Despite the fact that no court has ever found a Markush group improper, the USPTO continues to search for an analytical framework to realize such a finding. As we wound down our New Year’s celebrations, the USPTO “snuck” into the ninth edition of the Manual of Patent Examining Procedure (“MPEP”), which was released in January, a new section 706.03(y), entitled “Improper Markush Grouping.”
The new section has some significant flaws, which are unlikely to decrease the discord between examiner and applicant.
According to the new section, in the case where a claim defines alternative chemical compounds and the alternatives do not belong to a recognized physical or chemical class, the grouping will be improper if either (1) the members of the Markush group do not share a “single structural similarity” (emphasis added) or (2) the members do not share a common use. Further, the members will be considered to meet (1) and (2) simultaneously when “the alternatively usable compounds share a substantial structural feature that is essential to [the] common use (emphasis added).”
Building the new section around the requirement of a “single structural similarity” is not the best choice for a couple of reasons. First, although the new MPEP section attributes this requirement to In re Harnisch, this was actually the phraseology of the underlying board decision that the court in Harnisch reversed, and the phraseology was not adopted by the court as a determinative criterion of its own.
Second, and more importantly, the phrase “single structural similarity” is an oxymoron, and therefore, fraught with ambiguity. The examining corps will no doubt focus on the adjective “single,” and search for and require that a single invariable core group be identifiable in all of the grouped compounds of a proper Markush grouping. This will be in spite of the fact that the phrase concludes with the word “similarity,” which by its very nature permits variability.
Fortunately for prosecutors, the new section includes a subsection IV, which sets forth a number of examples and helps to shed some light on how this analysis should actually proceed. Although the new section indicates the “examples illustrate Markush groupings that have been found to be proper and improper (emphasis added),” I contend, as will be explained below, that all of the Markush groupings are proper.
Example A is the Harnisch case, in which the compounds claimed all contained the coumarin nucleus and, therefore, shared a single invariable core. In Harnisch, the Markush grouping was found to be proper.
Example C is taken from Ex parte Hozumi, and gives one pause due to the shallowness of the analysis. The claims in that case related to compounds of the formula, shown above, which were disclosed as anticancer and antimicrobial agents. The board decision noted that “the claimed compounds are phosphoric acid diesters in which one esterifying moiety is derived from a poly(ethylene glycol) monoether and the other is derived from a beta-aminoethanol.” And, after further noting the variability in the repeating oxyethylene groups (i.e., the parenthetical grouping modified by “n”) and on the nitrogen atom, the board emphasized “the relatively large portion of the compounds in the claimed class which is common to the entire class.” However, the board did not actually identify the large common portion, nor did the board provide any analysis how the large common portion was deemed essential to the shared use. Here again, the Markush grouping was found to be proper.
Assuming I’ve correctly identified the single structural similarity in Hozumi, both Harnisch and Hozumi provide ammunition for the single invariable core camp, as in both cases a single invariable core group was present.
However, Example B, Ex parte Dams is another kettle of fish altogether. As was the case with both Harnisch and Hozumi, the Markush group in Dams was found to be proper. At issue were compounds of the formula,
where, importantly, Y is “an anionic hydrophilic polar group comprising at least one carbon, nitrogen, oxygen, or sulfur atom (emphasis added).” As conceded by the MPEP, the board found this Markush grouping of compounds shared “a common structural feature disclosed as being essential to the disclosed utility of being an anionic surfactant (emphasis added).”
Indeed, it turns out that the CO-Y group—which is highly variable—was proven to be essential to the disclosed utility of the compounds as anionic surfactants. This remains true whether the single structural similarity is CO-Y itself or CO-Y combined with some adjacent other portion of the molecule. Thus, it is impossible to find in this Markush grouping of compounds a single invariable core to which one can assign the anionic responsibility and, consequently, Dams provides support for those who argue that the structural similarity need not be a single invariable core.
While the new MPEP section provides Dams as exemplary, curiously the analysis of the decision does not discuss this important point.
Collectively, these examples, properly analyzed, reveal the single structural similarity need not be a single invariable core, but can be generic, and thus the correct emphasis is on the “similarity,” not that the structural similarity is “single,” i.e., invariable in nature. Provided the alternative compounds share close structural similarity, and this close structural similarity can be viewed as essential to the shared use, this should be sufficient to establish the Markush grouping as proper.
