false
Catalog
2024 Webinar: One ALT is not like the others
One ALT is not like the Other
One ALT is not like the Other
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
ALT. Hello and welcome to today's webinar. One ALT is not like the other. My name is Lily Dara and I'm a hepatologist at the University of Southern California in Los Angeles and chair of the Hepatotoxicity Special Interest Group of ASLD. This activity is designed to define the clinically significant ALT measurement differences in the U.S. by ALT analyzers and assays. This activity will be offering CME credit as both a live activity and enduring material for one year. Financial disclosures for this program have been reviewed. Any potential conflicts of interest have been mitigated. This activity is designed to be interactive. In order to receive credits, we ask that you complete the evaluation which will include some self-assessment questions. The evaluation will be emailed to all learners tomorrow with instructions on how to complete your evaluation and receive your certificate. If you will require mock credits, you will need to submit your ABIM number and your birth month and day. If you do not submit this information, your mock credits will be delayed. Thank you for participating in today's webinar and we hope you find this information extremely useful. Okay, let's get started. First, on behalf of the American Association for the Study of Liver Diseases Hepatotoxicity Special Interest Group, I'd like to welcome our two speakers, Dr. Christine Hunt and Dr. John Toffoletti, both from Duke University School of Medicine. Dr. Hunt is a hepatologist and adjunct professor of medicine at Duke University. She has a long track record in basic and clinical hepatology research. Dr. Hunt has initiated and led national VA multidisciplinary teams to prevent hepatitis B reactivation and to advance mass clinical research and care. She has published widely, including 12 studies on ALT. Dr. John Toffoletti received a BS from the University of Florida in Gainesville, followed with his training in clinical chemistry at the University of North Carolina, Chapel Hill, where he earned a PhD in biochemistry, followed by a postdoctoral fellowship in clinical chemistry at Hartford Hospital. He has worked in the clinical laboratories at Duke University Medical Center since 1979, where he is now professor of pathology, director of the blood gas laboratory, the clinical pediatric laboratory, and several outpatient laboratories. He is also the chief of clinical chemistry at the Durham VA Medical Center. His research interests include sample collection, pre-analytical errors, analysis, and the use of these tests clinically. Welcome, John. Welcome, Chris. Please take it away. Thank you. Thank you for the kind introduction, Dr. Diera. I will be discussing ALT assays, analyzers, and studies, and then we'll turn it over to Dr. Toffoletti to discuss ALT reference range and some very helpful studies he's done to examine clinical studies to examine ALT reference range. My disclosures are shown. None apply to this talk. As many of you know, ALT increases as body mass index increases, and this has led many clinicians to be concerned that subjects with metabolic dysfunction associated steatotic liver disease, or MASL, may have been included in ALT laboratory reference range testing and increased ALT reference ranges. So this has led to the idea that we should identify a fixed ALT cutoff to identify healthy normal patients, but the story is not that simple. Let's see how this occurred and how it affects clinical care. We'll be discussing the clinical case, just presenting it now so you can mull over your recommendations as we share the information. In this, a primary care physician has referred Bill to you for a possible liver biopsy to identify MASL, because he's had a persistent ALT of 42 to 45 over the past six months. Bill is an overweight 42-year-old non-drinker with normal blood pressure who takes no medications or supplements. His ALT is 42 to 45, his AST is 31 to 34, both within the normal reference range, and he has a normal bilirubin, ALKFOS, albumin, and normal triglycerides, cholesterol, and glucose. His viral hepatitis serology and autoantibodies are negative, as are his liver disease markers. His FIB4 is normal, and his liver ultrasound is normal. So we will discuss following our, after our discussions of ALT, what you might recommend for Bill. Additional labs, liver transgenic elastography, liver biopsy, or none of the above. And you can mull over your choices, and we will just return to Bill after some other discussions. In most of the developed world, and much of the world overall, international standards for ALT are used in more than 100 countries. And in these assays, pyridoxal 5-phosphate is added to achieve consistent ALT measurements. For more than 20 years, the European Union has legislated these international standards to ensure ALT accuracy, and recognizing that non-equivalent labs have caused medical errors affecting patient outcomes. They wanted to avoid that with standardization and harmonization, ideally throughout the world. In contrast, U.S. and Canada, ALT assays are not standardized. They vary notably. And just to, Dr. Toffoletti will provide additional details, but just to emphasize that ALT assay is not like a red blood cell count or a protein count. It's actually a fluid dynamic activity. ALT transfers amino amino groups in the cytosol, and changes, because it's an enzyme, the assay changes with changes in temperature, pH, or the presence or absence of cofactors. So just keeping that in mind, that within U.S., FDA-approved ALT assays use multiple different methodologies, including spectrophotometric, chemiluminescent, and fluorescent. And U.S. ALT assays are FDA-approved with and without added pyridoxal 5-phosphate. In addition to marked variation assays, ALT analyzers vary, and even within the same analyzer, different assays are used to further complicate the issue. Analyzers can add, or assays can add or not include pyridoxal 5-phosphate. And there's five different analyzers used throughout U.S. and Canada, and these analyzers yield significantly different