This is an appeal from a decision of the Board of Appeals of the United States Patent Office which affirmed the examiner's rejection of the only claim of appellant's application for a patent on "Isoalloxazine Compound."

1. 6, 7-diethyl-9-(D-1'-ribityl)-isoalloxazine, having the structural formula:

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

The references relied on by the examiner and the board are:

  Tishler et al.  2,350,376  June  6, 1944
  Flexser et al.  2,456,395  Dec. 14, 1948
  British patent    628,410  Aug. 29, 1949

As the appealed claim indicates, this application relates to a single organic compound. The issue is whether this compound would have been obvious in view of this prior art at the time the invention was made.

The application has been amended so that the sole portion thereof relating to utility now reads:

"6, 7-diethyl-9-(D-1'-ribityl)-isoalloxazine is a potent riboflavin inhibitor or antagonist and is useful for producing riboflavin deficient laboratory experimental animals such as rats and mice. As compared with the prior art procedure of maintaining experimental laboratory animals on a riboflavin deficient diet to induce riboflavin deficiency, it has been found that by placing these animals on a diet containing 6, 7-diethyl-9-(D-1'-ribityl)-isoalloxazine, riboflavin deficiency can be more rapidly induced thereby significantly shortening the preparatory time to secure animals in this condition. These riboflavin deficient animals are used experimentally for nutrition studies." According to the record, riboflavin2 has the following structure:

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

It is evident that the claimed compound differs from riboflavin only in having ethyl groups, C₂H₅- or CH₃CH₂-, at the positions where riboflavin has methyl groups, CH₃-. Apparently these compounds are both "isoalloxazines."

Of the claimed compound, appellant also states in his brief:

"* * * It differs from riboflavin also in being an antimetabolite rather than a metabolite. In other words, it competes for riboflavin in the metabolic processes and thus interferes with metabolism. Thus, the compound is capable, among other things, of producing avitaminosis, a condition characterized by a deficiency of vitamins."

The Tishler et al. patent discloses a process for the preparation of isoalloxazines with the generic formula:

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

wherein R₁ is selected from the group consisting of hydrogen, alkyl and aryl; R₂ is selected from the class consisting of hydrogen, alkyl, polyhydroxylalkyl, and polyacyloxyalkyl; * * *.

The Tishler et al. process is not the same as appellant's process. In each of the four specific compounds disclosed, the top and bottom R₁ groups are hydrogen and the two side R₁ groups, corresponding to the 6- and 7-positions in riboflavin, are methyl. R₂ in each of these four examples is different, being variously hydrogen, methyl (a specific alkyl), ribityl (a specific polyhydroxyalkyl), and tetraacetylribityl (a specific polyacyloxyalkyl). Tishler et al. express no other specific preference for R₁ and R₂ and disclose no utility for isoalloxazines prepared by their process.

The Flexser et al. patent has been considered by the examiner and the board as "corresponding to" the British patent.3 Both patents disclose a process for the manufacture of isoalloxazines which is similar to though not identical with appellant's process. No generic formula for isoalloxazines is disclosed in either patent although both disclose as a starting product a compound expressed generically as:4

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

wherein R is a member of the group consisting of alkyl, polyhydroxyalkyl, and polyacyloxyalkyl, and wherein R' and R" are members, not necessarily identical, selected from the group consisting of hydrogen and lower alkyl groups which contain not more than 8 carbon atoms.

It appears that isoalloxazines, including riboflavin, are produced according to these two patents by "condensation" of this starting material with another compound, alloxan, and that the hexagonal ring shown above becomes the left-hand ring of the isoalloxazine structure as exemplified by riboflavin, supra. The N in -NHR of the starting material becomes the 9-position in the final isoalloxazine, R remaining attached to the N. It is noted that in the generic starting material structure, R' and R" are not attached to any particular angles of the hexagonal ring. Appellant refers to R' and R" as "floating" radicals. By that we understand R' and R" may be variously disposed on the ring of the starting material so that any two of positions 5, 6, 7 and 8 of the final isoalloxazine structure may be occupied by R' and R". Riboflavin and tetraacetylriboflavin are the only specific isoalloxazines disclosed and no other preference for specific R, R', and R" groups or specific positions for the R' and R" groups is expressed in either patent. There is no utility disclosure in the British patent. The only statement in Flexser et al. relevant to utility is:

"* * * This class of compounds has recently assumed technical and therapeutic importance, riboflavin or vitamin B₂ being an outstanding example thereof."

Although the examiner "deemed that applicant's compound is anticipated by the British Patent" and is "generically taught by the Flexser et al. and Tishler et al. patents," the board agreed with appellant that his claimed compound is not specifically disclosed in the patents and assumed that "the examiner intended the rejection to be based upon the British patent on the ground of lack of patentable invention, since he [the examiner] admits that the claimed compound is not specifically disclosed in the British patent but falls within the purview of the teaching therein." After noting the examiner's further rejection in view of the Flexser et al. and Tishler et al. patents, the board decided it would "consider the rejections as based solely on the ground of lack of patentable invention [i. e. obviousness] over any one of these patents." Thereupon the board sustained this rejection.

