Thank you very much for this kind introduction. I have to admit, this year I almost considered skipping, after 14 years, the first time the ASID meeting to attend the celebration for the 30th anniversary of the fall of the Berlin Wall in my home city, Berlin. But then I received that invitation, I quickly changed my mind to talk about the disease very close to my heart. Forward, I guess, my disclosures. And I also will start with a talk, with a case. A 12-year-old boy with serology negative autoimmune hepatitis in remission, now with ALT elevation, GGT elevation, alkaline phosphatase elevation, and a low serum albumin. The repeat liver biopsy does not show interphase hepatitis, but instead it shows advanced fibrosis and increased bile duct profiles. I will go through what probably goes to at least every pediatrician's mind when you hear that story, that it's most likely an over-lip syndrome of autoimmune hepatitis and primary sclerosing cholangitis, coined autoimmune sclerosing cholangitis. And I highlight here what I will discuss throughout the talk, the difficulties in actually distinguishing these phenotypes of autoimmune liver disease in pediatrics. In this particular case, the next step would probably be an MRCP to evaluate for cholangiopathy. And as I will point out later on, once you establish that there is an over-lip syndrome, it is important to screen for IBD. And I will talk later about the sensitivities of the different screening modalities. And then, of course, in pediatrics, we always have to consider that we may be completely wrong. And that would not be the first patients who is diagnosed at a later stage with MDR-3 disease, PFIG-3, or even cystic fibrosis. So always be humble in your diagnostic accuracy. My talk, as mentioned, will review the phenotypic spectrum of autoimmune liver disease. We will discuss tools for precision medicine, which are sort of in the beginning in pediatrics, predictors for response of therapy, the diagnosis of biliary disease and pediatric autoimmune liver disease will be center stage of this talk. And then, in the end, I will rather briefly discuss recommendations for treatment of pediatric-onset autoimmune hepatitis and PSC. So to start off, I will show you a picture from one of the few publications showing population-based data in pediatric-onset autoimmune liver disease. Most of the studies are retrospective studies from tertiary referral centers and probably biased. In this cohort from two hospitals in Utah, you can see that autoimmune hepatitis and PSC, as expected, is a rare disease, especially if you compare it with IBD in pediatric patients. And you can see that the age of onset in autoimmune hepatitis and overlap between autoimmune hepatitis and PSC tends to be younger than in PSC. You have the male predominance in the PSC patients and the female predominance in autoimmune hepatitis. And you have association with IBD in those patients who have PSC and ASC. ASC in this cohort overlaps to 30% with PSC and to 21% with the autoimmune hepatitis cohort. So how to diagnose pediatric autoimmune hepatitis? Probably many of us use the simplified autoimmune hepatitis criteria to guide the diagnostic evaluation, which requires exclusion of viral hepatitis at the onset to detect autoantibodies, ANA, LKM, SLA, and to establish hypergamma-gallulaminemia, and then a liver biopsy, which is either compatible or typical if it does show interface hepatitis. But it's important to mention in this context that a recent meta-review in pediatric autoimmune liver disease patients actually showed that it has only a sensitivity of 77% to detect autoimmune hepatitis. So it is important to remember that the score was developed as a research tool, and it's probably justified to start treatment if you meet the criteria, and it is very helpful in avoiding delay of therapy, but it's not useful in excluding autoimmune hepatitis, and in my personal opinion, just repeat the serologies in patients because they may turn positive after the first initial presentation. Modified scores are currently being published, and these modified scores de-emphasize the hypergamma-gallulaminemia. They put more weight on the liver biopsies, and they also include the cholangiogram. But we need to wait for the next wave of validation studies to see whether that modified score will help us more than the current simplified criteria. So now I would like to talk about the phenotypic spectrum. Autoimmune hepatitis can present in various forms. I would like to distinguish the acute presentation from the more chronic indolent presentations. The acute presentation is usually associated with jaundice and various degrees of coagulopathy, and it's, of course, important to distinguish autoimmune hepatitis from other causes of acute liver failure, from other causes of acute severe hepatitis to consider that systemic lupus can occasionally present with acute hepatitis, and in the infant stage, giant cell hepatitis with autoimmune hemolytic anemia is an important cause for autoimmune hepatitis. On the other side, those patients who have asymptomatic elevation of AST and ALT, they often have... We have to consider evaluation for underlying PSC. We need to consider primary immunodeficiencies, and I will go through that. Unfortunately, because of time constraints, I cannot elaborate on the difficulties of diagnosing autoimmune hepatitis in patients with NASH. Positive autoantibodies, elevated IgG levels