Bijal D. Shah, MD, discussed the challenge of treating this highly heterogeneous disease, how to choose the optimal treatment regimen, and the ongoing role of BTK inhibitors.
Bijal D. Shah, MD
The current state of treatment for patients with mantle cell lymphoma (MCL), continues to be challenging, according to Bijal D. Shah, MD.
MCL accounts for about 6% of all NHL diagnoses in the United States, and high-risk patients comprise 10% to 15% of MCLs. An analysis of the National Cancer Institute SEER database showed an age-adjusted annual incidence of 1.01 per 100,000 population from 1995 to 2013, with a steady annual increase of 2.56%.1
The FDA has approved a number of agents to treat patients with MCL since the late 1990s including rituximab (Rituxan), bortezomib (Velcade), temsirolimus (Torisel), bendamustine, lenalidomide (Revlimid), ibrutinib (Imbruvica), and acalabrutinib (Calquence). As a result, the risk for MCL-specific death declined significantly from 1995 to 1998 and again from 2009 to 2013 (HR, 0.58; 95% CI, 0.53-0.81).2 However, only patients with advanced disease demonstrated a significantly reduced risk for death (P <.0001) from SEER data sets.
Nonetheless, Shah, an associate member in the Department of Malignant Hematology at Moffitt Cancer Center, who delivered a presentation at the OncLive® State of the Science Summit™ on Hematologic Malignancies, said physicians and investigators are struggling to advance the field, in part because novel therapies are rarely evaluated in MCL-specific trials.
In an interview with OncLive, Shah discussed the challenge of treating this highly heterogeneous disease, how to choose the optimal treatment regimen, and the ongoing role of BTK inhibitors.Shah: MCL is a clinically and genomically heterogeneous disease. One of the reasons I first got excited about MCL is that it's a very rare lymphoma. I thought I could figure this out and contribute something major, because how much variability can there be in this one lymphoma that makes up 6% to 8% of B-cell lymphomas? I learned very quickly that there is quite a bit of variability, in fact. One of our first papers was a review paper titled, “Mantle cell lymphoma is a clinically heterogeneous disease.” That's the first thing I would want people to take home: there is no one-size-fits-all [approach].
We are starting to learn those features that drive a worse outcome. Whether we talk about this in terms of the Mantle Cell Lymphoma International Prognostic Index (MIPI) or whether we talk about this genomically, when you look at high-risk patients, they are typically more proliferative and also commonly carry these P53 mutations. This was very nicely described by the MCL35 assay. This gene-expression profiling tries to capture more proliferative versus less proliferative MCLs, and then overlap that with the MIPI and the combined MIPI (MIPI-c), which includes a measure of Ki-67 and P53 mutation status. It's highlighting this intersection of the clinical presentation and the biological behavior.
As we step back, we think about those patients who are more proliferative, the ones who carry these P53 mutations—the way that Dr Martin Dreyling and others would describe this as a distinct subtype of MCL. When they started looking at their outcomes on the MCL2 and MCL3 trials, the high-dose induction followed by stem cell transplant studies, what they very clearly saw was these patients—in spite of being younger—did much, much worse. That's another thing I need to drive home: high-dose chemotherapy/stem cell transplant-based approaches are not applicable to this subset of MCL.Lenalidomide and rituximab is one novel way of approaching these patients without cytotoxic chemotherapy. Bortezomib is another interesting approach. One of the things that has challenged us in the use of bortezomib is administering it chronically, particularly if we're talking about giving it as a single agent but even in the context of a CHOP-like regimen. What are we really entailing in terms of neuropathy, not just in terms of severity but also duration? That's not to say we don't use CHOP with bortezomib, but we have to be cognizant of what we entail when we do so.We talked about ibrutinib and acalabrutinib during the presentation, and the first thing I wanted to do was compare and contrast. The order that I picked was on purpose. I wanted to make it look like acalabrutinib was a “slam dunk, super winner, clearly the best BTK inhibitor, and we should be using it,” before coming back to say that these are 2 very distinct trials. Even though they did use the ibrutinib trial as a model, we already knew what ibrutinib had to offer in MCL. We were enrolling these patients much earlier in their treatment course. They were less heavily pretreated, less likely to be proliferative, and less likely to have these P53 mutations. Having been a part of that study, I can say that it tended to be a better bunch of patients, at least the ones we enrolled.
