Board of Patent Appeals and Interferences
Patent and Trademark Office (P.T.O.)
*1 EX PARTE BHARAT B. AGGARWAL, PATRICK W. GRAY AND GLENN E. NEDWIN
Appeal No. 90-3041
January 14, 1992
HEARD: November 13, 1991
Application for Patent filed May 9, 1986, Serial No. 732,312; which is a continuation-in-part of application Serial No. 616,503, filed May 31, 1984, now abandoned. Recombinant Lymphotoxin.
Carolyn R. Adler et al. for Appellants
Primary Examiner--Garnette D. Draper
Before Goolkasian, Tarring and Meros
This is an appeal from the examiner's final rejection of claims 57 through 66. Claims 1 through 53 remain pending but have been withdrawn from consideration as directed to a non-elected invention.
Appellants' invention concerns the treatment of tumors in animals by the administration of a therapeutically effective amount of a cytotoxic protein known as lymphotoxin. The claims describe the lymphotoxin as "a product of expression in a recombinant host cell".
Mammalian lymphotoxin is a glycoprotein, i.e., the amino acid molecules of the protein are glycosylated. [FN1] However, when the protein is recombinantly produced in transformed E. coli, the resultant lymphotoxin expressed is unglycosylated because the E. coli microorganism is incapable of glycosylating a protein. It is part of appellants' discovery that the unglycosylated lymphotoxin expressed by E. coli is cytotoxic in vivo.
Claim 57 is illustrative of the claimed invention and reads as follows:
57. A method for the treatment of tumors comprising administering to an animal in a physiologically innocuous dosage form a therapeutically effective amount of a cytotoxic lymphotoxin as a product of expression in a recombinant host cell selected from the group of cells from multicellular organisms, eukaryotic microbes and prokaryotes.
The references relied on by the examiner are:
Aggarwal EPA 0100641 Feb. 1984
Khan et al., Human Lymphokines, "Pre-Clinical and Phase I Clinical Trials with Lymphotoxin," 1982, pages 621 through 629.
Ransom et al. (Ransom (S)), JNC1, "Lymphotoxin Prevention of Diethylnitrosamine Carcinogenesis In Vivo," Vol. 69, No. 3, September 1982, pages 741 through 744.
Evans et al. (Evans (V)), Cellular Immunology, "Comparative Effectiveness of Lymphotoxin Anticarcinogenic and Tumor Cell Growth-Inhibitory Activities," Vol. 76, 1983, pages 295 through 303.
Ransom et al. (Ransom (R)), Cancer Research, "Molecular and Biological Characterization of Anticarcinogenic and Tumors Cell Growth Factor-inhibitory Activities of Syrian Hamster Lymphotoxin," Vol. 43, November 1983, pages 5222 through 5227.
All of appellants' claims stand rejected for lack of utility under 35 U.S.C. § 101, lack of enablement under 35 U.S.C. § 112 and obviousness under 35 U.S.C. § 103.
Before discussing the rejections, it is helpful to consider the subject matter and scope of appellants' claims.
The words "animal" and "tumor" are used by appellants in their broadest sense, i.e., any animal and any type of tumor. In its broadest reasonable sense the term "animal" means something other than plant and includes dogs, cats, cows, birds, humans, etc. In its broadest reasonable sense, the term "tumor" designates any tumor, whether malignant or benign. From a utility and enablement viewpoint, we construe appellants' claims as being very broad in scope and directed to treating any tumor (malignant or benign) in any animal.
"Lymphotoxin" is defined in the specification as a "biologically active polypeptide having a region demonstrating substantial structural amino acid homology with at least a portion of the lymphotoxin amino acid sequence shown in Fig. 2a" (page 11). The term "biologically active" is defined as including not only cytotoxic activity but immunological cross reactivity with a cytotoxic lymphotoxin, or the ability to compete with cytotoxic lymphotoxin for lymphotoxin cell surface receptors. The lymphotoxin of the claims is limited to lymphotoxin having cytotoxic activity.
