The Director of the United States Patent and Trademark Office (USPTO) has requested comments on the standard for determining whether an interference-in-fact exists. See Notice of December 20, 2000, in the Federal Register, Vol. 65, No. 245, 79809-10, entitled "Standards for Declaring a Patent Interference" ("Notice"). The current standard employed by the Board of Patent Appeals and Interferences ("Board") requires a two-way patentability analysis for the Director to be of the opinion that an interference-in-fact exists. Winter v. Fujita, 53 USPQ2d 1234 (BPAI 1999), reh'g denied, 53 USPQ2d 1478 (BPAI 2000). Recently, some commentators have suggested that the standard, at least in some cases, should be a one-way patentability analysis. After due consideration by the Executive Committee of the IPO Board of Directors (the “Executive Committee”), and the Chairs and Vice-Chair of the Patent Interferences Committee (the “Interference Committee”), the IPO officially recommends that the Director use a one-way patentability analysis to determine whether an interference-in-fact exists. The remainder of this paper is the Interference Committee's (1) response to the request for comments, (2) answers to the questions posed in the Notice, and (3) answers to the questions posed by the Board subsequent to the Notice. This paper may not reflect the opinions of all members of the IPO or the Interference Committee.

The one-way analysis will assure that the PTO will not issue two or more patents to different parties for the same patentable invention. A one-way standard is sure to create a conflict in the PTO between or among the conflicting parties, requiring the PTO to resolve priority issues. In contrast, a two-way test may result in no declaration of interference, with the consequence that the PTO will issue two or more patents, one or more of which may be invalid under 35 U.S.C. §102(g). A case or controversy may never arise between the patentees that could lead to a resolution of the issue, leaving third parties with no choice other than to accept multiple licenses or to bring separate lawsuits against each patentee, possibly in separate judicial districts. This is a poor option in view of current jury trends of awarding large damage awards, and in view of the fact that many persons and businesses have insufficient capital to fund such litigation.

The Director shall be responsible for providing policy and management supervision for the Office and for the issuance of patents. The Director shall perform these duties in a fair, impartial and equitable manner. 35 U.S.C. §3. Interferences are declared under 35 U.S.C. §135(a). Section 135(a) provides:

(a) Whenever an application is made for a patent which, in the opinion of the Director, would interfere with any pending application, or with any unexpired patent, an interference may be declared and the Director shall give notice of such declaration to the applicants, or applicant and patentee, as the case may be. ...

Thus, the Director may declare an interference when, in his opinion, an application is made for a patent that would interfere with either a pending application or with an unexpired patent. 35 U.S.C. §135(a). Once declared, the Board shall determine questions of priority and may determine questions of patentability. Id. In all cases, however, these duties are to be performed in a fair, impartial and equitable manner. 35 U.S.C. §3.

In 1984, the Commissioner [now Director] adopted a new set of rules, i.e., 37 C.F.R. §§1.601-1.688 (replacing old rules 37 C.F.R. §§1.201-1.288), to improve interference procedure so that the rights of parties in interferences are determined at an early date and the overall process of examining patent applications that become involved in interferences is simplified. 49 FR 48416 (December 12, 1984). As part of the rules adopted in 1984, the Commissioner defined when an interference-in-fact exists in 37 C.F.R. §1.601(j). This rule provides, as it did in 1984, that:

An 'interference-in-fact' exists when at least one claim of a party which corresponds to a count and at least one claim of an opponent which corresponds to the count define the same patentable invention.

The phrase "same patentable invention" is defined in 37 C.F.R. §1.601(n) as follows:

(n) Invention 'A' is the 'same patentable invention' as a invention 'B' when invention 'A' is the same as (35 U.S.C. 102) or is obvious (35 U.S.C. 103) in view of invention 'B' assuming invention 'B' is prior art with respect to invention 'A'. Invention 'A' is a 'separate patentable invention' with respect to invention 'B' when invention 'A' is new (35 U.S.C. 102) and non-obvious (35 U.S.C. 103) in view of invention 'B' assuming invention 'B' is prior art with respect to invention 'A'. (emphasis in original)

Accordingly, there must be at least one claim of an application and at least one claim of another application or an unexpired patent that define the same patentable invention for an interference to proceed. The issue addressed herein is how to decide if two (or more) claims are drawn to the same patentable invention. Thus, we face the fundamental question of whether to use a one-way or a two-way analysis to determine if a claim of a party and a claim of an opponent "define the same patentable invention" for the purposes of determining whether an interference-in-fact exists.

