Tisagenlecleucel Shows Response With Myc Expression and Tumor-Infiltrating T Cells in the Phase 2 JULIET Trial for R/R DLBCL

Videos

Rapid Readouts: Phase 2 JULIET Trial

 What I want to hear about is patients with central nervous system disease when we were developing C19-directed CAR T cells. What’s your experience with CNS?

: It’s extranodal relapse of disease that happens to be in the CNS. A lot of groups are starting to look at this. There were some data presented at ASH from The University of Texas MD Anderson Cancer Center of using axi-cel for patients who had a history of or active CNS disease. I know there are trials that are currently in development for both secondary and primary. But overall, we assume the risk factors for CRS and neurotoxicity are more derived not so much from the T cells, but from the innate immune activation and inflammatory cascade that comes thereafter. Theoretically, the patients with secondary and primary CNS should be at lower risk, given lower disease burden and less immune inflammatory priming in the systemic circulation. What we’ve seen, at least the first 8 patients whose results we published, was about 50% overall response rate, 3 of whom are still alive and in CR today. We’ve got about 30 in total we treated, and we’ll be putting that series together. They’ve actually started treating primary CNS lymphoma on an investigator-initiated trial, and we’ve got about 4 patients accrued. We’ll have those data available for this year. Overall, the biggest issue with systemic lymphoma…is getting the patient safely to the therapy, meaning that they’re not progressing too rapidly. Their disease should be under control. Going back to, I think, prior points that Caron, Loretta, and everyone else made was that steroids in these patients, if you need them, you give them, and it doesn’t necessarily have that deleterious of an effect.

 Yes, I did read your paper, Matt. And you say so nonchalantly, it’s another extranodal site, the brain. But those patients of yours, these were not extra-axial CNS lesions. These people, they had parenchymal extranodal brain lesions. I would have expected more toxicity than you observe. But like I said, I’ve certainly been wrong before.

: Anecdotally, the toxicity you can see is there, I’m not going to say pseudo-progression, but a little bit of inflammation in the critical part of the brain can cause a lot of problems. And I think that’s where I don’t want people to run off and start giving everybody CAR T cells for secondary CNS. Disease control going into the actual therapy is the most critical component, because if you lose control midtherapy, it doesn’t matter if the patient responds in 4 weeks; they’re not going to be alive in 2.

That disease control also underlies the success of CAR T-cell outcomes in systemic lymphoma therapies, as well. And I think David’s going to have a blood test for disease control soon. The CT DNA.

: Matt, the products that you reported on was tisa[-cel], right?

 We may have different results using different agents. I know, Caron, you’ve been working on some of that as well. I want to throw out 1 other unmet need that I think we all want to push for as well, and that is patients with HIV. HIV and lymphoma have been a problem. Because of CD4 attacking they are more at risk for the lymphoma in the first place, they’re not necessarily going to be good autologous CAR T-cell donors. I mean, we can talk about it. We can argue about it. This is a place where allogeneic CAR T cell holds some promise. These patients are eligible for that study that was just presented. I encourage you to consider that when referring your patients with HIV to the Allogene sites. I’m bringing 1 on right now.

It’s certainly not something that is excluded by the label or included in the label. But Kite has a standard operating procedure and will manufacture autologous CAR T cells for HIV-positive patients. And we’re doing that. But Novartis doesn’t, and so you can’t do tisagenlecleucel.

 Steve, I’ve also made CAR T-cell products for patients with HIV. Unfortunately, the 2 patients didn’t do well. One never got to the therapy with massive disease progression.

: We’ve treated and publishedresults on 2, both of which are in CR.

 And because of your publication, Matt, we’re now doing it, too. Again, maybe enough of this gets done. Somebody’s going to have to put a series together.

: This is going to be a big component, just to make a plug for regulatory. A lot of the reasons why patients aren’t being treated is based on how the products were approved and the regulatory hoops that people had to jump through for the manufacturing process. For populations such as this, I would ask the FDA and other regulatory bodies to help expedite getting access to patients, because biologically, it makes sense that we’re not manufacturing for them.

Ongoing Issues in CAR T-Cell Therapy and Lymphoma

There’s another possibility here that we haven’t discussed with our viewers. That is, that using a CAR [chimeric antigen receptor], which is foreign, is also potentially a good thing, because while we’ve often thought we want to have the CARs persist and provide antitumor responses, especially as ALL (acute lymphocytic leukemia) patients, the investigators have frequently advocated. On the other hand, if you hit them hard, and you wipe out the lymphoma, perhaps the best thing that happens is that the recipient’s immune system does reconstitute, eliminates that CAR T cell, and allows normal B-cell recovery, T-cell recovery, and CD4 cells. I’m going to point again to Michael Jain and Fred Locke’s work published in 

now, where they show the CD4 counts being less than 200 in half the patients beyond a year. Many people wonder if that’s the thymus in the older adults, the high-dose fludarabine, or perhaps B-cell aplasia contributing to a damaged germinal center. We don’t know. It’s an exciting experiment.

It is. I’d love to do the experiment. I had the chance to read the

 paper, as well. And that’s unique to that product. We have different data with tisa-cel [tisagenlecleucel]. For some reason, the immune system seems to regenerate much more readily. You see earlier B-cell recovery, immunoglobulin recovery, and cellular recovery. I don’t know a hypothesis that would make me think it was the CD28 costimulation, although I guess some kind of cytokine-mediated thing. God knows, a lot of hand waving in immunology.

 The Allogene [Therapeutics] experience is an experience of 2 things. It’s the experience of lymphodepletion. We went back to reconsidering lymphodepletion and all aspects of it. Increasing the dosing of the Cytoxan, increasing the dosing of the monoclonal antibody that’s against CD52, staggering the lymphodepletion, putting the lymphodepletion at the same times. Studying the absolute lymphocyte count has turned into the most interesting experiment that I’ve done in a long time, which is not very elegant, Steve, but incredibly important, as they’ve shown in some of their data slides that they presented at the meeting today. The expansion of the CAR and the clinical response rates correlated with the extent of the lymphodepletion that was achieved.

Yes, that makes sense, because there has to be room for expansion. I don’t necessarily think chemotherapy is the smartest way to do it. I don’t know at what stage of lymphocyte development CD52 becomes expressed. I’m sure somebody does. And I’ll wager that there are precursors that are present if you use CD52 alone to deplete cells, that you’ll have faster lymphocyte recovery than if you use a chemotherapeutic approach that is more likely to hit primitive, or more primitive, lymphoid precursors. But I don’t know the answer.

