GenFit Head of Global Diagnostics, Suneil Hosmane, joins the Surfers to discuss how we can improve the methods used to diagnose patients with NASH or NAFLD and to evaluate the efficacy of drugs in development today. The conversation considers the needs of drug developers, clinicians and patients, and highlights opportunities for improving diagnostic options.
Suneil Hosmane (00:00): This is Suneil Hosmane EVP and global head of diagnostics at GENFIT, and you are listening to Surfing the NASH Tsunami on the NewPodCity podcast network.
Drug developers, investors, researchers, and corporate executives wrestle weekly to understand what is happening in commercial development of NASH medications. Join hepatology researcher and key opinion leaders Steven Harrison, C-Suite veteran Peter Traber, and forecasting and pricing guru Roger Green as they discuss the issues affecting the evolving NASH market from their own unique perspectives on this week’s edition of Surfing the Nash Tsunami.
Roger Green (00:37): For everybody with an interest in NASH or more broadly fatty liver disease, surfs up. Episode eight of Surfing the NASH Tsunami starts now.
Roger Green (00:45): Today’s topic, Diagnosing Diagnostics part one, we’ll look at the good and bad in NASH and NFALD diagnostics today. We’ll look at the processes used to diagnose, classify, and monitor patients, as well those used to prove the drugs and development actually work or not. On a personal note, diagnostics is the topic that attracted me to NASH and NAFLD in the first place, so I’m very excited about this conversation.
Roger Green (01:06): Part two, the Future of Diagnostics will come in two weeks on June 18th with the fourth chair most frequently occupied by Louise, occupied instead by Dr. Mary Rinella of Northwestern University. Next week, June 11th, we will recognize International NASH Day by discussing patient advocacy in different parts of the world, probably another patient advocate in addition to Louise. Stay tuned for that.
Roger Green (01:28): Two notes before we start. Number one, last week I used a quote about looking at a chessboard that I always heard attributed to the chess champion Bobby Fischer. I liked the way that worked so well that I went looking for it online so I could turn it into a meme and couldn’t find it. So maybe he said it, and maybe it’s just that someone once told me he said it, but it’s a great quote, and I intend to keep using it with the caveat about whether it came from Fischer or not. Two, actually this is our second attempt to tape this podcast. Normally we tape on Tuesday afternoons, but yesterday afternoon we had a technical issue that effectively wiped out the recording half an hour into the podcast.
Roger Green (02:01): We’re starting again, this is now Wednesday morning, but unfortunately, Steven Harrison could not adjust his schedule to join us. Fortunately, Suneil Hosmane the EVP and head of global diagnostics at GENFIT, and a person with vast knowledge and experience in liver diagnostics was able to step in in his place. Suniel, welcome. Thank you so much for joining us today. Please say hi to our listeners.
Suneil Hosmane (02:22): Hello, everyone.
Roger Green (02:23): Thanks, Suneil. That’s good. Now onto our podcast, this week we have Suneil as noted, and Peter Traber, and Louise Campbell here with us. To open, let’s discuss one professional highlight from the week. Peter, why don’t we start with you?
Peter Traber (02:37): Yeah. I would say that my professional highlight was I’ve initiated working with a really brilliant group of scientists who are approaching combination therapy in NASH utilizing cell-based human assays. We’ve all talked about combination therapy as being critical in NASH because no one drug seems to have robust enough effect, and because of the heterogeneity patients. But we all know that doing every combination clinical trial is not very feasible, so these in vitro models of looking at combinations is a very exciting area, and I’m very pleased to be helping through that effort. So that’s my professional highlight for the week.
Roger Green (03:25): So Peter, first thanks, that’s an exciting highlight and congratulations. Second, you’re reminding me that we probably should put on our list for future episodes a case for combination therapy where we talk through why combination’s such a good idea, and how to think about it most intelligently as we develop drugs going forward. So thank you, that’s great news for you, and I think it’s going to be a gift over time for all our listeners. Louise, why don’t you go next.
Louise Campbell (03:47): I don’t have a professional highlight for the last week because I took the week off, but my personal highlight was I had a shed that needed building, and I managed with my husband to build it over three days. So we’re sorted now for the summer.
Roger Green (03:59): Excellent. That means that you can spend the summer at work, not worrying about whether you still have a shed to build. I think this is a good thing.
Louise Campbell (04:06): Absolutely.
Roger Green (04:06): Okay. Finally, Suneil.
Suneil Hosmane (04:08): My professional highlight, probably the highlight of the year and maybe it’ll change, but so far it takes the cake. As you may know, my group at GENFIT’s been working very hard for the last several years to generate, evidence, data to support some development that we’ve done in the diagnostic space and all of that work culminated to the acceptance of our manuscript. So we just got the acceptance a few days ago, so big accomplishment professionally for myself, but also for my team.
Roger Green (04:34): Congratulations, that is a huge milestone for the product, and the team and the company, and certainly for yourself as well. I think that’s fantastic. My professional milestone for the week, I think is about this podcast on two levels. Number one is as I mentioned I’m tremendously excited that we’re doing this event and we’re doing this topic because I think it’s important. We are now, I think, starting to step into a different level of thinking about liver diseases compared to merely talking about the drugs. I think that’s a good thing.
Roger Green (04:58): Second, we now have six weeks of data on the podcast, and it appears that we are still way ahead of the pace anybody expected in terms of attracting listeners. We’re frighteningly close to if not over 1,000 subscribers to the podcast which is a really strong number for a topic this limited in scope. I think it speaks to the thirst in the community among all different people, patients, and caregivers, and academics, and commercial people for a common source to learn and talk about the disease. If that’s the benefit we’re providing, I’m thrilled it’s a lot bigger than we thought we were going to get to when we started. Thanks for listening, thanks for subscribing. Please keep sending notes, keep sending questions, keep sending comments. Help us get better at what we do.
