...We Welcome You To The Resverlogix HUB withIn The AGORACOM COMMUNITY!
Message: RVX AHA Presentation 191116 - CORRECTED AGAIN (same as last correction in other message)
BETonMACE presentation at the American Heart Association Conference – 11/16/2019
Link from Lightmyfire2. (Thanks).
Link from BDAZ i think. (Thanks.)
Post-talk discussions on AHA channel of YouTube.
Dr. Kausik Ray: https://www.youtube.com/watch?v=NUYW0a3N_OQ&app=desktop
Dr. Svati Shah: https://www.youtube.com/watch?v=tPY75HxPrtE
Dr. Joseph Hill & Dr. Aruna Das Pradhan: https://www.youtube.com/watch?v=wOLODW_tPgg&pbjreload=10
So we’ll move on to the last clinical trial. A bromodomain inhibitor named Apabetalone, presented by Dr. Kausik Ray. Svati Shah will set the stage
Dr. Svati Shah:
We all know that DNA forms the backbone of the genome. But we have an increasing understanding of the role of epigenetics, or the so-called “second genome”. Epigenetics refers to modifications that occur on top of the DNA, that don’t actually change the underlying DNA code, but that result in regulation and gene expression.
Looking at this diagram, we know that DNA is formed by A, C, G, and T nucleotides. Those organize themselves into a double helix, which then wraps around histones. These secondary structures then compose the 23 pairs of chromosomes.
Epigenetic changes include methylation, which can occur anywhere along the DNA, and histone modifications.
So honing in more closely on the histones, epigenetic marks occur on the histone tail. Addition, removal, or recognition of these epigenetic marks results in differences in gene expression. And these epigenetic marks include phosphorylation, methylation, and acetylation.
It’s important to note that epigenetic changes are normal mechanisms by which the cell regulates activity of genes. But epigenetic changes can also occur due to in-utero environmental effects, as well as environmental effects that happen over a lifetime, for example related to smoking, pollution, and lifestyle. And we know that epigenetic changes can increase your risk of cardiovascular risk factor development, as well as cardiovascular disease.
Epigenetic modulation is also a potential therapeutic intervention; in fact, there are many drugs that exist that work through epigenetic mechanisms, including drugs that inhibit DNA methylation and drugs that inhibit histone acetylation. And the drugs themselves can be small molecules, monoclonal antibodies, or new genetics-based technologies like siRNA and gene editing.
So I told you that acetylation of histones is a mechanism of epigenetic regulation of gene expression. One way that acetylated histones can result in gene expression changes is by interaction with a set of proteins called bromodomain and external terminal, or BET proteins. BET proteins serve as a scaffold binding acetylated lysine residues on the histone, as well as binding transcription factors. And then they recruit more transcription factors and activate RNA polymerase-induced transcription, which then results in gene expression.
BET proteins are part of normal cellular processes and really in a normal cell relate to housekeeping genes, including cellular identity, cellular differentiation, and cellular proliferation. But there are disease states that can modify these acetylated histones and result in targeting of BET proteins to inappropriate genes.
BET inhibition has been pursued in oncology, but there is data to suggest that BET inhibition has cardiovascular effects, including effects on cardiac hypertrophy and vascular smooth muscle cell proliferation.
Apabetalone is an oral small molecule that binds the BD2 domain of the BET proteins BRD 2, 3, and 4, and by binding this BET protein, the BET protein can no longer bind the acetylated histone residues, and thus cannot induce transcription, and transcription is inhibited.
We are now going to hear the results of the BETonMACE trial, which evaluated Apabetalone in high-risk patients.
Dr. Kausik Ray:
Thank you. If I could have my slides, please.
So it’s my privilege and pleasure to present on behalf of the investigators, the steering committee, all the other committee members, and the patients, the results of the BETonMACE trial. This was a cardiovascular outcomes trial, the first that compared Apabetalone against placebo. The lists of the committees are shown here, the clinical steering committee at the top, the clinical events committee, and the data and safety monitoring committee here. This was a study that was multi-center, multi-country, conducted in 13 countries with 195 sites, and the national lead investigators are all shown here, and we’d like to thank them.
As you’ve seen in the introduction that was provided by Svati, one of the challenges that we have with this mechanism is that there are no traditional biomarkers or things that we can usually measure to guide what an outcome study might look like. So a lot of work has actually gone into the background, to look at the effect of Apabetalone. And in pre-clinical studies, we know that it has favorable effects on the complement system, on vascular inflammation, reverse cholesterol transport, acute phase proteins, vascular calcification, and the coagulation cascade.
