Digital pathology has the potential to revolutionize the delivery of faster, more accurate diagnosis by virtue of making it easier and faster to share diagnostic quality images. The sharing of images and data associated with these images makes the access to subspecialists/experts more rapid and realistic than ever before, particularly from remote sites. The use of digital pathology has become more prevalent over the past five years as the cost of technology has gone down and its adoption has increased. Larger laboratories are starting to use it for quality assurance, consultation and second opinions, education, archiving, research, and even intraoperative consultations and on-site evaluation of fine needle aspiration procedures.
Yet pathology’s adoption of digitization lags behind. Radiology has been completely digital for 15 to 20 years now. Pathology is similar to radiology in many ways, yet it is not using digital imaging to its full extent. Why? The answer is that the barriers that have prevented pathologists from fully incorporating digital technology—the cost of technology, the ease of adoption, and the regulatory environment—still stymie laboratories today. More important, there is a fundamental difference between radiology and pathology: radiology, when digitized, does not need films anymore. Pathology still does, and digitization of slides is an additional but important step.
Barriers to adoption
While the cost of technology has decreased over the past five years, it has not decreased to the point that all labs are able to implement the necessary technologies. In addition, labs still have to deal with the issue of ease of adoption and the regulatory environment.
The major obstacle to ease of adoption is the lack of standardization in the imaging format. In radiology, a great deal of effort went into standardizing the storing, accessing, and managing of images. An image taken at the Cleveland Clinic appears the same in Pittsburgh and California. Anyone could read those images because all the data is embedded in the image. That process, known as digital imaging and communication in medicine (DICOM), is standard for handling storing printing images. Pathology has not completely adopted DICOM, so vendors who make whole slide images for, say, Philips or GE, make different imaging formats. They don’t have the same DICOM standards. Pathology has to create DICOM standards for pathology images in order to move closer to full adoption of digital imaging. Work has started in this arena, but progress remains very slow to date.
Then there is the regulatory barrier, which is particularly important for whole slide imaging. When a manufacturer makes scanners for pathology, premarket approval from the FDA is required to market it. A scanner is a Class III device, which is classified by the FDA as the highest risk. In comparison, a microscope is a Class 1 device. There is no solid reason for these devices to be classified as Class III. Due to this higher risk classification, progress, including adoption and innovation in these areas, has been slow, and several vendors have either given up on the pursuit of FDA approval or have left the space due to constraints in resources and negative market forces resulting from regulatory issues.
Advances in cancer diagnosis
One important use for digital pathology is cancer diagnostics. Digital pathology has the potential to enhance cancer diagnostics in several ways. As cancer treatment becomes more specialized, it requires more careful review of the slides and more molecular testing, making it necessary for pathologists to integrate more information from multiple sources as well as consult experts. Often, a pathologist has a very basic knowledge of molecular testing. Pathologists are increasingly turning to subspecialists now to ensure the diagnosis is correct. Sharing digital images rapidly has the power to enable these activities, which depend on rapid consultations and integration of key information to treat the cancer patient.
When a patient is treated with a drug, digitized slides can easily compare today’s diagnosis with yesterday’s, with better access to previous images. As cancer diagnostics become more personalized, and there is an increase in companion diagnostics, the data driven from those tests is very objective. But with digital pathology, we can quantify those slides and better interpret and respond to the data. The question is no longer simply, “Is it cancer?” Rather, the questions are, “How much cancer is there? How big is the cancer? How deep is it? What is the stage and the grade?” With digital pathology, labs and clinicians have tools that can share images and provide answers to those questions.
Training the next generation
Digital imaging and informatics are making their way into residency program curricula. The College of American Pathology (CAP) has partnered with the Association for Pathology Informatics (API) and the Association for Pathology Chairs (APC) to create a program, Pathology Informatics Essentials for Residents (PIER), for pathology residents that focuses on informatics, of which digital imaging is one important component. Digital pathology implementation for clinical and research use relies heavily on pathology informatics because of the infrastructure and interoperability needs of acquiring, accessing, rendering, storing, and using these large data sets. It is essential to have expertise in pathology informatics, laboratory information systems, and data integration to enable the creation of clinical grade pathology imaging systems which are seamless in operation and enable the creation of the tools that future pathologists will need to stay in the forefront of cancer diagnosis and pathology reporting. Many residency programs do not have a formal informatics curriculum yet. More work needs to be done to train residents to give them the tools in technology.
The University of Pathology Informatics (UPI), an educational initiative of ASCP and API, is a self-paced online certificate program leveraging case-based learning to develop critical knowledge of informatics concepts, including digital pathology. Participants learn how informatics can be leveraged to increase quality, specimen throughput, and patient safety while controlling costs in the laboratory. The four core competencies the program covers are Information Fundamentals, Information Systems, Workflow & Processes, and Governance & Management. Courses explore the requirements and opportunities for supporting digital pathology across the healthcare enterprise as well as current best practices, which are based upon the broad experiential input from actual first-adopters and the vendors which are continually building solutions for these users. The program is also offered to Resident Program Directors to use in their curriculum.
The challenge is that digital pathology and informatics is not a priority for residents. When they come to a training program, they learn how to treat the patient and read slides on a microscope. It’s never a priority to learn these new technologies, and it’s not an inherent part of the training program. Until pathology educators emphasize the vital importance of those tools, residents will not see the value. There is a gap between what is expected today and what is expected tomorrow, and the gap must be closed.
Anil Parwani, MD, PhD, MBA, FASCP, is a Professor of Pathology and Biomedical Informatics at The Ohio State University Wexner Medical Center, where he also serves as the Vice Chair of Anatomic Pathology and Director of Pathology Informatics and Digital Shared Resources. Dr. Parwani serves as the Association for Pathology Informatics (API) Program Committee Chair-Elect and Co-Chair of the API Publications Committee.
