logo/base Created with Sketch.

Users stories: interview with Stephen Harrison


Many of our trials then also include FibroScan®  along the way, either as a baseline into treatment or a baseline midpoint into treatment, noninvasive test parameter as a secondary or exploratory endpoint.

Dr. Stephen Harrison| M.D., gastroenterologist and hepatologist, medical director of Pinnacle Clinical Research, and visiting professor of Hepatology, University of Oxford.


FibroScan® is an integral part of our pre-screen practice. When we identify a patient who from demographic criteria could potentially have fatty liver, we then look at the liver biochemistry tests, specifically an aspartate aminotransferase (AST), and then we perform a FibroScan® with the controlled attenuation parameter (CAP) function as well as the kilopascal (kPa) score.

At that point, we do two things:
1) We look individually at the AST and see if it is above 20, and we look at the kPa and determine if it is greater than 8.5. and that the CAP is greater than 280.
2) If that indeed happens, then we move the patient forward into the queue to be considered for a clinical trial at Pinnacle.

We also generate the FibroScan®  FASTTM Score by adding the CAP™, kPa plus AST.  If it is greater than 0.67, we put them into our F2 for higher studies. If it is greater than 0.35 but less than 0.67, we put them either in a noninvasive trial or a study that will take F1 patients in addition to F2s and F3s –but they span the spectrum usually from F1 to F3. That is how we use FibroScan® up until the point that they screen for a trial.
Many of our trials then also include FibroScan® along the way, either as a baseline into treatment or a baseline midpoint into treatment, noninvasive test parameter as a secondary or exploratory endpoint.

Part of the issue with the high screen fail rate is that in many of our studies, particularly late phase clinical trials, we have to go through three different gates to get a patient qualified for enrollment.
The first one is labs as well as meds and consent.  We lose a percentage of patients based upon these issues. 

For instance, we just lost a patient today when they came back with an HbA1c at 14.5 and almost every study excludes patients with HbA1c more than 9.5. We would not have known that prior to seeing the results because we did not have any historical A1c to draw from and the patient thought she was much better controlled than she actually was, which is a common occurrence.
Then we move forward to an imaging gate, usually an MRI.  We’ll lose a certain number of people based upon the magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) or the magnetic resonance elastography (MRE), which are requirements for some studies and the early phase clinical trial work that we are doing.
The last gate is the liver biopsy gate. Historically, pre-FibroScan®  and liver biochemistry analysis, liver biopsy screen fail rates were between 60-65%, just for that gate alone –not to mention the lab and the initial screen fails that you get prior to even going to liver biopsy.
So that contributes to the high screen fail rate. Where we have been able to improve that screen fail rate is twofold, using FibroScan®:
The first is the CAP function. If a study has an MRI in three criteria for a certain amount of liver fat based on MRI-PDFF, we can adjust the CAP to help us achieve that certain MRI-PDFF threshold with about an 80-85% accuracy. That minimizes the screen fail rate on MRI to somewhere between 15-20%.  Subsequently, by combining in a sequential fashion a liver biochemistry test with FibroScan®, we have been able to drive the liver biopsy screen fail rate down from 60-65% to 30-35%.

These are incremental advances, no doubt. But overall, this has improved our screen fail numbers. However, we still have work to do and we continue to refine the system.

We are hopeful that FibroScan®-AST (Fast) may be the next incremental advance in allowing us to do that.

The FastScore is a combination of three components: two physical biomarkers: liver stiffness by VCTE and CAP, estimating fibrosis and steatosis from an examination with FibroScan®, plus AST, a readily available blood marker of inflammation.

FibroScan® and tools like Fast support diagnosis efficiency at the point of care, improving diagnostic accuracy and reducing costs. Fast is inexpensive, reducing unnecessary, invasive assessments and the need for expensive, complex blood tests. It’s simple to determine and interpret – requiring just three numbers entered into a calculator on the myFibroScan app. Furthermore, the clinician can set the threshold for referral based on their goals, balancing sensitivity vs. specificity.

I think this tells practices that are interested in considering NASH clinical trial research that it is not for the faint of heart – even for those of us that do this for a living and enjoy the challenges that go along with it.

But you can equip yourself with tools like FibroScan® that can help you be successful in this arena –much like you use vital signs because they’re vital, I look at FibroScan® as a sixth vital sign to help us achieve success by pre-identifying patients at a greater likelihood of passing through the gates that I’ve alluded to earlier.

I look at all noninvasive tests through the prism of distinct contextual issues.

