claims to the “species” would clearly

anticipate Rolabo’s genus claim if the Medichem patent were assumed to be prior art.

Id. at 934-35. Thus, we held that the first prong of the two-way test was clearly

satisfied. Id. at 935. However, we remanded to the district court for a determination of

whether the second prong was also satisfied—namely, whether Rolabo’s genus claim, if

prior art, would either anticipate or render obvious Medichem’s species claim. Id. at

4
Rolabo’s effective filing date is February 26, 1997 and Medichem’s is May
30, 1997.

05-1179, -1248 4
935. We explained that “[a]s the ‘827 patent contains genus claims and the ‘100 patent

contains species claims, an arrangement that assumes that the ‘827 patent is prior art

does not necessarily anticipate or make obvious the narrower claims of the ‘100 patent.”

Id.

On remand, the district court held that “assuming arguendo [pursuant to the two-

way test] the priority of the ‘827 patent, claims 1 and 17 of the ‘827 patent clearly

anticipate and render obvious the adding of a tertiary amine, as in the ‘100 patent.”

Medichem III, 2004 WL 2674632 at *7. Although the court went on to explain its holding

on obviousness grounds, it was silent about the reasons underlying its apparent

determination that Rolabo’s genus claims would also anticipate Medichem’s species

claim. Instead, it improperly recharacterized our remand instructions as “reduc[ing] to

the question of whether it would be ‘obvious’ to add tertiary amine to a McMurry reaction

to make loratadine.”5 Id. (emphasis added).

The court then correctly stated that:

Determining obviousness requires consideration of two factors: 1) whether
the prior art would have suggested to one of ordinary skill in the art that he
should carry out the claimed process; and 2) whether the prior art would
have also revealed that in carrying out the process, one of ordinary skill
would have a reasonable expectation of success.

Id. The district court proceeded to articulate factual bases for its obviousness holding,

which included (1) an article that pointed to the use of amines to improve yields in

coupling reactions, (2) testimony by Rolabo’s expert about additional such prior art, and

5
In so doing, the court appears not to have separately considered the
question of whether the ‘827 patent, if taken as prior art, would anticipate the ‘100
patent.

05-1179, -1248 5
(3) evidence that such prior art had actually motivated Medichem’s inventor’s to try

adding tertiary amine to the reaction mixture. Medichem III, 2004 WL 2674632 at *7-8.

Having found the two-way test’s second prong to be satisfied on both anticipation

and obviousness grounds, the district court concluded that the Medichem and Rolabo

patents interfered, a finding that gave it jurisdiction over the priority dispute pursuant to

35 U.S.C. § 291. It awarded priority to Medichem, after finding that the invention

claimed in the ‘100 patent was reduced to practice prior to the constructive reduction to

practice date of Rolabo’s invention. See id. at *10-11 (referring to Medichem I and

stating that the court “reinstates and reaffirms its former priority ruling”).

In finding reduction to practice, the court neither explicitly discussed the legal

requirement that reduction to practice be corroborated by independent evidence, nor

made a factual finding of corroboration. However, it dismissed Rolabo’s argument that

Medichem’s inventors were not credible as a result of having fraudulently backdated

documents that it had offered to show reduction to practice in 1995. The court thus

affirmed its finding in Medichem I that Medichem had provided adequate proof of

reduction to practice in 1996. The court did so notwithstanding its previous observation

that “the willingness of Medichem to fraudulently backdate [evidence of reduction to

practice in 1995], coupled with Medichem’s less than punctilious recordkeeping

practices . . . does convince the Court that it cannot place the same reliance on

plaintiff’s testimony and documents as it might otherwise have.” Transcript of Verdict at

658, Medichem I. However, the court apparently adhered to its view that Medichem’s

fraudulent backdating was “chiefly a belated attempt to deal with their noncompliance

with [certain] regulatory requirements.” Id. The Medichem III court therefore reaffirmed

05-1179, -1248 6
its award of priority to Medichem, and Rolabo appealed on February 9, 2005. This court

has jurisdiction pursuant to 28 U.S.C. § 1295(a)(1).

As an aside, we wish to note that in parallel with the district court proceedings

under 35 U.S.C. § 291, the Board of Patent Appeals and Interferences (“Board”) has

been considering essentially the same interference and priority issues pursuant to

35 U.S.C. § 135. See Stampa v. Jackson, 65 U.S.P.Q.2d 1942 (B.P.A.I. 2002)

(involving an interference between Medichem’s then-pending reissue application and

both Rolabo’s patent and a pending continuation application thereof, giving rise to

Patent Interference Nos. 105,069 and 105,212). The Board held that the district court’s

holding in Medichem I did not bar the Board proceedings on grounds of issue

preclusion. See id. at 1945-47.