ALT measurements and reference ranges. So it's really quite a jungle of ALT assays and analyzers. Recognizing this issue, Canadian lab professionals collaborated on a national study where they sent pooled human blood samples to 40 labs across Canada and measured the results of the multiple assays in place across all five analyzers. And because it was lengthy, I just focused on a single analyzer, the Siemens analyzer. And this bar graph depicts the reference ranges. And as you can see, reference ranges vary from 10 to 40 to 10 to 80 across the single analyzer, revealing a two-fold difference across the Siemens analyzer with different assays in this Canadian analysis. This prompted the Canadian lab professionals to plan actually a harmonization of methods and then completing a large study in healthy control to harmonize the ALT reference range throughout Canada. Should this happen, when this happens, this will be helpful for U.S. Now, you've noticed that a local hospital consistently reports a lower ALT than what you find for the same patient at your medical center. And you were just curious, and you asked your laboratory pathologist, what's causing this ALT difference? Well, the pathologist responds that there's many different ways that you could see this difference. One of them is that pyridoxal 5-phosphate, or vitamin B6, is a coenzyme for alanine aminotransferase. And in patients who are nutritionally depleted, as in alcoholic hepatitis, as shown by Dr. Annamaydeel, or frailty, it results in a low ALT. When pyridoxal 5-phosphate is supplemented, ALT increases 38% in these patients. And Dr. Maria Tudor-Cresper, in roughly 50 healthy controls, found that adding pyridoxal 5-phosphate to an assay increased the ALT nearly 30%. So overall, pyridoxal 5-phosphate in assay increases ALT roughly 30%. So the pathologist shared that the pyridoxal 5-phosphate may not be included in the local hospital assay, resulting in lower and inconsistent values. And additionally, they point out that ALT analyzers and assays are not standardized across U.S. or Canada, and ALT measurements vary significantly, particularly within normal. So either of any of these factors could have influenced the difference in labs. So we're now up to a polling question to test your knowledge. Please select one response for the following prompt. Pyridoxal 5-phosphate is a coenzyme for ALT included in all ALT assays in U.S. to address... oh, here, I'll stop talking. It's presented... it should be appearing on your screen, and please fill in the response that you choose. Oh, that's good, that's good. So, people were very sharp. The correct answer, very impressive. Everybody, whoops, excuse me. Okay, I need to get rid of this, sorry. Close. Everybody got the answer right. Congratulations. Paradoxal 5-phosphate is a coenzyme for ALT, which is missing from many assays in US, yielding inconsistent results. I'm having trouble. Okay. So, just to… Canadian laboratory professionals are pursuing harmonization. However, US does not plan ALT standardization in the near future, as reported in the clinical chemistry white paper, which did not include ALT as a high-priority test to harmonize. Our next test your knowledge question is shown. Please select one response from the following prompt. US ALT analyzers and assays are… and if we could please show the poll. If we could show the polling questions, please. Great, thanks. Oh, that's the… That was a paradoxal. This is a second polling question, please. Yes, it's down at the bottom. It appeared at the same time. I apologize. Oh, that's okay. So, should I just go forward? Let people choose? The poll is closed now. They answered both. Oh, they answered. So, I can't see it. 91% got it right. Oh, perfect. Okay. Oh, for some reason. Great. So, excellent job. US ALT analyzers are not standardized and have significant measurement variation. Many of you are aware that fixed ALT thresholds are used in US hepatology and GI guidances. And much of this arose from a very influential study by Dr. Daniel Pratti in Italy 20 years ago, where he was interested in… He was concerned that patients with subclinical liver disease, such as NAFLD or hepatitis C, may have been included in reference ranges. And he performed ALT measurements in 4,000 Italian blood donors, from whom he selected patients with normal BMI, triglycerides, glucose, and cholesterol, and excluded those with hepatitis B, C, HIV, and medication use. Let's see. Oops, excuse me. And found in this select population, using ALT international standards assay, that the ALT upper limit of normal was 30 for men and 19 for women. However, international ALT thresholds are inappropriate for US clinical care research because of the marked variation in lab assays and analyzes throughout the US. And in a later paper, Drs. Valenti and Pratti actually recognized this and shared that clinicians, when reviewing labs, particularly the ALT for their patients, they should use the laboratory reference range to determine normal. This…Pratti's fixed normal ALT was popularized quickly and embedded in US guidelines, with these thresholds first included in the 2016 ASLD chronic hepatitis B guidance and two-fold elevation beyond these thresholds as a trigger for the use of normal ALT. These thresholds were also incorporated in the 2017 ACG clinical guideline on abnormal liver chemistries, along with several other ALT international standard studies from Europe and Asia, which are not appropriate for use in the US. This guideline recommends that a healthy normal ALT is 29 to 33 in men and 15…excuse me, 19 to 25 in women, and values above this should be assessed by physicians. These thresholds were then incorporated in the 2018 ASLD chronic hepatitis B guidance and also included, along with the Pratti thresholds, in the 2022 and 2023 ASLD and American Association for Clinical Endocrinology Guidelines, which state that an ALT greater than 30 should be considered abnormal. With guidelines suggesting the use of fixed ALT, US research also adopted these thresholds. For example, a NIH study reanalyzed healthy volunteers within NIH trials using the Pratti thresholds and disregarding the NIH laboratory reference