Approaches to Biliary Atresia Care at Cincinnati Children’s - Alex Miethke, MD - Advances in establishing an early diagnosis

So, I have the very distinct privilege of introducing a, a person I've known for a long, long time who is a Cincinnati lifer. He did his GI residency, uh, I'm sorry, his residency, his GI fellowship, and his hepatology fellowship all here at Cincinnati Children's, Doctor Alex Smithy, uh, who has been the medical director of, uh, the liver transplant program here in Cincinnati for the last 4 years and is the director of the Center for Autoimmune Liver Disease. Uh, Alex has a basic science lab effort where he studies the mechanism, uh, the mechanisms of immune cell mediate cholestatic and inflammatory liver disease, including BA. Um, Alex is also active in clinical multi-center studies pertaining to pediatric liver disease, and today he's going to be speaking about advances in, in establishing an early diagnosis in BA. Alex, thank you so much for that very kind introduction. Uh, you know, um, oh, let me see. Um, these are my disclosures. Um, actually, they are related to, um, uh, uh, treatments and beatresia, uh, IBET inhibitors, so we may actually, uh, touch upon that later in the discussion, but not during this talk. Um, and, you know, I, I, I take the transition actually from Aki's talk about the genetics. Um, the genetics are super important. I'm glad that it's brought up. I think it's a super exciting topic, but it should not delay making the diagnosis. So this talk here is all about making a timely diagnosis of biliotresia, and I will spend the first half of, uh, my talk about showing why it is so important to make it in a timely fashion. And I would argue that genetics are not helpful in making a timely diagnosis of bili atresia. Um, they are very important for, for other aspects. Uh, and, uh, why that timely diagnosis is actually a rationale for screening efforts, which I also will, uh, touch upon a little bit. Um, I will review, uh, how we commonly, uh, uh, uh, work up neonatal cholestasis. This is a very educated audience, so we are not going to spend a lot of time on that. Um, then I review a little bit on diagnostic tests and I will focus on, uh, MMP 7 because we believe in Cincinnati here that, that has been, um, Helped us to make the diagnosis earlier and improve our outcomes. Um, so AI already, uh, presented, of course, very nicely about, uh, the etiologic factors, uh, and the relevance of biliaresia here. I just want to highlight the, the difference in prevalence. It is a, uh, by far more common condition in the Far East and I have the, uh, incidents in Taiwan compared with North America where it happens in 1 in 20,000. Uh, we have, uh, um, Billy and Treasure, of course, the, the reason to do the timely diagnosis is, uh, to, um, Do the CSI procedure as early as possible and there's so many surgeons here on the panel, I won't go into the details of the surgery. Suffice to say that in the end, we want to re-establish a bile drainage by uh um uh uh uh establishing uh a hepatoportoenderostomy, uh, which you see on the very right. And that surgery, of course, was, uh, first established by Doctor Kasai in Japan, and this is one from the First papers in 1968. And as you can see in this, uh, very early table, he showed that out of the 24 patients who underwent that surgery, uh, 5, he called as cured, um, and you can see that those who were cured, um, they were, they underwent the surgery between 2 and 3 months of life. None of the ones who got the surgery at 4 to 5 months of age survived the surgery. Um, that was in the pre-transplant era. Uh, and then I look back at our own, uh, transplant outcomes. Uh, we have performed, uh, I have a hard time actually seeing my slide here, but I think we have about 53 patients, uh, who underwent, um, biliard treasure, uh, uh, uh, transplant surgery for a biliardresia, and 42% of them were unpalliated. And I, I find that scandalous that In the 2000s in North America, there's such a high percentage who were diagnosed at the time, uh, which did not allow them to undergo palliative surgery and have the chance of avoiding transplant. So what is the overall outcome in North America? Um, Uh, Alex Bordik, uh, mentioned that, um, I'm leading now here efforts in our multi-center study. We are part of the Childhood Liver Disease Research Network. You can see, uh, uh, 12 centers participating. The, the Blue Flag is a, uh, um, data