I would like to thank the organizers for inviting me. I feel really honored to be here. So, these are my disclosures. The case I will be discussing with you is a 32-year-old woman who comes to the hepatology outpatient clinic already with the results from genetic testing, which shows she has a heterozygous ABCB4 variant or mutation. Her complaints were that she has chronic abdominal pain and previous intermittent jaundice. She also experienced pruritus towards the end of both her pregnancies. And so, her primary care provider suggested that she underwent broad genetic testing and now here are the results. So, the question I would like to answer together with you are can a heterozygous ABCB4 variant explain this patient's symptoms? If yes, what do we need to look for in the future? Do we need to surveillance or offer monitoring and or treatment to this patient? And what is the potential differential diagnosis? So, are there other genes that could explain this phenotype? So, I will start with the first question, are there other genes? You're looking at two adjacent hepatocytes and ABCB4 is the gene that encodes the phospholipid floppase called also MDR3. So, this is the phospholipid transporter. We need this to form the mixed micelles to keep the cholesterol fluid. So, variants in the ABCB4 gene underlie intrahepatic cholestasis of pregnancy. It's the most common genetic cause of ICP. They may also lead to familial gallstone disease called ALPAC syndrome, which stands for low phospholipid associated cholelithiasis syndrome. It's quite easy to explain. You don't have enough phospholipids, so you have a supersaturation relatively of cholesterol and cholesterol gallstone formation. And it may also lead to what's called progressive familial intrahepatic cholestasis type 3. This is often early onset, so in infancy, progressive disease to end-stage liver disease, often necessitating transplant. But however, it can also present in adults. As the name already says, if you have pig type 3, you should have 1 and 2. So, there is actually more genes. So, ATP8B1 is the gene encoding FIG1, and it may also underlie ICP. It may underlie benign recurrent intrahepatic cholestasis type 1. This is intermittent cholestasis, often triggered by upper respiratory tract infections, and then the patient has episodes where he or she is completely normal until the next episode. Most patients do not progress to liver fibrosis and cirrhosis. And then, again, there is the very severe type of FIG type 1. ABCB11 is the gene that encodes the bile salt export pump, and this is the only bile acid transport in humans, transporting bile acids into bile. So, if you don't have BSEP, this is not really compatible with a normal life. This is FIG type 2, presents early in infancy, and often needs a transplant. But ABCB11 milder variants may underlie ICP, may predispose to drug-induced liver injury, and also BRIC type 2. And for a very long time, those three genes were the three cholestasis-associated genes. But over the last five years, novel genes have been described that may underlie very similar phenotypes. One is encoding tight junction protein 2. This can lead to BRIC and also FIG. We also have the gene encoding FXR, which is the main regulator of bile acetomyastasis and also BSEP transcription. It has been recognized for ICP and also for FIG. There's only five patients described. Two are transplanted and three dialed early. So, this is a very severe form. And myosin 5B, the pediatricians might notice from intestinal disease. There is also a subgroup of patients with only liver disease. They may present as BRIC and also as FIG. So, we have six genes now that can lead to a very similar phenotype, and there's probably more to come. So, how often are these disease phenotypes? So, if you go to a pediatric transplant unit, 10 to 15% of the patients presenting with neonatal cholestasis can be attributed to FIG type 1 to 3. In the adult patients, this is more difficult. So, ICP is found in up to 2% of pregnancies in Europe and the US, and the number is increasing with hormonal stimulation. So, just to give you a little bit more insight into the molecular function, ATP8B1 is a flippase. So, it helps to keep this lipid bilayer asymmetric for the lipids. This is called the liquid ordered state. We need this because those bile acids are very toxic, and they might dissolve the cholesterol from this membrane and then impair ABCB11 or BSAP function. So, if ATP8B1 is deficient, cholesterol is getting out of this membrane and BSAP is dysfunctional. So, it does mimic BSAP disease, and the same is true for myosin 5B because you need this to anchor BSAP towards the apical membrane, and FXR also mimics BSAP disease. ABCB4 we need for the phospholipids, for the mixed micelles. If you don't have this, again, you have toxic bile acids and biliary damage, and you have a cholesterol transporter. So, variants in those six genes leads to a very broad phenotypic picture ranging from only elevated liver enzymes to end-stage liver disease and liver cancers. If you look at this genetically, we find single allele variants on the milder end of the spectrum, missense variants, and common polymorphisms, whereas on the severe end, we see both alleles affected either by homozygous variants or compound heterozygous variants, and we see a lot of variants that lead to complete absence of the protein. However, there is exceptions. So, we looked at a very mixed cohort, both of pediatric patients, adult patients referred by the primary care