Shortly after the district court’s remand decision in Medichem III, the Board

resolved the interference in favor of Rolabo, reaching a conclusion opposite to that of

the district court. See Stampa v. Jackson, Inter. Nos. 105,069 & 105,212, 2005 WL

596770 (B.P.A.I. January 25, 2005). Central to its decision was Medichem’s failure to

corroborate its account of an alleged actual reduction to practice with evidence

independent of its inventors’ testimony. Id. at *19-20. The Board noted that “[a]ll of the

evidence regarding an experiment on May 7, 1996 which is said to have obtained

loratadine via a process of the count and conducted by [non-inventor] Lola Casas and

said to be recorded [in her notebook] is based on the testimony of [Medichem

inventors].” Id. at *15. Significantly, Medichem did not produce any testimony from

Casas, a failure that the Board perceived as sufficient to permit the inference that

Casas’ testimony would have been adverse to Medichem. Id. at *20. However, the

05-1179, -1248 7
Board declined to apply such an adverse inference on the grounds that “[Medichem’s]

case is so weak, we find it unnecessary to draw an inference one way or the other.”6 Id.

While appellant does not argue that the Board decision has a binding effect on this

court, Board decisions nevertheless represent the views of a panel of specialists in the

area of patent law. Medichem has appealed the Board’s decision to this court. See

Stampa v. Jackson, appeal docketed, Nos. 06-1004 & -1029 (Fed. Cir. Oct. 6, 2004 &

Oct. 24, 2004).

DISCUSSION

There are three issues in this case—namely, whether the district court (1) erred

in finding the existence of an interference-in-fact; (2) committed reversible error in failing

to formally define a count corresponding to the interfering subject matter; and (3) erred

in awarding priority of invention to Medichem based on the oral testimony of Medichem

co-inventors, testimony that Rolabo claims was not corroborated by independent

evidence, and thus should not have been credited in the final determination of whether

reduction to practice was established before the critical date.

A. Existence of an Interference-in-Fact

For the reasons explained below, we agree that under the second prong of the

two-way test for obviousness, Rolabo’s genus claim renders obvious the Medichem

species claim. We therefore affirm the lower court’s finding of an interference-in-fact

6
A final judgment on the merits was issued the same day. See Stampa v.
Jackson, Inter. Nos. 105,069 & 105,212, 2005 WL 596771 (B.P.A.I. January 25, 2005).
The Board later denied Medichem’s request for rehearing, stating inter alia that “[t]he
importance of Lola Casas’ testimony is manifest. She is the principal, if not the only,
corroborating witness on the issue of whether an actual reduction to practice took
place.” See Stampa v. Jackson, Inter. Nos. 105,069 & 105,212, 2005 WL 1541082
(B.P.A.I. June 27, 2005).

05-1179, -1248 8
without needing to review the district court’s unsupported factual finding that the second

prong of the two-way test was independently satisfied on anticipation grounds.

1. Standard of Review

In reviewing a district court’s finding of an interference-in-fact pursuant to the

two-way test, this court reviews, where necessary, both the subsidiary findings of

anticipation and/or obviousness as they relate to the application of the test. See

Medichem II, 353 F.3d at 932 (articulating the standard of review for findings of an

interference-in-fact under 35 U.S.C. § 291). Here, because we agree with the district

court’s subsidiary finding of obviousness, which is sufficient to support its finding of an

interference-in-fact, it is not necessary for us to review the court’s finding of anticipation.

Obviousness under 35 U.S.C. § 103 is a legal conclusion that is reviewed de

novo; however, it is based in turn on underlying factual determinations which are

reviewed for clear error. Id. Under the clear error standard, a reversal is permitted

“only when this court is left with a ‘definite and firm conviction’ that the district court was

in error.” Ruiz v. A.B. Chance Co., 357 F.3d 1270, 1275 (Fed. Cir. 2004) (quoting Amhil

Enters. Ltd. v. Wawa, Inc., 81 F.3d 1554, 1562 (Fed. Cir. 1996)).

2. Obviousness

The ultimate determination of whether an invention would have been obvious

under 35 U.S.C. § 103(a) is a legal conclusion based on the factual Graham findings,

e.g., “(1) the scope and content of the prior art; (2) the level of ordinary skill in the prior

art; and (3) the differences between the claimed invention and the prior art.” Velander

v. Garner, 348 F.3d 1359, 1363 (Fed. Cir. 2003) (citing Graham v. John Deere Co., 383

U.S. 1, 17 (1966)).

05-1179, -1248 9
 This court has held that if all the elements of an invention are found in a

combination of prior art references:

a proper analysis under § 103 requires, inter alia, consideration of two
factors: (1) whether the prior art would have suggested to those of
ordinary skill in the art that they should make the claimed composition or
device, or carry out the claimed process; and (2) whether the prior art
would also have revealed that in so making or carrying out, those of
ordinary skill would have a reasonable expectation of success.