Roger Green (05:42): Okay, so with that said, I want to move onto a question. We’re taking a question that came in two or three weeks ago from Dr. Tim Brown at the Abdominal Transplant and Liver Disease Clinic in Milwaukee. He wrote three things. First, that I want to thank him for his kind words about the podcast, he wrote, “I think the show is amazing. Really, really good. Production quality, content, panelists, [inaudible 00:06:02]. Great stuff, please keep it up.” Doctor, it’s three weeks later, obviously, we’re keeping it up, and we intend to do so for a long time to come.
Roger Green (06:07): Number two, he asked, “Is this show edited at all? I’m just curious because I rarely hear one stutter or even a breath of hesitancy in anyone’s speech.” This is not because we are the most articulate people on the planet, Doctor Brown, actually it’s because at the end of every episode when I thank Frank Sasso, our engineer, for making us sound so good, this is exactly what I’m talking about. His job is to make us sound this good instead of like the “um” marginally articulate “um” frequently umming people that we are in everyday life.
Roger Green (06:36): Third and last, and this is the question he wrote, “I am hoping no offense is taken by this, I mean no ill will towards anyone. That being said, with all due respect, I disagree with Doctors Traber and Harrison from the May 19th podcast episode discussing Wild Times in NASH-ville,” by the way, he liked the title, “in which it’s loosely mentioned that ‘no cure’ is possible right now pertaining to NASH.” Dr. Brown goes on to discuss how he’s reversed nine patients with probiotics and vitamin E and finally concludes that, “To suggest that finding the right drug is what we need and that there is no cure sends a message that we as doctors are lazy and just waiting with hope that something may turn up one day.”
Roger Green (07:13): So first of all, as moderator I try to listen carefully, I didn’t hear anybody say that. I spoke with Dr. Harrison about this, we were speaking about it yesterday afternoon, and he noted that obviously, that was not what he meant. That in fact, his comment was that you get to point in disease where it becomes hard to reverse. But up until then, the fact that the liver can regenerate and people have the ability as a result by managing their diet to improve their health says in fact, that things can be turned around. Now, Steve is not with us to speak for himself today, he’s a lot more eloquent, he says a lot more. I’m going to stop there for him. Peter, what did you have in mind when we talked about that on the 19th?
Peter Traber (07:47): Yeah, I think Dr. Brown’s point is very well taken, and this is an example of unintended consequence of what we may have said on the podcast. In fact, Dr. Brown is quite right that there is a lot that individual physicians and their patients can do to ameliorate and even reverse NASH. Physicians should be focused on that, and the more diligent a physician is in that regard, the more likely they’ll successes. So I really appreciate Dr. Brown bring this up.
Peter Traber (08:19): For instance, we know from studies that if one loses significant weight, and it’s not huge amounts of weight, that up to 10% or greater of weight loss can reverse NASH in many individuals with mild disease and can even reverse degrees of fibrosis in patients. So 10% weight loss is an important goal for physicians and their patients to target Even patients who have less weight loss that also include other lifestyle changes such as either cardiovascular or resistance exercise can also improve their NASH. So there’s a lot that physicians and their patients can do.Other ancillary things like vitamin E can certainly be helpful.
Peter Traber (09:06): The real issue here is that while in some groups of patients, up to 90% of people can have a reversal of NASH with a 10% weight loss. In practice, less than 10% of people actually achieve that. So that may be different for diligent physicians as Dr. Brown sounds like, but other individuals don’t have as much success. So when we’re talking about drug development, we tend to forget the baseline lifestyle changes that should be done by everybody, and it’s a great reminder that Dr. Brown gives us with this question.
Roger Green (09:44): Thanks, Peter. Louise, do you have any thoughts that you want to add or any comments about that?
Louise Campbell (09:48): Yeah. I think it’s not that patients want to rely on hope so much. I’d argue that it’s probably in a lot of those patients that get scared, and hope is what they hang onto, it’s because of the late diagnosis. Liver disease in any format struggles very much with early location, and I think we’re going to go on in this episode to discuss some of the diagnostics for that and the reasons. But I think when you’re diagnosed late, pharmacological solutions are the quick win that patients want, changing and reversing that weight is very difficult to do after 20, 30, 40 years of a lifestyle.
Louise Campbell (10:29): Going back to an earlier episode, people do have lack of confidence in their own ability to make these change. That paradigm that Peter rightly commented on, one of the triangles that Steven put together, it shows that you can reverse simple steatosis with as minimal as 3% weight loss, which is achievable and is a confidence boost. So just starting small in those lifestyle changes that Peter was talking about are absolutely key. So I congratulate Dr. Brown on being able to engage the patients to change behavior, and I think it’s vital. We’ve got to use all avenues.
Roger Green (11:05): Thanks, Louise. Suneil, thoughts?
Suneil Hosmane (11:07): I echo a lot of the same sentiment that’s already been discussed and want to congratulate Dr. Brown for the excellent work that he’s been to do, not just in terms of the field, but for his patients. As a general comment I think it’s not about one or the other, I think it’s about the totality of the options that are available to the physician and the patient, and having that provide once again, different therapeutic strategies for the patient given their needs.
