Radiology’s Role in Precision Medicine - Dr. Rajarshi Banerjee, CEO of Perspectum

Michael Carrese: Hi everybody I'm Michael Carrese. Precision medicine is most often associated with the field of genetics, but today we're going to explore how the use of advanced imaging technology is also enabling providers to customize treatments. With us to explain is Dr. Rajarshi Banerjee -- also known as Dr. Banjo -- CEO of Perspectum, a global precision health company focused on improving the diagnosis, treatment, and management of metabolic diseases and cancer. Prior to co-founding Perspectum, in 2012, Dr. Banjo was a research fellow at the University of Oxford and developed MRI techniques for rapid liver assessments while there. Dr. Banjo also has many years of experience in running clinical trials, and we're really happy to have you join us on the program today. Thanks for coming.

Dr. Rajarshi Banerjee: Thank you very much, Michael, for having me.

Michael Carrese: We always like to start with learning more about our guests, what first got them interested in medicine, and in your case, as I understand it, you were drawn to heart and liver disease early on. Tell us how all that came about.

Dr. Rajarshi Banerjee: Sure. So, I grew up in the UK, and I'm the only child of a father who was an engineer and a mother who was a computer scientist. So clearly, I grew up unable to do anything apart from games on a computer, and unable to change a light bulb. But I was okay at biology, and I was okay at sort of...I guess you would call them the caring aspects of professional behavior. And so, I went to do medicine in Oxford and now I've ended up spending half my life here. But yeah, think of me as a failed engineer and a failed computer scientist.

Michael Carrese: I'm sure your parents are happy anyway, right? And imaging...how did you drift over to that space?

Dr. Rajarshi Banerjee: If you look at the history of medicine, going back two hundred years, sight is one of the key senses to use in making a diagnosis, whether it's trying to determine whether or not you have liver disease by looking to see evidence of jaundice, or yellowing in the whites of your eyes; pallor in your mouth, or around your lips as a sign of anemia; scars to suggest old surgery; or looking for lumps and bumps. So, our eyes are very important to us in medicine and increasingly since the 1970s, experimentally, and in the 1990s in major mainstream practice, we've been used to CT and MRI scans to see inside us. I became very interested in that because when learning the practice of cardiology, scanning is very, very important because it determines who does or doesn't have heart disease and what kind of heart disease they have, and that's where I really fell in love with scanning,

Michael Carrese: Because it enabled you to do a better job of diagnosis, or you just loved the technology piece of it, or some combination?

Dr. Rajarshi Banerjee: Well, remember Michael, I'm a failed engineer, so it's never going to be about the technology for me. 

Michael Carrese: (laughs)

Dr. Rajarshi Banerjee: For me, it was always being able to make the diagnosis and being able to explain to a person what was going on. There's a technique we use in cardiology called ultrasound, or echocardiography, where you use an ultrasound probe to see the heart beating in real-time, and you can explain to a patient what's going on. For conditions like heart failure, where the heart doesn't pump as well as it should, you can show what's happening to a patient and explain to them why they have to take certain types of medicines -- ACE inhibitors, beta blockers, and others -- and how that can over time improve the quality of cardiac functional or pumping function. That sort of dynamic feedback really can only come from images. You can't do it with biochemical outputs or lists of numbers. Pictures paint a thousand words, and make the art of diagnosis much more palatable.

Michael Carrese: As a company that works with images, we completely agree with that. So how did all of this lead you on the path to Perspectum? How did that come together, and what was your original thought?

Dr. Rajarshi Banerjee: So, for all the lovely things I've said about echocardiography, one of the problems with echocardiography is it's very manual. You have to physically put an ultrasound probe on someone's chest and move them around to see different aspects of the heart and there is a better technique than ultrasound using magnetic resonance images. These are the big machines that have no radiation, but they're quite loud, so patients have ear defenders on when they go into them. MRI is really the gold standard for cardiac imaging and it was designed originally so that you could scan babies' hearts and children's hearts in great detail without any radiation. 