Id.

The first requirement, the motivation to combine references, serves to prevent

hindsight bias. See McGinley v. Franklin Sports, Inc., 262 F.3d 1339, 1351 (Fed. Cir.

2001) (“To prevent hindsight invalidation of patent claims, the law requires some

‘teaching, suggestion or reason’ to combine cited references.”) (quoting Gambro Lundia

AB v. Baxter Healthcare Corp., 110 F.3d 1573 (Fed. Cir. 1997)). In making

obviousness determinations, the test is “whether the subject matter of the claimed

inventions would have been obvious to one skilled in the art at the time the inventions

were made, not what would be obvious to a judge after reading the patents in suit and

hearing the testimony.” Panduit Corp. v. Dennison Mfg. Co., 774 F.2d 1082, 1092 (Fed.

Cir. 1985). Whether such a motivation has been demonstrated is a question of fact.

See Winner Int'l Royalty Corp. v. Wang, 202 F.3d 1340, 1348 (Fed. Cir. 2000).

Evidence of a motivation to combine prior art references “may flow from the prior art

references themselves, the knowledge of one of ordinary skill in the art, or, in some

cases, from the nature of the problem to be solved.” Brown & Williamson Tobacco

Corp. v. Philip Morris Inc., 229 F.3d 1120, 1125 (Fed. Cir. 2000).

When a piece of prior art “suggests that the line of development flowing from the

reference’s disclosure is unlikely to be productive of the result sought by the applicant”

05-1179, -1248 10
the piece of prior art is said to “teach away” from the claimed invention. In re Gurley,

27 F.3d 551, 553 (Fed. Cir. 1994). As with other subsidiary obviousness inquiries,

“[w]hat a reference teaches and whether it teaches toward or away from the claimed

invention are questions of fact.” Winner, 202 F.3d at 1349 (internal quotations omitted).

However, obviousness must be determined in light of all the facts, and there is no rule

that a single reference that teaches away will mandate a finding of nonobviousness.

Likewise, a given course of action often has simultaneous advantages and

disadvantages, and this does not necessarily obviate motivation to combine. See id. at

1349 n.8 (“The fact that the motivating benefit comes at the expense of another benefit,

however, should not nullify its use as a basis to modify the disclosure of one reference

with the teachings of another. Instead, the benefits, both lost and gained, should be

weighed against one another.”). Where the prior art contains “apparently conflicting”

teachings (i.e., where some references teach the combination and others teach away

from it) each reference must be considered “for its power to suggest solutions to an

artisan of ordinary skill. . . . consider[ing] the degree to which one reference might

accurately discredit another.” In re Young, 927 F.2d 588, 591 (Fed. Cir. 1991).

As stated above, an obviousness determination requires not only the existence of

a motivation to combine elements from different prior art references, but also that a

skilled artisan would have perceived a reasonable expectation of success in making the

invention via that combination. While the definition of “reasonable expectation” is

somewhat vague, our case law makes clear that it does not require a certainty of

success. See In re O'Farrell, 853 F.2d 894, 903-04 (Fed. Cir. 1988) (“Obviousness

05-1179, -1248 11
does not require absolute predictability of success. . . . [A]ll that is required is a

reasonable expectation of success.”).

However, to have a reasonable expectation of success, one must be motivated

to do more than merely to “vary all parameters or try each of numerous possible choices

until one possibly arrived at a successful result, where the prior art gave either no

indication of which parameters were critical or no direction as to which of many possible

choices is likely to be successful.” Id. at 903. Similarly, prior art fails to provide the

requisite “reasonable expectation” of success where it teaches merely to pursue a

“general approach that seemed to be a promising field of experimentation, where the

prior art gave only general guidance as to the particular form of the claimed invention or

how to achieve it.” Id.

The district court’s finding of a reasonable expectation of success is a question of

fact, which we review for clear error. See Ruiz, 357 F.3d at 1275 (explaining that the

obviousness determination rests on “various factual findings that this court reviews for

clear error following a bench trial”); Brown & Williamson, 229 F.3d at 1129 (reviewing

the district court’s finding of reasonable expectation of success under the clear error

standard); see also Velander v. Garner, 348 F.3d 1359, 1376 (Fed. Cir. 2003)

(reviewing the Board of Patent Appeals and Interferences’ finding of a reasonable

expectation of success under a “substantial evidence” standard).

3. Analysis

Rolabo argues that the district court erred in finding that the Medichem invention

(which uses a tertiary amine) would have been obvious over the broader Rolabo

invention (which does not require it). Specifically, it appears to argue both that the prior

05-1179, -1248 12
art contained no motivation to combine references so as to have encouraged one

reasonably skilled in the art to have added a tertiary amine to a McMurry reaction and

that an artisan, even if motivated to add a tertiary amine to Rolabo’s process, would

have had no reasonable expectation of succeeding in making loratadine via a McMurry

reaction in the presence of a tertiary amine.

