The ODYSSEY Trial Ends Well— But Will It Be Enough?

Sanofi and Regeneron achieved a modest victory with the much anticipated ODYSSEY Outcomes trial of its cholesterol lowering drug Praluent (alirocumab). The trial met its primary endpoint and even reported a significant improvement in all-cause mortality. But the mortality finding has an asterisk attached to it and it is far from clear whether the overall trial is positive enough to significantly alter the cost-benefit concerns that have so far been the chief barrier to wider use of alirocumab and its close cousin, Amgen’s PCSK9 inhibitor evolocumab (Repatha).

The ODYSSEY Outcomes trial was presented on Saturday morning in Orlando, Florida at the annual meeting of the American College of Cardiology.

ODYSSEY compared alirocumab (Praluent, Sanofi/Regeneron) with placebo in 18,924 patients with a recent (1-12 months) heart attack (MI) or unstable angina. The trial met its primary endpoint, reducing the primary endpoint (the combination of MI, stroke, death from coronary heart disease, or unstable angina requiring hospitalization) by 15% (HR=0.85, CI: 0.78-0.93, p=0.0003).

Individual components of the primary endpoint:

  • CHD death: 2.2% for alirocumab versus 2.3%, p=0.38
  • Non-fatal MI: 6.6% versus 7.6%, HR 0.86 (CI 0.77 – 0.96), p = 0.006
  • Ischemic stroke: 1.2% versus 1.6%, HR 0.73 (CI 0.57 – 0.93), p = 0.01
  • Unstable angina: 0.4% versus 0.6%, HR 0.61 (CI 0.41 – 0.92), p = 0.02

The investigators also reported a significant reduction in overall mortality, from 4.1% in the placebo group to 3.5% in the alirocumab group (HR=0.85; CI: 0.73-0.98, p = 0.026). Because, in accordance with the predetermined statistical analysis plan, the reduction in CHD mortality did not achieve statistical significance, the reduction in overall mortality was considered an observational finding. As a result, this means the company will not be able to make a mortality reduction claim without qualification. Sanjay Kaul (Cedars-Sinai) commented that the mortality finding  should not be considered robust.

There will almost certainly be a significant amount of discussion about the subgroup of patients who had high levels of LDL at baseline (100 mg/dl or above). Nearly all the benefit in the trial was observed in this subgroup. MACE was reduced by 24%, from 14.9% in the placebo group to 11.5% in the alirocumab group (HR=0.76, CI 0.65-0.87) and overall mortality by 29% (HR=0.71, CI 0.56-0.90). Regeneron and Sanofi estimate that about 1.3 million people in the US fall into the subgroup of ACS patients with persistent LDL levels above 100 mg/dl despite taking statins.

Another important finding is that no new safety issues emerged in the trial. In particular, there was no excess of either diabetes or neurocognitive side effects in the trial, thereby helping to settle an important lingering question about the PCSK9 inhibitors. However, as expected, patients in the alirocumab group did have more injection site reactions.


ODYSSEY will likely be seen as both consistent with FOURIER, the earlier cardiovascular outcomes trial (CVOT) with Amgen’s evolocumab, and an incremental advance on that trial. Despite differences in patient populations and trial design, both trials resulted in a significant but modest 15% reduction in the primary endpoint. But the ODYSSEY mortality results and the subgroup results in patients with baseline LDL levels over 100 mg/dl appear to confer a decided advantage for ODYSSEY. The results will likely be seen as confirming the view that Amgen may have shot itself in the foot with FOURIER.  In an attempt to gain a competitive advantage over Sanofi/Regeneron, Amgen increased the size of FOURIER to accumulate more endpoint events in as short a period of time as possible. As a result FOURIER was able to report results one year before ODYSSEY. But the downside is that FOURIER may have been unable to reflect the full benefit of evolocumab, since many events in the trial occurred before the beneficial effects of the drug had time to emerge.

It is not at all clear whether the modest incremental benefit of ODYSSEY will lead to significantly increased uptake of the drug. Insurance companies have raised nearly insuperable barriers to reimbursement. (The drugs now have a list price of more than $14,000 per year but are usually discounted to about $9,000.) Many cardiologists have tried and failed to get these drugs approved for their high risk patients with persistently high cholesterol levels even after taking or failing to tolerate statins. ODYSSEY does not appear likely to substantially alter the cost benefit equation.

Some experts speculate that Regeneron and Sanofi will use the results of ODYSSEY to capture market share from Amgen. Regeneron and Sanofi reported worldwide alirocumab sales in 2017 of only $194 million, $131 million of which was in the US. Amgen’s evolocumab had worldwide sales in 2017 of $319 million, $225 million of which was in the US. Regeneron and Sanofi could also use the trial to try to expand the overall marketplace for PCSK9 inhibitors.

