Why The FDA Got It Wrong And Why It Will Approve Vascepa For ANCHOR SNDA Submission

Nov.19.13 | About: Amarin Corporation (AMRN)

Disclosure: I am an academic physician and a clinician scientist with on-going clinical and research activities in obesity, lipid and inflammatory disorders, and health outcomes. I have served as both a member and chair of NIH and VA grant review study sections and have reviewed over 1,000 research grants during my academic career.

I. Overview

In this article, I examine the merits of the FDA's controversial decision to rescind Amarin Corporation, PLC's (NASDAQ:AMRN) ANCHOR special protocol assessment (SPA) agreement. On October 23, 2013, the FDA unilaterally rescinded the ANCHOR SPA by declaring that a "substantial scientific issue essential to determining the effectiveness of Vascepa in the studied population was identified after testing began." The FDA specifically cited that the results from the three recent outcome trials, the ACCORD-Lipid, the AIM-HIGH, and the HPS2-THRIVE, "fail to support the hypothesis that a triglyceride-lowering drug significantly reduces the risk for cardiovascular events in statin-treated patients with mixed dyslipidemia and residual high serum triglyceride levels (200-499 mg/dl)".

The purpose of this article is to provide an in-depth review and analysis of the above clinical outcome trials cited by the FDA as the reason for rescinding the ANCHOR SPA. It is intended to educate the readers as to the adequacy and fairness of the FDA's review/analysis; and finally, to determine whether the FDA's decision to rescind the SPA was justified based on the "substantial scientific issue" argument.

The major conclusion of this analysis is that the FDA erred in its review and in its interpretation of the ANCHOR results. There were important omissions of data, and leap-of-faith extrapolation from those trial findings, well beyond the accepted standards of evidence-based medicine. This led to the FDA's unsubstantiated and faulty conclusion that lowering high triglyceride (TG) levels would not lead to clinical benefit. In the US, the NIH CTEP national guidelines group TG levels partly on the basis of CVD risk: normal (< 150 mg/dl); borderline high (≥150 and <200 mg/dl); high (≥200 and <500 mg/dl); and very high (≥500 mg/dl). Apparently, not noted by the FDA, most of the patients enrolled in the 3 trials had either normal or borderline high TG levels. These trials were neither designed nor powered to detect clinical benefits of triglyceride-lowering therapy in patients with defined high TG levels (≥200 and <500 mg/dl); that is, the ANCHOR study population.

The fundamental flaw in the FDA's analysis was the assumption that failure of triglyceride-lowering therapy in trials where the majority of patients were in the low to borderline TG range, could somehow translate to demonstrating that triglyceride-lowering treatment in patients with high TG levels would not be beneficial. The FDA conclusion is wrong, and cannot be derived from the above outcome trials having mostly normal or borderline TG patients. Thus, the FDA reason for rescinding the SPA has absolutely no scientific merit.

Unexpectedly, the most relevant findings culled from these reports, as they relate to the ANCHOR SPA, is the information obtained from the patients' baseline TG levels. The post-hoc sub-group analysis, based on these TG levels, indicates that patients with normal or borderline TG levels do not benefit from triglyceride-lowering drugs. On the other hand, the sub-group of patients covered by the ANCHOR SPA, i.e., those with high TG levels and low HDL (GOOD)-cholesterol, had a marked reduction in CVD complications, defined as nonfatal myocardial infarction (MI), nonfatal stroke (CVA), or death. The 28% (ACCORD) and 37% (AIM-HIGH) decrease in CVD complications in these trials were quite compelling. Thus, in direct contradiction to the FDA opinion, these results indicate that TG lowering therapy was highly beneficial in these patients.

The final conclusion of this analysis is that the FDA decision to rescind the ANCHOR SPA was not justified and not supported by the three clinical trials. On the contrary, sub-group analysis of these trials provides compelling evidence that the patients covered by the ANCHOR SPA, those with high TG and low HDL-cholesterol levels, had the greatest clinical benefit from triglyceride lowering therapy.

