“Prostate Cancer Biopsy – Digital Pathology”
Interview
May 9, 2017
Darryl Mitteldorf, LCSW, Executive Director of Malecare Cancer Support
Russ Granzow, General Manager at Philips Digital Pathology Solutions
Esther Abels, Director of Quality, Regulatory and Medical Affairs at Philips Digital Pathology Solutions, and the current chair of the Digital Pathology Association Regulatory Taskforce
Darryl M.: When patients have a biopsy, tissue samples are harvested from their prostate and someone in a laboratory carefully puts a slice of that flesh onto a glass slide and a physician known as a pathologist looks at it. Why does the use of glass slides need improvement? Isn’t having tissue on glass a perfect way for a pathologist to actually image a sample from a patient and understand where their pathology might lead their treatment?
Russ Granzow: It’s a good question. In fact, a glass slide and putting a piece of tissue, a very thin slice of tissue on that slide is a very good way to image the tissue and look for suspected prostate cancer or benign or malignant tissue and we’re not actually changing that part of the procedure. What we are changing is that an area that needs vast improvement and that’s how one interprets the images on that slide, from that slide. The traditional way is a hundred and fifty year old technology of looking through a microscope.
It’s an analog approach and what we’re doing is really enabling a digital approach to medicine that’s the last field of medicine that has to be digitized, and this enables more effective and more accurate diagnosis, better workflow and faster turn around times for patients and better understanding of the disease because you can do a lot of downstream digital analysis which helps to decide what types of drugs patients should receive, what types of treatments they should receive. So, it’s really about how the images are analyzed, not so much about how the slides are prepared.
Darryl M.: How does digital imaging of the slides actually helps analysis. One could imagine you could magnify an image to a greater size than under a microscope. What else can you do with it?
Russ Granzow: Yes, the biggest opportunity really lies in the fact that once you digitize, you have information about every single pixel, every single component of that image now in digital format. And you can apply algorithms and technical filters, you can apply mathematical equations to those images to extract much more information than a pathologist could extract simply by looking at the image with his naked eye. You can do all sort of convolutions, you can compare slides, you can send those slides digitally to your collaborators around the world for second opinion and additional advice on a difficult case.
So really it enables this ability to work in a digital world by very sophisticated mathematical methods but also to collaborate much more effectively than sending glass slides through the post.
Darryl M. : It sounds like there are three specific advantages. One is the speed that a pathologist could look at a sample and come up with an analysis of it. Second, using algorithms and basically using a computer as a sidekick to a pathologist, and third around the ease in getting second and third opinions. Malecare strongly suggests that patients get second and third opinions on their pathology. By having second and third opinions, it increases patient safety by a quantum level and gives patients a broader view from many doctors.
Russ what you’re suggesting is patients and their physicians could email their biopsies to whomever they want around the world and ask for a second, third, fourth or fifth opinion, is that correct?
Russ Granzow: Yes, as long as it’s done under the purview of the medical community and done with the support of the physician and oncologist, absolutely. I mean, one of the reasons you see this need for second or third opinions is because the mind is not a perfect thing. The eyeball is not a perfect thing. People see different things when they look at images, just like you see different things when you look at artwork. And that variability has led to this requirement for second and third opinions. Now that you can have more quantitation, now that you can have second and third reviews done digitally, it will happen faster and it will be more effective, and I think it will ultimately improve care for patients.
Darryl M. : You mentioned the idea of using algorithms and computers and perhaps artificial intelligence or what I believe Philips is calling computational pathology. Can you describe that a bit more?
Russ Granzow: Yes, I think the simplest way to discuss it is in the context of the prostate cancer patient. I’m sure many men with prostate cancer are familiar with their Gleason score, and Gleason is a fairly complex, multivariate analysis done by pathologists on tissue samples, biopsies for prostates, and it’s a fairly convoluted, rather empirical approach that has a lot of variability and in fact the nature of how it’s developed, been developed over the years, and these things ultimately can move towards what we would think of as a digital Gleason, and just be more accurate.
Those things aren’t available yet, of course, but we’re working hard in that area and looking forward to bringing that to the FDA for clearance at a future date. But digital pathology enables the ability to analyze and lead to digital Gleason scores actually in the future.
Darryl M. : How?
