CEO SUMMARY: Digital pathology is considered to be one of the more disruptive technologies now finding acceptance in anatomic pathology. Since founding Aperio Technologies, Inc., of Vista, California, in 1999, President Dirk G. Soenksen, M.S., M.B.A., has been in the forefront of this important trend. In part one of this two-part interview, Soenksen discussed the most significant forces now reshaping the profession of surgical pathology. Now, in this concluding part two, Soenksen addresses some of the barriers to the adoption of digital pathology technology. He also explains why partial adoption is the preferred course for most anatomic pathology laboratories.
“Anatomic pathologists have a bright future, because they will benefit from rapidly developing technologies in gene sequencing, molecular analysis, digital pathology, and integrated informatics.”
–Dirk G. Soenksen, M.S., M.B.A.
Founder, President and Futurist, Aperio Technologies, Inc.
Second of Two Parts
EDITOR: Today, I would like to discuss the widespread myths and misunderstandings about digital pathology, particularly those around full adoption and partial adoption. However, as we start part two of this interview, Dirk, I’d like to remind our readers of the key points you emphasized during part one of this interview.
SOENKSEN: Let’s do that.
EDITOR: First, we discussed major trends in healthcare that you see as reshaping anatomic pathology. One such trend is the ongoing expansion in knowledge of the human genome, along with the technologies needed to speedily sequence whole genomes at a cost that shrinks steadily.
SOENKSEN: This trend creates an opportunity for pathologists to advise physicians about how to interpret genetic information and how to develop a proactive wellness and care plan for the patient.
EDITOR: One essential point you emphasized was that, even in the age of whole human genome sequencing, healthcare will still need pathologists to identify abnormal tissue.
SOENKSEN: Yes. Pathologists will continue to do everything they currently do, particularly as it relates to tissue. The new opportunity for pathologists is to become the integrator of all the diagnostic and clinical information on an individual patient—to provide a holistic view of the patient—and then advise that patient’s care team.
EDITOR: You and I also discussed how the exploding number of molecular and genetic testing technologies opens new doors for pathologists. You pointed out an important fact that makes a good springboard for launching part two of our interview.
SOENKSEN: Does that fact deal with the rapid increase in molecular and genetic testing in surgical pathology, which means that pathologists now find themselves spending more time per case?
EDITOR: That’s correct. It’s the productivity issue. As I recall, although you noted that while this additional time per case translates directly into a more accurate diagnostic answer for the patient, it reduces the productivity of surgical pathologists.
SOENKSEN: This trend is simple to understand. Many types of cancers and diseases require the pathologist to make the primary diagnosis of positive or negative. Then, if the patient is positive, the pathologist will spend additional time in two professional activities associated with each individual case.
EDITOR: I suspect one activity is to conduct the follow-on tests that are done after the primary diagnosis. The need to identify the various types of lymphomas and leukemias would be one example.
SOENKSEN: Yes. The need to do follow-on testing is one factor that adds time to the case. The second factor is that pathologists more frequently must pull together all the diagnostic and clinical information, then participate in consultations with the referring physician and the patient’s care team. That requires substantial amounts of time and reduces the productivity of pathologists—but for worthwhile reasons.
EDITOR: Tumor boards are one such activity that shows how pathologists increasingly sit at the table with other physicians to discuss patient cases.
SOENKSEN: Both of these activities show how new technologies and new clinical knowledge can reduce the productivity of pathologists—even as these developments help them deliver more value to the physician and his or her patient.
EDITOR: This is a timely place to shift our conversation and discuss the adoption of digital pathology. You are on record as stating that many misconceptions exist across the pathology profession about when and how to acquire and implement a digital pathology system.
SOENKSEN: Many aspects of digital pathology are poorly understood. That is understandable, since this technology is new and its capabilities are advancing rapidly.
EDITOR: I am curious as to whether you see more acceptance of digital pathology by young pathologists coming out of their residencies and fellowships, in contrast to pathologists who may have been practicing for a decade or longer?
