3D view of a molecule of a microRNA (miRNA) molecule

by Albert Wright  PhD, a retired British research scientist with investments in life science and biotech companies. His approach is to try to understand the complexe science underlying the development of biotech companies and to make this information available to readers in a form that they can most easily understand. In some cases the science is new to him and requires considerable research before publishing an article. Please fell free to contact him on

Genfit's biomarkers : from Nash diagnosis to patient classification ?



Genfit a annoncé en Septembre 2015, le dépôt d'un brevet sur un test sanguin pour définir la sévérité de la maladie dans des patients souffrant de Nash et qu'il serait utilisé dans l'étude Phase III à venir. Ce test est basé sur les micro-ARN (miARN), molécules qui sont produites par l'ARN du patient et circulent dans le sang. Genfit n'a fait aucune tentative pour expliquer les micro-ARN, ce qui fait que l'annonce n'était uniquement compréhensible par des spécialistes de la biologie moléculaire ou de la génétique. Les analystes financiers n'ont fait aucune tentative pour comprendre l'annonce et l'ont ignoré.

L'importance de cette annonce ne peut être reconnu que si on comprend le rôle essentiel joué par les ARN dans le métabolisme lipidique et l'obésité. Les microARN régulent l'expression des gènes. Un seul miARN peut réduire les niveaux d'expression de centaines de gènes. En particulier les miRNAs sont des inhibiteurs de PPARa, le récepteur nucléaire qui régule le métabolisme des lipides. Cela fonctionne en opposition à Elafibranor de Genfit qui est un puissant activateur de PPARa et stimule le métabolisme des lipides.

En utilisant les données de l'étude de phase IIB, Genfit a également identifié deux groupes de patients, appelés Répondeurs et Non-répondeurs. Les intervenants ont donné une très forte réponse positive à Elafibranor en enregistrant réversion de Nash après 52 semaines. Les Non-répondeurs enregistrés très peu de changement au traitement, mais sont restés stables.

La différence entre ces deux groupes de patients réside dans leur constitution génétique, et par conséquent dans le métabolisme des lipides contrôle de l'expression génique et régulée par les micro-ARN.

En mesurant les profiles de micro-ARN spécifiques aux deux groupes circulant dans le sang, Genfit pourrait classer les facteurs génétiques qui contrôlent le métabolisme lipidique des patients avant le début du traitement et avant les biopsies. Cela donnerait un avantage significatif à Genfit pour classer la réponse des patients au traitement la Nash et pour définir des critères supplémentaires du régime d'alimentation afin de surmonter le handicap génétique des patients plus résistantes.



Full Text : 

Genfit recently announced that it had patented a method for defining the degree of severity of NASH in patients on the basis of a simple blood test. This diagnostic test was reported to be more reliable than that based on liver biopsies and would be beneficial to Genfit in selecting patients for the next Phase III study.

I will try to explain that this was more than a simple announcement related to a new patent application, but contained a strategic message intended for specialists only that Genfit had achieved a major breakthrough in understanding how Nash develops and how to treat it.

First of all I will try to explain how I interpret the announcement before looking at the science mentioned in the text and I will conclude with my analysis of what I believe is the real message that has been completely ignored by the market and the financial analysts. 

Genfit has not given details of its bio-marker research but we can make some informed guesses based on the company's motivation and the methods used. Genfit has stated several times that it has a large and competent bio-informatics team that analysed the phase IIB data and defined the programme to create the bio-marker algorithm for Nash. The result is a bio-marker method based on a multi-criteria blood test that defines a patient's Nash condition better than the 8 point NAS score based on liver biopsies. This is not really surprising since the NAS score is an oversimplified scale based on an unweighted and unprioritised sum of three liver conditions visually observed on the biopsy. It is sensitive to sampling errors, operator bias and does not take into account liver function.


This is what Genfit announced in September 2015 :


GENFIT has launched an R&D initiative in the field of NASH diagnosis biomarkers based on its expertise in transcriptomics applied to small circulating non-coding RNAs, in particular miRNA.

GENFIT has constituted a large bank of plasma samples from NASH patients with a liver biopsy. This patient cohort is extremely well-characterized (complete anthropometric and biochemical data, centralized biopsy reading), and covers a wide spectrum of NASH disease activity and severity.

After developing and validating a reliable method for the systematic measurement of different miRNAs in the plasma, GENFIT has introduced these measurements into the data set and challenged their diagnostic values versus over 70 variables using two independent biostatistical approaches. Two methods were used to generate thousands of cohorts from the initial patient population in order to mimic real-life NASH variability and assure the translatability of the results to the global NAFLD/NASH population. These parallel approaches independently identified the same two specific miRNA species within the top 3 most powerful diagnostic markers of NASH.

The two resulting algorithms combine the identified miRNAs and known markers of liver damage. A comparative study demonstrates that these algorithms are more powerful than existing scoring systems for the identification of NASH patients that deserve to be treated. New patent applications have been filed for the use of the technologies developed by GENFIT for the diagnosis of NASH.