Suneil Hosmane (11:32): So lifestyle and these changes which take a long time and will also require truly an adoption of a change of lifestyle, not just a quick fix in terms of a diet and things like that. These things take time, and I think this is where if necessary and if indicated, other therapeutic solutions can help in making that overall change in that person’s life and ultimately prevent progression of disease. I think that’s the overall goal. So in many ways, I think we’re all talking about the same thing, and just thank him for his work.
Roger Green (12:03): Okay. Thanks, Suneil. Thanks, everybody. I agree what Dr. Brown wrote, I agree with everything everybody said. I do want to look at this from a slightly different perspective, which is a point everybody made is it’s about change in lifestyle. Let’s talk realistically about what that means. The late comedian Jackie Gleason once describe himself as the greatest weight loss expert of all time because over the course of his life he lost 1,000 pounds. Unfortunately for him, it was the same 50 pounds 20 times. In fact, when we talk about lifestyle change, what we talk about is something that’s more systematic and requires more time than people give it, number one.
Roger Green (12:39): Number two, I like to think of life as having an illness paradigm, which is about helping people heal and a wellness paradigm, which is helping people stay well. The things that help you heal and the things that help you stay well may not be the same things. So in medical school, my understanding to this date is that there’s not tremendous teaching about the behavioral techniques and psychological things that you would need to do to help a patient stay well after they get well. Certainly, it’s not as much of a focus there as it might in different degrees or different companies that focus exclusively on that.
Roger Green (13:11): What we see now, we see a lot of advertising these days for a company called Noom. I don’t mean just to talk about Noom, they’re not the only ones that do this. But Noom is basically a wellness company, and what they are about is helping people learn the behaviors that they need to learn over time to maintain changes in behavior. I’d like to think as a result Dr. Brown, that the good work that you’re doing will be easier for people to access without necessarily having to go to a medical specialist or a clinic who’s really good at that stuff the way you are, but that they will have more ways to find that through apps and some available to that end.
Roger Green (13:42): I’d like to believe that over time that will lead people to be more successful in losing weight, and as a result, drugs will play less of a role. That said, one of the comments that I know Steven makes is that you never see a sequel to the program the World’s Biggest Loser because tend to gain all that weight back. So we need to help people do a better job of keeping the weight off so that they can have those successes, but at the same time, we need to acknowledge the idea that many people will not be able to do that.
Roger Green (14:05): The point that I believe Steven and Peter made on the original program accurately is that when you get to a point where cirrhosis turns into portal hypertension, at that point it does become very difficult to reverse, impossible really to reverse disease. But up until then, drugs can do it, and lifestyle modifications sticks can do it, and weight loss and eating different foods certainly can do it. Hopefully, that’s a helpful answer Dr. Brown. If you’ve got more you want to say or write about this, write back. We continue this dialogue in writing or bring it up on the show a couple times, we’ll see. I think it’s a good topic.
Roger Green (14:34): With that said, let’s jump on to today, which is about diagnosing diagnostics. What we’re going to do, and I will go first, is each take three, four minutes and share our perspective on diagnostics today and what’s working and what isn’t, and what we might need to think about wanting to do differently as we go on.
Roger Green (14:51): When I first became interested in liver disease, it was because I’d been asked by some clients and some friends in the investment community, Yasmin Rahimi who was on our first couple episodes, notably one of them, to take a look at business models and see what made sense for liver disease as a commercial issue. As I got into it though, I started more and more at diagnostics, and particularly the NASH tech meeting in 2020 featured a talk by Dr. Michael Charlton and another one by Steven that pointed out the idea that biopsy, the supposed gold standard was in fact a deeply flawed technique.
Roger Green (15:23): So first of all, just to give the audience an understanding, when we historically talk about biopsy, we’ve been talking about biopsy for cancer. In cancer biopsy, you know exactly where the target tissue is that you’re shooting into to look and see what you want to find. In liver, we don’t have target tissue, we have a liver. So you take a biopsy out of the liver pretty much in a random location, and there’s no reason to be certain that the location that you’re taking the biopsy from would have the same profile that you would have if you looked somewhere else in the liver. So issue number one is the variability of disease within the liver might mean that the slide that you’re taking is not the best representation of the liver per se, number one.
Roger Green (15:58): Number two, if you think about a drug study, we take one biopsy at the beginning and another after a predesigned period of time, which people have been taking active medication or a placebo group. Three months, six months, 12 months, 16 months are the four most common. Then we try to lay these two slides side by side, and we compare them. If the comparisons are really good and really accurately read and consistently read, even so, we know what’s different, but we don’t know what happened because we have no real progression that enables us over time to understand what’s happening in the liver.
Roger Green (16:28): A lot goes on, livers spontaneously get worse and get better. We see in virtually every phenomenon in life that there are thresholds that you cross, and until you get to the threshold you can go back, but after you hit the threshold you can’t. We don’t really know what goes on within the liver that constitutes the threshold, and we don’t know what happens as you approach it, particularly because we don’t have those kinds of tools present today. So even when we read these things consistently, we know what changed, we don’t know what happened.
Roger Green (16:53): Then on top of that, Dr. Charlton’s paper was about the idea that goes at this point blood tests might be more accurate than biopsies at figuring out progression of disease, which makes sense if you think about what I just said because you can track blood tests over time. Steven’s paper was about the idea that if you ask two different people to read the same slide, the probability that they will read it the same way is somewhere between 50 and 80%, while 40 to 70% actually two different people. If you ask the same person to read the same slide at two different points a year apart, the probability that they’re going to read it the same way as they did a year earlier is only about 50 to 80% depending upon what you’re looking at.
