Genfit SA (GNFTF)

[Wildbilly]

Genfit: Misunderstandings And Propaganda About Phase 2b Results Provide A Good Entry Point - Genfit (OTCMKTS:GNFTF)

Summary

After the release of its phase 2b preliminary results in NASH, Genfit’s stock saw a nearly 45% decline following extremely negative interpretations of its results.
With a cool head and some analytical perspective, I believe that Genfit’s results fell victim to a widespread misinterpretation and offer great perspective in NASH.
When compared against its biggest competitor’s drug, Intercept’s OCA, Genfit’s GFT505 demonstrates very good efficacy results and a much better safety profile.
After Friday’s sell-off, which I believe is unjustified, Genfit still presents an excellent investment opportunity and current share price is a very good entry point.
Genfit (OTCPK:GNFTF), a French biotech listed on Euronext Paris, which stands in the race to develop one of the first treatments against non-alcoholic steatohepatitis or NASH, released on March 26 the preliminary results of its "GOLDEN-505" phase 2b clinical trial, conducted in 274 patients.

The complete press release, which can be found here (pdf file), gives some very important indications on Genfit's treatment, GFT505. Here are some important extracts (highlights are added):

Due to the unexpected rate of resolution of NASH in patients randomized to placebo who had early NASH (NAS of 3, placebo response rate >57%), along with the high number of sites for a limited sample size, the study as initially designed did not enable the trial to meet directly the primary endpoint. With correction for this baseline severity and site heterogeneity by a standardized statistical analysis, GFT505 120mg meets the primary endpoint: Reversal on NASH without worsening of fibrosis, as detailed below.

Treatment with GFT505 provides a significant beneficial effect on the primary endpoint (GFT505 120mg vs placebo, p=0.016, RR=2.03) in the global randomized population (n=274, full analysis set), where patients without an end of treatment biopsy were considered as non-responders. The primary endpoint was also achieved in the evaluable population of patients who underwent both baseline and end of study liver biopsies (n=237, ITT; p=0.027 vs placebo; RR=1.94). In the evaluable patient population, GFT505 120mg also has a beneficial effect of a decrease of NAS-score =2 (p=0.04 vs placebo).

Early NASH patients with NAS=3 were not included in other recent NASH trials. If the same is done in the GOLDEN-505 study, keeping patients with more severe disease defined by NAS=4 (n=202), GFT505 120mg demonstrates a doubling of responders on the primary endpoint (22.4% vs 12.7%, p=0.046, RR=1.9), thus providing evidence of a clinically meaningful benefit in patients with more advanced disease.

(...)

Even on top of various standard of care therapies, GFT505 provides additional improvements vs placebo on cardio-metabolic risk factors, commonly found in NASH patients:

-lipid profile: TG, LDL-C, HDL-C

-glycemic indices/insulin resistance in Diabetics: HbA1c, FPG, Fasting insulin

-inflammatory markers: Haptoglobin, Fibrinogen, CRP

The safety assessment of this one-year study demonstrates a very favorable profile, which is consistent with the conclusions of the DSMB reviews throughout the study. There were no cardiac events, signal on cancer, nor death in the GFT505 treatment groups. Weight remained stable, and no signal for edema was observed. A mild dose dependent increase in creatinine was noted (< 5%; GFT505 120mg vs placebo), which is a known reversible effect of GFT505. The most common adverse events were of gastrointestinal nature of mild intensity.

So why then would the stock plunge nearly 45%, wiping off $550 million of market cap in one single trading session on the day after this quite enthusiastic press release?

The trick here lies in the very first paragraph of the announcement: in the patients that were only moderately ill (patients with a NAFLD Activity Score or "NAS" of 3) there was such a huge placebo effect (close to 60%) that those results alone would have caused to study, as initially designed, to fail.

