Goodbye biopsies, non-invasive diagnostics for fatty liver disease are here

From last night’s happy hour drinks to this morning’s greasy breakfast burrito and ibuprofen, the liver has seen it all. As both factory and filter, the liver is a jack of all trades. It removes harmful toxins from the blood, creates bile to carry away waste, metabolizes drugs, stores iron and other vitamins, and even synthesizes some proteins. 

While the liver bravely faces whatever people decide to put in their mouths, it is not infallible. Whether due to drinking too much alcohol or from living with conditions such as type 2 diabetes and obesity, people can begin to accumulate fat on their livers.

“One out of three adults in North America, for example, has excess fat in their liver,” said Arun Sanyal, a fatty liver disease researcher at Virginia Commonwealth University and the chair of the Foundation for the National Institutes of Health (FNIH) initiative Non-Invasive Biomarkers of Metabolic Liver Disease (NIMBLE) consortium. “The real question is not whether you have excess fat in the liver, but is the fat injuring the liver and putting you at risk of developing something bad from a liver point of view like cirrhosis?”

Most people who have non-alcoholic fatty liver disease (NAFLD) simply have some excess fat on their livers with no additional liver damage. Lifestyle interventions such as eating healthier foods and exercising more can prevent further fat accumulation. A small number of people with NAFLD, however, also have liver inflammation, an indicator of the more serious disease, non-alcoholic steatohepatitis (NASH). At more advanced stages, people with NASH can develop liver fibrosis and eventually cirrhosis, leading to liver failure or liver cancer.

The problem is that early stages of NAFLD and NASH are asymptomatic. Symptoms only start to emerge at advanced stages of NASH, which currently has no treatment.

“These patients are picked up mostly incidentally. They will have a blood test with their family doctor, and that will measure some abnormal liver function tests. Or they will go to have an ultrasound scan for whatever reason, if they have aches or pains in the tummy, and that measures fatty liver,” said Michael Pavlides, a gastroenterologist and expert in liver imaging at the University of Oxford.

While blood tests and imaging tests like ultrasounds and MRIs can help diagnose someone with NAFLD, these tests are not very good at determining the exact stage of liver disease. For that, the gold standard is a liver biopsy.

“Imagine we have 200 million something adults in the United States, and a third of them have excess fat in the liver,” said Sanyal. “To do a liver biopsy in 60 million people is just impractical.”

Liver biopsies are also invasive. If a doctor wants to prescribe a patient a treatment, they want to be able to see if the treatment is working or not. To do this, doctors need to take an initial liver biopsy before treatment and another one a year to 18 months later.

“It's very hard to convince a patient to go through two liver biopsies. It's even harder to get them through three liver biopsies,” said Sophie Cazanave, a research director at Glympse Bio, a biotechnology company developing a non-invasive blood-based biomarker test for NASH.

Liver biopsies can also give variable results depending on where in the liver the biopsy is taken. In addition to being invasive and painful, they also have risks of complications, including death.

“A diagnostic test that can itself kill you should be a ‘never’ event,” said Sanyal.

To find an alternative for invasive liver biopsies, fatty liver disease scientists have formed partnerships across academia, industry, and government to test the accuracy and variability of currently available and newly developed non-invasive tests to determine the progression of NAFLD and NASH. The validation and development of non-invasive diagnostics will allow NASH patients to be more easily monitored by their doctors and will help clinical trial administrators see how effective a particular medication is in patients at different stages of fatty liver disease, accelerating the development of new treatments for NASH.

Breaking boundaries for non-invasive NASH diagnostics

While people with NAFLD or early stages of NASH may appear perfectly healthy, their livers tell a different story. They may be decorated with fat globules or may already be starting to stiffen with fibrosis.

Non-invasive tests for NAFLD and NASH are already available to patients and clinicians, but many of these tests lack the sensitivity to accurately identify a person’s stage of liver disease. These tests may also give variable readouts from day to day or pick up markers from other commonly co-occurring conditions like obesity or cardiovascular disease.

“When the test is positive, we want to be certain that the patient has a problem. That has been a major issue with liver disease diagnostics, that even if the test is called positive, the positive predictive value is low,” said Rohit Loomba, a NAFLD and NASH diagnostics researcher at the University of California, San Diego.

Not being able to accurately assess the disease progression of a patient makes it difficult to determine how to treat a patient and to tell if those treatments are even improving liver health. This ambiguity also makes clinical trials for new NASH treatments difficult.

