CEO SUMMARY: For years, the adoption of digital pathology has lagged behind the predictions of its advocates. That has encouraged one digital pathology company—Inspirata, Inc., of Tampa—to come up with a different business strategy designed to help anatomic pathology labs address the barriers and capital costs involved with establishing a digital pathology system. The Ohio State University Comprehensive Cancer Center is working with Inspirata on this pioneering approach.
THIS SUMMER, an ambitious project to convert the entire department to a digital pathology system is launching at the Department of Pathology at the Ohio State University Comprehensive Cancer Center. The goal is to digitize all current pathology images, along with thousands of archived glass slides going back at least five years.
What adds interest to this large-scale effort to implement digital pathology is that it involves an unusual long-term partnership. The agreement was announced last month and calls for pathologists at the academic center to work with Inspirata Inc., a company in Tampa that aims to transform cancer diagnostics and provide turnkey digital pathology workflow systems. (See TDR, August 3, 2015.)
Under the agreement, Inspirata will install all necessary hardware and software, and implement services required to run a fully-automated, end-to-end digital pathology system. One unusual aspect of the deal is that Inspirata also will provide the staff to manage the project and operate the digital imaging equipment at the Arthur G. James Cancer Hospital and Richard J. Solove Research Institute (OSUCCC-James) at The Ohio State University Wexner Medical Center in Columbus.
This digital pathology project will be a sizable undertaking. The pathology department has 35 pathologists and 45 other staff who together handle about 70,000 cases annually.
“The initial plan is to digitize 10 years’ worth of glass slides in our archives. Those are the slides that we use the most, because we have the most need for them clinically,” stated Anil Parwani, MD, PhD, a Professor of Pathology and Vice Chair of OSU’s Anatomic Pathology Department.
“There are two aspects to our desire to move to digital pathology,” he continued. “First, we see it as an opportunity to improve workflow. Second, we can use digital images to improve patient care and enhance the education of our trainees.
“We have slides that go back several decades,” said Parwani. “But obviously, we will not digitize everything. Instead, we will digitize select groups of cancer cases that reach back more than 10 years.
“We also want a workflow that allows our pathologists to do a look-back on certain cases and to use those images as an archive for cancer research,” he explained. In addition to his work for the Pathology Department, Parwani also is the Director of Pathology Informatics and Director of the Digital Pathology Shared Resource at OSUCCC-James.
Digitizing Archived Slides
Along with the workflow benefits from adopting digital pathology at OSUCCC-James, another goal of the project is to digitize, analyze, and archive millions of retrospective and prospective pathology slides, specifically to advance patient care.
“This system will make it easier for us to catalog how each patient’s cancer progressed or regressed over time while looking at the treatment protocols their OSUCCC-James oncologists followed,” noted Parwani. “For research purposes, we will be able to match those diagnoses and treatments with their associated patient outcomes.
“We envision that using the digital pathology system and images will enable us to perform analyses that we cannot do now, given the limitations we have with glass slides,” he said. “Inspirata’s digital pathology system will digitize the glass slides we have in our archives and create digital images from slides of our current patients.
“Another problem we expect to solve with this new system is improved integration with the lab information system (LIS),”added Parwani. “Lacking integration with the LIS adds 10% to 15% of the time it takes to prepare a glass slide and present it to the pathologist.
“With the Inspirata system, we expect integration that leverages all the various information systems in the OSUCCC-James Pathology Department while at the same time reducing the time it takes to create a slide and an image.
“In essence, we are creating a new workflow to fix that problem that pathologists face with most glass slides: the lack of associated information,” he emphasized. “That’s why we say this new system is all about what I call information management. It is a prerequisite for managing all the data in the lab: the slides, the assets, the blocks, the paper requisitions; all of which are useful and needed information for both current cases and for research.
“Once our current slides are digitized and we digitize our archive of pathology cases, we will have created a rich, highly annotated image database,” noted Parwani. “Our researchers can then tap into that database to create very powerful and predictive algorithms.
“As an example, say I am looking for 600 cases of breast cancer with a specific histological subtype, a specific demographic, or some other feature of the tumor or patient type,” he stated. “I now have access to that data and the corresponding digital images. As a next step, I can train the algorithms to create an image dataset that supports multiple diagnostic and research purposes.
Matching Data With Images
“This type of work is already being done on a small scale in other countries and in other clinical settings,” explained Parwani. “Here at OSUCCC-James, we expect to create a large dataset—of cancer cases mostly—and link that data to other vital information about cancer and patient type, tumor type, and treatment protocols, and any other information that may be useful. That’s what’s referred to as the Cancer Data Trust.
“Our goal is to be able to share these images easily,” he emphasized. “We have several network or affiliate hospitals, and with this platform we will be able to move these digital images among pathologists for consultation.
“Like any database, the larger the pool of images you have, the more value it will have in terms of its ability to be predictive,” continued Parwani. “We are leveraging the predictive power of numbers. To boost that power, we can potentially invite other institutions to participate as well.
“If we have 10 cases here at OSUCCC-James of a certain type of cancer and Memorial Sloan Kettering has 20 cases of that same type, and Johns Hopkins has 30 cases, we can share these images and allow our scientists to work on these datasets collectively,” he noted.
Inspirata’s Digital Pathology Contract with OSU Means Vendor Must Hit Specific Milestones
OVER THE PAST DECADE, adoption of digital pathology systems has not been as fast as predicted by its proponents. Cost has been a major reason for the slow adoption.
