THE NASH MARKET (part 1- Segmentation)

orange-188082 1920




here is an approach for the segmentation of future NASH market, 

We will focus on North American and European markets in the first instance.

Countries included in the survey

The prevalence of NASH is still debated; according to the party, its goes now from 3% to 12% of the population in industrialized countries. The most advanced laboratories on the subject reported in their submissions a consensual prevalence of 12%, Gilead speaks of 10% 20%, The NIDDK him, always refers to a document, dated 2006, reporting a prevalence between 3 and 5%. 

The FDA for its part raised a prevalence between 12% and 17%.

The range is wide; I tend to think that the laboratories are generally optimistic about the market size they are targeting in their communications to investors.

However the lowest assessment from NIDDK date nearly 10 years and was not revised in light of recent studies.

To assess the market prudently, I chose an intermediate prevalence I estimated ¼ of patients with NAFLD (25%), so, according to published studies, a prevalence of NASH about 5% of the adult population.

This population is not homogeneous, all NASH are not necessarily of the same nature. There are NASH strongly linked to obesity, some do not. Some are associated with identified cardiovascular risk, others with type 2 diabetes. Some are associated with a combination of these three factors or correlated with a genetic typology.

It is important to understand this, because medications would target some typologies but will be effective only on one or more segments.

Furthermore NASH is considered the hepatic manifestation of the metabolic syndrome for which current treatments are limited.

NASH is a silent disease because it has no clear symptoms for much of its evolution and becomes apparent at the onset of symptoms associated with the disease it causes, such as liver fibrosis.

Therefore,  what the laboratories presents as a treatment for NASH; is often the treatment of only a single factor, or more symptoms thereof.

This further complicates the analysis of the potential market.

Sorting between laboratories whose molecule specifically targets the accumulation of fat in the liver, which is not yet NASH  but NAFL which can potentially evolve to NASH, those targeting liver cirrhosis and advanced fibrosis that are not a direct treatment of metabolic NASH but the consequences of the latter, and the 'pure NASH players’ curing the liver and metabolic derangement which is the main characteristic of NASH, is not easy.

So we should not deviate from the study the "NASH side players" targeting the upstream and downstream of the disease because, if spreads, there will be very few companies in the market studied, one or two..

Liver fibrosis is one of the monitored parameters in NASH patients,  although mortality generated progression to fibrosis is largely inferior to that related to increased cardiovascular risk.

This is quite understandable because it is easier to focus in clinical research on an objectively measurable factor in a relatively short time, as the advancement of fibrosis, rather than a multi factorial result, as the complex mechanism increased cardiovascular risk.

It is quite simple, based on histological observations, to classify patients according to the progress of their hepatic fibrosis on a scale of F0 to F4.  F0 meaning no fibrosis, and F4 meaning mainly cirrhosis.

Companies like GILEAD with GS4997, Conatus, Galectin or TOBIRA (ALLERGAN now ) thus concentrate their studies of patients with advanced fibrosis (F3) or cirrhosis (F4) from NASH. These companies present their drug candidate as a treatment for NASH, but in fact they target the symptoms of the final stages.

A sub market segmentation, integrating the different stages of fibrosis is therefore needed to take into account the specific typology of these targets.

So I tried to segment the market with the following parameters

  • Patients with NAFLD or NASH with a NAS score partitioned between 3 and 8.
  • Patients with cardiovascular risk proved or not.
  • Patients with T2 diabetes or not.
  • Patients divided according to their fibrosis score between F0 and F4


This ambitious segmentation, NASH divides the market into 140 segments, which is difficult to represent.

I tried to evaluate the prevalence of these 140 segments in the adult population, which is not easy, because no study to date has reached this degree of accuracy.

So i proceeded empirically, with the only evidence I could find.

In addition to many articles, each with statistics and sometimes, contradictory patients distributions, there is a base, the FLINT study, and the PIVENS study, although small, which can help.

On a population basis with NAFLD, with a prevalence of 25%, and a fifth who have NASH,  we can extract from the FLINT study  a segmentation of fibrosis scores.

But as the FLINT recruitment  departed cirrhosis, we must recover the cirrhosis rate in the NASH population in another study and allocate percentages of fibrosis scores on the rest of the population.

Once this is done, we have a distribution of fibrosis scores on the whole sample of NASH, with a NAS> = 4.

There is another published study that disaggregates fibrosis score between NAS <= 4 and those> 4 which shows that statistically fibrosis progresses with the worsening of the disease.

