Hearts & Minds Blog
A Closer Look – June 2018June 5, 2018
As we approach the halfway point in 2018, our Lipid Management Team remains confident and encouraged by the consistently positive results from four Phase 2 and Phase 3 clinical studies of bempedoic acid, all of which met their primary endpoints. In this blog post, I’d like to take the opportunity to reinforce the strength of our program as well as clear up any concerns or misinterpretations of the data, particularly as they relate to the safety and tolerability of bempedoic acid.
Esperion and the bempedoic acid program were born from the need of the medical community for new, complementary, once-daily oral therapies that provide the LDL-C lowering and hsCRP reductions that their high-risk patients need, the value that payers appreciate, and the convenience and tolerability that patients want and deserve.
Keeping in line with these goals, our Lipid Management Team remains on track to complete our two remaining pivotal studies later this year and submit our NDAs no later than the first quarter of 2019. We’d like to offer a closer look at various aspects of our program and reinforce our path forward. Here’s a recap of what we’ve learned to date:
1. Efficacy of bempedoic acid
2. Safety and tolerability of bempedoic acid
3. Liver Function Tests
4. Numerical imbalances in deaths in Study 1
5. Mechanism of Action (MOA) for bempedoic acid
6. Phase 2 diabetes study
1. Efficacy of Bempedoic Acid: LDL-C Lowering and hsCRP Reduction
First, it is unequivocal that bempedoic acid lowers LDL-C and reduces hsCRP. The Phase 2 and Phase 3 efficacy results reported to date are highly consistent at 12-weeks (the FDA-agreed-upon time-point for measuring efficacy) using an “apples-to-apples” on-treatment analysis which is most similar to the modified intent to treat analysis used in Phase 2. This on-treatment analysis is predictive of an individual patient’s response to a therapy and is therefore most useful to physicians. The ITT – or intent-to-treat – analysis is used in Phase 3 clinical testing, and the results of this analysis will be included in our label.
To summarize the efficacy of bempedoic acid:
- 20-24% additional LDL-C lowering on low-, moderate, or high-intensity statins with statistically significant reductions in hsCRP, regardless of statin intensity
- 26 – 30% LDL-C lowering as mono therapy or when added to background therapy of ezetimibe or pcsk9i’s, also with significant reductions in hsCRP
- 50% LDL-C lowering when combined with ezetimibe, the world’s first and only once-daily, oral, non-statin, LDL-C lowering drug which provides high-intensity-statin-like levels of LDL-C lowering and hsCRP reductions.
- Patients taking bempedoic acid achieved reductions in hsCRP of 22-40%, regardless of background therapy
- Consistent with every LDL-C lowering therapy developed, LDL-C lowering with bempedoic acid is less using the more rigorous ITT analysis and also less over time, but remains clinically meaningful even at 52 weeks.
2. Cumulative Phase 2 and Phase 3 Safety and Tolerability Results
- We continue to set new standards for transparency, and in connection with the results from Study 3, we reported cumulative safety and tolerability results for the bempedoic acid clinical program as a whole. We were surprised to see that some in the investment community did not draw the conclusion that the most relevant safety results were the cumulative safety and tolerability results in over 4000 patients, many treated for 52 weeks. So what is the cumulative safety and tolerability profile for bempedoic acid?
- There were no clinically relevant differences between the bempedoic acid and placebo groups in the occurrence of AEs with 67.8 percent and 66.1 percent, respectively; or SAEs with 9.2 percent and 8.9 percent for the bempedoic acid and placebo groups, respectively.
- Discontinuations due to AEs were 9.5 percent and 6.9 percent, respectively, for the treatment and placebo groups; importantly, there was no single class of events or specific event responsible for the difference.
- Bempedoic acid showed no meaningful difference from placebo in muscle-related AEs.
- The occurrence of fatal adverse events was 0.5 percent and 0.2 percent for the treatment group and placebo group, respectively. None of these events were deemed to be related to study treatment by the study investigator.
3. Let me start with the rate of increases in Liver Function Tests as this issue was raised in March when Study 4 results were announced.
