ATLANTA — To just say that Mary-Claire King, PhD, has simply been “instrumental“ in the way physicians look at diagnosing breast cancer would be an understatement. King was a pioneer in helping the medical community understand the BRCA1 gene and its role of genetics in the development of breast and ovarian cancer.
The University of Washington geneticist spoke to a large audience during the opening day of the American Association for Clinical Chemistry's (AACC; Washington) annual meeting held here in the Georgia World Congress Center.
The geneticist was given The Wallace H. Coulter Lectureship Award by the AACC for her work in developing diagnostic strategies for cancer.
King rose to prominence back in 1990 when she demonstrated that a single gene on chromosome 17 (BRCA1) was responsible for a great deal of the many breast and ovarian cancers. The discovery of that gene changed the course of study of numerous other common diseases.
The technique King developed to identify BRCA1 has since proven valuable in the study of many other illnesses, and King has built on to that research by identifying BRCA2, and extending her technique to other diseases and conditions.
Speaking from in front of the podium to the crowd King gave a look that suggested her work was far from complete.
“Knowing what we know about BRCA1 and BRCA2, why do we have any women who have BRCA1 and BRCA2 that go on to have breast cancer,“ said King, who was posing a question to the audience.
She said that researchers face several challenges in identifying patients that are at risk for breast cancer.
• The mutational spectrum is extremely broad.
• Many women with the mutations have no family history or breast cancer.
• There are more culprits than just BRAC1 and BRCA2.
These challenges led to King and her team developing the Broca test – named after Paul Broca, a 19th century French physician that dealt with the histology of cartilage and bone, but also studied cancer pathology, the treatment of aneurysms, and infant mortality.
“What we did was develop a one stop genetic test to simultaneously capture and sequence all genes known to [be associated with] ovarian and breast cancer,“ she said. “We know that targeted enrichment for all known genes that are predisposed to breast and ovarian cancer combined with next generation sequencing is accurate and sensitive for detecting inherited mutations in these genes.“
King displayed the results of an early study of the test that identified 15 small mutations from 10 different genes. She added that there were zero false positive calls for mutations in each sample.
Since the test is gaining traction she told the audience what she envisioned next for the test.
“We hope that the speed which is fast and the cost which is low, will allow for the immediate translation into clinical labs,“ she said.
But there are some issues that she said the test could run up against and she listed several key concerns she had with the application moving forward.
“If we can technically detect all of the mutations in all known genes for inherited breast and ovarian cancer, what should we be concerned about at the level of the big picture.“
She said that responsibility for vetting these new tests was a primary concern.
“I really worry about this class of testing getting into the hands of people who don't know what they're doing,“ she told the audience.
She added that she was part of a CLIA Certified lab, but wondered what would happen as the test became more mainstream and what effect that could have on the accuracy of results.
“The actual competency carrying out the test is one issue . . . but another issue is the quality of the test, she said. “I hope that [AACC] will think about the application of these very deep sequencing testing tools and think about how they might best be brought under proper regulation so that they are useful to people.“
King said that there should also be a mechanism that included a provision for insurance coverage for testing and follow-up.
King's presentation could be thought of as the kick off to the nearly week long conference that will run throughout Thursday. This year's theme is a telling one – “Bringing New Value to Diagnostics“ The conference boasts taking a deeper look at personalized medicine and more diagnostic applications.
Combo molecular tests best at detecting thyroid cancer
A picture displaying what appears to be part of a female's abdomen pops up on the screen – with the words next to it “The fastest growing cancer among women isn't what you think.“
“That isn't the side of the abdomen and chest wall – that's actually the chin and neck,“ Mira Milas director of the Thyroid Center at the Cleveland Clinic told a jam-packed room at the Georgia World Congress Center during a panel session at the AACC meeting. “While Thyroid cancer cases will never amount to the total amount of cases that are diagnosed from breast cancer – in terms of the rapid rise in incident rates – [thyroid cancer] occupies the number one spot for women and the number two spot for men.“
Milas spoke in what was part of a panel titled “Molecular Diagnostics and Molecular Genetics of Thyroid Cancer.“
“I think its a topic that is incredibly important and is here to stay and will be part of the daily care of our patients,“ Milas told the audience. “So what I hope to illustrate are some of the real life scenarios with applications of molecular markers and highlight some that are available nationwide.“
Doctors often biopsy suspicious thyroid nodules, to evaluate the potential for cancer. Typically thyroid nodules are biopsied using a needle, in a procedure known as “fine needle aspiration biopsy“ (FNA). Some patients have a surgical biopsy, where the nodule, or the thyroid gland itself, is removed surgically.
Milas pointed out that historically there has been some difficulty in diagnosing thyroid cancer because it was heterogeneous.
“Thyroid cancers are heterogeneous and a single reliable picture that would classify all of them molecularly as cancers would be just as elusive as their picture under a microscope,“ she said.
This very issue led investigators teaming up with Veracyte (San Francisco) to develop the Gene Afirma Expression Classification test.
