Best therapies for relapsed diffuse large b-cell lymphoma

MD Anderson Cancer Center
Date: 07/28/2014


 [Background Music]

Lisa Garvin: Welcome to Cancer Newsline, a podcast series from the University of Texas MD Anderson Cancer Center. Cancer Newsline helps you stay current with the news on cancer research, diagnosis, treatment and prevention, providing the latest information on reducing your family's cancer risk. I'm your host, Lisa Garvin. Today, we're talking with Dr. Jason Westin who is an assistant professor in the Lymphoma and Myeloma Department. He's one of our returned guests. And we're going to be talking about a particular type of lymphoma. Welcome, Dr. Westin.

Dr. Jason Westin: Thank you, Lisa.

Lisa Garvin:  Now, we are talking about diffuse large B-cell lymphoma.

Dr. Jason Westin: Yeah, it's a mouthful, isn't it?

Lisa Garvin: Yes. Tell me what that's about.

Dr. Jason Westin: Diffuse large B-cell lymphoma or large-cell lymphoma for short is the most common lymphoid malignancy seen in adults. So, it's a pretty common disease. It's about 30,000 patients are newly diagnosed each year in the United States.

Lisa Garvin: And I believe that's classified as an intermediate grade type of lymphoma?

Dr. Jason Westin: Yeah, it typically is. It has more aggressive features than what we call indolent lymphomas. So, it's not the most aggressive lymphoma which would a Burkitt's or some other type of extremely aggressive lymphoma but it has aggressive features and is generally treated with aggressive therapies.

Lisa Garvin: And, you know, I was looking at some old publications in there where as a regimen called CHOP that was the standard treatment of about 10 years ago.

Dr. Jason Westin: Well, unfortunately, you could the read from today and that would be the same treatment. That treatment has been around for about 35 years. The only modification we've made is adding rituximab about 10 years ago. So, it used to be just CHOP alone about 10, 15 years ago. Since then, it's been R-CHOP but it's basically a similar therapy, minor change by adding rituximab.

Lisa Garvin: And what are the four--what is CHOP, C-H-O-P, what are the drugs?

Dr. Jason Westin: So, we try and make things confusing. As doctors, we don't want our patients to know too much. I'm just kidding, but the C is cyclophosphamide. The H is a drug called doxorubicin, makes sense. It used to be called hydroxyl-daunorubicin, which is where H actually came from. O is a drug called vincristine, used to be called oncovin. And P is prednisone. And R is rituximab. So now, we commonly give patients R-CHOP for newly diagnosed diffuse large B-cell lymphoma.

Lisa Garvin: So, but it sounds like that the paradigm is about to change a little bit as far as treatment regimens. So, even though CHOP is still standard treatment today, what are you looking into now that's so different?

Dr. Jason Westin: Well, we're trying to change the paradigm. It's very difficult to change. There was a trial about 20 years ago that looked at four different treatments including CHOP and they all worked about the same, you could say equally well or equally worse. So, depending on your perspective, but we've learned recently that large cell is probably a couple different diseases went together. So, investigators here at MD Anderson and at other cancer centers are designing approaches to try and treat these sub-types of large cell lymphoma as unique diseases. They have different biology, different response to targeted therapies. So, we want to try and manipulate that by using our knowledge of the biology to design clinical trials to evaluate these target therapies.

Lisa Garvin: So, I--You were saying that you've got what you're calling a smart trial or an adaptive clinical trial that kind of shifts with the, you know, what's happening in the preliminary results and so on--

Dr. Jason Westin:  That's right.

Lisa Garvin: --and so forth. Explain how that works.

Dr. Jason Westin: Well, we're looking at doing what's called an adaptive randomized clinical trial and we will probably do this in the relapse setting first. So, not patients who are receiving CHOP but patients who have previously received CHOP and unfortunately, the disease came back. But what an adaptive randomized clinical trial is, as you--if you use the term smart clinical trial. This approach was pioneered by investigators here at MD Anderson, mainly in the thoracic medical oncology group. They can now do something called the BATTLE trial. Our breast medical oncology group also did something called the I-SPY trial that helped pioneer this. So, this is an idea that generated from MD Anderson. But now, it's used at many different centers. And what happens here is we use the data we learned during the clinical trial to help assign patients to the correct treatment on the clinical trial. And what I mean by that is it's different from a normal clinical trial where the rules are set at the initial start of the trial. And we follow them regardless of what happens during the trial until the trial is complete. An adaptive trial allows us to set rules which have flexibility at the beginning of the trial to then smartly modify which patients are assigned to which treatment based upon previous patients on the same trial.

