Digital/Computational Pathology – Ready For Primetime?

“Digital pathology empowers the pathologist. We are no longer chained to the location of the histology lab. On top of that, digital pathology opens up a whole new set of informatics and image analysis tools that we are just now starting to create.”

Jonhan Ho, Director of the Dermatopathology Division and the Fellowship Director, University of Pittsburgh Medical Center, and Founder of KiKo, Knowledge in Knowledge out.
Image Credit: Shutterstock.com

In this digital age, it may come as a surprise that in one of the most advanced healthcare markets in the world, primary pathological diagnosis remains conducted by pathologists peering down the lens of a microscope. Digital pathology is, by now, indisputably the future of the field. Although some are still reluctant to accept its place in the lab – and despite its slow adoption in many institutions – the digital slide image is slowly replacing the glass slide on the microscope’s stage. The flexibility, accuracy, and potential to deploy ever-improving informatics and image analysis tools make digital pathology an attractive option.

Digital pathology is not a new concept. Its history goes back around 100 years when specialized equipment was first introduced, and images from a microscope were captured onto a photographic plate via a camera. Telepathology emerged around 50 years back; however, over the past decade, pathology has undergone a real digital transformation, gradually shifting from an analog to an electronic environment. Here’s a timeline showing the development of digital pathology:

Image Credit: Nature.com and Pathologist.com

My former colleagues at Sg2 and I have been talking about the benefits of implementing a digital pathology solution in health systems as far back as 2012. At that time, professional societies like the College of American Pathologists (CAP) were skeptical. But our extensive experience in the development of radiology digitization convinced us that the lessons we learned in implementing Picture Archiving and Communications Systems (PACS) in radiology could be transferred to pathology. The two disciplines share a lot in common. Here are some of those areas:

  • Both are visual disciplines that rely on the skill of the reader in the interpretation of the study.
  • Both have become increasingly more sub specialized through the years.
  • Both are critical components in determining a diagnosis.
  • Both are facing shortages of skilled professionals sufficient to meet the demand for services. Pathologists have the second highest percentage of active practitioners aged 55 or older – and while many are beginning to retire, fewer new pathologists are moving into the field to offset the shortfall. According to one study, by 2030, the number of active pathologists may have dropped by 30 percent compared to 2010. In some parts of the globe, the shortage of pathologists is shocking. Workload demands are also on the rise. The number of new cancer cases managed by pathologists rose 41.73% in the U.S. between 2007 and 2017. In contrast with physician specialties, “…a pathologist’s workload is not capped by a specific number of patients [who can be seen and treated], but rather expands to encompass all case materials generated by their clinical colleagues”. This results in overwork, diminishing quality, and diagnostic variability.
  • Both have a requirement to store large amounts of information that must be accessed frequently during the patient’s care pathway.
  • Both have a requirement that the data be available in multiple locations for review by multiple members of the care team.

Before examining the benefits and challenges in implementing digital pathology solutions, the projected market growth, and the major players, some basic information on the workflow differences between the traditional approach and the digital approach is essential.


Comparing traditional versus digital pathology workflows

A typical traditional pathology workflow is a complex series of events, including a manual review of glass slides that ultimately results in diagnosis. However, analyzing stains on a glass slide is lengthy and complicated. This whole process might take several days to weeks and might cause damage to slides during transit. It also becomes difficult for the pathologist to search for old slides, images, and case information in the archives of hospitals. In addition, the results depend on the specialty and not on the personal judgment or mental state of the pathologist, as it may lead to a lowering the level of accuracy.

Digital pathology has been introduced to overcome numerous limitations associated with traditional pathology. Digital pathology is a computerized, image-based (digital slides) system for managing and interpreting information. The digital slide is a complete representation of a glass microscope slide, which can be viewed at any magnification, including intermediate magnifications not available on standard microscopes. The slides can be accessed remotely anywhere in the world to deliver consistent, rapid, and accurate results, as compared to traditional methods. The workflow difference between conventional and digital pathology is shown in the graphic below.