To obtain data for the trial evaluating Afirma – patients with sonographically confirmed nodules greater than 1 cm requiring FNA were enrolled at 49 U.S. sites. FNA samples were collected under US-guidance for usual cytologic assessment, with one additional aspirate collected for study analysis. Specimens were classified according to the Bethesda System for Reporting Thyroid Cytopathology.
The molecular classifier was trained using microarray data generated from 381 samples collected prior to the present study, and 45 FNAs accrued in the present study. The trained classifier, which uses signals from nearly 200 genes, was then locked and used to analyze the independent sample set described below. The results were compared to two blinded expert histopathology reviews to determine sensitivity, specificity and negative predictive value (NPV) of the molecular classifier.
In the initial results researchers found that among 66 thyroid FNA samples, including 43 indeterminate samples, for which surgical pathology results were known, Veracyte's test had a NPV greater than 95% when compared to the gold standard of two blinded expert histopathology reviews.
“There observation was, instead of trying to find a footprint for what is a picture of cancer, they were more successful in finding a footprint of what was consistent on a molecular level for a benign nodule.“
She added “Veracyte proposed the physician that does 240 FNA per year might identify 24 patients, who might potentially avoid surgery.“
Milas also spoke of the TSHR MRNA marker test, which the Cleveland Clinic uses. In October of 2008 the Cleveland Clinic approved this test to fall under its registry of home developed tests that are certified by the CLIA status of its laboratory. She added that the company performs about 150 tests a month and it is reimbursed by the insurance companies.
“To clinicians who are busy the test results come back in a week,“ Milas noted.
She said that TSHR MRNA yielded great results and was highly effective – but it could not be thought of as the magic bullet that would consistently diagnose thyroid cancer.
“If you used it as an interventional and only predictor of Thyroid cancer its diagnostic numbers would be moderately good,“ she said. “[But] each test has limitations and I think it would be very optimistic of us to think one and only one test would be the key answer to the holy grail.“
Milas also mentioned using FNA to classify BRAF mutations. She said that researchers first published the first foray into using the BRAF mutations to detect cancer back in 2009.
Commenting on results of their findings, she said that “if a mutation was present, [their method] had a 100% predictive value and in fact at surgery those nodules proved to be malignant, but the limitations were that not one single gene would carry the answer so you had to test for a panel of genes. Obviously there were some cancers that might have mutations that affect none of these genes.“
She added, “Encouraged by this approach even the American Thyroid Association [Falls Church, Virginia] incorporated consideration of that approach in its 2009 guidelines and since then a number of investigators have embarked on trying to validate a similar approach at their own institutions.“
Milas said that even though there were new tests coming out she saw that the future of detection for thyroid cancer was a combination of tests, which could be used eliminate any limitations of one single assay.
“There are at least three categories of tests that are readily available at the present time,“ she said. “My prediction is that excellent success will come in combining these various modalities because none may be perfectly able to predict the diagnosis itself, but all together these test will give us the opportunity to much better treat our patients.“
Benefits outweigh risks for DTC genomic testing
Is opening the door for direct-to-consumer (DTC) genomic testing actually opening pandora's box? A small panel took on that issue at the meeting.
Genomic testing represents a radical new approach to patient diagnosis and treatment. In the past few years there has been explosive growth in applications and firms that have gravitated toward this approach. As a major growth engine for clinical laboratories today, genomic testing along with molecular diagnostics is generating more than 30 million tests annually in the U.S., with dozens of new procedures introduced every year.
“The general public is very interested in the promise of personal genomics,“ Jerry Yeo, PhD, told the audience. “I think this is due to the convergence of two factors, the breakthrough in high throughput frequency as well as social networking.“
Yeo, a professor from the University of Chicago, moderated the panel. He pointed out a recent survey from Neuman ET al that said 70% to 88% of consumers are willing to pay for the genomic test even if it is not of any immediate clinical value.
“Now I'm not saying that there is a big rush of people going to get all of the genomic tests,“ he said. “I think the pickup is coming and that genomic testing hasn't caught a big wave yet, but clearly the companies are out there.“
“However on the academic side there is general consensus of calling for increased oversight and more regulatory control,“ Yeo told the audience.
He then introduced Jeff Gulcher, chief scientific officer of deCode Genetics (Reykjavik, Iceland) as the second speaker.
deCode is a firm that analyzes the human genome. By using its unique expertise and population resources, deCODE has discovered key genetic risk factors for dozens of common diseases ranging from cardiovascular disease to cancer. deCODE employs its capabilities to develop DNA-based tests and personal genome scans to better understand individual risk and empower prevention. It also licenses its tests, intellectual property and analytical tools to partners, and provides comprehensive value-adding genotyping, sequencing and data analysis services to companies and research institutions around the globe.
Gulcher said the company ran a clinical trial in the UK and asked if information derived from its genetic test actually adds to the positive predictive value of the PSA.