In support of its arguments, Rolabo cites the trial testimony of an expert witness

who explained that a seminal review article in the field showed that a tertiary amine

could have “a positive effect, a negative effect, and in some cases, both a positive and

negative effect” on the McMurry reaction. Rolabo goes on to cite prior art references

that disclose negative effects and essentially argues that the existence of prior art

references that teach away from the invention clearly negates the motivation to combine

and that the district court’s finding of motivation was clearly erroneous. We disagree.

Granted, it is clear that the prior art disclosed not only potential advantages of

using a tertiary amine in a McMurry reaction but also potential disadvantages. On the

one hand, some pieces of prior art taught that low concentrations of a tertiary amine

could sometimes be used to improve the yield of reactions or to avoid the formation of

undesirable rearranged products. On the other hand, other references reported that

tertiary amines could sometimes promote the formation of undesirable diol side-

products and that when they were used as the reaction solvent (i.e., when tertiary

amines are present at their highest possible concentrations), they could stop the

reaction completely.

We also note the ambivalence of Medichem co-inventor Dr. Onrubia toward the

introduction of a tertiary amine to the reaction mixture. On the one hand, she testified

05-1179, -1248 13
that she had added a tertiary amine “[b]ecause the literature said that it might be

possible to use tertiary amines in the reaction, that it wouldn’t interfere, that it wasn’t

incompatible, and it’s habitual in these circumstances to try various options until you get

the reaction to work.” On the other hand, when asked, “Is this purely hit or miss or is

there some logical cause . . . for believing that tertiary amine would add something?”

she responded: “Frankly, as an organic chemist I have no reason to say that there were

grounds for expecting anything from the addition of tertiary amine.”

As we have explained above, the fact that some teachings in the prior art conflict

with others does not render the findings of the district court clearly erroneous per se.

Rather, the prior art must be considered as a whole for what it teaches. We understand

the prior art, viewed as a whole, to teach that the addition of a tertiary amine sometimes

works to improve the yield of McMurry reactions, especially when a tertiary amine is

used in relatively low concentrations. In light of this, we cannot say that the district court

clearly erred in finding that the prior art would have provided the skilled artisan with a

motivation to combine references so as to use pyridine in the McMurry reaction. We

wish to emphasize that this is not a case where the prior art’s lack of definiteness or

certainty about the result of using a tertiary amine in a specific reaction system renders

the inventive subject matter “obvious to try” but not obvious. While we have made clear

that “‘obvious to try’ is not the standard under § 103[,] . . . the meaning of this maxim is

sometimes lost.” In re O‘Farrell, 853 F.2d 894, 903 (Fed. Cir. 1988). In O’Farrell, we

opined that:

[This] admonition . . . has been directed mainly at two kinds of error[,
namely where] . . . what would have been “obvious to try” would have
been . . . to vary all parameters or try each of numerous possible choices

05-1179, -1248 14
 . . . where the prior art gave . . . no direction as to which of many possible
choices is likely to be successful[ or] . . . to explore . . . a promising field of
experimentation, where the prior art gave only general guidance . . . .

Id. (citations omitted). In the instant case there are not numerous parameters to vary.

Rather, the principal parameter is the concentration of tertiary amine that should be

used, and the prior art teaches that if the tertiary amine were to have any positive effect

at all, it would be when it was present at low concentrations. Likewise, this is not a case

where the prior art gives merely general guidance. In contrast, the guidance is quite

clear—namely, that McMurry reactions of this kind can sometimes be optimized by

adding low levels of a tertiary amine.

For the aforementioned reasons, we find no clear error in the district court’s

determination that skilled artisans in possession of the Rolabo patent and the prior art

would have not only been motivated to add a tertiary amine but that they would have

possessed a reasonable expectation that they would succeed in optimizing the reaction.

Reviewing de novo the trial court’s application of these factual findings to reach the

legal conclusion of obviousness, we likewise find no error. Accordingly, we agree with

the district court’s determination that the addition of a tertiary amine to a McMurry

reaction would have been obvious in view of the Rolabo patent and the prior art.

Because this obviousness finding satisfies the second prong of the two-way test for an

interference-in-fact, we affirm the district court’s determination that an interference-in-

fact existed.

