Wesley Gale Irons appeals from a decision of the Board of Appeals affirming the examiner's rejection of claims 5 to 17 in his application1 relating to an antiguanidine polypeptide factor having the formula:

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE where n is an integer and each R is selected from the group consisting of the hydrogen atom and the monovalent radicals CH₂CH₂COOH, CH₂COOH, CH₂CH₂CH₂NH — C.NH — NH₂ and CH₂CH₂CH₂NH₂, the values of n and R being such that the molecular weight is between 10,000 and 11,000 * *.

Appellant claims the process for making the polypeptide, the polypeptide product in terms of the process for making it and the process of using the polypeptide. Claims 10 and 14 are illustrative:

"14. An anti-guanidine polypeptide factor produced in a process from beef liver tissue after removal of proteins, albuminoids, histones, basic amino acids, proteoses and peptones to leave a peptide-containing extract by the improvement comprising adding to said extract a water-miscible solvent selected from the group consisting of water-miscible alcohols, ketones and ethers so that the concentration of the solvent does not exceed 60% by volume, allowing the material to stand at a low temperature for some hours until it forms a top liquid layer of an orange-red colour separated by a solid wafer of a gray or olive-green colour from a lower liquid layer, separating the upper liquid layer, adding thereto a water-miscible solvent selected from the group consisting of water-miscible alcohols, ketones and ethers and cooling to a low temperature for home hours to form a flocculent, slightly pinkish precipitate, separating the solid from the liquid, macerating the solid with an aqueous liquid having a pH of between 7 and 9, separating the liquid obtained and subjecting it to dialysis, and thereafter heating the dialysed liquid to a temperature of 43 to 49° C. until a heavy flocculent precipitate forms, separating the liquid and subjecting it to distillation at a reduced pressure and a temperature not exceeding 43° C. until a solid residue of said factor is formed."

"10. A method of treating human arthritic diseases comprising the administration of antiguanidine factor of Claim 14 to said human."

The sole ground of rejection is lack of utility under 35 U.S.C. § 101. The basis of the examiner's rejection on utility can best be understood by considering the actions of the appellant and examiner in chronological order. The application contained the following statement of utility:

"This invention relates to novel compounds and the production thereof, and is particularly concerned with new substances suitable for the treatment of arthritis and related diseases."

There is a detailed explanation of the method of administering the claimed compound followed by 5 examples of clinical results obtained by using the compound. A typical example is reproduced below:

"* * * A female aged 59 suffering from Arthritis Deformans was given by injection one dose per week for five weeks of a preparation according to the invention consisting of 5 ccs. of liquid containing 1.5 mgm. of the solid produced according to Example 1. No adverse effects were observed and there was complete relief of pain and marked reduction of joint swellings."

In response to the examiner's rejection for lack of utility, the appellant submitted the following evidence of utility:

(1) A letter from Jack R. Knudson, M.D., to appellant stating that 370 patients, 310 of whom had rheumatic disease, had been treated with appellant's compound and that:

"There was complete subjective and objective improvement in many cases and profound improvement in the majority of cases, far superior to any previous therapy. In about 5% of the cases, the response was minimal, attributed to either inadequate therapy, or the presence of chronic infection."

(2) A letter from M. P. Ream, M.D., to appellant stating that 518 patients, 361 of whom were arthritic patients, were treated with appellant's compound and that "Less than 5% failed to receive some relief." (3) An article in the American Institute of Homeopathy2 giving 10 case histories of the treatment of arthritis patients with appellant's compound. In all cases pain was relieved and in 9 cases other symptoms such as swelling or mobility were improved.

The examiner rejected appellant's proofs, stating

"The tests set forth in the record of this application have been carefully considered, however since the tests are without controls, the uncontrolled tests do not convince the examiner that applicant's composition has the utility alleged for it. Double blind tests with the protocol of the experiment set forth should be performed and presented as proof of utility."

In support of his position, the examiner cited the following references:

Beecher, "Appraisal of Drugs Intended to Alter Subjective Responses, Symptoms," 158 A.M.A.J. 399.

Beecher, "The Powerful Placebo," 159 A.M.A.J. 1602.

The Washington Daily News, May 6, 1959, p. 18.

The Daily News article reports that of 426 arthritis patients treated at the Arthritis Clinic of Cook County Hospital in Chicago with sugar pills and salt water shots, 80% "responded favorably." The sugar pills and salt water shots or any other pharmacologically inert substances are known as placebos which are used to make patients think they are actually receiving something good for them.

The Beecher articles point out that placebos have an average effectiveness of about 35.2% in the relief of pain. It is pointed out that because of this effect, very special controls are necessary to appraise drugs having as their chief purpose a change in symptoms. Beecher contends that when the placebo effect is involved, hundreds of uncontrolled cases and years of effort may be valueless in evaluating a drug. The necessary controls suggested by Beecher involve use of a "double blind test." In the double blind test each subject is given a placebo, a standard reference drug and the drug being tested in random order. Neither the subject nor the observer knows what or when test agents are employed.