"* * * Appellant's principal contention is based upon the ground that the compound which is claimed exhibits an unexpected result, namely as a riboflavin antagonist or antimetabolite."

The board then referred to "an article in the Journal of Nutrition, Vol. 47, No. 4, August 1952, pages 539-560." Although this article has not been made part of the record in this court, it appears that it was written by appellant and another, that it discloses the claimed compound, and that it was first cited by appellant in a letter to the examiner, after the latter's final rejection, for the reason that the article "confirms emphatically the position previously set forth [by appellant] that the claimed compound has startlingly new and unexpected properties."

"* * * The article which is referred to by appellant in the Journal of Nutrition clearly indicates on page 540 that two analogs of riboflavin, including as one the 5-6-dimethyl-9-(D-1'-ribityl) isoalloxazine, also displayed the inhibiting action in rats and that some of these compounds contain the riboflavin activity in the nutrition of L-casei and B. lactis acidi while other [sic] did not. These analogs all fall within the purview of the disclosures of the patents relied upon by the examiner. The property of inhibition of riboflavin activity is, therefore, not a property which is solely possessed by the compound which is specifically claimed herein, but is also a characteristic of certain other members of the class. It would be expected and not surprising, therefore, that other analogs, not specifically spelled out in the patent [patents?], would possess either of the two properties. It is our conclusion, therefore, that the reliance upon one of the characteristics which is known to be possessed among the members of this class does not lend patentability to a compound which is impliedly conceded as suggested by and is clearly within the purview of the prior art references. * * *"

The Journal of Nutrition article appears to be the only source of evidence considered by the board in deciding that the anti-riboflavin activity of the claimed compound "would be expected and not surprising." It is noted that the appealed application was filed October 20, 1955 as a continuation-in-part of a then pending application, Serial No. 253,000, filed October 24, 1951. Although the examiner was informed of the facts that the article discloses the claimed compound and had been published more than one year before the filing date of the appealed application, he did not consider the article a bar to the claim. Accordingly, appellant must have been given the benefit of the filing date of his parent application. For this reason and in the absence of any other reasons to the contrary suggested by the record, we believe the article should not be considered further by us. Again we point out that the article itself is not a part of the record of this case. As far as we can tell, it was appellant's own knowledge that the board relied upon for its position and there is nothing in the record to indicate that appellant was merely reciting the state of the art. As in all causes brought before us, we must decide the issues from the record as presented to us.

Furthermore, we do not find in the record any other factual basis in support of the board's contention that the antiriboflavin activity of the claimed compound "would be expected and not surprising," or that it was known before the filing date of appellant's parent application (which date was before the date of the article) that this property is a characteristic of other members of the class to which the claimed compound belongs, insofar as this record reveals.

Turning next to the question of obviousness in view of the three patents of record, we believe that appellant's compound would not have been obvious at the time the invention was made to a person having ordinary skill in this art. We reach this result in two steps, first treating the patents as a disclosure of riboflavin, and second, treating them as an additional disclosure of certain classes of compounds encompassing both appellant's compound and riboflavin along with a very large number of other compounds.

The board considered the claimed compound to be "the next higher homologue of riboflavin (vitamin B₂)." There has been much discussion in the record and briefs relating to "homology." A comparison of the structural formulas of these two compounds shows clearly that there is substantial structural similarity. But more appears from the facts of this case than structural similarity, facts which raise genuine questions as to the real significance of such bare structural similarity, whatever label may be attached to it.

It appears that riboflavin is an essential part of the diet of normal healthy rats. Appellant has found that rats fed his compound develop riboflavin-deficiency symptoms even though their diet also includes riboflavin. There is no evidence in the record which would lead one skilled in this art to expect that the differences in molecular structure between riboflavin and appellant's compound would cause this difference in properties. The former compound is a vitamin, the latter an antivitamin; the former is a metabolite, the latter an antimetabolite; the former acts to promote the well-being of the animal, the latter acts to its detriment. We find it difficult to conceive of a better example of a difference in kind than is presented in this case and we also find in view of this record that this difference was unexpected and unobvious.

The solicitor urges that "there is no specific evidence in the record of differences in chemical and physical properties of the prior art compound [riboflavin] and the claimed compound." In view of the biochemical differences which we have just discussed, we can only assume the solicitor is urging that while differences in chemical properties might be persuasive of patentability of the claimed compound, differences in biochemical properties are not to be considered. We see no reason to distinguish between chemical and biochemical properties and no reason or authority for this position has been presented to us.