As a final matter, we note that we find no merit in Rolabo’s contention that we

should exclude from the subject matter of the interference that portion of its invention

that is directed to running reactions where titanium is present in specific concentration

05-1179, -1248 15
ranges (claims 10 and 11 of the ‘827 patent). Claim 10 requires a relative titanium

concentration of 1.5:1 to 4:1, and claim 11 requires a ratio of 2:1 to 3:1. The district

court relied on the testimony of Medichem’s expert witness, Dr. Finney, in holding that

all of the various claims of the ‘827 patent were “essentially identical to one another and

substantially the same as claim 2 of Medichem’s patent.” See Medichem III, 2004 WL

2674632 at *4. Rolabo argues that Finney’s expert testimony was “conclusory” and

therefore insufficient to establish an interference. However, it is clear from the record

that Finney’s testimony was far from conclusory. In fact, Finney provided a solid factual

basis for his opinion, stating that

“[c]laim 10 says that you should have between, a ratio of one and a half to
4 to 1 titanium to dibenzosuberone. Claim 11 states the range should be
2 to 1 to 3 to 1. These are both perfectly normal ranges. And in fact, the
patent examples in the ‘827 [Rolabo’s] patent specify I think about a 2.2 to
1 ratio. . . .”

Indeed, other evidence of record also supports the conclusion that these are normal

ranges. The Banerji reference discloses ratios of 2:1 and 1:1, Ishida discloses ratios of

1.5:1, 2.5:1 and 5:1, and Lenoir discloses a ratio of about 1:1.

In short, it is clear that Rolabo’s claims 10 and 11 are directed to titanium ratios

that are entirely within the range of the prior art, and this fact is dispositive. This court

has held that “[s]electing a narrow range from within a somewhat broader range

disclosed in a prior art reference is no less obvious than identifying a range that simply

overlaps a disclosed range.” In re Peterson, 315 F.3d 1325, 1330 (Fed. Cir. 2003).

Moreover, when “the claimed ranges are completely encompassed by the prior art, the

conclusion is even more compelling than in cases of mere overlap.” Id. We have

explained that the “normal desire of scientists or artisans to improve upon what is

05-1179, -1248 16
already generally known provides the motivation to determine where in a disclosed set

of percentage ranges is the optimum combination of percentages.” Id. Therefore,

because Rolabo’s claims 10 and 11 are directed to ratios that are entirely within the

prior art, the district court properly held those claims to be part of the interfering subject

matter pursuant to the two-way test.

B. Identification of Interfering Subject Matter

Having affirmed the district court’s determination that an interference-in-fact

exists, and that it properly includes those claims directed to specific titanium ratios, this

court turns to address Rolabo’s procedural argument that the district court erred when it

failed to comply with the Board’s practice of articulating a precise count of the

interference prior to making priority determinations.

This court has not yet addressed “whether district courts handling interfering

patent suits under § 291 must define this interfering subject matter in a way similar to a

count.” Slip Track Sys., Inc. v. Metal-Lite, Inc., 304 F.3d 1256, 1264 (Fed. Cir. 2002).

Nevertheless, we have made clear that at least “a single description of the interfering

subject matter is necessary for a determination of priority.” Id.

That said, Slip Track does not require a court to refer explicitly to the interfering

subject matter as a “count,” and we believe that in this case the district court was clear

about the identity of the interfering subject matter, stating in its opinion “all the various

claims of the ‘827 patent are essentially identical to one another and substantially the

same as claim 2 of Medichem’s patent.” Medichem III, 2004 WL 2674632 at *4.

Moreover, to the extent that the district court may not have been clear about whether

the tertiary amine limitation was part of the interfering subject matter, we can resolve

05-1179, -1248 17
this issue on appeal. See Slip Track, 304 F.3d at 1264-65 (holding that where “the

parties . . . dispute only whether one limitation is part of the interfering subject matter,

and determination of this issue is dependent upon issues of law alone, we will resolve

this issue on appeal.”) Accordingly, we hold that the interfering subject matter in this

case does not include the limitation of the tertiary amine, and corresponds to claim 17 of

Rolabo’s ‘827 patent. See id. 1265 (“Since the claims of the ‘760 patent do not include

a wallboard . . . the wallboard cannot be an element of the interfering subject matter in

this case, even though it is a limitation in the claims of the ‘203 patent.”).7

C. Priority of Invention

Finally, we review the district court’s award of priority of invention to Medichem.

Because the Medichem ‘100 patent issued from an application that had a later effective

filing date than did Rolabo’s ‘827 patent application, see supra note 4, Medichem bears

the burden of establishing priority by a preponderance of the evidence. See Eli Lilly &

Co. v. Aradigm Corp., 376 F.3d 1352, 1365 (Fed. Cir. 2004) (“[U]nder 35 U.S.C. § 291,

a party that does not have the earliest effective filing date needs only to demonstrate by

a preponderance of the evidence that it was the first to invent if the two patents or

applications at issue were co-pending before the PTO . . . .”). Medichem bears no

7
We note that in parallel interference proceedings, pursuant to 35 U.S.C.
§ 135, the Board reached a similar definition of the count. See Stampa v. Jackson,
65 U.S.P.Q.2d 1942, 1948 (B.P.A.I. 2002) (defining the count as Jackson’s (Rolabo’s)
claim 17).