coordinating. Center in Michigan, uh, one of the red flags right below it in the middle of the country, that's Cincinnati, that's us. If you want to send us research fellows to do research, that's where they would end up, not on the coast, not the West Coast or the East Coast. But in that, um, multi-center study. Uh, as you can see, 136 patients were enrolled. Those who didn't undergo surgery, the 23, they did not, uh, survive without a transplant beyond 1 year of life. Uh, out of the ones who did a drainage procedure, 46%, um, survived the 1st 2 years without death or transplant. Uh, and in that, uh, larger multi-center prospective study, there were several, several factors which were predictive of, um, survival with the native liver for the first, uh, two years of life. Uh, uh, the surgeons, of course, know that the type of the billiard treasure. It is very important. Um, presence of ascites is a, a, a bad predictor because it is, uh, associated with advanced portal hypertension. Biatrisia splenic malformation is a predictor of a worse outcome, but in this study, it also showed that, uh, undergoing the surgery, uh, at age over 75 days of life had a 1.73 hazard ratio of, uh, requiring a transplant within the first two years of life. I think the best study showing the age dependence in outcomes after CSA comes from France. There they did a retrospective study of 740 patients between 1986 and 2002. Um, it's a great study. It has a horrible graph. You can barely see the colors. What you need to see is that the first, uh, line in white are the patients who underwent the CASai at less than 30 days of life, and, uh, at gray, that's the one between 31 and 45 days of life. And you can see that these two age groups, they had a 50% chance of surviving the 1st 12 years after CASAi. And their outcomes are much better than Um, the ones older between 75 and 90 days, or of course, the ones, uh, in orange, um, not undergoing any surgery. So that study in my mind, it supports the notion that the targeted age for Kasai should be less than 45 days of life, and this, it should, that should direct our efforts in establishing, uh, a diagnosis. Uh. Importantly, here, um, how you can make the best diagnosis and that previous paper I just showed, um, nicely developed the rationale. The only way of, uh, uh, reliably, uh, diagnose biliardresia early is through screening processes and, um, the screening, the first efforts came from Taiwan, like so many things, uh, in Villardtrisia research, and they use the stool card, uh, and they could show that they could decrease the age of diagnosis. Um, through these efforts. In, in North America, one of the, um, uh, best studies, uh, uh, came from the Baylor Group, uh, in Texas, and they, uh, implemented screening with, uh, of conjugated bilirubin. 120,000 infants from 14 Texas, Texas hospitals were enrolled into this trial between 2000. 15 and 2018 and the uh direct bilirubin was measured uh within the 1st 60 hours of life and then, uh, those who had an elevated bilirubin, they underwent a confirmation at two weeks at the Welch check and they identified seven patients with biliattrisia uh through the screening efforts. And you can see uh that post-implementation, um, the age of referral to a specialist was greatly reduced, um. Uh, compared with pre-screening efforts. A recent meta-analysis compared, uh, the, uh, the effects of these screening efforts. The stool, uh, card, which was used in several Asian countries, Taiwan, China, and Japan. And you can see that uh these efforts reduced the average age at CASA from about 60 to 7 days age to less than 60 days. And in several of these studies, it was a significant difference between pre and post-implementation. Um, and then the group from um Texas showed that it, it reduced the age from 56 per site to uh 36 days. So, a highly significant and very efficacious, but that's the uh me review already identified that. Uh, while it, it is effective, it is still a very expensive way of screening for biliarres, so more research is needed. So now I would like to just review briefly of how we're making the diagnosis of uh biliotresia in patients who present with neonatal cholestasis. So no screening effort but diagnostic effort, uh, of course, you know that the most important part is the differentiating cholestatic disease from physiologic jaundice. And here, I cannot uh overemphasize the efforts by our own society, NASA, uh, Um, emphasizing that any