The Evolution of Apabetalone, Redux
in response toby
Top-Line Data Expected September 2019 for BETonMACE Phase 3 Cardiovascular Outcomes Trial for Apabetalone
The understanding of the molecule known as apabetalone (aka RVX-208), as well as its clinical development has evolved over time. The Phase 2 ASSERT, SUSTAIN and ASSURE trials were either completed, or near completion, when apabetalone was first announced as a bromodomain-2 selective BET inhibitor about 7 years ago. The ongoing Phase 3 BETonMACE cardiovascular outcomes trial (CVOT) is the first major clinical endeavor Resverlogix has pursued since announcing this new mechanism of action. What was once thought to be just an apo-AI/high-density lipoprotein (HDL)/reverse cholesterol transport (RCT) drug is now known to reach into many more facets of biology and affect many other diseases beyond cardiovascular disease due to its epigenetic mechanism of action. A previous 'The Evolution of Apabetalone' post of mine skimmed over much of this. Here, I go into some more detail.
ASSERT, SUSTAIN and ASSURE
ASSERT (https://clinicaltrials.gov/ct2/show/NCT01058018) was a 12-week trial that started in 2010 and tested 3 different doses (50, 100 and 150 mg, twice a day) of apabetalone in patients with normal HDL-cholesterol (HDL-C); there was no low HDL-C requirement. All study sites were in the U.S. The primary outcome in ASSERT was % change in apo-AI for each dose vs. placebo. Although no one particular dose achieved statistical significance for % change in apo-AI vs. placebo, there was a significant dose-dependent increase in apo-AI. Additionally, there was a significant dose-dependent increase in HDL-C, HDL particle size (by NMR), number of large HDL particles (by NMR) and number of large alpha-1 migrating HDL particles (by 2D electrophoresis). Many of these apo-AI and HDL parameters still seemed to be increasing by the end of the 12-week study, suggesting that the full potential of apabetalone-mediated increases had not yet been realized. From the ASSERT top-line data news release: "Resverlogix Senior Vice President of Medical Affairs Dr. Jan Johansson stated, 'In patients who received the newer class of statins and had baseline HDL below 45mg/dL, an important high-risk subpopulation, the middle dose of 200 mg saw the most pronounced increases of 12% in ApoA-I (p<0.002), 21% in HDL cholesterol (p<0.015) and 32% in large particle HDL (p<0.018). We are delighted by these results and now have a much better understanding of what doses to use and what patient population to target moving forward in our ASSURE Phase 2b trial.'" Here is the published ASSERT paper.
Subsequent trials, including but not limited to SUSTAIN, ASSURE and BETonMACE, used the optimal dose identified in ASSERT (100 mg, twice a day), with a longer treatment period and in patients with lower HDL (<45 mg/dL females; <40 mg/dL males) than in ASSERT. So although ASSERT technically failed to meet its primary outcome, it accomplished the main goals of finding the optimal dose and refining the trial design for SUSTAIN and ASSURE and beyond.
SUSTAIN (https://clinicaltrials.gov/ct2/show/NCT01423188) was a 24-week trial that started dosing in 2011. It was an improvement over ASSERT in several ways. It used the optimal apabetalone dose (100 mg, twice a day), lengthened the treatment from 12 to 24 weeks, limited statin type to either atorvastatin or rosuvastatin, and required patients to have low HDL-C (<45 mg/dL females; <40 mg/dL males). It also switched the primary outcome from change in apo-AI vs. placebo to change in HDL-C vs. placebo, the latter of which was more consistently and robustly modulated in ASSERT. All study sites were in South Africa. SUSTAIN absolutely nailed the primary and secondary outcomes related to apo-AI and HDL, confirmed the findings of ASSERT, and suggested that the chosen dose, duration and patient population were correct for the plaque reduction ASSURE trial. From the SUSTAIN top-line data news release: "RVX-208 significantly increased HDL-C (p=0.001)......Apo-AI (p=0.002) and large HDL particles (p=0.02)."
I don't think that the full SUSTAIN results were ever published on their own in a scientific journal. However, a combined analysis of SUSTAIN and ASSURE was published in the Atherosclerosis 2016 article, which found that apabetalone significantly increased (vs. placebo) apo-AI, HDL-C, large HDL particle number and average HDL particle size. There was also a trend (p=0.07 vs. placebo) for increase in total HDL particle number. Of note, the combined analysis of SUSTAIN and ASSURE patient data showed much greater differences for the apo-AI and HDL metrics between the apabetalone and placebo groups than what was found in the ASSURE study data alone. Since the placebo group in ASSURE showed unexpected increases in these apo-AI and HDL metrics, it is reasonable to conclude that the biggest difference in the SUSTAIN and ASSURE data sets for apo-AI and HDL metrics was the respective placebo groups. By combining the two placebo groups (88 placebo patients in SUSTAIN, 81 placebo patients in ASSURE), the combined placebo data set effectively diluted out the unexpected increase that the ASSURE placebo patients experienced.