 A new proprietary diagnostic tool will be used in the Phase 3 trial for Elafibranor.


What can we understand from the above statements ?

Whereas much of this analysis was based on patients from the phase IIB comprising about 280 patients, Genfit states that it has a large bank of plasma samples from extremely well characterised Nash patients. In order to build up thousands of cohorts (selected related groups with common characteristics), Genfit clearly has many blood samples from the same patients at different stages of the phase IIB trial as well as from patients at different stages of the disease who were not included in the IIB trials.

This data was tested in a computer model using 70 variables to describe Nash in all its stages and the model was refined (values and weightings for each variable were progressively adjusted by computer) to to give the best match of the model to the data. The aim here was to find the best model but also to identify which elements of the data were most sensitive to the model and which are not relevant. The final selected data included known liver damage markers and two specific miRNA species within the top 3 most powerful diagnostic markers of NASH. Two algorithms were identified as giving excellent results, which likely means that two data sets selected were not identical or had different weightings or ratios.


To whom was this announcement intended ?

Certainly not to the financial analysts or markets, nor to its shareholders since Genfit made absolutely no effort to make its announcement understandable to them. Indeed the market reaction was limited in both magnitude and in time. The share price registered a sharp rise (maybe based on an unknown market valuation for a biomarker kit in the distant future) and then fell back to pre-announcement levels. The financial analysts, being unable or too lazy to understand it, largely ignored it.

Indeed the full scope of this announcement could only be readily understood by specialist researchers in molecular biology and/or genetics, that is by researchers in universities or in companies like Genfit, its competitors, potential partners or the FDA. This was therefore a strategic announcement, but what was the strategy ? This can only become clear once we at least partially understand the announcement.

Having said that, there were enough indications in the text to enable a curious researcher to get a sense of the message Genfit was sending out.


This message is: “Genfit is now able to use micro-RNAs to classify Nash patients”

The new elements in this announcement relate to micro-RNAs or miRNAs of which two specific miRNA species were identified within the top 3 most powerful diagnostic markers of NASH. So what are micro-RNAs and why are they different from existing known markers of NASH ?  The following links give an insight into the relation between miRNAs and Nash.


1) Micro-RNAs are biological entities whose main function is to down-regulate gene expression

MicroRNAs regulate gene expression. A single miRNA can reduce the expression levels of hundreds of genes. The mechanism by which miRNA molecules act is through partial complementarity to one or more messenger RNA (mRNA) molecules, generally in 3' UTRs. The main function of miRNAs is to down-regulate gene expression. They can also up-regulate gene expression by down-regulating inhibitors to gene expression. 

Reference 1 (


2) Micro-RNAs are not simply markers of NASH as Genfit has announced, they are also essential players in the biological processes controlling health and disease.

This means that they regulate biological functions as opposed to simply signalling their condition.


3) Micro-RNAs down-regulate the nuclear receptor PPARa which in turn can down-regulate some miRNAs

a) MicroRNA (miRNA) expression profiling demonstrated that activated PPARα was a major regulator of hepatic miRNA expression. Of particular interest, let-7C, an miRNA important in cell growth, was inhibited following 4-h treatment and 2-week and 11-month sustained treatment with the potent PPARα agonist Wy-14,643 in wild-type mice. 

Reference 2 (  

b) miR-141 suppressed HBV replication by reducing HBV promoter activities by down-regulating PPARA. This study provides new insights into the molecular mechanisms associated with HBV-host interactions.

Reference 3 (


4) MiRNAs play a role in lipid metabolism through targeting PPARa

a) Peroxisome proliferator-activated receptors (PPARs) function as transcription factors regulating the expression of different genes. Most important roles PPARs play in higher organisms are regulation of cellular differentiation, development, and tumorigenesis. They are expressed throughout the body. Three types of PPARs (alpha, gamma, and delta) have been identified to date. Transcriptional activity of the various PPARs isoforms in physiological and pathological situations is regulated by numerous and complex epigenetic, transcriptional, and posttranscriptional mechanisms. Interestingly, a further level of complexity in these regulatory mechanisms has recently emerged and involves an epigenetic regulation of PPAR isoforms by microRNAs or vice versa, which can either degrade or repress PPAR mRNAs at the translational level. The interaction of PPARs and microRNAs in cells influences tissue and organ functions and also plays important roles in physiological and pathological processes in inflammation and cancer. MicroRNAs (small noncoding RNA of 18-24 nucleotides) are involved in gene expression by translational repression or cleavage of the mRNA targets. PPARs expression/activity in different cells/tissues could regulate expression of hundreds of microRNAs, which have potentials to act as oncomirs or tumor suppressor. It will be a very interesting aspect of study to reveal PPARs mediated regulation of microRNAs, as it could be used as future therapeutic weapon in cancer.  

Reference 4

b) Accumulating evidence supports the effects of miRNA in lipid metabolism, providing a potential linkage between certain miRNA and non-alcoholic fatty liver disease (NAFLD). We aimed to investigate the miRNA expression pattern in a steatotic L02 cell model and explore the function of certain miRNA target pairs.