provider, as well as specialists, and what we found is that in 10 to 35% of patients we could identify a disease-associated variant, and if you add those disease-associated variants with the common polymorphism, this was quite striking. We only had two patients which didn't carry either variant in ABCB11 and only 16 patients which had neither a disease-associated variant or a polymorphism in ABCB4. If you look at the age of presentation for the pediatric patients, this is very, very early for ATP8B1 and ABCB11, a little bit later for ABCB4, and about 30% of patients with ABCB4 present in adulthood. So, these are the patients we see as adult hepatologists, and this is exactly the mean age of our patient in this case. Now, looking at all six genes, so we added those three here, you see in the pediatric cohort you have homozygous, compound heterozygous variants, so the classical picture of an autosomal recessive disorder. Picture looks different if you look at the adult cohort. They are much older, and you don't see compound heterozygous or homozygous. You see one disease-associated variant, and you see more than one gene affected, and very often ABCB4 is affected. So, why is ABCB4 so interesting? There was this nice study from Iceland where they actually whole genome sequenced 2,600 individuals and then looked at the variants and the disease registry, and there's only one gene that came up with liver disease, and this was ABCB4, and then they re-sequenced the first and second degree relatives, and they found two missense variant, one frame shift, and one very common polymorphism that does not even change the amino acid sequence of the protein, but interestingly, what you see here, even the common variant is associated with elevated liver disease and predisposes to gallstone disease, and it's actually the wild-type allele that's the reason why those numbers here are negative. If you look at the missense variant and the disease-associated variants, you have an increase in gallstone disease, ICP, and it also predisposes to cirrhosis and liver cancers, and this is not only found in Iceland, but this is a study from Germany. Frank Lemers Group looked at another ABCB4 variant that is very common. One percent of the Europeans have it, and if you perform fibrous scan, patients with at least one allele of this variant have a higher fibrous scan, so again, this does predispose to fibrosis. Now, coming back to our case, she had abdominal pain, intermittent jaundice, and pruritus in pregnancy, so the most likely diagnosis is ABCB4-associated gallstone disease and ICP during pregnancy, so what is ALPAC defined as? So this is the gallstone disease. It's defined clinically by symptomatic cholelithiasis and at least one of the three following, symptom onset before the age of 40, this is true in our patient, recurrence after cholelithiasis, and on ultrasound, you find sludge or microlithiasis, and if you have this diagnosis clinically, 50% will carry a disease-associated variant in ABCB4, and they will respond well to urthodeoxycholic acid treatment, so what are we gonna tell our patient? So her heterozygous ABCB4 variant may predispose her to elevated liver enzymes, to gallstone formation early in life, to recurrent ICP if she becomes pregnant again, but also to biliary fibrosis, maybe even cirrhosis, and also higher risk of cholangiocarcinoma and HCC development, so what are we gonna offer her? We would recommend urthodeoxycholic acid because it helps in LPAC and PICC type 3. We don't have targeted therapies yet, and we would recommend surveillance once yearly if she doesn't have cirrhosis yet, to do blood, ultrasound, and elastography, and early monitoring for cancers. If she has a cirrhosis, we would do everything we do in our cirrhotics. So the difficulty for us as clinicians is if we order genetic testing or the patient comes with the results, how are we gonna interpret the findings? So what are we gonna do? The problem with the cholestasis-associated genes is you don't have the one hotspot variant. Here, for example, you see ABCB4, the variants we detected in Düsseldorf, and this is the protein. They are spread through the whole gene and the whole protein, so everybody has its own or her own variant. Only the very common ones are found in more than one family, so you're facing trouble because you don't know whether this variant is disease-causing or not. So what are you gonna do? You can go to the Internet, look in the literature. You can look in databases, see how common this variant is. Is it a rare one? You can use online prediction tools and protein modeling. In vitro analysis is not something for everyday clinic. This is for selected cases, but what you can do is use everything you have from your patient. Maybe you have liver tissue. You can go for staining. You can do bile acid profiles and use your clinical data. For example, this is a liver biopsy, severely cholestatic pregnant patient in the ninth week of pregnancy, and you see B-sub staining here is severely reduced, and she had a variant, and she also had an ABCB4 variant. So prediction tools, for example, for novel missense variants you see here. In many cases, different prediction tools come to the same result. However, you have variants like this one here and ABCB4 where one prediction tool says everything's fine and the other one says it's disease-causing. So what are you going to tell your patient? So in this case, as we went to protein modeling because we have colleagues who can do this for us, and