 That’s part of the experiment, you can use this targeted therapy against the cells of the recipient’s lymphocytes that is leaving the stealth CAR-T cells immune. And likewise, Steve, you can go back at day 14 or day 30 and pump the cycle again and give another dose of anti-CD52, knock down the recipient’s T cells again, and allow expansion of the CAR, if you think you need it, without redosing and without having allosensitization, which is potentially exciting.

 And we’re planning, if we can ever get out of COVID-19 mode and now riots in the street, to actually begin to work with Allogene at Penn, as well. In our discussion, there may be differences in terms of the importance of persistence, because you don’t have the persistence with Allogene that you would with an allogeneic product like this, that you have with an autologous product. But there may be differences between diseases in terms of the importance of your system. Anybody want to comment on that? It seems as though resistance is less important in lymphoma than in ALL, based on what we’ve seen in the clinical studies. And that may be right.

 …In lymphoma, these are mature cells. You just need to wipe them out 1 time. You don’t necessarily need CARs to persist if they’re effective at eradicating the tumor cell, whereas in acute leukemia, that might be a little bit different.

The cells are more stemlike, they’re repopulating, or just those young people can get up again and again and again. You keep knocking them down, and they keep getting up.

Developing CAR T-Cell Trials

In this analysis of the sponsored Allogene [Therapeutics] work, which has an interesting history, these were the founders of Kite who then moved on to a new project. Today, we’re seeing the first published presented observation in lymphoma that we might be able to take out of the can and have super donor cells, allogeneic and prepared to become stealthlike. The stealthlike notion of this is deep lymphodepletion of the recipient in the setting of making the CAR [chimeric antigen receptor] itself resistant to the lymphodepletion by removing CD52 with technologies. Avoiding the cell against the recipient risk of graft vs host disease by removing the α chain of the TCR [T-cell receptor]. Both of these deficiencies are only limited by the current level of technology, which is not better than about 85%.

What is interesting is that the 2 genetic modifications are occurring in the same cells. The cell that gets TCR knocked out is the same cell that has the CD52 removed, and therefore now is resistant to an antibody against CD52 and lymphodepletion.

 If there are some cells that didn’t successfully have 52 knocked out, they get knocked out by the lymphodepleting alemtuzumab-like antibody that’s used in part of the protocol.

 You’re right, you get a mixture. When you make these cells, you get a mixture of cells that have knockout of CD52, cells that have knockout of TCR-α, and cells that have knockout of both. Most of them, of these 3 variants, are going to have a CAR receptor. This is a multiplex genetic engineering event, which the 2 of us with gray hair get excited about.

: I want to know what the younger people think, particularly Loretta. 

 The most striking thing is when we identified a patient and treated within 2 weeks, which I think really gets to the point of you’ve got 1 shot at this. Time matters. Get it in quickly. One of the potential caveats to this specific agent, though, is they had rituximab built in as a safety, which was part of their earlier development. And you had to have patients who didn’t have recent rituximab, or at least not much circulating rituximab. This did limit some of the application, but that’s going to go away in the next iteration.

: Yes, because that doesn’t happen in Philadelphia, if you have a lymphoma patient that’s failing that has no Rituxan

 in their system. I don’t know about other places.

 What do people feel about lymphodepletion with the CD52 antibody? David mentioned it’s deep lymphodepletion.

: With the toxicity profiles, you would predict a lot more viral reactivation. There were some CMV [cytomegalovirus] and reactivation reported. With the advent of letermovir, that could mitigate some of the issues that we’re seeing. But it’s like bendamustine. It does have implications for long-term, deep lymphodepletion, although it took a couple of months before you saw ALC [absolute lymphocyte count] recovery. That would be sufficient for harvesting for an autologous product in that case.

 Full disclosure: I’m an investigator, right? I’ve seen some of this. I will only speak to the 23 patients who are presented at American Society of Clinical Oncology today. There is no neurotoxicity so far, zero, zip, nada, nothing. There’s no GVHD, none. There’s massive expansion of these CAR T

 cells in the patient with, again, greater than 60% overall response rates, greater than 50% complete response. And they’re showing some durable responses in the total cohort that are CR [complete response] greater than 37% out of the can with low doses. If you go to the higher dose, infusion is 360 million per recipient. The 3 dose cohorts were 40, 120, and 360. The 360, those numbers are greater than 50% CR rates, which is something. Matt, what you’re saying is that if you knock out other cells besides T cells using a CD52, you may be removing a lot of the neurotoxicity triggers, and the monocyte and myeloid cells may be getting profound depletion in addition to the T cells. 

: Would you expect that you’re not going to see much IL6 at all, because you’re killing all your macrophages with the anti-CD52. That’s a good agent to treat patients with refractory CRS [cytokine release syndrome] with.

: Everybody said you have to pay a price for that.

: Lymphodepletion, and that’s infection and potential viral reactivation. And you’re mentioning strategies to get around it.

Allogenic CAR T-Cell Therapy in Lymphoma

: The next issue is an in-depth analysis of the future directions, and any data that are driving those future directions, anything that you guys saw that was exciting between the last ASH [American Society of Hematology annual meeting] and the current ASCO [American Society of Clinical Oncology] meeting?

What I thought was exciting was that they didn’t use chemotherapy. I’ll bet you don’t see the same kind of cytopenias when you don’t give Cy/Flu [cyclophosphamide, fludarabine]. I’m speaking as somebody who also has seen a lot of CLL [chronic lymphocytic leukemia] in my life, and I’m sure Loretta has seen a lot of CLL where she lives. And [cyclophosphamide, fludarabine] is tough on patients, particularly if they’ve had other prior treatments, including transplants. And I’m not at all surprised that there are prolonged cytopenias. Yes, there’s some impact of having an inflammatory state, particularly if the disease involved the marrow. That’s going to contribute. But believe me, [cyclophosphamide, fludarabine] is not vitamin B12.

: The other thing I thought was interesting about that abstract is they had very low rates of grade 3 or higher CRS [cytokine release syndrome] or neurotoxicity, strikingly low.

: In some data I’ve seen using the tandem bispecific CAR [chimeric antigen receptor] is similar. I don’t know what to make of that, Loretta, history being important. But that struck me as well. 

