DNA Testing for Breast Cancer Risk: What the Current Medical Evidence Tells Us

The advent of predictive genetic screening for mutations in theBRCA1,BRCA2, andPALB2genes was a tremendous breakthrough that’s transformed cancer prevention and early detection. That’s the good news. The bad news: While predictive genetic testing is useful and even crucial for many populations, new medical advances are routinely overhyped in the media and misunderstood by consumers and providers alike.

Indeed, physicians who graduated from medical school before 2002 weren’t even educated about the genetic testing tools in wide use today, as DNA tests weren’t part of the standard of care. As a result, health and medical reporters often end up relying on less-than-fully informed medical sources and information from test developers when reporting.

The upshot is that advances in DNA testing—such as panels—are touted as the latest game-changing breakthrough, and often marketed directly to both consumers and physicians. While it’s no wonder that consumers want the latest tests, providers get caught in the middle, as they’re often unclear about when a given test is warranted.

Stress-Testing the DNA Test

Here’s a case in point: Recently, theJournal of Clinical Oncologypublished a study¹evaluating whether patients with breast cancer in the United States are being appropriately recommended for a full panel of DNA testing for hereditary breast cancer.

The researchers developed a “multi-part question,” reportedly based on the National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guidelines),²to sort patients with breast cancer into those who qualified for testing and those who didn’t. Then, germline genetic testing with a multi-cancer panel of 80 genes was performed on all the patients who hadn’t previously undergone testing—thus heavily weighting the group with those who hadn’t previously met the criteria for genetic testing.

Of the 959 individuals tested, 83 harbored a pathogenic or likely pathogenic variant. The researchers posit that in the real world, based on their response to the multi-part question, 38 of these 83 women would never have been tested—leading to potential missed diagnoses. The authors conclude by recommending that all previously or currently diagnosed breast cancer patients undergo expanded panel testing.

The issue with this line of reasoning is that the clinical utility of many of these found variants for predicting recurrent cancer or breast cancer risk in family members is actually unclear.

When the study was picked up by mainstream media, reporters put their own spin on the findings. The study’s conclusions were summarized with inaccurate, provocative statements such as (I’m paraphrasing) “current genetic tests miss as many patients with hereditary cancers as they find.” Articles like these did not address the fact that the study excluded patients with breast cancer who had already been tested, as they had met qualifying testing guidelines. At least one publication went further, reporting that breast cancer guidelines were “out-of-date and unusable,” given the recent proliferation of multi-gene panel tests.

In truth, the clinical guidelines for breast cancer treatment are kept current. The NCCN updates its genetic testing guidelines annually (and commonly more frequently, as needed). The Version 2.2017BRCAtesting criteria used in the study listed 21 distinct and concrete criteria, all of which are understandable and applicable by clinicians. NCCN also provides criteria for other breast cancer-related genetic syndromes. A multi-gene panel table lists specific genes associated with hereditary breast cancer, coupled with breast cancer risk management recommendations for these genes.

Another article reported that genetic testing is now so inexpensive, hovering around $250 for 80 genes, that all women should be tested. While it’s true the cost of multi-gene testing panels has declined, most tests for hereditary breast cancer retail for between $1000 and $3000. Broader panels can cost more. The 2018 Centers for Medicare & Medicaid Services (CMS) rates for completeBRCA1/2testing (CPT codes 81162 and 81211) were listed at $2253 to $2396 respectively, depending on which deletion/duplication variants are included.

Conflating points of view with research-based evidence is not uncommon, but it does have consequences. Asserting that every woman should be tested simply because we have the technology is an irresponsible idea. Genetic tests are not unequivocal, and there are well-known unintended consequences to their widespread, unmanaged use.

We must be discerning about who is tested, why, and when, remaining cognizant of the test’s purpose. This issue is not simply about preventing the runaway costs that would be incurred by genetically testing every woman in the United States. That’s secondary. The primary concern is how indiscriminate testing would impact outcomes. If predictive test results are misinterpreted or lead to unnecessary preventative actions, what care value is this genetic information actually providing?

Clinical Significance and Selection Bias

Let’s take another look at that study. It classified patients as meeting or not meeting NCCN guidelines based on a clinician’s yes or no response to a multi-part question developed by a genetic counselor. As NCCN did not actually have concrete criteria for multi-gene panels in their 2.2017 guideline, the study’s subjects were classified as having met or not met NCCN guidelines based on subjective clinical opinion—possibly coupled with the criteria forBRCA1/2testing.

Diving into the data reveals that fewer than 1% of subjects who did not meet NCCN criteria forBRCA1/2testing had a pathogenic or likely-pathogenic variant. In other words, 1000BRCAtests would be required to identify just 6.3 additional women who may be at elevated risk for breast cancer. Shouldn’t those 6 women have been tested, regardless of the cost? Isn’t the information about their elevated risk clinically significant?