direct bilirubin of more than 1 mg per deciliter, not, uh, more than 20% of total, not more than 2, not more than 1.5, it is more than 1 mg per deciliter, that should, um, uh, trigger a workup for cholestatic liver disease for biliotresia. Um, we, we need diagnostic tools which are sensitive, rapid, non-invasive, and ideally, the initial workup can be done by the primary care physician, by the, uh, pediatrician, and that is directly related to the outcome of surgery and the question whether that patient will need a transplant or can survive with the native liver for long term. Uh, there are many, uh, billiard treasure. There is not one test which gives us a diagnosis. It is something which requires a history and physical, laboratory values, imaging, and liver biopsy. Um, of course, you know that acholic stool is a major red flag. Um, I think you will also see that in many of these studies, uh, GGT, uh, levels are actually as predictive, uh, of billiard treasure as many of the other more sophisticated tests. So, uh, GGT, uh, is very important in the initial workup. Um, not so, uh, reliable in, in, I would say in, in our hands here, but, uh, other studies show that ultrasound can be very powerful, and I will go a little bit into detail. Smaller studies show that Um, they, uh, and it's hospital-dependent. They use HIDA scan, MRCPs, ERCP, or they go straight to interoperative cholangiogram. I show a picture of the liver biopsy, of course, classic features with ductive proliferation stromal edema. But the liver biopsy, while, uh, highly specific and sensitive, it is an invasive test and we need other tests, uh, prior to that. This is a study from Iran, enrolling 64 patients of which, um, uh, um. Uh, what, um, there's a, uh, of which, you know, eventually, half of them were shown to have biliotresia. The other ones had intrahappening in neonatal cholestasis. They all underwent the same, um, diagnostic tests, and they showed the sensitivity and specific. of diagnosing bili atresia. Uh, and as you can see there, um, the sensitivity and specific ranges between 40 and 80% for clinical evaluation of liver enzymes, ultrasound is only 50% sensitive. That actually matches up my own experience. Uh, HIDA scan is sensitive but not specific, and liver biopsies is both, but it is invasive. So, what are the non-invasive tests? So there was a nice paper in Nature communication, showing Uh, a Chinese effort to use, uh, AI, deep learning tools, um, and they basically had, uh, radiologists take screen pictures from their sonograms, upload them in an app on their smartphone, and send them for central review using AI technology. Uh, uh, and that, that system, the AI was built with a large data set using um Ultrasound images from 330 patients with biliotresia and 800 patients without biliotresia. Uh, that was a discovery cohort and in a prospective trial, uh, that was validated in 102 patients with biliotresia and 196 without biliatresia, and the outcomes are phenomenal. They had an area under the receiver operating curve of uh 0.95 for the AI model to diagnose biliardresia using those smartphone, um. Ultrasound pictures essentially of the gallbladder. Uh, and that was better than what the experts could do in this study. Uh, I want to focus on our efforts here to establish, uh, MMP 7 as a diagnostic test. That goes back to one of the very first, uh, transcriptomic studies led by Georgie Bizera doing, um, Affymetric gene chip on liver tissue. And one of the first hits was actually MMP 7 in the study 20 years ago. Um, and MMP 7 is a, uh, matrix metaloproteinase which is, uh, important in degrading macromolecules, so it is directly related to fibrosis. And, uh, uh, 15 years later, uh, Georgie and his team received, uh, serum samples from the Childhood Liver Disease Research Network from patients with biliotresia and from those with, uh, intrahepatic cholestasis, all obtained, uh, at an age about 2 months of life when, uh, bilia. diagnosis occurs, and he used the SOMA scan, which simultaneously, um, identifies 1200 proteins. And out of this, uh, SOMAScan, proteomics approach, MMP7 and ENPP7, they turned out to be, uh, the most sensitive for diagnosing it. Um, here's a discovery cohort, how the, uh, how MMP 7 distinguish, um, from, uh, indiabetic cholestasis the healthy controls. MMP 7 is still sort of an emerging field, so it's important to look at the data, uh, critically, um, and, uh, you can see that, um, Uh, the intrahepatic