“For drug development, not only do you need to be able to measure something, but you need to be able to measure something in a very well-understood, well-characterized, and standardized way,” said Anthony Samir, an expert in ultrasound imaging for liver disease at Harvard Medical School.

To tackle this problem, groups in both North America and Europe assembled teams of the top fatty liver disease minds in academia, industry, and government — essentially assembling the Avengers of NASH. The North American contingent is called NIMBLE, while the European group is Liver Investigation: Testing Marker Utility in Steatohepatitis (LITMUS). By assessing the variability and sensitivity among different non-invasive tests, both the NIMBLE and LITMUS groups hope to identify biomarkers for the progressive stages of NAFLD and NASH that can be approved by the appropriate regulatory agencies, accelerating the process of NASH drug discovery and development (1-2).

“There’s no way that two big competing pharma companies can just get together and do something. We have to have a structure in place, and the FNIH provides that [for NIMBLE],” said Samir, who is also a member of the NIMBLE consortium. “The field will not advance if academia, industry, and government aren’t all talking to each other, and they don't talk to each other if they're left to their own devices.”

Both NIMBLE and LITMUS have already made some exciting progress validating established non-invasive diagnostics and developing new ones, laying the groundwork for developing new treatments for NASH.

A picture perfect liver

Physical changes to the liver due to disease have been documented since the time of Hippocrates (3). A hardened or stiff liver was a clear sign that something was amiss. Now, scientists have sophisticated tests like ultrasound and MRI that can measure the physical transformations occurring in the liver due to fatty liver disease.

Clinicians often use ultrasound technologies to measure liver fibrosis and fat content. How ultrasound waves echo and scatter off of liver tissue gives a measurement of the fat present. For fibrosis, clinicians use a technique called ultrasound elastography, which measures the stiffness of an organ or tissue. Livers with more fibrosis like those in patients with advanced NASH are stiffer than healthy livers and respond differently to an ultrasound wave, which gives the clinician an idea of the patients’ liver fibrosis level.

Ultrasound elastography accurately determines if someone has no fibrosis or advanced fibrosis, but it is less accurate when diagnosing intermediate stages of fibrosis. But, Samir explained, “That’s useful because fibrosis is actually the best predictor that a person will go on to develop cirrhosis.” Clinicians want to identify people with this advanced stage of NASH so that they can try to intervene with medications to slow its progression.

Compared to ultrasound, MRI is much more accurate at measuring liver fat content and slightly more accurate as a measure of fibrosis. It is, however, more expensive and less accessible than ultrasound.

“There are just a lot more ultrasound machines than there are MRI scanners, so when you look at technologies that are suitable for imaging one third of the population, then you're probably looking more at ultrasound,” said Samir. “But when you're looking for technologies where accuracy is really important and where looking for fine-grained change is really important, then it's almost certain that MRI has a very significant role to play as well. They're both very important and complementary.”

To assess the usefulness of multiple different imaging tests in diagnosing different stages of NAFLD and NASH, Pavlides recently performed a meta-analysis of 70 different NAFLD and NASH imaging studies as a part of the LITMUS consortium (4).

“We wanted to see which of these techniques could be used to diagnose what's termed advanced fibrosis,” Pavlides explained. “These are the patients who are more at risk of progressing and running into problems with their liver disease, so we're particularly interested in them.”

After investigating results from studies of four different imaging techniques, Pavlides and the LITMUS investigators reported that two tests, magnetic resonance elastography (MRE) and point shear wave elastography (pSWE), detected advanced fibrosis with acceptable sensitivity and specificity. MRE also accurately staged patients with liver cirrhosis. The LITMUS group is now validating the results of this analysis by tracking NAFLD and NASH patients prospectively.

Monitoring microRNAs

As fatty livers undergo a physical metamorphosis, they also secrete evidence of those changes into the bloodstream, which — if validated as specific markers for certain stages of NAFLD and NASH — allows for a diagnosis via a simple blood test.

Some microRNAs (miRNAs), non-coding strings of about 22 nucleotides that can regulate gene expression, associate with NAFLD progression and are currently used as blood-based biomarkers for certain stages of NASH progression (5).