So, the shortcut from here is big as a boulevard and truckloads of "analysts" and press agencies did not hesitate to go down this obvious path, declaring Genfit's study "a failure" and some even resorting to thinly-veiled propaganda. There's no secret here: at the moment, it all comes down to a battle of numbers and PR between the two first-tier contenders in the NASH race that are Genfit and Intercept (NASDAQ:ICPT), which published in November 2014 the results of its phase 2b "FLINT" trial in which it tested obeticholic acid or OCA in NASH patients.

But since this rough trading day is now behind us, let's have a cold, hard look at facts and draw relevant comparisons where it is needed to provide some balance and set the record straight about Genfit's results. My analysis, detailed in 6 points, is provided hereunder.

Useful links:

- Genfit's GOLDEN-505 preliminary results.

- GOLDEN-505 clinical trial

- Intercept's FLINT results (Lancet publication)

- FLINT clinical trial

1. The placebo effect and revised population are coherent with other NASH trials

That is the spark that ignited it all: isn't it cheating to exclude a certain patient population when such patients would otherwise render the study unable to meet "directly" the primary endpoint? Let's look at the facts.

First of all, to understand the following explanations, you have to know the basics of NAFLD histological diagnosis. Nonalcoholic Fatty Liver Disease (NAFLD) or hepatic steatosis is the disease stage that precedes NASH (nonalcoholic steatohepatitis). Basically, it means that the patient's liver is saturated with fat (at least 5% of hepatocytes presenting visible lipid deposition). NASH is only diagnosed when, in addition to NAFLD, both inflammatory infiltrates as well as ballooning and liver cell injury are present. The severity of NASH is then determined according to a numerical score, which is called the NAFLD Activity Score or NAS. Accordingly, NAS is the sum of the separate scores for steatosis (0-3), hepatocellular ballooning (0-2) and lobular inflammation (0-3), with the majority of patients with NASH having a NAS score of =5. NASH patients all have a NAS comprised between at least 3 (the less ill patients) and up to 8 (the most severely ill patients), with at least a score of 1 in each of the three indicators (steatosis, ballooning and inflammation). (source: Medscape)

With that in mind, let's get back to the exclusion of "early-NASH" patients who were only marginally affected by the disease (NAS=3) in the revised analysis of Genfit's GOLDEN trial. While the initial patient population was 274, only 237 patients underwent the two compulsory liver biopsies that were defined as the primary assessment method for the NASH-reversal endpoint, and excluding all NAS=3 patients, the final population included in the analysis was 202 patients. So, by excluding only early-NASH patients, Genfit still has 85% of evaluable patients available for its final analysis after 52 weeks: that still is a significant patient sample and it is sufficient to validate the data obtained from this population -- for comparison, in the FLINT study the number of patients who actually completed the final 72-week biopsy was 200.

Then, the placebo effect in those "early-NASH" patients was extremely high (close to 60%). What this means above all else is that this less severely ill patient population does not seem to require any kind of specific medication to experience improvements in their disease state, be it Genfit's or any other -- a healthy diet and some exercise might be just enough. This is obviously a small flaw in the French biotech's trial design, as this should have been anticipated by excluding those patients right from the start -- just as Intercept did it, by accepting only patients with a NAS of 4 or greater in its FLINT trial. So, to put it simply, Genfit's revised analysis only aligned its target population to match the one of Intercept's FLINT trial. If nothing else, that makes results even easier to compare.

On the other hand, of the very few actual subgroup numbers released by Genfit's management during March 26 conference call, CEO Jean-François Mouney stated that in the most severely ill patients (NAS >6), where a placebo effect of 0% was observed, a response rate of 15% was confirmed in NASH patients with the 120mg dose, further confirming the efficacy of the drug in this target population.

Besides, Genfit's GOLDEN trial was conducted in 56 international centers in Europe and in the US, whereas Intercept's FLINT study only recruited from 8 US centers -- a much more homogeneous population for Intercept but a much more robust sample for Genfit. Even so, Intercept itself concedes in its Lancet publication that the FLINT trial "showed more favorable OCA-mediated results on NASH resolution when excluding patients found not to have had NASH at baseline". So, basically, everybody seems to have better results when excluding those patients who were not severely affected by the disease.