CAR T-Cell Therapy’s Future for Lymphoma

: Let’s discuss adverse events. Matt, you said everybody knows about steroids and toci [tocilizumab]. Does anyone in this group feel as though steroids are absolutely contraindicated in CAR [chimeric antigen receptor] T-cell therapy, which used to be the mantra that the immunologists told us? I remember the early days, where, if we put steroids onboard, it wasn’t going to work. And again, it wasn’t data-driven, but it was theoretical. What’s been the cumulative experience here with steroids to treat toxicities?

: Loretta, I think [the University of Texas] MD Anderson [Cancer Center] had a retrospective review of steroids that was presented at ASCO [the American Society of Clinical Oncology annual meeting], was it?

: Paolo Strati, [MD,] looked at it. Across the board, there’s been early introduction of steroids, and it does not seem to impact efficacy, at least in the short term. What Paolo looked at, again, retrospective single-center; he demonstrated that if you had high doses and a particularly prolonged course of steroids, that may have a negative impact on outcome. So again, everyone should be looking at strategies to mitigate toxicity, particularly if you can prevent them from going from grade 2 to grade 3 or higher. But you do need to be mindful of having these patients on high-dose steroids, particularly for prolonged periods.

: This reminded me of how we frequently discuss graft-versus-host disease [GVHD]. It’s 50 years into allotransplant, and we still argue about what is the definition of steroid-refractory acute GVHD, how do you best manage, what is the best way to taper? If anybody thinks there’s going to be a simple answer to the management of steroid CAR T, I want to take that out of your heads right now.

:Unless you prevent the toxicity in the first place.

: I do think it’s a little bit easier, because there’s a starting point. And I do think that allogeneic products might give us the ability to look at more uniform products being put into different people and how they’re managed. But this will be the problem that we’re going to be trying to argue about for a long time, I suspect.

: I’ll be retired by then, by the time it’s solved. But I’ll tell you that I think, Matt, it really is dose and duration. It’s not like steroids should be anathema if you’re getting CAR T cells. A short course of therapeutic for, I should say pharmacologic doses of steroids, I’ve never seen it affect outcome, and we’ve had to do it for neurotoxicity and even CRS [cytokine release syndrome]…. But yes, the same way it’s not very effective in killing the T cells that cause GVHD, it doesn’t really kill the CAR T cells that well either.

: The counterarguments are that correlation of steroids and decreased long-term efficacy could be that you’re having to treat patients who have higher-risk disease that’s more inflammatory in the first place.

: Yes, and there are data showing the sicker patients have worse outcomes. We used to think the sicker you got, the better the result was going to be. In the early days of this, I have to tell you, I had patients who didn’t get sick, who were going into remission. I was treating these patients as outpatients, and they’d come in and say, “Doctor, it didn’t work.” And I’d say, “How do you know it didn’t work. We didn’t do a scan.” They’d say, “Maybe I had some backache or a little—but I didn’t get sick.” And it’s like, no, you don’t have to get sick for this to work. Because ALL [acute lymphocytic leukemia] was the first disease treated, there was a folklore that developed. And honestly, I had patients who were disappointed they didn’t get sick. But in reality, the data suggest that those who get sick in the end may have worse outcome.

: There’s a toxicity, though, that none of us gave enough credibility to up front. And I want to hear your opinion, Steve, with the experience of [tocilizumab]. But in the axi-cel [axicabtagene ciloleucel], the cytopenias that are developing in patients, beginning usually 3 weeks to 6 weeks initially, with neutropenia being a major problem with patients around day 28, 50% of whom are getting G-CFS [granulocyte colony-stimulating factor] support. And even looking out at a year now, in a study that I saw from Michael Jain, [MD, PhD,] in the Moffitt [Cancer Center] group, and we’re doing similar studies at Stanford [University], showing that at least 10% of the patients beyond a year of [axicabtagene ciloleucel] have intermittent neutropenia. So cytopenias remain a bit of a problem, which is especially a little scary to the community physicians because every time the patient has a fever, do they need to bring them in? This remains a question.

: Yes. We keep focusing on things like CRS and neurotoxicity. In the real world, cytopenias are a problem. How long do you need to get platelet transfusions for and check blood counts constantly, and I guess the propensity to develop infections early on after this? And I think there are differences. There are probably going to turn out to be some differences between the products, and also between the type of lymphodepletion that you use. If you look at our prolonged cytopenias with the low-dose Cy/Flu [cyclophosphamide with fludarabine] that we use with tisagenlecleucel or bendamustine, it’s less than with the 900 [mg] cumulative for Cytoxan and 90 [mg] cumulative for fludarabine. That’s why I was so excited when I saw the abstract from ASCO 2020 where they were using pembrolizumab, not chemotherapy, prior to their T cells, and ended up with response rates that were like a complete remission [CR]. Yes, follow-up sure, but a 55% CR rate. Come on, it’s not bad.

CAR T Cells and Lymphoma Adverse Events

: This is a great question that was posed. I don’t know the answer to it, but I’d certainly love to hear the opinions on it. We know there are 2 efficacious agents out there. And we know that there’s 1 coming. Let’s assume it’s here. How are you going to choose? Matt?

: I’m of the mindset that depending on the logistics and the circumstances, it’s down to certain strategies. The CIBMTR [Center for International Blood and Marrow Transplant Research] registry data that were presented at ASH [the American Society of Hematology annual meeting] this past year did touch on this little bit, where you tended to have younger patients who didn’t have as many comorbidities getting access. All this is center-specific, because I know some people look at it as getting the sicker patients to an infusion as quickly as possible. [Tisagenlecleucel] patients tended to be a little bit sicker. They tended to have more morbidities, more cardiac disease, more involvement with secondary CNS [central nervous system] lymphoma. That’s one component. We’ve got to analyze the liso-cel [lisocabtagene maraleucel] data once they are available and reviewable. That’s where the registries become important to be able to compare. But ultimately, it comes down to optimizing the patient.

For us, at least how I approach it is, for the double-hits primary refractories, regardless of age, are pretty much going to something like axi-cel [axicabtagene ciloleucel], because you need fast, reliable manufacturing that’s going to get back to you quickly. And also [tisagenlecleucel], for those patients who tend to be a bit older, who are a little bit frailer, for whom neurotoxicity means 4 weeks in a rehab facility tied to a bed, I’m not sure that’s in their best interest.