ASSURE (https://clinicaltrials.gov/ct2/show/NCT01067820) was a 26-week trial that started dosing in 2011. ASSURE was designed very similar to SUSTAIN for dose, duration and patient population, except the primary outcome was change in plaque (percent atheroma volume) as measured by intravascular ultrasound (IVUS). All seemed good leading up to ASSURE top-line. And then the bomb dropped, aka ASSURE top-line new release. Not only was there no significant reduction in percent atheroma volume (PAV) in the apabetalone group vs. baseline (primary outcome), but there were no significant differences between the placebo and apabetalone groups for change in PAV, total atheroma volume (TAV), HDL-C or apo-AI. While not the only reason for these observations, there was an unexpectedly strong beneficial change in apo-AI, HDL-C, PAV and TAV in the placebo group. Here is the published ASSURE paper.
Moving on! Post-hoc Analyses
The above ASSURE trial results were of course devastating to Resverlogix and its shareholders. However, just when you thought all was lost and that apabetalone was heading for the dumpster, several post-hoc analyses were performed that gave renewed hope. The first reported post-hoc analysis found evidence of synergy between apabetalone and rosuvastatin (Crestor) as compared to atorvastatin (Lipitor), and identified those with below median HDL (<39 mg/dL) treated with rosuvastatin as a responder population. "The responder population (i.e. HDL <39 mg/dL taking Rosuvastatin and RVX-208) exceeded the primary endpoint and also surpassed secondary endpoints reflecting regression in coronary atherosclerosis. These measures included total atheroma volume (TAV) and changes in the 10 mm most diseased segment of the coronary arteries, we noted marked regression versus baseline of -12.3 mm3 ( p< 0.0001) and -4.3 mm3 (p<0.0001), respectively. Other secondary endpoints assessed in this population were biomarkers of reverse cholesterol transport (RCT), including: HDLc, ApoA-I and large HDL particles which increased by 18.2% (p<0.0001), 16.4% (p<0.0001) and 74.7% (p<0.0001), respectively, vs. baseline." If you want to see some impressive responder population plaque reduction data that is stratified by statin type, statin dose and HDL-C level, see the patent linked in this post. In another post-hoc analysis, a sub-population with high baseline hsCRP (>2 mg/dL) experienced robust plaque regression, reductions in hsCRP levels, and reductions in MACE (discussed further below). However, this reduction in hsCRP level was not evident in the combined SUSTAIN and ASSURE populations (2016 Atherosclerosis article), likely because that analysis did not limit to those with baseline hsCRP >2mg/dL.
Although the failure to achieve the primary plaque reduction outcome overshadowed the ASSURE trial, the incidence of major adverse cardiac events (MACE) was a pre-specified secondary outcome in ASSURE (5-point MACE: death, myocardial infarction, stroke, coronary revascularization, hospitalization for ACS or heart failure). In the news release that reported the low-HDL with rosuvastatin responder population, it was noted that apabetalone-treated patients in ASSURE showed a trend for reduced MACE. Then came the post-hoc analyses of the Phase 2 programs that combined either SUSTAIN and ASSURE trials, also published in this Atherosclerosis 2016 paper, or combined ASSERT, SUSTAIN and ASSURE trials. Long story short, apabetalone reduced 5-point MACE in the total population, but even more so if one looks at just the diabetics or just those with elevated baseline hsCRP >2mg/dL as revealed in post-hoc analyses (here and here).
Two big caveats for the MACE analysis of the Phase 2 programs are #1: It is based on very few MACE events; and #2: Very few of these were the strict 3-point MACE events (cardiovascular death, non-fatal myocardial infarction, non-fatal stroke).
From Failure in ASSURE to Hope in BETonMACE
ASSURE failed. There is no sugar coating that fact. Why did the placebo group in ASSURE experience much more benefit than the placebo patients in ASSERT or SUSTAIN for apo-AI and HDL metrics? One unproven possibility is due to geographic, genetic, or other environmental differences between the trial populations. Study sites for ASSURE were in Argentina, Belgium, Brazil, Hungary, Netherlands, Poland, Russia, Spain, whereas ASSERT and SUSTAIN study sites were only in U.S. and South Africa, respectively. Perhaps we will never know for sure. The BETonMACE population is more geographically similar to ASSURE than it is to ASSERT or SUSTAIN.
This doesn't necessarily portend the demise of BETonMACE. Post-hoc analysis of ASSURE identified a low-HDL, rosuvastatin responder population that experienced robust and significant plaque reduction. Furthermore, post-hoc analysis of Phase 2 trials identified MACE reduction as a significant apabetalone-modulated endpoint, especially in diabetics. BETonMACE has wisely focused on a sub-group of patients who responded robustly for the MACE and plaque reduction outcome: low-HDL, diabetics with a pre-specified comparison of rosuvastatin vs. atorvastatin. See here for more info on primary outcomes, secondary outcomes, and pre-specified comparisons.
Beyond apo-AI, HDL-C and Reverse Cholesterol Transport
Apabetalone elicits effects on many biological pathways of benefit to cardiovascular disease due to its epigenetic mechanism. This may be a double edged sword. On the one hand, having effects beyond apo-AI/HDL-C and reverse cholesterol transport is great! There is published evidence that apabetalone:
The Resverlogix poster library has many other examples of these effects too.
On the other hand.....with an epigenetic drug that affects so many pathways there could be some adverse events related to apabetalone treatment. So far, the only one that has been noted in previous trial is the transient elevation of liver transaminases in some patients. But these past trials have only dosed for up to 6 months. BETonMACE patients will have been dosed for about 1 to 3 years. The safety and adverse event profile for BETonMACE will be very important to see.