In the early 2000s, when I was in training, the only way to really learn cardiac MRI was to do it as part of a Ph.D. program. So, I came back to Oxford from London where I was practicing to do a Ph.D. program in cardiac MRI, and while here, I fell in love with some of the new techniques that were emerging as to how you could map the hearts to look for disease and apply them to other organs, and that gave birth to Perspectum, which was a spin-out company from the University of Oxford in 2012.

Michael Carrese: And Perspectum was focused on the heart and liver from the beginning, or what was the focus?

Dr. Rajarshi Banerjee: Well, the idea was, again -- going back two hundred years in medicine -- we have often in the medical professions used biopsy as the definitive way of working out what's wrong with any part of Michael. Twenty, thirty, years ago, if I thought there was inflammation in Michael's heart, I would do a myocardial biopsy. I would literally take a little bit of your heart out, typically from the central chamber, put it on a microscope and look at it to see if there's evidence of inflammation or fibrosis or anything else. Now that's pretty invasive, pretty painful, and not risk-free, as you can imagine. And so cardiologists, being fairly innovative, developed scanning mechanisms that allowed them to diagnose myocardial fibrosis, or myocarditis, without doing a biopsy of the heart. 

Now, these methods, usually using MRI, have gained a lot of public attention recently as myocarditis has come into the public eye in conjunction with some of the mRNA vaccines and certainly also with long COVID. But the critical thing for Perspectum is, if you can remove the need for a biopsy of the heart, can you remove the need of the biopsy from other organs like the liver or the kidney or the prostate? So Perspectum tried to do computer-assisted imaging with AI to replace biopsies and make it therefore a lot safer and more scalable to diagnose inflammation or fibrosis or other diseases of many organs in patients.

Michael Carrese: And you managed to do that?

Dr. Rajarshi Banerjee: We've definitely managed to do that in the liver. We're now in guidelines for liver disease. If you have liver inflammation of any cause, you can have a liver multi-scan to determine how good or bad it is. If you have autoimmune hepatitis -- which is a waxing and waning condition of the liver, triggered by the immune system -- you can use imaging to guide how much steroids or immunosuppression you have. And you can use imaging to determine if you have a liver transplant, or whether the transplant is being rejected. All of these indications that I've described historically would need a biopsy to diagnose them, which creates a barrier and a delay for a patient getting the right diagnosis and therefore the right treatment. Diagnosis is only really part of the journey, right? Working out what's wrong with Michael is the first step to then hopefully, making him better, or, in some cases, counseling him as to how he can make himself better.

Michael Carrese: Well, you can see, of course, with chronic diseases, how this would also be helpful because as you're monitoring the progression of a disease, you can also do that without an invasive procedure.

Dr. Rajarshi Banerjee: Absolutely, and one of the nice things about scanning -- going back to my heart failure examples to start with -- is when you can show a patient what's wrong, and if they can see it and understand it, which is easier with images, you then kind of make yourself redundant if you do your job properly. Most patients want to understand what's wrong with them and giving them images that they can see -- to empower them, to give them the knowledge to work out what's wrong -- they can then take action to deal with it themselves. Whether that's taking the right tablets on a regular basis because it's less of a chore and more of a benefit; whether it's abstaining from alcohol if they've got alcohol-related liver inflammation; or altering their diets; or in the case of patients with gallstones and pancreatitis, deciding whether or not they want to have a gallstone removed. There are lots of indications where actually the decision-making rests with the patient. They just need to understand the parameters of that decision.

Michael Carrese: You can see from a patient education perspective how that'd be pretty powerful. So, give me some other examples of the precision medicine piece of this. As I said at the beginning, it's something people will automatically associate with genetics. For providers using your products, how does it help them tailor things?