Roger Green (17:25): As a marketing researcher, we do open-ended questions where somebody gives a two-sentence answer, and we’ve got to convey to our client what that answer means. If my coders were as inconsistent in coding as we are scientifically reading liver biopsies, I’d get fired. So I look at all this and what I see is all right, we’ve got point A to point B, we don’t know what happened in the middle, and we have some real error in terms of what we’re reading in the beginning and the end. We’re also, by the way, asking for four separate judgments, three to decide what NASH is and one for fibrosis level.
Roger Green (17:52): So between the inability to talk about what happens as we go through the process and the low CAP ratings, what we wind up with is an inefficient static rating that we’re using to try to figure out what’s going on in the dynamic process. That’s not a good way to learn. I think it hurts us in terms of figuring out which drugs are working when and why. I’m confident in it hurts people like Louise when working with patients trying to help them understand what’s going on in their own bodies and how to make it work better.
Roger Green (18:19): To me, whether we’re talking about figuring out whether a drug is working, getting initial diagnosis on a patient, or helping treat them over time, we need better metrics, we need dynamic process, and we need to do that in a way that’s readily interpretable for the patient. That’s my schpiel. Peter, why don’t you go next?
Peter Traber (18:35): Thank you, Roger. I think your global perspective about testing and how the liver is changing is really critical. I’m going to take a bit of a 30,000-foot view also of this issue before we get into the details of specific tests or new omics tests, metabolomics, or micro RNAs, et cetera. I want to just take an overall look just as review, about one in four people in the world have NAFLD. It’s really important to distinguish between NAFLD, simple fatty liver disease, NASH, and NASH with fibrosis because those with NASH, about 20% will progress to cirrhosis.
Peter Traber (19:17): We know that the fibrosis stage is most predictive of liver mortality where stage three, four advanced liver disease has a higher liver-related mortality. So those are things that we know, but the problem we have is that our diagnosis and our way of following patients critically is based on liver biopsy as Roger said. Now let me just give you two analogies from other important organs.
Peter Traber (19:43): Let’s talk about lung fibrosis. IPF is a relentlessly progressive fibrosis of the lung resulting in lung failure. Biopsy is virtually never done. Diagnosis with a high-resolution CT and symptom complex is made, and then you can monitor lung function; forced vital capacity, diffusion capacity, and you can follow treatment. That’s why we have a couple of drugs approved for IPF.
Peter Traber (20:07): Same thing with kidney fibrosis. The diagnosis is made in many ways, and biopsy is sometimes done for diagnosis. But for following treatment, you can monitor kidney function with GFR. The problem is in NASH and liver disease in general, we don’t have a good simple way of monitoring liver function in patients. We measure a lot of things that are affected by liver disease, but not specifically liver function. So the diagnosis is based on liver biopsy, ability to get a drug approved is based on improvement of liver biopsy, which is as Roger said a very imperfect test.
Peter Traber (20:45): Now again, before we talk about advanced approaches, I want to just mention that not all is nihilistic in this area. Yes, we have lots of problems. There is a lot of research, and there is a lot of information out there for practicing physicians and patients on how one can diagnosis fatty liver disease, and there are basically two ways. One are clinical prediction scores like the FIB-4 and NAFLD fibrosis score, and the other is elastography, either ultrasound or MR elastography.
Peter Traber (21:18): I mention these two because various scenarios have been put together that can rule out advanced fibrosis, not necessarily rule in NASH or advanced fibrosis, although MRE is probably the best test for that. Every physician and every physician group should sit down and discuss how they’re going to use these tests in their clinical practice because they have a very good high negative predictive value. Not so reliable in terms of positive predictive value, but they can be used as an algorithm to narrow down people at this point that may need further evaluation or testing, such as a liver biopsy. So I don’t want people to come from this that all is nihilistic, but in fact, there are ways that we can today evaluate patients with fatty liver disease.
Peter Traber (22:09): Now I’m not going to get into this too much until we hear from the others, but the most challenging thing is hepatocyte ballooning or apoptosis is the hallmark diagnostic of NASH, and we don’t really have a good test to measure that. CK18 has been studied a lot but hasn’t been sorted out to work all the well, so therefore we have two big consortium efforts to take lots of different biomarkers into account to try and see longitudinally over time how various biomarkers may work in NASH.
Peter Traber (22:45): One is litmus, the EU, and the other is NIMBLE here in the US. Both have received letter of intent from the FDA and are rapidly moving forward. We hope over the next few years to gain some of this longitudinal data on standard biomarkers that are already in use and more advanced biomarkers such as microRNA and other things that Suneil may talk about. So that’s my overall view, Roger, of where we sit, and I thought it would be useful to review that before getting into more detail.
Roger Green (23:18): Thanks, Peter. That’s great. Louise, why don’t you jump in next?
Louise Campbell (23:22): Yeah. Find it interesting listening to everybody talking, we’re all coming from our own perspectives. Now I’m coming from obviously more of a holistic and patient perspective. I think the liver is a very complex organ, and when we look at testing it and finding diseases like NAFLD or any liver disease to be fair, we’re left with an organ whose whole basis of being is not to show us any signs and symptoms to of being unwell. Yet we’re asking people to try and diagnose early so that we can stop pathways and progress of diseases that take numbers of years to develop and cause problems.
Louise Campbell (24:01): But also with fatty liver disease, as Peter said, one in four people because everybody’s a patient eventually, has NAFLD. Now that’s a massive global population. We don’t have in any country that I’m aware of a national screening program to locate liver disease. We do have, for example, and I can comment on this, in a woman we have screening programs for breast cancer. If you said one in four people in the world got breast cancer but we’re not going to screen for it, I don’t think many countries would tolerate that sort of lack of screening.
