Soon after its appointment in 1959, the Cancer Advisory Council learned that a biochemical test for the diagnosis of cancer was in use in California. Because the test had apparently not gained wide acceptance since it was initially reported in the literature (1955), and since the Council members were aware of the search for a reliable diagnostic tool for early diagnosis of cancer, it was decided to engage in a study to determine the merits of the test. A study was proposed in which R.H. Beard, Ph.D., one of the authors of the test, would perform tests on 200 specimens; simultaneous tests on the same specimens would be performed in the laboratories of the State Department of Public Health for comparison.
The test as reported by Beard and his associate, T.C. Terrell, M.D., both of Fort Worth, consists of the use of quantitative determinations of chorionic gonadotropin which they claim is produced in all cases of malignancy and excreted in the urine, as diagnostic and prognostic aids in cases of malignancy. The technique of the test is as follows: an acetone precipitate of an aliquot of a 24-hour specimen of urine is redissolved in a phosphate buffer; to one measured amount of the buffer-precipitate solution, anthrone-sulphuric acid reagent is added directly and the intensity of the resulting color due to the presence of chorionic gonadotropin, anterior pituitary gonadotropins and polysaccharides, is measured in a colorimeter. To eliminate one of the elements contributing to color density, chymotrypsin is added to a second portion incubated overnight and then treated as above. The two colorimetric readings are converted to milligrams and the difference between them determined. The difference represents the chorionic gonadotropin fraction which has been responsible for the greater color intensity of the undigested specimen and which has been destroyed during the incubation period. From this figure, the amount excreted in 24 hours is calculated.
The critical difficulty with a test of this nature is the introduction of exceptions—false positives—which in this case included pregnancy and a variety of clinical conditions. Beard denoted the conditions which might generate false positives and these were excluded when selecting patients for study, but in no phase of the comparative studies did the results satisfy the standards adopted by the National Research Council: that false positive tests in apparently normal individuals under age 30 should not exceed 5%, nor that fewer than 90% positive tests be recorded in individuals with cancer. These values for specificity and sensitivity are observed throughout the nation, but the studies performed by Beard failed to meet these standards.
Dr. Beard agreed to the objective criteria and methodology recommended by the Department; specimens were furnished by the Cancer Research Institute of the University of California School of Medicine. Dr. Beard cooperated fully with the study design, but during its course introduced some deviation in the basic technique.
After approximately one-half of the specimens had been tested, an examination of findings in the Department Laboratory revealed that the digested specimens on cancer patients consistently showed a greater color concentration (less transmission) than the undigested specimen precisely the opposite reaction expected with the test. In only 2 of the 23 cancer patients did both samples of a split specimen demonstrate the expected trend. Because the results in the Department were so inconsistent and unpredictable, this phase of the study was suspended.
Although the actual tests on patients performed by the Department Laboratory were worthless, several desirable objectives were realized: chiefly that the color reaction in the authors’ standard must be due primarily, if not exclusively, to lactose and not to chorionic gonadotropin. Most significant, however, was the fact that the Anthrone test was completely unreliable.
Dr. Beard’s results did not satisfy the rigorous scientific standards recommended by the National Research Council. On a single specimen of a patient’s urine sent the same day as received, positive tests were obtained in 59% of cancer patients and 47% of controls, and on specimens sent a day apart, the percentages shifted to 52% of cancer patients and 49% of controls. When results on paired specimens were examined without regard for time of submission, there was disagreement in 38% of specimens from cancer cases and in 48% of the controls. The inconsistencies established in Beard’s own laboratory served to reduce any credence given to this diagnostic technique.
Action of the Cancer Advisory Council
In addition to evaluating the results of these studies, the Council requested the opinions of the Deans of California Schools of Medicine and of medical experts, Ralph W. Weilerstein, M.D., and William A. Atchley, M.D. The Anthrone test is not regarded by any of these individuals as valid or suitable for instruction. The failure of the test· to meet criteria for sensitivity or specificity implies that the Anthrone test is completely unreliable and unacceptable as a diagnostic or prognostic test for cancer.
On August 17, 1964, the Cancer Advisory Council concluded its studies by finding that the Anthrone Test was of no value in the diagnosis of cancer and recommended to the Director of the State Department of Public Health that appropriate action be taken to prohibit the administration of said test or one substantially similar thereto.
This article was posted on September 13, 2014.