Russ Granzow: It allows you to apply a mathematical routine to images and come up with a quantitative, numerical scores that then can be associated with the disease, prevalence of disease, the impact of disease on the patient.
Darryl M. : Digital pathology is a wonderful idea. I could imagine that the standard of care may be a digital interpretation of pathology first, followed by supervision by a human pathologist. In my mind, this opens the entire world of patients in lesser developed countries as well as in lesser developed neighborhoods in the United States. Access to healthcare is a challenge for many people. The monetary savings of a digital pathologist – having your slides read by a computer rather than by a human – creates opportunities of excellence for so many more people than exist today. Is that a fair way to present the hope for the future around this?
Russ Granzow: Yes, I think we firmly believe, both as a company and I think, individually. Pathologists will, certainly never be replaced. Their role is evolving and changing. The demand on them is ever growing because there are less and less pathologists going into practice. There are more and more complex diseases, subtypes of cancer, and there’s an ever increasing amount of drugs to treat those diseases, those cancers, so the complexity of what one calls personalized medicine is increasing logarithmically and there are less and less pathologists trained to deal with it so, we see digital pathology as really helping to deal with that complexity and do it in different and creative ways.
Certainly access to care both in settings where it may not be available, be it rural settings or be it in rural countries, will be benefited by digital pathology. That the ability to send an image to a pathologist taken from somewhere in Sub-Saharan Africa to an expert at Johns Hopkins or Massachusetts General Hospital will be enabled by the effective technology. We are already starting to see that kind of deployment so we won’t be replacing pathologists any time soon. All of these solutions are used as an aid to improve workflow efficiency and improve the effectiveness of the decisions by the pathologists.
Darryl M. : We touched a bit on the technique of imaging and magnification, understanding structure down to a pixel. Is 3D imaging an added innovation around biopsy interpretation?
Russ Granzow: Yes. Some structures and some diagnoses like cytology require 3D analysis to really look at that 3D architecture of the cell, so there certainly are opportunities and we’re developing 3D technologies ourselves to really apply that kind of technical approach to those analyses. In the case of most of the work that pathologists do today, it’s still in the 2D domain and will remain so, but then there are certainly opportunities for the future.
Darryl M. : One can imagine novel, new and as yet unimagined biomarkers to be discovered as a consequence of this.
Esther, is whole slide imaging is currently available in Europe, is that correct? What are the challenges in bringing it to the United States?
Esther Abels: Yes.. It’s grown to be available in Europe already for a few years, and indeed we have some discretion especially with the FDA on bringing in the device to the market in the U.S. and what were the challenges there is that, of course, the FDA has their obligation to look at the benefit, what can it bring to the patients, but in the meantime, also look at the safety and effectiveness of the device. That’s something that we as manufacturers have to prove. We have to come up with the evidence for that.
And that’s something that we have been working on together with the FDA and also with the Digital Pathology Association to get clarification on what that whole trajectory would be. Also, what the concerns are of the FDA, what we have to … what kind of evidence we have to provide and that really ranges from showing that your device is performing the same over and over again. Also if you use, if you have different users using it, you have to show that each user can use it, you have to provide some guidance on how to use it and train them, and that’s something that you have to show that the training is effective.
And next to that is also all the technical aspects that are in the works. Do we have designs, how we have designed the device for example that you look at. The colors that we are using. How the scanner is working. How it creates an image that the whole processing of that images don’t correct it over and over again.
And next to that of course we have to show that you actually can replace a microscope with a whole slide imaging device and that a pathologist is able to give the same diagnosis as he would have been doing on the microscope. So that you can just interchange the modalities, that you really can replace the microscope that’s a whole slide imaging device. And that’s something that we had to prove. That we have to come up with evidence. So we studied a lot of organs, a lot of diseases, and we have proven that indeed the device, the whole imaging device is comparable to the microscope.
Darryl M. : Is it available in the USA?
Esther Abels: It’s available in the U.S.
Darryl M. : It is available today in the U.S. as you just said. But saying that is different than it actually being available. If a patient has a biopsy in June 2017, how reasonable could that patient expect to have a digital presentation of their biopsy available?
Esther Abels: Right. And that all depends of course on the users. it’s actually revolution in medicine. And as you had mentioned before, there’s a transformation phase for pathologists. They have to get used to this. So there’s a transformation and a phase that people are having time to get the devices installed at the laboratories and start using it according to its intended use. And that takes some time. So how likely is it? It could be that it is available in one lab, but not yet in the other lab.