Dr. Rajarshi Banerjee: Well, historically, when people have looked at precision medicine, they've generally looked at genetic markers of rare diseases or histopathological markers of, for example, breast cancer to say, “Well, for this kind of breast cancer, we're going to do this type of treatment based on its genetic profile and pathological profile.” You can also look at a person or a person's organs with cross-sectional imaging and get a very accurate characterization of what's wrong. The best examples outside of the liver and the heart are probably in the prostate and the breast. I can work out with incredible clarity what kind of prostate cancer someone has, and what they're likely to respond to, just from a scan. Precision medicine is really all about -- for a given condition with an array of possible treatments -- an array of possible hammers with which to hit one nail. Which hammer works best? If there's only one hammer, you don't really need a precision medicine decision tool because you’ve got one hammer, you’ve got one nail, just do it. 

But when you have a variety of treatments -- for example, in prostate cancer or many solid organ tumors; in liver disease, where there are many, many potential therapies for liver inflammation -- it is better for the patient to get a tailored treatment. It's possible to tailor a treatment on genetics, of course, on pathology, but also on imaging. Actually, if you project forward to when our kids are adults -- and my kids are five and seven, so we're talking about fifteen or twenty years down the line -- I suspect that they will have integrated diagnostics that tailor their treatments in all three domains; imaging, pathology, and genetics.

Michael Carrese: So right now, there really are no systems that integrate that seamlessly or without much trouble for providers?

Dr. Rajarshi Banerjee: There are no technology systems that do that. Good doctors do that all the time. When patients see physicians, the really great doctors don't wow you by how complicated they describe something. At least for me, they wow me by how simple they make things out. So, I may go into a clinic, or into a medical meeting with a complex case of "condition X" and a patient who also has other comorbid conditions A and B, and has had these odd symptoms of P and Q, and someone really good, Michael, will be able to say to me, "Hmm, this fits with that and that fits with this.” This reduces all these spinning plates to two or three questions that need asking, and if the answers are this and this, then the treatment is this or that. That degree of simplicity, almost creating the story for each individual patient based on them, rather than a textbook...that is the art of really great medicine.

Michael Carrese: Yeah, the communication piece is so important. So, we talked about biopsies. Are there other traditional testing methods that your technology can replace?

Dr. Rajarshi Banerjee: The other thing apart from biopsies that I think is ripe for a little bit of disruption, and a little bit of disruption is always good, is how we assess patients with multi-organ disease. Historically, going back maybe now just thirty or forty years, it's always been a story of specialization. I remember here in Oxford, we used to have five or six consultants running the whole hospital Department of Medicine, and now we have maybe two hundred. There's much more specialization. So, whereas previously you had a cardiologist who did all cardiac medicine and quite a lot of other types of medicine, you now have cardiologists who just do interventional cardiology, or just do imaging or just do electrophysiology, or just do rhythm disorders or just to genetic disorders. Problems occur when you have all these specialists, but you don't have someone, as a human, who can integrate across disciplines. When a person comes in with, for example, breathlessness, a good internist will assess their lungs, their hearts, their other internal organs that may lead to breathlessness -- or for example, muscle diseases, pancreatitis, inflammation -- and work out on balance what's wrong. 

Specialists often find this difficult to do because they're great in their domain, but outside of their domain, it's a little bit risky for them to opine. As we bring together technologies that integrate across technology -- we've talked about imaging + pathology + genetics, for the human -- actually what we need to do is integrate across organs...lungs + heart + kidneys, for example. 

There are some physicians who are very good at this naturally and there are some specialties that are very good at this naturally as well. Anesthetists, or anesthesiologists as you call them in the U.S., internists and intensive care doctors...they're used to looking after a whole patient rather than just one organ or just consulting on one organ. And so that's an area where Perspectum is quite deeply entrenched as we try and address what we call in England “the multimorbid patient” -- someone who's got more than one thing wrong with them -- and therefore, it's a balanced scorecard approach as to what you treat and how you treat it.