Image Credit: Pathkids.com

Next, it is helpful to understand the critical drivers in favor of the adoption of digital pathology solutions versus some of the barriers that have stood in the way of widespread adoption to date. I have found this graphic to be an excellent summary of those elements:

Image Credit: Signify Research

For the US market, two of the most significant barriers to adoption are highlighted above – both revolve around the significant costs associated with digitization. However, the costs of network architecture and storage have decreased exponentially in recent years. And as I mentioned above, many organizations can leverage the existing network infrastructure and archiving solutions in radiology to support the digital pathology program. Finally, as will probably come as no surprise to those familiar with the U.S. health system, lack of reimbursement is the other barrier to adoption. While single-payer markets like Western Europe and the UK, Nordics, and the Netherlands, in particular, established digital pathology networks and workflows via large-scale governmental funding and legislative aid, the US market has historically received no such support.

But change may be on the horizon on the reimbursement front. The recent announcement of new add-on CPT digital pathology codes has caused a stir. After many years of lobbying, the College of American Pathologists (CAP) successfully petitioned the American Medical Association (AMA) to create 13 category III codes. Currently, pathology is reimbursed via a technical component (TC), professional component (PC), or a “global” combination of the two. If industry speculation is to be believed, additional reimbursement will come in the form of an addition to the technical component of the review. These codes went into effect on January 1st, 2023, and while they currently offer no additional compensation, they do offer officials the chance to begin tracking and reporting the impact of digitalization.

A note of caution, however, Category III codes do not have an assigned relative value. Therefore payments for these services or procedures are based on the policies of payors and not on yearly fee schedules. Category III codes also rarely receive national pricing from Medicare. While reimbursement for digital pathology may be possible soon, it’s not guaranteed. Several unanswered questions remain, such as how long will it take for payors to review the data accumulated by the codes, and are any more (for example, Code I) being considered? Crucially, the most important of these questions, how much reimbursement will be, is yet to be answered.

The current lack of reimbursement does not necessarily mean that a business case cannot be made to deploy a digital pathology solution. One recent study was conducted at Poplar Healthcare, an anatomic pathology group in Memphis, with 25 pathologists. Poplar is beginning to realize the benefits of digital pathology and WSI in its daily workflow. Here are the five benefits they have learned from their implementation:

  • It lowers costs by eliminating the need to ship glass slides to remote pathologists.
  • It improves the productivity of remote pathologists, because the whole-slide images can be sent instantaneously to a pathologist along with the case information, and both are viewable at the same time within their case viewer.
  • It allows remote pathologists to share digital images for second opinions, consults, or to refer difficult cases back to their subspecialists for review or for interdepartmental review.
  • It provides faster turnaround time for results, which helps them gain new clients.
  • It provides a platform to increase revenue by delivering services to customers seeking to reduce their histology costs and incorporate whole-slide imaging. Under a model in which Poplar Healthcare works remotely with anatomic pathologists located within physician practices, some of its revenue comes from doing the technical component (TC), some comes from the professional component (PC), and some come from doing both TC and PC (or global). They’re currently reviewing additional revenue streams by providing services overseas, starting with dermatology.

Also, in a paper published by the National Center for Biotechnology Information (NCBI), a study analyzed how implementing a digital pathology system can significantly cut costs. Five years after implementing a digital pathology system, expected savings were estimated at $12.4 million for a laboratory with 219,000 accessions per year. For a large academic-based healthcare organization, there was approximately $18 million in savings upon fully implementing a digital pathology solution.


Companies in Digital Pathology

The global digital pathology market size was valued at USD 926.9 million in 2021 and is expected to expand at a compound annual growth rate (CAGR) of 7.5% from 2022 to 2030.

Image Credit: Grand View Research

The market is segmented into hospitals, diagnostic labs, biotech and pharma companies, and academic and research institutes based on end-use. The hospital’s segment accounted for the largest revenue share of more than 36% in 2021, owing to the high adoption of digital scanning techniques in hospitals for faster diagnosis and better patient compliance.

Image Credit: Grand View Research

The market is usually sub-segmented into hardware, software, and storage segments. You can see the breakout by sub-segment in the graphic below.