Darryl M. : So, when do you imagine this would be universally available for all patients in the United States?
Esther Abels: Wow that’s a real difficult question. I truly believe in digital pathology. And I really hope the sooner the better. Because as it opens doors. Also now with digital pathology you can start making use of second opinions and also having cases referred to sub-specialists who have much more knowledge on a special disease in a shorter time frame. So your turn around time to get a diagnosis is shorter. And that will ultimately will benefit the patients.
So I truly believe in digital pathology. But it just takes time. So how long that will take that everybody will be transferred? Yeah, I cannot really answer that. But I hope it will be in a year from now. That’s my personal hope. But of course we also know that labs have to go through a transformation.
Darryl M. : It’s been in use for two or three years at least in most of Europe. Esther and Russ, are you able to say that pathologists and their patients who are using this in Europe are having a higher quality experience or more accurate understanding of their cancer diagnoses as a consequence of whole slide imaging?
Russ Granzow: Sure. I think what we are certainly seeing is that the pathologists that deploy digital pathology are seeing the benefits immediately because of workload. I’ll give an example that’s relevant to prostate cancer. When a lot of biopsy cores are taken, there could be 10 or 15 biopsy sampling from the patient and those are put on slides … 10, 15, 20 glass slides with tissue on it. And the workload of doing that for a significant number of patients every single day becomes quite material and if a digital pathologist system can order those slides simply by scanning them before the pathologist even arrives in front of the monitor to review them such that they are ordered in order of which one has tissue and which one doesn’t, which one has tissue that may be relevant to look at first as opposed to fatty tissue or something.
It makes the pathologist’s day a lot more streamlined, a lot more effective and turn around time can be improved. I think the first thing that patients will see, improved turn around time, improved confidence around the diagnosis, the ability to have a rapid second opinion rather than shipping slides hundreds of miles away and then getting a two or three day turn around. This could be done in the same day, potentially. So the patient will see these benefits immediately. We also see benefits longer term in improving quality … those studies are still being done, we’ll start publications around those over time.
Esther Abels: Yes. So I think I absolutely agree with Russ. And also what you see happening is that pathologists are now, they already collaborated of course, but then they indeed have to ship the glass slides to their colleagues to ask for their opinion, second opinion maybe or maybe even a consult as a sub-specialist. And what you see now, they can just send a link to their colleague and ask for their opinion directly. So the speed of the diagnosis as Russ said, the turn around time is much shorter. And that’s really beneficial for patients because we can all imagine everybody has prostate cancer to refer back to that, you want to know what the diagnosis is, where you are and how you will be treated. And so that’s indeed something that we see happening already in Europe that is faster.
Darryl M. : I think I misunderstood the process itself. It’s not so much that you are e-mailing the imagery. You’re actually emailing a link to wherever the image is stored, is that about right?
Esther Abels: Yes. You are sending as a pathologist who has scanned the slides, they have access to the images. And they can send that link to their colleague. And then the pathologist clicks on it and logs into that server of that hospital.
Darryl M. : Is there anything else you would like to tell? What other innovations can you tell us are in the Philips pipeline for cancer patients over the next year or two or three?
Russ Granzow: I think as you pointed out, the real future of this whole industry now that we have FDA cleared and regulatory cleared, digital pathology. That’s really just the first step. Digitization is not the end-game. The reason to do this is to do all one can do with digital pathology. Our focus now is 100% related to really enabling the next steps of using digital images in different ways, more effectively, moving those images around in the cloud, enabling access through digital images that have been transformed from slide and then analyzing those in all sorts of different ways such that we can have better tests, faster turn around times, more accurate tests.
The pathologist is the one that diagnoses your cancer. And you don’t have it until he or she says you do. So they are a very important part of the process. A lot of people hardly ever meet their pathologists, I think will be more and more patient outreach of pathologists to their patients, they will be part of the dialogue for the pathologist. And digitization will help them communicate what they do, how they do it. And I think everyone will be better for it.
Darryl M. : That’s great. So thank you very much. Today we have been talking with Russ Granzow and Esther Abels, both from Philips Digital Pathology Solutions about the new way of, the new world of digital imaging of biopsy samples. Thank you for being with us, see you next time. Bye.