What is encompassed by "substantial structural amino acid homology" is described as "greater than about 60 percent ... of the amino acid residues in the polypeptide are the same or conservative substitutions [FN3] for the corresponding residue(s) in the sequence of Fig. 2a" (specification, page 12). [FN4]
Appellants' claims are of very broad scope. Indeed, while the protein depicted in Fig. 2a is a characteristically human lymphotoxin, the claims are not limited to the use of human lymphotoxin type protein on homologous (human) tumors but are sufficiently broad to include within their scope the use of human lymphotoxin on heterogenous tumors of heterogenous species, i.e., dogs, cats, etc.
Claim 57 uses product-by-process terminology to describe the lymphotoxin used in the claimed process as a "product of expression in a recombinant host cell." Importantly, the host cell may be selected from any of the various types of cell lines and microorganisms which may be used to express recombinant lymphotoxin. Among the cell lines said to be useful is the human HeLa cell line (page 25, line 25 of the specification). When human cells, other animal cells, or even yeasts are used to express recombinant proteins, the resultant proteins are expected to be glycosylated. Accordingly, appellants' claim 57 is of sufficient breadth to encompass the treatment of tumors with human cytotoxic lymphotoxin which (a) is the recombinant expression product of a transformed human or mammalian host cell, and (b) is expected to be glycosylated.
We consider first the rejection under 35 U.S.C. § 103. It is the examiner's position that appellants' method for treating tumors utilizing recombinant lymphotoxin would have been made obvious by the teachings of any one of Aggarwal, Ransom (R or S), Evans (V) or Khan. It is the examiner's position that appellants' claims call for the treatment of a tumor with human lymphotoxin regardless of whether the lymphotoxin was made recombinantly or purified from natural tissues. The examiner takes the position that whether extracted from humans or prepared recombinantly, lymphotoxin is essentially the same material or a minor modification thereof which would have been expected to effectively treat at least the tumors described in the references, all of which are concerned with natural lymphotoxin extracted from mitogen-stimulated lymphocytes.
*3 We have carefully considered all of appellants' arguments but are in complete agreement with the examiner on this point with regard to (1) claim 57 which includes the use of lymphotoxin prepared from recombinant human cells, (2) claim 58 wherein the recombinant host cell is a yeast cell, [FN5] (3) claim 59 wherein the recombinant host cell is a mammalian, non-human cell, and claims 62, 63 and 64. We affirm the examiner's rejection of these claims. Claims 60, 61, 65 and 66, however, refer to the use of lymphotoxin prepared using recombinant cell lines which are incapable of glycosylating the expressed protein. We reverse the examiner's rejection of these claims.
Appellants argue that one skilled in the art would not have reasonably expected that lymphotoxin produced in recombinant cell culture would have the same tumor fighting characteristics as native lymphotoxin because recombinant expression would be reasonably expected to vary (or to eliminate) the native glycosylation thought to be critical for the activity of lymphotoxin. Appellants base this argument on the teachings of the Proctor reference cited by appellants during prosecution to teach that glycosylation affects the conformation of the protein. It is appellants' position that this knowledge would have led one skilled in the art to recognize that a non-glycosylated lymphotoxin would not have been expected to have the requisite activity. Appellants also rely on the teachings of Ransom et al., WO 85/04662, and other references for a teaching of the importance of proper glycosylation. (Ransom et al teaches that treatment of lymphotoxin with neuraminidase, which cleaves glycoproteins at sialic acid residues, inactivates the lymphotoxin.)
We are unpersuaded by these arguments as they concern claims 57 through 59, 62, 63 and 64 because these claims are not limited to non-glycosylated material and are directed to the use of recombinant lymphotoxin which either has the same glycosylation as normal lymphotoxin, especially if prepared using a human host cell, or is a minor glycosylation variant if produced in cells such as yeast, Chinese hamster or monkey kidney.
What is at issue is the reasonableness of the examiner's conclusion that at least some lymphotoxin encompassed in the aforesaid claims will be glycosylated such that it is either the same as or a minor variant of the glycosylated purified human lymphotoxin of the prior art. In re Best, 562 F.2d 1252, 195 USPQ 430 (CCPA 1977); Ex parte Skinner, 2 USPQ2d 1788 (BPAI 1986). Having reviewed the art, we consider the examiner's position to be quite reasonable with regard to claims 57 through 59, 62, 63 and 64 which include the use of lymphotoxin prepared using cell lines known to be capable of glycosylation.