Furthermore, this last sentence should raise on itself some serious questions about Intercept's data. Indeed, it should be noted from Intercept's last investor meeting at the AASLD Colloquium that the company itself admitted that "two independent observers (...) made a determination of whether a patient actually did or did not have NASH and according to these two experts Elizabeth Brunt and David Kleiner about 20% of the patients actually didn't have NASH according to their criteria so we have removed those patients who had crazy NASH or didn't have NASH from the study from these analysis". If you understand this correctly, it means that not only did Intercept actually include nearly 20% of patients in its trial that were not formally diagnosed with NASH (in total contradiction with its inclusion criteria, so what should the FDA think about that?), but in some of its post-hoc analysis it did just the same as Genfit by excluding those "crazy NASH" patients to embellish its data.

So, the bottom line is that these results themselves should not in any way be worrying as the most important target population has always been the more severely ill patients and those results seem to corroborate those of other NASH trials.

2. The standardized statistical analysis was discussed with the FDA beforehand and will guide the design of the phase 3 trial

Another argument used by Genfit's detractors is the purported "mess" created by a post-study change of the statistical analysis plan. While the details of the analysis are not known yet, making those affirmations nothing more than conjectures, Genfit's press release clearly states that "using the initial protocol analysis, statistically significant improvement of the following liver related biomarkers was noted: decrease of ALT, GGT, ALP, and improvement on various NAFLD composite scores (Steatotest, Fibrotest, Fatty Liver Index, NAFLD Fibrosis score)".

And since from the start, Genfit has been in close contact with the FDA to discuss clinical trial designs and statistical analysis plans -- and even more since the agency granted Fast Track designation to GFT505 in 2014 -- Genfit explicitly confirmed during its conference call that the applied statistical plan was discussed and agreed upon with the FDA before the collection of the results. So again, why would this adjustment be a problem if statistical power is not overly diminished from the somewhat smaller population?

Basically, one of the main purposes of phase 2 trials is to evaluate the relevance of dosing and target populations before starting bigger, more expensive phase 3 trials. That is exactly what Genfit is doing: while confirming the company's intention to start a phase 3 trial in NASH before the end of 2015, Genfit now has collected a precious amount of data that will help it design a more targeted and efficient late-stage study. Before releasing all the detailed data to the public, Genfit anticipates the publication of a manuscript in a major scientific journal, which demands to place a partial embargo on the release while review is ongoing. Nothing weird in this, really -- after all, Intercept "only" took about 10 months to publish the full results of the FLINT trial, and Genfit anticipates a more detailed presentation at the EASL congress in April this year, in about a month from now...

3. On a similar patient population, GFT505 demonstrates very good efficacy results compared to Intercepts' OCA…

So, back to NAS scores defined above in this article: in the FLINT study, primary endpoint was defined as "no worsening in fibrosis and a decrease in NAFLD Activity Score (NAS) of at least 2 points", which basically means that patients would only need to improve on their current disease stage rather than being effectively cured. For example, severely ill patients could improve from an overall score of 8 to 6 and be counted as "responders" in the FLINT trial, while still suffering from steatosis, ballooning and inflammation.

On the other hand, GOLDEN-505 primary endpoint was defined as the "disappearance of steatohepatitis (i.e. patients no longer meeting the criteria for steatohepatitis) without worsening of fibrosis", meaning that patients would not be considered to suffer from NASH anymore, as they would have a 0 score in at least one of the three NAS criteria described above.

While this frontal comparison is admittedly a bit simplistic, the point was even made by some analysts, prior to the announcement of Genfit's results, that the company's "disappearance of NASH" endpoint was too restrictive and would be significantly harder to reach than OCA's 2-point improvement...