As a clinician scientist who has dedicated his life to advancing medical science and discovering new medical therapies to save lives, I am deeply concerned with this inappropriate and unjustified action taken by the FDA. The FDA action, if it stands, is likely to be harmful to the public's interest and possibly result in thousands of lives being lost that may have been saved by this highly efficacious, safe and preventive therapy. I am optimistic, and believe that, on closer inspection, the FDA will reverse its decision to rescind the ANCHOR SPA, and approve label-expansion for Vascepa for patients with high TG levels. The scientific evidence from the outcome trials cited by the FDA clearly show that TG lowering in the ANCHOR sub-population with high TG and low HDL-cholesterol leads to a reduction in CVD complications.

II. Background

A. The Company

Amarin Corporation, with global headquarters in Dublin, Ireland, and US Headquarters in Bedminster, New Jersey, is an advanced clinical stage pharmaceutical company with an emphasis on treating lipid disorders. The long-term objective is to extend its expertise in lipid management to reduce CVD risk. In July of 2012, the United States Food and Drug Administration (FDA) approved the company's leading product, Vascepa for the treatment of very high triglycerides (> 500 mg/dl), otherwise known as the "MARINE" indication. These patients are very vulnerable to cardiovascular events (CVE). Major pharmaceutical companies, such as GlaxoSmithKline (NYSE:GSK) have approved products such as Lovaza for similar indications. An estimated four million Americans are in the very-high TG category, and treatment costs for CVD run in the billions of dollars.

As noted above, the long-term plans of the company are to serve the "mixed dyslipidemia" market, described below. An estimated 40 million Americans, and up to 120 million individuals worldwide, are in the high-triglyceride category. To this end, the company submitted a supplemental NDA (sNDA) for ANCHOR in February, 2013 for an expanded-label use approval for Vascepa® (icosapent ethyl) capsules as an adjunct to diet in the treatment of adult patients with high triglycerides and mixed dyslipidemia (two or more lipid disorders).

B. The ANCHOR Clinical Trial

The ANCHOR trial was a "multicenter, placebo-controlled, randomized, double-blind 12-week study," which investigated the efficacy of Vascepa ultra-pure eicosapentaenoic acid (EPA) (AMR101), at doses of 2 gm and 4 gm, for the treatment of high TGs in 702 patients with mixed dyslipidemia on background statin therapy who were at high risk of cardiovascular disease " (Amarin Corp. press release, 2013).

"The ANCHOR trial met its primary endpoint, defined as the median percentage change in TG levels from baseline after 12 weeks of treatment. This endpoint was achieved at doses of 4 grams and 2 grams per day with median placebo-adjusted reductions in TGs of 21.5% (P<0.0001 value) for 4 grams and 10.1% (P=0.0005) for 2 grams." The efficacy of pure EPA (Vascepa) was dose dependent, with 4 gm/day having significantly greater effect than 2 gm, showing that the higher dose was necessary for optimal TG lowering.

The Vascepa treatment also resulted in a "statistically significant 6.2% reduction in LDL (bad) -cholesterol in patients optimally treated with statins". The reduction in LDL-cholesterol was an important beneficial effect of Vascepa, as other triglyceride-lowering drugs, including fenofibrates (numerous manufacturers) and Lovaza (GlaxoSmithKline) increase LDL-cholesterol. Vascepa had no identifiable side effects compared with placebo.

C. FDA Justification for Rescinding ANCHOR SPA

Per FDA regulations, SPAs are binding agreements and can only be broken when a company fails to comply with the provisions of the agreement or "if the [FDA] director of the review division determines that a 'substantial scientific issue' essential to determining the safety or efficacy of the drug has been identified after the testing has begun". In the case of sNDA submission for ANCHOR indication, Amarin fulfilled all aspects of SPA, including the pre-defined primary endpoint, i.e. lowering TGs in patients in the high TG category. (Amarin Corp FDA Briefing Document, 2013).

This surprising and controversial action by the FDA has created a great deal of confusion and concern, and question whether such an action, as I investigate herein, was justified. The FDA's unilateral dissolving of a binding SPA sets a troubling precedent, and could have important ramifications for the FDA, the pharmaceutical industry and the public.

