ACC 2017 Scientific Sessions (American College of Cardiology)

March 17-19, 2017; Washington, DC; FOURIER Results – Draft

Executive Highlights

The ACC’s 66th annual scientific sessions featured groundbreaking new data right off the bat – the meeting kicked-off this morning with a presentation of the FOURIER trial for Amgen’s PCSK9 inhibitor Repatha (evolocumab). Full results from this CVOT have been highly-anticipated since Amgen released topline data in February, and fittingly, Dr. Marc Sabatine spoke to a P-A-C-K-E-D conference hall of very excited attendees. See below for our deep dive into FOURIER (importantly, the very first outcomes trial to report for a PCSK9 inhibitor!). But first, a few key takeaways:

Top Five Highlights

1. Risk reduction for the composite primary endpoint was 15% for people treated with evolocumab (p<0.001 vs. placebo). This primary endpoint consisted of CV death, non-fatal MI, non-fatal stroke, hospitalization for unstable angina, and coronary revascularization. Risk reduction for a key secondary endpoint of three-point MACE (non-fatal MI, non-fatal stroke, and CV death) was a remarkable 20% for people treated with evolocumab (p<0.001 vs. placebo). Note that, fitting for a cardiovascular drug, the CVOT was powered for superiority on both composite endpoints.

2. The driving factors for cardioprotection were coronary revascularization (HR=0.78 in favor of Repatha, p<0.001), non-fatal MI (HR=0.73, p<0.001), and non-fatal stroke (HR=0.79, p=0.01). Notably, there was no statistically significant reduction in risk for CV death associated with evolocumab, though we imagine this may have been a consequence of the relatively short follow-up (median follow-up of 26 months).

3. CV benefit appeared to increase over time through the course of follow-up. Risk reduction for the primary endpoint was ~12% in the first year and ~19% in the second, while risk reduction for the secondary endpoint grew from 16% to 25% between year one and beyond.

4. No sub-analysis for participants with baseline diabetes was presented, but if this sub-group (10,081 individuals) mirrors the overall FOURIER results, we wonder what the possibilities will be for synergistic CV benefits with Repatha and Jardiance (Lilly/BI’s SGLT-2 inhibitor empagliflozin) or Repatha and Victoza (Novo Nordisk’s GLP-1 agonist liraglutide). Co-administration of Repatha and Jardiance could be particularly interesting – the latter showed statistically significant risk reduction for CV death likely driven by heart failure benefit in EMPA-REG OUTCOME, while evolocumab’s CV benefit was driven by reduced risk for MI and stroke.

5. Cost is an inescapable theme when it comes to any discussion of any PCSK9 inhibitor. Indeed, during Q&A following the FOURIER presentation, cardiology experts emphasized the role of risk stratification as a means of identifying and specifying which patients stand to benefit most from evolocumab therapy. Of course, this introduces a dilemma – how can we pursue chronic disease prevention while also creating a system where you have to do worse to get a better therapy? We only hope that significant risk reduction for CV events, as demonstrated in this trial, will bring forth better reimbursement of Repatha. The ODYSSEY Outcomes CVOT for Sanofi/Regeneron’s Praluent (alirocumab), the other major PCSK9 inhibitor on the market, is scheduled to complete in February 2018.

Detailed Discussion and Commentary

Late-Breaking Clinical Trials

Primary Results of the Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk (FOURIER) Trial

Marc Steven Sabatine, MD (Brigham and Women’s Hospital, Boston, MA)