If the Office were to use a one-way analysis, an interference-in-fact would exist if at least one claim of either party was directed to the same patentable invention as at least one claim of an opposing party. As long as one party's claim is determined to be directed to subject matter that is the same as or is obvious in view of the subject matter of the opposing party’s claim, assuming the subject matter of the opposing party’s claim to be prior art, the Office should declare an interference-in-fact. Under a one-way analysis, an interference-in-fact would always exist between one party's claim to a genus and an opposing party’s claim to a species within the genus. The principle is well-established that the disclosure of a species in a cited reference is sufficient to prevent a later applicant from obtaining a "generic claim." In re Gosteli, 872 F.2d 1008, 10 USPQ2d 1614 (Fed. Cir. 1989); In re Slayter, 276 F.2d 408, 125 USPQ 345 (CCPA 1960).

If the Office were to use a two-way analysis, an interference-in-fact would exist only if at least one claim of either party was directed to the same patentable invention as at least one claim of an opposing party and vice versa. First, the subject matter of one party's claim would be assumed to be prior art and a Board member would decide whether the subject matter of the opponent's claim is the same as or is obvious in view of the subject matter of the party’s claim. Likewise, the subject matter of the opponent's claim would be assumed to be prior art and a Board member would decide whether the subject matter of the party's claim is the same as or is obvious in view of the subject matter of the opponent’s claim. An interference-in-fact would exist only if the Board member finds anticipation or obviousness in both determinations. Further, under a two-way analysis, a party’s claim to a genus would not interfere with an opponent’s claim to an unobvious species within the genus (i.e., there would be no interference-in-fact).

Until recently, no precedential decision of the courts or the Board has explicitly stated whether a one-way or a two-way test is appropriate for determining whether an interference-in-fact exists. However, in Winter v. Fujita, 53 USPQ2d 1234, 1243 (BPAI 1999), reh'g denied, 53 USPQ2d 1478 (BPAI 2000), the Board explicitly stated:

The claimed invention of Party A is presumed to be prior art vis-a-vis Party B and vice versa. The claimed invention of Party A must anticipate or render obvious the claimed invention of Party B and the claimed invention of Party B must anticipate or render obvious the claimed invention of Party A. When the two-way analysis is applied, then regardless of who ultimately prevails on the issue of priority, * * * [USPTO] assures itself that it will not issue two patents to the same patentable invention.

This statement spurred much public debate.

The contention between a one-way analysis and a two-way analysis is not always an issue of importance to parties. The following examples, taken from the Notice, illustrate situations in which either test results in an interference-in-fact.

Example 3: Application E contains patentable claims 1 (engine), 2 (6-cylinder engine), and 3 (engine with a platinum piston). Application F contains patentable claims 11 (engine) and 12 (8-cylinder engine). Claims 1 and 2 of application E and claims 11 and 12 of application F define the same patentable invention. Claim 3 of application E defines a separate patentable invention from claims 1 and 2 of application E and claims 11 and 12 of application F. If an interference is declared, there would be one count (engine). Claims 1 and 2 of application E and claims 11 and 12 of application F would be designated to correspond to the count. Claim 3 of application E would not be designated to correspond to the count.

Example 4: Application G contains patentable claims 1 (engine), 2 (6-cylinder engine), and 3 (engine with a platinum piston). Application H contains patentable claims 11 (engine) and 15 (engine with a platinum piston). Claims 1 and 2 of application G and claim 11 of application H define the same patentable invention. Claim 3 of application G and claim 15 of application H define a separate patentable invention from claims 1 and 2 of application G and claim 11 of application H. If an interference is declared, there would be two counts: Count 1 (engine) and Count 2 (engine with a platinum piston). Claims 1 and 2 of application G and claim 11 of application H would be designated to correspond to Count 1. Claim 3 of application G and claim 15 of application H would be designated to correspond to Count 2.