III. What We Know About Hypertriglyceridemia and Risk of Cardiovascular Disease

Many large epidemiologic studies have shown a correlation between elevated TGs and increased risk for cardiovascular disease (CVD) (Sawar, Circulation, 2007; Tirosh, Annals of Internal Medicine, 2007; Patel, Circulation, 2004; Liu, Lipids Health Disease, 2013). Studies from around the world have found the triglyceride-associated CVD risk to be present in all races (Jeppesen, Circulation, 1998; Patel, Circulation, 2004). Other studies also confirmed it in both genders. The aggregate studies showed elevated TG levels as an important risk factor, both in healthy subjects and those with an underlying CVD risk. (Tirosh, Annals of Internal Medicine, 2007; Liu, Lipids Health Disease, 2013; Honkanson, J Cardiovascular Risk, 1996). Most importantly, these studies demonstrated that the increased CVD risk was directly proportional to the TG level elevation. A 2013 meta-analysis by Liu et. Al. was a huge study involving 726,030 patients. It determined that every 10 mg/dl increase in TG level predicted a 1.4% increase in CVD risk (Liu, Lipids Health Disease, 2013). Recognizing the importance of TG as a risk factor for CVD disease, major medical societies, including the American Heart Association (AHA), American Diabetes Association (ADA), American Association of Clinical Endocrinologists (AACE), and American College of Cardiology (ACC), recommend treatment of patients with TG levels over 200 mg/dl. Current guidelines recommend lifestyle modification, i.e., diet, weight loss, and exercise. If the TG level still remains above 200 mg/dl after lifestyle modifications, a treatment with triglyceride lowering drugs (TLDs), including fibrates, niacin, or omega fatty acids, should be considered. The ANCHOR SPA addresses this patient population exactly.

IV. Analysis of the Clinical Trials Cited by the FDA

In this section, I review the results of the clinical trials cited by the FDA (ACCORD-Lipid, AIM-HIGH, and HPS2-THRIVE trials) to see whether their outcomes provide the "substantial" new scientific evidence claimed by the FDA, to justify rescinding the ANCHOR SPA. The key question is whether or not lowering TG in patients in the ANCHOR sNDA range (≥200 and <500 mg/dl) would not be clinically beneficial.

The specific questions that will be addressed in this review for each of the three clinical trials are:

1) What specifically did the clinical trial investigate, and what was its primary endpoint;

2) Did the study examine whether lowering TG in the ANCHOR population leads to a reduction in CVD risks;

3) Did the trial outcome support the FDA assertion that the trial results provided substantially new scientific evidence that triglyceride-lowering therapy would not improve clinical outcomes in patients with high TG levels.

A. ACCORD-Lipid Clinical Trial:

What specifically did the clinical trial investigate, and what was its primary endpoint?

To determine whether the "addition of fibrate to statin therapy compared with statin monotherapy alone reduces the risk of CVD in patients with type 2 diabetes mellitus (T2D)." (Ginsberg HN, NEJM, 2010). The primary outcome was whether or not fibrate therapy reduced the risk of CVD. The reported outcome showed that fibrate therapy did not reduce the risk.

Did the study examine whether lowering TG in the ANCHOR population leads to a reduction in CVD risks?

No. This clinical trial was not designed or pre-specified to investigate triglyceride-lowering drug (TLD) effects on patients with high TG levels. In this trial, patients were enrolled without consideration of their TG levels, and the enrolled patients had a median TG level of 162 mg/dl; and most of the patients had TG in the normal or borderline ranges. Thus, the primary endpoints were non-informative. When a sub-group analysis was performed to determine whether fibrate therapy had clinical benefit in patients with differing TG levels, it was found that fibrates did not reduce CVD complication in patients with normal or borderline high TG levels (<200 mg/dl). However, patients that had high TG (>204 mg/dl) and low HDL (<34 mg/dl) levels (the sub-population covered by the ANCHOR SPA) had an impressive 28% reduction in CVD risks, prompting the investigators to conclude that the patients with high TG level (> 204 mg/dl) and low HDL level "appeared to benefit from fenofibrate" therapy.

Did the trial outcome support the FDA assertion that the trial results provided substantial new scientific evidence that triglyceride-lowering therapy would not improve clinical outcomes in patients with high TG levels?

No. As the study population included mostly patients with normal and high TG levels, no conclusion can be derived as to whether fibrate therapy would not benefit patients with a high TG (HTG) level. Directly relevant to the ANCHOR SPA, the sub-group analysis showed that fibrate therapy produced a 28% reduction in CVD risk in patients with HTG (>204 mg/dl) and low HDL-cholesterol (HDL-C).

B. The AIM-HIGH Clinical Trial

What specifically did the clinical trial investigate, and what was its primary endpoint?

