PCSK9 inhibitors have ‘the dubious distinction of being the most expensive preventive therapies by far in the history of cardiovascular medicine’

Two new economic analyses conclude that PCSK9 inhibitors are far too expensive to be cost effective. Both studies incorporate data from FOURIER, the first and still the only large cardiovascular outcomes trial with a PCSK9 inhibitor.

The first paper, published in JAMA, is an update of an earlier 2016 paper which concluded that in order to be cost effective the price of PCSK9 inhibitors would need to be reduced by two-thirds.

After incorporating the results of FOURIER into their analysis, the new paper, by Dhruv Kazi and colleagues at UCSF, goes further, concluding that even greater price reductions are required in order for the drugs to be cost effective.

The authors calculated that the price would now need to drop by 71% to reach the conventional level of cost effectiveness.

Further, they pointed out that a rebate model being proposed by one manufacturer, in which the cost of the drug would be refunded in the event a patient suffered a cardiovascular event while taking the drug, would be “unlikely to meaningfully reduce drug expenditures given the low overall MACE rate (approximately 3% per year).”

“What should be done about a therapeutic innovation that provides incremental clinical benefits relevant to a large group of patients, appears, so far, to have acceptable safety, and yet is neither cost-effective nor affordable?” ask Dan Mark and Kevin Schulman (Duke University), in an accompanying editorial.

They argue that “draconian restrictions on access to drugs that are priced for profit maximization out of proportion to any value proposition and budget tolerances may continue to be the only way medicine can send a strong signal” that we need to fix “a deeply flawed system.”

The second paper, by Fonarow et al, is published in JAMA Cardiology. The study was sponsored by Amgen and the authors are FOURIER investigators and Amgen employees.

It is therefore even more remarkable that this group is unable to present a convincing case for PCSK9 inhibitors.

In one of the more optimistic models they found that for treating patients who met the FOURIER entry criteria the price would need to drop to $6,780 to be cost effective. But even this scenario assumes that the short term benefits of FOURIER will eventually lead to a long term improvement in cardiovascular mortality.

However, “in the scenario that no cardiovascular mortality benefit was to emerge,” the authors said the drug would be even less cost effective, resulting in a cost of $483,800 per QALY gained, “substantially exceeding current intermediate value and cost-effectiveness thresholds.”

 In a remarkable coincidence, Dan Mark was also the first author of an editorial about the JAMA Cardiology paper. Mark, along with Ilana Richman and Mark A. Hlatky, writes that the scientific “breakthrough” represented by the PCSK9 inhibitors has “been overshadowed by another breakthrough of PCSK9i therapy: its high cost.”

 The negative conclusions reached by Fonarow et al in the JAMA Cardiology paper “likely represent best-case scenarios, with several model assumptions that are more optimistic than the available data warrant.

For example, the investigators assume that PCSK9i-treated patients will experience a survival advantage after 5 years of therapy, although FOURIER found no evidence of an effect of PCSK9i therapy on either cardiovascular or all-cause mortality.”

Sanjay Kaul (Cedars Sinai) had a similar criticism of the paper. He also pointed to the CV mortality problem, noting that the assumptions used by the authors may have unduly favored PCSK9 inhibition.

“The model assumes a time-dependent increase in MI benefit supported by the landmark analysis, but ignores the time-dependent increase in CV mortality (the magnitude of benefit and harm being similar) based on the same landmark analysis.

So, if one is modeling larger projections for MI benefit over time, then one also has to project greater mortality harm over time, not the benefit as this model wrongly assumes!” Kaul also said that the “utility values for MI and stroke prevention are not justified because we don’t know the types of MI or stroke events or their clinical importance.

For example, we don’t know if treatment prevented large disabling strokes or prevented large MIs for which we can reliably assign credible dysutility (or utility).”

Mark et al wrote that although different groups of investigators have used different models, “these independent analyses reinforce the conclusion that PCSK9i therapy currently does not provide good value for the money.”

 In their analysis Fonarow et al wrote that the cost effectiveness could be enhanced by— obviously— lowering the cost of the drug, or, alternatively, targeting a smaller but higher risk subset of patients.

This statement was echoed by Mark et al in their JAMA Cardiology editorial: “Solutions proposed thus far for unraveling this Gordian knot include a radical lowering of the price (which so far seems very unlikely) or a radical restriction of therapy to patients with a markedly elevated risk.”

“With list prices more than $14,000 per year, these drugs have achieved the dubious distinction of being the most expensive preventive therapies by far in the history of cardiovascular medicine,” wrote Mark and co-authors.

Update: The Spin Zone

Amgen is trying to persuade the world that the Fonarow analysis makes a good case for increased use of Repatha. Unfortunately for the company, although the paper is less negative than the JAMA paper, it hardly delivers the evidence the company needs to convince anyone outside the Amgen orbit that its drug is cost effective.

If you had only read the Amgen press release you would think the Fonarow study presented a glowing endorsement of Repatha. Here are the key sentences in the press release:

Repatha was found to be a cost-effective treatment for patients with established ASCVD in the U.S. when the net price is at or below $9,669 per year…. This estimate of the value-based price range for Repatha is in line with the discount range typical for biologic medicines in the U.S. market when applied to the U.S. list price of Repatha.

Unfortunately for Amgen, this is not the message the authors— all of whom either work for Amgen or have had their research funded by Amgen— delivered in their paper. Here is their conclusion:

At its current list price of $14 523, the addition of evolocumab to standard background therapy in patients with atherosclerotic cardiovascular disease exceeds generally accepted cost-effectiveness thresholds. To achieve an ICER of $150 000 per QALY, the annual net price would need to be substantially lower ($9669 for US clinical practice and $6780 for trial participants), or a higher-risk population would need to be treated.

Now it should be acknowledged that economic analyses can be a fertile grove for cherry pickers, and it’s always possible to generate a model that can support increased usage.

But we now have these 2 analyses, along with the ICER analysis, and the recent analysis looking at a drug price that reflects the real world cost based on discounts and refunds, all showing that PCSK9 inhibitors are not generally cost effective.

There’s an awful lot of variation in these analyses, but the fact that even the most industry-friendly group expresses strong reservations about the cost effectiveness of the drugs means that we all need to bring a high level of skepticism to any effort to commit significant resources to these drugs at the present time.