: As we were discussing how we’re getting better, the 2 drugs that we’re talking about for managing the toxicities are toci [tocilizumab] and steroids. And we’ve been talking about these same drugs for a very long time. As we start understanding mechanisms of toxicity and employing other agents, it’s going to get better and potentially allow sicker patients, or patients we previously wouldn’t treat, or change the paradigm for how we’re doing product selection. Efficacy is ultimately a product of overall response rates, but also the patients surviving the treatment that you’re giving.

: There was a quote that’s become my most famous line, that “we treat them all because that’s what the patient needs.” And it’s not right to withhold a therapy that’s so effective. But price and opportunity for especially those patients on Medicare has been a big driver, and no matter what anybody says, has been impacting how patients get their therapy. The recent announcement from our administration that they’re seeking the…really big DRG [diagnosis-related group] is great news for the patients to remove price from how we’re managing their care. And I’d like to see therapies in the outpatient setting as well. But some of the therapies are just requiring inpatient toxicity management.

: There are 2 ways that a patient can be sicker. They can be sicker from their lymphoma, and they can be sicker from other medical problems. And so I think it depends how they’re sicker that will pick the product that they’re going to use at this point.

If they’re sick enough, either way, they end up in the hospital. But I’ll tell you that external factors other than money, believe it or not, sometimes influence choices. Early in the COVID-19 [coronavirus disease 2019] epidemic, when we were expecting to have inadequate numbers of beds, we started using tisagenlecleucel almost exclusively because we didn’t want to commit patients to even a 10-day hospital stay or a 7-day hospital stay. But unfortunately, as it materialized, the peak of COVID-19 in Philadelphia was not so great, and we ended up having a surplus of beds and space. But not every city had the same experience. I think every now and then, you don’t have a choice.

: Also with [tisagenlecleucel], sometimes you can use a cryopreserved product. And in some circumstances, having the ability to cryopreserve opens other doors in terms of bridging therapies or logistical circumstances that the patients are dealing with. It’s not just the efficacy reporting from the clinical trials. It’s how you get them there.

: It’s interesting, and correct me if I’m wrong, but in Europe, [axicabtagene ciloleucel] can be made from cryopreserved cells.

: How cohort 4 of the early steroid use was done is every patient was cryopreserved and then manufactured…I believe.

 Well, I’m hoping that that crosses back over the Atlantic to us, the ability to cryopreserve, regardless of product. That being said, Matt, what I think is that the reason you need a higher lymphocyte count for [tisagenlecleucel] is because you do cryopreserve and thaw, and you lose cells that way. And so there is some rationale for shipping a fresh product.

Low counts. And that goes the same with [lisocabtagene maraleucel] as well, where we ship fresh products. They can almost produce lymphocytes out of thin air, in my experience. It’s because you’re sending them a fresh product. It makes a huge difference.

: But even for products that, for failed manufacturing runs from a fresh product, there’s usually a frozen backup that can also be used. They do have a good success rate with that.

 In patients we were treating with [axicabtagene ciloleucel], they went to the backup product and did it. In fact, we were impressed by that enough, that anybody that we get more than enough cells for at Penn [Abramson Cancer Center], we freeze cells and keep them at our institution as backup. I don’t know how long that’ll last, but at least it’s what we’ve been doing.

LBCL Therapy: The Right Therapy

: The next part of the discussion is based on our experiences and what is being captured in the literature. These are points that are practical for those physicians who are in practice and treating patients with lymphoma but are not at a center that is necessarily doing CAR [chimeric antigen receptor] T-cell therapy. And that relates to identifying who’s the right person for CAR T; well obviously, we can only go by the label at these prices. But timing is a more important issue. Identifying and timing are related because you can identify when you’re too late. Let’s talk about timing and identification of the patients you would refer to CAR T-cell therapy.

: You started with the obvious one. You can be too late, to not get the referral going early when a patient is showing a refractory response to induction chemotherapy. Even though you must give that second round of salvage by label, you know that’s not likely to work, and you better be calling your referral center. And secondly, the patient who has no evidence of response after 1 round of salvage, it’s not going to get better after the second. You’re still going to give that therapy, but please call. Because low disease burden, good performance status, low LDH [lactate dehydrogenase], and low inflammatory state are 4 of the 5 now repeatedly identified poor prognostic signs. They’re all related to waiting too long.

: Yes, and that extra cycle of chemotherapy further depletes the lymphocytes in patients for whom you may need to manufacture cells. You’re not doing a patient a favor by waiting.

: Not only that, but some of the newer therapies that are coming out that are competing with the same indication have impacts in terms of the types of quality of cells you’re getting, your likelihood of manufacturing these cells. And so dragging your feet and being pulled to give a therapy that may contain something like bendamustine will impact your ability to collect, manufacture, and get a patient to infusion quickly.

: Matt, you touched on this earlier. One of the biggest issues we have is we know identifying patients early and introducing this is great. But we partner with community oncologists oftentimes. And these patients get to us, and they’re in a state that we’re nervous about toxicity and efficacy. How do you optimize these patients? We now have things like bendamustine, polatuzumab, rituximab, though I agree, I have significant concern about giving them bendamustine before collecting those cells. We may have something like tafasitamab, which is a naked CD19 antibody, in combination with lenalidomide. There is a huge unmet need for how we bridge these patients, because we know those who need bridging are going to do worse, based on the fact they need to be optimized. That’s where I struggle. What are your thoughts?

: One thing that’s very important is the bendamustine/polatuzumab issue, because the data had been presented. There’s a nice response rate in people that are refractory to the common salvage regimens. And bendamustine is a wonderful lymphodepleting agent. We use it at Penn [Abramson Cancer Center] more often than CyFlow [flow cytometer]. Results are not worse than other people’s, and I have lymphocyte data to substantiate what it does to the lymphocyte count. It makes no sense to be giving that and then sending them to a place where you’re expecting them to pherese and make CAR T cells. That’s clearly, before you give bendamustine to somebody….

: Most people on this talk and most people doing CAR T-cell therapy are happy to get a phone call or an email or any kind of communication to give freebie advice on how to strategize. Because one thing we can touch on here is that there’s a strategy of getting, at least in the autologous T-cell setting, a strategy of getting a patient to an infusion safely while still simultaneously optimizing them for success.

: Recognizing you still have other hurdles to get over—scheduling, insurance authorization. The 5 of us probably take referrals for the next day, because we know that the timeline is short for that patient, and the success is dependent on getting it going fast.

