Commentary: New Cancer Diagnostics Are Being Misdiagnosed in Their Value

by Eric A. Klein and Tomas J. Philipson


Various methods of cancer screening, such as mammograms and colonoscopies, have been highly beneficial to patients across the world by allowing more effective treatments to be applied earlier. The most recent innovations for diagnosing cancers include genomic blood tests that can detect more cancers at earlier stages than existing screens. If reimbursed and adopted widely, they offer a great potential advance in the war on cancer, but unfounded critiques of them misdiagnose their value.

Currently, about 600,000 Americans die every year from cancer which makes up the second leading cause of death. About 70 percent of these deaths are for cancers outside of the 5 big single cancers we typically screen for in the colon, breast, prostate, cervix, or lungs. The cancers we typically don’t screen for are often caught too late when symptoms appear and thereby more deadly.

In contrast, the new genomic blood tests for cancer cast a much wider net – one such test can detect a cancer signal across more than 50 kinds of cancer including many that currently have no available screens or for which the USPSTF recommends against screening. Beyond detecting signals for cancer types that current screens miss, such as ovarian and pancreatic cancers, the new genomic tests have the potential to detect signals for cancers at earlier stages and with greater efficiency.

These new tests offer previously unimaginable breakthroughs in the war on cancer. For example, almost 90 percent of pancreatic cancers are detected when they have already spread, and treatment applied too late offers little help. But imagine a world where most of them are found earlier as evidence suggests that 78% of cancers currently detected at a late stage could be detected and treated earlier with the new tests.

Despite the potential benefits of molecular cancer tests, some argue against them due to the risks that false positive tests may induce unnecessary care and anxiety.

But false positives are not unique to the new genomic blood tests. Their estimated error rate of about 0.5% is well below the 7 percent rate for existing single cancer screens in aggregate. Furthermore, one large study showed that the cumulative false positive rate for repeated screening with currently recommended tests can be up to 60% for men over 3 years.

Unlike existing tests, which only seek to maximize the chance of detecting a cancer when indeed present, the new genomic tests are calibrated to keep the false positive rate under 1%. While adding new genomic tests as an adjunct to current screening tests will increase the absolute number of false positive tests, for the whole population they will actually make screening more efficient by reducing the number of people needed to be tested to find one cancer.

In addition, any negative consequences of false positives must be seen relative to the value the diagnosis of true positives provides. One study estimated that use of the new tests could reduce all cancer mortality in the US by 26% which seems a fair trade-off for any additional false positives.

Any screening effort in cancer, whether with established or novel technologies, is based on the premise that detecting cancer in earlier stages improves treatment outcomes. If not, simply discovering a deadly cancer without being able to treat it better would induce much unneeded anxiety during its growth. This principle is important to assess the value of the new genomic blood tests because much anxiety  associated with cancer screening in the past, although often short-lived, is based on cancer diagnoses historically translating to a death sentence. However, secondary to advances in cancer treatment, earlier cancer detection using these new tests enables earlier and potentially more efficient and effective treatment. Put simply, the new tests enable testing into treatment rather than testing into a death sentence.

In general, the value of testing into treatment rather than death means that improved treatment often drives demand for screening. This was evidenced with the surge in HIV testing following its successful treatments in the mid-1990s. Therefore, anxiety associated with earlier cancer detection using genomic blood tests may well be less than the anxiety associated with traditional screening.

More importantly, most patients understand that screens may result in false positive or false negative outcomes and may understandably choose to use the screens regardless. Before taking the screens, they understand that a false positive may result but are willing to live with that issue given the large benefits from true positive or negative results. This is an important issue because the new cancer tests are so preferable (a simple blood draw vs. a colonoscopy, for example) that they will democratize the screening decisions to the patient level more. Third parties outside the doctor-patient relationship steering some patients against their individual tradeoffs then seems misguided.

Appropriate value assessment of any new technology, including the new cancer diagnostics, must evaluate the relative size of any costs and benefits that both always are present. Critiques of genomic blood tests do not follow the evaluation science this way but tend to lend a disproportionate focus on harms without regards to the more than offsetting benefits. It seems likely that correct value assessments will indicate that the benefits are larger than any potential risks or costs. Such evaluations are needed to better inform the adoption of these promising tests.

– – –

Klein is Emeritus Professor and Chair of the Glickman Urological and Kidney Institute. Philipson is an economist at the University of Chicago and served in 2017-2020 as a member and acting chair of the President’s Council of Economic Advisers. Both are consultants for GRAIL Inc.
Photo “Mammogram” by Defense Visual Information Distribution Service.




Content created by RealClearHealth is available without charge to any eligible news publisher. For republishing terms, please contact [email protected].

Related posts