Claims 60, 61, 63 and 66 are directed to the use of lymphototoxin prepared in non-glycosylating cell lines. We have reviewed the Proctor, Delente and Ransom references relied on by appellants and have also reviewed the Toth and Granger reference discussed at page 3, line 28 through page 4, line 4 of the specification. We note that there appear to be differences in the teachings of the references regarding the effect of glycosylation. Toth and Granger, for example, taught that the removal of sialic acid from lymphotoxin containing lymphocyte supernatants by neuraminidase treatment did not adversely effect the in vitro lytic activity of the lymphotoxin. However, Proctor teaches that glycosylation plays a definite role in the conformation of the protein and is necessary for utility. Having weighed the Toth and Granger statements versus those of Proctor and the other references cited by appellants, we are in agreement with appellants that the evidence of unobviousness weighs more heavily and, accordingly, we reverse the examiner's rejection with regard to the claims involving use of the unglycosylated protein.
*4 The examiner has requested this Board follow Ex parte Gray, 10 USPQ2d 1922 (BPAI 1989) and affirm the examiner's rejection in its entirety. We have followed Gray, with regard to the obviousness of using a glycosylated recombinant lymphotoxin for treating tumors. However, this case is distinguishable because, unlike Gray, there is evidence of record herein which indicates that the unglycosylated lymphotoxin would not have been expected to retain its cytotoxic activity. The Proctor reference clearly states that glycosylation is necessary for providing the proper conformation to the protein. Accordingly, one skilled in the art would have preferentially utilized yeast or a mammalian cell line to prepare the recombinant lymphotoxin protein and would have been dissuaded from utilizing a procaryotic cell line such as E. coli. Similar evidence was not before this Board in Ex parte Gray. The different factual situation herein warrants a different decision at least with regard to the use of cell lines which do not glycosylate proteins.
Claims 57 through 66 stand rejected under 35 U.S.C. § 101 and 35 U.S.C. § 112 for lack of utility and enablement. It is the examiner's position that the invention as disclosed and claimed is inoperative over the broad range of cancer/tumors set forth therein, not useful throughout its scope (35 U.S.C. § 101) and not enabled (35 U.S.C. § 112). We treat this as one rejection [FN6] and affirm the rejection.
In re Langer, 503 F.2d 1380, 1391, 183 USPQ 288, 297 (CCPA 1974), requires the Patent and Trademark Office to consider a specification disclosure, which corresponds in scope to the subject matter sought to be patented, as sufficient to satisfy the utility requirement of 35 U.S.C. § 101, unless "there is reason for one skilled in the art to question the objective truth of the statement of utility or its scope" (emphasis added).
A review of appellants' specification reveals that although the specification contains many broad statements regarding utility, and describes administration of lymphotoxin via virtually all known routes of administering anti-cancer substances, the actual exemplification of utility is sparse.
Examples 4 and 5 of the specification describe preparation of recombinant lymphotoxin and evaluation of activity by in vivo tumor necrosis assay using mice as the host. The lymphotoxin produced was somewhat active by this test. This is the only test data of record.
Accordingly, though appellants' specification is replete with "generalities" regarding the use of lymphotoxin, it is virtually devoid of teachings that indicate the broad scope of lymphotoxins claimed has the broad scope of utility asserted. See In re Gardner, 427 F.2d 786, 166 USPQ 138 (CCPA 1970).
*5 It is the examiner's position that the treatment of tumors is an essentially unpredictable activity and that, at the time the invention was made, the activity of lymphotoxins was not well understood and not predictable. In support of the rejection the examiner has set forth a list of reasons why there is a strong likelihood that recombinant lymphotoxin, a protein, may not be effective, e.g., proteolytic degradation before the lymphotoxin ever reaches the target, etc. (Answer, page 5, paragraph 1). The examiner's reasoning is set forth below in the footnote. [FN7] The examiner's statement was not challenged by appellants in the Reply Brief.
Appellants have also taken the position that the activity of lymphotoxin is heterogenous in nature and that human lymphotoxin or some material similar thereto would be effective against any tumor in any species of animal. Aside from broad statements in the specification, there is no evidence supporting this position which is counter to general beliefs. Lymphotoxin is a protein and most organisms generate antibodies to invading foreign proteins.