Louise Campbell (24:33): But it is an organ that shows you a lot of problems. It regenerates, it’s there to cope with everything we chuck at it, do to it whether we eat too much, drink too much, have viral infections, genetic. Your liver is not there to show you too many signs, so looking for algorithms and blood test aren’t necessarily the right ways, so what we’re left with is really late diagnosis, people feeling that their options for treatments are limited. I think if we don’t start bringing diagnostic services forward, then we are missing massive populations and opportunity to change lifestyles, outcomes, and prevent type 2 diabetes in a lot of people, the cardiovascular mortality that is significant in NAFLD and other associated lifestyle illnesses because we’re looking through the liver to tell us when it’s not well, and it doesn’t do that.
Louise Campbell (25:28): I think if you look at the Portillo-Sanchez study back in 2015 when they looked at the prevalence in type 2 diabetes with NAFLD and abnormal AST ALT, when you got to a BMI of over 40, 90% of those patients had normal ALTs and AST. But if you got to the non-obese, 36% of them had normal liver tests.
Louise Campbell (25:54): So we are by not screening and with our lack of diagnostics that tell us the full picture, actually failing the population, one in four of the globe. So that leaves us with as I said lack of opportunity we’re learning into pathways of medications to try and help those patients recover, but because they’ve had that long lifestyle, it’s very difficult to change that around. Again, we backload our care. We don’t put in psychologists, dieticians, weight loss programs until patients are coming up to bariatric surgery really have a problem or we’re target their type 2 diabetes, we’re not targeting their NAFLD and their liver; or we’re targeting their heart health with a good diet.
Louise Campbell (26:40): Well, that good diet changes your liver health which then changes and is party to changing your heart health, but there’s very little cross-communication in that from patient perspective. So joining the silo healthcare that we have, of you’re either diabetic or you’ve got heart disease, or you’ve got hypertension, or you’ve got cholesterol, I think we need to look at what is common. And we do know, we’ve got multiple research studies out there that show that fatty liver disease is very common in heart disease, for example. But if I go on to several of the heart charity sites, there is no negotiation or discussion of the impact of fatty liver in heart disease for their patients. So you get stuck with the I’ve only got one disease, whereas we can look at diagnostics that can help both.
Louise Campbell (27:29): I think going back to the question we were asked a couple of weeks ago, and I reversed into who shouldn’t we be screening? By screening the liver to find that the liver’s either stiff or too much fat should guide us to do tests for the liver to find out what that cause is, rather than waiting for the liver to give us a test to show us to start screening and find it because it’s not cost-effective to allow patients to get liver disease late. I think in the U.K. a lot of our patients are diagnosed late. In casualty, only 25% of liver diagnostics or diagnosis takes place in a GP surgery because we’re looking for blood tests to find that disease, and we’re not really getting it.
Louise Campbell (28:11): So I would again, still reverse that and say who should we not be testing for liver disease, and particularly fatty liver disease? So that’s where I’m going to hold that there before we move on.
Roger Green (28:22): Well held, Louise. Thank you. Okay, Suneil.
Suneil Hosmane (28:22): Wow, what a complex and fascinating topic to discuss and just taking a step back, touching so many areas spanning from just philosophically the approach impacting clinical development, impacting utility in the clinic and what that means for identification of patients, and then of course treatment selection and various other things. Of course, that’s just on the science and technical side that you get into financial impact and what this means to overall cost of care. It’s a phenomenal topic.
Suneil Hosmane (28:52): From my standpoint, once again, I’m not saying that it’s impossible feat by any stretch, but just to add a little bit to what Peter was saying when he was describing his perspective. The liver is a very complex organ, and I say that only because all of that complexity impacts the way in which we can identify disease and diagnose disease. So for example, it’s one of the most metabolically active organs in the body, and why is that important? Because it’s production of biomarkers and it’s regulation immune system, in fact the way in which it’s critical for metabolizing drugs. This is a very active organ, and therefore you can imagine that the biology does have some ebb and flow to it, and that’s the noise one has to overcome when we’re developing blood-based biomarkers.
Suneil Hosmane (29:35): When you switch gears and you talk about imaging, eating food can increase stiffness, changes in the physiology of the patient that alters blood flow, that can change stiffness. So there are some complexities just with regards to some of the physical properties of the liver and in how that may influence diagnostics. It’s also an incredibly iron-rich organ, and you can imagine that your ability to image certain things in your liver are also impacted by the level of iron within the organ.
Suneil Hosmane (30:02): So these are just some of the things in the background that constitutes some of that noise when we’re developing diagnostics. Of course the other noise that does occur is we’re developing against a reference standard that’s discreet. We always talk about things in terms of scores and fibrosis stage, but in reality, to continue in the actual organ and we try to for the ease of staging purposes, we try to categorize in the bucket, and that leads to some noise too.
Suneil Hosmane (30:30): But I fundamentally agree with a lot that’s been said here. I also agree with Roger that looking at it in a static view is misleading. I think about things that I did during my PhD, I did a tremendous amount of microscopy and time lapse imaging, and the one thing that you begin to see when you’re looking at biological systems is if you take a single frame, which is what the biopsy is oftentimes misleading to what the full story is and it’s just a limitation of the system. Even if I was a patient, I don’t know how soon I’d be willing to do my next biopsy if I just got one today. So there’s the complication of that.