05-1179, -1248 18
heightened burden, because neither patent enjoys a statutory presumption of validity.

See id. (“[T]he presumption of validity is nonexistent and the preponderance of the

evidence burden is appropriate even if both of the patents have issued by the time a

section 291 interference proceeding is initiated in a district court.”).

We have held that “priority of invention goes to the first party to reduce an

invention to practice unless the other party can show that it was the first to conceive of

the invention and that it exercised reasonable diligence in later reducing that invention

to practice.” Cooper v. Goldfarb, 154 F.3d 1321, 1327 (Fed. Cir. 1998). Here, because

neither party relied on a date of conception, priority is properly awarded to the party that

was the first to reduce its invention to practice, either actually or constructively. Rolabo

relies on its date of constructive reduction to practice, namely its February 26, 1997

effective filing date. Medichem, on the other hand, alleges that it achieved an actual

reduction to practice in the spring of 1996, a date which if proven would antecede

Rolabo’s filing date, and thereby entitle it to priority. See supra note 4 (effective filing

dates).

In order to establish an actual reduction to practice, Medichem must establish

three things: “(1) construct[ion of] an embodiment or perform[ance of] a process that

met all the limitations of the interference count; [ ] (2) . . . determin[ation] that the

invention would work for its intended purpose,” Cooper, 154 F.3d at 1327; and (3) the

existence of sufficient evidence to corroborate inventor testimony regarding these

events, see id. at 1330 (“In order to establish an actual reduction to practice, an

inventor’s testimony must be corroborated by independent evidence.”). The key issue

05-1179, -1248 19
on appeal is the last one, namely whether Medichem provided adequate corroboration

of the inventors’ testimony regarding the alleged actual reduction to practice.

For purposes of conceptual clarity, as well as clarity of language, it should be

noted that no similar condition of “corroboration” is imposed on an inventor’s notebook,

or indeed on any documentary or physical evidence, as a condition for its serving as

evidence of reduction to practice. See, e.g., Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572,

1577-78 (Fed. Cir. 1996) (explaining that “[t]his court does not require corroboration

where a party seeks to prove conception through the use of physical exhibits [because

t]he trier of fact can conclude for itself what documents show, aided by testimony as to

what the exhibit would mean to one skilled in the art”); Price v. Symsek, 988 F.2d 1187,

1195 (Fed. Cir. 1993) (“Only the inventor’s testimony requires corroboration before it

can be considered.”). Of course, the credibility (and therefore the corroborative value)

of an inventor’s notebook may vary. Nevertheless, a notebook, unlike the oral

testimony of an inventor, may be weighed, for whatever it is worth, in the final

determination of reduction to practice. However, in a case involving reduction to

practice, an unwitnessed notebook is insufficient on its own to support a claim of

reduction to practice. See Reese v. Hurst, 661 F.2d 1222, 1232 (CCPA 1981) (“The

inventors’ notebooks are accorded no more weight than the inventors’ testimony in this

instance, since they were not witnessed or signed and were unseen by any witness until

after this interference was declared.”); Hahn v. Wong, 892 F.2d 1028, 1033 (Fed. Cir.

1989) (stating that “affiants’ statements that by a certain date they had ‘read and

understood’ specified pages of Stephen Hahn’s laboratory notebooks did not

corroborate a reduction to practice . . . because they established only that those pages

05-1179, -1248 20
existed on a certain date . . . [and] did not independently corroborate the statements

made on those pages); Singh v. Brake, 222 F.3d 1362, 1370 (Fed. Cir. 2000) (stating

that Hahn v. Wong did not nullify the value of laboratory notebooks in corroborating

conception because “the standard of proof required to corroborate a reduction to

practice [is] more stringent . . . than that required to corroborate a conception.”).8 Once

properly admitted into evidence, documentary and physical evidence is assigned

probative value and collectively weighed to determine whether reduction to practice has

been achieved. This is what is meant by the maxim that documentary and physical

evidence do not require “corroboration.”

1. Corroboration

Credibility concerns undergird the corroboration requirement, the purpose of

which is to prevent fraud. See Chen v. Bouchard, 347 F.3d 1299, 1309 (Fed. Cir.

2003) (“[T]he purpose of corroboration . . . is to prevent fraud, by providing independent

confirmation of the inventor’s testimony.”) (internal quotations omitted). As such, the

corroboration requirement provides an additional safeguard against courts being

deceived by inventors who may be tempted to mischaracterize the events of the past

through their testimony. See Mahurkar, 79 F.3d at 1577 (“While perhaps prophylactic in

application given the unique abilities of trial court judges and juries to assess credibility,

the rule provides a bright line for both district courts and the PTO to follow in addressing

the difficult issues related to invention dates.”).