Michael Carrese: You mentioned long COVID before, and this has emerged as an application, as I take it, for what you're talking about because it does affect multiple systems.

Dr. Rajarshi Banerjee: Exactly. Approximately 6% of U.S. patients who have symptomatic COVID still have symptoms at three months, and 15% of them, it's almost 1%, have symptoms at one year. That's 1% of all Americans who've had COVID. That's a lot of people. And these people have symptoms really in three domains: they have breathlessness, fatigue, and brain fog or cerebral impairment. If you have breathlessness, you can see three or four different types of doctors. Brain fog, do you see a neurologist, do you see a psychiatrist, do you see a counselor, do you see a rehab doctor? And fatigue...I mean, there are many Sundays, I'm sure Michael, you wake up fatigued.

Michael Carrese: (laughs) Just Sundays? I could throw five or six other days in there, too.

Dr. Rajarshi Banerjee: So, it becomes a real challenge as you try and tease out which patients need which specialties, if at all. Add to that the American Disabilities Association has defined long COVID as a disability if you can have demonstrable organ impairment. So, we've worked really hard with the different agencies -- including Health and Human Services -- to produce a test that's scalable across the United States that assesses multiple organs at the same time. We've received a billing code and we're in discussions about how we roll this out over the next few months, mainly to make it accessible to the wide variety of patients that have long COVID. But yes, it's a big challenge. It's a multi-organ challenge, and it's a multi-specialty challenge, and so standardizing some aspects of it by how we look at the body, I hope, will be useful.

Michael Carrese: So, this is an MRI scan that you folks are able to interpret in a way for the provider where they can see what's going on in various organs?

Dr. Rajarshi Banerjee: Right. It's a forty-minute MRI scan that looks at your lungs, your heart, your liver, your pancreas, your kidneys, and your spleen. Based on that, in the best-case scenario, we can say, "Michael, you have no significant organ abnormalities. You may have long COVID. But hopefully you're recovering. There's nothing that we can do right now other than rest and rehabilitation." If that’s not the case -- if we pick up damage in one or more organs -- we can at least direct you to the right, we would say here in the UK "ologists.” 

If we have problems with your lungs, we'll direct you to the pulmonologist; if you have problems with your heart will direct you to a cardiologist; if you have problems with both, we'll say go to one, but bear in mind that the other one is also impacted as well. This makes it a lot more efficient for the payer, and also for the patients. Because remember, these patients are starting out with either brain fog or breathlessness or fatigue. So, multiple visits to multiple specialties is not really top of their agenda. 

Michael Carrese: What are you trying to tackle next? I mean, not that that isn't enough to keep you busy for a long time, but where do you see this technology going?

Dr. Rajarshi Banerjee: Michael, as you came out with that question, I thought my mother had fed it to you. (laughs)

Michael Carrese: (laughs) I'm sorry about that.

Dr. Rajarshi Banerjee: No, no, don't be. I love my mother. She's always pushing me. The thing I'm proudest about in Perspectum is we have a lot of scientists who do great basic science that's applicable to the assessment of human health and we also have great clinical scientists who addressed tough questions. Long COVID is a case in point. When we started doing long COVID research two and a half years ago, it wasn't sure that it was a real disease. No one really wanted to know about it. I mean, acute COVID was bad enough. Who wanted to deal with the long-term consequences? And in order to address it, we would have to work with a variety of different stakeholders. This all seemed very challenging in a world where you couldn't meet people or travel freely. But I'm very proud of the guys for taking that on and doing it. 

When you ask what we want to do next, it's the next big challenge in digital health, which is really all aspects of health but addressed in a quantitative digital function. Personally, I think the biggest areas that need attention, certainly in the U.S., are metabolic disease and the cancers that arise from metabolic risk factors, which are colon cancer, breast cancer, and liver cancer. The rise in the rate of obesity and metabolic diseases means that unless we do something about them, there is going to be a fourfold rise in breast, colon and liver cancer in the next two decades because of the rising risk factor profile in the coming generations. 

