mathematics-1044089 1920

What happened on LDL levels during FLINT, for a non spécialist like me,  is still a kind of magic box regarding the data published.

As the adverse effects of OCA are screaned now to evaluate their potential impact on the chance for Intercept to reach the NASH market, it is interesting to try to understand the variations of lipid balance under OCA treatment.

This analysis is an interpretation of the known datas by a non specialist, this could lead to a misinterpretation of the datas, or a not accurate analysis because of missing datas. One can comment directly at the end of this page to amend or criticize this analysis.

We surely do not have all the datas on the cholesterol evolutions during the FLINT study.

Mainly we have :

  • The data published in the Lancet and the Annexes 
  • An Intercept Pharma News on a post hoc study including new data on the LDL evolution during the study.

The data presented in the post hoc analysis news are only on LDL so we will focus on the LDL levels evolution.

In the Lancet publication we can find in the annexes a graph presenting an average LDL levels evolution including all the patients of the study.

We do not have the exact population concerned on the figure , but ,as the figure  is just followed in the annexes by the AST levels figure mentioning the all population treated in the study (141), we consider that is the same for the lipids but in the table published in the article they mention 126 patients under OCA ? 

We can remark that the treated population average LDL level presented a peak just after the beginning of the treatment (OCA) and then decreased slowly and stabilized.

The data above are based on 126 patients under OCA and not the full population. it could be explained by the a sub selection or other reason but as many things in this study it is not clear why 141 patients were included in the AST graph and 126 only in the table figure.

One is maybe able to explain the 126 patients selection using the trial profile  below. I was unable to do it.

Nevertheless , the graph at week 72 is coherent whith the +0,22 value of average LDL level you can find in the table.

In april 2015 Intercept published a news 

Intercept Announces New FLINT Trial Data Showing OCA Treatment Increases Fibrosis Resolution and Cirrhosis Prevention in High-Risk NASH Patients

In this  news, a small paragraph is dedicated to the new finding on LDL levels in the post hoc analysis of the FLINT data’s.

 "The impact of statin use on LDL cholesterol was also evaluated in the FLINT trial population (n=283). In this post-hoc analysis, OCA-treated patients who initiated statins during the FLINT trial (n=26) experienced a rapid reversal of their observed mean LDL increase to below baseline levels, with a mean decrease after 72 weeks of treatment of -18.9 mg/dL. In contrast, other OCA-treated patients with no reported initiation or change in statin therapy experienced an increase in LDL that peaked at week 12 and was sustained over the 72 week treatment period. Patients treated with statins at baseline who maintained statin treatment over the duration of the study (n=50) experienced a mean LDL increase of 8.7 mg/dL at 72 weeks. Patients not treated with statins during the study (n=65) experienced a mean LDL increase of 16.0 mg/dL. Treatment related LDL increases in all groups reversed with treatment discontinuation. This post-hoc analysis suggests that the OCA associated LDL increase appears to reach a maximum peak and plateaus soon after initiation of therapy and that concomitant statin use in NASH patients receiving OCA may ameliorate any treatment-related LDL increases."


First remark, the population mentioned is the full population treated in the study (141 patients) but we dont know if some discontinued the treatment (8 mentioned and 2 death mentioned) and if yes, how they are integred in the 141 patients average levels , it is still a magic box for me but i am a neophyte.

Second remark, we discovered, with this news, that 26 patients in the OCA arm started a statin treatment during the study.

It is only 18% of the patients mentioned in the news but 25% of the patients included in the histological results published. It surely impact seriously the results, but no any word on it.

Third remark, the data are now in mg/dL,  not anymore in mmol/L, it will not disturb a scientist but could perturb a financial analyst … 

We mentioned that the initial figure published in the annexes on 141 patients  is coherent with the +0.22 published in the table on 126 patients. But now, a new effect of the magic box .. there is another distortion

at the end of the 72 weeks we have :

  • 26 patients with a new statin therapy who dropped their LDL to an average of  -18,6 mg/dL  soit -0,481 mmol/L
  • 50 patients under an pre existing statin therapy who upped their LDL to an average of  +8,7 mg/dL  soit +0,225 mmol/L
  • 65 patients with no statin therapy who upped their LDL to an average of  +16 mg/dL  soit +0,414 mmol/L

A basic pondered average method should give us an average evolution of the full population.

(26*-0,481 + 50*0,225 +65*0,414)/141 = 0,182 mmol/L

A smaller LDL increase than the data published initially in the Lancet ? what a magic box ..

the explanation could come from the different size of the population 126 vs 141 

the 15 patients added should have an average LDL decrease of -0,14 to explain the gap ! no way to understand what happened in the magic box during the post hoc study .. Intercept announced the data but did not publish it fully.

Their conclusion was the possible demonstration of pathway to reduce the aggravation of the cardiometabolic profile induced by the OCA, using statines. 