cholestasis group was primarily idiopathic cholestasis, not allergy and not alpha-1 antitrypsin deficiency. The validation cohorts, uh, this is a large study from China, showed a high sensitivity and specificity, but again, the, um, comparison was primarily CMV and idiopathic cholestasis, not, uh, um, Eliel or uh um uh in um alpha-1 antitrypsin deficiency. We see a further validation uh data here, high AOC for MMP 7 and GGT not too bad. Uh, I, I have a table here comparing the studies from Taiwan, um, validation, uh, of MMP 7, again, high sensitivity and specificity, and we may discuss, um, afterwards why those studies from UK and Iran, uh, but I want to mention them, they did Not show a high sensitivity and specificity. So very important to see, uh, who is doing the test in which context it is. Uh, and we are, we are, uh, we continue to learn from it. MMP 7 was also recently shown to be a predictor of outcomes. This is a, a study from Taiwan, and there, they measured MMP 7 at 6 months after CASAi, and it nicely separated those who, uh, went on to require transplant versus those who, uh, survived, uh, for the 1st 2 years. Uh, And MMP 7 level of more than 10 was predictive, um, of, uh, poor outcomes. So that's how I integrate, uh, MMP 7 in my, uh, workup. Um, unfortunately, we currently don't do, uh, newborn screening to make a diagnosis early, but we certainly advocate for fractionating the bilirubin and we use a cutoff of, of direct bilirubin of more than 1 mg per deciliter to distinguish cholestasis from physiologic jaundice. Uh, and then, we integrate MMP 7 in the first line workup for, uh, cholestasis in the outpatient setup. Uh, together with NFTs with GGT and CMV is one of the most important ones to rule out. Um, and, uh, we still confirm diagnosis, um, of biliotresia with liver biopsy. And the goal is to do all that within the 1st 35 days of life so that, uh, the CASA can be done within the 1st 45 days. Uh, thank you for your attention. I look forward to the question. All right, Alex. Thank you so much for another wonderful talk, um, summarizing quite a bit of, um, data. And we, just like Aki, we have a poll question to share with the gang. Um, but as we are going through that, a question for you is, do you think we'll ever get to a point where Um, we won't need any invasive testing before we get to intraoperative changiogram and Kasai. You know, the. For us, as you've shown everyone, we do a liver biopsy, which is obviously an anesthetized procedure for an infant. For example, the Toronto Protocol, um, that they've published, they do an IR guided cholangiogram before CAII. So that's another sedated procedure. Do you think we'd have achieved enough sensitivity with the, all the things you've talked about or even adding Opry score to that workup to get the sensitivity closest, you know, as close to 100% so that You're going to the OR for just a cholangiogram and a casta, and that's the only sedated procedure we'll have. I think um. Uh, I think we are probably pretty close with MMP 7. I don't think that we have quite dared to make that jump, but I think we can, uh, uh, in the combination of using GGT, using Uh, likelihood, uh, knowing the family history, having knowledge about, uh, is there a high likely of, uh, allergy syndrome, um, uh, I think, uh, with the MMP 7 level, um, go skip the liver biopsy and go to, um, an intra intraoperative. io gra m I think it is, it is, uh, a possibility. I, I know you and Greg, uh, uh, you always tell us that when, once you do, uh, a, uh, a Kasai at age 30 days of life, these are pretty small children. So, having, having enough workup done. Uh, upfront and increasing the pre-test probability of what that intraoperative cholangiogram is going to show, that is useful, but I think, uh, the combination of non-invasive tests is as good as a liver biopsy at this point. Awesome. Um, so we have a nice robust response to your poll question just to repeat, it's based on a large retrospective study from France. The optimal age for Kasai portoenterostomy is um A, between 45 and 75 days, B, between 75 and 90 days, C, less than 45 days, and B, there is no age dependency. And it looks like we have uh 80% uh responding less than 45 days of life. Alex, what do you think? I, I'm surprised that uh there are not more who called it D, so there must not be a lot of surgeons on that webinar. I thought, I thought your, your personal fight would trump whatever you learned or