All Malignancy Tests are not Benign
November 1, 1999 7 minutes read
Special Feature
All Malignancy Tests are not Benign: The Case for and against Anti- Malignin Antibodies
By Thomas J. Smith, MD, FACP
True story:
A 40-ish-year-old woman went to one of the burgeoning "anti-aging" clinics in the sunny southwest, to evaluate it for her firm. As part of her trial run through the clinic process she underwent a complete testing—complete history and physical; a multitude of lab tests; including all automated chemistries, heavy metals, trace elements, and about 15 types of steroid hormone metabolites; bone density screening; and an anti-malignin antibody screening test.
She was panicked when she got the call from the physician at the clinic—the test was positive, and she might have cancer, or she might develop it.
This was a surprise since she was healthy, exercised strenuously and ate well, and had no risk factors for any malignancy.
Her gynecologist had never heard of the test. She called me, and I thought that I might have heard of the test a long time ago. It wasn’t in DeVita, and it wasn’t in the last major review of tumor markers in Seminars in Oncology, and it had not crossed my desk when we looked at virtually thousands of pages on tumor markers to create the ASCO guidelines. What would you tell her?
The Anti-Malignin Antibody Screening Test
Not wanting to be too "allopathic," I tried to objectively review the information the clinic sent her on the test. After all, I am personally an exercise and fruit/vegetable fanatic. The test was backed up by references in the Lancet and elsewhere. It sounded too good to be true.
In 1977, Bogoch described a 10,000 molecular weight protein found in glial cells, malignin.1 In a 1981 letter to Lancet Bogoch reported a seven-year study of cancer patients showed that concentrations of anti-malignin antibody were 59 mcg/mL in normal people and 237 in patients with cancer.2 Of those with levels over 135, 83% were dead in 12 months; of those alive at four years, 90% had levels less than 135. There is not much detail in a letter to the Lancet, nor is the information "refereed" to have an external judge of quality. The letter states that the information will be published elsewhere: it appeared in the Journal of Medicine in 1982.3 The study is sketchy on the clinical details of the patients, and does not specify if any cutoff point for "positive" was predetermined. There were no false-positives comparing normal, healthy outpatients to active cancer patients, and 93% of the cancer patients showed a positive test. But, 5.4% of the non-cancer outpatients had an elevated AMA test, and 8.8% of the non-cancer inpatients had a positive test.
The only other large-scale test was reported in Cancer Detection and Prevention.4 Tests were ordered when either signs or symptoms suggested cancer or the patient requested the test, or to monitor therapy. Two hundred and eight tests were done on 181 patients. Of 125 patients with cancer in the differential diagnosis, AMA was elevated in 21—all of whom had cancer (of many types.) AMA was normal in 97—none of whom had cancer (although there was one "initial test false-positive" repeat normal). Seven of the 125 (5.6%) suspected cancer patients had persistent elevations but had not developed cancer. Transient AMA elevation occurred in 3%.
The test was positive in a variety of patients, from those with 1 cm resected breast cancer to acute leukemia. In general, there is not enough clinical detail presented to make sense of the correlation to disease status; many patients with positive values are listed as "clinical evidence of residual disease" but I cannot tell what that means. Metastatic? Recurrent? The cancer patients with normal values were considered false-negative; if however, they died within a year, they were counted as terminal cancer (these patients were not expected to have elevated values, the explanation was that the immune system gears down near death).
What can we learn from this data?
There may be something here, not to be dismissed. Perhaps the test is like the C-reactive protein (prognostic in myocardial infarcts), or sedimentation rate—allows one to detect illness, just not what type of illness.
But the data are almost fatally flawed. The premise, that all cancers elicit one type of antibody response, makes little sense. Or that it always goes down in the year before death. Therefore, these are true-negatives, not false-negatives.