The primary objective of the AIM-HIGH trial was to examine whether treatment with an extended-release niacin-type TLD leads to a reduction in CVEs in patients with pre-existing CVD (The AIM-HIGH investigators, NEJM, 2011). Complications were defined as death, nonfatal MI, CVD, hospitalization for an acute coronary syndrome, or coronary or cerebral revascularization. The rationale for the study is as follows:

In patients with established cardiovascular disease, residual cardiovascular risk persists despite the achievement of target low-density lipoprotein (LDL) cholesterol levels with statin therapy. It is unclear whether extended-release niacin added to simvastatin to raise low levels of high-density lipoprotein (HDL) cholesterol is superior to simvastatin alone in reducing such residual risk.

Thus, the primary intent was to examine whether raising HDL-C reduces complications in patients with established CVD, and did not study lowering TGs per se. A total of 3,414 patients were randomized to receive niacin (n= 1,718) or placebo (n= 1,696). The study had to be abruptly terminated after 3 years, as patients on niacin therapy had an unexpected increase in CVAs. At the time of the trial termination, niacin therapy did not reduce CVEs in the study population.

Did the study examine whether lowering TG in the ANCHOR population leads to a reduction in CVD risks?

No. The study was neither designed, nor pre-specified, to assess whether TG lowering would reduce CVD risk in patients with HTGs, for similar reasons as in the ACCORD study. The failure of niacin in the AIM-HIGH trial is uninformative as to whether triglyceride-lowering drug (TLD) use in HTG patients is clinically beneficial. Only a small proportion, 12.9%, of the enrolled patients had HTGs and the median baseline TG level was 166 mg/dl. The failure of niacin therapy to reduce CVD risk in the AIM-HIGH trial is non-informative as it stands. But, similarly to the ACCORD trial, a post-hoc sub-group analysis was performed, and the analysis indicated that niacin did not reduce CVD complication in patients with normal or borderline high TG levels (Guyton JR, Journal of the American College of Cardiology, 2013). However, as it relates to the ANCHOR population, patients with high TG and low HDL-Cholesterol (HDL-C) had a marked reduction (37% reduction) in CVD risk (p=.017). These findings are in agreement with those of the ACCORD trial.

Did the trial outcome support the FDA assertion that the trial results provided substantial new scientific evidence that triglyceride-lowering therapy would not improve clinical outcomes in patients with high TG levels?

No. The AIM-HIGH trial, as designed, was uninformative. The failure of niacin to reduce risk in patients with pre-existing CVD does not provide insight into the ANCHOR population. As in the ACCORD-Lipid trial, sub-group analysis based on TG levels again showed no risk reduction with normal or borderline TG levels. However, there was an impressive 37% reduction in CVD risk in patients that had high TG/ low HDL-C; again showing us that triglyceride-lowering therapy in this group is greatly beneficial.

C. HPS2-THRIVE Clinical Trial:

What specifically did the clinical trial investigate, and what was its primary endpoint?

The purpose of the study was to assess the effects of niacin (2 g) in reducing CVD risk in patients with occlusive arterial disease (HPS2-THRIVE Collaborative Group, European Heart Journal, 2013). Occlusive vascular disease was defined as a, "history of MI, CVD, and peripheral artery disease; or in diabetes mellitus associated with any of the above." CVD risk was defined as "coronary death, nonfatal MI, CVA, or coronary revascularization." A total of 25,673 patients, already on simvastatin, were randomized to receive either niacin or placebo. The final results, as reported, indicated that niacin did not reduce CVD risk in patients with occlusive arterial disease.

Did the study examine whether lowering TG in the ANCHOR population leads to a reduction in CVD risks?

No. As in ACCORD, this trial was not designed to examine the effects of TG lowering in the ANCHOR population. The patients were enrolled based on the presence of occlusive arterial disease, without regard to their TG level. The median baseline TG was, in fact, 125 mg/dl, i.e., normal. Due to sparse numbers, a sub-group analysis of HTG patients is unlikely to be reported; again reinforcing a lack of connection between the FDA cited study to the ANCHOR population.

Did the trial outcome support the FDA assertion that the trial results provided substantial new scientific evidence that triglyceride-lowering therapy would not improve clinical outcomes in patients with high TG levels?

