Do Pathologists “See” Breast Cancer In Benign Biopsy Samples?

new study published in the Annals of Internal Medicine suggests that while pathologists usually agree in diagnosing cases of invasive breast cancer, they’re more likely to contradict one another when looking at biopsy slides that show precancerous lesions or non-invasive tumors. The study, titled “Variability in Pathologists’ Interpretations of Individual Breast Biopsy Slides: A Population Perspective” was released online on March 22, 2016.

Pathologists “Overcall” Breast Cancer Risks, Study Finds

Led by Joanne Elmore, MD, a team of academics from the University of Washington, Stanford and Dartmouth took 240 individual biopsy slides and gave them to a group of three experts in pathology. These doctors served as a “reference group,” interpreting the slides together to accurately diagnose whether each patient had:

  • invasive breast cancer,
  • ductal carcinoma in situ (DCIS),
  • atypical hyperplasia or
  • benign papilloma without atypical hyperplasia.

After recording those accurate diagnoses, Elmore’s group sent the biopsy slides along to 115 different practicing pathologists for diagnosis. Overall, the practicing pathologists “verified” the experts’ diagnosis in a little more than 92% of cases. In other words, they got it right in a majority of patients. But the number of diagnoses the practicing doctors got wrong varied widely depending on the condition displayed in the slide.

Atypical Hyperplasia “Overintrepeted” In More Than 50% Of Cases

Once cancer cells had broken away from their site of origin, and begun spreading into surrounding breast tissue, the pathologists almost always returned the correct diagnosis, agreeing with the experts on almost 98% of the invasive breast cancer cases. Their judgement was less reliable, however, when it came to cancerous tumors that hadn’t migrated yet, in the case of ductal carcinoma in situ, or precancerous cell accumulations like atypical hyperplasia.

When presented with cases of “atypia,” a condition in which cells within the mammary ducts have begun to proliferate abnormally, the pathologists “overinterpreted” the slide 53.6% of the time, calling it a cancerous tumor. 8.6% of the patients, however, would have been told that their precancerous condition was in fact benign, based on the pathologists’ “underinterpretation” of their biopsy. Only a little more than 1 out of every 3 slides deemed “atypia” by the expert panel were verified by individual practicing pathologists.

As for ductal carcinoma in situ, the most common type of non-invasive breast cancer, according to the American Cancer Society, 9.5% of those slides were diagnosed as benign, while 9% and 11.8% were diagnosed as atypical hyerplasia and invasive breast cancer respectively.

When The Gold Standard Isn’t Perfect

In an editorial accompanying the study, Doctors Alexander Borowsky and Laura Esserman note that, while histopathology, the microscopic inspection of cellular structures, is often considered the “gold standard” of diagnosis, Elmore’s study has revealed several major “problems,” most notably “a tendency toward ‘overcalling’ the risk level of disease.”

Although that may be true, Borowsky and Esserman urge caution in interpreting the results of Elmore’s research (a caution, one would hope, shared by pathologists interpreting biopsies). For one, it’s altogether unclear whether or not the reference diagnoses, those determined through consensus by the three cancer experts, are actually accurate. Experts, they write, “can learn to agree more often,” but that’s not the same as getting a diagnosis right.

Elmore herself has voiced some reservations. The pathologists involved, she told Medscape on March 30, 2016, were most consistent in their diagnoses on cases of invasive breast cancer and benign tissue, which are “by far the most common” conditions seen by pathologists in a real-world setting. The danger of a misdiagnosis, then, to real patients is probably not as high as her numbers would suggest. Even so, she and her colleagues ended their paper by writing that “efforts to reduce diagnostic variability need to be considered,” even floating the idea of mandatory second opinion policies.

The researchers reported good news in relation to slides that depicted benign, or non-cancerous, growths. In 97.1% of cases, the practicing pathologists correctly diagnosed the slides. Of course, that means an estimated 2.9% of the patients without breast cancer, or any precancerous cell growth, would have been told otherwise.

How Can We Help?