Example 5: Application J contains patentable claims 1 (engine), 2 (combination of an engine and a carburetor) and 3 (combination of an engine, a carburetor, and a catalytic converter). Application K contains patentable claims 31 (engine), 32 (combination of an engine and a carburetor), and 33 (combination of an engine, a carburetor, and an air filter). The engine, combination of an engine and carburetor, and combination of an engine, carburetor, and air filter define the same patentable invention. The combination of an engine, carburetor, and catalytic converter define a separate patentable invention from engine. If an interference is declared, there would be one count (engine). Claims 1 and 2 of application J and claims 31, 32, and 33 of application K would be designated to correspond to the count. Claim 3 of application J would not be designated as corresponding to the count.

In each of the above examples, the parties are in an interference, can contest whether their opponent is entitled to priority, can contest whether their opponent's claims are patentable, and can rely on all of their proofs to establish priority. If, however, the above examples are revised to limit one applicant to a separately patentable species not disclosed by its opponent, the dichotomy between the one-way analysis and the two-way analysis is apparent. Take the following example:

Example 6: Application E contains patentable claim 1 (engine). Application F contains patentable claim 11 (8-cylinder engine). Claim 11 of Application F defines a separate patentable invention from claim 1 of Application E. Application E does not disclose an 8-cylinder engine. Application F does not disclose engines generically.

In Example 6, above, the species claim anticipates the genus claim, assuming the species is prior art. However, under the two-way analysis, there would be no interference-in-fact between a party's claim to the genus and an opponent's claim to an unobvious species within the genus.

Interferences are declared to determine which of at least two conflicting applicants (or applicant and patentee) made the common invention first without abandoning, suppressing or concealing the invention. 35 U.S.C. § 102(g). In Example 6, above, the facts presented (wherein the species “8-cylinder engine” is assumed patentable over the genus “engine”) raise two troublesome scenarios if a two-way patentability analysis is applied.

A. Species Applicant Files First

1. Examiner Refuses to Consider Rule 1.131 Affidavit

In one scenario, Applicant F files first, disclosing and claiming the species of an 8-cylinder engine, but not the genus of an engine; and Applicant E files second, disclosing and claiming an engine, but not disclosing or claiming an 8-cylinder engine. If application F issues as a patent, then application E should be rejected in view of the patent under 35 U.S.C. §102(e).

In this scenario, the generic applicant might attempt to overcome the rejection by filing a declaration under 37 C.F.R. §1.131. However, a declaration under 37 C.F.R. §1.131 is not effective to swear behind the filing date of a reference that claims "the same patentable invention as defined in §1.601(n)." Further, when 37 C.F.R. §1.131 was revised in 1988 to include the above cited language, the USPTO made it clear that:

The purpose of the amendment to §1.131(a) is to ensure that an applicant who is claiming an invention which is identical to, or obvious in view of, i.e., the same patentable invention as claimed in a domestic patent, cannot employ an affidavit under §1.131 as a means for avoiding an interference with the patent. To allow an applicant to do so would result in the issuance of two patents to the same invention.

See 53 FR 23728 (June 23, 1988). Thus, if Examiners follow the commentary of the Commissioner in (1) applying the patent as the prior art (a one-way test) and (2) preventing two patents to the same invention from issuing (using the one-way test), then Applicant E will not be permitted to overcome the rejection by showing an earlier date of invention with a declaration under 37 C.F.R. §1.131 (Applicant F is claiming the same patentable invention as Applicant E because an 8-cylinder engine anticipates an engine). Moreover, if the Board uses a two-way patentability analysis to determine whether an interference-in-fact exists, Applicant E cannot provoke an interference because the species is patentable over the genus.

The USPTO, however, cannot prevent an applicant from overcoming a reference with a declaration under 37 C.F.R. §1.131 on the grounds that the reference claims applicant's invention and, at the same time, deny applicant an interference on the grounds that the claims of the application and those of the reference are not for substantially the same invention. See In re Eickmeyer, 602 F.2d 974, 202 USPQ 655 (CCPA 1979).

2. Examiner Accepts Rule 1.131 Affidavit

Assume the same facts as above. If the Examiner uses a two-way patentability analysis to determine whether Applicant E can file a declaration under 37 C.F.R. §1.131, the Examiner must conclude that the 8-cylinder engine is not the same patentable invention as the genus claim to the engine, and that an interference-in-fact does not exist. Again, for purposes of this example, it is assumed that the species is patentable over the genus. Thus, Applicant F and the public may be harmed since two patents will issue even though Applicant F may have invented the species before Applicant E invented the genus. If Applicants E and F are content not to challenge the validity of the other’s patent under 35 U.S.C. §102(g), a member of the public wishing to make an 8-cylinder engine will be put in the position of having to seek out two licenses. In the area of double patenting, such a result is avoided by conditioning enforceability on common ownership. Why a less protective outcome in the interference context?

