Low Histopathologic Agreement for DCIS, Atypia Breast Cancer

Pathologists disagree with one another’s interpretations about 8 percent of the time when diagnosing a single breast biopsy slide, according to a study published March 22 in the Annals of Internal Medicine. The present study applied previous discrepancy findings from the Breast Pathology Study (B-Path) to patient populations. Discordance between individual pathologists and a reference consensus diagnosis was more likely in cases of ductal carcinoma in situ (DCIS) or atypia, with higher levels of overinterpretation of disease risk. The authors say that this “diagnostic grey zone” needs to be considered in clinical management decisions. Histopathological diagnosis remains the gold standard of breast cancer diagnosis, despite concerns about the variability in specimen interpretations in clinical practice. B-Path previously found that one in four breast biopsy results were discordant with expert reference consensus diagnosis. But, experts note the study included higher proportions of cases of DCIS and atypia than typically seen in clinical practice. However, the study did not assess population impact.

The present study estimated the effect of interpretation variation from the perspective of U.S. woman having a biopsy. The researchers calculated predictive values using Bayes’ theorem, combining results from B-Path with published data of the population-based prevalence of breast pathology diagnoses in women aged 50 to 59 years. B-Path compared 115 pathologists’ interpretations of a single biopsy slide (6,900 total interpretations from 240 distinct cases) versus a reference consensus interpretation from three experts.

The researchers found that the reference panel members’ review showed concordance with the final consensus diagnoses 90 percent of the time versus participants’ 75 percent concordance. Overall, using one representative slide per case, 92.3 percent of breast biopsy diagnoses would be verified by reference consensus diagnoses, with 4.6 percent of discordant results overinterpreted and 3.2 percent underinterpreted.

Most U.S. women undergoing breast biopsy in clinical practice receive a benign without atypia diagnosis. For these women, diagnostic agreement with the reference panel would be high (97.1 percent). Verification of invasive breast cancer was also highly probable (97.7 percent). The likelihood that a diagnosis of atypia or DCIS would be verified by the reference consensus diagnosis was low. Diagnostic agreement with the reference consensus panel for atypia was less than 50 percent regardless of the pathologists’ desire for a second opinion or whether they noted that the case was borderline. Using a single slide for the cases interpreted as DCIS, the reference consensus panel would interpret 9.5 percent as benign without atypia, 9.0 percent as atypia, and 11.8 percent as invasive breast cancer.

“Women with borderline breast lesions that are difficult to categorize, such as atypical ductal hyperplasia and low-grade DCIS, may benefit from revised guidelines for clinical treatment and management given the degree of diagnostic variability and biological overlap between these diagnostic categories,” write the authors led by Joann Elmore, M.D., from University of Washington in Seattle.