And if, as I contend, close structural similarity is sufficient, then “single” has no role in the phraseology “single structural similarity” and, thus, should be dropped. Presumably, a grouping of compounds that share multiple close structural similarities (and, therefore, are conceptually even closer structurally than a grouping of compounds that share only a single close structural similarity) could not be found improper on this basis. Consequently, the presence of “single” in the USPTO’s new framework only serves to confuse the issue.
A final example D is provided, but the analysis here also leaves much to be desired. The claim is drawn to a herbicidal composition consisting essentially of a mixture of (a) a first herbicide, 2,4-D; and (b) a second herbicide selected from a Markush group of various herbicides. The analysis concedes “[a]ll members of the Markush grouping have a common disclosed use as herbicides,” but alleges the alternatives (b) are not all members of the same physical or chemical class and are not art-recognized equivalents. However, as discussed earlier, where the claim defines alternative chemical compounds, and the alternatives do not belong to a recognized physical or chemical class, the grouping will be improper only if either (1) the members of the Markush group do not share a “single structural similarity” or (2) the members do not share a common use.
A common use is already conceded, so the propriety of the Markush group depends on the presence of a single structural similarity.
Further, in this case, all of the alternative chemical compounds share the component 2,4-D. According to the analysis, the members of the Markush grouping do not share a substantial structural feature because “the members of the Markush group defining component (b) represent a plurality of chemical classes with varying structures.” This may be, but this ignores the admonition from Harnisch that “in determining the propriety of a Markush grouping the compounds must be considered as wholes and not broken down into elements or other components (emphasis added).” The failure here is the failure to view every embodiment of the claimed composition as a single chemical compound, albeit produced by combining individual component compounds. Every such embodiment will contain 2,4-D, which is essential to the herbicidal activity claimed, and, therefore, links all of the embodiments, providing unity of invention. Indeed, there certainly is more structural commonality here than there was in Dams.
In short, this Example D is wrongly decided, and all of the examples provided should be considered to illustrate proper Markush groups.
While I certainly can appreciate the enormous burdens the examining corps faces in examining claims containing large Markush groups, this new MPEP section is sufficiently flawed that it will not be of any help in this regard. Although the court in Harnisch hypothesized that there might be an improper Markush grouping, the fact is that patent prosecutors are sufficiently well-versed in Markush practice by now, nearly 40 years post-Harnisch, that the chances are low they will introduce claims containing an improper Markush grouping. If this new MPEP section was intended to equip patent examiners with a new tool to even the Markush playing field, more work needs to be done, and the USPTO will probably need to look elsewhere to lessen the examiners’ burdens.
If you have any questions about this post, please contact me at email@example.com.
 Actually, the revision was made in August 2017 without much fanfare, then made its way into the newly released MPEP revision.
 206 USPQ 300, 305 (CCPA).
 The court did treat this as an “admission” by the board that the compounds, which the court repeatedly emphasized were all coumarin compounds, were closely structurally similar. Id.
 One can expect the examining corps also to place emphasis on the “a” in “a substantial structural feature”.
 3 USPQ2d 1059.
 Unpublished decision mailed September 13, 2000, during the prosecution of U.S. Patent No. 6,201,122.
 Comparison data in the patent specification involving compounds wherein CO-Y was not anionic proved the resulting compounds were not anionic surfactants, thereby establishing CO-Y must be anionic in order for the compounds to function as anionic surfactants.
 Whereas Examples A-C are all taken from appeal decisions, this Example D is taken from examination guidelines.
 “2,4-D” is an abbreviation for 2,4-dichlorophenoxy acetic acid.
 206 USPQ at 305.
 According to the Harnisch court, inventions lacking “unity of invention,” which, thus, would be drawn to “inventions which are truly independent and distinct (emphasis in original),” might present improper Markush groups. See 206 USPQ at 306.
 MPEP 803.02 has long been on the books for exactly this purpose, but probably because it still entails examination of the entire scope of the Markush claim as originally presented, its use is disfavored/discouraged in practice. The new section would have the examiner maintain the improper Markush rejection “until (1) the claim is amended such that the Markush grouping includes only members that share a single structural similarity and common use; or (2) the applicant presents convincing arguments why the members of the Markush grouping share a single structural similarity and common use.” If the examiner maintains applicants’ arguments as unconvincing, applicant either must appeal (and wait years for a decision) or amend to some scope the examiner will accept as clearly containing a single invariable core group.