No. Most of the patients in this trial had normal or borderline TG levels and, thus, the trial was neither designed nor powered to assess the beneficial effects of TG lowering in the high TG population.

V. Final Analysis

After an in-depth review and analysis of the three clinical trials referenced by the FDA, my conclusion is that the FDA erred in its review, by extrapolating that the failure of these trials to achieve their primary endpoints equates to a "substantial scientific issue". This appears to be a huge leap-of-faith by the FDA and not supported by the results of the trials. As most of the patients enrolled in the trials had either normal or borderline TG levels (< 200 mg/dl), the trials were not designed nor intended to assess the risks of CVD in the high TG patients (≥ 200 and < 500 mg/dl), the ANCHOR SPA population. The FDA's decision to rescind the ANCHOR SPA based on "substantial scientific issue" from these trials simply has no scientific merit and is not supported by the outcomes of the trials.

Over the next month, there will be numerous discussions between Amarin and the FDA regarding the merits of the FDA's action. Amarin will have the opportunity to refute the FDA's flawed argument, and convince the FDA to reconsider its conclusion and approve the ANCHOR sNDA. Based on my review, I can only conclude that the FDA's assertion is baseless. The FDA needs to acknowledge its error if it is to maintain its credibility and a perception of impartiality. I am optimistic that, in the end, a compromise will be reached that will lead to the approval of Vascepa for ANCHOR indication.

Another important factor that will greatly impact ANCHOR SPA discussions is the sub-group analysis from these trials actually showing a reduction in CVD complications in sub-group of patients with high TG levels. These data are also consistent with the published findings from the only large prospective clinical trial with pure EPA, which examined the efficacy of purified EPA in preventing CVD risks in hypercholesterolemic Japanese patients, referred to as the JELIS trial (Yokoyama, Lancet, 2007). In the JELIS trial, 18,645 patients were randomized to receive either EPA (1.8 gm) and statin or statin alone, and followed for a 5-year period. At the end of the study, the patients treated with EPA had a significant 19% reduction in CVD complications. A sub-group analysis also showed that the patients with elevated TG (≥150 mg/dl) and low HDL-cholesterol (< 40 mg/dl) levels had the greatest benefit with an impressive 53% reduction in CVD risks (Saito, Atherosclerosis, 2008).

Together, sub-group analyses of ACCORD-Lipid, AIM-HIGH, and JELIS trials provide compelling evidence that TG lowering therapy in patients with a high TG level (≥200 mg/dl) and low HDL level (< 40 mg/dl) leads to a reduction in CVD complications. If Amarin and the FDA cannot come to an agreement on ANCHOR SPA, an obvious compromise that would clearly benefit patients with high TG and one that the FDA should not object to (since the sub-group analyses in their cited trials support this) would be to narrow the Vascepa approval to patients with high TG levels (TG ≥200) and low HDL cholesterol (< 40 mg/dl). Ultimately, I believe this may be the final outcome upon which all parties can agree. As the studies cited by the FDA showed a reduction in CVD risk in patients with high TG and low HDL cholesterol, it would be hypocritical of FDA not to approve Vascepa for this sub-group of patients. Money lost may be recovered eventually, but human life lost is lost forever.

Financial Impact and Collateral Damage

The financial impact of the FDA's decision to rescind the ANCHOR SPA has been staggering to Amarin, a small pharmaceutical company. Amarin lost over 80% of its market capitalization, falling from $1.5 billion to $270 million, following the 8-K SEC filing regarding the FDA decision. Almost overnight, Amarin went from being one of the most promising biotech companies, with the best-in-class drug, to one that is facing questions about its future viability and financial solvency. There is a rightful concern that Amarin may not be able to complete the all-important REDUCE-IT outcome trial. The financial future of Amarin is intricately linked to the FDA's decision on ANCHOR SPA. Ultimately, I believe that a compromise will be reached and ANCHOR sNDA will be approved in some form. Although Amarin's stock price has plummeted as a direct consequence of the FDA action, there are compelling reasons to believe that the FDA will reverse its decision to rescind the ANCHOR SPA. At the current stock price ($1.55) the denial of the ANCHOR sNDA has been priced into the stock price, and for an astute investor investment in Amarin could pay off handsomely with only a limited downside risk.

Disclosure: I am long AMRN. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am a long-term supporter of and investor in a wide variety of biotech companies that are developing novel therapies to improve the quality of life.