comscoreStudy Looks at Accuracy of Breast Biopsy Results

Study Looks at Accuracy of Breast Biopsy Results

A study looked at the accuracy of pathologists' interpretations of breast biopsies. The results have generated some controversy.
Mar 18, 2015.This article is archived
We archive older articles so you can still read about past studies that led to today's standard of care.
A biopsy is a small operation done to remove tissue from an area of concern in the body. The tissue sample is examined by a pathologist (a doctor who specializes in diagnosing disease) to see if cancer cells are present. If cancer is present, the pathologist then looks at the cancer’s characteristics.
Biopsy is usually a simple procedure. In the United States, only about 20% of the women who have breast biopsies turn out to have cancer.
A study looked to see how accurate pathologists’ interpretations of breast biopsy samples were.
The research was published in the March 17, 2015 issue of the Journal of the American Medical Association. Read the abstract of “Diagnostic Concordance Among Pathologists Interpreting Breast Biopsy Specimens.”
To do the study, the researchers divided 240 breast biopsy slides (one slide per case) into four sets of 60. Each set of 60 slides was sent to each of the 115 pathologists in the study. The pathologists came from eight states and regularly interpret breast biopsies in their practices. The pathologists looked at the slides, made a diagnosis, and returned them to the researchers. The set was then sent to the next pathologist. The study took 7 years to do.
The slides included:
  • 23 cases of invasive breast cancer
  • 73 cases of DCIS (ductal carcinoma in situ)
  • 72 cases of atypical hyperplasia (abnormal-looking cells are growing faster than normal); atypical hyperplasia is benign breast disease (not cancer) but it does increase the risk of being diagnosed with breast cancer in the future
  • 72 benign (not cancer) cases
Each pathologist evaluated the slides without any contact with the other pathologists in the study.
The researchers then compared the pathologists’ diagnoses with each other and with the diagnoses of a consensus panel. The consensus panel was made up of three internationally recognized breast biopsy experts.
Overall, the diagnoses of the pathologists in the study agreed with the expert panel’s diagnoses 75.3% of the time. While this seems concerning, it’s important to know that the diagnoses of the three members of the consensus panel agreed with each other only 75% of the time until they met and discussed the cases.
The researchers also looked to see how closely the diagnoses of the pathologists in the study and the diagnoses of the consensus panel matched for particular cases:
  • 96% of the invasive breast cancer diagnoses matched
  • 84% of the DCIS diagnoses matched
  • 87% of the benign diagnoses matched
  • 48% of the atypical hyperplasia diagnoses matched
On the one hand, it’s reassuring to know that the pathologists, even working in isolation, were very good at diagnosing invasive cancer. On the other hand, some of the numbers seem alarming, especially the 48% match of atypical hyperplasia diagnoses.
It’s important to know that the design of this study didn’t allow the pathologists to interpret the slides and make diagnoses as they do in the real world.
In the study, the pathologists could only look at one slide from each case. They also couldn’t talk to other doctors about the cases if they were unsure or had questions.
In the real world, pathologists can look at a number of slides from each biopsy. If one slide isn’t enough to make a definitive diagnosis, the pathologist can look at others.
In the real world, pathologists also regularly consult with each other on diagnoses, especially if there are questions or uncertainty. It’s worth noting that the experts on the consensus panel only agreed with each other’s diagnoses 75% of the time until they talked to each other about the cases.
In the real world, a pathologist who is uncertain about the results of a biopsy can tell a woman’s doctor that the biopsy results are unclear and that a second opinion is needed. The biopsy slides can then be sent to another pathologist to read.
If you’re having a biopsy because a mammogram or other test found a suspicious area in your breast, you’ll receive the results in a pathology report. If you have questions about anything in your pathology report, ask your doctor or nurse navigator to explain these areas to you. If your pathology report diagnoses you with DCIS or benign breast disease or is in unclear in any way, you may want to ask for a second opinion. If you’re thinking about a second pathology opinion, call your insurance company to make sure this service is covered. Sometimes, insurance companies only pay for a physician to give a second opinion about your original pathology results.
When you and your doctor have the most accurate information, you can make the best treatment decisions for you and your unique situation.
For more information on when and how to get a second opinion, visit the Getting a Second Opinion pages.

— Last updated on February 22, 2022, 10:03 PM

Share your feedback
Help us learn how we can improve our research news coverage.