The answer is, “not entirely.” Cancer penetrance estimates—or, in layman’s terms, how much having a mutation increases a patient’s risk for cancer—are generally based on women who were tested because they already have additional risk factors, such as a family history of cancer. It’s not clear to what extent aBRCAmutation alone might increase the risk in women who don’t have these additional risk factors.

Indeed, the study acknowledged its selection bias, stating women who met the NCCN testing criteria were only included in the study if they had not yet been tested. It is likely that the study subjects who met the NCCN guidelines were at a lower-than-average risk of having a pathogenicBRCA1/2variant, or they would have been tested already. After all, these were the women who had not been referred for testing by their initial diagnosing provider.

This hypothesis is supported by the fact that only some 2.5% of these women had mutations. In referred populations, the average rate of mutations is 2 or 3 times higher. For this reason, the study’s comparison between the guideline-met and guideline-not-met groups cannot be taken at face value. Moreover, 21% of the NCCN-guidelines-met group did not have a family history of cancer.

In addition to various family histories of cancer, NCCN criteria allows testing based on a personal history of cancer (eg, other types of cancer, young age at onset, or a mutation found during tumor testing); types of breast cancer (eg, triple-negative or multiple primaries); sex (male); and ethnicity (Ashkenazi Jewish). What do the NCCN guidelines recommend for this population? Are decision support tools useful here?

Personalized Care for Population Health

This is where Change Healthcare’sInterQual®has an important role to play. It was one of the first decision-support tools to provide criteria for genetic testing, and it provides guidance based on clinical utility for each individual patient for over 800 tests. The InterQual clinical development team, of which I am a member, synthesizes the best and most current evidence into a fully referenced decision-support tool that helps mitigate consumer pressure to over-test, overprescribe, and overtreat. InterQual also helps identify underutilization for patients who should be treated.

The volume of medical research published every day can be overwhelming for clinicians to process, and it’s nearly impossible for one individual to properly validate medical advances in real time. InterQual’s clinical developers serve as curators of the research, identifying and critically appraising the most relevant studies; synthesizing results across studies; and weighing the comparative benefits and risks of various interventions. InterQual helps physicians balance patient safety and risk with medical necessity.

The tool’s objectivity is crucial, as is the fact that its clinical guidelines are not only current but comprehensive. For example, in addition to the NCCN criteria, InterQual criteria also accepts the results of the well-validated BRCAPRO and BOADICEA prediction models. If either of these has been applied and indicates at least a 10% chance of aBRCA1/2mutation, testing is recommended.

If more than one hereditary syndrome is suspected, InterQual criteria recommend multi-gene panels targeting the relevant actionable mutations. The criteria also provide Limited Evidence recommendations for broader, more comprehensive multi-gene panels, which might well be more cost-effective than targeted panel testing. Limited Evidence recommendations are based on one or more of the following:

• Research-to-date has not demonstrated this intervention’s equivalence or superiority to the current standard of care
• The balance of benefits and harms does not clearly favor this intervention
• The clinical utility of this intervention has not been clearly established
• The evidence is mixed, unclear, or of low quality
• This intervention is not standard of care
• New technology is still being investigated

Panels and other advances in DNA testing are certainly revolutionizing diagnostics and treatment. But they are not changing the rules of clinical engagement. The mission remains the same: to treat each patient as an individual who is entitled to receive the best possible care. For each patient, that care might be different, as what defines appropriate care is based on what the clinical evidence indicates will lead to the best outcome.

That’s not as simple or easy as recommending genetic panel testing for every woman in the United States. Genetic tests aren’t like vaccinations—they’re not useful for everyone. Clinical guidelines remain the foundation for how we evaluate and provide care. That’s the only way providers can help ensure the most appropriate care to support better outcomes for each individual patient.

Teresa Benson, MA, LP, is the chair of the InterQual Evidence-Based Medicine Committee, and senior clinical information specialist for Change Healthcare.

References:

1. Beitsch PD, Whitworth PW, Hughes K, et al. Underdiagnosis of hereditary breast cancer: are genetic testing guidelines a tool or an obstacle?J Clin Oncol.2018;37(6):453-460. doi: 10.1200/JCO.18.01631.
2. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines). Genetic/Familial High-Risk Assessment: Breast and Ovarian. Version 2.2017. Published December 7, 2016. nccn.org/professionals/physician_gls/pdf/breast.pdf. Accessed March 1, 2017.
3. Are the Guidelines for Breast Cancer Genetic Testing Out of Date? Healthline. Published December 19, 2018. https://www.healthline.com/health-news/are-breast-cancer-genetic-tests-out-of-date. Accessed April 16, 2019.