“Can genomic testing diminish the number of unnecessary biopsies,“ he asked the audience. “Because it turns out the positive biopsies in the U.S. are about 20%. That's it. About 8% are negative biopsies usually prompted by a rectal exam or more commonly a high PSA level. Or can we actually diminish the number of men who have a prostate cancer biospy by ignoring those men who are a much lower baseline risk, despite having a borderline level PSA? Conversely men of higher genetic risk, maybe those are the ones we do screen between the ages of 40 and 50 or maybe earlier. Maybe those are the men who we'll watch and wait and look to see if there is a dramatic increase in the PSA level and that could quite potentially diminish the number of prostate biopsies.“
Gulcher then pointed out the potential risks and benefits of these DTC applications.
He spoke about an experience in his own life recounting a time he took one of the companies tests to determine his risk for prostate cancer. The test pointed out that due to his background there was risk for developing prostate cancer.
“Had I not received this genetic test I wouldn't have gotten screened until my 50th birthday,“ he told the audience.
The other example was hypothetical but a typical argument opponents of DTC genomic test use.
“What if a patient finds their mother has a lower risk of breast cancer,“ he said. “The danger is she's not going to get her mammograms. Lower risk does not mean zero risks right – and we say that again and again through all of our reports.“
“This is a major challenge for the physicians already. She may decide not to get a mammogram because she's lower than average risk for breast cancer. This is a common problem and we can solve it by making sure we . . . educate patients and emphasize the risk,“ he said.
Gulcher also addressed the notion that taking these tests could cause some high anxiety in patients, if the person finds out they are at high risk for a disease.
“There doesn't seem to be any dramatic increase at least in the utilization of healthcare resources,“ he told the audience pointing out data collected from past studies. “Patients seem to be comfortable with the concept of risks . . .“
Gulcher concluded, “The benefits stem from the medical community encouraging individuals to better take of themselves. These test help define the individual's risk factors, and someone who has a bigger risk is going to be much more active in trying to mitigate that risk. That is exactly what we want to happen, to make our healthcare system more efficient.“
Look for unusual suspects in glucose monitoring errors
Back in 2010, the FDA held two meetings to discuss the accuracy and safety of glucose meters. During the meetings the agency pushed for tightening system accuracy criteria specified by current glucose meter standards and considered whether FDA review criteria for these devices should be changed to promote the public health. The FDA solicited feedback from clinical experts about the clinical requirements for blood glucose meter accuracy and precision, and the benefits and risks of using glucose meters to achieve and maintain tight glycemic control.
The subject that the FDA raised in 2010 was the focus of a session at the AACC meeting.
Martha Lyon, PhD of Calgary Laboratory (Calgary, Alberta) spoke during the session, which had limited availability, about eliminating patient related glucose meter errors.
The U.S. is the largest market for glucose meters and strips according to the 2010 market research report “Blood Glucose Meters and Strips: A Global Strategic Business Report“ developed by Global Industry Analysts.
The rapidly growing market has consumers, clinicians and researchers faced with the question, what's the best way to ensure an accurate reading, when a wrong one could cause serious harm?
Lyon discussed possible answers to the question and displayed a slide that was used at the March 2010 FDA meetings – regarding factors that could affect glucose meter performance.
“I like this slide because at the center it shows lots of factors that can influence performance of glucose meters,“ she told the audience. “I know we can talk about user error – expired strips, hand washing all that kind of stuff. We can talk about environmental factors. But what about the other factors that could influence glucose meters?“
She said some of those factors, which often aren't discussed are patient related issues.
“What's the patient's disease,“ she asked. Are they a burn patient, or a renal patient? So patient-specific factors can influence glucose meters as well as different treatments and different drugs.“
Lyon went on to suggest there needs to be a greater emphasis put on patient differences when evaluating the meters.
She said these errors can happen because of changes in the patient and that could guarantee that laboratorians could run quality control tests until they were “blue in the face“ and an error still wouldn't be detected.
She then discussed ways to confirm if a glucose meter is functioning properly.
“We've got a couple of tools available to us,“ she said. “One is to be involved with an external proficiency test. When we do those tests we have a sample that is given to us and we're trying to simulate a patient testing situation. What's interesting to note is that the material ... is not whole blood. And, if we're looking at a glucose meter we're going to be testing whole blood. So that's an important thing we want to look at and keep in mind.“
She pointed out that the other method of testing – which is analyzing quality control materials put device makers and those who use glucose meters in a similar position.
“The other thing we can do to figure out if our meters are working is analyze quality control materials,“ Lyon said. “The take home message is ideally, you would like the [sample] to be the same matrix you're running with the patient sample. Again we're hooped because with our quality control materials, they're not whole blood and what we're running is an aqueous solution of glucose.“
This point then brought Lyon to what she said was her key editorial comment on the subject.
“These things are not detecting patient related factors that could influence the performance of a glucose meter,“ she told the audience. “So just because we enroll in CAP [an external proficiency test for the meters] and just because we so diligently do our quality control, it doesn't mean we are devoid and we won't have errors that we'll still have to figure out. It's our responsibility as laboratorians to investigate the effects of patient related errors on the performance of glucose meters.“