First, recall that a liver function test is a blood test that measures substances including alanine transaminase (ALT) and aspartate transaminase (AST) enzymes as well as the bilirubin and alkaline phosphatase biomarkers. These enzymes are primarily but not exclusively produced by the liver. An increase in a liver function test (LFT) is when the level of AST and/or ALT is greater than the upper limit of normal. LFT measurements are a standard safety measure in almost all drug development programs. Low rates of LFT increases are expected with oral, LDL-C lowering drugs – an “on-target” effect of the drug lowering LDL-C in liver cells. In general, LFT increases occur infrequently (between 0.2% and 2.3% for approved LDL-C lowering drugs) and have been observed in the development and use of ALL approved and successful oral lipid-regulating drugs. Importantly, LFT increases with oral lipid regulating drugs have NOT been associated with, or predictive of, liver injury. The protocol to address an LFT increase of >3X ULN with these drugs, in clinical practice or clinical studies, is simply to withdraw drug therapy for a few days to a couple of weeks and then restart therapy. This protocol has been highly effective and remains the industry standard.
Here are a few additional thoughts for context: the liver is the primary site for controlling the production and metabolism of cholesterol in the body. As a result, most of the drugs developed and approved for managing cholesterol have activity in the liver and have been associated with increases in LFTs, and are regarded as “on-target” effects – in the case of bempedoic acid, it is a pro-drug and only active in the liver. If you want to better understand LFT increases with LDL-C lowering drugs, we encourage you to speak with a lipidologist or cardiologist – and not an equity research analyst. You’ll conclude, like our Lipid Management Team and our KOLs, that the rate of LFT increases associated with bempedoic acid is very low overall and consistent with ALL other previously approved LDL-C lowering drugs. Having reported all LFT elevations (a rate of 0.56% – a rate similar to the placebo rate of LFT elevations in statin labels) from our Phase 2 and Phase 3 completed studies, even for on-going and still-blinded Phase 3 studies, we’re confident we can all put this issue behind us now.
4. Next I want to briefly cover the “numerical imbalance in deaths unrelated to study treatment” in Study 1: 13 in the treatment group, and 2 in the placebo group (recall the study randomized patients 2:1).
- I want to emphasize that the deaths were assessed as “unrelated to study treatment” by the investigator who was “blinded” as to whether the patient was receiving bempedoic acid or placebo. This is extremely important – because in every case the investigator determined that the patient’s death was NOT associated with our study treatment.
- Second, recall that the Clinical Events Committee (CEC), comprised of a group of highly experienced and well-regarded physicians, independently reviews and adjudicates the cause of death for each fatality in the Phase 3 program based on the clinical findings. As a result, we (and all of you) can be confident that the cause of death is correct.
- Third, recall that the Data Monitoring Committee (DMC) reviews the unblinded data from every study in the Phase 3 program and from CLEAR Outcomes study each quarter. The DMC has consistently recommended that the Phase 3 program and CLEAR Outcomes continue as planned, with the most recent meetings in May 2018.
What do we know about these deaths that are unrelated to study treatment and that make our Lipid Management Team feel confident about the continued good safety and tolerability profile of bempedoic acid?
- In study 1, the overall death rate was 0.9% in the treatment group and 0.3% in the placebo group. This compares very favorably to the 1.8% one-year general fatality rate for the general population aged 65 years old in the U.S., as cited by the CDC’s Morbidity and Mortality Weekly Report. Patients were very well cared-for in this long-term study.
- It’s also important to note that the recently approved pcsk9i’s had similar death rate imbalances in their long-term studies – and still received approval – since regulatory authorities evaluate the cumulative safety and tolerability profile for entire programs. In addition, in a recent ~27,500 patient cardiovascular outcomes trial, a pcsk9i demonstrated a 15% CV event risk reduction but a 1.05 hazard ratio in CV death. Despite this hazard ratio, which was not statistically significant, the pcsk9i received an expansion of their indication statement to include CV risk reduction.
- A few other important facts provide additional perspective on the unrelated nature of these fatalities:
- Seven deaths – all in the treatment group – occurred in the first 12 weeks; while 13 of the deaths occurred within the first 26 weeks after randomization into the study. If treatment with bempedoic acid were antagonizing fatalities, it is reasonable to expect that treatment group deaths would be weighted to the second half of the study. The opposite was observed.
