A study recently published in the British Medical Journal has found that, when assessing breast abnormalities, getting a second pathologist’s opinion decreases the number of misdiagnosed cases.
When an abnormality, such as a lump, is found in the breast, doctors often recommend what is known as a biopsy. Such procedures involve the removal of a small tissue sample from the breast that can then be examined under a microscope by a specialized doctor called a pathologist. Ideally a pathologist can look at a biopsy and determine what type of abnormality is present and, importantly, whether it is benign or cancerous. However, breast abnormalities can be quite difficult to interpret which leads to cases of misclassification. When this happens some patients are unnecessarily given invasive treatments while others do not receive the treatment or monitoring they need. The authors of a study recently published in the British Medical Journal postulated that seeking a second opinion regarding the diagnosis of breast abnormalities would decrease the number of misdiagnosed cases. Because of this, they wanted to find out which strategy of seeking second opinions would yield the most benefit.
To carry out their study, 240 breast biopsies were obtained. Each of these were examined by three experienced pathologists who, after individually assessing the samples, worked together to reach a consensus on the diagnosis. This diagnosis could be one of four categories; benign, atypia, ductal carcinoma in situ (DCIS), or invasive cancer. Atypia is a benign breast disease that is known to have the potential to progress to cancer while DCIS is thought to be a very early stage of breast cancer. This was taken to be the reference diagnosis against which additional diagnoses were compared. Subsequently, 115 pathologists were recruited and each were asked to diagnose a set of 60 biopsies. This data was collected and a simulation was created to assess the impact of twelve different second opinion strategies on the variability in breast pathology diagnosis. In the simulation, if the second opinion diagnosis was the same as the initial diagnosis for a given patient, that was taken to be the final classification. If the two diagnoses were different, a third opinion was taken. The result of this simulation was compared to the reference diagnosis previously established. If the diagnosis generated by the simulation differed from the reference diagnosis it was termed a misclassification.
After carrying out their simulation, the researchers found that 24.7% of breast abnormalities were misclassified when only one doctor carried out the diagnosis. Unfortunately, no second opinion strategy completely eliminated this misclassification. However, when a second opinion was established for all biopsies regardless of the diagnosis made by the first pathologist, the number of misclassified tumors decreased to 18.1%. Additionally, it was found that when second and third opinions were determined by pathologists with high work loads, and therefore presumably more experience, accuracy increased. It was found that in eleven of twelve second opinion strategies, the number of misclassifications were reduced in comparison with a single interpretation. The one strategy for which this was not the case was when a second opinion was only requested when the initial diagnosis was cancer. This indicates that seeking a second opinion only for biopsies diagnosed initially as cancer is not beneficial in terms of improving the number of breast pathology misclassifications. Interestingly, this is common practice in some medical laboratories.
The findings of this study indicate that the accuracy of breast abnormality diagnosis could be improved by seeking second opinions on all breast biopsies from experienced pathologists. Although this is an interesting finding, and does have the potential to guide diagnostic protocols, it may not be feasible to establish second opinions on the millions of breast biopsies taken each year. However, the findings of this study do indicate that the majority of misclassifications occur when dealing with atypias and DCIS, two conditions that are on a somewhat ambiguous continuum between benign and malignant. Because of this, it may be particularly important to seek a second opinion on breast biopsies for which these conditions are indicated by the initial interpretation.
Written By: Katrina Cristall, BSc (Hons)