Can Pay-For-Results Drug Pricing Work? A Q&A With Michael Sherman

Xconomy Boston — 

[Editor’s note: This is part of a series of posts sharing thoughts from industry and technology leaders about 2018 trends.]

It’s been almost a year since Harvard Pilgrim Health Care signed what’s known as a “value-based agreement” with Spark Therapeutics to cover Luxturna, Spark’s $850,000 gene therapy for a form of vision loss. In doing so, Harvard Pilgrim became the first insurer in the U.S. to try this new kind of payment model for gene therapy—one-time treatments that are meant to last for life, but aren’t guaranteed to. In the deal, Harvard Pilgrim pays for Luxturna, but gets rebates if the treatment wears off after a certain period of time.

This cemented Harvard Pilgrim’s position as a leader among insurers in crafting deals that tie payments for a drug to its performance. These value-based agreements, or VBAs (also known as outcomes-based agreements), are becoming more common between drugmakers and payers, and are touted as a way to control skyrocketing drug costs. Since 2015, Harvard Pilgrim has signed around 17 of these deals with companies like Spark (NASDAQ: ONCE), Alnylam Pharmaceuticals (NASDAQ: ALNY), and Amgen (NASDAQ: AMGN).

While such agreements are growing in number, it’s unclear whether they’ll make much a dent in drug spending, or whether they’ll be more widely adopted.

Over the last year, Michael Sherman, Harvard Pilgrim’s chief medical officer (pictured at an Xconomy event this year, second from right), says he has had more conversations with companies trying to find new ways to be paid for their drugs, especially for treatments with six- and seven-figure price tags. Xconomy recently spoke with Sherman about what he’s learned about VBAs, how they can be more effective, and whether they can really keep drug costs down. The conversation has been edited for length and clarity.

Xconomy: What trends do you see with VBAs?

Michael Sherman: I’m seeing more [of them] in the areas where we need these types of agreements. By that, I mean high-cost things like gene therapy, cancer therapy, the CAR-Ts [personalized cell immunotherapy for cancer], treatments for orphan diseases and others where there are high-cost therapeutics.

X: Why do you think VBAs are most needed in these areas?

MS: I think these are the areas where we can do the most good to ensure access and also to reduce spend in areas where the drugs are less efficacious. So gene therapy, for example, that’s an area where we as payers should be seeking to facilitate access, but doing so in a way where we’re also good stewards of the healthcare dollar.

Based on a lot of discussions and comments that have been made, my expectation is that some of these [gene therapies] will exceed $2 million for one-time treatment. Now that creates challenges to access for two reasons. The first is that you’ve got a huge upfront budget impact, where the benefit and cost offsets may be downstream.

So the question is how do you get over that hump up front? Paying over a number of years may be part of the solution.

The second concern is the uncertainty [of the treatment’s long-term effect]. The other point is that most treatments don’t work in 100 percent of individuals.

With gene therapies, to the extent they work in some and not in others or do not have long-term durability, doing innovative agreements— where we’re paying over a set number of years commensurate with ongoing demonstrations of effectiveness—allows us to align those aspects by allowing us to stop paying if the drug stops delivering results.

X: What about areas where VBAs haven’t been as effective in lowering costs?

MS: We have signed a number of [VBAs] for conditions from heart disease to [high cholesterol] to diabetes. Several of them have been successful in that they showed we can actually put in place an outcomes-based agreement and operationalize it. But in many cases, the actual impact on [the amount] spent has been minimal.

In some cases, that’s because these are already competitive [drug] categories, so you don’t have pricing which is very out of whack. Second, there tends be a greater comfort level [among insurers] with the [expected] results [of these drugs]. And third, the agreements are [often] built on preferred drugs within our formulary, which means there’s [already] a base rebate [a price discount given by drugmakers to insurers] and an [additional] amount that is at risk [for possible refunds to insurers].

Generally, pharmaceutical companies have only been willing to put a small [additional] amount at risk because of the base rebate.

Although I will say in some cases there is a benefit to thinking of [VBAs] as a performance guarantee.

X: VBAs have been criticized for not really lowering drug costs though.

MS: So one of the caveats here is that if you start with a drug that’s priced three times too high and have a VBA [where only a small amount of money is up for possible refund]—that, in itself, doesn’t necessarily help a lot.

