Nice piece by Jeannie Crofts from Northwestern University Medill Reports Chicago and Liron Pantanowitz at UPMC on recent work at Stanford University using advanced technologies for tumor analysis in breast cancer.
Nov 16, 2011

Despite advances in treatment, breast cancer remains the most common cancer for women in the country, claiming 40,000 lives last year alone. A team of Stanford University engineers and doctors is hoping advanced technology will better evaluate breast cancer and potentially save lives. The study in the Nov. 9 journal of Science Translational Medicine showed how researchers trained computers to analyze tumors, which they say is more accurate than human analysis. Click here to read the full story.

Medill News Service asked a top anatomical pathologist to explain the challenges of evaluating breast cancer, what he thinks of the study and where medicine and technology go from here. Dr. Liron Pantanowitz is also an associate professor of pathology and biomedical informatics at the University of Pittsburgh Medical Center and a fellow of the College of American Pathologists.

Q. What are the challenges when it comes to diagnosing or evaluating breast cancer?
A. When pathologists evaluate breast cancer with the use of their microscope, they essentially have to perform two tasks - both with inherent difficulties. The first is to make an accurate diagnosis based mainly on the structure or form of the cells.  Once they have made a diagnosis of cancer, their next task is to provide some predictive information about the tumor’s behavior and prognosis, as well as information about that tumor’s likelihood of responding to specific targeted therapy. If this latter task is performed with the light microscope and human eye, then such scores may vary. However, if this scoring is performed using image analysis by examining a digital image, then it is likely that the results will be more reproducible. In other words, if every pathologist is given the same yardstick to perform these measurements, the results will be more accurate.

Q.  What do you think of this Stanford University study?
A. I think the study is a really well conducted one and is important for two reasons. First, there are likely many factors about cancers that may be important for prognosis. This study uncovered one such prognostic factor [A characteristic that can be used to estimate the chance of recovery from a disease or the chance of recurrence]. Secondly, it showcases very nicely the added value (what I call 'magic' or potential) of a digital pathology image. The data in the digital images of this study helped uncover the important prognostic feature of breast cancer. It is feasible that in the future pathologists may use this C-Path system (computer tool) to measure and report out this important factor in breast cancers to better manage patients.

Q. Do you think there is some promise with a computerized way of analyzing breast cancer?
A. Yes, definitely. As I mentioned above, computer-aided tools allow pathologists to provide more reproducible and accurate diagnoses. In fact, similar tools have been used for several years now by pathologists for breast cancer, some of which are FDA approved for this purpose. However, we still need to see some long-term studies to determine the real value of these tools.

Q. Where do you see medicine and technology going from here?
A. While tools like this are definitely helping us move towards practicing better personalized medicine, I do not believe that they will replace the pathologist. Rather, in this new era of digital pathology, pathologists will be able to use more and hopefully better tools. In radiology, for example, even though their images have been converted to digital format for many years now, radiologists are still required to interpret these images, called upon to apply appropriate computerized tools, and are in the best position to interpret the results from these analyses. I think the future of digital pathology is bright and with studies like this I am pleased to see that we are headed in the right direction.