Based on those early studies and an intra-vascular outcomes study, which although it did not show significant reduction in plaque volume, on virtual histology it showed favorable effects on plaque morphology. But that’s great, looking in pre-clinical studies you still want some kind of signal that this might work in man.
If we look at the clinical studies in the Phase II trials – we’ve pooled those together – we see that overall, Apabetalone compared to placebo seemed to reduce major cardiovascular events and there’s a broad basket there. Moreover, among specific populations, such as those with diabetes, shown in the bold lines, the effect of Apabetalone on MACE appeared to be greater than in those without that particular characteristic. This was also seen in those with a low HDL and a high CRP. These were populations with established coronary disease, and that kind of gave us a scenario on which to design a cardiovascular outcomes trial.
So the inclusion criteria were as follows: type II diabetes, a recent acute coronary syndrome, between 7 and 90 days [prior], and a low HDL cholesterol.
The exclusion criteria, shown on the right, are planned further revascularization beyond the index PCI, severe heart failure, coronary artery bypass surgery planned in the next 90 days, severe renal impairment, or severe liver disease.
This was an event-driven trial. The primary endpoint was the time to first occurrence of cardiovascular death, non-fatal myocardial infarction, or total strokes, including hemorrhagic strokes. In studies like this it’s always difficult to know what to do with deaths of undetermined origin, but because this was an ACS population, they were presumed to be cardiovascular deaths. We did pre-specify a safety analysis which excluded those deaths.
These are the pre-specified secondary endpoints, including a broader MACE that included hospitalization for other events. We also had total events, and at the bottom you’ll see hospitalization for heart failure. And this is in hierarchical order.
Time to first 4-part MACE (primary endpoint + hospitalization for CV events)
Total (first and recurrent) non-fatal MI or stroke and CV death
Time to first [sic] CV death or non-fatal MI
Time to first coronary heart disease death or non-fatal MI
Individual components of primary endpoint [cardiovascular death, non-fatal MI, stroke]
Hospitalization for congestive heart failure
When we estimated the sample size, 2,400 patients needed to be randomized to provide 80% power based on the following assumptions: that we would accrue 250 events, that the event rate in the placebo arm at 18 months would be 10.5%, and that we might see a 30% relative risk reduction, providing a 7.47% event rate in the Apabetalone group at 18 months. [Also on slide, 2-sided type 1 error rate a=.5%.]
This is the study design. Patients had to be on high-intensity statin therapy in a blinded comparison. Patients were randomized to either Apabetalone 100mg twice daily (b.i.d.) or matching placebo plus standard of care until 250 events had accrued. There was a safety observation period at the end of the trial after cessation of study drug. Again, something that is not uncommon.
This is the patient disposition. So about 4,000 patients were screened to randomize 2,425 patients. The reason for screen failure are shown here [HDL-C, Bilirubin, Triglycerides, Withdrew consent, Other], and we can see that about 90% concluded the trial protocol [89.8%]. [On slide, reasons for non-completion on placebo side were adverse event 3, withdrew consent 24, lost to follow-up 13, died 72, other 15. On Apabetalone side, reasons for non-completion were adverse event 2, withdrew consent 31, lost to follow up 7, died 61, other 26.] Most importantly, 98.7% of those randomized patients had full ascertainment of the primary outcome through the planned observation period or had vital status known as well. So vital status was known in about 99.2%, so that’s quite important.
Baseline characteristics are not dissimilar for a study like this. Everybody had diabetes; the duration of diabetes was just over 8 years. You will see that the majority of patients had PCI (percutaneous coronary intervention) for the index ACS. The time from index ACS was about 38 days.
Patients were well treated with standard of care, with about 90% on high-intensity statins, good use of ACE inhibitors, Beta blockers, and anti-platelets. And you’ll see that the use of SGLT2s or GLP1 agonists was low, but none of these agents have been shown to be beneficial in this setting.
These are the baseline parameters and importantly I’ll show you that LDL was well controlled by protocol. HDL was low. HbA1c was just over 7.
If we move to the efficacy endpoints now, there are no real biomarkers that we can measure. If we look at differences in biomarkers between Apabetalone and placebo, HDL, EGFR, and Alkaline Phosphatase were statistically significant, but these are small and it is uncertain what the clinical significance is.