Lisa Garvin: 'Cause we're talking about several different biomarkers for the same disease, correct?

Dr. Jason Westin: That's correct. And unfortunately, to date, we've not had great biomarkers for lymphomas to help assign therapies. Many of our targeted therapies attack things that are important for normal B-cell or normal lymphoid cells and therefore, there's not a really good marker that says which patient should or should not receive that. And other clinical trials such as the BATTLE trial, lung cancer patients have mutations which can be easily assessed. And therefore, patients could be assigned to a treatment based upon a yes or no type biomarker like mutation, would there be a mutation there or not? For lymphomas, the biomarkers we've used as far are a little bit more murky. They're not quite as clear as yes or not type question. So, myself and others in my team are working to help define these biomarkers so that our clinical trial can be defined and set up in a smart way as possible to have the cleanest breakdown between different patients.

Lisa Garvin: So, how often do people with this type of lymphoma relapse?

Dr. Jason Westin:  By definition, they're all relapsed. So, if we're talking about relapse large cell lymphoma, they've already relapsed once. And if your disease is already relapsed once, the chance you will again is significantly higher than it was at initial diagnosis. So, the new diagnosis of diffuse large B-cell lymphoma is treated with R-CHOP. And response rates there are relatively high compared to the relapse setting where patients as I said have already relapsed once. They're more prone to relapse again. So, in the relapse setting, the response rates to aggressive chemotherapy are somewhere between 50 to 60 percent depending on different patient factors. So, it's essentially a coin flip in the relapse setting if someone is going to respond or not. And unfortunately, we have a difficult time of predicting who these responders are. And therefore, we have to treat everybody to see who responds and who doesn't. We can't--We have a difficult time predicting this before we do the therapies.

Lisa Garvin: And of course, aggressive chemotherapies come with lots of side effects?

Dr. Jason Westin: A lot of side effects. So, these are very toxic chemos we're talking about. So, patients who get these aggressive chemotherapies are generally fairly fit patients and we do these to try and set them up for something called an autologous stem cell transplant. So, when somebody relapses for large cell lymphoma. And the first question the doctor thinks of is, is this patient somebody that could tolerate this aggressive therapy and then tolerate an autologous stem cell transplant? So, when I'm discussing these aggressive therapies, there's a sub-population of relapse large cell that gets these, not everybody does, somebody who's frail, somebody who has a lot of other medical conditions may not qualify for a stem cell transplant. Therefore, they wouldn't get these aggressive chemos. But in the population of patients who do get these aggressive chemos, about half the patients will respond to it and can eventually go--on go--undergo autologous stem cell transplant. And what an autologous stem cell transplant is, is basically a way for doctors to give a very large dose of chemotherapy safely. So, generally, our chemotherapies will reduce blood counts for about two to three weeks, which is why the chemos are given every three weeks. So, the blood counts have recovered enough that the patient is ready to go for another round. The autologous stem cell transplant is designed in such that the dose of chemo would knock down blood counts, somewhere in the neighborhood of eight to 10 weeks, a significantly longer time which result in the higher risk of infection or other complications. But we found a smart way of getting around this which is to give a patient a transfusion of their own stem cells after the chemotherapy is done. These stem cells then home back to the bone marrow and can set up shop and the blood counts return in about two to three weeks. So, the autologous stem cell transplant is basically a way for us to deliver a very large dose of chemotherapy safely. That implies the patient has to be chemotherapy sensitive for us to do that with a chance of success. So, the relevant cut point for us to make a decision of go, no-go decision about stem cell transplant is does somebody respond that initial chemotherapy?