- Recall that our Phase 2 program included over 800 patients treated with bempedoic acid, most for 12 weeks, with only one patient experiencing a fatality that was deemed potentially related. A rate of 0.01%.
- Five of the unrelated deaths were due to neoplasms (cancer).
- Four were lung cancer deaths – all in current or former smokers – in patients on treatment for an average of 120 days (range of 3 – 235 days).
- However, SAEs associated with neoplasms were balanced between the treatment group and the placebo group at ~1%.
- Similarly, in our Phase 2 program, which treated over 800 patients with bempedoic acid, the rate of neoplasm SAEs was also low overall and balanced between the treatment and placebo group at a rate of ~0.25%.
- Finally, after reviewing the blinded fatalities in the on-going Phase 3 studies (particularly Study 2, a 52-week study in ~800 patients), we can report that there have been NO deaths associated with neoplasms.
- We are comfortable asserting that the numerical imbalance in unrelated deaths associated with neoplasms was not clinically meaningful, and indeed, is a spurious finding.
- Six of the unrelated deaths were due to cardiovascular disease (5 in the treatment group and one in the placebo group). With 2:1 randomization, a balanced result for cardiovascular deaths would have been 4:2 – so the “numerical imbalance” is one – a very minor imbalance indeed.
- Note that three deaths in the treatment group were heart attacks (MIs) which occurred within the first 61 days of treatment in patients that had a history of heart attacks (these patients average time on treatment was 43 days (range of 28 – 53 days). The other two were due to heart failure in patients with a history of heart failure prior to randomization into the study.
- Importantly, when evaluating adjudicated cardiovascular events from Study 1 – 5-component MACE, 4-component MACE, and 3-component MACE – ALL favored bempedoic acid – meaning the rate of cardiovascular events was lower in the bempedoic acid group than in the placebo group. Importantly, we know from recently completed cardiovascular outcomes studies that the Kaplan-Meier event curves only begin to separate after a year of treatment. The conclusion is that while cardiovascular deaths in the study were slightly imbalanced (by one – 5:1 vs 4:2, bempedoic acid:placebo), the cardiovascular event rates clearly favor bempedoic acid in this study providing us confidence that bempedoic acid had a positive effect on cardiovascular event rates in Study 1.
- Four of the unrelated deaths in the study were due to various other causes (3 in the treatment group and one in the placebo group). With 2:1 randomization, a balanced result for other unrelated deaths would have been 3:1.5 – so the numerical imbalance is very slight – only “one-half”.
- This category of unrelated deaths included acute pancreatitis, severe ischemic apoplex, multiple organ failure and sepsis.
- There is no common cause so our Lipid Management Team is comfortable concluding that the very slight imbalance in these other unrelated deaths is not clinically meaningful.
- Finally, some have asked about the statistical chances that the imbalance in neoplasm or cardiovascular deaths could occur. Our bio-statistician first performed a chi-sq calculation to determine the odds of getting the 5:1 imbalance in events in a 2:1 randomized study since a few external parties have attempted to perform their own statistical analysis. A two-sided test using chi-sq statistics gives you a p-value=0.387 for 5:1. While this is usually an acceptable approach, in this case the event (deaths) rate is very low (leading to very small cell count) and so the normal approximation (chi-sq test) doesn’t work very well. Instead she recommended a more appropriate exact test (Fisher) to evaluate the results which compares the two proportions. The two-sided p-value is 0.67 for the 5:1 imbalance. The conclusion is that there is a very good chance (67%) to see this imbalance even when the underlying event (death) rates are actually the same!
5. Bempedoic Acid Mechanism of Action and Supporting Genetic Validation Results
Esperion worked extensively with the Cardiology team at the Cleveland Clinic to design a robust cardiovascular outcomes trial (CVOT). The CLEAR Outcomes CVOT is enrolling ASCVD and/or HeFH patients with elevated LDL-C who are also considered statin intolerant. These patients are taking less than the lowest starting dose of a statin or no statin at all. The mean baseline LDL-C level for these patients is as expected – 135 mg/dL because these patients will not have benefitted from prior statin therapy (e.g., equal to or greater than approved daily starting doses of statins).