range. And an additional two studies, one in adult survivors of childhood cancer and acute COVID, appropriately used the laboratory reference range along with the Pratti threshold. And in the acute COVID study, using the laboratory reference range, 52% of patients had ALT elevations, whereas using the Pratti threshold, 76% of patients had ALT elevations, revealing that the fixed normal ALT misclassifies patients as diseased, and it's… …and physicians should use their laboratory reference range to determine normal. And lastly, a study…a genome-wide association study of ALT as a proxy for NAFLD was performed using 30 to identify healthy controls and an ALT of 40 in men persisting for six months or greater after excluding underlying liver disease to identify as a proxy for NAFLD. This study did not use the laboratory reference ranges, and when studies use erroneous ALT measurements, it's…it's difficult to interpret, replicate, and apply the research findings. So, to summarize, there's been a misinterpretation of ALT measurement in GI and hepatology guidances and guidelines, which has a direct effect on clinical care and has also affected multiple research studies. And to highlight, ALT…in the U.S. and Canada, ALT assays differ. They have different methods. The…even among FDA-approved ALT assays, they can have…they may or may not include pyridoxal 5-phosphate, and ALT analyzers differ significantly, yielding clinically important and statistically significant differences. So, these are the differences in ALT and reference ranges. To further evaluate ALT across the five analyzers used in the VA, Dr. Lauren Best led a multiyear study of VA labs. At the VA, all labs are CLIA-certified, and the VA goes one step above for quality checks using independent assessment with external samples from the College of American Pathology reference standards, which range from an ALT of 21 to 268. She analyzed more than 20,000 ALT samples at more than 200 labs. The key aim of this study was to assess how much variation…what were the ALT measurements found across analyzers throughout the VA, and the VA analyzers are shown, and secondly, to identify how would use of the fixed thresholds in guidelines affect clinical care? I'm only focusing on the normal range to share results because that's where the greatest difference was seen, and it decreased at an ALT greater than 50. However, in this study, an ALT…a College of American Pathology cap sample of 28.5 in these quality-checked labs in an ortho analyzer on top revealed an ALT of 41, and these quality-checked labs in a Roche analyzer yield an ALT of 24. So, a marked difference within normal in these quality-checked labs. A mean of 17-unit difference was seen within the normal range across analyzers, and statistically significant and clinically important ALT differences within normal were seen by analyzer, and even when the highest-reading analyzer was removed, these significant differences persisted. More than one in three VA analyzers were Siemens or ortho vitros, and they yield the highest ALT readings, and this is also found in the Canadian study. So, the second goal of the study was how does a fixed normal ALT in U.S. guidelines affect patients? Well, nearly one in five patients within normal had an ALT greater than 33, exceeding the ACG threshold. So, patients could be misinformed of liver injury, become anxious, you know, when told of these values, or undergo needless testing. This study shows that providers can best assess ALT normal values using the ALT reference range, and should not use fixed thresholds to trigger clinical actions, as one in five VA patients were misclassified by these thresholds. I'd like to now turn the talk over to Dr. Topoletti. Thank you. Okay. Well, thank you, Dr. Hunt and Dr. Dara, for including me on this talk. I have to say, as always, you always learn things when you have to prepare for a talk like this. Let me start out by just showing you some of the data I came across, studies I came across. One of them is from 2004, and this figure here is about reference ranges for serum ALT reported by 11 Nash CRN labs. I had no idea what the Nash CRN labs were, but then lo and behold, when I looked at it, Duke is one of them, and many other major medical centers. But the point of this is to show that the ALT reference range has varied. If you look at some of these, the ones where the male and female had the same values, they just lumped them all together, which several hospitals do. And then on the right, it notes that the female upper reference limit varied from 31 for Lab B, that's the dark bar, up to 55 for Lab E. While the male reference intervals varied, the upper limit varied from 35 with Lab A to 79 with Lab K, and both J and K were fairly similar. So you see right away, even in 2004, there were a lot of variations. And I just did a quick check of several institutions that I could find the reference interval that they would publish, and it was Iowa University, UC San Francisco, Cleveland Clinic, LabCorp. And most of these are actually fairly similar when you look at it. The University of Iowa has both male and female, so 33 and 41. If you took an average in there, that'd probably be around 36, which looks like what San Francisco did. They put all sexes together. LabCorp had both very similar, since they separated male and female, very similar to what University of Iowa had. Cleveland Clinic appeared to be the outlier, just because they were higher, clearly higher on the upper end, 56. And I did note that LabCorp looked at over 260,000 different samples. I don't know how they got those. I'm sure they did not get 260,000 volunteers. But there are these ways of looking at your data and screening out the outliers and coming up with normal reference intervals based on overall results from patients. Next slide, please. In case you're interested, here is the reaction. And I showed here both the Duke and VA methods, and I did not know it, but the VA includes pyridoxal phosphate in their assay, while Duke does not. And so the old ALT reaction is you get alanine and alpha-cleatoglutarate, which are added as reagents, and they're included in both VA and Duke methods. And