: And now with telemedicine, you can even see the patients without going to the office. There’s no excuse not to see them the next day. You can even see them at night before you go to bed, David. Yes, there is no excuse to not send the patients. We can see them right away.

R/R NHL: Candidate Identification and Referrals

 When are we going to have it for follicular, bad follicular lymphoma and bad marginal zone lymphoma, because as we said earlier, these do exist?

: Yes. I’m optimistic that 2021 may be a year for follicular lymphoma at least, so that we presented the results of ZUMA-5 (NCT03105336) at [American Society of Clinical Oncology] on Friday. This included up to 125 patients with follicular lymphoma, 80 of whom were evaluable for efficacy, because I had at least 9-month follow-up. Then we added additional 16 patients out of a planned 35 patients with marginal zone lymphoma, all third-line and beyond. They all had to have a CD20 monoclonal antibody, as well as an alkylating agent. These patients had median number of lines of prior therapy, 3 or more, 70% of patients have had 3 or more prior lines of therapy.More than half the patients had a high FLIPI (Follicular Lymphoma International Prognostic Index). They had high tumor burden by GELF (Groupe d’Etude des Lymphomes Folliculaires) criteria. It’s a high-risk population. The overall response rate for follicular lymphoma was 95%, and the CR [complete remission] rate was 81%. And if you looked to the median follow-up for that cohort at 15.8 months, 80% of patients who achieved a CR maintained their CR. Again, [as] with mantle cell lymphoma, we need much longer follow-up to understand what the natural history of this disease will be after CAR T cells, but definitely longer than other available options for patients in this line of therapy.

CAR T-Cell Therapy for Follicular Lymphoma

:It’s difficult to think of 11 lines of therapy for lymphoma. But I agree with what you said. Thinking of myeloma always makes me think of the lymphoma counterpart of myeloma, because a lot of the cases have the t(11;14) translocation, and there are also other molecular characteristics that were discussed at the current meeting.Proteasome inhibition works in mantle cell and works in myeloma. Lenalidomide works in both. There are clearly biologic and genetic pathologies.

: Yes, but so are follicular lymphoma cells, which proteasome inhibition doesn’t do anything like large cell B cells and follicular B cells. There’s something about those mantle cells. Mark my words, maybe they’re memory B cells that are closer to the myelomas. I don’t know. I’m not a biologist. But the things that work in myeloma tend to work in mantle cell. Let’s move on to mantle cell lymphoma, because when Dr Michael Wang published in the 

what I think is a landmark article, if anybody would like to recap it, using a version of axi-cel [axicabtagene ciloleucel] where the T cells are separated from any circulating B cells to minimize CAR T-cell or malignant B-cell contamination of CAR T-cell product. It’s called KTE-X19, the ZUMA-2 trial (NCT02601313).

: I treated a number of patients here, and I have some clinical experience. The patients failed BTK inhibitors in mantle cell lymphoma because there was not effective therapy. We’ve been looking at a number of different targeted agents. But this really does fit a need, particularly for those [who] are ibrutinib-resistant. But Caron, what was your impression of the study?

: I was with a group of patients [who] had mentioned before, their survival was on the order of weeks to months. CR [complete remission] rates to subsequent lines of therapy were single-digits. And with KTE-X19 and multiply relapsed mantle cell lymphoma that had failed chemoimmunotherapy as well as a BTK inhibitor, you’re seeing CR rates of 67% with, at 12 months, you’re seeing somewhere around 60% of people maintaining their response, which is phenomenal, that’s really moving the bar forward. And we talked a lot about how long the follow-up is for large-cell lymphoma to see if these are durable responses. Mantle cell lymphoma is a different disease. It is a disease marked by inevitable relapses. We need longer follow-up of the series to know whether this is going to be a definitive therapy for a subset of patients or whether it’s a bridge to something like an allogeneic stem cell transplant. This moves the needle forward for a group of patients with a huge unmet medical need.

: Caron, about one-third of those patients are blastoid variants, correct?

: Yes. There are a number of patients with blastoid variant, a number of patients with p53 aberrations, complex karyotype, it’s a very high-risk group of patients, for sure.

t’s worth pointing out, though, that there were significant challenges to managing the patients. They were able to get quite sick with very large tumor burdens, aggressive forms of CRS and neurological toxicities. The management on that trial being performed at centers that had already been experienced through the earlier ZUMA studies, ZUMA-1 (NCT02348216), those were still patients [whose cases were challenging}. And we all had to share our experiences in order to help some of those patients get through some of those toxicities.

: When the bone marrow is involved and sinusoidal organs like spleen are involved, there’s much more toxicity with CAR T cells in any of the B-cell lymphomas. They’re much more like ALL [acute lymphoblastic leukemia]. And that’s mantle cell.

ZUMA-2 Trial Overview: Mantle Cell Lymphoma Impact 

Loretta, please share your thoughts on the 

 [Journal of Clinical Oncology] publication 2 weeks ago. What are the predictors of how patients did on simple things such as LDH.

: When you have this available in your toolkit, you’re going to start to apply it in [populations that are] sicker, frailer….And so, 43% of patients would not have met the eligibility criteria in ZUMA-1 [NCT02348216]. Despite that, the safety looks good. When we try to predict who did best, meaning who had the best outcomes, LDH at lymphocyte depletion and ECOG performance status were very robust clinical prognostic factors, which tease out those patients [who] have highly proliferative tumors that you don’t have the luxury of time or a cellular therapy that you’re manufacturing from an autoproduct.

: David has a blood test for performance status, don’t you, David? Do you want to tell us about that?

: We don’t have a blood test for performance status. One has often said that I never get much below 2. But we do have a blood test looking at the tumor burden, which is another part of this discussion of how much disease somebody has. There was work produced by Dr Matthew Frank in our group and Dr Fredrick Locke and at ASH [American Society of Hematology], where we looked at circulating tumor DNA using the soluble DNA in the plasma blood using the clonoSEQ (assay) from the company Adaptive Biotechnologies. In that 80-patient study, it was highly informative, both looking at the amount of tumor going in, much like LDH pre-LD-amount of tumor, or some of the diameters of the tumor, or I hear metabolic tumor volume at Dr Locke’s group now. There [are] a lot of ways to read out how much patients [with tumors] have going in. And then, of course, it tells a lot about modality of when patients are losing the benefit of the CAR T, which is the most important thing about this frequent measurement, because we could see in that analysis that everybody was hitting the nadir of tumor at 21 to 28 days, arguing that, what you said, Steve, hitting them hard up front might be a very important thing for axi-cel [axicabtagene ciloleucel]. It may be different in other CAR T-cell therapies. But with axi-cel, you needed to get the big bang up front.