B. Genus Applicant Files First

In another scenario, Applicant E files first, disclosing and claiming “engine,” but not disclosing or claiming an 8-cylinder engine, and Applicant F files second, disclosing and claiming the species of an 8-cylinder engine, but not the genus of “engine.” If application E issues as a patent, then application F should not be rejected in view of the patent under 35 U.S.C. §102(e) or 35 U.S.C. § §102(e)/103. Again, for purposes of this example, it is assumed that the species is patentable over the genus. Thus, once again, two patents will issue with no USPTO determination of priority.

A two-way patentability analysis simply ignores the expectations of applicants and the public that the USPTO will issue valid patents. The United States has not adopted a policy of granting patents based solely on filing dates; rather, the issue is who was the first to invent the subject matter in question. United States law resolves the issue by considering the acts of conception and reduction-to-practice. A two-way test may never allow these facts to be examined.

Some Board members have argued that a party is not harmed when the Commissioner declines to declare an interference. The argument stems from a comment from the Supreme Court in Ewing v. United States ex reel. Fowler Car Co., 244 U.S. 1, 11 (1917), wherein the Court mentioned that two conflicting patentees have the option of litigation to resolve priority disputes. However, reliance on a decision from 1917 in today's fast-paced, technology-driven world is not fair or equitable. The cost of District Court litigation is often ten-fold more than the cost of an interference. Instead of a cost of hundreds of thousands of dollars, a cost of millions of dollars is typical in District Court litigation.

Moreover, this argument does not take into account the uninformed third party who is faced with jury awards exceeding tens if not hundreds of millions of dollars. Third parties have no clue from whom to request a license or whether to request a license at all. Should a third party proceed with no license and face lawsuits and possible damage awards from two or more patent owners including potential findings of willfulness? Should a third party request a license from one patent owner and face a lawsuit from one or more other patent owners? Should a third party take a license and pay royalties to all patent owners when it is possible that there is only one valid patent a patent that may not even cover the third party's activities?

Clearly, if the USPTO has the ability to issue only valid patents, it should do so. The resource consequence to the USPTO is minor compared to the burden on patentees and third parties.

If a one-way patentability analysis is sufficient to establish an interference-in-fact, would it be possible to have an interference with two counts as set out in Example 4, reproduced above?

An interference with two counts would be possible. Comments were received regarding 37 C.F.R. §1.601(n) which defines "same patentable invention" and "separate patentable invention" when the rule was proposed. 49 FR 48416 (Dec. 12, 1984). One written comment suggested that §1.601(n) be deleted. In view of the nature of the comment and the importance of the definitions in §1.601(n), the written comment and response are reproduced, in part, below:

Comment: The paragraph to be deleted contains an explicit definition of both 'same patentable invention' and 'separate patentable invention'. It requires application of a two-prong test for determining 'sameness' or 'separateness'. Two inventions are considered to be the 'same patentable invention' when they are the same under 35 U.S.C. 102 or obvious variants under 35 U.S.C. 103 when one is considered as 'prior art' with respect to the other. Application of this standard may lead to results which are anomalous or in conflict with statutory requirements if, as the proposed rules apparently assume, the award of priority as to counts embodying 'separate patentable inventions' will inherently and finally resolve the question of entitlement to a patent as between the parties on involved claims corresponding to those counts. Three specific problems with the definition of Rule 601(n) have been identified as follows:

(1) If the standard of separate patentability as between two counts is applied on a mutual basis (i.e., Invention A must be separately patentable from Invention B considered as prior art and Invention B must be separately patentable from Invention A considered as prior art), then the conduct of interferences with separate counts directed to both a generic invention and an included specific embodiment thereof (i.e., a species) are frustrated. If, however, such mutuality is not applied (i.e., if Invention A is the existing count, then a count directed to Invention B is proper so long as Invention B would be considered patentable over Invention A as prior art), then a number of irreconcilable anomalies in count modification through preliminary motions practice will be manifest.