that's converted. The important thing is down, the pyruvate that's produced from that is then reacted with NADH, which absorbs light at 340 nanometers. But when it's converted to NAD+, charged NAD, then that absorbance goes away. So that's what you're monitoring, you're monitoring the decrease in absorbance. And as Dr. Hunt said, you know, these reactions depend on pH, cofactors, temperature, things like that. So I'll just say now, and I may say it again, that enzyme assays are notorious for having difficult standardization, or they're difficult to standardize in all cases. Next, please. Now, I say this, I'll never do this again, but it seemed like such a good idea at the time. You know, do a reference interval study, and I got high praise for it when it started, and I think the old joke about I got severely criticized by the end of it. You might notice when it stopped, it was 2021. And of course, what happened there was COVID. So 2018 to 2021 at Duke Medical Center, we collected blood samples from 106 healthy volunteers and analyzed these samples for many chemistry and hematology tests. We did, you know, probably 40 or 50 different chemistry and hematology tests. And we collected like six or seven tubes from the volunteer, each volunteer. And so any changes to our reference ranges were based on data from our study. But not just that, we also looked at the package insert from the manufacturers because for CAP regulations, they almost always say, did you follow manufacturer's recommendations? Now that can apply to a lot of things, but one of them could be reference intervals. And so you should always look at those. I don't think you have, I think if you can justify changing them, that's fine. And I think by our study, we could do that, but you should always take them into account. Maybe look at other large medical centers and maybe find out what instruments they use and always look at existing published literature. So in October of 21st of 2023, after lengthy discussions, we updated our reference range for 12 different chemistry and hematology analytes. So out of all the tests we did, only 12 did we change. And we felt that if it was close enough, there was no need to change it. It just is a sidelight there. Okay, next please. Next slide. And here's the data we got from that study. This has some kind of interesting sidelights to it as well. So on the left, you see the reference, the ALT range of the values. Okay, so less than 14 and all these different intervals up to greater than 45. Now the females, we got 77 volunteers. Males, we got I think 29. So if you look at the values, you can see clearly the females are the only ones that had values less than 14. When you get into the 14 to 30, they're more or less similar. And I put the percentage in here because there's such differences among the total of that number. So it's 49 versus 41%, 19 versus 17%, 9 versus 7% in that range of 14 to 30. So those are quite similar. Then when you get above that, you can see that there are more males in the upper groups like, you know, as Dr. Hunt was saying, when you get above 30, there were still in our study, about one third of them had values above 30. So clearly just saying a limit of 30 is correct, might be okay for females, maybe, but certainly not for males. This also has interesting thoughts here. Like why were there so many females versus males? Is it that males are never scared to get a venipuncture, you know, with all the blood? Or they did not want to divulge their health information? And all those are possible where any medications are taken. So I don't know, didn't look into that, but they clearly, we had a harder time getting those. And starting in, you know, it was going fairly well until COVID hit. And I think for the next year and a half, we only got like five or six volunteers. And so we just stopped the study. So COVID had an effect on that as well. So next slide, please. But you can clearly see that there were a lot of values above 30, and the males and females are clearly different. So here's the results of our study. The former Duke Medical Center reference interval was females went up to 54, this on the left column, males up to 63. The reference interval study based on the EP evaluator said females should go up to 39, males should 55. The manufacturer said up to 41 for both as a general number for combining both sexes, male 45 to 55. And so we followed our reference interval study exactly 10 to 39. For males, we went a little bit lower because, you know, it's 55. We went down to 50 because the manufacturer said 41. And, you know, we thought that 50 would just be more appropriate. So we definitely did change our reference intervals based on that study. So next slide, please. The VA Med Center study was much simpler. I know Lainey Bell, who's the lab manager at the time, got 40 samples from, they were collected at a non-urgent clinic. They were assuming that these would be, you know, fairly healthy people, no serious metabolic diseases. And I wasn't going to say about whatever. And we also assumed they were mostly male. So from that data on 40 samples, we got a lot of, again, did a lot of chemistry tests. So I'm just highlighting the ALT ranges here. So next slide, please. Oh, that's the Med Center. Yeah, go for it. There it is. There it is. So this is what we got in the VA study. Again, less than 15, there were none. So that would fit with being males. The 16 to 20 up to 30, we got, that was about 55%. If you look at the right column, that's the percent of the total. But then above 30, we got, you know, large numbers. I mean, almost 40%, whatever, a very large number of samples above 30. So we certainly, clearly would not support stopping at 30. And next, please. So this is how we ended up changing our reference intervals back, you know, over 10 years ago. We used to be up to 75. And just one reference interval for males and females. Again, most of the VA population is males, not all of us, of course, but most. We changed that at the time down to 60. And as of today, we are getting a new analyzer, which is the one you see there. And their literature says only up to 49. So we'll have to look at that and decide what