To identify new miRNA biomarkers for NAFLD and NASH in an unbiased manner, Ann Daly, a pharmacogeneticist at Newcastle University, and her LITMUS colleagues, profiled more than 2000 miRNAs from patients with a confirmed diagnosis of NAFLD or NASH via a liver biopsy and healthy controls (6).

Daly and her team identified multiple miRNAs that had increased expression in patients with advanced stages of NASH as diagnosed from their liver biopsies. The signal from one miRNA in particular, miR-193a was the most consistently increased at advanced NASH stages, and it was more abundant in the blood serum compared to some of the other miRNAs with increased expression.

“Abundance is not such a problem when you're doing sequencing. But we don't want to be doing sequencing all the time,” said Daly. “For that reason, miR-193a seemed to be the easiest in terms of detection by real time PCR,” making it a potentially useful marker to detect in a blood sample from a patient.

Daly added that while miR-193a was statistically elevated in patients with advanced liver fibrosis and fat levels, there was some variability in the data. She and her team are now looking into whether exosomes, which are membrane-bound vesicles that can carry proteins and nucleic acids from cell to cell, might be more enriched for miRNAs from fatty livers. 

“We're looking at the possibility that exosomes derived from the liver are higher in patients with NAFLD, and that isolating miR-193a from those might give better specificity than we see at the moment,” Daly said.

In parallel with their miRNA work, Daly and her LITMUS colleagues are searching for new blood biomarkers that are slightly larger than miRNAs: proteins.

Peptides and proteases point to NASH

As NASH livers become more fibrous and liver cells inflame, proteins and proteases that regulate those processes get secreted into the bloodstream. There are multiple non-invasive tests that use these proteins to distinguish between different levels of fibrosis, fat, and inflammation in NASH (7).

While these tests have proven useful for identifying different stages of NAFLD, many still need to be validated in terms of their sensitivity and specificity for discriminating between early and more advanced stages of NASH. Scientists at the biotechnology company SomaLogic have been working with the LITMUS team to compare their new non-invasive NASH protein test called SomaScan with other non-invasive tests for NASH.

SomaScan uses machine learning to determine which combination of proteins out of approximately 5,000 proteins in blood characterize the different stages of NAFLD and NASH based on liver biopsy data. They have now expanded their test to measure approximately 7,000 proteins at once.

“We decided to make four tests, one to mimic each independent component of the liver biopsy, so steatosis, inflammation, ballooning of hepatocytes, and fibrosis,” said Stephen Williams, the chief medical officer at SomaLogic.

Williams’ team trained their model on data from a natural history study of NASH patients and two prospective clinical trials that investigated the effectiveness of two potential treatments for NASH: the Pioglitazone, vitamin E, or placebo for nonalcoholic steatohepatitis (PIVENS) trial and the Farnesoid X nuclear receptor ligand obeticholic acid for non-cirrhotic, non-alcoholic steatohepatitis (FLINT) trial (8-9). Overall, they based their algorithm on about 2,500 samples from 636 individuals enrolled in these studies (10).

“We actually had three active groups and two placebo groups,” said Williams. “The training and the validation would be applied to people on and off drug, with and without diabetes so that we thought that the results would be more generalizable.” 

Because both the PIVENS and FLINT trials collected seven blood serum samples from the participants during the course of the trial, the SomaLogic scientists were able to test how well their model tracked an individual’s response to the drug during the course of the trial, and they found that their SomaScan test picked up the effects of drug treatment within weeks of the initial dose. 

“We get a very nice pharmacodynamic profile for each of the components of the liver biopsy independently,” said Williams. “There's nothing else that measures all four components, and it's all from a single source, all non-invasively.”

Williams is excited about the information SomaScan can provide to scientists and companies performing clinical trials for potential NASH treatments. The test could inform researchers on how long it takes people with a certain stage of NASH to respond to a treatment and which features of NASH improve. Williams and his team at SomaLogic have submitted this work for publication, and it is currently under review.

Taking a complementary approach, scientists at the biotechnology company Glympse Bio developed a new non-invasive biosensor for NASH based on the proteins that chew up other proteins: proteases.

“Both progression of fibrosis and regression of fibrosis are associated with unique patterns of enzymes that are activated, and by looking at the products of those enzymes, you can then get an idea of whether you're in active fibrosis mode or in regressive fibrosis mode,” explained Sanyal.