* * *

The following examples are provided as indicative of perceived difficulties in the application of the definition embodied in Rule 601(n):

Example A: Rule 601(n) is applied in a mutuality sense to an interference where the existing count (Invention A) is a broad genus over organic compounds. Both parties disclose a species (Invention B) which one of the parties alleges represents a "selection invention', i.e., is neither anticipated by the genus nor obvious therefrom. The party proposing to add a species count to Invention B must demonstrate that the genus was neither the same as nor obvious in view of the species, an impossibility since disclosure of the species anticipates the genus under 35 U.S.C. 102. In effect, no party could ever propose sub-generic or species counts included within broader generic counts. Since all potential patentability and priority questions with respect to possible sub-generic and species claims are intended to be resolved in the unified patentability/priority proceedings, loss of such a broad generic count might be construed as estopping the losing party from arguing entitlement to a patent for species and sub-genera within the lost count. Since proof of a single species is dispositive of the question of priority as to a genus, no matter how large, this is a more severe result than would appear to be proper or intended.

In response to the comments, the USPTO stated the following:

With respect to paragraph (1) of the comment, the standard of patentability will not be applied 'on a mutual basis.' Thus, if a species is patentable over a genus, the species is a 'separate patentable invention' from the genus. Compare In re Taub, 348 F.2d 556, 146 USPQ 384 (CCPA 1965) (fluorine species might be patentable over genus of Markush group of hydrogen and halogen). A first count to a genus and a second count to a species which is patentable over the genus may properly appear in an interference. See e.g., Example 4. The comment suggests that if 'such mutuality is not applied * * * then a number of irreconcilable anomalies * * * will be manifest.' The urged 'irreconcilable anomalies' are not readily apparent to the PTO.

* * *

Analysis of Commentator's Example A. Example A does not describe any practice under these rules, because 'same patentable invention' and 'separate patentable invention' under § 1.601(n) are not intended to be 'applied in a mutuality sense.' Where a first count is to a genus and a second count is to a species within the scope of the genus, there may be two counts if the species is separately patentable from the genus. The species is 'invention A' referred to in § 1.601(n); the genus is 'invention B' referred to in § 1.601(n). (Emphasis added).

From the above, it is clear that one does not apply a two-way test to determine whether a second count is appropriate.

How would having an interference between claim 1 of Application G and claim 15 of Application H of Example 4 square with the holding of Nitz v. Ehrenreich, 537 F.2d 539,543, 190 USPQ 413, 416-17 (CCPA 1976)?

There is no conflict between having an interference between claim 1 of Application G and claim 15 of Application H of Example 4, and the holding of Nitz v. Ehrenreich, id. In Nitz, the Court held that no interference-in-fact existed with respect to Count 1. Count 1 required, as did Nitz's claims, a "substantially all carbon friction article." According to the Court, "substantially all carbon" meant that at least eighty percent by weight carbon was present. Thus, the Count was internally inconsistent since on the one hand it required at least eighty percent by weight carbon by recitation of the phrase "substantially all carbon" and, on the other hand, it permitted only fifty-two percent by weight carbon by recitation of the phrase "up to about 48% by weight" of modifier. The Court could not proceed with such a count and ruled that no interference-in-fact existed because allowing the recitation of 48% by weight of the modifier was an improper modification of the Count which already required the friction article to be "substantially all carbon."

The Court in Nitz did not apply a one-way or two-way patentability test to determine whether there was interfering subject matter between the claims of Nitz and the claims of Ehrenreich. Instead, the Court merely considered the appropriateness of the language of the count and found it to be inappropriate.

Moreover, it was not until 1977 (the year following the Nitz decision), that the CCPA announced that the test for interference-in-fact involved whether two sets of claims are patentably distinct from each other. In Aelony v. Arni, 192 USPQ 486, 490 (CCPA 1977), the Court stated:

Sections 102, 103 and 135 of 35 U.S.C. clearly contemplate - where different inventive entities are concerned - that only one patent should issue for inventions which are either identical to or not patentably distinct from each other. Hence, we view the phantom count practice as a proper procedural tool established by the PTO in order to ensure, where necessary, what the statute intends, viz., that only one patent should issue for one inventive concept. Moreover, we believe that there is ample precedent from this Court for framing the test of interference in fact in terms of whether two sets of claims are patentably distinct from each other.

