John M. Mandrola, MD


March 11, 2018

Professor P Gabriel Steg, MD, from the Coronary Care Unit, Hôpital Bichat-Claude Bernard, Paris, France, presented the ODYSSEY-Outcomes trial in the opening late-breaking clinical trial session here at the American College of Cardiology (ACC) 2018 Annual Scientific Session.

ODYSSEY investigators asked whether alirocumab (Praluent, Sanofi/Regeneron), a monoclonal antibody against PCSK9, would reduce cardiovascular (CV) morbidity and mortality after recent acute coronary syndrome (ACS) in patients who had elevated levels of atherogenic lipoproteins despite intensive or maximum-tolerated statin therapy. The study is not yet published.

By now, you have heard that ODYSSEY delivered positive results, and the Wall Street Journal reported that Regeneron and Sanofi have proposed lowered prices for alirocumab.

Here are my initial thoughts after the first day of ACC:

1. ODYSSEY was a massive and truly global trial. Almost 19,000 patients were randomly assigned from 1315 centers in 57 countries. The average length of follow-up was 34 months (8 months longer than FOURIER).

2. ODYSSEY enrolled patients after ACS. This is a higher-risk group than those enrolled in FOURIER, which studied patients with stable coronary heart disease. The mean age of patients in ODYSSEY was 58 years, and 25% were women.

3 In addition to ACS, patients enrolled in ODYSSEY had to have high levels of atherogenic lipoproteins despite statin therapy. Steg said that more than 90% of enrolled patients qualified on the basis of low-density lipoprotein cholesterol (LDL-C) levels ≥70 mg/dL, and a similar percentage took high-dose atorvastatin or rosuvastatin. ODYSSEY investigators also titrated drug dosing to achieve an LDL-C level <50 mg/dL.

4. A vital statistical consideration in the ODYSSEY trial design was that secondary outcomes would be analyzed in hierarchical fashion, with all-cause death analyzed after coronary heart disease and CV events. More on this later.

5. Alirocumab reduced LCL-C levels by about 50%, and this translated to a 1.6% absolute risk reduction in major adverse cardiac events (9.5% vs 11.1%; 15% relative risk reduction). Similar to FOURIER, nonfatal events (myocardial infarction, ischemic stroke, unstable angina) drove the reduction in the primary composite endpoint. Alirocumab did not reduce the "hardest" component of this endpoint: death from a coronary event.

6. For secondary endpoints, alirocumab reduced all-cause death by 0.6%, but statistical testing revealed a "nominal" P value of .026. The problem with making conclusions from this finding was the fact that alirocumab did not reduce death from coronary events, or death from any cardiovascular causes. If CHD and CV death aren't lower, how can one explain alirocumab's effect on all-cause death?

7. There were no safety issues during the nearly 3 years of follow-up. Combining the lack of safety issues in the two trials provides reassurance on drug safety, at least over the course of 3 years.

8. The most discussed aspect from ODYSSEY was the subgroup analysis by baseline LDL-C. Here the investigators analyzed outcomes based on 3 levels of LDL-C: <80 mg/dL, 80 to 100 mg/dL, and 100 mg/dL or higher. Alirocumab did not significantly lower either the primary outcome or all-cause death in the first two groups, but did show significant reductions in the group with LDL-C levels of 100 mg/dL or higher.

Specifically, in this group, alirocumab reduced the risk for the composite primary endpoint by 3.4%, and all-cause mortality by 1.7%. The P value for interaction was not significant, at 0.09, but Andrew Foy, MD, from Penn State University, Hershey, Pennsylvania, told me that the interaction with baseline LDL-C would have been significant if the investigators had split the patients into two groups, rather than three (eg, <100 mg/dL vs 100 mg/dL or more). Foy's point was that this is real interaction and likely represents a heterogeneous treatment effect.

9a. Two experts I contacted had differing views on the magnitude of ODYSSEY's positive results. By email, Sanjay Kaul, MD, from Cedars Sinai, Los Angeles, California, explained that he was less impressed with the mortality reduction. Kaul called the 0.6% absolute risk reduction in overall mortality "not robust," adding that the lack of coronary heart disease death and CV death further weakens the mortality argument. He also wrote that per the statistical rules of engagement, when a lower hierarchical endpoint reaches significance but the ones above do not, this is an exploratory finding. Indeed, Steg used the phrasing "suggests a reduction in all-cause mortality" during his presentation.

9b. In an email, Professor Darrel Francis, MD, PhD, from Imperial College, London, United Kingdom, was more impressed. He suggested that the 15% relative risk reduction in the primary endpoint may be an underestimate of the cumulative benefit of the PCSK9 inhibitor.

Francis said ODYSSEY delivered "a startlingly positive result for CV events and all-cause mortality. Moreover, the hazard ratio reduction develops at 1 to 1.5 years, which means that during a lifetime of treatment, the long-term hazard ratio is likely to be even better.... Any doubt I had that LDL is bad for you is now dispelled."

10. Cost matters a lot: Kaul showed me some sobering economic figures: An absolute risk reduction of 1.6% in the primary endpoint translates to a number needed to treat of 64. Using the current price of $14,500 per year, Kaul calculated that preventing one event over the trial period of almost 3 years would cost about $2.6 million.

The numbers are a little better for the subgroup with a LDL-C level of 100 mg/dL or higher. Here the number needed to treat is 29 and the cost per major adverse cardiovascular event prevented is $1.18 million. Cutting the cost of the drug by half still leaves a cost of $590K per event prevented.


It feels weird when a winning trial results in lower prices for a novel drug. One might think lower prices means more patients could get these medications. Market penetration would increase, right?

I think it could be the opposite. Even with major reductions in price, the PCSK9 inhibitors will remain expensive, especially when one considers the duration of therapy.

The stronger force limiting market share for these drugs may be the subgroup findings. If experts or payers deem that PCSK9 inhibitors are a "value" only in high-risk patients post-ACS with LDL-C levels at or above 100 mg/dL receiving standard lipid-lowering therapy, then only a very small group of patients will be eligible for them.

Perhaps it took us two massive trials of these novel drugs to bring us back to the old idea of getting LDL-C levels below 100 mg/dL. In the vast majority of cases, that can be done with an inexpensive generic statin.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.