However, not only does GFT505 120mg demonstrate a near doubling of responders on the primary endpoint (22.4% vs 12.7%, p=0.046, RR=1.9), but in a nod to Intercept's results, Genfit also specifies that "GFT505 120mg also has a beneficial effect of a decrease of NAS-score =2 (p=0.04 vs placebo)".

From the Lancet publication, the FLINT trial demonstrated that 45% patients in the OCA group had reached primary endpoint compared with 21% of patients in the placebo group (relative risk 1.9, 95% CI 1.3 to 2.8; p=0.0002). In other words, OCA "improved" NASH scores twice more than in placebo patients, while GFT505 actually "cured" nearly the same proportion of patients, relative to placebo.

What's even more telling is that one of Intercept's trial secondary endpoint was actually "NASH resolution" -- which is as close as you can get from the GOLDEN primary endpoint -- and those results, while displaying a numerical trend in favor of OCA, were actually not statistically significant (22% vs. 13%, p=0.08).

Finally, Genfit reports that GFT505 efficacy is proven to be dose-dependent (i.e. better results in the 120mg arm than in the 80mg arm), which provides supplementary pharmacological proof that the drug's efficacy is not due to chance. The same cannot be said from the OCA phase 2 trial as there was only one tested dosing against placebo.

The bottom line is that, according to available efficacy data at this point in time, there is absolutely no reason to think objectively that OCA results in NASH would be any better than GFT505, and blatantly stating the contrary can only be interpreted as a distortion of scientific facts.

4. … but Genfit's treatment has a much better safety profile!

While OCA's problems with bad cholesterol and other drug related side-effects such as pruritus (or itchy skin) are well-documented -- pruritus was confirmed in 23% of patients receiving OCA vs. 6% in the placebo arm, which is significantly higher -- Genfit's drug, given on top of other standard of care therapies, actually provides additional improvements on cardio-metabolic risk factors, including lower levels of bad cholesterol (LDL-C) and insulin resistance in diabetics. All adverse events in the study were mild, with the most common being of gastrointestinal nature and of mild intensity.

On the other hand, in the FLINT study, while most severe adverse events in OCA-treated patients were judged to be unrelated to therapy, five severe or life-threatening adverse events in the patients receiving OCA were judged to be possibly related to the treatment (3 severe pruritus, 1 hyperglycaemia and 1 dysarthria and dizziness possibly due to cerebral ischaemia).

Even though, according to recent findings, Intercept's investigators seem to be suggesting that patients taking OCA could add statins to the mix to effectively control their cholesterol levels, Genfit's overall advantage is nonetheless undeniable on the safety front. So really, when presented with two effective NASH treatment options, when patients will be given the choice between the pills that will give them itchy skin and increase their bad cholesterol, or those which will actually improve their cholesterol and heart condition at the same time that it cures their disease, will it be a hard, thoughtful choice?

5. And then, there's the reversal of fibrosis…

To this point, as with all the detailed data from the other histological biomarkers, there's no specific mention, yet, of the rate of improvement in fibrosis scores from the GOLDEN study. However, Genfit's preliminary announcement mentions that "statistically significant improvement of the following liver related biomarkers was noted" including "NAFLD Fibrosis score". More details will probably be released in the coming months.

Besides, even if GFT505 has previously demonstrated anti-fibrotic properties in animal models, as fibrosis is basically a form of scarring, 52 weeks is a short period of time to actually witness such reversal through biopsies. Genfit's management estimates that, whatever the treatment is, longer studies (18 to 24 months) should be conducted to really confirm this positive effect -- however, additional data should shed some light on this issue.

6. … but what about the FLINT trial limitations?

To be fair, as we don't have access to the detailed results or the peer reviewed publication of the GOLDEN results, it is impossible to avoid all speculations about the consequences of Genfit's phase 2b trial design. There could be some flaws, and there could arise some questions at the reading of the complete publication.