SOENKSEN: A majority of medical schools use digital slide images today, and so, yes, there is a sense of comfort among younger pathologists with the use of digital pathology. Having said that, there also are quite a few pathologists with 15 years of practice experience who are embracing digital pathology as well. I regularly meet pathologists who have practiced for 15 years who are ready to adopt digital pathology. They recognize that their profession is going digital and that they will have to use these new tools sometime during their career, so why not start now and be among the first? The point I’m making is that the correlation between a pathologist’s age and readiness to adopt digital pathology is not as clear as you might expect.
EDITOR: Let’s discuss the cost to implement digital pathology and other impediments to adoption. Cost certainly must be high on the list.
SOENKSEN: Yes, that’s partly true for this reason. When many pathologists think about adopting digital pathology systems, they immediately think about full adoption. Full adoption means digitizing all cases and every glass slide, then reading all these images on a monitor. But the belief that the pathology laboratory must immediately scan 100% of their daily case flow—full adoption of digital pathology—is a false argument.
EDITOR: So your point is that there are smaller steps any pathology laboratory can take to acquire the capability to digitally scan slides, then work with those digital images.
“It is important to understand the difference between scanning speed and through-put through the digital pathology system.”
SOENKSEN: Yes. We can talk about the benefits of partial adoption in a moment. But when a pathology group believes its only choice is full adoption, they have created a false impediment in making a sound clinical and financial decision.
EDITOR: What causes so many pathologists to look at digital pathology as a “full adoption” decision?
SOENKSEN: When they assume that they must digitize 100% of the glass slides they currently process in their lab, they then make another leap. They believe—falsely, I might add—that scan speed is the most important criteria. They believe that today’s digital scanners aren’t fast enough and so they can’t adopt digital pathology.
EDITOR: I’ve heard that argument. Why do you consider it fallacious?
SOENKSEN: This is the argument that many vendors want pathologists to believe. These vendors tell pathologists that scan speed is the impediment for adoption. Yet, nothing could be further from the truth. Scan speed is not the primary metric to tie to adoption of a digital pathology system. That’s just false.
EDITOR: Can you explain why pathologists should consider other factors, and not exclusively the scan speed?
SOENKSEN: For many digital pathology vendors in the field, it is common for them to discuss the speed of digital scanners and the cost of the instruments that make up their system. But—and this is an important distinction—they rarely discuss throughput.
“Most pathology laboratories opt for phased adoption of digital pathology. This is true for both academic center pathology laboratories and community hospital-based pathology groups.”
EDITOR: Would you explain why speed and throughput are not exactly the same?
SOENKSEN: It is important to understand the different concepts of “scanning speed” and “throughput” when discussing the performance of a digital pathology system. Take the example of an instrument that has a 60-second advertised scan speed—which you might think means you can scan 60 slides per hour—but whose actual throughput is only 10 slides an hour. Some instruments take only 60-seconds to capture the image, but then require up to an additional five minutes to post-process the digital slide image to get it into a viewable form.
EDITOR: That’s an important distinction, because it represents the productivity of the scanner in actual clinical operation, and not just the time required to capture the digital slide image and store it in some temporary memory in the computer.
SOENKSEN: This is why it is important to understand the relationship of “scan speed” to “throughput” for any digital pathology system. If pathologists are bombarded with information about scan speed and they don’t pay attention to throughput, for example, it will be more difficult to properly evaluate how the proposed digital pathology solution can benefit their pathology practice.
EDITOR: Do you have a metric that is more important than scanning speed?
SOENKSEN: In my experience with pathology clients, the metric that matters the most is the cost per slide. To calculate the cost per slide, it is necessary to include all costs associated with the instrument, the software, the labor, and image storage. Even the instrument with the highest throughput may not provide the lowest cost per slide if, for example, it has a high rescan rate that requires lots of technician time to do the rescanning; or if it uses a file format that results in larger images that will consume more storage.
EDITOR: Dirk, this brings us to the subject of phased adoption of digital pathology versus full adoption of these systems. What do you see unfolding in the pathology marketplace?
SOENKSEN: Most pathology laboratories opt for phased adoption of digital pathology. This is true for both academic center pathology laboratories and community hospital-based pathology groups.
EDITOR: When implementing digital pathology in phases, what clinical activities offer the quickest benefits?
