The panel elaborate on how they’ve incorporated RWE into their decision-making processes for their patients with MM.
Ryan Haumschild, PharmD, MS, MBA: Roy, there aren’t many people with greater years of experience in the payer area than you. You bring a good perspective to real-world evidence [RWE] and where it fits in. I’d like to get you to share your experience with real-world evidence and where you think it’s headed. At the end could, maybe you can touch on when we think about diversity, equity, and inclusion and health equity and creating care pathways for small patient populations. Is that going to be doable in the future, from a payer perspective?
Roy Beveridge, MD: Real-world evidence is important. Muhammad and Josh [Richter] were eloquent when they spoke. Everyone knew a particular regimen was superior, but it took 8 years to prove it in a randomized trial. That doesn’t do anyone any good. Whether you’re Jay [Weaver], me, or anyone, payers want to pay for appropriate treatments. They just want to make sure from an actuarial standpoint, that what they’re putting into their calculations and charging as premiums are valid. I don’t think waiting 8 years is good for patient care; it’s not right.
At the same time, RWE is inherently imprecise for all the reasons everyone has just spoken about. There has to be a balance in using it appropriately to push for certain things. Depth of response is important; we all believe that. But sometimes you have to go to smaller populations to answer questions. Josh, Muhammad, you know this way better than I do because I stopped seeing patients some time ago. Most of these patients with myeloma are older, with a significant number of comorbidities. What we saw at Humana, when we looked at RWE, was that there was a reason—Muhammad, you’re exactly right—the vast majority of these people couldn’t follow and get to the same end points, based on the FDA trials. That’s because they’ve got kidney function, which is inherently significantly worse, and LFTs [liver function tests] that aren’t good. So it’s expected. That’s where RWE should be used, to advance things that are obvious. We may still need to do randomized clinical trials, but sometimes we’re going to take the RWE data and say this makes sense.
The proviso with all that is that there’s some agreement among the very smart clinical experts, so the payers don’t get confused. Not that Jay would ever get confused, but I always got confused. I didn’t know sometimes. I’d call up my friend and say, “What do you think about such and such,” and they’d say, “This is the way I think about it.” Then you’d talk to another person and get a diametrically opposed point of view. That’s not what we need. The NCCN [National Comprehensive Cancer Network] Guidelines have to be maintained appropriately and quickly. We all acknowledge that it’s very complex, but having a standard is what’s going to move care faster for the majority of people.
With certain diseases, if we don’t have plasma cell leukemia and things that, we’re never going to have enough data. That’s the time you pick up the telephone, speak with the medical director of whatever institution, and say, “Here are the data.” They’ll verify that this is what’s going on. That’s where we have to get to because we’re never going to have pathways to include every person. We may get there at 80%, and that’s great; 20%, we’re going to spend some time thinking about. Most payers, as I said, are very comfortable with 20% off pathway. Ryan, you must see that with your pathways and things because you can’t fit everyone in every time. That’s not going to happen.
Ryan Haumschild, PharmD, MS, MBA: That was well put. I like the way you said that. I’m a big believer in that you’ve got to have some consensus, but there are always exceptions. You can plan for that. Muhammad, I want to ask you a question. As you’re practicing, you see some of these one-off examples with real-world evidence, or something of trend, that you’re seeing in practice and would like to see incorporated. But you might run into a barrier in terms of prescribing, or it might not be the consensus or a payer. You might have to do an appeal every time you want to do this. Talk me through it: do you have any examples or areas where you feel like you can incorporate this in your own practice?
Muhamed Baljević, MD, FACP: I’ve had the privilege of representing the last institutions where I’ve worked on NCCN Guidelines. I had an patient with MGRS [monoclonal gammopathy of renal significance]. As everybody knows, MRGS is a recent entity. I had trouble getting 1 of my insurance patients to autologous transplant consolidation. The person was losing renal function rapidly, spilling a lot of proteins. With adequate treatment, we reversed a lot of that and normalized their kidney function. It was natural to consolidate them, but I couldn’t do it. Consciously, we simulated all the data and, to Roy’s point, made sure that we kept them updated and included the whole section on MGRS. We explained that even though it’s an esoteric entity that requires a particular criterion to diagnose, you manage it like a malignancy, depending on whether it’s going down the myeloma path or CLL [chronic lymphocytic leukemia]. The reality is that you have to manage it aggressively, like a formula, because it’s compromising your end organ.
We’ve learned from these N-of-1 trials. I’m noticing that we’re getting better at realizing that we have huge data holes and data gaps. Our ability to perform randomized controlled trials on some of these infrequent diseases is a realistic challenge. I don’t know that we will necessarily cross it, despite the best intention.
I wanted to mention something Roy mentioned earlier, and that’s the fact that our patients are older. Median age about 69, 70 years old. More than half our patients will be fairly older, and it’s important to understand that and respect that. I have to admit, some of these data, I wasn’t even fully aware of it. But once you read the studies that were done from the real-world side, you quickly learn that the attrition rate and your ability to give subsequent lines of therapy to a patient with myeloma is fairly staggering. Some data argue that even from 1 line to second line, you may lose up to 35% of patients. By the end of the second line, you lose another 25% or so. By the time you’ve given somebody 2 lines of therapy, just over half of the patients that you started with will have either progressed or died or had toxicity that precluded them from being considered for subsequent lines of therapy.
That brings into focus very acutely how you choose your therapies and whether you should be thinking about putting your best foot forward. There were even some provocative data at ASH [American Society of Hematology Annual Meeting] from Arizona arguing that incorporation of some of these CD38 antibodies has survival benefit if you put them in the first line vs if you wait and do it in the second line. This is a pain point. We have a newly diagnosed, transplant-ineligible patient population with 2 category 1 recommendations—VRd [bortezomib, lenalidomide, dexamethasone] on 1 side, DRd [daratumumab, lenalidomide, dexamethasone] on the other—yet we don’t have a head-to-head trial.
What do you choose? What’s better for 1 group of patients vs the other? This is where data modeling, real-world experience, and data pooling can help us look at the greater numbers of patients and try to extrapolate valuable pieces of data. We’re working hard at doing the studies we think we have huge gaps in knowledge, but all this takes a long time in a disease where PFS [progression-free survival] is. Survivals are getting so much longer that we’re rapidly forced, for a good reason, to consider surrogate end points.
Josh pointed out earlier that the FDA is considering that a valid end point in myeloma. Myeloma is a poster child for that. I don’t know if we’re going to touch on MRD [minimal residual disease]. We could have another session on that. That’s 1 of the critical things. And all these things cost money. Running a study for many years costs hundreds of millions of dollars. There’s an incredible amount of competing interest and values for us to learn these things. We’ll draw from real-world evidence as much as we can. It’s going to be important because otherwise we have a huge lag and a huge catchup by the time we produce something down the conventional path.
This transcript has been edited for clarity.