Image Credit: Fortune Business Insights

Major players, such as Danaher and Hamamatsu Photonics, Inc., held a dominant market share in 2021. These companies are increasingly undertaking various strategies, such as new product development, extensive collaborative strategy, and M&A, to gain a higher market share. For example, in April 2021, Leica Biosystems partnered with Paige to integrate its digital pathology platform with the latter’s AI-enabled research oncology software. This would extend access to computational pathology products for translational and clinical in select countries across North America and Europe regions. In April 2021, Philips partnered with Ibex Medical Analytics to combine its digital pathology solutions with the latter’s AI-powered Galen platform, thus enhancing its offerings. Leading American biotechnology company Inspirata Inc., Dynamyx™ is the company’s flagship end-to-end digital pathology solutions streamlining image digitization and delivering essential case allocation, outsourcing functionality, and real-time collaboration. In March 2022, the FDA awarded Inspirata clearance for Dynamyx software for primary diagnosis in place of a traditional glass slide. Oxford University Hospitals Cellular Pathology department completed 100% digitization of histology slides with the help of Philips. And recently, FUJIFILM Corporation (President and CEO, Representative Director: Teiichi Goto) announced the company has entered into an asset purchase agreement to acquire the global digital pathology business of Tampa, Florida-based Inspirata, Inc. Upon completion of this agreement, Inspirata’s Dynamyx® digital pathology technology, employees and customers will become part of Fujifilm.

Then you have the lab companies like Labcorp and Quest Diagnostics. Both offer a “digital pathology as a service” business model. Labcorp provides both bright-field and fluorescent digital scanning services using slide scanning platforms to generate high-quality whole-slide images of your glass slides. The images can then be used for various endpoints ranging from pathologist review to digital image analysis and archiving. Quest Diagnostics (NYSE: DGX) and Paige announced a collaboration designed to unlock the potential of artificial intelligence (AI) to improve and speed the diagnosis of cancer and other diseases that rely on pathologic assessment. The collaboration involves analysis using Paige’s proprietary machine learning expertise of pathology diagnostic data and digitized slides from Quest Diagnostics and its AmeriPath and Dermpath businesses to uncover cancer markers and other diseases. Using these insights, the parties intend to develop new software products, which, following regulatory approval, will be marketed to pathologists, oncologists, and other providers to support disease diagnosis. Near term, the parties also intend to license the insights to biopharmaceutical and research organizations to aid biomarker discovery, drug research and development, and companion diagnostics.


The Future of Digital/Computational Pathology

Technology, both in scanning and IT, has finally met the requirements for digital pathology in terms of software maturity and network, computing, and storage capability. Whereas digital pathology enables more efficient workflows, computational pathology will take the field one step further, allowing pathologists to use digital images in more varied and efficient ways. There are many definitions of Computational pathology; however, the one cited in the article “Computational Pathology: An Emerging Definition” is a holistic way of looking at this space. The author defines computational pathology as an approach to diagnosis that incorporates multiple sources of data (e.g., pathology, radiology, clinical, molecular, and lab operations); uses mathematical models to generate diagnostic inferences; and presents clinically actionable knowledge to customers. This vision goes beyond an informatics-centric view and leverages the core competency of pathology and the ability to communicate clinically actionable knowledge effectively. In the future, smart image recognition algorithms could help streamline pathologists’ workflows and help them to focus on the things that matter most. Pathologists will use their hands less and their eyes more, analyzing more data in less time with improved efficiency and accuracy combined with improved prognostics and therapeutic diagnostics across a global network.

“Digital pathology represents the most transformational change in anatomic pathology since the introduction of the light microscope.”

Joaquin Garcia, M.D., Mayo Clinic Division of Anatomic Pathology, Chair, Digital Pathology Practice Subcommittee in Laboratory Medicine and Pathology.

Digital pathology isn’t a replacement for the pathologist; it’s an enhancement. Not only to the practice but to the individual pathologist and the industry. In a recent post, I offered this equation as a framework for determining whether action will occur when introducing new technologies:

Image Credit: Graphic, H. Soch, based on a concept created by Alex Lindsay, Office Hours. Global

The gist of the equation is that action occurs when the vendor can demonstrate that the possibilities in implementing the solution exceed the current circumstances. How does that play out in the case of digital pathology? The key “possibilities” in digital pathology implementation include the following:

  • Revolutionizing the way we treat cancer and what we know about it.
  • Improving the time and accuracy of diagnosis.
  • Greater management and effectiveness of lab staff
  • Automizing tedious tasks without sacrificing the quality of work
  • Enhanced collaboration and sharing via virtual meetings and cloud components
  • Digital storage can transform the workspace of the laboratory
  • Making the discipline more quantitative

So, it is way past time to implement digital/computational pathology in daily clinical practice. I think that mainstream digital pathology adoption is, therefore, no longer a question of “if” but “when” and “how.”

Leave a Reply