Mammography screening for detection of breast cancer produces a lot of false positive findings. This means that many women are subjected to unnecessary breast biopsy procedures. This news release reports on a study in which the authors argue that false-positives prior to biopsy may be reduced by using a procedure called diffusion kurtosis imaging. This technique relies on information provided by MRIs performed as a supplement to mammograms. Diffusion kurtosis imaging provides information on tissue structures at a microscopic level that can be read via a software algorithm rather than by human technicians. Given that malignant lesions affect tissue structures at this level, it is suggested that this technique might serve as a relevant marker of changes.
False positives are frightening for patients and lead to increased diagnostic costs. Reducing the number of them seems like a laudable goal, and a software algorithm to assess imaging would be less invasive than biopsy procedures. The news release is careful to state that the new approach is not being presented as a substitute for current MRI protocols that use contrast-imaging to detect malignancies. The lead author emphasizes that diffusion kurtosis imaging is simply another diagnostic option, and that findings must be subjected to larger scale studies. All of this is straightforward. It does, however, beg the question of why these findings are worth reporting in a news release.
At HealthNewsReview.org we regularly see releases and articles describing results of early stage trials and pilot studies which may not hold up in larger trials and which are years away from widespread implementation. Information in such studies is certainly worth disseminating to the scientific community as knowledge about a particular issue is painstakingly being constructed. However, to flood the broader public with preliminary results runs the risk of raising false expectations or inuring them to significant developments when they arise.
Although it seems self-evident that reducing the number of unnecessary biopsies could save on medical costs, the release doesn’t elaborate on the role of MRIs as an adjunct to mammography. In what proportion of cases are MRIs ordered prior to tissue biopsy? Does diffusion kurtosis imaging add to the cost? Cost of this procedure simply isn’t addressed.
Readers learn that after the technique was developed, the radiomics analysis reduced false-positive findings by 70%, while detecting 60 of 61 malignant lesions in a test set of 127 women, or 98%. In order to give full meaning to these statistics, we need to know what the number of false positives was in the comparison group, rather than being given only the percentage of change.
Similarly, a brief statement is made about the advantage of the consistency provided by a software algorithm in comparison to readings of images by human technicians. That statement, too, is not quantified. How much error does result from the variability introduced by human reading of images? To what extent is that a problem?
Also, the publication stated that the rate of malignant lesions was higher than expected in this group so they would need to replicate this technique to make sure they didn’t have bias within their sample.
No specific description is provided of potential harms. One potential harm is that this technique still misses some malignant lesions.
This news release is based on a retrospective study in which a software algorithm was created based on 222 women at two independent study sites who had suspicious mammography findings. The algorithm was then pilot-tested on another 127 cases retrospectively, and found to be effective. Given the very preliminary nature of these findings, the news release is appropriately careful not to over claim benefits. The release does assert that the new technique might cut down on the number of false positive findings, and therefore on unnecessary biopsies. However, it also states that definitive information on that awaits large-scale clinical trials.
No disease mongering is evident in the news release. Breast cancer is a serious disease, and false positives create a very real level of anxiety in patients.
No statement is made regarding the funding source for the study.
The lead researcher is quoted explaining that the new approach is not being presented as a substitute for current MRI protocols that use contrast imaging to detect malignancies. Diffusion kurtosis imaging simply adds to the spectrum of diagnostic tools. A statement detailing why the software algorithm might be superior to use of gadolinium-based contrast agents that need to be injected intravenously would have made the comparison clearer.
This report is based on a retrospective study early in the process of testing the new imaging procedure. It is too early to comment on availability. Also, it appears this study was conducted in Europe so it’s unclear if the algorithm would become available to use in US clinics.
Diffusion kurtosis imaging is about a decade old technique for examining diffusion of water in human tissue. However, the application to identifying breast cancer malignancies appears to be novel.
The lead author notes: “The new approach is not intended to replace current contrast-enhanced breast MRI protocols in general, Dr. Bickelhaupt emphasized, but to expand the spectrum of options available for answering specific clinical questions.”
No unjustifiable language was observed in this news release.