■ site map
BLOG: June 2010 - December 2013
III - ALTERNATIVE BREAST CANCER SCREENING TESTS
5. AMAS test
Is there a single screening test that would cover all the cancers? Sounds too good to be true but, in fact - there is such test. It is called anti-malignin antibody screen or, for short, AMAS. Although often referred to as "new test", it is about as old as the standard X-ray mammography: it was FDA-approved for cancer screening as far back as 1977.
Part of its lasting novel appeal is in being fairly obscured for decades, apparently - for no good reason. Nearly all of the research on its effectiveness for over two decades since the mid 1970s was led by its discoverer, Dr. Samuel Bogoch, MD, PhD, and his wife Eleanor Bogosh, MD, with the first wide scale trial in 1982 (Determination of anti-malignin antibody and malignin in 1,026 cancer patients and controls: relation of antibody to survival) supplied the evidence that
one specific antibody (protein) is indeed elevated in the presence of almost any cancer.
This was confirmed by several authors' and independent studies after that. The test is in clinical use for over three decades, supplied by Oncolab, Boston MA.
The test is based on determining the level of this specific protein - named malignin by Dr. Sam Bogoch - created by the body in the presence of almost any malignancy. Bogoch's data on thousands of patients with various forms of cancer indicate high test sensitivity (93%, i.e. 7% false-negative rate) and specificity (95%, i.e. 5% false-positive rate).
In 1994, a study of the AMAS test used in the actual clinical practice of 42 physicians from 11 states (nearly 200 patients total) recorded near 100% sensitivity and specificity (Early detection and monitoring of cancer with the anti-malignin antibody test, Abrams et al.).
In 2000, an independent small study of test's efficacy with breast cancer come to essentially the same numbers as Bogoch (Thornthwaite JT. Anti-malignin antibody in serum and other tumor marker determinations in breast cancer). And the most recent, small study, found that test's efficacy, with 62% sensitivity and 67% specificity - is insufficient (Harman et al. Discrimination of Breast Cancer by Anti-Malignin Antibody Serum Test in Women Undergoing Biopsy, 2005).
Assuming no bias, Bogoch's initial test efficacy numbers carry more weight than the smaller studies, simply for much larger number of cases. Two of the small studies fall completely in line with Bogoch's initial results, so it is only Harman et al. that suggests the opposite, i.e. insufficient test sensitivity and specificity. Taking a closer look at this study, its main limitations are obvious:
(1) number of participants - 71 breast cancer cases scheduled for biopsy - is too small to be statistically significant,
(2) the only exclusion criteria was prior history of malignancy, and
(3) there was no screening to determine whether a "false" positive test result was caused by the presence of some other form of cancer.
Insufficiency of a single exclusion criteria is due to the test limitations reported by Dr. Bogoch, namely that body's malignan production
becomes irresponsive to the presence of cancerous growth after a significant cancerous mass is present for about 3 years, or longer, or when the immune function is grossly suppressed.
The latter commonly results from conventional chemo and radiation treatments, but can also be caused by some other medical treatments, or by medically-unrelated factors. Since only 17 participants had biopsy-confirmed breast cancer, the presence of undetected longer-term cancerous growth of any other form in only a few participants, could alone significantly degrade study's test specificity results, making its false-positive rate much higher than what it was.
On the other hand, relaying on the conventional breast cancer diagnoses induces another obvious flaw, since it entirely neglects overdiagnosis. Knowing that 1 in 4 to 1 in 3 of conventionally diagnosed breast cancers are pseudo-disease implies that a significant portion of the diagnoses could have been pseudo-disease and, therefore, not true false-negatives.