are commonly seen in patients with fatty liver disease, but what we do with that information is unknown. A recent publication from Italy compared acute versus non-acute presentation in 479 patients, including pediatric patients, and they compared those which presented with AST, ALT elevation more than 10 times the upper limit of normal or elevated bilirubin levels to those without these biochemical features, and the take-home message from this is, actually, despite the impressive differences in biochemistries, many features are actually very similar. Both groups had similar levels of elevation in autoantibodies, ANA, and both groups had similar stage of fibrosis, Escher stage 3, and both, importantly, responded to standard therapy with prednisone and azathioprine similarly. So if you do diagnose autoimmune hepatitis, you have to treat it, even if it occurs in acute presentation. So I mentioned that in pediatrics, it is important to consider primary immunodeficiency, especially in patients who have a very prominent presence of extrahepatic autoimmunity, and those who have a family history of autoimmune liver disease or other liver disease and who present with an early onset. Autoimmunity relates from two primary processes in fracture tolerance. One is the central tolerance, which is established by the thymus and deleting autoreactive T-cells, and patients with mutations in the gene AIR developed a well-described syndrome of autoimmune polyendocrinopathy, candidiasis, and ectodermodystrophy. That syndrome is usually associated with a type 2 autoimmune hepatitis, and these patients are at risk for adrenal insufficiency, Addison crisis, hyperparasitism. Anti-interferon antibodies are usually elevated and used on a research basis. The spectrum in patients who have a genetic predisposition for loss of peripheral tolerance is actually much more variable and probably more common. Those autoreactive T-cells, when they escape the deletion in the thymus, in the appropriate MHC-restricted background, can orchestrate a T-cell-mediated injury to the hepatocytes, and that would actually happen quite commonly and is probably unavoidable if you didn't have the population of regulatory T-cells expressing FOXP3, preventing those autoreactive immune-mediated hepatocyte injuries. Many genes have been implicated, but related to autoimmune hepatitis, the most common one is probably FOXP3, which predisposes to IPEX, an autoimmune disorder associated with type 1 diabetes, andropathy, polyendocrinopathy, and autoimmune hepatitis in 30% of the patients and infants. GATA2 was found in a 26-year-old adolescent presented to Birmingham with autoimmune hepatitis with a low frequency of regulatory T-cells. And I would like to just mention LRBA deficiency, a gene encoding a protein which is important for CTLF4 trafficking. Patients with that mutation have a low expression level of CTLF4 on their regulatory T-cells, and they are prone to develop common variable immunodeficiency with multiple immune-mediated injuries, which usually develop autoimmune hepatitis in their teenage years. And perhaps important for this audience is that in that particular disease, CTLF4 fusion protein, abetasub, was very effective in controlling all presentations of autoimmunity, including the autoimmune hepatitis, and it may be a tool for refractory autoimmune hepatitis in the future. Let me finish the discussion on primary immunodeficiency with one example of my own patients. I think seven years ago, one of the two siblings, the boy, presented to me with thrombocytopenia, splenomegaly, and elevated LFTs. And on the basis of elevated IgG levels and anti-LC1 antibodies and a biopsy which showed autoimmune hepatitis features, but also cirrhosis, we diagnosed that patient with type 2 autoimmune hepatitis. He responded very well to prednisone and azathioprine, but then over time, similar to our initial case, the GGT started to rise, and most recently an MRCP showed overlap with PSC. His sister presented at the same time. She was 9 years of age at that time, and she had low platelets and elevated LFTs, similar serology pattern with elevated anti-LC1 antibodies, and the liver biopsy was slightly different from what we typically see in teenagers with autoimmune hepatitis that had prominent giant cell hepatitis. But she also responded nicely to prednisone and azathioprine, but then developed a very different flair of extra-hepatic autoimmunity compared with her brother. She developed arthritis and genital ulcers. So I did what every pediatrician would probably do if you have siblings with the same disease. Many years ago, I did whole-exome sequencing, and it was negative. And then I did what only pediatricians do who attend the postgraduate course for many years. I did the same thing again. I did another whole-exome sequencing using new technology, new analysis pipelines. And this time, we found a deletion of the entire coding region in TNF-alpha-inhibitory protein 3, heterozygous, which encodes A20, an inhibitor of NF-kappa-B and a negative regulator of NLRP3 in flammasome, which has been associated with the autosomal-dominant form of batched-like autoinflammatory syndrome. And only if you present at this prestigious annual meeting, you have the luck that just a few weeks before the presentation, a paper is published from Japan showing that in a cohort of 360 patients with autoimmune hepatitis in Japan, that particular mutation in that gene was