The first one is, diagnosis of at-risk NASH patients—those who will likely progress if therapy is not provided. That is the group of patients we want to key in on. I think utilizing app with FibroScan® helps us to exclude most patients that do not fit that category. Then combined with routine liver biochemistries, we can better identify that group that is likely to progress.

The second context of use is monitoring for therapeutic efficacy of the drug in question. Here we have early preliminary data from the Regenerate trial through a post-talk analysis recently presented by Jérôme Boursier group, showing that we can utilize FibroScan®, the liver stiffness measurement, to correlate with improvement in histopathology. Previous work has shown that a 25% reduction in FibroScan® kPa correlates to improvement in histopathology.

I am hopeful that as we move forward and we utilize different mechanisms of action to treat this disease, studies will be utilizing FibroScan® as a noninvasive tool along the way, such that similar work can be generated like the Regenerate trial, showing that improvement in histopathology is correlated to improvement in FibroScan® kPa. When the time comes to prescribe the drug, providers will feel comfortable in utilizing vibration-controlled transient elastography (VCTE) to assess both compliance with the drug and overall efficacy.

Focusing specifically on patients with Type 2 diabetes, we know from elegant work generated by Zobair Younossi’s group, that the overall prevalence of fatty liver is around 55-60% among Type 2 diabetics, no matter what part of the world you live in.
Specifically amongst diabetics, we often see the more aggressive form of fatty liver, nonalcoholic steatohepatitis (NASH). In fact, up to one-third of patients with diabetes may present with this more advanced form of fatty liver disease.
Maybe, and most importantly, about 17% of people that have diabetes have underlying advanced liver fibrosis. So, if all this begins with identifying a Type 2 diabetic patient with fatty liver, then we can use the CAP successfully to do just that. We know that a CAP score greater than 280 in the United States correlates well with identifying patients with fatty liver.
If we have a Type 2 diabetic patient and we utilize VCTE, determine the CAP score and identify that it is greater than 280, then we can also look at the FibroScan® kPa and see if that score is less than six or greater than 8.5.
If it is less than six, but they have fatty liver, I tell the patient, “Well, good news of great joy for today in the city of San Antonio, while you have Type 2 diabetes with fatty liver, it has not precipitated significant scarring in your liver. I want you to maintain better control, tight control of your Type 2 diabetes. I would like for you to focus on losing about 10% of your body weight through aggressive diet and exercise. And I want you to focus on eating more of a Mediterranean type diet and limiting your processed carbohydrates.”
On the other hand, if your CAP score is 280 or greater and your kPa is greater than 8.5, regardless of what the serum liver biochemistry test panel shows us, you are now identified as an individual that is potentially at risk for scarring due to fatty liver and Type 2 diabetes.

A more aggressive management approach should be brought forward to include consideration of evaluation for a clinical trial, further evaluation with MR elastography and/or liver biopsy.  
If you are in the range where the CAP is elevated above 280, which the kPa is between 6 and 8.5 This is where I ask the patient to look at the liver biochemistry panel, as well as family history of liver disease. And if you are in the at-risk group there, then I would proceed as if your kPa were greater than 8.5.

If we can prevent the progression of fibrosis at a minimum and better than that, show regression of fibrosis and resolution of NASH, then I think we will affect long-term outcomes –not just liver related events, but cardiac, renal and malignancy related events.

If we’re able to continue to generate positive research as it relates to VCTE, I would love to be able to point to a point of care test that’s readily available in my clinical practice that can actually be linked to a long-term outcome.  This gets me back to the third context of use of a noninvasive test, which is the ability of that noninvasive test to predict how a patient feels, functions or survives over the long-term.

We’re in the middle of data generation now, but we’re hopeful that not only can we use VCTE for the first and second context of use that I’ve highlighted, but also for that third one, so that we can provide a full service point of care test to our patients at the time they come into clinic and reassure them along the way that they’re either getting better or that at worst case, they’ve stabilized their disease.

It does several things. Again, the biggest one is if you are able to do VCTE in your clinic, then you’re able to exclude 80% of the patients who are not at risk of disease progression but have fatty liver.

Another way to say that is this: there are a roughly 100 million Americans with fatty liver, but only about 20 million have NASH and are at risk of disease progression over the long-term.

We have identified a test that can really reassure us that you are in that 80% category. You are in the group that has fatty liver but doesn’t have NASH with fibrosis. That is a patient where we can institute lifestyle modification and one that does not to move forward with expensive testing and medication.

Simply through foundational lifestyle modification, these patients can begin to affect a positive outcome. In addition, if we’re able to identify those patients with fatty liver, we can correlate that fatty liver to a risk of developing diabetes or other potential comorbidities.  With aggressive management, we may now prevent the development of those long-term comorbidities down the road.