SOENKSEN: We see pathology labs set up a digital pathology system and then use it, as appropriate, for remote frozen sections, for image analysis, for sending selected cases out for consultations and for tumor boards. They will also flag selected slides to be imaged and archived so they easily retrieve them. In academic center labs, certain pathology slides will be scanned and used in medical education.
EDITOR: In October, I was able to tour the pathology laboratory at the University Hospital Network (UHN) in Toronto, Ontario. They showed me how they are using their digital pathology system to do remote frozen sections.
SOENKSEN: Yes. It was as early as 2004 when UHN pathologists began to digitally scan frozen section slides at the different hospitals they cover, then read those digital slides at their main office.
EDITOR: Now the UHN pathologists have a frozen section service with a rural hospital that is located 425 miles north of Toronto. Also, they regularly use their digital pathology system to provide consults to at least two other hospitals in Ontario that are located are hundreds of miles from Toronto.
SOENKSEN: This demonstrates the power of digital pathology to bring digital images to the right pathologist.
EDITOR: The remote frozen section service also shows how digital pathology can boost the productivity of pathologists, since they don’t have to physically be present at the site where the specimen is harvested.
SOENKSEN: This was one major benefit from the story you published in THE DARK REPORT last summer about Northwest Pathology, located in Bellingham, Washington. In an example of phased adoption of digital pathology, it provides a remote frozen section service to 49-bed Ketchikan General Hospital, which is in a remote area of Alaska.
EDITOR: That is an excellent example of phased adoption. Not only does this digital pathology relationship cross state lines, but it allows the hospital to schedule more surgeries. That is revenue positive for the hospital and allows more patients in Ketchikan to stay in town and be served by their local hospital.
SOENKSEN: This shows how partial adoption of digital pathology gives a pathology laboratory the capability to develop additional clinical services as new opportunities present themselves.
EDITOR: While we are talking about adoption of digital pathology, we would be remiss to not address how digital pathology can position pathologists to benefit from the rapid advances in informatics technology and software algorithms that are capable of doing sophisticated image analysis in research settings.
SOENKSEN: It will take more time before surgical pathologists will be able to work routinely with these types of image analysis solutions.
EDITOR: Can you talk a bit about how image analysis will move in a parallel adoption curve with digital imaging and the wider use of digital pathology systems in clinical care?
SOENKSEN: You are correct that image analysis is a big driver for pathologists and scientists who want to do research. But in terms of routine usage by surgical pathologists, I believe image analysis for clinical applications—beyond the use for quantifying digital IHC for ER/PR and Her2, where we do see adoption—is far out into the future and will not be a major driver of clinical adoption.
EDITOR: Why is this true?
SOENKSEN: There are two reasons. First, routine use of image analysis requires widespread usage of digital slide images in the workflow, which is only beginning to happen now. Second, the time frame associated with clinical validation studies, FDA clearance (or approval), and reimbursement is very long.
EDITOR: As we conclude this interview, Dirk, what advice do you have for working surgical pathologists in a community hospital setting? What are the essential elements those pathologists would need to take advantage of digital pathology?
SOENKSEN: Our opinion is that pathology is going digital and it is getting there quickly. Our business is built upon this belief. Thus, unless a surgical pathologist is so close to retirement that he or she won’t need this technology, embracing the use of digital images will bring significant benefits.
EDITOR: How does this tie in to pathologist productivity and the ability to deliver greater value to referring physicians and their patients?
SOENKSEN: We’ve discussed a few ways that digital pathology can contribute to improved productivity of pathologists. The more skills that individual pathologists have, the more efficiency they will gain.
EDITOR: At a time when the Generation X and Generation Y pathologists want more balance between work and play, how is digital pathology important to them?
SOENKSEN: Any resource that contributes to greater efficiency gives pathologists of all ages the option to devote that time to their priorities. For some, it may be to spend more time with family and to pursue hobbies like golf. Or it could be to make more money by reading slides, for example, digital slides from consultations.
EDITOR: Let’s end our interview by considering the pathologist who is the physician business leader of his community-hospital- based pathology group. What recommendations or insights would you offer to this pathologist who is seeking to keep his or her group on the front edge of clinical services while also maintaining financial sustainability?