didn't learn from that talk. So, so it is, it must be all hepatologists and gastroenterologists. It is interesting. That is interesting that there's not more D. So I, yeah, go ahead, Alice. I'm sorry. I was just gonna read uh another question. Did you wanna comment on that? No. All right. So, um, Try Meyer says, um, given the previous discussion about the different etiologies and clinical manifestations of biliary atresia, is there data that MMP 7 values differ in these different groups? Could that explain the variation in specificity and sensitivity in different regions? I think it's a, I, I think it's an excellent point. So far, we haven't really seen that. It, I think it's, it's surprisingly homogeneous across the biliard region the splenic malformation, across the, the, the, the genetics, across whether it's a virus associated. I do think that, um, Uh, what is the composition of, uh, non-biliarresia intrahepatic cholestasis that, that may have an impact if you, for instance, see the UK study, uh, which had a very poor sensitivity. They had a lot of patients with alpha-1 antitrypsin deficiency, and we know that that disease, for instance, can. Fool us even with the liver biopsy. It can cause bile plaques and others. So I think if you have a large, and I think that's, that is important to realize that MMP 7 test is for the outpatient setup to differentiate um villiardrea from the large group of idiopathic neonatal cholestasis. As the patients who actually just get up, get better eventually without um any chronic liver disease. It's not really validate to, to separate from allergic syndrome and so you, you still have to do your little bit of workup, but I think it, it, it actually, uh, most of the time, it turns out to, to, to separate. Um, in the NICU setup, the kids with uh TPN cholestasis, it separates quite well the allergy, it does it, but the, the, the studies which show the very high sensitivity and specificity, they separate biliard re for non-BA without these rare genetic causes and so forth. I think one other population to highlight is the congenital heart disease patients where it can also be a little confounding, where the level, levels tend to run high. And I think that's still where MMP 7 is a useful adjunct, but it's specificity and role is still evolving. It ties to screening. I think one of, uh, you highlighted on one of the, the, the figures you show where people are trying to develop it for, um, you know, blood spots, you know, in the neonatal period, which would make it, you know, that much more helpful in terms of, uh, bringing that patient to the attention of the primary care team. The pediatrician, and potentially the hepatologist. So those are all parts of the evolving effort to utilize something that came out of the lab. I think, um, uh, to, to mention the new, newborn screens because I do think that that will, uh, be a huge opportunity for us going forward. Uh. Um, the direct bilirubin, of course, I mean, it's, it's already great. It, it reduced from 120,000 to 1200 who screened positive in the first test. But then they still had 120 patients who tested positive on the second test, and they all needed a workup. So, MMP 7 could be way more specific for biliard lesion on the dry blood spot. Great. Um, let's see. Uh, Maria Shulkkoff had a question that she said that at their hospital, MMP 7 takes 3 weeks to turn over, but that hasn't necessarily been our experience. Have you, can any of the hepatologists comment on that? Um, I think has it been refined enough that it's Emily, I, I, I, I think the, the important thing is, yeah, I think I see Emily, uh, sent something in the chat on where to send it. It is only useful if it has a quick turnaround. If the MMP 7 doesn't work, this is all, this, this doesn't help to make ultimately the diagnosis of any, it's, it's to help to reduce the time for diagnosis. If there's no turnaround of 2 or 3 days, it's probably not worth using that vendor. And Emily added in the chat that um the lab at UT Southwestern in Dallas at Cook Children's is working to run these MMP 7 tests daily, so that you can return results the following day. Um, labs are working on establishing it in, in, in their own centers because it really is a, a, a, a study that can be done by any kind of, um, um, lab setting. It is clear certified, um, the, um, the, um, the one that was being done here, um, and now, you know, with Georgie leaving, he took it with him to UT