to do. So we're going to be lowering it even more as of the next several months. Next, please. Now, if you think that ALT is the only one, or even enzymes, I just got this data from Fred Apple. He talks a lot about troponins. He's one of the experts on it. This is a high sensitivity troponin. And I don't think we need to belabor all the data in here. But if you look at the manufacturer on the left, in the left column, and just look at the 99th percentile, that's basically the upper reference interval. And just look at the all category. That's combining males and females. That just makes the point here. As you go down, 26, 18, 29, 11, 14, 59, and 46. So very different upper reference intervals for high sensitivity troponin. And this, after billions of dollars of study trying to normalize or harmonize these different ones, one that is not going to harmonize is the Roche, because that's troponin T. The others are troponin I. Uh, something else you might find interesting is the enormous number of people that they had to get for these reference intervals by the companies. You know, at least 500 on up to 2000. And I believe starting with troponin, that's when the FDA started requiring, not just comparing your new method to an old method, but you had to actually do clinical studies on the clinical evaluations on each of the patients, which means you had to have like sometimes two or three MDs review the results. And you can just imagine the cost of that. It's enormous. So that's what they had to do for troponin. I think that's kind of becoming more common. I believe that's my last slide, but it shows clearly that standardizing reference intervals is a challenging task. I believe for enzymes, especially because they depend on so many factors and, you know, they're based on activity really. And, uh, so that, that can vary depending on what, like Dr. Hunter said, the pH, the temperature cofactors or not. Okay. So that's it. Thank you a lot. Thank you very much. And now let's return to Bill. Um, so people can mull over how they would address his, the question that the primary care, um, who referred him asked, you know, does he need a liver biopsy? What's going on? So Bill's a 42 year old non-drink, non-drinker, uh, slightly overweight with an ALT of 42 to 45 and AST 31 to 34, both within the normal reference range. He has a normal metabolic panel, liver ultrasound and FIV4 and negative liver disease markers. What tests do you recommend for Bill? And this one is not a polling question. This is just mull over in your, in your, um, in your head, um, would you want additional laboratory tests or liver transgen elastography or magnetic resonance elastography, liver biopsy or none of the above? So you can select in your head the single best choice. Sorry. Sorry. It's not a polling question and I'll give a couple of seconds. Okay. And reveal the answer. None of the above. Bill's liver chemistries are normal. They're within the normal reference range and he should not, he is not a candidate for liver biopsy for these results. If you put Bill's blood in two different, as we, as was done by Dr. Best and others in the VA, uh, if you assay Bill's blood and say a Siemens analyzer, the ALT might be 45 and within the normal reference interval as, as he had, if you put it in, if you analyze using a Roche analyzer, the ALT might be 28. And again, within the normal reference range. So the normal, a normal ALT lies within the reference range. And that's the best way to determine it in the U S at this time until, um, until ALT assays are harmonized. So the key message is ALT is best interpreted using the reference range because statistically significant and clinically meaningful ALT differences by analyzer we're seeing, um, you know, the 17 unit difference across analyzers, particularly within the normal range. Fixed ALT thresholds misclassify patients as disease as shown in the VA study, as well as the acute COVID study and are inappropriate as a trigger for clinical action or for use in research. And lastly, if you have questions about, um, your reference ranges or the subjects included, um, talk it over with your laboratory professionals, um, and, and also consider including laboratory experts in your research studies and guidelines. I'd like to acknowledge the, the work of all those on the slide, um, to, to analyze and, and report the ALT study at the VA, as well as to interpret the VA study. And, uh, we'll close now and open it up for, for questions. Thank you very much. And thanks for your attention. Thank you both. Very informative. I like to say that, you know, this, although it's sponsored by our SIG, because you are a member of our SIG, this really applies to every single, um, hepatologist, no matter their discipline or their niche, because, uh, it is our most used biomarker for liver injury, most commonly used biomarker. There's a few questions in the chat. I'm just going to read through them. Uh, first attendee wanted to know, and I think maybe John can help with this. Do you know, uh, if Quest and LabCorp use pradoxal 5-phosphate, uh, when they measure tests? The answer to that is easy. I don't know. Uh, like I said, I, I even looking up, uh, the Duke and VA, that was, uh, I never even thought about it really, but, uh, you know, the VA has it in there and the Duke Beckman method does not. And if I understood you correctly, John, it's not just about the, the pradoxal phosphate. Even if you add the same amount of pradoxal phosphate, but use two different analyzers, you could get two different results. Absolutely. Absolutely. Nazia says the VA has a thanks to Dr. Robert Dufour, I believe my mentor. That's right. Did you know, did you say you knew him? Nazia knew him, uh, one of the, one of the participants. I know him. Yeah. He's, he's, he retired a few years ago. Chicago Cubs fan. Yeah, he, he was a real, real, he actually promoted ALT standardization in US. He included in a paper with people from the American association for clinical chemists. So, uh, he's sort of a hero. That would be him. Yes. More work to be done, huh? Get the chemists to agree with us. I guess they're still stuck on troponin. So we had a long way to go until they