The second step in our reasoning involves the additional disclosures of the reference patents. The board stated with regard to the properties of the claimed compound:

"* * * We are not persuaded, however, that a showing of unobvious or unexpected property must necessarily lead to the conclusion that the claimed compounds are patentable. As held by the Court of Customs and Patent Appeals in the decision In re Finley, 36 CCPA [998] 999; 1949 C.D. 284; 624 O.G. 363; 174 Fed.(2d) 130; 81 U.S.P.Q. (BNA) 383, the showing of unforeseen properties is not conclusive of patentability and may require the consideration of other factors. * * *"

The "other factors" considered by the board were the teachings of the Journal of Nutrition article which are not before us, and the board's opinion that the claimed compound is "suggested by and is clearly within the purview of the prior art references." The solicitor interprets the latter "factor" to be based on the following:

"(1) that the reference patents disclose a class of compounds, the di-lower alkyl 9-(D-1'-ribityl) isoalloxazines, having the alkyl groups in any two of the 5, 6, 7, and 8 positions; (2) that the class is merely exemplified in the prior art by the 6, 7 dimethyl compound (riboflavin); and (3) that the claimed compound is clearly a member of that class."

The solicitor has also restated his position as follows:

"* * * the prior art applied against the appealed claim discloses generically a limited class of compounds which clearly includes and suggests the compound defined in the claim on appeal * * * and in addition specifically discloses a compound which is a close lower homologue of that compound. * * *"

We do not agree that the "limited class" suggested by the solicitor is disclosed in any of the reference patents. Moreover, we do not believe that any class of compounds disclosed in these patents is sufficiently limited to "suggest" appellant's compound.

Referring to our discussion of Tishler et al., supra, we note that "lower," as in the solicitor's "limited class," is not disclosed as a limitation on "alkyl." Although one of the specific examples of Tishler et al. recites a d-ribityl group at the 9-position as in the solicitor's "class," this one example is not enough to constitute a class. Further, although the four specific examples of Tishler et al. all include methyl groups at the 6-and 7-positions and hydrogen at the 5-and 8-positions, the 9-position is occupied by a different group in each example. Any limited class to which these four examples belong clearly would not be the "limited class" suggested by the solicitor. The only class expressly disclosed by Tishler et al. is that whose generic formula is given supra. Although we agree that this broad class includes appellant's compound, we do not agree that this compound is obvious.

Referring next to our discussion of the British patent and Flexser et al., supra, we agree that isoalloxazines produced from compounds within the scope of the generic starting material, presented supra, will have lower alkyl groups (containing 1 to 8 carbon atoms) in any two of positions 5, 6, 7 and 8, thus satisfying that requirement of the solicitor's "limited class." However, R in the generic starting material, appearing at the 9-position in the final isoalloxazine, may be alkyl, polyhydroxyalkyl and polyacyloxyalkyl. We believe that the class of compounds disclosed by Flexser et al. and the British patent is much broader than the "limited class" suggested by the solicitor since the variation in R of the patentees' starting material and hence the group at the 9-position in the final product is essentially limitless. Although the broad class suggested by these two patents includes appellant's compound, we do not believe that this compound is obvious.

The solicitor cites In re Rosicky, 276 F.2d 656, 47 CCPA 859, in support of his contention that the concept of appellant's compound would be obvious from the disclosures of the prior art. In that case, Rosicky was claiming xanthene compounds with a particular "side chain," -CH₂CH₂-Z, where Z could be a piperidine or pyrrolidine ring. We held these compounds to be unpatentable in view of a reference patent which disclosed xanthene compounds with a similar "side chain" expressed generically and attached in the same position. It was disclosed specifically in the patent that the ethylene radical could be one part of that side chain as in Rosicky's compounds, and that piperidine could be the other part of the side chain also as in the compounds claimed by Rosicky. We held that the claimed compounds were within the scope of the limited generic disclosure of the patent and that it was of no moment that Rosicky's specific compounds were not named in that patent. Further, we pointed out that the compounds of the patent were disclosed therein to have one of the two properties alleged by Rosicky for his compounds.

Clearly In re Rosicky is distinguishable from the situation now before us where no specific preference for ethyl is expressed in the patents, where no specific pair of positions for the two ethyl groups is expressed, and where the properties of the claimed compound differ from those of riboflavin in an unexpected manner.

It is true that appellant's compound is encompassed by the broad classes of compounds disclosed by these reference patents but so are many many other compounds. We do not think that progress in the useful arts would be promoted by permitting broad theoretical disclosures as these to preclude appellant from obtaining a patent for his invention.

It is also true that appellant has invented a compound with close structural similarity to another known member of these broad classes, namely riboflavin. But consideration of this matter returns us to the first step in our reasoning, i. e. consideration of the dissimilar properties of the compounds being compared. At the most, we think one skilled in this art would be taught by the reference patents that other groups than those present in the riboflavin structure can be attached to the parent isoalloxazine structure. We doubt that this fact is a noteworthy addition to the knowledge of an organic chemist of ordinary skill because he knows this is true of all such parent chemical structures. Though this would be obvious to him, it does not follow that all new compounds so produced would be obvious in the sense of the patent law.

The decision of the board is reversed.

Reversed.