So, we either deal with it with a public health policy and good diagnosis of the chronic disease and mitigate the risk factors, or we deal with the cancers. More likely, we'll have to do both because I don't foresee our public health efforts completely reversing the trend in obesity-related disease and in all of these indications. Imaging along with blood tests, but probably not necessarily genetics, are going to be the key drivers of early diagnosis, treatment, and monitoring.

Michael Carrese: What about the access question there? This obviously seems like terrific technology to have. What kind of penetration are you getting? What are you hoping for? We hear a lot of guests talk about how long it can take for adoption on the part of providers or hospital systems. Talk a little bit about that side of it.

Dr. Rajarshi Banerjee: We're very fortunate in a way, and it's serendipity. It's always easier to get an MRI scan than it is to get a biopsy. If I told you I was going to stick you in the side with a nine-centimeter needle and I could do that locally, or you could drive fifty miles to an MRI scan and have a scan, which was as good, I'm fairly certain that you would cover the costs of that petrol for that journey.

Michael Carrese: You're right. (laughs)

Dr. Rajarshi Banerjee: So, access is partly determined by risk, and the risk of a biopsy is much higher than the risk of a scan. Secondly, the United States is very well-resourced with regard to scanning equipment. There are approximately thirteen thousand MRIs in the United States, which is also coincidentally the same number there are of McDonald's franchises.

Michael Carrese: (laughs) That's a great statistic. Oh my gosh.

Dr. Rajarshi Banerjee: For our kind of metabolic imaging, it’s sort of a one-to-one ratio. We've worked with the manufacturers -- especially General Electric and Siemens -- and we know that they're open to having more access to MR scanners, whether it's mobile systems to service rural areas; or whether it's renewing the equipment and existing hospital-based systems in the cities in the metropolitan areas. They're very open to it. That is their business model. But what we can do is add AI to the scans to make them shorter, more efficient, and more quantitative, so that you can see the tissue characteristics as well as the anatomy. 

But you're right. Access is still an issue across many types of healthcare, even just seeing your first doctor who may say that you might have liver disease. That's why we need to work with agencies like Health and Human Services and CMS to make sure that our tests are appropriately coded and covered by Medicare so that they are accessible to people. I have to say that our U.S. team has worked incredibly well with those agencies to raise the profile of access for chronic disease.

Michael Carrese: That's great news. As we wrap up here, we always like to have our guests provide a little advice to our audience -- which is mostly students and early career professionals -- about dealing with this moment in time, but also just generally looking forward to their career. What do you tell young docs?

Dr. Rajarshi Banerjee: To young people in medicine...firstly, you're in a very, very privileged career. You get to meet some wonderful people, both professional colleagues, but probably most of all patients. You'll hear the most amazing stories from people who you get the privilege to talk to as part of your job. But the biggest thing I can say is, you must never forget that it's a caring profession and actually being a great healthcare professional is not about how much you know, or how much you can do, or what procedure you feel that you can do better than anybody else. The best physicians I've met, and I've been lucky enough to speak to many, are the ones that care the most. That in turn ensures they have the highest standards. That in turn ensures that they have the best outcomes, which in turn defines that they are the best. But it all starts with caring.

Michael Carrese: That's a great point. We spend a lot of time on this program talking to entrepreneurs like yourself about digital health and people can get kind of distracted -- maybe that's not quite the right word -- by all the technology and the AI and focused on all these tools and bells and whistles at the risk of losing sight of really what it's about, which is what you were just talking about. So, that's a great note to end on. Thank you so much for spending time with us today, Dr. Banjo.

Dr. Rajarshi Banerjee: No trouble. Thank you, Michael, and good luck with the show.

Michael Carrese: Thank you. I'm Michael Carrese. Thanks for checking out today's program and remember to do your part, to Raise the Line and strengthen the healthcare system. We're all in this together.