this is this variant, and they told us this is near the substrate entry gate and therefore interferes with substrate translocation, so this could be disease-causing. How about variants on more than one gene? May they also explain the phenotype? Just to give you an example, there is a very common ABCB11 polymorphism called V448. It's found in 50% of the population of European descent, so I'm a heterozygous. It does not impact B-sub transport activity, but if you have it, you have lower B-sub amount in your liver. Under normal circumstances, this does not affect the patient. However, if you have high gestational hormones or take certain drugs, you may become cholestatic, and it aggravates the ABCB4 phenotype. So why ICP and why is this so interesting? Because those gestational hormones and their metabolites inhibit B-sub function from the bile side and also expression, and so if you're in a genetically predisposed individual, this may trigger that you are not sufficient to transport bile acids enough anymore and therefore become cholestatic. So usually they present late in pregnancy, earlier if they have twins, earlier if there is hormonal stimulation, and they present with pruritus and elevated bile acid levels. Why is this important to pick up? Because there is a risk for the fetus, especially if maternal bile acid levels exceed 40 micromolar. There is preterm delivery, respiratory distress, bile acids interfere with surfactant, and if bile acid levels are really high, there's also the risk of interuterine death, and this increases late in pregnancy. What about the mother? The long-term outcome for the mother, she has an increased risk of gallstone disease, fibrosis, HCC, and cholangiocarcinoma. So this sounds very familiar because it's the complete spectrum of the ABCB4-associated diseases. So just to give you an example, this lady presented in her third pregnancy. She already had ICP and severe liver disease in her first. She miscarried the second one, which is also quite typical, and she had bile acid levels up to 260, so really severe ICP. We found a homozygous ABCB11 variant. This is the common polymorphism, and a rare ABCB4 variant that is heterozygous. We put her on ERSO. However, she had a preterm delivery. The APGA was slow due to asphyxia. However, the child luckily recovered completely, and she completely normalized her liver function tests and serum bile acid levels within four weeks. But this shows you just what can happen in these patients, and a recent review showed that ERSO doesn't do anything for the fetal outcome, but it might help the mother. So what are we going to do? Recurrent ICP has a very high predictive value that you get a positive genetic test result. Severe variants, and if you have variants on more than one gene, are associated with early onset and severe disease cause. You have an increased risk for the fetus. This is the reason why we tell them better to deliver in a hospital with a pediatric ICU, and you should consider lifelong surveillance and maybe ERSO therapy for the mother if you do detect an ABCB4 variant. So how about treatment? This is really fast because there's not much to tell. We have liver transplant as the only curative treatment. We more or less try ERSO in all the patients. There's the biliary drainages for the children, and you might try rifampicin. But I think the future looks bright because there's quite some studies preclinical and also starting clinical, especially the ASBT inhibitors, look very good in the ABCB4 animal model, so hopefully there's something in the future. So how about genetic testing? When do we recommend it? We offer genetic testing to our patients with severe ICP, recurrent ICP, and early onset ICP. We also do it in cryptogenic liver cirrhosis if they don't have any signs of nephaldein NASH, especially if they have high gamma GT because it's quite typical of ABCB4 disease. In a positive family history, an exclusion of other causes of liver disease, and this is different in pediatrics. In pediatrics, you go for genetic analysis early because it might guide your treatment and it might also guide your donor selection if you go on to transplant, so different in adults and pediatrics. Always offer appropriate education and counseling for the family, and we don't really go for whole genome because you come back with a lot of variants which you then have to explain, so this is reserved when you have more than one generation affected or, again, if you look at a pediatric patient. So in summary, variants on different genes may predispose to adult cholestatic liver disease. Adult onset forms may present as BRIC, ICP, ALPAC, chronic cholestasis, biliary fibrosis, cirrhosis, but also with HCC and cholangiocarcinoma. The gene to look for, especially, is ABCB4. Consider long-term surveillance and even erythrodeoxycholic acid therapy if you have a disease-associated ABCB4 variant. Disease-associated variants and common polymorphisms on more than one gene may influence the disease cause and also the disease phenotype, so look for these. NGS panel analysis and next-generation sequencing allow a really rapid diagnosis, but a dilemma is the interpretation of the results which may prove very difficult in the adult patient. So what we need for future therapies is we need a better characterization of the molecular consequences and the mutation classes as a prerequisite, and I thank you for your attention. Thank you.

ABCB4 variant

liver conditions

genetic testing

cholestasis

ursodeoxycholic acid therapy