Lisa Garvin: Going into now this trial, this adaptive randomized trial. Is it open?

Dr. Jason Westin:  No, it's not open. It's something that we are in the planning stages and hope to open in the near future but this is something that takes a lot of work to define these biomarkers and we're currently doing that.

Lisa Garvin:  Have you had any--Have you done any work with animal models?

Dr. Jason Westin: We do a lot of work with cell lines, that's a very relevant model for us in terms of mirroring or modeling the patient outcomes. Animal models are a bit more challenging because you can treat a lot more cell lines with a lot more different combinations and you can with animal models. So, we're at a phase right now where we're pioneering these in cell lines which is basically cancer cells that grow in a culture and plastic, in a Petri dish in the lab. And then once we get to having a more clear combinations and biomarkers we want to use, we'll then tries these at animal models. But animal models have limitations and are not the same as treating patients.

Lisa Garvin: But are you seeing in your research on these cell lines as--There must be something encouraging there that's leading to the formation of the trial?

Dr. Jason Westin:  Absolutely. We found in the cell lines that the thing that's encouraging there is that number one, they have nice response to the target therapies that we see in the clinic. These are good models for what we see in the clinic. Number two. We can treat cell lines with an infinite number of different therapies. So, in patients, you can't do that. Taking a patient to a clinical trial, it takes a long time for us to set up a clinical trial and open it and to enroll enough patients to get conclusions. And the cell lines, we can do this quickly and at a much different scale so we can learn a lot of things from treating cell lines. And then quickly, take that to the patients once we found something that's very promising.

Lisa Garvin:  Because this is an adaptive trial, how will you go about figuring out which patient should get a particular targeted therapy?

Dr. Jason Westin:  Well, we need to have good biomarkers to begin to assign patients with a particular abnormality in their cancer to receive a therapy that they're likely to respond to. And many of these assignments to particular biomarkers with a particular therapy will come from preceding trials where patients have been treated with a single agent targeted therapy and found out which patient is responding and which patients did not. The way clinical trials are set up right now is that patients with relapse disease are generally treated with either chemotherapy or under clinical trial. And we've talked in previous podcast about how it's a shame that very few patients go on clinical trial. So, we have a hard time developing new drugs and learning where they work. But I think a big problem with the way we evaluate these drugs is that we are evaluating these drugs as single agents and patients with very heavily pre-treated disease where the chance of response is a lot lower. And one of the things we've learned is that attacking the most common abnormality or attacking the most easily identifiable abnormality or biomarker may not alone be sufficient to induce responses in patients with lymphomas or cancers in general. We've learned a lot about complex networks and about how complex systems act over the past decade or so. And part of this has actually come from the study of the internet of all things, but basically with the advent of powerful computers, we've been able to do a lot more work of looking at complex systems. Cancer is a complex system. That's kind of an obvious statement but cancer cells by their nature have a lot of things wrong with them that normal cells do not. Generally, cancer cells do not have one thing wrong with them. There's multiple abnormalities. So, there is a one particular type of cancer that has a biomarker that we've used that to assign therapy and it's been a blockbuster. It's been a great result, and that's CML or chronic myelogenous leukemia. So, the story of imatnib or Glivec has been a homerun for oncology. But I think that using that drug and the biomarker that we found to correlate response with that drug is the wrong approach moving forward for cancer. I don't think we're going to see a lot of single agent chemotherapy or targeted therapies have that type of response. CML is relatively unique and that really has one significant driving abnormality and that's it. Other cancers generally have at least two, if not, up to eight of these abnormalities which could be driving things, perhaps even more than eight. That's a conservative estimate. And taking out the most relevant part of the network may not alone be sufficient. So, cancer networks have very strong similarities to other networks, other things called scale-free networks. And an example I like to use of this is looking at the airline distribution network. So, how do airline flights--How are they allocated? And you look at the map in the back of the magazine when you're on the airplane, you see that airlines use something called hubs. So, here in Houston, we have IAH or Bush. In Chicago, there's O'Hare. In Atlanta, there is Hartsfield-Jackson Airport that basically these are the way the airlines are the able to send patients--or not patient, well, to MD Anderson, yes. Patients travel by flight, but how they travel their passengers to go from one part of the country to another from small airports through a hub and then to another airport. They use these as connecting points. The reason why this is relevant for us is that there are certain parts of their airline network which have hugely out of proportion, number of connections. So, these hubs I mentioned have hundreds of different flights coming in and out. The vast majority have one or two flights. In cancer cells, the vast majority of the signaling network is not critically important but there are few parts of it which are dramatically out of proportion to the other in terms of their importance, number of connections.