Southwestern, and they are trying to turn it over more rapidly because without a very timely, uh, assessment, it, it doesn't. You know, improve the efficiency. But I, I would like to emphasize that, Greg, um, when you look at these studies, which didn't, uh, reproduce a high sensitivity and specificity, they all use different EISA tests. They, while it is theoretically possible, it is highly dependent on the right Eliza and how you do it. So, if you want to have a rare disease, which in most centers happens between 3 or 7 times a year. Uh, you, you want to make sure that the test works as well as it was published. And, uh, I think that is the primary reason why these others couldn't put it. They all didn't use different analyzers with, and you see the cutoffs for the MMP 7 across the different studies, uh, largely varies. So, uh, it's, it's really, it's really important to have enough cases to make robust cutoff for these measures. Um, another question for you, Alex, here as the Q&A period, um, kind of shortens up, but Doctor Troy Gibbons has a question. Can doing a liver biopsy too early add to diagnostic confusion? I, I, it's a great, I think the, the goal to do it less than 45 days and, and you see, in Texas, many of these kites were done at less than 30 days of life. The earlier you do it, the more shaky I would say, not that our surgeons are ever shaky, but it, it, it makes a lot of phone calls from the OR. Are we sure that this is what it is? And Um, while studies show 100% sensitivity and specificity for a liver biopsy, we have done two liver biopsies in several patients. Yeah, when you do it early, you can have an incomplete picture. And while these three features of ductal proliferation, bile plaques, and stromal edema, uh, they don't always occur. And the earlier you do it, you may not have the bile plaques, you may only have the stromal edema. So a liver biopsy also needs to be reviewed by an expert pathologist, and the earlier you do it, the harder it gets to separate that from other causes. So, I wouldn't say it's too early, but it, it gets harder and it's probably something you want to send out for consultation if there's any doubt. Well, and I know Doctor Sparina has also published a paper calling into question the utility of liver biopsy, sort of, sort of general preoperatively. So I think it's a completely valid question, especially when you, just like you guys were talking about the MMP 7, if you don't have the right person necessarily interpreting the test or running the test, it could be confusing. And I will say, when we've been fortunate and had these, you know, two early biliary atresia patients come to us who have had elevated direct bilirubin, negative alpha-1, acolic stools, an ultrasound that's consistent, and we're seeing them at 23 weeks of age, um, we've met with the family and we've planned a liver biopsy somewhere between 21 and 28 days of life, so around that 30-day mark. Um, so that we can aim for Casa shortly after that if the liver biopsy is consistent. So if we see somebody at a week of life, we're not biopsying necessarily the next week, but we do take some time to do the workup and, and have come to a sort of group consensus on how we manage these early patients, which is actually a, a good problem to have. I think one important point to, to highlight as this, as we wind down the session here though is the one diagnosis that we want to be most sensitive to. is not doing a eye on is allage patient. Cause if you do an alloys patient, you will, um, precipitate more, uh, a higher likelihood of the need for transplant. So this really is that the diagnosis that is, gives us the most angst because also when you're doing the cholangiogram on, uh, on a baby like that, if you're, if you are stuck in the OR, um, um, with a case like that, um, it can be very difficult to differentiate because the ducts are so hypoplastic. So they're in. is where you really have to part, the surgical side really has to partner with the, the, the, the medical side to try to be as sure as you can. Uh, the, the challenge for allage Gs, of course, is the genetic, um, workup, the testing for that takes several weeks. And so this all goes to the sooner that patient gets plugged, gets plugged into being evaluated, the more likely that you will to have be in the OR with less anxiety when you're going to do your procedure. OK. I think in the interest of time, we'll move on to our next speaker.