SOENKSEN: We find that just by being open-minded, digital pathology can deliver immense value. Whenever we visit a community hospital and talk about how digital pathology fits into the workflow, if the pathologist is open minded, he or she absolutely sees the value in adopting digital pathology for some specific application (not for full adoption). They see opportunities to improve their value to physicians, along with the financial benefits that would also result.
EDITOR: Thank you, Dirk.
SOENKSEN: You’re welcome!
Dispelling a Myth, Expert Says Few Laboratories Are Ready for Full Adoption of Digital Pathology
MANY PATHOLOGISTS AND LAB DIRECTORS incorrectly believe that, when adopting digital pathology, they have to start with full adoption. This would require scanning 100% of the slides, then viewing all sides on a monitor.
“The idea of having to start with full adoption is problematic because very few labs are actually ready today to support the full adoption of digital pathology,” said Dirk G. Soenksen, M.S., M.B.A., the Founder and President of Vista-California based Aperio Technologies. “For full adoption of digital pathology, four requirements must be met.
“First, the pathology lab must have bar code capability,” he noted. “Second, the digital pathology system needs to be integrated with the laboratory information system (LIS). Third, the workflow must be as continuous as possible. Finally, the lab must be willing to validate its digital pathology implementation.
“Many pathology labs use bar codes today, and some have integrated their digital pathology systems to their LIS,” he continued. “But if the pathology lab lacks small batch and continuous workflow, particularly in how it processes tissue, then full adoption of a digital pathology system will likely be cost-prohibitive.
“Overnight batch processing of tissue specimens often leaves the pathology laboratory with a small time window during which to digitize all glass slides,” said Soenksen. “A batched workflow produces a large number of glass slides at one time, typically in the early morning. In the case where there may be only a four-hour window during which all glass slides would have to be digitized, the lab would require six times as many scanning instruments compared to having a 24-hour scanning window.
“The cost-per-slide for a digital pathology system decreases dramatically with a longer scanning window,” noted Soenksen. “That is why continuous workflow is required to make it cost effective for a pathology lab to opt for full adoption of digital pathology.
“How tough is it to achieve full adoption?” he asked. “Currently, Aperio has more than 850 digital pathology installations worldwide. Our most fully adopted customers are in Sweden. These are the only examples of near full adoption that we have seen anywhere in the world.
“In addition to having validated their digital pathology solution, the most advanced pathology labs in Sweden have implemented continuous workflow and bar codes,” continued Soenksen. “They have also integrated their LIS with our image management software.
Shortage of Pathologists
“It was the shortage of pathologists in Sweden that motivated these laboratories to fully implement a digital pathology system,” he observed. “Sweden provides an example of how external pressures and the need to optimize the productivity of individual pathologists can encourage full adoption of a digital pathology system.
“It is important to recognize that the example of full adoption in Sweden is quite unusual,” he added. “Everywhere else in the world, the other 849 current installations using an Aperio system are embarked on the path of partial adoption of digital pathology to realize the benefits and productivity contribution that results from selective use of a digital pathology solution.”
Digital Pathology Used for UCLA-China Consults
EARLIER THIS YEAR, Aperio announced that its digital pathology system was being used to support subspecialty pathology consultations between Ronald Reagan UCLA Medical Center in Los Angeles, California, and the Second Affiliated Hospital Zhejiang University, (ZHU) in Hangzhou, China.
UCLA has more than 30 sub-specialty pathologists who provide remote consultation services to pathologists and clinicians at the 2,000-bed ZHU hospital. The natural follow-up question is: Are more international, cross-border arrangements in the offing?
“We are certainly aware of a number of projects involving academic programs to support certain countries,” stated Dirk G. Soenksen, M.S., M.B.A., the Founder and President of Aperio. “We are working with a variety of institutions that want to put a scanner in Africa or the Middle East to provide pathology services to those regions.
“Also, we absolutely believe that a viable, reliable secure pathology consultation network will be a significant driver of adoption of digital pathology,” he noted. “This also could foster the growth of outreach programs that use digital pathology to support diagnosis and consultations.”