get to the ALT. Literally, I called troponin the golden child of lab tests. He gets all the money, all the attention. And, you know, it's, it's the perfect lab test, but as you can see, it's not quite perfect. You're a hepatologist. We're used to it. Underdog. Given the very high prevalence of steatotic liver disease in the US population in general, as well as the high prevalence of substance use disorder in the veteran population, how was normality determined in the absence of actual liver pathology in patients exceeding the thresholds of normal? John, how did you define, uh, who is normal? Like, did you just assume they were normal or did you charge? We asked people, you know, that you had any kind of serious disease, chronic illness, not to volunteer if you're diabetic, uh, things like that. That's all we did. We certainly weren't going to do any kind of, uh, in-depth biopsy or even ask about that. So, you know, people wanted to not divulge that they were free to, um, we, we, we kept all the information confidential on their sheet, but they, like I said, many did not want to volunteer for whatever reason. So that, that was, you know, if the person's walking around and has no apparent illness, that's kind of defines what you'd call apparently healthy. And that's probably the best you can do. And then you think about it, reference intervals are difficult to do because you've got to get volunteers. You've got to get IRB approval these days, and you've got to ultimately have people volunteer and you've got to get a range. It's nice to have a range of ages, a range of sexes, and it's not easy. It's not easy. Another one asks, um, how would you interpret or evaluate an ALT that is lower than the limit of normal? It's also out of the range lower. I'll take a stab at it, but it'd be like alkaline phosphatase. You know, is there a genetic deficiency of that, um, the gene that codes for ALT? I don't know that. And any guesses there, Dr. Hunt? Do you know if they dilute the samples or they just put them in like... Well, they're always diluted in the assay, but they're usually not diluted. No, they go on the analyzer as serum or plasma. They're usually spun down with automated systems. Now they automatically have centrifuges that automatically spin them down and they sample from the plasma or serum layer, usually plasma. And then that definitely has to get diluted once it gets into the assay. And the assay could be lacking pyridoxal phosphate, you know. Indeed, you're depending on the, of what's in the, what's in the... Oh, now that would be another obvious. They're, they're very deficient in pyridoxal phosphate. That's a much simpler answer. When you add the pyridoxal phosphate, John, the ALT registers as higher? It would, if there's, especially if they're deficient. I don't know what level you have to be at. You know, if you have, there's usually with cofactors, you'd need a certain amount and if you had, you could add more and it doesn't have any effect. So I don't know what that is. Yeah, I know. It's always been hypothesized to be the reason with the ASC ALT ratio and alcohol. Right, right. Right. Okay. So next question. Also often upon evaluation of patient for chronic liver disease, lab results obtained throughout months and years and across institutions are taken into account. What do you suggest when patients have values that are elevated per one lab threshold, but not another? What if the question is high, but your local lab is not? Well, I'll just say, I expect that they would be fairly close to the upper limit. If somebody had one of 28 and another had 300, I'd say that's something is drastically going on there. And some mix up of the sample, some other possible lab error, collection error. But if it's like 28 versus a 36, I could say that could happen very easily. And as we've talked about, the different methods, I think you have to look into that. And if one is out of the range and one is not, I guess you just repeat the test. Always a good idea, if there's any question. And then Don asks, Dr. Rockey, in both the VA and Duke populations, can you be sure that all these participants were truly normal and did not have underlying unrecognized MASH, MAFLD? All we can do is we ask them, OK? You can make jokes about it and say, is any Duke person really normal? I mean, the UNC might want to say that. Seriously, we asked that. We asked several questions. I'd have to review them what they were now. But basically, if you had any metabolic disease, any medical condition, then please don't volunteer. Yes, all your best. And then one of the attendees, Christine, was asking what Bill's FIP4 score was. I'm guessing his platelets were normal and it had a normal FIP4. Yeah, he was 1.04. He was normal. Normal. And the anonymous attendee said, none of the above, as in he has normal tests. Brett Tetre says, I don't think that we can assume that people seen in outpatient clinics did not have risks for MAFLD. The PRADI study excluded patients with risk for MAFLD, which should be done for any study to define healthy ALT upper levels. I think, Brent, what Dr. Toffoletti is saying is that they asked patients if they had risk factors, excluded diabetes, to the best of their ability. Let me clarify something, because you said it and maybe that's the other person that asked the question. These were not patients. These were people walking around. We were employees of Duke and others. So we got no data from patients. We got nothing from patients. They were totally not part of the conversation. I think that answers the question. Canadian laboratory professionals are planning a study. They want to try to harmonize methods first and then do a large study in healthy controls where they will exclude risk factors for MAFLD and then ideally harmonize ALT reference ranges in Canada. And that will undoubtedly help US. I wonder if they're going to do it tomorrow. That's not going to happen tomorrow. Yeah. But Canada has really taken the lead in this for years. Dr. Adeli, Castro Adeli, he's on there. And he's a wonderful person, by the way, if you've ever met him. But he's done that Caliper study, which is Canadian, something on pediatric reference