Lisa Garvin: Now, you just recently published an article on this subject. Let's talk about that.

Dr. Jason Westin: That's right, thanks. I did recently publish an article on that. It's called "Busting Robustness." Kind of play on words-type title, but my thought on this article was that cancer can be thought of in the same context as airlines. So, if an airline is dependent on only one hub such as CML is dependent on one significant aberration. If you close that hub, if you shut down the airport, the airline can't compensate. So, if you take out Bcr-Abl as you do with imatnib or Glivec, CML is out of business. It's basically functionally cured. Other airlines, other airlines may have more complex flight networks. So, if you look at the example of BRAF mutant melanoma. BRAF is a very common abnormality found in patents with melanoma. And if you give a target inhibitor of BRAF, you'll see responses in about 50 percent of patients. None of them are cured. The disease comes back in every one of them. So, this--the analogy here would be an airline that depends upon the hub airport but has other ways of getting around. So, if you close down the hub, they can reroute and may actually be stronger for having more hubs from their rerouting strategy. The third example that I used in this paper is BRAF mutant colon cancer, same general mutation of BRAF is seen in melanoma. But the response rates to BRAF inhibition are about 5 percent here. The reason why it's so much lower is that the network is a lot more complex and there's other abnormalities such as eGFR which are critical. And if you inhibit both of those, you may see significantly higher response rates. So, the analogy there would be an airline that has several hubs that uses. So, if you close one, it's not really inconvenienced versus CML is out of business and BRAF mutant colon cancer--I'm sorry, BRAF mutant melanoma would be shut down for a short period of time and then be able to reroute. So, I think that we need to evaluate targeted therapies in this context of looking at other complex systems such as scale-free networks and the analogy of airlines just sort of an easy one to conceptualize. But when we're looking at setting up these clinical trials for lymphoma, we're keeping this in mind that this single agent response rate of targeted therapies may not tell the whole story that this may be a significant part of the story but just because a drug doesn't have a blockbuster single agent response rate doesn't mean it couldn't be a critical part of a therapy that we derive in the future.

Lisa Garvin: But so, your adaptive trial is going to address this issue?

Dr. Jason Westin:  Our adaptive trial will address this issue and that we will use targeted therapies in patients we think that are likely to respond based upon biomarkers. And then begin to enrich for responses by increasing the allocation of patients to one arm of the trial or another based upon responses. So, yes, we will plan to use this strategy to hopefully improve outcomes for our patients.

Lisa Garvin:  I don't want to put you on the spot but in closing, I hate to make people prognosticate, but it seems like there's a little bit of excitement here. Do you feel like we're moving towards the next generation of standard treatment?

Dr. Jason Westin: I think that we need to move towards the next generation. So, as I said before, R-CHOP has been around for 30 plus years. These salvage chemotherapies we use now with autologous transplant have been around for 20 years. Think of all the advances in modern society and in cancer therapies and in our knowledge of the biology that have happened on the past 30 or 20 years. It's outstanding that all things have had happened and the fact that we're still using the same therapy is extremely disappointing. So, yes, there is excitement that to move beyond what we're currently doing and to make these things happen and places like MD Anderson and other big cancer centers are working as quickly as we can to make these things happen.

Lisa Garvin: Well, hopefully, we can talk to you in a year and you'll have good news for us.

Dr. Jason Westin:  Hopefully, so.

Lisa Garvin: Thank you, Dr. Westin. If you have questions about anything you've heard today on Cancer Newsline, contact to ask MD Anderson at 1-877-MDA-6789 or online at [Background Music] Thank you for listening to this episode of Cancer Newsline. Tune in for the next podcast in our series.