However, about what we do know on Intercept's trial design, it is interesting to read what Dr Vlad Ratziu, Professor of Medicine at the Pierre and Marie Curie University and leading world expert in NASH -- as recognized by Intercept itself since a quote of Dr Ratziu was included in its own November announcement -- has to say about the design of the FLINT trial:

Although obeticholic acid had an effect on the primary endpoint, it did not cause a significant proportion of patients to cross the threshold from a histological diagnosis of definite or borderline non-alcoholic steatohepatitis to a diagnosis of not non-alcoholic steatohepatitis. One explanation is that the improvement in the NAFLD activity score might reflect a decrease in the severity of disease but not to the point of resolution of non-alcoholic steatohepatitis. (...)

The benefits of improving the histological features of non-alcoholic steatohepatitis and reducing serum aminotransferase concentrations without suppressing disease activity to the point of resolution need to be shown.

(source: Vlad Ratziu, "Starting the battle to control non-alcoholic steatohepatitis", The Lancet Journal, November 2014)

In clear, while Dr Ratziu rejoices at the significant efficacy results demonstrated in the FLINT trial, he also raises the question whether the initial primary endpoint chosen by Intercept (a reduction of 2 points in NAS) will provide enough benefits to actually improve significantly patients' health (i.e. "curing" them of NASH), and whether the improved histological features (i.e. the parameters which Intercept chose to monitor) are actually relevant to conclude that the disease activity has diminished enough.

On the other hand, what the GOLDEN trial has proven is that GFT505 effect is dose-dependent, and that it has an excellent safety profile -- clearly opening the way for longer, chronic treatments. So while Intercept is evaluating smaller dosing of OCA in some of its trials to compensate for treatment-induced adverse events (severe pruritus mainly), Genfit could actually design a phase 3 trial testing at least one treatment arm with a dosing of more than 120mg, with the perspective of increasing even more the already demonstrated efficacy -- especially since the safety of GFT505 in healthy volunteers is already proven up to at least a 300mg dosing.

CONCLUSIONS

Obviously, details of the GOLDEN trial are not available yet, and all data will need to be carefully analyzed upon release.

However, I not only believe that investors' initial reaction on Friday (after the announcement of Genfit's result) was an absolute overreaction, but I am also convinced, after thoroughly reviewing available data, that GFT505 definitely hinted at superiority over OCA in the severely ill patient population. Even so, I believe that there are enough opportunities in the NASH market to support commercialization of several drugs, and that even if OCA succeeds in getting market approval, Genfit's drug is currently worth a significant share of this global market, estimated by some at $35 to $40 billion by 2025.

Therefore, I believe that current "post-shock" valuation of around $30 per share provides an excellent entry point, before investors gradually realize the terrible misunderstandings, and sometimes propaganda, that led to this stock action on Friday.

Current price targets, with phase 2b efficacy demonstrated, stand at around $100 per share -- see details in this excellent SA article -- meaning a nearly 250% upside on Friday's share price. With the intention to start a phase 3 before the end of the year, Genfit's management should be quick to discuss with the FDA and provide additional details of the study, meaning that revaluation could happen very fast -- I anticipate that before year's end and at least upon announcement of the start of a phase 3 trial, a $100 price per share should be reached.

Risks

As always with development-stage biotechs, there exists a risk that clinical trials could fail and/or that the FDA could refuse the start of a phase 3 clinical trial, or demand additional studies before authorizing further trials, which could drive the stock further down. Besides, Genfit will need to finance further clinical trials and has not, at present time, enough cash on its own to fund a complete phase 3 trial in NASH -- partnering is in discussion, but should the company decide to go on its own, it will need to raise additional capital, which could lead to some stock dilution.

However, I believe that in the current situation, scientific reason will prevail over investors' feelings and PR battles and that Genfit has now significant data to proceed with its trials and bring its product to the market.


http://seekingalpha.com/article/3038226-genfit-misunderstandings-and-propaganda-about-phase-2b-results-provide-a-good-entry-point?auth_param=vfef:1ahlan0:42a2546c16ffc18c49408d4283843d59&uprof=44