What’s next with this evolving technology?
Whole slide imaging is a powerful tool when used appropriately. But in terms of clinical use, the technology is still in its infancy because the speed it provides pathologists is something akin to a toddler’s tentative gait.
“We are just beginning to see adoption for reading slides without a microscope,” said Michael Feldman, MD, PhD, FASCP, associate professor of pathology and laboratory medicine at the Hospital of the University of Pennsylvania, Philadelphia. Feldman, who presented a live webcast on “Whole Slide Imaging and Beyond” for the American Society for Clinical Pathology, noted that the real challenge in the proliferation of whole slide imaging lies in the fact that the available imaging systems to date really don’t improve the speed or functionality of what a pathologist can do with a microscope.
Digital Speed Bumps
“Right now, using a microscope is much faster than using a digital slide system,” said Feldman. The reason for the time lag is the massive amount of data contained on a whole slide. “The size of the image is very large – 30 or 40 monitors worth of data” said Feldman. “A monitor can only display a couple million pixels at a time, but when there are a billion pixels to look at, they have to be pushed through screen by screen. It’s a process that is about 50-60% slower than viewing through a microscope. If you slow your pathologists down, that adds considerable cost to the diagnostic process.” Feldman noted vendors are working tirelessly to enhance the technology and are focusing on ways to improve workflow, process and performance as well as to add computational image analysis into the anatomic workflow to enhance efficiency and speed to diagnoses and drive down the costs of digital pathology adoption. “But we’re just not there yet,” he added.
However, there are places were adoption makes sense-places where the business driver to adopt trumps the time challenge. “For example, remote places where there are not many pathologists available, or where a subspecialist may not be available,” explained Feldman. “In such a case, you can send the slide and share expertise remotely without having to fly someone hundreds of miles to your location. Pathologists and sub-specialty pathologists are limited commodities; there’s an important business driver in being able to digitize a slide and move it across a distance.”
Bringing whole slide imaging to a dollars-and-cents reality is critical, given the expense of the undertaking. “It costs $250,000-$300,000 per scanner, and you can’t have just one. A very busy lab might need 8 to 10 scanners, plus a couple of lab technicians to run them. A facility could be looking at millions of dollars in scanner costs plus hundreds of thousands of dollars in recurring full-time employee costs. Add to that data storage costs. It’s very expensive.”
Beyond Whole Slide Imaging
Whole slide imaging already offers tremendous advantages in remote consultations and telehealth, and will be even more compelling in other uses when speed is increased and cost is decreased. Whole slide images tend to be brightfield images – as if a standard microscope image were placed inside a computer.
Another technology – multispectral imaging – moves beyond that visual concept. “Multi-spectral imaging allows us to interrogate tissue in a multiplexed, multichannel way through the use of antibodies and fluorophores,” explained Feldman.
What, exactly, might be detected on multispectral inspection that could not be seen on whole slide imaging? Feldman answered, “We will be able to dissect oncogenic pathways, immune responses; it will depend on what markers are available on the protein level. It’s very much like flow cytometry for the slide – it allows us to investigate one or more proteins in each and every cell, and allows us to understand the biologic questions of cell proliferation and cell death.
“The way we do things today is one color at a time, and then we try to visually compare things. The reality is that it is very hard to do,” Feldman said. “But being able to stain tissues for four, five or more antibodies at a time so that we can co-localize the information brings us to another level of sophistication in diagnosing disease and prognosticating outcomes. It has huge potential.”
Right now, multispectral imaging is primarily a research tool, requiring a multispectral technical platform (Feldman said PerkinElmer offers an “excellent platform” which is in use at the University of Pennsylvania) as well as expertise in the specialized staining.
As for clinical use, it’s not far away, said Feldman. “We are having great progress. I expect to see this first in hematopathology where it will be very useful. At Penn, we believe we may be able to bring up four- or five-color staining as a routine analytic test in the next two years.”