Dr. Marc Sabatine presented data from the first-ever completed outcomes trial of a PCSK9 inhibitor. Underscoring the great interest in these results, FOURIER was simultaneously published in NEJM. Evolocumab reduced risk for the composite primary endpoint (CV death, non-fatal MI, non-fatal stroke, hospitalization for unstable angina, or coronary revascularization) by 15% (p<0.0001) and for the key secondary endpoint of three-point MACE (CV death, non-fatal MI, or non-fatal stroke) by 20% (p<0.00001). Dr. Sabatine underscored that these risk reductions were both clinically meaningful and highly statistically significant. CV benefit was driven largely by decreased risk for coronary revascularization, MI, and stroke – no other individual endpoints (CV death or hospitalization for unstable angina) showed statistically significant risk reduction with evolocumab vs. placebo. Coronary revascularization was 22% less likely in the treatment arm (p<0.001). There were 468 cases of MI among evolocumab-treated patients (3.4% event rate) and 639 cases among placebo-treated patients (4.6% event rate) – the active agent reduced risk for MI by 27% (p<0.001). Looking specifically at stroke, there were 207 events in the evolocumab arm (1.5% event rate) and 262 events in the placebo arm (1.9% event rate) – Repatha reduced risk for stroke by 21% (p=0.01). The therapy was associated with a 59% mean reduction in LDL, from a baseline 92 mg/dl to a median 30 mg/dl, which corroborates earlier clinical evidence showing the drug’s lipid-lowering efficacy. Dr. Sabatine shared that 25% of evolocumab-treated participants achieved an LDL below 20 mg/dl – in other words, well below what current guidelines recommend. The overarching implication of these numbers, he explained, is that aggressive LDL-lowering with evolocumab, even beyond current targets, offers a substantial cardioprotective benefit. FOURIER now provides published evidence to back-up the added benefits to treatment with a PCSK9 inhibitor on top of moderate- or high-intensity statins – not only are further substantial reductions in circulating LDL possible, but this reduction in LDL level has a measurable impact on CV outcomes.

  • Study design: FOURIER enrolled 27,564 participants (all adults between 40-85 years-old with clinically-evident atherosclerotic CV disease on a background of moderate- to high-intensity statins), randomizing 13,784 to the evolocumab arm and 13,780 to the placebo arm. Participants hailed from 49 different countries, making for a truly global subject pool. Median follow-up was 26 months, which resulted in 59,865 patient-years of follow-up. In total, there were 2,907 primary endpoint events during the trial (1,344 evolocumab-treated patients vs. 1,563 placeb0-treated patients) and 1,829 MACE events (816 evolocumab-treated patients vs. 1,013 placebo-treated patients) – for context, there were only 254 MACE events in SUSTAIN 6, 1,302 in LEADER, and 772 in EMPA-REG OUTCOME (all diabetes CVOTs). Patients were given a choice between twice-monthly 140 mg injections vs. once-monthly 420 mg injections of evolocumab, and results were similar regardless of chosen regimen (both in terms of LDL-lowering and CV outcomes).
  • The Kaplan Meier curves for the primary and secondary composite endpoints show increasing divergence over time, which suggests a greater CV benefit with longer duration of evolocumab treatment. Risk reduction for the primary endpoint was ~12% in the first year and ~19% in the second, while risk reduction for the secondary endpoint grew from 16% to 25% between year one and beyond. As Dr. Sabatine articulated, “it takes time for LDL-lowering to translate to a clinical benefit” – prior studies have demonstrated this, and the trend held true in FOURIER. Median follow-up during the trial was 26 months. Longer-term data is necessary to confirm enhanced cardioprotection in years three, four, and five of evolocumab therapy, which will hopefully come from future outcomes research on Repatha.