When you start comparing the progression-free and overall survival, you have to look at that carefully and determine if it makes sense. That's why I said in the presentation, "Ask yourself: does it make sense that a drug that is more selective for BTK would be more effective?" It really doesn't. One would intuit that broader enzymatic inhibition, particularly when we talk about Src kinases and the other kinases that we hit, should come with a better outcome.
The second point I wanted to make was related to adverse events (AEs). It's very tempting to say that acalabrutinib comes with less bleeding. Certainly, there was less severe bleeding relative to the ibrutinib-treated patients, but there were also slightly fewer patients who were anticoagulated. When we start thinking about minor bleeding rates, contusions, petechiae, epistaxis, and so on, they look the same. When you start asking whether one drug is better than the other when it comes to bleeding, I would argue that they're pretty similar. I wouldn't decide based on that.
Another very important observation is that roughly half of the ibrutinib-treated patients were still on anticoagulation and around 40% were on it on the acalabrutinib study. This means that if you need to anticoagulate, you still can. You just have to be aware of the risk. That is something that gets overlooked commonly.
As we think about more specific AEs, we saw more headache with acalabrutinib and more atrial fibrillation on the ibrutinib study. Otherwise, [rates of] myalgias and diarrhea are all pretty much in line with each other. I'm not saying they're exactly the same, but the rates are pretty similar.Not a whole heck of a lot. When we start thinking about how these drugs play out in these biologically heterogeneous populations, we're seeing the same thing. The low-risk, less-proliferative, P53 wild-type patients do great. These are the same patients who did great with transplant. However, the P53-mutated, more-proliferative patients aren't doing so hot. They're progressing quickly and when they progress, their survival is exceptionally poor; survival is around 2 months in our study. There is this tail of around 20% we can salvage, but the other 80% is who we really struggle with.
These patients are often progressing in the first few months of taking ibrutinib. That opens up so many opportunities when we think about using minimal residual disease to guide the addition of other drugs, to start fixing this problem. “What can I put into this space once I know ibrutinib isn't going to cut it?” Therefore, now we not only have predictive things like P53 and proliferative status, but we also have something else we can use before we get to that 6-month endpoint where we are seeing those poor outcomes. One or 2 months in, let's see what we're accomplishing, particularly in high-risk subsets.One of the challenges of looking at the ibrutinib/venetoclax dataset is distinguishing it from the acalabrutinib data, the single-agent ibrutinib data, the ibrutinib/rituximab data, or the ibrutinib/lenalidomide/rituximab data. How do we differentiate these studies because they're all coming out right about the same? What these data are telling us is that even the addition of venetoclax isn't saving us when it comes to these high-risk subsets. My guess is we're not going to see, especially in more-proliferative MCLs, rituximab or even lenalidomide/rituximab overcome that risk feature. Therefore, we need to start thinking critically about how we're going to address this high-risk subset.
High-risk patients are less common. However, But what we observed as we started sequencing patients is P53 mutations at baseline at a much higher rate. If we take all patients we have sequenced over the past couple of years, about 50% are coming to us with P53 mutations, either in the de novo or relapsed setting. That's a big problem. The number of patients who come to us with proliferative disease overlaps closely with this P53 mutation status, and this is a big problem. Of the 100 or so patients we're going to see in the next year, half are going to be in a bad place and we need to come up with novel approaches to address that.Not a whole lot. There are about 50 ongoing trials that are actively recruiting right now, 15 of which are MCL-specific and only 5 that are not based around BTK inhibitors. Only 2 of those include patients who have progressed on a prior BTK inhibitor. That's bad. If we move from the chemotherapy or novel agent space into the chimeric antigen receptor (CAR) T-cell space, we've only got 1 contender that is targeting MCL specifically: the ZUMA-2 trial. Everybody else just lumped MCL patients in. Are we going to have meaningful endpoints from the other 45 trials? [What about] from the other CAR T-cell therapy trials? Or, is this going to be one of those, “Well, this might work. We saw in the 5 patients we enrolled that there was this much or this little activity.” How do we grow from that?
One of the first things I encountered when I wanted to focus on MCL was: how do I tease out from all of these novel agent trials where there is a signal of efficacy when they only enroll a handful of patients? The answer is: I couldn't. There was no meaningful signal. If you saw 2 out of 5 people respond, you had no idea what that meant. Therefore, if we really want to push the bar forward, if we really want to develop novel therapies in MCL, we're going to have to carve out MCL-specific trials. Otherwise we're going to have that same biological heterogeneity play out with mixed endpoints that won't inform anything in the clinic.