In answer to the examiner's demand for controlled tests, appellant then submitted an affidavit of Robert W. Hillman, M.D., a professor at the State University of New York, College of Medicine and lecturer in Fundamental Biostatistics and Applied Statistics (Medical). Dr. Hillman analyzed the data presented in the letters by Dr. Knudson and Dr. Ream and in the Ream et al. article. The data was analyzed in accordance with a text, Hill, "Principles of Medical Statistics" (7th ed. 1961). The data was compared with data in five publications reporting experiences with rheumatoid arthritis and osteoarthritis. Dr. Hillman's calculation indicated that appellant's data was statistically significant.

The examiner was not convinced by the affidavit because the data upon which it was based were merely the conclusions of Doctors Ream and Knudson. The examiner also pointed out that the Hill text states:

"In practical statistical work the occasions upon which such past experience is available as a safe and sufficient guide are relatively rare. As a substitute for past experience the experimenter takes a control group and uses it as the standard of comparison against the experimental group."

The examiner further relied upon the following article:

"Clinical Therapeutic Trial of a New Drug," Johns Hopkins Hospital Bulletin, Sept. 1949, p. 221.

"It may be well to mention two or three types of `controls' that are not satisfactory. One is the historical control, for example the observation that in the pre-drug days, only m per cent reacted favorably and with the drug in use this is changed to n per cent. If everything remained constant in the situation, this would be all right, but there is no assurance that this is the case. There are differences in treatment aside from the specific drug, difference in type of patients presenting themselves, often due to the very fact that a drug is available; difference in the general economic and health set-up of the patients."

With regard to statistical analysis of results, the Bulletin states:

"In dealing with clinical studies of this type, it is not at all uncommon for the clinician to seek statistical aid after the results are in, to find out whether differences are significant, or in some other way to have assistance in the analysis. If such a move is contemplated it is well to seek this assistance at the outset for there are no statistical maneuvers that will rescue a study that has not been well controlled. Statistical constants can be computed and it can be determined whether a difference is significant, but whether or not the difference is associated with the therapy or with the unknown or unthought of factor, such as ballast in the ship, is anybody's guess."

Relying upon the articles described above, the examiner finally concluded that appellant's evidence would not convince the medical profession that the claimed product was effective in the treatment of arthritis and thus did not convince him.

The board, in affirming the examiner, indicated that it was influenced by the fact that it was "dealing with a disease whose treatment depends as much on subjective symptoms as it does on objective symptoms." In addition to agreeing generally with the examiner's position, the board stated:

"In our opinion, evidence should have been presented which showed either objective improvement, such as, for example, decreased swelling of the hand or X-ray observation showing improvement in the bone structure, or alternatively, demonstrated subjective improvement, by means of properly controlled tests."

Although the above statement is apparently intended to establish some criterion for acceptable proof of utility, it fails to do so in at least one respect. The board indicated that evidence showing decreased swelling would be acceptable in place of controlled tests demonstrating subjective improvement. Appellant points to 15 case histories describing objective improvement such as increased flexion and decreased swelling. This apparently is the exact type of evidence the board was looking for and yet they failed to mention it. It may be they failed to consider it. The solicitor argues that the board found that 15 case histories were not enough. We do not, however, find anything in the board's opinion that would even imply such a holding. We avoid deciding the case on this issue because the board's position is unclear.

The board has made its position clear on the need for controlled experimentation and on this issue we disagree. We acknowledge that the placebo effect may be appreciable and should be taken into account when subjective symptoms are being treated. However, we cannot agree that under the particular facts in this case controlled tests such as double blind tests are the only means of establishing utility under 35 U.S.C. § 101. The Patent Office has failed to show why the historical control used in the Hillman calculations would introduce error. The solicitor relies upon In re Novak, 306 F.2d 924, 49 CCPA 1283. In the Novak case we held that when utility as a drug, medicant, and the like in human therapy is alleged, it is proper for the examiner to ask for substantiating evidence unless one with ordinary skill in the art would accept the allegations as obviously correct. The solicitor argues that the substantiating evidence must be of a type that would convince one skilled in the art. We agree that the proofs of utility should be convincing to one skilled in the art, but we cannot agree with the degree of proof required by the Patent Office.

The solicitor contends that past experience which was used in Dr. Hillman's statistical analysis cannot replace a control group. The Johns Hopkins Bulletin previously quoted suggests that the historical control is less desirable because factors such as the methods of treatment and type of patient may not remain constant. Each change in test conditions apparently will affect the certainty and accuracy of the test results. There is apparently little doubt that a double blind control is more reliable than a historical control. But, the evidence clearly indicates that both types of control are accepted. In fact, the Beecher and Johns Hopkins articles were apparently written to persuade those skilled in the art to use controlled experiments in place of whatever less acceptable techniques they are now using in evaluating the performance of drugs. Thus it would appear that tests of the type conducted by appellant are convincing to many skilled in the art. We feel that the 888 tests found to be statistically significant by Dr. Hillman establish that appellant's compound possesses anti-arthritic utility to a degree sufficient to satisfy the patent statute.

The burden the Patent Office would place on appellant would, in effect, require proof beyond a reasonable doubt that the claimed compound possesses the alleged utility. The Patent Office has required one of the most rigid and complete tests known to the medical profession. We feel that in view of the factual situation presented by the record before us, appellant has met the burden of showing compliance with section 101.

The decision of the board is thus reversed.

Reversed.