  • Components of the composite endpoints: The 15% risk reduction for the primary endpoint and 20% risk reduction for the secondary endpoint in FOURIER can be attributed to reductions in risk of coronary revascularization (HR=0.78, p<0.001), non-fatal MI (HR=0.73, p<0.001), and non-fatal stroke (HR=0.79, p=0.01). CV death trended in the wrong direction, favoring placebo (HR=1.05; 251 deaths in treatment arm; 240 deaths in placebo arm; 95% CI=0.88-1.25), but importantly this correlation did not reach statistical significance (p=0.62). Dr. Sabatine underscored that CV death due to acute MI and due to stroke specifically trended in the right direction, favoring evolocumab (hazard ratios=0.84 and 0.94, respectively), so only “other” CV death showed a higher rate in the treatment group (HR=1.10). We assume a significant portion of deaths in the “other” CV death category were related to heart failure and it’s not a major surprise that evolocumab appears more effective on endpoints related to atherosclerosis given its LDL cholesterol-lowering mechanism of action. The hazard ratio for all-cause mortality, 1.04, also favored placebo, but this did not reach statistical significance (p=0.54). The hazard ratio for hospitalization for unstable angina was 0.99 in favor of evolocumab, which was not statistically significant (p=0.89).
  • Results were consistent across a variety of subgroups in FOURIER: evolocumab conferred similar CV risk reduction regardless of baseline CV disease, baseline statin intensity, or baseline LDL. This latter analysis was a major focus of Dr. Sabatine’s remarks: He pointed out that even participants in the lowest quartile of baseline LDL (<80 mg/dl) experienced a consistent CV benefit on the primary and secondary endpoints. By the end of the trial, a lower achieved LDL level was associated with lower risk for MI, stroke, or CV death (p<0.0001), which contributes to the notion that “lower for longer is better.” A question arose during Q&A on whether percent or absolute reduction in LDL is more meaningful – Dr. Sabatine expressed his view that the absolute LDL level achieved is key, again emphasizing that we could be lowering LDL cholesterol below current targets.
    • We’d love to see a sub-analysis of participants with diabetes at baseline, which was a substantial subset of the participant pool – 10,081 individuals, or ~37%. We’d be very curious to see the hard data reflecting whether or not patients with diabetes mirrored the overall results of FOURIER. This is especially important given the immense residual CV risk associated with type 2 diabetes, even with statins and other CV therapies. Additionally, taken together with the several recent positive diabetes CVOTs, there are exciting possibilities here. Could people with diabetes at high-risk for CV events reap synergistic benefits from SGLT-2 inhibitor empagliflozin and PCSK9 inhibitor evolocumab? This combination of therapies would be particularly interesting, since the three-point MACE CV benefit in EMPA-REG OUTCOME was driven by risk reduction for CV death likely from a heart failure benefit, while the benefit to the three-point MACE in FOURIER was largely driven by MI and stroke. Also along those lines, could there be promise in co-administration of GLP-1 agonists liraglutide  or semaglutide and evolocumab? How might we identify the best candidates for these various cardioprotective regimens (especially considering the enormously high cost and poor reimbursement of PCSK9 inhibitors currently)?
  • Safety findings: Injection site reactions were more common in the evolocumab group – 2.1% or 296 cases vs. 1.6% or 219 cases in the placebo group (p<0.001). All other adverse events – including allergic reaction, muscle-related events, rhabdomyolysis, cataracts, adjudicated new-onset diabetes, neurocognitive events, laboratory-confirmed aminotransferase level >3x upper limit of normal, and laboratory-confirmed serum creatinine >5x upper limit of normal – occurred with similar frequency in both study arms. In total, there were 10,664 adverse events among evolocumab-treated patients (77.4% event rate) and 10,644 adverse events among placebo-treated patients (77.4% event rate). Serious adverse events occurred in 3,410 evolocumab participants (24.8% event rate) vs. 3,404 placebo participants (24.7% event rate). Excluding individuals with pre-existing diabetes, the incidence of diabetes was 8.1% in the evolocumab group (677 new cases) vs. 7.7% in the placebo group (644 new cases), and this difference was not statistically significant – this was reassuring given the correlation between statins and new-onset diabetes. Discontinuation from the study was infrequent, and did not differ between groups. Dr. Sabatine highlighted that binding or neutralizing antibodies (which would mitigate Repatha’s efficacy) developed infrequently. In fact, there was no evidence of neutralizing antibodies developing in any evolocumab-treated patients.
  • During Q&A following the presentation, cardiology experts suggested risk stratification as a means of identifying which patients stand to benefit most from evolocumab therapy. This is not an unusual idea, considering the very high cost of PCSK9 inhibitors. Of course, this also introduces a dilemma – how can we pursue chronic disease prevention while also creating a system where you have to do worse to get a better therapy? We only hope that significant risk reduction for CV events, as demonstrated in this trial, will bring forth better reimbursement of Repatha. The ODYSSEY Outcomes CVOT for Sanofi/Regeneron’s Praluent (alirocumab), the other major PCSK9 inhibitor on the market, is scheduled to complete in February 2018.

-- by Payal Marathe, Helen Gao, and Kelly Close