Related to this, the number of positive (malignant) and suspicious biopsy results - 17 the former and 12 the latter - indicates selection bias toward women likely to have breast cancer. The usual rate of positive biopsy result is between 5% and 10%. Assuming that half of suspicious biopsies were, actually, breast cancer, puts the rate of positive biopsy results in the study at 32%. Relatively large number of participants with existing breast cancer would actually increase the chances that the portion of those with longer-term tumors will be closer to its statistical average. But if uncontrolled, as it was in this study, it can be a source of unknown sample deviations, invalidating study results.
The supplied data on study participants is, in general, very scarce, leaving open the possibility of some forms of selection bias. For instance, there is no data on age, except a hint in the Discussion section that study's participants were middle-aged women.
Also, the study does not use repeated AMAS test when the initial result is unclear, which is the usual practice for obtaining better indication in borderline cases, or to rule out so called transient malignan elevation. Obviously, such study approach does not allow for the test to reach its full potential.
Harman et al. do acknowledge the limitation resulting from not checking for the presence of other forms of cancer. It offers a quick "what if" calculation based on some research data on false negative breast cancer biopsies, which would, in the "worst case scenario" increase the AMAS study specificity from 67% to "only" 78%. That, of course, is not appropriate. Presence of any undetected cancer form, not only false negative biopsy for breast cancer, would result in a false "false positive" AMAS result.
Despite the study acknowledging one of the limitations related to its low number of participants - specifically, the possibility that its particular group of participants had "an unusual number of patients with malignant disease but low AMAS titers" - and stating that, consequently, "only further studies with larger numbers of patients would elucidate this possibility" the lead author does not refrain from publicly criticizing AMAS test as having insufficient sensitivity and specificity (hence, "If you're going to do a mammogram, why do this?").
In addition to its obvious design flaws, this sort of questions author's impartiality, while clearly suggesting author's preference. At the time when the standard X-ray mammography's cult status seems to be threatened from all sides, the mighty industry behind it is
likely to try to downgrade any competition.
Creating studies fitting that purpose is no news in the realm of profit- and status-driven medicine.
But the study does touch a very important distinction and limitation of the AMAS test: it is a cancer test, but it is not cancer-specific. In other words, it indicates the presence of malignant growth, but does not specify its type, location or size/stage. For that, additional, different tests are needed.
According to Oncolab's site, the test was used successfully for screening in selected high-risk populations (e.g. chemical workers) and preclinical detection of cancer in a relatively small portion of medical-surgical cases. However, it has never been tested for screening broad populations. Obviously, it is not a test that can be used to screen for any specific cancer form, but considering that
breast cancer makes only about 15% of all cancers in
it seems very worthwhile to have it done, from about age 50 on, for the price comparable to that of standard mammography.
At present, Oncolab sends a free kit for taking blood sample. The test alone is $165, and has to go through your doctor, who receives the test result. If it is negative, the chances that you don't have neither breast, nor any other cancer, are about 13 to 1. For comparison, with the standard mammography it is less than 4 to 1 for breast cancer alone.
If the test is positive, it is about 19 to 1 that you have malignant growth - generally in an early stage - somewhere in the body. Or, vice versa, the chances of a false alarm are 1 in 20. Again, with the standard mammography it is about 9 to 1 for breast cancer alone that the positive test is true, or about 1 in 10 chance of a false alarm.
In short, what standard mammography does for breast cancer screening, AMAS test does for screening all cancers - only better. It would certainly make more sense to
screen with AMAS test for all cancers,
and in the case of confirmed presence of malignancy move on to what seems to be the most logical cancer-specific test. Since AMAS test detects malignant growths in its early, preclinical stage - generally earlier than standard mammography in the case of breast cancer - it buys enough of bonus time for additional tests.
Being a simple blood test, AMAS itself is essentially non-invasive. Its limitations affect relatively few, mainly those with advanced cancer forms, and those with severe immuno-deficiencies, whose bodies may lose the ability of producing this specific immuno-response to the presence of malignancy.
And, if breast cancer is primary concern, the positive AMAS test could be best followed with the good old physical breast exam. More on this test on the following page.