Despite press releases that indicate otherwise, there is not yet a reliable blood test that improves our ability to detect early breast cancer. Even if there were a useable test, today’s standard for evidence would eventually mandate proof of a mortality reduction above and beyond that provided by mammography. That said, the introduction of such a test into clinical practice might not be that far away. Currently in my 28th year of pursuing this line of research, I’ve amassed a long list of failed efforts after sending over 10,000 specimens (tied to a database that includes MRI results) to multiple, international collaborators in academics and industry, along with several failed prospective clinical trials. Here’s my Top Ten List of what I’ve learned so far:
Biomarkers for early breast cancer are likely to be different than those in later stages
Unblinded test sets are subject to tweaking and don’t hold up well in blinded studies
Benign proliferative disease and atypia are the great spoilers in blood test performance
Biomarkers could well be unique for each biologic type of breast cancer
Menopausal status can be a difficult barrier to overcome, perhaps prompting 2 tests
Cell free DNA (“liquid biopsy”) has its strength in specificity, but sensitivity is a problem
Protein biomarkers are stronger in sensitivity, but specificity is the problem
A single biomarker is unlikely, so combining multiple results is a challenge
Performance characteristics should be judged with and without DCIS.
Sensitivity, specificity, accuracy, NPV and PPV do not stand alone with a blood test unless it is used in women who do not get screening mammography (under age 40, refusal of mammography, countries without mammography infrastructure). Thus, these performance characteristics must be merged into imaging performance characteristics as well as combinations of imaging modalities. This merging of performance is a challenge with screening mammograms where density levels dictate a wide range of performance. Thus, if blood test performance is independent of breast density, then the contribution of a blood test increases as density increases.
One of the main issues in this discussion is: Which is more important for a screening blood test – sensitivity or specificity – as we can rarely have both. To cast light on that discussion,here is an article that elaborates.