limits. And they've done a wonderful job getting reference intervals on pediatric patients. Healthy people, healthy kids. Yeah, they've always been ALT activists. Yeah, ALT activists. Rah-rah ALT, get it right. Let's protest. He's even added, tried to put in factors for if you have a different analyzer. So they would try to get from different analyzers and put a factor. Let's say instrument A tends to run 20% higher. Then they would put a factor in there and say, apply this reference range by 1.2 or whatever. So it's really a fantastic job that they've done. Well, that sounds exciting. Chris, Eric Mailer has a question. Is ASLD, ACG listening to your strong case for them to modify their recommendations and guidelines to no longer use a standard ALT, at least in the US? I can't speak to ASLD listening. But I know that the co-authors are aware of the ALT information and the VA study. So I can say that at least the co-authors are aware. ASLD is too large a scope. I think most people in ASLD actually believe we should be using fixed ALT thresholds. And that was the whole point of the talk. Or we should not be. A lot of people do believe that you should use them. However, I think there's abundant data to not use them. We should not be using them in the US, because they misclassify, mostly because they misclassify patients. Yeah. I'll tell you, that's how I was trained. We used 19 and 30 when I was training. And you know what the problem is? You know me from the hepatotoxicity world. But I also do a lot of autoimmune and PBC. And I see a lot of people being misclassified with PBC as having overlap, or having AIH variants, just because they're ALTs. If you use 19, a lot of people are going to end up having five times the upper limit of normal of ALT. Whereas we just really have just a little more aggressive PBC. So it really matters if you use 19 as the cutoff, or 35 when you're talking about these low-level elevations. Yeah. Yeah. People need to use their lab reference range. And ALT is normal if it's within that reference range. Well, you know, Chris and Lily, you also mentioned Dr. DeFore. And maybe it's time to revisit those. There's been so much effort on proponent. I mean, I hear talks, like I said, they're the golden child of lab tests. And I think we've focused on that so much that we've kind of forgotten some of these other tests. And it might be time to refocus on ALT, being that this, I did not realize there was such an intense interest in the subtleties of this until Dr. Hunt asked me to look into this. So like I said, I've learned a lot myself. And maybe it's time to, maybe that'll be something I'll take up, maybe. Maybe get Dr. DeFore out of retirement. Oh, he would be all over harmonization. I have heard him speak on harmonization. Has he talked about this recently, Chris? This is probably at least 15 years ago. OK. But it would be good to see. I'd be interested to see if I ever see him. And he sometimes shows up at meetings to see what he concluded from all that or where we need to go with that. So I'm just going to try to, because we only have like eight minutes left, I want to get through the rest of the questions. So Dr. Skip Hayashi, a colleague of ours at the FDA, says the variation analyzer seems to be 40% at most. This doesn't seem to fully explain the range of upper limit of normal from 35 to 80 within the NASH CRN. This is more than twofold, so more than 100% higher than the upper limit of normal when you go from 35 to 80. What is the rest of this variation? Should we still feel comfortable accepting such a high upper limit of normal, such as ALT of 80 in clinical practice? If the upper limit of normal is 80, should we still accept it if that's what the range that the lab gives you? I don't know. I'm not sure what you would do. You can ask them how they got their reference intervals. And I think part of it comes down to the paradoxical factor. What I didn't do is look and see which ones of those with their methods were, and did they include paradoxical phosphate? The person that's saying that would not explain all of it, but I don't know that. It would not explain all of it. So it's the difference in the analyzers and the paradoxical phosphate. My understanding is some reference ranges have been, as Dr. Toffoletti was saying, that they change over time. And if somebody is using an earlier assay, maybe a broader reference range. So that would be a discussion with the lab professionals to find out what kind of assay is being used and have there been healthy controls assessed. Also, that study was from the early 2000s when the bell curve of the normal for each lab was defined. Maybe NASH-NAFLD was less. Maybe there were some actually fatty liver metabolic. Maybe that was not an actual truly normal population in the early 2000s. Maybe now we have more and more careful about who we include in the normal, as everyone in the chat has been suggesting. Kido Hoshitsuki asks, do you think the new FDA final rule on laboratory-developed tests, LDT, will help standardize the ALT tests across labs because they're already FDA-approved tests? Additional standardization is needed? Oh, I don't think that would. I would not want to go to the root of an LDT. I'm not sure, you know. I'd have to know more about what the question asked. What would you do to make it a lab-developed test? And he's very correct about that's a serious issue now. Although I think they've made it so that the FDA does not have control over it, I believe. I haven't followed it very closely. But at the end of the day, they're not standardized, right? You're saying that each, they can get FDA approval, but the ranges are all over the place. What it is, if a manufacturer gets FDA approval, then that's it, as long as you use it. If you modify it, which can include a lot of different things, you know, you can use different, if it's for plasma serum and you use a CSF or some fluid, that should be validated. If you modify the reagents anyway. I mean, it's possible even with the representable, if you just decided to modify it. Although if you have data, like I hope we would, that would, you know, that we try to validate the ranges, then we should be okay. I don't