A breast biopsy which illustrates the grey zone of pre-cancer (Courtesy of Dr. Michael J. Misialek)
By Michael J. Misialek, MD
If you’re a woman who has ever had a breast biopsy, you may be asking yourself a few serious questions:
“How do I know if my breast biopsy is completely accurate?” And, “Who is the pathologist reading the biopsy, and what is their level of training?”
Many more patients are asking these and similar questions following widespread media coverage on a Journal of the American Medical Association (JAMA) study, which casts doubt about the accuracy of interpreting these biopsies.
Let’s break the study down and ease some anxiety. Perhaps most importantly, this provides a great opportunity to learn about one of the lesser know medical specialties, pathology…which is what I do.
The JAMA study, “Diagnostic Concordance Among Pathologists Interpreting Breast Biopsy Specimens,” revealed the following key finding:
• Overall agreement between individual pathologists’ interpretations and that of an expert consensus panel was 75 percent, with the highest agreement on invasive breast cancer and lower levels of agreement for ductal carcinoma in situ (DCIS) and atypical hyperplasia.
What this means is that the agreement between a general pathologist and an expert was excellent for breast cancer (those with the ability for metastasis), but varied significantly for early cancers and high-risk pre-cancers.
While the study’s findings may not be surprising to physicians who understand the challenges of diagnosing complex breast cases, news of the article could lead to unnecessarily heightened anxiety for patients and the public as breast cancer is a highly publicized and pervasive disease.
The study confirmed that the majority of breast pathology diagnoses, especially at either end of the spectrum (benign disease and invasive breast cancer) are accurately made by practicing pathologists regardless of practice setting. The overall rate of agreement for invasive breast cancer cases was 96 percent.
Issues with diagnostic disagreement mainly center on the borderline cases, between atypical hyperplasia, that is, pre-cancer, and DCIS, early cancer.
Why does this matter? Overdiagnosis can lead to unnecessary surgery, treatment and anxiety. Underdiagnois can lead to a delay in treatment. The bottom line is that experience matters.
Factors that contributed to greater disagreement included: a low case volume, small practice size, nonacademic practice and high breast density.
The study has many weaknesses. Chief among them was that only a single slide per case was given to each pathologist. As a practicing pathologist, this never happens. I will review multiple slides, often ordering several additional deeper sections and ancillary special stains, studying each carefully. This practice was prohibited in the study.
Additionally, the study cases were a mixture of core biopsy and excision specimens. A core biopsy is obtained using a needle, often by a radiologist, in which a small core of tissue is removed. An excision is a “lumpectomy” which is done in the operating room where a large section of breast tissue is removed. Diagnostic criteria vary between a needle core and excision. Often times it is not necessary to render an exact diagnosis on the core biopsy, but rather recognize an abnormality and recommend an excision for which additional tissue will clarify the diagnosis.
Even the experts disagreed in the study (75 percent initial agreement then 90 percent after discussion).
This illustrates the fact that pathology is both a science and art. Experts may stress slightly different criteria in their pathology training programs. The “eye of a pathologist” is a difficult measure to quantify and is dependent on multiple factors that best function in real time, not an artificial study.
Another weakness is that there is no evidence that the experts were more accurate in predicting outcomes than test subjects. Perhaps most importantly, a second opinion was not allowed in the study, even when study participants indicated uncertainty. These are in fact the very cases that would most likely have been shown around, sent out for consult and further worked up.
It is not realistic to introduce such a large caseload of breast biopsies that are heavily weighted towards atypical hyperplasia and DCIS. Since these borderline cases represent only a small fraction of breast biopsies in actual practice, diagnostic agreement in routine practice is higher than that reported in this study. No clinical information other than patient’s age was given to the study pathologists, and no imaging findings were included. In actual practice, integration of the clinical setting and imaging findings is routinely used in making a diagnosis.
The findings are not unique to pathology. All of medicine has grey zones, where controversy often exists. The study does have an important message for pathologists. As noted in the accompanying editorial, it should serve as a “call to action.” A better, more reproducible definition of atypical hyperplasia is needed.
The article highlights the need for an active quality management program in surgical pathology that includes targeted review of difficult or high risk cases. The College of American Pathologists (CAP) and the Association of Directors of Anatomic and Surgical Pathology have been developing an evidence-based guideline expected to be released in May to provide recommendations to reduce interpretive diagnostic errors in anatomic pathology.
The CAP is proactively addressing educational opportunities through advanced breast pathology training programs designed to provide a route for pathologists to demonstrate their expertise regardless of the setting in which they practice.
Patients can take steps to help ensure their breast biopsy is read accurately:
o Inquire about the pathology laboratory that will examine your tissue sample. Is the laboratory accredited? The CAP accredits more than 7,600 laboratories worldwide and provides an online directory for patients. Continue reading