8
Cf. Stern v. Trs. of Columbia Univ., No. 05-1291, slip op. at 5 (Fed. Cir. Jan. 17,
2006) (“[R]egardless of the contents of the notebooks, unwitnessed laboratory
notebooks on their own are insufficient to support his claim [of conception, and
therefore] of co-inventorship.”).

05-1179, -1248 21
 Sufficiency of corroboration is determined by using a “rule of reason” analysis,

under which all pertinent evidence is examined when determining the credibility of an

inventor’s testimony. See Price v. Symsek, 988 F.2d 1187, 1195 (Fed. Cir. 1993) (“‘A

rule of reason’ analysis is applied to determine whether the inventor’s prior conception

testimony has been corroborated.”); Berges v. Gottstein, 618 F.2d 771, 776 (CCPA

1980) (“In the final analysis, each corroboration case must be decided on its own facts

with a view to deciding whether the evidence as a whole is persuasive.”).

The requirement of independent knowledge remains key to the corroboration

inquiry. See Reese v. Hurst, 661 F.2d 1222, 1225 (CCPA 1981) (“[A]doption of the ‘rule

of reason’ has not altered the requirement that evidence of corroboration must not

depend solely on the inventor himself.”). “Independent corroboration may consist of

testimony of a witness, other than the inventor, to the actual reduction to practice or it

may consist of evidence of surrounding facts and circumstances independent of

information received from the inventor.” Id. One consequence of the independence

requirement is that testimony of one co-inventor cannot be used to help corroborate the

testimony of another. See, e.g., Lacks Indus. v. McKechnie Vehicle Components USA,

Inc., 322 F.3d 1335, 1350 (Fed. Cir. 2003) (opining that the Special Master rightly

refused to accept cross-corroboration of oral testimony as being adequate).

Despite the importance of the independence requirement, however, “[t]he law

does not impose an impossible standard of ‘independence’ on corroborative evidence

by requiring that every point of a reduction to practice be corroborated by evidence

having a source totally independent of the inventor. . . .” Cooper v. Goldfarb, 154 F.3d

at 1330 (internal quotations omitted). Similarly, “it is not necessary to produce an actual

05-1179, -1248 22
over-the-shoulder observer. Rather, sufficient circumstantial evidence of an

independent nature can satisfy the corroboration requirement.” Id.

When an inventor claims a process for making a chemical compound rather than

the compound itself, it is the successful reduction to practice of the process that must be

corroborated, and not merely the successful production of the compound per se. Thus,

spectral evidence that might be sufficient per se to corroborate a claim directed to the

product will generally not be sufficient to corroborate a claim directed to the process, in

the absence of some evidence to corroborate that the product was produced via that

process.

2. Standard of Review

Whether or not corroboration exists is a question of fact, the district court’s

determination of which we review for clear error. This is true because “issues of

conception and reduction to practice are questions of law predicated on subsidiary

factual findings,” Eaton v. Evans, 204 F.3d 1094, 1097 (Fed. Cir. 2000), and

corroboration is properly viewed as a subsidiary factual finding. See Singh v. Brake,

222 F.3d at 1368 (implying that corroboration is a question of fact by holding that

“substantial evidence supports the Board’s finding that this [notebook] entry alone was

insufficient to corroborate Singh’s testimony . . . .”)(emphasis added).

Before reviewing the determination of the court below, we note that it is true that

corroboration is fundamentally about “credibility,” see supra Discussion, Part C.1, and

that in reviewing factual findings under the clear error standard, this court “gives great

deference to the district court’s decisions regarding credibility of witnesses.” See

Ecolochem, Inc. v. S. Cal. Edison Co., 227 F.3d 1361, 1378-79 (Fed. Cir. 2000)

05-1179, -1248 23
(internal quotations omitted). Indeed, such deference is appropriately accorded to

assessments of witness credibility because “only the trial judge can be aware of the

variations in demeanor and tone of voice that bear so heavily on the listener’s

understanding of and belief in what is said.” Anderson v. Bessemer City, 470 U.S. 564,

575 (1985).

Nonetheless, such deference is often of little consequence in a corroboration

inquiry because the raison d’etre of the corroboration requirement is our refusal to base

priority determinations on a court’s uncorroborated assessments of a testifying

inventor’s credibility. Even the most credible inventor testimony is a fortiori required to

be corroborated by independent evidence, which may consist of documentary evidence

as well as the testimony of non-inventors. To the extent that a district court’s finding of

corroboration rests on its assessment of the credibility of non-inventor testimony, we

apply the deferential standard of review stated in Ecolochem. To the extent that it rests,

as it does here, on the district court’s assessment of documentary, as opposed to

testimonial evidence, we still apply clear error review; however, clear error is less

difficult to establish.