Suneil Hosmane (31:09): That’s where blood-based biomarker technology, imaging-based biomarker technology has value and they provide differential value, and they provide differential parts of the story. I think the way that we’ll eventually move forward is a combination of those two. Now the specifics are what’s being figured out today, and as more and more data emerges I think we’ll have a better sense of sequence order, how to use this data. To a comment that Louise just said, when we talk about screening or identification, or these type of identification programs I think what really important is defining in more specificity who those people should be and giving that guidance.
Suneil Hosmane (31:48): I think that there’s knowledge there, there’s things that we know. We know about the impact and the association of obesity and type 2 diabetes, for example. But giving a little bit more color so that clinicians have the tools to better identify who should be further evaluated today versus tomorrow, and as that data grows, we can expand, and we can make a more sophisticated process. But I didn’t want to get into any specific technologies, but I just wanted to provide some perspective on where I think some of the gaps and challenges are, and at least some of the things I think about being on the development side of diagnostics as well.
Roger Green (32:23): Thanks everybody for an array of fantastically interesting and diverse answers. I have a couple of observations, and then what I’d like to do is I’d like to invite each of us to throw out a question to the others, and then try to see if you can keep your answers to two minutes per answer.
Roger Green (32:39): Thought number one, this becomes complicated from say a payer point of view in the US, maybe anywhere else in the world as well. You’re talking about a disease where progression of fibrosis stage might be somewhere between three and eight percent of patients a year depending upon the level of fibrosis. So if you treat everybody at F2, for example, the odds that any one of those people progress to F3 is fairly low, it’s really low. Without the ability to do what Suneil just described, which is can we figure out who at F2 is likely at a point where treating them will yield a more immediate benefit. If we go back to where Louise was, and in fact some ways where I think Suneil was as well, one of the things they teach you in organizational behavior is that people do what’s inspected not what’s expected.
Roger Green (33:24): So because we know how to measure diabetes, because we know how to measure cholesterol, because we know how to measure blood pressure, it’s easier to look at those things. It’s harder to figure out how to measure something complex like the liver, but the liver might be the organ that underlies a lot of those issues. So if you could help the liver function better, that might have an effect on all that; on glycohemoglobin, blood pressure, cholesterol. Now we’re starting to think antiviral response on COVID based on some data that we’ve seen that goes only to the level of fatty liver disease.
Roger Green (33:50): So we don’t know how to really measure the overall performance of the liver, and that makes it hard for payers who are accustomed to taking a look and saying, “Well, here’s how much it cost me to get to this metric and what the outcome benefit of doing that is.” Very hard for them to figure out how to reimburse, and I think in a world where dollars are likely to be scarcer for the next decade than they’ve been for the last 15 years, that’s something we’re going to need to wrap our arms around.
Roger Green (34:14): Let’s go in reverse order now. Suneil, question you want to ask any member of the group about something that they were talking about that you’d like an answer to? We’ll see if we can get one answer per person, that would be good.
Suneil Hosmane (34:24): I think I’ll pose this to everyone from a pragmatic standpoint, and I know we have a perspective from the patient, we have a perspective from both a physician and a member in drug development. We’ll look at HbA1c as an example. HbA1c is a pretty good marker, it’s by no means perfect, I think we really want to get into specifics there are much more [inaudible 00:34:44] techniques to assess insulin resistance and glycemic control, but it’s a pretty good marker, it’s easily doable, loads a lot of kits, and it’s a low price.
Suneil Hosmane (34:53): So we’ve taken a trade-off, we’re willing to sacrifice at least in that disease state some level of performance to have something that’s robust, that’s fellable. When it comes to NASH, and NASH and fibrosis or cirrhosis for that matter, how much are we willing to trade off in terms of clinical performance to have a test or tests that would be scalable and implementable from both a patient perspective, from a cost perspective, for example, and from a clinical perspective?
Roger Green (35:22): That’s a great one. Who wants to jump in first?
Louise Campbell (35:24): I’ll jump in first. Thank you, Suneil, that’s a great question. I think from my perspective, if we’re going forward it needs to be something that engages a patient to be interested and to find that they can achieve something. I alluded to Steven Harrison’s details of how you change liver fat, or your weight can change your liver fat and liver fibrosis quite markedly. I think we’ve got multiple studies in bariatric surgery that prove that. I think from my perspective, transient elastography or fiber scan does that, it’s also something that is repetitive; weekly, monthly, yearly, wherever, and it can be done in primary care, which is not where it’s where primarily located at the moment making it quite an expensive test.
Louise Campbell (36:06): But I think patients engage with it very, very well, and I think that’s unlike biopsy or any of the paradigms. It’s not perfect, and it needs to be used with other devices but it’s a guide, but it is what the most powerful tool that patients engage with. Whether or not you’re talking about reducing steatosis in alcohol withdrawal, and it’s ability to cross multiple specialties from endocrinology to cardiology to rheumatology because underlying fatty liver I think makes it patient-friendly. I think going forward, whatever we choose has to be patient-friendly to guide us to more accurate diagnostics.
Roger Green (36:44): Thanks, Louise. That’s a great answer. Peter?
Peter Traber (36:46): Louise makes a profound point here, and that is that both the physician, the community, and the individual patient need to be engaged in the testing that’s done. To answer Suneil directly, I believe that we do have to make a compromise between the perfect and the good when it comes to thinking about diagnostic tests that are going to be clinically useful. I think that if the ideal test is going to give a patient and physician some goals to shoot for and I think that for instance, hemoglobin A1c, what Suneil used as an example, patients can shoot for better glycemic control and their hemoglobin A1c comes down.