This is the primary endpoint. This occurred over 26 months. It occurred in 12.4% of the placebo group and 10.3% of the Apabetalone group; hazard ratio of 0.82; upper boundary of the confidence interval 1.04. This was the pre-specified sensitivity analysis excluding deaths of unknown origin; hazard ratio 0.79; upper boundary 1.01; consistent with the primary endpoint.
Given that the primary endpoint is not significant, everything else that I give you should be considered nominal, hypothesis-generating, or exploratory. There are consistent trends for all components of the primary endpoint with the exception of stroke. And there was a nominal difference in hospitalization for heart failure.
[NOTE: this is the slide with the first Forest diagram. Summarized below.]
1. First occurrence of primary endpoint. Apabetalone better, hazard ratio 0.82, upper boundary of confidence interval 1.04, p value 0.11.
2. First occurrence of primary endpoint with undetermined death excluded. Apabetalone better, hazard ratio 0.79, upper boundary of confidence interval 1.01, p value not given.
3. First occurrence of primary endpoint or hospitalization for unstable angina or urgent or emergency revascularization procedure. Hazard ratio 0.85, Upper boundary of confidence interval 1.06
4. First and recurrent primary endpoint events. Hazard ratio 0.79, Upper boundary of CI 1.01
5. Cardiovascular death or non-fatal myocardial infarction Hazard ratio 0.79, upper boundary of CI 1.02
6. Coronary heart disease death or non-fatal myocardial infarction Hazard ratio 0.79, Upper boundary of CI 1.02
7. Non-fatal myocardial infarction Hazard ratio 0.80, Upper boundary of CI 1.08
8. Cardiovascular death Hazard ratio 0.81, Upper boundary of CI 1.19
9. Stroke Hazard ratio 1.01, Upper boundary of CI 1.98
10. All-cause mortality Hazard ratio 0.88, Upper boundary of CI 1.24
11. First hospitalization for congestive heart failure Hazard ratio 0.59, Upper boundary of CI 0.94 *
If we look at subgroups – again this has to be taken with some caution given the primary endpoint. There was generally no effect modification by baseline characteristics with the exception of low LDL at baseline or a lower EGFR.
[Note: This was the second Forest Diagram. Primary endpoint in specified subgroups. Summarized below.]
a. Female Hazard ratio 0.79, Upper boundary of CI 1.36, p value 0.70
b. Male Hazard ratio 0.84, Upper boundary of CI 1.10
a. Rosuvastatin Hazard ratio 0.86, Upper boundary of CI 1.22, p value 0.67
b. Atorvastatin Hazard ratio 0.78, Upper boundary of CI 1.09
a. < Median Hazard ratio 0.60, Upper boundary of CI 0.86, p value 0.024 *
b. ³ Median Hazard ratio 1.06, Upper boundary of CI 1.46
4. Hemoglobin A1c
a. < Median Hazard ratio 0.88, Upper boundary of CI 1.28, p value 0.79
b. ³ Median Hazard ratio 0.82, Upper boundary of CI 1.12
5. Estimated Glomerular Filtration Rate (EGFR)
a. <60 Hazard ratio 0.50, Upper boundary of CI 0.96, p value 0.032 *
b. ³ 60 Hazard ratio 0.94, Upper boundary of CI 1.22
To finish with safety, adverse events occurred with a similar frequency in the Apabetalone group vs. placebo. Adverse events leading to discontinuation were somewhat greater. If we look at SAEs these were identical. If we look at the reasons for discontinuation, transaminase elevations were more common, and there were no differences with respect to Hys laws or bilirubin levels. Discontinuation due to LFTs are shown on the bottom, about a two percent excess.
So in summary, Apabetalone did not significantly reduce the primary endpoint. The observed event rate was slightly lower than what we had anticipated (9.7% vs. 10.5%), but most importantly the study was powered on a 30% event reduction, and was thus underpowered to detect a smaller difference in events. Apabetalone was generally well tolerated with an overall incidence of adverse events similar to that in the placebo group. However, discontinuation due to liver function abnormalities was more frequent.
So we think this first proof-of-concept trial really shows promise for epigenetic modification with Apabetalone. Favorable trends were observed for all components of the primary endpoint with the exception of stroke, with a nominal difference in heart failure hospitalization. And we feel that based on these results that further studies of this approach are probably warranted, Thank you.
So Apabetalone: what did we know before BETonMACE? We know that epigenetic modulators have effects on many different downstream genes. And accordingly, Apabetalone has had effects on vascular calcification, inflammation, monocyte recruitment, reverse cholesterol transport, and thrombosis.