3. Analysis

The parties in this case dispute whether or not there was adequate corroboration

of the inventors’ testimony that Medichem had actually reduced to practice the process

of the claimed invention before Rolabo’s effective filing date. Medichem put forward two

05-1179, -1248 24
principal types of corroborating evidence: documentary evidence generated by

inventors and that generated by non-inventors.9

In the first category, it produced a documented request for the analysis of a

sample, purported to have been produced via the claimed synthetic route, which was

sent by one co-inventor to another. Also in this category were the NMR spectral data

obtained by the co-inventor pursuant to that request. These spectra were consistent

with loratadine, and the accuracy of that chemical identification is not being challenged.

Finally, this category includes the original laboratory notebook of co-inventor

Dr. Rodriguez. In the second category, documentary evidence by non-inventors, there

is the original laboratory notebook of former Medichem employee, and non-inventor,

Lola Casas.

This court now turns to consider the corroborative value of the three principal

pieces of potentially corroborative evidence: the NMR spectra, the notebooks of

Medichem’s inventors, and the notebook of non-inventor Casas. We note at the outset

that the problem with the dated NMR data is that at most they corroborate that the

inventors were in possession of the chemical loratadine as of that date; they do not, in

themselves, adequately corroborate the claimed process, as they do not establish

whether the sample that was analyzed was actually produced by that process. If this

case dealt with a claim to a composition of matter, rather than to a process, the NMR

evidence might very well take on a different relevance in this regard. As far as the

corroborative value of the inventors’ notebooks is concerned, they were not witnessed,

9
This patent bore a number of co-inventors, many of whom testified at trial.
As we have noted above, the testimony of one inventor cannot be corroborated by the
testimony of co-inventors.

05-1179, -1248 25
and they do not provide an “independent” source of authority on the issue of reduction

to practice. Hence, they have minimum corroborative value.

It is clear to this court, therefore, that Medichem’s claim of corroboration stands

or falls with the modicum of additional corroborative value that can properly be assigned

to non-inventor Casas’ notebook.10 However, Casas did not testify regarding the

notebook or the genuineness of its contents. In addition, although Casas’ notebook was

dated, it was neither signed nor witnessed, and inventor Rodriguez testified that she

and Casas had made entries in each others’ notebooks. Rodriguez characterized these

occasions as not out of the ordinary. As a result, the district court was clearly reliant on

the inventor to help to identify the author of specific entries made in Casas’ notebook,

because in a reduction to practice inquiry, only those passages of the unsigned,

unwitnessed notebooks authored by non-inventor Casas could possess significant

corroborative value. In addition, without testimony from Casas, the court lacked any

non-inventor testimony regarding the genuineness of the notebook’s contents.

We also note that Medichem admitted fraudulently backdating certain documents

relating to a purported 1995 reduction to practice. Even though the backdating of the

1995 documents was unrelated to the critical pages in Casas’ notebook, which purport

10
When an inventor attempts to offer into evidence the notebook of a non-
inventor as evidence of corroboration, evidentiary issues might be implicated. For
example, the notebook is likely to be hearsay, and if so, there may be an issue as to
whether or not it falls within an exception to the hearsay rule, such as the business
record exception. Indeed, in Chen v. Bouchard, this court affirmed the decision of the
Board of Patent Appeals and Interferences to exclude as inadmissible hearsay a non-
inventor’s notebooks, which had been offered to corroborate reduction to practice
where, as in the instant case, the non-inventor did not testify. 347 F.3d 1299, 1308
(Fed. Cir. 2003).

05-1179, -1248 26
to establish a reduction to practice in 1996, the district court found that the credibility of

the Medichem inventors was accordingly diminished.

Where a laboratory notebook authored by a non-inventor is offered into evidence

pursuant to authentication by an inventor, where the author of the notebook has not

testified at trial or otherwise attested to its authenticity, and where the notebook has not

been signed or witnessed and has not been maintained in reasonable accordance with

good laboratory practices sufficient to reasonably ensure its genuineness under the

circumstances, then the corroborative value of the notebook is minimal. Given the facts

of this case, Casas’ notebook should therefore not be accorded much corroborative

value. In view of the minimal corroborative value of the inventors’ notebooks and the

limited value of the NMR spectrum, we conclude that the evidence, evaluated as a

whole under the rule of reason, is insufficient as a matter of law to corroborate

Medichem’s reduction to practice.

The district court did not specifically address corroboration in its obviousness

inquiry, a fact that might, in some circumstances, hamper our ability to conduct clear

error review. Here, however, the facts of the case admit of only one conclusion as a

matter of law, and we therefore decide the case without remanding to the district court

for an explanation of why it implicitly found corroboration to be present. We hold that

corroboration is absent and that the district court therefore erred in reaching its legal

conclusion that Medichem had reduced its invention to practice in the spring of 1996.

Accordingly, we reverse the district court’s award of priority to Medichem.

AFFIRMED-IN-PART, REVERSED-IN-PART

No costs.

05-1179, -1248 27