Our review of the prior art relied on by the examiner and appellants confirms the examiner's skepticism as follows.
(a) The article published by Evans and Heinbaugh titled "Lymphotoxin Cytotoxicity, a Combination of Cytolytic and Cytostatic Cellular Responses," states emphatically, page 354, "[T]umor cells while usually inhibited in their growth or destroyed cytolytically by homologous lymphotoxin can be unaffected or even stimulated to proliferate in the presence of heterologous lymphotoxin" (emphasis added).
(b) The Kahn reference relied on by the examiner does describe previous work showing regression of malignant melanoma in a dog following local injections of partially purified human a-lymphotoxin. However, the bulk of the paper reveals that lymphotoxin was relatively ineffective against breast cancer and ineffective against other cancers in in vitro stem cell assays, and showed "no overall reduction in the tumor mass" or "no antitumor effect" when the lymphotoxin was injected in humans in vivo. Kahn theorized that the in vitro tests measured only the direct cytotoxicity of lymphotoxin and that participation of other (unknown) components of the cellular and immune response was necessary to achieve success in vivo. Kahn suggests further experimentation to determine how to use lymphotoxin.
(c) Published PCT application No. WO 85/04662 to Ransom et al, discussed by appellants at page 16 of the Brief, describes tests conducted on lymphotoxin supplied by one of the inventors and states:
(1) Lymphotoxin ... while readily lysing murine alpha L929 tumor cells, does not possess any detectable antihuman tumor cell activities (page 4).
(2) Human lymphokine and lymphotoxin, however, had little or no cytolytic activity towards human leukemia, sarcoma, and carcinoma cells (page 27).
On the record before us, then, the practical utility of lymphotoxin when used against cancer is disputed by those skilled in the art. The evidence of record is such that there is considerable doubt that those skilled in the art would be willing to accept appellants' in vitro tests and in vivo tests as established models predictive of utility against tumors in humans. See In re Jolles, 628 F.2d 1322, 206 USPQ 885 (CCPA 1980). [FN8]
*6 The examiner had more than adequate reason to doubt the objective truth of the broad statement of utility set forth in appellants' specification. We are also of the opinion that the examiner has provided adequate evidentiary support for the assertion that appellants have not complied with the utility requirements of 35 U.S.C. § 101 and enablement requirements of 35 U.S.C. § 112 with regard to human lymphotoxin itself and also with regard to the broad generic class of lymphotoxins encompassed by appellants' claims and the broad generic class of tumors set forth therein.
Appellants urge that the Sherwin Declaration under 37 CFR 1.132 evidences the usefulness of the claimed invention. We are not persuaded by this declaration. Paragraph five of the declaration, which admits, inter alia, that cancer chemotherapy is an empirical art, confirms the examiner's position. The declaration admits that it not possible, particularly in the early stages of development of a given candidate chemotherapeutic agent, "to predict in advance whether a selected tumor is susceptible to treatment with the agent."
The Sherwin declaration does not state that lymphotoxins are useful against tumors or that practitioners skilled in the art know how to use lymphotoxin. All that is said is that (1) oncologists will "not be misled" by the assertion of broad anti-tumor activity, (2) routine and conventional clinical studies will be conducted to "more fully refine the activity of the agent," and (3) though "extensive and burdensome," the studies involve nothing more than "routine manipulations."
While appellants would identify these clinical studies as "routine and conventional", the evidence of record in this case is to the contrary, at least with regard to lymphotoxin. The Kahn article and Ransom patent publication indicate that at the time the instant application was filed, lymphotoxin had no practical utility against tumors.
There is no question that appellants have made an important discovery with regard to chemical compounds (proteins) which are the subject of serious scientific investigation but of unverified and speculative utility. Appellants urge that in such situations it is in the public's interest that patent applications be filed early rather than waiting for what may be a long period of experimentation. A similar situation was before the Supreme Court in Brenner v. Manson, 383 U.S. 519, 148 USPQ 689 (1966) wherein, with full consideration of the public's interest, the Court decided against the grant of a patent. The Brenner court warned that "a process patent in the chemical field, which has not been developed and pointed to the degree of specific utility, creates a monopoly of knowledge which should be granted only if clearly commanded by the statute." The court noted that "such a patent may confer power to block off whole areas of scientific development, without compensating benefit to the public".