Peter Traber (37:28): The example of using ultrasound elastography is a good one because you can measure fat through the CAP score, you can measure stiffness, and you can see those things go down, and they’re easily explainable to the patient as to what’s going on there. The same thing could be for a serum test for NASH, whether it’s a combination test or a single test, but I do think that compromises are going to need to be made.
Peter Traber (37:54): I would make one last point, and that is we are already compromise dramatically in liver disease by calling liver biopsy a gold standard. I have looked at testing for many years in liver disease, and the AROC never goes over a particular point because we’re comparing it to liver biopsy, which has a poor reliability in terms of diagnosis. So we get to a certain level of diagnostic reliability with a non-invasive test, and then we hit a brick wall because we’re comparing it to an imperfect standard. So therefore, we are going to have to make concessions using the biologically and clinically most focused test to induce changes in our patients and changes in physician practice.
Roger Green (38:43): Thanks, Peter. That’s also I think really an excellent point and a different perspective, let me try a third. There are debates about exactly what cholesterol metric is best, but that debate centers around outcomes. We look at blood pressure levels that centers around outcomes. One of the real problems in liver disease, in addition to the one Peter mentioned, which is you reference everything against a bad test you can’t really get a great improvement. It’s that we don’t really know where the outcomes are in this.
Roger Green (39:07): So what we settle for instead, and I think Louise and Peter both got this right, which is if we can educate the patient how to do better and the patient then does better, that’s a good thing. But until we can look at a patient with some confidence, and by the way on the other side look at a payer and the doctor with some confidence, and say this measure will enable us to predict this improvement in overall health, mortality, morbidities, whatever it is with this level of confidence we’re going to have a hard time holding the standard up; and given the complexity of the liver, that’s going to be a hard thing to do.
Roger Green (39:37): I think at the intersection of what can a patient engage and what can we prove as an outcome is where we’re going to need to look, and at that point we’ll make whatever compromises I think we need to make in order to have that standard. Suneil, do you want to answer your own question before we go on to the next one?
Suneil Hosmane (39:50): Sure. I agree with comments by everyone. I do think that the engagement with the patient, it’s really a teaching moment. If for nothing else, I think it leads to an incredible opportunity to educate on both sides on the severity of disease and where that patient sits at that moment, and that can be elicited through many ways. I think stiffness and stiffness stage techniques will absolutely have their place, that coupled with some better understanding of biology in terms of some serum markers, I think that adds a different independent color. Whenever you’re doing science, you want to triangulate and achieve the same meaning, but through different mechanisms, so I think that’s another piece of it.
Suneil Hosmane (40:28): Roger, to what you said, I think the turning point for these diagnostics is actually very much linked to a comment that Peter made as well, there’s a fundamental limit as you start to compare against this reference standard. I think what’s going to really take everything over the top is once we begin to generate more outcomes data and being able to associate these markers with that so that you essentially have a correlation association with outcomes, and that will change where we are in terms of leveraging and using these non-invasive technologies both in development, but also in the clinic.
Peter Traber (41:00): Roger, can I make one more point here quickly?
Roger Green (41:03): Please, Peter, go ahead.
Peter Traber (41:04): I want to just emphasize a critical point that Louise brought up, and that is the enormous capacity of the liver to disguise when it’s hurting and how long it takes for the disease to actually cause problems for the patient. Now, this is not related to comorbidities like heart disease and other things, but related to liver-related disease per se.
Peter Traber (41:28): I used as an example IPF, well the survival in IPF is between three to five years after a diagnosis. The survival to liver-related outcomes in NASH is three to five decades. So we have a very different timeframe, and I think an important issue is that we don’t want to wait to the end when people are severely ill and have cirrhosis, as Louise mentioned. So it’s an added issue in talking about outcomes in a disease such as NASH.
Roger Green (42:00): I’d like to ask Suneil a question. We’ve had this discussion Suneil about the “gold standard” of liver biopsy and the fact that diagnostic accuracy is never going to get above a certain level with non-invasive biomarker tests when comparing to liver biopsy. What do you think is the solution to this conundrum and how we can most effectively look at non-invasive tests, and how we can eventually get away from liver biopsy?
Suneil Hosmane (42:30): I think that’s a phenomenal question, probably something that keeps me up, not every night, I would say probably say every other, other night. But the way that we at least are thinking about it in a more pragmatic sense is looking at it and generating data against outcomes. Now outcomes come in many flavors, there’s hard clinical events which could lead to liver decompensation, complications with cirrhosis. But we wanted to say as [inaudible 00:42:57] after that which takes a tremendous amount of time and generate data for that, we wanted to start looking something a little bit more near term.
Suneil Hosmane (43:05): So in the case where a patient is in a non-cirrhotic state, fibrotic or not, that progression to cirrhosis we have trials that are conducted that enroll patients with F1, F2, and F3 fibrosis. Something that’s very interesting, we have a very small data set, but when we went back and looked at our phase two clinical trial we saw that there are patients that within one year progressed from F2 to F4. So now is that a variability of the biopsy? Perhaps. Is that real? Perhaps. But I think getting data and showing the prognostic ability to determine who these patients are that could move quickly, I can begin to generate the kind of data to support that transition, answering what you asked. So I think that’s the interim solution, I think the long term solution is the hard clinical event.