To gain additional insight into the potential for success of the CLEAR Outcomes CVOT, Mendelian randomization studies were conducted by a team, led by Dr. Brian Ference, to genetically validate ATP Citrate Lyase inhibition (ACLi), the enzyme targeted by bempedoic acid. Results from the Mendelian randomization studies presented at ACC in March 2017 and demonstrated that genetic variants of ACL have the same effect on the risk reduction for cardiovascular events per unit change in LDL-C as variants that mimic the effects of HMG-CoA reductase (statins), pcsk9i’s and NPC1L1 (ezetimibe). Dr. Ference and colleagues concluded that an ACL inhibitor such as bempedoic acid should reduce the risk of cardiovascular events by the same amount as statins do per unit reduction in LDL-C.
The importance of genetic validation of targets for LDL-C lowering and cardiovascular disease risk reduction was highlighted at the pcsk9i’s Advisory Committee meetings in June 2015. Mendelian randomization studies have previously provided genetic target validation for several LDL-receptor-mediated LDL-C lowering therapies including HMG-CoA reductase, the enzyme target of statins, NPC1L1, the enzyme target for ezetimibe and pcsk9i’s and have been published in scientific literature over the past 15 years.
Published in November 2016 in Nature Communications was research conducted by Esperion elucidating the bempedoic acid mechanism of action. Bempedoic acid works two steps up from HMG-CoA reductase in the cholesterol biosynthesis pathway; specifically targets ATP Citrate-Lyase (ACL); is focused only in the liver; and does this without apparent off-target effects. Bempedoic acid and statins both inhibit the same cholesterol biosynthesis pathway, resulting in an upregulation of LDL receptors. The specific enzyme target is different.
We, and most lipid experts, know that statin use is not limited by the capacity of the cholesterol biosynthesis pathway but by the ability of patients to tolerate their prescribed doses of statins. Recall, for example, that simvastatin was tested at doses much higher than the current maximum approved dose of 40 mg, including at least 160 mg. The pathway has considerable additional capacity even at the highest approved doses of the most potent statins. The ability of bempedoic acid to reduce LDL-C by an incremental 20-24% on top of statins illustrates this point because when you double a statin dose, you only reduce LDL-C by a further 6% (the “Rule of Six”). The ability of bempedoic acid to achieve this incremental reduction would be roughly equivalent to 3-4 doublings of the statin dose.
6. New Phase 2 study in patients with Type 2 Diabetes
- As shown on ClinicalTrials.gov, this is a 12-week study to assess the LDL-C lowering efficacy, tolerability and safety of the bempedoic acid / ezetimibe combination pill compared to ezetimibe and placebo in patients with type 2 diabetes (T2D) and elevated LDL-C. The study will also explore the effect of the bempedoic acid/ezetimibe combination pill on diabetes endpoints of interest at 12 weeks, including HbA1c and fasting glucose.
- The rationale for the study is the observation in previously completed Phase 2 and Phase 3 studies that the use of bempedoic acid has not been associated with new onset or worsening of diabetes, or any negative effects on diabetes endpoints. As you may recall, the use of statins, particularly high-intensity statins, has been associated with increases in HbA1c and fasting serum glucose levels (this is referenced in the labeling for all statins).
- We look forward to sharing the topline results for this study with you in 2019.
Drug development – especially cardiovascular drug development – is what I call a “long game”; success is not measured in days, weeks or months – success is measured in years.
There are countless ups and downs and wins and losses along the way. It’s a lot of work. It takes grit. It takes persistence. It takes resiliency. Most of all it takes team work. I constantly remind our team “drug development is a team sport.” At Esperion, we’ve assembled a team that is among the most successful and experienced in developing lipid-regulating drugs anywhere in the world – our Lipid Management Team. Our Team works incredibly hard, is passionate, and is committed to the success of the bempedoic acid franchise and Esperion. To paraphrase Margaret Meade: “Never doubt that a small group of thoughtful committed people can change the world. Indeed, it is the only thing that ever has.”
I encourage you to stay tuned. We are working hard for patients who can’t and won’t take statins as well as those that can’t achieve their LDL-C lowering goals on statins. We’re working hard to give physicians new, convenient, once-daily, oral, accessible LDL-cholesterol lowering drugs. We’re almost there. We’re confident that if we get it right for patients and physicians, our investors will be rewarded.