: Going back to something Loretta said, too, that it’s important that a good number, if not half of these patients would not have met the criteria. And we’re still seeing efficacy. And even in the patients who had poor prognostic factors, patients still did benefit. The question of selecting patients, at least in my opinion, it’s more about an optimization question than an exclusion, they’re acting on the considerations in terms of how are we paying for these therapies. But benefit was seen across the board. And I don’t think there was a specific population that uniformly did horribly 100% of the time.

: Quite frankly, Matt, the toxicity in the real world is lower than the toxicity in the clinical trials. What does that mean? It means history is important. We’ve learned something over the years. And the patient selection, as Caron can tell us, the patients had very poor characteristics in relatively speaking to the clinical trial patients. And yet toxicities were lower than expected. Maybe in the real world, doctors don’t see toxicities and report them.

: I don’t even know that we’re just improving toxicity. If we take a look at some of the real-world data, we may even be improving in efficacy, as well, because when you do tease out the patients [who] were eligible versus ineligible for clinical trials, the group that’s eligible is actually lifting as a whole, they’re the same. But when you tease it out by 2 different groups, the group that was eligible is actually lifting up the curve, or it’s looking better in multiple series now, compared [with] the pivotal clinical trials. To Matt’s point, we’re not only getting better at toxicity, we can treat people with more comorbidities, but we’re also getting better at selecting the patients and getting patients in faster, which I think is a really important point…to meet these patients earlier in their disease course, earlier in their relapse, so that we can get them in and get their cells faster when they’re in the best position to respond.

: Three studies showed that [women] who are over 65 or 70 do the best. Isn’t that something?

That may be, as we were saying earlier, a niche that autotransplant couldn’t serve, which is you’re talking about not only does CAR T-cell [therapy] save some patients [who] relapse after autotransplant, but now you’re talking about niches of patients [who] might not be eligible that you can treat. Clearly, and the real-world data are showing this. I’m waiting for Caron’s abstract next year. I want to see if this holds up in the telehealth world, whether we get these same good results.

: It is shifting our transplant practice. When you look at the patients who are having what we would now consider suboptimal responses or partial responses following salvage therapy, we’re thinking about those patients differently now in the world of CAR T. The trials don’t necessarily reflect that, and some of the randomized studies that’ll be coming out and the phase 3 studies. But overall, moving forward, we are shifting the needle in terms of transplant versus CAR T in this population, just knowing the burden of treatment and taking a patient who may still have chemorefractory disease.

: I’m eagerly awaiting the results of the randomized studies in second line, because for some of these patients [who] we don’t think are transplant-eligible, right now, the label makes us treat them with a second line of chemotherapy, and it would be nice to be able to take those patients directly to CAR T cells. We have 3 randomized studies looking in the second line, and I think that those will be important to further this along.

: I agree 100% with Caron. I agree because I’m a convert to this opinion that you just expressed. I totally thought this would just get rid of autotransplant completely. I was so impressed with how well we were doing with CAR T cells nowadays, and how poorly our outcomes have become in the rituximab era with autotransplant. And most of our patients are progressing in a year, and they do terrible, but guess what? Okay, so we’re expressing opinions. I’m going to share observations. These come from my practice. I have patients in my practice [who] were part of 1 such trial, who progressed after getting CAR T second line and went on for salvage by transplant. And I have early [University of Pennsylvania] data on that exact situation, progressing and failing CAR T cells despite expansion, and then having a splenectomy and go. Because this particular patient, that was probably 1 of the reasons he was failing. And getting an autotransplant, he’s 5 years in remission now. And then the converse is the patients [who] have an autotransplant and stay in remission. When you do this randomized study, I think we need these studies to be done, because there is going to be a subpopulation of patients [who] will still benefit from transplant, even autotransplant, as primitive as it sounds. And…I know for a fact that it will even salvage some CAR failures. So we’ll wait and see.

: I think we’ve all had those. Whether you resensitize a tumor to chemotherapy following CAR or whatever the different selection pressures are, we’ve all probably taken a couple patients to allo[transplant] following a CAR T failure once you’re able to get the patient back into remission. And there are people who are still walking around doing that.

: But I’m talking auto, not allo, Matt. And in this particular case, the CAR T-cells were probably being sequestered in a massive spleen. And taking the spleen out was the key.

: The point is, keeping score is very important. Sponsor randomized trials are very powerful. But having real-world data and CIBMTR [Center for International Blood and Marrow Transplant Research] registry data going forward where we can track the people who were on those trials, who went on to another trial, and [then] went on to another trial. We’ve been talking in our own center about the problem of competing risks. This is a good problem to have. In the old days, competing risk was if the therapy didn’t work, and you died of something, we blamed the therapy. Now…we’re becoming multiple myeloma docs, right? We have multiple therapies, and patients are benefiting, living longer. Good news for the patients.

: This is also going to be a paradigm for multiple myeloma. We’ve all seen the patients who are getting to the CAR T therapies with BCMA [B-cell maturation antigen] and other antigens. They receive average 5, 6, 7 lines of myeloma therapy. Just like we can with patients prior to going to CAR T with lymphoma, we can do the same thing in myeloma. And if it does hold true, that we do get durable responses in myeloma eventually, moving it up in earlier lines is only going to benefit patients.

: It’s pretty difficult to think of 11 lines of therapy for lymphoma. But I agree with what you said.

LDH and ECOG Performance Tests

: We’ve talked a lot about axi-cel (axicabtagene ciloleucel). Tisagenlecleucel is the 4-1BB-driven CAR (chimeric antigen receptor ) developed at Penn [University of Pennsylvania], and we’ve played a major role in its clinical development. We have data now that have been presented at 1 year, which was about one-third of patients who were responders [and] were still in remission at 1 year. I can tell you that that hasn’t changed at 2 years, although I’m not sure that that’s yet appeared in abstract form. But again, this is the same construct that was licensed from Penn to Novartis. We have follow-up in large-cell lymphoma now, median follow-up that’s 5 years. These responses are durable. At Penn, we received better than one-third. We were actually between 40% and 50% still in remission at 5 years, not because our vector was better at Penn, the one that was licensed to Novartis, or we made it better. It’s because in every single study, there [are] differences among patients, selections, and centers. There are so many variables that can influence the outcome, and I’m only comfortable saying it’s between 30% and 40% for any of these products in terms of durable remission. There are too many variables to sit down and compare studies side by side.