In a recent Science Translational Medicine study, Cazanave and her colleagues at Glympse Bio identified proteases that were differentially regulated between early and advanced stages of NASH by isolating RNA from liver biopsy samples (11). They narrowed down the number of proteases to 13 that could distinguish between early and advanced NASH regardless of a patients’ obesity or diabetic status, indicating that their test was specific for NASH.

The team at Glympse Bio designed a peptide-based biosensor that would be cleaved by the 13 differentially regulated proteases that they identified. Originally, they planned to inject the peptide sensor into the bloodstream where it would directly measure protease activity in the liver, and then the peptides would be released into urine for collection and analysis.

“We realized that a lot of the proteases that we were studying were secreted… and we could detect them from the blood,” said Wendy Winckler, the chief scientific officer at Glympse Bio. “Once we were able to convert this to a blood test, it enabled us to go much faster, to look at many more patients, and to have, I think, something that will be a clinically more tractable product.”

In a presentation at the American Association for the Study of Liver Diseases meeting last November, Sanyal reported data from the Glympse Bio team showing that their sensor could detect changes in protease activity in blood samples as early as one week after treatment in a rodent model of NASH (12).

“That's the benefit of working with activity. We're not looking at a static marker that is only detectable over time,” said Winckler. The process of fibrosis occurs over months to weeks in an animal and would take even longer in a human, so by having a quick readout for changes in fibrosis activity, Glympse Bio’s test would be useful for monitoring NASH patient progression.

So far, the team at Glympse Bio has reported that their sensor can differentiate between healthy and NASH in human blood samples, and they are planning to release data for their sensor’s ability to differentiate between different stages of NASH soon.

Whether it’s through a drop of blood or the image of an echo, non-invasive replacements for liver biopsies — that are just as accurate and precise — are closer to the clinic than they’ve ever been. Through partnerships across research institutions and borders, liver scientists are revolutionizing how clinicians monitor and treat NAFLD and NASH patients. These new diagnostic platforms will speed up clinical trials, allowing for easier monitoring and faster development of new treatments for a disease that affects millions of people around the world.

References

1. Sanyal, A.J. et al. Non-Invasive Biomarkers of Nonalcoholic Steatohepatitis: the FNIH NIMBLE project. Nat Med (2022).
2. LITMUS. https://litmus-project.eu/ (2017).
3. Mousa, O.Y. & Kamath, P.S. A History of the Assessment of Liver Performance. Clinical Liver Disease  18, 28-48 (2021).
4. Selvaraj, E.A. et al. Diagnostic accuracy of elastography and magnetic resonance imaging in patients with NAFLD: A systematic review and meta-analysis. Journal of Hepatology  75, 770-785 (2021)
5. Harrison, S.A. et al. A blood-based biomarker panel (NIS4) for non-invasive diagnosis of non-alcoholic steatohepatitis and liver fibrosis: a prospective derivation and global validation study, Lancet Gastroenterol Hepatol  5, 970-985 (2020).
6. Johnson, K. et al. Increased serum miR-193a-5p during non-alcoholic fatty liver disease progression: Diagnostic and mechanistic relevance. JHEP Reports  4, 100409 (2022).
7. Castera, L. et al. Noninvasive Assessment of Liver Disease in Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology  156, 1264-1281.e4 (2019).
8. Sanyal, A.J. et al. Pioglitazone, Vitamin E, or Placebo for Nonalcoholic Steatohepatitis. N Engl J Med  362, 1675-1685 (2010).
9. Neuschwander-Tetri, B.A. et al. Farnesoid X nuclear receptor ligand obeticholic acid for non-cirrhotic, non-alcoholic steatohepatitis (FLINT): a multicentre, randomised, placebo-controlled trial. Lancet  385, 956-965 (2015).
10. Williams, S.A. The “Liquid-Liver Biopsy”: Predicting NASH and NAFLD with serum protein biomarkers. (2020). https://somalogic.com/webinar/the-liquid-liver-biopsy-characterizing-nash-and-nafld-with-serum-protein-biomarkers/#webinar 
11. Cazanave, S.C. et al. Peptide-based urinary monitoring of fibrotic nonalcoholic steatohepatitis by mass-barcoded activity-based sensors. Sci Transl Med 13, eabe8939 (2021).
12. Sanyal, A.J. Accurate Diagnosis of NASH Using Novel Protease Based Liquid Biopsy. AASLD (2021). https://glympsebio.com/presentation-accurate-diagnosis-of-nash-using-novel-protease/