In Nitz, the Court did not make a determination that the parties' claims were patentably distinct from each other. Instead, the Court found that material limitations within the count were inconsistent with each other.

If a one-way patentability analysis is sufficient, what would it take to establish that there is no interference-in-fact?

To establish that no interference-in-fact exists one would apply the test set forth in 37 C.F.R. §1.601(n) for same patentable invention. Both Invention A would have to be patentable over Invention B assuming Invention B is prior art, and Invention B would have to be patentable over Invention A assuming Invention A is prior art for an interference-in-fact not to exist. If either Invention A or Invention B is unpatentable over the other, then an interference-in-fact exists.

Subsequent to publication of the Notice, the Board has asked two additional questions. The additional questions follow.

First, in a genus/species context, what should be the count? (Assume the genus applicant does not disclose the species). Second, under what circumstances are both parties entitled to their claims.

At the outset of the interference, the count should be the species, i.e., the interfering subject matter. That would give the parties two options.

One, the parties could accept the narrow count. If accepted, and the generic applicant wins on priority, the generic applicant would be entitled to its entire generic claim. The species applicant would not be entitled to its species under 35 U.S.C. § 102(g). If the species applicant wins on priority, the generic applicant would not be entitled to its generic claim under 35 U.S.C. § 102(g). The generic applicant could, however, return to ex parte prosecution and obtain claims that exclude the species assuming that such claims are (1) separately patentable from the species and (2) supported by the specification. The species applicant would be entitled to the species.

In the above cases, there is no violation of any interference rules. The species claim could be designated to correspond to a species count as could a genus claim. In such case, one asks whether the claim of the party is patentable over the count assuming the count is prior art. In this case, the species count anticipates both the species claim and the genus claim.

Obviously, however, the genus applicant may complain that it is unfair to limit him to a count that does not encompass his entire genus and, in some instances, this may be true. However, the interference count need not be sufficiently broad to encompass the broadest corresponding patentable claim of each party. Lee v. McIntyre, 55 USPQ2d 1406, 1411 (Bd. Pat. App. & Int. 2000). Relevant to this issue is a decision of the Commissioner, reproduced in part in Godfredsen v. Banner, 598 F.2d 589, 592, 202 USPQ 7, 10 (CCPA 1979):

Where, as here, the parties both disclose the same three species, that fact does not justify including those species in a single count as member of a Markush group if the Examiner has determined that the three species are patentably distinct inventions. If such a count were permitted, then the party who proved the earliest date of invention as to any one of the members of the group would be awarded priority as to the entire count, i.e., as to all three members. It is not considered that such a result would be consonant with the primary purpose of an interference or with the intent of 35 U.S.C. 135, since there would be no determination of priority as to each of the common inventions claimed by the parties [emphasis in original]. Theeuwes v. Bogentoft, 2 USPQ2d 1378 (Comm'r Pat. 1986).

Thus, in the genus/species context, the starting point should be the species, and, the burden of broadening the count should be placed on the party who wants the broad count. See also Example 36 in 49 FR 48416 (December 12, 1984).

The burden on the generic applicant to broaden the count is analogous to that placed on an applicant trying to swear behind a reference under 37 C.F.R. §1.131. Under such practice, proof of prior completion of a species different from the reference species may be sufficient to overcome a reference indirectly if the reference species would have been obvious in view of the species shown to have been made by the applicant. In re Clarke, 356 F.2d 987, 148 USPQ 665 (CCPA 1966); In re Plumb, 470 F.2d 1403, 176 USPQ 323 (CCPA 1973); In re Hostettler, 356 F.2d 562, 148 USPQ 514 (CCPA 1966). Thus, a party should be entitled to broaden the count to include those species that would render the species of the count obvious.

If the count is broadened, and the generic applicant wins on priority, the generic applicant would be entitled to its entire generic claims. The species applicant would not be entitled to its species under 35 U.S.C. § 102(g). If the generic applicant loses on priority with respect to the broad count, the generic applicant would not be entitled to its generic claim but could return to ex parte prosecution and obtain claims that exclude all of the species of the broadened count assuming that such claims are (1) separately patentable from the species and (2) supported by the specification. The species applicant would be entitled to the species.