History in the Making…or Naught?

On June 17, 2015, the first patient (perhaps the first anywhere) was enrolled in a “proof of concept” trial that has the potential to alter forever how we screen for breast cancer – by using a blood test. The intent of a screening blood test would be to use it in tandem with mammography. Then, if the mammogram is negative, but the blood test positive, the patient would proceed to a breast MRI, the latter highly sensitive for the detection of breast cancer (capable of finding 2X to 3X the number of breast cancers found on mammograms in head-to-head trials of MRI and mammography).


Currently, screening with breast MRI is limited to high-risk patients only, but there are two problems here. Even at the highest known risks (mutations in BRCA or comparable genes), the MRI yield is only in the 3% range. That is, 100 MRIs to find 3 cancers. This is barely cost-effective, so risk-based screening is already at its limits of practicality. The other problem is that this approach excludes the vast majority of women who are headed for breast cancer, given that most who are eventually diagnosed have no major risk factors.


Yet, even so-called “tumor markers” are unreliable when it comes to diagnosing recurrent cancer, so how can we be discussing a blood test that detects breast cancer at its earliest stages? The answer comes through new technologies, one of which allows us to detect tiny quantities of autoantibodies that develop in response to the earliest forms of breast cancer. In fact, a blood test for early detection may not even work for detecting late stage disease. No one knows at this point.


The enrollment of this first patient in a trial expected to last several years is the culmination of 24 years of dedication to this single agenda. And, a special nod to the women of Oklahoma who let my team draw blood for basic scientists around the world, wherein we have shipped over 10,000 samples during my tenure at Mercy Hospital–OKC. University researchers as well as biotech companies have tried diligently to develop such a test, but have been cut short by failure. Out of many collaborations over the years, only 3 have made it to formal clinical trials, and 2 of those failed. One remains.


Provista Diagnostics, Inc (New York/Scottsdale) is completing an 11-site formal clinical trial using a blood test in a diagnostic role, to help radiologists decide whether or not to proceed with biopsy. Through a proprietary algorithm, 10 biomarkers, half of which were developed at Dana-Farber/Harvard by Dr. Karen S. Anderson et al, are combined into a “yes” or “no” report suggesting the need to proceed with biopsy or not. Mercy Breast Center contributed over 500 samples in the development phase of the test, and this multi-site diagnostic trial of 1,000 patients will draw to a close in August 2015, having included samples from the Cleveland Clinic, Mayo Clinic, Scripps and other busy sites.


Now, enter my “proof of concept” study being conducted at Mercy Hospital – OKC, which reflects a different use of the same blood test, for screening asymptomatic women rather than diagnostic help for the radiologist. Here, blood will be drawn on patients who are already scheduled for breast MRI in our high-risk screening program, under an approved Institutional Review Board protocol. Thus, there will be no further implications of a positive blood test result – the patient will already have had the definitive test, i.e., the MRI. After several hundred patients have been enrolled, we will be able to calculate what the outcomes would have been if we had acted only on the blood test results combined with mammography. Mathematical modeling already shows us that MRI yields, even with a modestly good blood test (say, 80% sensitivity and 90% specificity), should vastly improve upon the cancer yields of 3% currently being achieved with the risk stratification approach, with yields possibly up to 10% or more, highly cost-effective.   Plus, a screening blood test will open up the option of MRI screening to all women rather than only those at high risk.


Now, jump one step further into the future. With a very good blood test (95% sensitivity and 95% specificity), one could screen on the basis of the blood test alone. That is, no mammography at all, unless the blood test is positive. A positive test would then require mammography first to localize and diagnose the tumor. If mammograms were negative, then the patient would proceed to ultrasound or MRI. Even further into the future, the blood test could be improved to the point of detecting only those cancers that are biologically a threat to the patient, thus avoiding “overdiagnosis.”


Why hasn’t this been a top line agenda item for research funding? Because of the false belief that mammography detects 90% of breast cancers. “Why search for a blood test when we already have mammography?” But in recent years, the oft-quoted 90% has been adjusted down to 80% and “lower in younger women with dense tissue.” But that’s still not enough adjustment. The true sensitivity is lower still. Mammograms need help. The first time I heard of anyone else working on a blood test for early diagnosis of breast cancer was the 1st government grant (in 2004) issued for this same purpose, awarded to a researcher at the Fred Hutchinson Cancer Center in Seattle. I sent blood samples in collaboration with the PhD recipient of the grant, but the proposed line of study dead-ended. But I’m getting ahead of myself…


In 1991, I attended a breast conference in Dallas, Texas where a young Dr. Steve Harms showed the audience his experimental images using MRI to detect breast cancer. One case of breast cancer after the next, he reported, “and the mammograms were negative.” This was one of several turning points in my career, but I began a personal search to try and determine the incidence of missed cancers on mammography. My first academic paper was one of the early analyses as to why some cancers don’t appear on mammography. Of course, in the back of my head was one overriding concern – once breast MRI becomes available, how are we going to recommend such a cumbersome test to the entire population? Even with upcoming improvements in 2016 that are going to shorten the MRI study into a 10-minute test, there’s still the IV to be started, injection of a contrast agent, etc.


While still in academics, I started a weekly research meeting where those scientists at the OU Health Sciences Center campus interested in breast cancer research met and exchanged ideas, an exercise that continued every Friday the entire decade I was at that location. In 1991, shortly after I had seen the presentation by Dr. Harms on MRI, one of the senior scientists who attended that meeting (Paul McKay, PhD, who had been one of my professors in medical school) showed me an article about a proposed blood test for breast cancer. The light bulb turned on, and it’s never gone out.


Dr. McKay was based at the Oklahoma Medical Research Foundation, and he managed to get funding to bring the author of the paper (Dr. Chaya Moroz) from Tel Aviv to Oklahoma City, so that I could assist her in validating and bringing her blood test to the clinic. I had explained to Dr. McKay how the blood test could be the spark needed to select patients for additional imaging beyond mammography, especially MRI. In actuality, Dr. McKay and OMRF provided half the funding to launch this relationship with Dr. Moroz. The other half was raised by an army of volunteers, led by Cheryl Browne and the late Patricia Browne, who had devoted themselves whole-heartedly to the establishment of Oklahoma’s first multidisciplinary breast center that was under construction when the blood test opportunity arose. It was a match made in heaven – a novel breast center with a novel research agenda – but with one problem. In spite of my trip to NYC to meet with the senior partner of Penny & Edmonds intellectual property law firm handling the rights to the test, as it turned out, the blood test did not pan out. 10,000 samples and 24 years later, we are on the cusp of finding out if we can revolutionize breast cancer screening…or naught.