think anybody would complain about a slight change, like from, let's say 40 to 43 or something like that. I just don't think that would ever be a problem. I hope not. We have some comments about low ALT. Mercedes Martinez is saying low ALT could represent malnutrition or lower muscle mass. And Andrew was asking, why are ALT levels lower in patients with renal failure? And I guess maybe Mercedes was saying, maybe it's because of malnutrition or lower muscle mass. Of course, muscle also makes ALT, right, John? Chris? Say it again, please. Muscle makes ALT too. That would be correct, yeah. Yes, so does kidney. So maybe in renal failure, is there a reason why ALT levels are lower? Have you ever thought about that in renal failure? They are lower, and I honestly can't tell you why. I mean, that's well-established. Did you consider measuring a secondary liver marker like microRNA 122? It would be highly specific for hepatocyte damage. Simple answer is no. Good question, though. A very specialized test. Yeah, in addition to excluding mass ALT, how did you exclude alcohol use in the reference sample? SEPH reported or biochemistry? I think we asked that, but again, I don't know if people are gonna be that honest about it. I don't think we rejected anybody because of that. Honestly, I can't remember. Good question, though. And if you look closely at that data, there were some that, well, you know, on the males, there were some that had very high ones that we did not count in determining the reference sample. There were three that were above 50, 58, and 65. And they may have had some condition. They may have been, you know, drankers, heavy drinkers, I don't know. You may have read that. Just looking back at our data, I have it right here. You know, from 40 to 45, there was zero, no males. But above 45, there were those three. So those could be things you might exclude, statistically and maybe medically as well, clinically. Okay, so the whole very interesting concept, and Nadia's gonna check back with Dr. DeFore and get back to us. And as we wrap up, I'm just gonna throw a little bomb out there that we talked about earlier before the webinar started. And this is all well and good, but, you know, in patients with cirrhosis, the AST-ALT ratio predominates, you know, AST, and that's what we cover. Is that its own can of worms when we talk about standardizing ALT? Do the same concepts apply to AST as well? Would we have to separately also standardize that? If you're asking me, I guess, in the ideal world, yes. You know, what I was gonna suggest is that there are a lot of committees that are formed for a lot of things, and one of them would involve, like, the manufacturers. And plus clinicians like yourself, laboratory people, and that are interested in harmonizing this. And, you know, it sounds like a lot of people are harmonizing this. And, you know, it sounds like the United States is the part of the culprit, I guess. And just looking at the two methods we looked at, one has pyridoxal phosphate in there, and one doesn't. So I think it would have to be involved with the manufacturers. I think if that person who asked about the laboratory-developed test, if that means we would go out and modify our test to put pyridoxal phosphate in one of the methods that does not already have it, that would be a lab-developed test for sure. And we try to avoid that at all costs. And then Skip's saying, in DILI, where we rely on the R-value, the upper limit of normal for alkaline phosphatase is always a big problem. The range is very wide, I agree. And, you know, with alk-phos-Skip, the gender and the age both factor in. With ALT, at least we let go of the age issue, but with alk-phos, it's like such an age-dependent enzyme. It's interesting you should mention that, because about, I don't know, 10 or 12 years ago, we had a big change with the lower limit of the alk-phosphatase, because as I understand it, in fact, I know this, we had 24 was the lower limit for all ages, because we didn't worry about it. The upper range were very different as ages go. But it turns out that the people with the hypophosphatase emia genetic disease and had very serious bone disease, there was really no treatment for it. So nobody really questioned it, but about whatever, 10, 12 years ago, maybe a little bit more, they came up with a treatment for it, and so it became very important to now diagnose it. So we modified our reference ranges for ALT-phos, mostly in the lower ranges. So children were now, the lower range was now 40 or 50 instead of 24. The adult was correct, we stayed the same, but so it can go both ways on that. Very interesting. Any final words as we wrap up? Just want to say thank you very much for the questions and people's attention and great ideas about moving this forward with manufacturers. I think that's an excellent strategy. Excellent questions, for sure. Yeah, yeah, great questions. And thank you so much. Yep, very good. Bye everyone. Thank you, John. Thank you, Chris. Thank you. Thanks.
Video Summary
In this webinar, the speakers discussed the variability in ALT measurements due to different analyzers and assays, highlighting the need for standardization. They emphasized the importance of using lab reference ranges to interpret ALT levels accurately and cautioned against relying on fixed thresholds. The speakers also presented data from studies at Duke University and VA medical center that showed inconsistencies in ALT measurements across analyzers, further highlighting the need for harmonization. They addressed questions regarding low ALT levels, exclusion of potential liver pathologies in reference samples, and the implications of the FDA's final rule on laboratory-developed tests for standardizing ALT measurements. They concluded by highlighting the complexity of ALT measurement and the importance of collaboration between clinicians, laboratory professionals, and manufacturers to improve standardization and accuracy in ALT assessments.
Keywords
ALT measurements
variability
analyzers
assays
standardization
lab reference ranges
interpretation
data inconsistencies
harmonization
low ALT levels
Hepotoxicity
×
Please select your language
1
English