Gregory Schwartz (University of Colorado), the principal investigator of ODYSSEY, said he didn’t know if the results would be strong enough to alter concerns about cost. “It will be our peers that will decide this,” he said. The mortality finding is “fundamentally different,” because it matters to both physicians and patients. “With all the caveats of the nominal p value this is plausible and fits with decades of our understanding of the biology,” said Schwartz.

Phillippe Gabriel Steg (Imperial College of London and University of Paris), the chair of the trial, said that “if the drug was very cheap then we would use it in everybody.” He said it was “the intuitively obvious choice” to favor usage in higher risk post ACS patients with LDL levels over 100 mg/dl.
Milton Packer (Baylor Health) pointed out that the results will make it more difficult for companies to study new lipid lowering drugs, since the findings diminish the urgency to lower LDL cholesterol in patients with LDL levels below 100 mg/dl. ODYSSEY, he said, may represent the end of the campaign to consistently drive LDL levels ever lower.
Sanjay Kaul said that in his opinion, “the trial results do not move the needle for PCSK9 inhibitors! I don’t see how the trial results will encourage the payers to be more favorably disposed to this drug unless of course they are convinced/persuaded of the mortality benefit!!”
James Stein said that it “makes sense with everything we know from pathophysiology and other clinical trials” that “people with higher LDL may have benefited more.” He said that PCSK9 inhibitors are safe and effective but that “the challenge will be figuring out the subgroups of who benefits the most and convincing payers to cover this medication at its current price.”In the meantime,” said Stein, “I will continue to add ezetimbe to high dose statin and maximize lifestyle changes, reserving these wonderful new medications for those with the highest LDL values or highest risk.  If their price were lower, however, it would be a no-brainer to use them much more.  We have proven efficacy and safety in the medium term now, so cost is the major barrier.”


  1. Viewed as a GROT this is unsatisfying. They administered the drug to very ill people (with a recent (1-12 months) heart attack (MI) or unstable angina) and yet “the company will not be able to make a mortality reduction claim without qualification. Sanjay Kaul (Cedars-Sinai) commented that the mortality finding should not be considered robust.”

    For the sake of argument let’s suppose it is right, though. I, and perhaps others, would appreciate a summary along the lines of “over a treatment and observation period of 60 months overall mortality was reduced by 0.6 %age points. From this we calculate that a one year treatment period would extend life by an average of two days” (or whatever number is, in fact, implied).

  2. “[S]aid Stein, “I will continue to add ezetimibe to high dose statin and maximize lifestyle changes, reserving these wonderful new medications for those with the highest LDL values or highest risk.”
    Let’s not forget that ezetimibe only helped diabetics and those over 75.
    See the IMPROVE-IT prespecified subgroup results hidden away on pages 40 & 41 of the Supplemental Appendix. &
    Patients younger than 75 or non-diabetic patients saw no benefit. (Nor did the USA patients…)
    Oh,well, so much for practicing evidence-based medicine…

    Would patients pay for alirocumab themselves if told “CHD death: 2.2% for alirocumab versus 2.3% for placebo”? (And even that is including all those >100 mg/dl-LDL patients…)

  3. Here’s a hand-waving sort of calculation: refinement or correction welcomed.

    Over one year the reduction in mortality is 0.6%/5 i.e. 0.12% or about one in a thousand. So after an extra one thousandth of a year elapses the treated group would have as many deaths as the untreated group had in exactly a year. So the average extra lifespan is roughly half a day. Moreover, it’s half a day in the life of a seriously ill individual, not half a day of being hale and hearty. Hell, if the poor chap is ill enough he might sleep all that half day anyway.

    That’s lousy value for $9,000, unless of course it’s $9,000 of somebody else’s money.

  4. Another shot. Rate of reduction in mortality = 0.12% per year of treatment. So imagine treating 800 patients in this ill group for one year. Absolute reduction in mortality = 800 x 0.12% = 1.

    You’d have to treat 800 patients to extend the life of one of them beyond the year in question compared to 800 people who went untreated.

    Hence cost of drug = $9,000 x 800 = $720,000. I think we may take it therefore that the cost of treatment will comfortably exceed a million dollars. It still looks to be lousy value.

  5. Oops, not $720,000 but $7,200,000. I am getting old: I never used to make that sort of howler.

  6. Placebo controled?

    2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment
    elevation states: Lipid-lowering treatment
    It is recommended to initiate high-intensity statin therapy [i.e. statin regimens that reduce low-density lipoprotein (LDL) cholesterol by ~50%] as early as possible after admission in all NSTE-ACS patients (in the absence of contraindications).

    2014 AHA/ACC Guideline for the Management of Patients With Non–ST-Elevation Acute Coronary Syndromes Cholesterol Management
    Class I
    1. High-intensity statin therapy should be initiated or continued in all patients with NSTE-ACS and no contraindications to its use (269-273). (Level of Evidence: A)

    Ethicaly questionable trial

Speak Your Mind