In pre-clinical studies, Apabetalone has been shown to induce hepatic synthesis of apoA-1 and increase the cholesterol efflux capabilities of HDL particles. In mice, Apabetalone decreases aortic lesion size, improves lipids, and decreases levels of inflammatory proteins including cytokines and IL-6.
Inflammation is likely to be a very important mechanism of Apabetalone’s effects. We know that in response to inflammatory stimuli such as LPS [lipopolysaccharide] there is translocation of the master inflammatory transcription factor NF-kB from the cytoplasm to the nucleus, where it binds with genetic regulatory elements. It then recruits BET proteins; and it recruits these BET proteins from their normal housekeeping genes to now induce inflammation. And we know that BET inhibition attenuates this inflammatory response. Therefore, Apabetalone was evaluated for potential clinical utility in targeting residual risk in patients who were already treated with evidence-based lipid therapies. There have been three clinical studies. They were all Phase II trials before BETonMACE. ASSERT was a safety and tolerability trial, SUSTAIN looked at changes in lipids, and ASSURE looked at CAD progression. In a meta-analysis of these Phase II clinical trials, a reduction in MACE was shown, with a more pronounced effect in high-risk patients, including type II diabetics, patients with high CRP, and low HDL. Importantly, the only safety concerns that were raised in these trials were a transient LFT elevation which was seen in less than 4% of patients.
So what did BETonMACE add to our knowledge? Well first of all I really want to congratulate Dr. Ray and his collaborators on a study that was extremely well-conducted. This was the first ever trial of BET inhibition on cardiovascular outcomes, with complicated patient recruitment, high-risk outcomes with a narrow window of enrollment after their MI or ACS event. So I really want to applaud the investigators for their success in recruitment and retention in this trial. And likely the biomarkers study will add important mechanistic information about the drug. But we have to remember that this was overall a negative trial. This was likely because of low power with an event rate in the placebo group that was lower than expected. And the model effect size of 30% with Apabetalone was not achieved in this study. Other potential reasons for a negative trial are that the drug does not actually work on MACE, unintended pleiotropic effects, or heterogeneity of effects in the background of revascularization. In pre-specified secondary endpoints, Apabetalone decreased heart failure hospitalization at a nominally significant association. And while we have to view these results with caution, it’s interesting to note that Apabetalone has known effects on cardiac hypertrophy and cardiac fibrosis markers.
In pre-specified subgroup analyses they showed benefit of Apabetalone in patients with LDL lower than the median. It’s not sure what that means, but perhaps it’s a surrogate for compliance, or relatedly, the LDL paradox. In patients with GFR below the median, they also showed a beneficial effect of Apabetalone. And, while again this has to be viewed with caution, there is biological plausibility because we know that this drug reduces alkaline phosphatase levels, and we know that alk phos levels are elevated in patients with kidney disease and induce vascular calcification. It’s important to note that although the secondary endpoints and subgroup analyses were pre-specified, that there was no adjustment for multiple comparisons and that these results should be viewed with caution.
And finally BETonMACE had a longer follow-up than the previous Phase Ii clinical trials and again confirmed no significant harm other than a transient elevation in LFTs, and importantly, because this drug might have anti-inflammatory effects, there was no increased risk of infections.
So in conclusion, this was the first ever phase III clinical trial of an interesting small molecule with a relatively novel mechanism of action. BETonMACE showed that Apabetalone did not show benefit for MACE in high-risk post-MI/ACS patients with diabetes and low HDL who are on evidence-based statin therapy. However, given the large body of pre-clinical and human studies, as well as a suggestion of effects on heart failure and in subgroups, and a concordance of non-significant effects across the secondary endpoints, there is cautious optimism for this drug. While it’s not ready for use in patients, I would say that these results suggest that an adequately powered clinical trial is needed. This raises the question of how one would design such an adequately powered randomized controlled trial. BETonMACE focused on high-risk patients and so a future clinical trial, if it continued to work on that patient population, would need to increase the sample size. Another alternative would be to focus on higher-risk subgroups, such as patients with heart failure or patients with chronic kidney disease. So again, my congratulations to Dr. Ray and his colleagues for a well-conducted trial. Thank you.
Moderator: We’ve got lots of great questions. ... etc.
Trial-specific questions were not directed to BETonMACE, so I did not bother to transcribe this section. However there was some general discussion of the possibility that a variety of combined drugs might be useful for patients with different disease drivers, e.g. inflammation vs. other factors. Discussion also touched on the need for guidelines and algorithms to direct such choices. A general statement was also made that statins are still the most valuable tool in the box at this point.
Please login to post a reply