First is the capital cost of acquiring the digital scanners, hardware, and software needed to produce whole slide images, store them, and work with them. Second is the ongoing cost of labor and related expenses to scan the hundreds of glass slides generated daily by a large anatomic pathology lab.
“Generally speaking, if a hospital wants to implement a digital pathology system, it must first answer several important questions,” stated Satish K. Sanan, Chairman and CEO of Inspirata, Inc., of Tampa. “If it wants to do it on its own, it must determine if it could afford the software, along with whether it should buy digital pathology software or write its own. It can require several million dollars of scarce capital to assemble the right digital pathology solution.
“Then comes the hardware and staffing that is required,” he continued. “What scanners should it buy? Who will scan the glass slides? Is there space in the lab to accommodate digital pathology equipment? And what about the need to store the digital data produced by the scanners?
Barriers to Faster Adoption
“Each question involves a specific barrier to faster adoption of digital pathology systems,” said Sanan. “To overcome these barriers and help labs adopt digital pathology solutions, Inspirata is pursuing a business strategy we believe better meets the needs of pathology labs interested in acquiring digital pathology systems.
“Inspirata offers a solution-as-a-service (SaaS) model that includes aspects of a managed service agreement with payments that are tied to client-approved milestones,” he added. “In addition, our SaaS model provides implementation services as well as the onsite project management, staffing for the scan lab, and overall program management. With our model, the institution essentially gets everything needed for an end-to-end digital pathology solution except the pathologists.”
Inspirata’s strategy in the OSUCCC-James project has three components. “The first component involves the digital pathology hardware, software, and workflow systems that a pathologist requires,” noted Sanan. “Inspirata has its own workflow solution, and, instead of manufacturing our own scanners, we have partnerships with slide scanner manufacturers, including Philips, Motic, Sakura, Mikroscan, Huron, and Objective Imaging.
“The second component is our milestone-based payment arrangement so that labs and health systems do not have to come up with the full price of the digital pathology systems at purchase,” he said. “This enables the lab to conserve cash. Instead, Inspirata offers the lab up to a 10-year pay-as-you-go plan.
“The third component is that Inspirata will hire the staff needed to scan and archive the glass slides and to manage that process,” stated Sanan. “This is important because existing histology workflow is preserved.
“Inspirata develops the financial plan and the plan to integrate the new digital system with the existing workflow, including all of the delivery, adoption, training, and change management,” he said. “Full implementation takes about 18 months.”
Another twist in Inspirata’s strategy is to reduce payments at the start of the multiyear agreement. “In this early period, payments are much less and don’t increase until the lab gets new revenue increases from the use of digital pathology,” Sanan stated. “This helps the lab achieve its desired return on investment and enjoy some positive cash flow.”
Use of Algorithms
“What is important is that, by having such a large dataset of a particular cancer type, we can use algorithms to quantitate the information in ways that were not possible before,” said Parwani. “Typically when a pathologist looks at a slide, he or she would apply special biomarkers to the tissue and then would manually quantitate the results.
“But with this system, a pathologist can apply algorithms to quantitate that case more accurately and more objectively,” he explained. “Moreover, these algorithms can separate certain features in the images that a human cannot separate easily and thus would be unable to quantitate.
“Algorithms also can find features that might not be obvious to the pathologist because those features actually are a combination of features and not just one feature,” added Parwani. “Taken together, this particular combination of features can be predictive and quantitative. These features act almost like an image signature to help us make a prediction. We can say that, based on these particular features, this patient will have a good outcome or this patient will have a bad outcome.
“We can also use algorithms to do risk assessment. We can build algorithms based on these different features,” he stated. “For example, there’s an algorithm today that can do image-based risk assessment for breast cancer. To build such algorithms, pathologists need a large image dataset, and now we will be able to create that type of a system.”
To build such a system, the Pathology Department will need to create a new work- flow. “What Inspirata offers is the ability to become a digital pathology lab that elimi- nates all the problems we have with glass slides today,” observed Parwani. “To do that, we will need a new way to work.
“We believe that digitizing glass slides will improve turnaround time signifi- cantly,” he said. “This will be particularly true for difficult cases that the pathologists want to show to colleagues, even someone who is not in the vicinity or is in a remote location. That digital image can be sent and an answer returned within minutes!
“This can’t be done if the lab is using only glass slides and doesn’t have digitized images,” he added. “Once installed, the Inspirata system will digitize slides rapidly. But it cannot cut down the time required to produce a glass slide. That’s beyond their control.
“Another reason this system will be useful for us is that it will allow the pathologists in the existing James Cancer Network to collaborate more effectively,” added Parwani. “The James Cancer Network is a group of community hospitals that have partnered with the OSUCCC–James to bring subspecializing oncology care to outlying commu- nities. It will allow us to share cases more readily, to do tumor boards when needed, and to have ongoing collaborations with pathologists in different facilities.
“Another long-term goal of great importance for us is to create additional revenue opportunities for Ohio State and for the James Hospital CCC,” he concluded. “We expect to achieve this by linking our capabilities to other facilities, not only here in Ohio, but to other places in the United States or anywhere in the world.”
Contact Anil Parwani at 614-293-7625 or email@example.com; Mark Lloyd at firstname.lastname@example.org or 813-570-8900.