Methods:The cell model was established by culturing L02 cells with a high concentration of free fatty acid. Micro-array and stem-loop reverse transcription polymerase chain reaction (RT–PCR) were utilized to detect dysregulated miRNA, whereas computational algorithms were used for target prediction. Real time RT–PCR, Western blot, luciferase activity measurement, and other techniques were employed for target verification.

Results:  Seventeen up-regulated and 15 down-regulated miRNA were found in steatotic L02 cells, while miRNA-10b was proven to regulate the steatosis level. Peroxisome proliferator-activated receptor-α (PPAR-α) was also found to participate in steatosis, as its protein level was decreased in steatotic L02 cells and its over-expression by transfection into the PPAR-α–pcDNA 3.1 vector could partially alleviate steatosis. We further found that PPAR-α is the direct target of miRNA-10b as it showed significantly changed protein expression, but a relatively unchanged mRNA level in steatotic L02 cells transfected with pre-miRNA-10b and anti-miRNA-10b. Moreover, the action of miRNA-10b on PPAR-α depends on the presence of a single miRNA-10b binding site, as the activity of a luciferase reporter carrying the mutant PPAR-α 3′-untranslated region was not reduced by the expression of miRNA-10b.

Conclusion:  The established miRNA profile of the steatotic L02 cell model and the novel effect of miRNA-10b in regulating hepatocyte steatosis may provide a new explanation of the pathogenesis of NAFLD.

Reference 5  ( 2009

So now we see that there is a close 2-way relation between micro-RNAs, gene regulation, PPARa lipid metabolism and Nash. PPARa is activated by ligands such as Elafibranor and unsaturated fats to control lipid metabolism but is down-regulated by micro-RNAs. In addition, activated PPARa can inhibit the action of some MiRNAs.


The dynamic equilibrium between food intake, micoRNAs and PPARa activity is key to controlling lipid metabolism and Nash.

If the above hypothesis is true, we can formulate the following mechanism whereby lipid metabolism and Nash are regulated by both external and internal factors

External factors raising lipid levels and Nash:  excessive lipid and sugar intake especially saturated fats and fructose), lack of regular exercise.

External factors regulating lipid levels and Nash: PPARa activators (Elafibranor, vitamin E, unsaturated fats, omega 3), regular exercise. 

Internal factors of genetic origin inhibiting lipid regulation:  micro-RNAs down-regulating PPARa activity.


What does this mean in terms of Genfit's biomarker results ?

Genfit used many cohorts to establish the best biomarker algorithm to create a diagnostic tool for Nash. In doing so it also was able to check out the results for two specific cohorts identified in terms of the response or not to Elafibranor.

These two cohorts have been labelled Responders and Non-responders.

Responders gave a very strong positive response to Elafibranor by registering reversion of Nash over the 52 week trial. Non-responders registered very little change to the treatment, but remained stable.

What is the difference between these two groups ?

The conditions of the trial imposed controls over the External factors so that these should have been largely equivalent in the two groups.

The difference then lies with the Internal factors, the gene expression controlling lipid metabolism and regulated by Micro-RNAs.

By measuring the levels and ratios of specific micro-RNAs circulating in the blood, Genfit is able to classify the Internal factors that control lipid metabolism of patients before starting treatment and certainly before performing biopsies.

To my knowledge Genfit has not stated which miRNAs will be used in the diagnostic test, but simply identified the two most specific miRNA species for diagnostic purposes from a large number that were tested. 

Is Genfit working on a Biomarker Test to classify patients according to their sensitivity to respond to treatment by Elafibranor ?


We now know however that Nash patients can be grouped into at least two different types “Responders and Non-responders”, so that by using the phase IIB data of the evolution of the miRNAs of responders and non-responders over 52 weeks, refinement of the computer model would generate different algorithms of markers compared to the standard Nash algorithm for the global diagnosis of Nash patients.

This analysis opens the way to classifying and selecting patients according to specific miRNA profiles, which correlate to their tendency to respond to treatment by Elafibranor.

It could lead to a protocol by which “responders” get a standard Elafibranor therapy whereas those classed as “non-responders” or partial responders would require a higher dose therapy combined with additional “external factors” such as severe dietary constraints to overcome their genetic handicap.

Could this be the strategic message that Genfit was really sending out to its competitors or potential partners? If so, it means that Genfit is very clearly leading the way in its understanding of how to care for Nash patients and is on the right track to providing the first effective drug to treat Nash

Albert Wright, October 2015


This document contains interpretations and extrapolations that are entirely the work of the author and may or may not be close to the real situation. They are however based on referenced published data and analysed in good faith according to the author's experience as a scientific researcher and author. The author invites the reader to make his own research and interpretation and to draw his own conclusions. The author is open to modification of this text after discussion with the reader. Please fell free to contact him on


The author is a small shareholder in Genfit.


WWW.NASHBIOTECHS.COM  -  Copyright G DIVRY 2015-2016  - Contact and TERMS OF USE