Indubitably, regarding the post Hoc data given in the News,  the patients under a statins treatment before the beginning FLINT study increased theirs LDL levels but less than patients with no statin treatment . ( +0,225 mmol/L vs +0,414 mmol/L)

but the OCA still increased the LDL levels

Intercept enhanced patients who initiated statins during the FLINT trial (n=26) experienced a rapid reversal of their observed mean LDL increase to below baseline levels.

 The animation below try to illustrate the evolution of patients LDL levels during the FLINT study

As we can see all the patients under OCA increased their LDL levels at the beginning, during the study, patients should have raised critical LDL levels and the physician following them decided, despite the -no change any treatment rule- required during a clinical study, to initiate a statin treatment. At this time there is a clear decrease of the LDL levels for those patients.


Regarding that, it is interesting to look again at the global evolution of  LDL levels published in the Lancet .


We can note a decrease of the AVERAGE levels of LDL in the ENTIRE population, INCLUDING those who initiated a statin therapy during the study and decreased their own average levels from +0414 to - 0,481.

We do not know when each patient initiated this new therapy but it is clear that the AVERAGE levels were impacted and this is maybe the explanation of the decrease of global average LDL levels during the study.

If those patients had respected the protocol and avoided to initiate an statin therapy the average levels would have been very different ..

Like :

  • 50 patients under an pre existing statin therapy who upped their LDL to an average of  +8,7 mg/dL  soit +0,225 mmol/L
  • 91 patients with no statin therapy who upped their LDL to an average of  +16 mg/dL  soit +0,414 mmol/L

The new average LDL level variation would have been +0,347 mmol/L,  it would have been 60% higher than the published one.

We do not know either what statins were used by patients, as they were only installed in the US one can imagine that they took one of the 6 statins: atorvastatin, fluvastatin, lovastatin, pravastatin, simvastatin and maybe rosuvastatin

However when you consider that the most prescribed statin in the US are the astorvastatin and simvastatin, existing statistics can be used and compare the theoretical decrease of LDL allegedly have these 26 patients, compared to that seen with OCA simultaneous treatment  .

according to a study published in 2003

Comparisons of Effects of Statins (AtorvastatinFluvastatinLovastatinPravastatin, and Simvastatin) on Fasting and Postprandial Lipoproteins in Patients With Coronary Heart Disease Versus Control Subjects 

The average decline observed in LDL in patients taking these statins is as follows

We see that if we take the intermediate statin dose, the decline in average LDL level observed with statins is of the order of - 1.08 mmol / L

Or patients who began treatment on statins during the study FLINT finally have obtained an average decrease of 0.48 mol / L, or less than half the gain normally expected. This shows the effect of the OCA, which is weighted by statins but who nevertheless neutralizes much of the expected effects.

OCA brings down the beneficial effect of statins:

  • 40% for low-dose statins
  • 56% for intermediate doses of statins
  • 59% for high-dose statins

This should have been constated  in the 96th week, after cessation of treatment with the OCA, but probably not of the statin therapy ..

These patients must have had a significant drop in their LDL levels, but as Intercept did not publish the numbers, we will not know.

It would be nice for Intercept to publish the full datas they presented on their post hoc analysis in a way to help the analysts to understand the cardiometabolic issues with reviewed data’s.

At this time, if we imagine that patients who initied statin therapy during the trial benefit from the full average benefit of statins on LDL levels (-1.08 mmol/L) it implie that the average LDL level of patients with 24 mg/ d OCA treatment but no any statin therapy increased to 0.6 mol/L 

This is fully coherent with the average +0,22 mmol/L  LDL increase of the entire population.

It means that the interest of a preventive statin therapy given to patient taking OCA is clear .. 

no statin therapy : +0.6 mmol/L of LDL

preventive statin therapy : +0.08 mmol/L of LDL

The statin therapy do not prevent the LDL to increase but minimize a lot the increasing.

This will maybe confirmed in the incoming studies 


This  time, it would not be acceptable to initiate statin therapy during the trial.

The best strategy would be to include a maximum of patient under a stable chronical statin therapy  in the trial. 

As we have seen, and if the data given by Intercept on their post hoc analysis are confirmed, their LDL levels increase only 1/7 the LDL levels of patients with no statin therapy.

Nevertheless the recruitment is supposed to represent the intent to treat population and if the ratios are respected, it will be difficult to propose at the FDA a recuitment with 90% of patients under statin therapy.  

Under an Half/Half the LDL increase expected could be just around+0,35 mmol/L

So Analysts should expect in the Phase 3 REGENERATE study  24 mg arm an average increase of LDL levels at 72 weeks higher than in the Phase 2b FLINT and it could be an issue for the Phase 3 recruitment . 

Share on StockTwits

WWW.NASHBIOTECHS.COM  -  Copyright G DIVRY 2015-2016