Ideally you want to start with a fair price to get to something that is reasonable. [But] if there’s a therapy for an unmet need which is priced significantly higher than what we would all agree is fair, I would still rather have an outcomes-based agreement with dollars at risk for lack of a result versus not having that agreement and just paying that high price without that benefit.

The more nuanced answer is: I think it’s hard to get to sustainability and fair value without some constraint on initial pricing.

X: What’s changed since the Luxturna agreement?

MS: We’ve been having a somewhat unusual degree of discussion with companies that are maybe two or three years out from having an approved drug. Generally, they don’t get involved with payers that early. Yet we are having a number of innovative and collaborative discussions with companies like [Bluebird Bio (NASDAQ: BLUE), a gene therapy company], [Sarepta Therapeutics (NASDAQ: SRPT), working on gene therapy and rare disease drugs], and others about what these agreements might look like in the future, and then also with traditional companies like Merck around cancer and the like.

The big benefit of the Luxturna agreement is that it was very visible. And I think that visibility is part of why other gene therapy companies are seeking out discussions with myself and others.

X: What do you think is driving these companies to come talk to you earlier than usual?

MS: One interesting observation is because these drugs [for example, Luxturna, and Alnylam’s Onpattro, for a rare liver disease] at least to date have no competition, one could argue that the manufacturers really have not had to come to the table with an outcomes-based agreement.

But I think that they understand that as prices go higher and higher, there’s going to be push-back, in cases where the drug is administered and there’s not a good response. So as the dollars go up, I think there is more of a need and more of an understanding that the company should price [its drug] in that fashion.

X: How has one of your VBAs actually played out in the real world? How about the one with Amgen on its cholesterol-lowering drug Repatha?

MS: We have an agreement with Amgen as you know, and it’s got a couple of components. One component was essentially tying payments [for the drug] to reductions in cholesterol level. Well, that seemed like a good idea at the time and it’s still in effect, but at this point, no one questions whether someone taking [Repatha] will experience significant reductions in cholesterol. The drug is incredibly effective in doing so.

Then there’s the aspect [of the deal whereby] an individual experiencing a heart attack or stroke would trigger a full refund. I think this is more interesting—one that we think can provide value at least for a small number of individuals.

If people do have a cardiovascular event and there’s a refund, that seems fair. But if the drug is highly effective and no one has a cardiovascular event, we shouldn’t be unhappy about that either.

X: Can you give an example of where you think VBAs won’t make much of a difference?

MS: One example would be with the CGRP inhibitors [a new class of FDA-approved drugs for migraines].

Given that there are three [of these drugs] that have just come to market, we’re negotiating base rebates [from the manufacturers] in return for preferred access. But my sense is if they did something outcomes-based [in addition to the base rebate for being a preferred CGRP inhibitor], I’m not sure that we really move the needle much. And it would be a lot extra work. So that would be an example of where I said, ‘Thank you, but I don’t really think we need to go there.’

X: How common will VBAs be?

MS: I’ve been in more than a few discussions over the past few weeks and months, where we’re talking about agreements and asked if we’d like to have an outcomes-based component and in many cases I’m saying no. We could be doing more in number but there are many cases where I just don’t think it’ll make a difference and I don’t think it’s worth the work.

X: What needs to be done to make VBAs more effective?

MS: There’s been a lot of work engaging with [The Centers for Medicare and Medicaid Services] and [The U.S. Department of Health and Human Services] about what it would take to drive change that would prevent [what’s known as] “Medicaid best pricing” from being a barrier [Editor’s note: this is a policy that requires drugmakers to give state Medicaid programs the best price given to other purchasers.].

In the absence of a remedy to [that] problem, we think that the pharma companies won’t engage in the kind of substantive agreements that we’re all looking for. Right now that’s the biggest barrier.

X: Can you explain what this problem is?

MS: Say we enter into an agreement with Bluebird, for example, and we’re paying over four years and the drug stops working after one year. Therefore we’ve only paid 25 percent of the cost. [In that scenario, we’re asking CMS] not to then require Bluebird to sell [the drug] to all the Medicaid plans at 75 percent off of list [price]. Without that, the pharma companies won’t rationally engage in [VBAs]. That’s what we’re asking CMS to fix.