And the tests reported are interesting but not conclusive. In the first study, was the 135 mcg/mL set as a cutoff point a priori, or only found to best fit the data? In the second, the clinical details and verification of clinical status are so sketchy that the study is rendered uninterpretable.
There are no data on:
• Screening asymptomatic populations
• The performance of the test under monitored, carefully controlled clinical settings that would allow external verification of diagnosis and clinical state.
Bogoch states that large, prospective determinations are being done in the United Kingdom for cancer screening, but I have not found any published information on Medline.
So, in my opinion (and that of other tumor marker mavens whom I asked), it sounds too good to be true.
Why do people want such tests?
We all want the perfect marker to diagnose and monitor. For the two new adjuvant breast cancer patients I saw today, such a test would be a godsend. "You have no cancer cells, get out of here! But, you have 1200 in your liver and bone marrow, and we’re going to try to get them with adriamycin and cyclophosphamide, then paclitaxel and tamoxifen." This would make my life easier, but we are just not there yet.
Why do people order such tests?
I can only speculate on why the anti-aging clinic ran all these tests. The clinic ran 118 separate tests on their Multi-Chem. I have never personally ordered "alkenals (total)" to assess oxidative damage; three different types of tocopherols; IGF1; urine melatonin; retinyl palmitate, lutein, cryptoxanthin BE and about 10 other "antioxidants" strontium, molybdenum, or 19 total trace metals. I was pleased to see that my new patient’s Coenzyme Q10 was normal. As far as I know, deficiency/excess states for most of them have not been described, and even less is known on the implications for treatment. We know from the CARET and other trials that giving large doses of antioxidant vitamins can worsen things for those at risk for lung cancer.
The requisition form for the test is intriguing. Both patient and physician must sign "As in all clinical laboratory tests, I understand that the AMAS test is, not by itself, diagnostic of the presence or absence of disease, and that its results can only be assessed as an aid to diagnosis, detection, or monitoring of disease in relation to the history, medical signs and symptoms, and the overall condition of the patient."
What are the consequences of a false-positive or negative?
We don’t really know. It is evident that there are false-positives, which my new patient is likely to be. We don’t have the requisite information to reassure her that the test is predictive.
For such imprecise tests, a note of caution. People with the BRCA 1 and 2 genes have been studied the most carefully. Knowing that you were negative for the gene should make you less worried, right? Not always. And finding out that you were positive should give you control over destiny, and less worry, right? Again, not always. A significant number of people who test positive or negative have enhanced distress. Those who choose not to have the BRCA1 test may also have enhanced distress. Those most distressed before the test are more likely to have worse distress afterwards.5-7 The counseling for such tests is complex, and should, at a minimum, include true and false rates and implications of the test. The elements of informed consent about tests with unclear implications have been worked out by genetic counselors over Huntington’s chorea and other hard-to- diagnose/impossible-to-treat illnesses.
If I were to send the test, I would certainly make sure that my patients signed the disclaimer.
Take Home Message
There are some tests that have not been rigorously evaluated, or even scrutinized by the FDA. These tests should be done with extreme caution, if at all, and only with informed consent. My wife the pediatrician once told me, "Doing a diagnostic test is like picking your nose in public. Before you do it, figure out what you are going to do if you find something."
Doing these tests may have little consequence if truly positive or negative. But even when the tests correctly point out a serious condition, there can be serious psychological distress. And when the test is false-positive, the distress will be palpable—as with my patient. Such tests are not benign.
References
1. Bogoch S. Natl Cancer Inst Monogr 1977;46: 133-137.
2. Bogoch S, et al. Lancet 1981;18:141-142.
3. Bogoch S, et al. J Med 1982;13:49-69.
4. Abrams MB, et al. Cancer Detec Prev 1994;18:65-78.
5. Croyle RT, et al. Health Psychol 1997;16:63-72.
6. Lerman C, et al. Oncology 1996;10:191-195.
7. Lerman C, et al. J Clin Oncol 1998;16:1650-1654.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content