This correlation is stronger with the progression of inflammation and ballooning as steatosis, which is logical, but should have been further subdivided into 3 new segmentations, which would have burdened the exercise beyond the reasonable.

It was therefore necessary to somewhat ‘empirically' fill some lines so that, taking into account the horizontal split NASH population (score NAS), and vertical (fibrosis score), the assembly is consistent on two axes and reflects a progression of fibrosis in most advanced patients.

This method of distribution is partly empirical, but consistent, in row and column with the ratios of the population of FLINT.

We must also consider what concerns the important and accessory in the selected segmentations. For example, the segmentation of cardiovascular risk is only of interest because some companies (especially INTERCEPT) had to deviate from the recruitment of their Phase 3 patients with a specific cardio vascular risk, and patients with a rate not mastered high LDL, potentially closing much of their market.

Similarly, companies with products increasing insulin resistance should not be prescribed to patients with type 2 diabetes.

Conversely some drugs appear to be effective only in patients with diabetes (as seen in FLINT study)

To assess the prevalence,  and regarding the last studies published we considered a rate of 60% of NASH patients with identified cardiovascular risk. 

For diabetes rate I took the last published 75 % prevalence of diabetes in the NASH population.

We can push this figure by averaging between the two studies and establishing the supposed prevalence of diabetes in the NASH population to 46%, in my opinion, this ratio should grow up with the publication of other studies.

The recalculated distribution of prevalence in the adult population gives an empirical picture of the prevalence by market segment, but consistent with the known studies to date.



To continue in the market assessment logic, it is necessary to introduce a second parameter, which, too, has an empirical side: the health care rates.

multiples announcement of the development of diagnosis techniques greatly changes the estimated diagnosis rate of the disease.

Some methods of companion diagnostic biomarker ( like Genfit’s one )  can identify with a simple blood test patients with NASH and likely to respond to treatment companion.

The cost of the diagnosis should be reduced to a few hundred dollars at the beginning of its commercialization, to quickly go below $ 100, which give the very suitable for mass screening.

The method used will assess an estimated average rate of care for each market segment, in order to assess the number of patients in the respective predictable general case. This care rate will evoluate in time with a better knowledge of disease by the medical world.

It goes without saying that a patient already followed for cardiovascular risk and diabetes, is more likely to be detected and treated as another. Similarly, the higher the pathology progresses, clinical signs become important and induce a higher rate of diagnosis (cirrhosis having the highest treatment rate).

Conversely, a NAFLD, NASH with NAS or 3, silent in a patient with no cardiovascular risk or proven diabetes, is unlikely to be detected by a physician during a routine visit

The following table try to address these issues by providing a treatment rate per segment between 2% and 60%,  it is very cautionnous ! for example in France this rate for diabetes (patients under treatment / patients with a diabetes, regarding prevalence )  is over 76% !!

With prevalence and care rate, we can apply that to the targeted population, USA + 6 of the most advanced EU countries, total 600 M of adults and we can estimate segment by segment the population supposed to be treated.

The third component that we must assess in order to present a segmented approach to the market is the estimated annual cost of treatment.

The logic used here is that the acceptable cost for the treatment of distressed patients, such as those with cirrhosis, is much higher than the acceptable cost of chronic treatment with a silent disease.

but in very advanced disease , this cost should be compared with the one of a liver transplant estimated to 739 000 $ in 2014


Nevertheless, these figures have been lowered compared to  first analysis because of recent controversies related to the costs of medical treatments. Funders and political organizations are becoming very vigilant on new medication costs.

The cost of treatment taken into account is within a reasonable range of $ 20,000 to $ 8,000 per year for the treatment of cirrhosis and from $ 15,000 to $ 6,000 per year for other patients. Treatments that impact only NASH with no fibrosis are estimated  between $ 3,000 and $ 1,200 

Prices are in a large range because of the previsible evolution of treatment with the raising of the concurrence.


In a previous study we estimated the global market value by segment based on a fixed cost, but we modified our model to integrate the market dynamics.

The market will take time to maturate, and at the same time , many actors are supposed to progressively reach the market and will impact the prices as concurrence increase. 

It means that the market will evoluate smoothly in value, despite maturation.

On this segmentation basis we estimated in Part 2 a dynamic model applied on 12 drugs and timed market shares and value

Link to Part 2

Share on StockTwits

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