We have 2 available home runs. And we’re waiting for a third home run to come in later in this year. And that’s liso-cel, or lisocabtagene maraleucel. That is another 4-1BB-driven CAR developed by Juno and licensed to Celgene and now to BMS, [Bristol Myers Squibb], and they’re all one big happy family. Jeremy Abramson presented data at ASH [American Society of Hematology] last year, which I thought were pretty exciting, [first because of] the patient population. Every study is a little different. And just the way [it was] in follicular lymphoma, I’ll be impressed by 2 years if I know the patient characteristics; you really have to know the patient characteristics [when] comparing even large-cell studies. The data presented by Jeremy included some histologies [that] weren’t included in other aggressive lymphoma studies. For example, transformed marginal-zone lymphomas, you were allowed to have had CNS disease. That was not done for either of the other products. The data that were presented showed the CR [complete remission] rate was 40%, similar to what was shown with the other 4-1BB CAR. And one-third of those patients had ongoing CRs at 1 year. Looking like tisagenlecleucel, with a very low toxicity profile, which may be a historical artifact, because we’ve learned a lot about CRs and how to manage and prevent it. And earlier intervention leads to lower grades in the end. Not only are there patient selection characteristics, but we’re all victims of history, as well. 

: We get different outcomes based on who the patients are that we put in, and, of course, if we do clinical trials that are very stringent, we get a more stringent outcome. If we do experiences, [such as] the real-world type of experiences where you can capture consecutive patients at large centers, then we get to have increased numbers and a little bit more variety. The retrospective analysis of those multivariates becomes very interesting.

Liso-cel Therapy in R/R Large B-Cell Lymphoma

The Importance of Antigen Loss in R/R Disease

Let’s talk about axicabtagene ciloleucel, or axi-cel. That was the first approved CAR [chimeric antigen receptor] T-cell therapy for lymphoma. About 4 to 6 months before that in pediatric ALL [acute lymphoblastic leukemia], there was approval of tisagenlecleucel. Technically, that smaller market was hit first by tisagenlecleucel. But in the big market, which is lymphoma, axi-cell [axicabtagene ciloleucel] was more common. We have mature data that Fred Locke published in 

 One thing I’ll mention about this is that it’s a win for science, and it’s a win for clinical researchers, because I have patients beating on the door trying to get on these phase 1 studies, and it was something that was so novel, but it did reinvigorate our clinical research program. As you mentioned, some of the early publications were criticized based on the relatively short follow-up. The questions about whether this would be a durable treatment in a onetime therapy for patients with large-cell lymphoma. What we saw with Fred’s publication with longer follow-up, at the time he had about 27 months of median 

 follow-up. We didn’t see any new late-toxicity signals, which was incredibly important. The overall response rate, which was the best response rate, settled in at about 83%. And 58% of patients had achieved a complete response. ASH [American Society of Hematology Annual Meeting] this past year updated that with 3 years of median follow-up and shows that the overall survival had about 47% still alive at 3 years of follow-up. That starts to answer the question that yes, there is a subset of patients, approximately 40%, for whom we do think this is a definitive onetime therapy. The question remains, how can you improve outcomes for those 60% of patients where that’s not true? 

 I’ll tell you that 5 years ago, when we saw our large-cell lymphoma patients who were refractory to immunochemotherapy go into remission and not die within a year, that was very exciting. Now we’re talking about, “Wait, we didn’t cure them. What are we going to do now?” There was a time where we didn’t cure anyone. And you’re right. What’s impressive is, 40% of those patients, for 2 to 3 years are in remission. Most people, including the FDA, agree that 2 years is a pretty good benchmark. About 80% to 90% of relapses occur in large-cell lymphoma within the first 2 years anyway. These are good data. As Loretta said, 2 years is certainly a benchmark, in my opinion, for aggressive lymphomas. But Caron, would you feel as enthusiastic at 2 years for follicular lymphoma if you’re sitting on the FDA committee reviewing the new drug application?

 I don’t think we can say anything about it being a definitive therapy for follicular lymphoma after 2 years. But if you think about the way it’s been tested in the third line and beyond, median 

 duration of responses to other available agents in that setting are under a year. If we continue to see more than half of patients maintain their response at 2 years, that’s a meaningful opportunity for patients with follicular lymphoma. I cannot say it’s definitive, though.

 No, you are spot-on. It really depends on the patient population, right? As we were saying before this conference in our informal discussion, there are those who we know early on have a terrible prognosis, 

 chance of being dead at 2 years. That’s within the first 5 years, rather. You’ll know quickly. At the other end, for patients you’ve been following for 10 to 15 years, us old doctors, we know when they get to the end, what the survival is like when their remission duration becomes very short to the existing agents, under 12 months in most cases. That’s a group of bad actors. If you look at those patients, I would be impressed with 2 years in some of those. There are a group of patients with follicular lymphoma, despite the fact that it’s “indolent,” that I would be impressed with 2-year data. They’re the patients Caron described. There’s a subset of bad actors. It really depends on the eligibility criteria.

R/R Large B-Cell Therapy in Managing Various Lymphomas

 We’re talking about finding a potentially curative regimen for people who failed a therapy that was potentially going to cure their lymphoma. There are a lot of lymphomas that we don’t have conventional cures for. And when we get to the second- and third-line treatments for those lymphomas, even though they have generally longer natural histories, response durations are short and people relapse, and we start to see decreased life expectancy. What we’ve already seen for mantle cell lymphoma, knowing that patients who relapse after BTK [Bruton tyrosine kinase] inhibitors have a short survival, on the order of months, has been impressive. Some data were presented in indolent lymphomas as well, especially late lines of therapy.

 A lot of that will be your data, and they are very important data. For indolent lymphomas, people are less impressed. There is a business of progression of disease within 2 years as a subset, maybe 20% or so of patients have a bad indolent lymphoma. But having been in practice for 40 years, I’ll say that when you’ve taken care of these patients for 10 to 15 years, and they’ve been through a number of treatments, guess what? They’re not progressing within 2 years. It’s a bad disease. And Caron’s right. We need something for those patients too. It’s just not captured the same way, because we’re looking at patients with much shorter follow-up. At the other end of the indolent lymphoma natural history, it gets bad.