Roger Green (43:56): Great question, great answer. I want to add an observation about outcomes that I think what Suneil said is really important. If you take a look at where the money is in disease, the money in treating liver per the healthcare system starts with cirrhosis. Before cirrhosis, not a very expensive thing to treat per se. We’ll come back to how expensive it really is in a minute. Any ability to identify patients who will go quickly to cirrhosis will dramatically improve the ability of the system to identify who needs to get treated most urgently in an effort to provide better outcomes. That I think in and of itself will be, I agree with you Suneil, a huge step forward.
Roger Green (44:30): Louise, do you have a question?
Louise Campbell (44:32): I have a question, and then I’ve got a comment. With Suneil and Peter both talking about biopsies being the gold standard or certainly the standard we try and compare everything to for clinical outcomes, for trials, and for diagnosis. Do you think given the variability of interpretation by histopathologists that we will move to AI as being the reader of these slides in future to get a consistent approach throughout for the diagnosis?
Suneil Hosmane (45:00): It’s an amazing question. In fact, we’ve done a number of different meetings with companies that are focused in the space. We also ave had some internal development in this space as well. What I would say is that used a philosophical back, true AI is not here yet. We’re moving in that direction. I think we’re still trying to understand the frameworking, which we want to conduct those type of studies.
Suneil Hosmane (45:24): But the general question is with enough data, big data, are we able to leverage that to have a much more precise reading of the liver biopsy? I think my answer is in the near term, maybe; but in the long term, yes. The reason I say maybe is because ultimately, these are machine learning type algorithms, so you feed a tremendous amount of data, and you annotate that data exceptionally well. You tell the machine what all the features are that you want it to learn on the slide. So things that you may not realize are important are then picked up by the machine, but also sometimes artifacts are picked up by the machine.
Suneil Hosmane (46:04): I’ll give you one example. So there’s been some AI analysis done in cellular research where they were trying to identify what would differentiate these two different kind of groups as cells, and a machine ended up picking up in very high precision. But when they actually digged down into the metrics, it was really picking up the type of glass or substrate that the cells were on because there were some very interesting features in the plastic or in the glass. That was the feature that it picked up on because it did it in a very unbiased way and that ended up being a core feature.
Suneil Hosmane (46:34): So my point being is it’s going to take a tremendous amount of annotated data and lot of other complications that we don’t have to necessarily go into on this call, but in order for you to feed and really get to that point, and I think that’s going to take a few years. So near term, the next one or two years, I would say more likely no than yes, three to five you getting in the right direction, in over five years I think we’ll be there. But that’s simply my opinion.
Roger Green (46:59): First of all, I want to thank everybody for a really fantastic wide-ranging discussion, interesting thoughts, creative questions, long term thinking. I think this has been a fantastic episode, and I’m delighted that we’ve been able to participate, and Suneil, thrilled you were able to join us. Let’s go to our last question now. What was the one thing you heard in the last section of this conversation, last half hour or so that surprised you, that you heard that you didn’t expect? I’ll let whoever has an answer jump in first on that.
Peter Traber (47:24): My surprising thought comes from Louise because I think she made a very profound point in how the patient connects to the diagnostic and how that may lead to behavioral change. I think that’s a perspective to put into diagnostic testing that is an important one.
Suneil Hosmane (47:43): I can say three. I echo the comment, Louise, regards to patient perspective. But the other two that came that made me think back was the analogy that Peter gave with regards to IPF. I think it’s good to periodically take yourself out of the space that you’re in and think about what’s happening in other disease areas, and have that influence your thinking; saying that was a very interesting comment. Then of course your comment, Roger, around the dynamic nature of disease, fast progression. These are things that I think about, sometimes they subside in terms of where they are in my thinking, and this helps bubble all of that back up to the top.
Roger Green (48:21): The thing that surprised me most that I hadn’t thought about really was Louise’s comment about the value in pushing this back into primary care as a way to help patients think about all the multiple implications of liver disease. That if it comes from the hepatologist, you’re not talking about heart disease, but if it comes from the PCP you might be. So part of getting the patient to understand the holistic role of the liver in the body is simply to have that discussion not come from hepatology later in the disease but from other specialists, maybe primary care, earlier in disease.
Roger Green (48:51): I think you said that, if you didn’t it’s how I interpreted what you said; like a smack on the side of the head, a good one. So thanks for that. Next thought.
Louise Campbell (48:58): I think it’s the fact that the overall level of agreement of how restricted we are on being able to look at liver function in the easiest formats that we have, which are blood tests that tell us very little. I think we’re all from four different backgrounds, and yet we all agree that we have an awful long way to go to try and fundamentally understand one of the most important organs in our body.
Roger Green (49:24): That’s a fantastic note on which to wrap up. Again, thanks everybody for a really far-ranging and fascinating discussion. I think we’ve covered as many points in this as we have in any of our previous episodes. So that ends Diagnosing Diagnostics part one. I want to thank our panelists, Peter, Louise, and our exceptional pinch hitter, Suneil. A special thanks to engineer and podcaster Frank Sasso that makes us sound so good as we discussed with Dr. Brown at the beginning, and to our social media master Eric Rounds. An extra special thanks to you our subscribers, and our listeners, who as I mentioned at the beginning have made us the fastest growing fatty liver podcast in the world.
Roger Green (49:55): We’ll be back next week talking about International NASH Day, key points around patient advocacy, and education. In two weeks we’ll come back to this subject once again. I expect Dr. Harris will be with us next, and we’ll figure out who else. Until then, everybody stay safe and surf on. Good day.
Speaker 5 (50:10): You’ve been listening to Surfing the NASH Tsunami. Send in your questions to surfingnash.com, and our panelists will spend the first five minutes of next week’s episode answering your questions. Visit us online today, surfingnash.com.