associated with cirrhosis at presentation in their cohort. So I hope I showed you that identification of monogenetic disorders predisposing to autoimmune liver disease are important, and they give us insight into mechanism-driving autoimmunity and perhaps novel therapeutic targets for autoimmune liver disease. But we cannot only do genetics. What are the more common predictors of response to therapy? There's a scarcity of literature in pediatric autoimmune liver disease, so we performed a retrospective review on our own patients. And I plotted there the AST levels over time, and you can see that about 61% of our patients with autoimmune hepatitis treated with prednisone, budesonide, and azathioprine responded within the first six months and normalized their AST-ALT levels, but 25% normalized initially but then relapsed. And the other 25% actually normalized at all and never went into remission. And we found in our cohort the predictors for nonremission to be a presence of cholangiopathy at the initial MRCP and higher IgG levels, emphasizing again how important it is to evaluate our patients with pediatric autoimmune liver disease for cholangiopathy because you certainly want to avoid over-immunosuppressing patients on the basis of elevated liver enzymes which may not be responsive to immunosuppressants. And predictors for relapse were, interestingly, at lower levels of AST. Ethnicity has been described very well in adults with autoimmune hepatitis, but so far there has only been one study which came from the group in Atlanta. They serve a larger community of African-American patients, and they compared African-American patients with non-African-American. They have a high incidence of end-stage liver disease at diagnosis, 28% and 75%. 80% in both groups were started on steroids, but importantly, 71% of the African-American remained on high-dose steroids at one year of therapy versus 57% in the other groups, suggesting that their response to steroid therapy is lower. The transplant-free survival was borderline lower in African-Americans than in non-African-Americans, and they had a significantly higher risk for recurrence of autoimmune hepatitis following liver transplant. Obviously, any of these predictors require validation and multicenter study to truly guide us in our individualized therapy in the future. So, I couldn't speak about autoimmune hepatitis and overlap without mentioning one of the most important studies in that field, which came from the King's College group and led by Dr. Miele-Vergani, in which they studied prospectively 55 patients with suspected autoimmune hepatitis with a cholangiogram, liver biopsy, and sigmoidoscopy, and they found an overlap with ASC based off an abnormal cholangiogram in half of their patients, and they compared those with overlap compared with those autoimmune hepatitis without overlap. Of course, they found an increased incidence of IBD in ASC patients. Interestingly, GDT and alkaline phosphatase were not useful in distinguishing the diseases. MRCPs are obviously important to establish the diagnosis, but it's not easy. Simon Lamb from our group reported yesterday and posted a form of blinded review of research MRCPs in a cohort of ASC-PSC patients versus autoimmune hepatitis. There was low inter-rater reliability in a blinded review of these MRCPs, but multiple focused strictures and enlarged portal lymph node stood out as good predictors for overlap with PSC, and there are now emerging technologies to quantify the strictures. We collaborated with Perspectum Diagnostics, and they did an automated quantification of the strictures and dilatations, and you can see two representative pictures of these cholangiograms, and in that study, we found that length of dilatation and the maximum diameter of the left hepatic bile duct had an AOC of 0.92 in distinguishing ASC from autoimmune hepatitis. Are there circulating biomarkers? As we mentioned, GDT and alkaline phosphatase are not very promising. We measured MMP7, a cohort of PSC patients, and found MMP7, which was found to diagnose biliatresia, a neonatal cholestasis, to be elevated in ASC-PSC patients. It had a higher AOC in diagnosing PSC-ASC compared with GGT or alkaline phosphatase. It is expressed by the cholangiocytes, the injured cholangiocytes, and we found that there was a nice correlation between MMP7 gene expression levels in the liver and serum levels. It was associated and correlated with a number of strictures and with the stiffness by elastography. So that may be an emerging biomarker to monitor our patients in the future. I also mentioned before how important it is to diagnose IBD, and a group from Toronto showed that in patients with PSC-IBD, they may be asymptomatic with a PUCAI of less than 10 and still develop endoscopic evidence of IBD. So the recommendation is to screen for IBD and with endoscopy, or at least with Fecal Carprotectin, but not use symptom markers. So am I already? Time is up. I guess we have to refer to you guys' memory of Mark De Nooth's presentation on the natural history, and I have to acknowledge that... Oops. Can we go back to at least my... Maybe not. As I showed you, that we have evidence in our pediatric patients for monogenetic causes. The simplified score is not sufficient, and we need multicenter collaboration to validate predictors.

autoimmune hepatitis

pediatrics

phenotypes

non-alcoholic fatty liver disease

genetic predispositions

precision medicine