 It’s important to remind the audience that we’re the gateway drug, right? We’re the gateway drug to cellular immunotherapies for the nonblood cancers or the diseases that need alternative therapies beyond chemotherapy. The lessons we learn in lymphoma, leukemia, mantle cell, and myeloma, like everything that’s ever happened in hematology, will be important guides for how you combine therapies, how you target. We’re the gateway drug.

 You’re right. Rituximab taught us that patients can respond to immunotherapy when they’re resistant to chemotherapy. You hit an external target with immunotherapy, and you can get a response in a disease that is resistant to all kinds of toxic substances that kill cells by getting into them. We saw that with Rituxan, but then there were Rituxan failures or Rituxan chemotherapy. Now we’re at the cellular end of immunotherapy, as opposed to serotherapy, and we’re seeing the same thing again. Patients who are resistant to those drugs that kill cells internally—like chemotherapy, small molecules—can respond to immunotherapy. And those are unresponsive to intracellular agents and rituximab or anti-CD20s. You can get at them with a T-cell–based therapy, another extracellular agent.

 We were trying to do a GVL [graft versus leukemia] and allogeneic transplant to a much larger degree with a shorter period of toxicity and fewer toxicities. It’s less of a bazooka approach of an allogeneic transplant and more targeted.

 What do you think? Is allogenic going to disappear, Matt?

 It depends on the disease, the context, and whom you talk to. There are going to be a subset of patients once we can risk stratify. I think there are going to be various clinical characteristics or drug characteristics or product characteristics in terms of persistence, and what may be important for 1 disease in terms of persistence may not be the same for another. Depending on what we see and how we risk stratify, I can see a world where we have low-risk AML [acute myeloid leukemia] that you can cure with consolidation therapy, and a low-risk lymphoma or leukemia that you don’t need a transplant for. For others, you may want to transplant at 3 to 6 months.

 There’s an important point to make here, though, and that is that the lymphocytes that give rise to lymphoma are long-lived cells. They’re far beyond the hematopoietic stem cell stage, sometimes a lifetime’s worth, whereas the cell contributing to AML and MDS, myeloproliferative disorders, that’s about 14 days down the pipeline. We are continually going to be facing the problem that the patient needs not only a cellular therapy, for AML or myeloproliferative disorder, but a lot of repair.

 Loretta is going to tell us about AlloCAR T therapies. There’s always going to be AlloCAR T if you believe in CARs. We’re not addressing stem cell diseases. We’re taking care of neoplasms of mature cells. There always will be diseases where we have to replace stem cells. The only way I can think of is CAR. I’m sure with the way technology is moving, we’ll be differentiating our skin cells. You’re going to need different stem cells, not your own, for certain diseases. It’s never going to disappear, in my opinion. And you know I believe very much in CAR T-cell therapy.

CAR T-Cell Therapy in Managing Various Lymphomas

 I’m Steve Schuster from Abramson Cancer Center at the University of Pennsylvania in Philadelphia. Joining me today for a lively discussion are my friends and colleagues Dr Matthew Frigault from the MGH [Massachusetts General Hospital] in Boston, Dr Caron Jacobson from Dana-Farber Cancer Institute in Boston, Dr David Miklos from Stanford University Medical Center in Stanford, California, and Dr Loretta Nastoupil from The University of Texas MD Anderson Cancer Center in Houston, Texas.

Today we’re discussing a number of topics related to the use of CAR [chimeric antigen receptor] T cells in patients with lymphoma. We’ll be making reference to ALL [acute lymphoblastic leukemia] as well, because that was the earliest data. Most of us have some experience. But we are primarily treating lymphoma patients right now. We’ll talk about the latest information that’s out there, some of which was presented in December at the American Society of Hematology Annual Meeting, and most recently at the meeting that just happened, which is ASCO [American Society of Clinical Oncology Annual Meeting].

Let’s begin with lymphoma and possibly discuss leukemia. Some of the leukemia data will come out as we discuss lymphoma, and we can compare and contrast diseases that are treated with the same agent but with different outcomes.

What is the rationale for CAR T-cell therapy? We know it’s approved for diffuse large B-cell lymphoma. There must have been a need if the FDA approved it. What do you think is the rationale for further development of diffuse large B-cell lymphoma? There will be new developments. Who wants to weigh in?

 It’s 1 of the first therapies we’ve brought to the market that’s a cellular gene therapy, and we’ve overcome barriers in terms of technology and regulatory hurdles. And we’ve seen it only accelerate. As remarkable as the responses have been, there’s still room to improve the technology, whether it’s in efficacy or toxicity or just having more access to patients.

 I agree. What you’re seeing right now are the first-generation cars, the Model Ts. It’s going to get better and better, and every manufacturer is going to say that they make the best car. That’s what happens in Detroit and everywhere else, right?

 If we consider what happens when the patient has progression and relapse after R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone], chemotherapy and other things that we call salvage on a goal toward going to a high-dose stem cell transplant with, at best, a 50% response to salvage and 50% durable responses after autotransplant. We can help many patients by bringing an effective immunotherapy forward, if for no other reason than what they endure going through mobilization, salvage chemotherapy, and an autotransplant. If you add it up, that is 5 months of their lives right there, let alone the problem of median 

time to survival of patients who have relapsed, since refractory chemotherapy-resistant lymphoma is only 5 to 6 months. 

 Just to reiterate, because they’re really important points. There are patients who would not be candidates for autotransplant who are candidates for CAR T. This relates to patients who are older and have comorbidities, because it’s a lot easier on patients, as we’ve learned, than high-dose chemotherapy. This is a benefit for those patients. We’ll wait for additional studies to see if we can move it to second-line for some groups of patients that are not transplant eligible. Also, our ability to salvage patients with autotransplant in the immunochemotherapy era has dropped significantly. I remember when we first started doing autotransplants in the late 1980s, prior to rituximab. And 50% of the patients were fixed, OK? In the current era, with rituximab and particularly those who fail within the first year of receiving a rituximab-containing induction regimen, the outcome with second-line chemotherapy and autotransplants is kind of abysmal. That’s the other big niche for CAR T in aggressive lymphomas.

Matthew J. Frigault, MD, Massachusetts General Hospital Cancer Center

Articles