*7 Case law subsequent to Brenner is receptive to early filing of applications in the biomedical field so long as the patent applicant, when properly challenged by the examiner, can provide evidence showing substantial activity in screening tests customarily used and accepted as predicative of human activity for the type of chemical tested. Of course, the evidence presented must be commensurate with the scope of utility asserted and the subject matter claimed. See In re Buting, 418 F.2d 540, 163 USPQ 689 (CCPA 1969).
The following comment by the court in In re Hawkins, 486 F.2d 569, 574, 179 USPQ 157, 161 (CCPA 1973) quoting In re Argoudelis, 434 F.2d 1390, 1395, 168 USPQ 99, 104 (CCPA 1970) is pertinent to situations such as this:
... for the satisfaction of the second aspect of section 112--that of establishing the filing date as the prima facie date of invention--'it is essential that there be no question that, at the time an application for patent is filed, the invention claimed therein [here, the use of lymphotoxin] is fully capable of being reduced to practice (i.e., that no technological problems, the resolution of which would require more than ordinary skill and reasonable time, remain in order to obtain an operative, useful embodiment).'
The examiner's rejection of claims 57 through 59 and 62 through 64 under 35 U.S.C. § 103 is affirmed. The examiner's rejection of claims 60, 61, 65 and 66 under 35 U.S.C. § 103 is reversed. The examiner's rejections of claims 57 thorugh 66 under 35 U.S.C. § 101 and 35 U.S.C. § 112 is affirmed.
No time period for taking any subsequent action in connection with this appeal may be extended under 37 CFR 1.136(a). See the final rule notice, 54 F.R. 29548 (July 13, 1989), 1105 O.G. 5 (August 1, 1989).
BOARD OF PATENT APPEALS AND INTERFERENCES
John T. Goolkasian
Henry W. Tarring, II
Edward J. Meros
FN1. Proteins are high molecular weight copolymers composed of a variety of amino acids linked by peptide bonds to form polypeptide chains. Some proteins contain covalently attached carbohydrate units bonded to oxygen or nitrogen molecules of some amino acids. These proteins are known as "glycosylated" proteins.
FN2. From an enablement viewpoint, there is a distinction between claims drawn to the use of a chemical compound and claims to the compound itself. See In re Bundy, 642 F.2d 430, 209 USPQ 48 (CCPA 1981) and its discussion of In re Gardner, 475 F.2d 1389, 177 USPQ 396 (CCPA 1973). The latter is controlling herein.
FN3. By "conservative substitution" is meant the replacement of one amino acid by another resulting in little or no change in the properties of the protein.
FN4. The paragraphs extending from line 30 on page 12 through line 8 on page 13 of the specification appear to broaden the description of "substantial structural homology" to the extent that greater than about 60 percent amino acid residue correspondence need not be achieved for the entire protein.
FN5. Yeast cells are known to be capable of glycosylating recombinant proteins expressed therein.
FN6. Reliance on either or both of 35 U.S.C. § 101 and the first paragraph of 35 U.S.C. § 112 as the statutory basis for this rejection has been judicially approved. In re Fouche, 439 F.2d 1237, 169 USPQ 429 (CCPA 1971).
[A]ppellants have not shown that the protein would not be inactivated (before producing the desired effect) by such mechanisms as proteolytic degradation; immunological inactivation; poor biodistribution due to a short serum half-life of the protein (rapid excretion/poor bioavailability); deactivation of the LT by the liver; hydrolysis in serum; binding to plasma protein; inability of the protein to reach the target site because the protein may not be able to cross the mucosa or is absorbed by fluids, cells or tissue where the protein has no effect; that the circulation to or in the target area may be insufficient to carry the drug; or a large enough effective local concentration may not be capable of being established, particularly as its relates to ensuring that adverse side effects do no [sic, not] occur.
FN8. The in vitro and in vivo test conducted by appellants have not been shown to have been recognized by the art as predictive of success with lymphotoxin in the treatment of tumors. This is unlike the situation in Cross v. Iizuka, 753 F.2d 1040, 224 USPQ 739, 747 (Fed.Cir.1985), wherein the evidence of record showed that there was a reasonable correlation between the disclosed in vitro activity and an in vivo activity.