Transcript: Neurology Journal: Dr. Bevel Interviews Dr. Liau
Dr. Linda Liau joins Kate Neville on Neurology Journal Podcast, Mar 2, 2023
0:27
Dr. Nevel: Hi, my name is Kate Nevel and I'm a neuro-oncologist at Indiana University and today I'm delighted to be talking to Linda Liau, professor and chair of the Department of Neurosurgery at UCLA, about her recent study and publication “Association of Autologous Tumor Lysate-loaded Dendritic Cell Vaccination with Extension of Survival Among Patients with Newly Diagnosed and Recurrent Glioblastoma,” a phase 3 prospective externally controlled cohort trial published in JAMA Oncology in November of 2022. Dr Liau, thank you so much for being here today and talking with me.
Dr. Liau: Thank you for having me.
Dr. Nevel: So I remember when this paper was published because the neuro-oncology community was at a national meeting and you know kind of wondering what this would mean, and we'll get to the details later, but since this paper has been published I've have had patients calling emailing me asking me about DCVax-L. So for our listeners out there what's the most important thing you think clinicians should tell patients when they call asking about this study?
Dr. Liau: This was a phase three non-randomized prospective cohort trial of autologous vaccine for brain cancer that has shown some promising results. However it is not yet FDA approved so it is not currently available in the U.S outside of clinical trials. And you know further clinical trials are ongoing using this agent.
Dr. Nevel: I think in order to better understand the results of the study, which we'll get to, we first need to go over some of the methods and design and how you analyze the results. So do you mind giving us some background on that the study was initially designed as a phase 3 double-blinded randomized controlled trial with the crossover arm. However as the trial is underway we began to realize that because of the crossover arm, eventually about 90 percent of all patients received DCVax either at new diagnosis or at presumed recurrence. There are actually several patients that you know subsequently were probably found to have pseudoprogression, so in the end it was depletion of the placebo control arm. Therefore before the trial was ever unblinded a prospective Statistical Analysis Plan (SAP) was designed to compare overall survival in the newly diagnosed cohort and the recurrent glioblastoma cohort of patients who are treated with DCVax with contemporaneous match external control populations from control groups of other formal randomized clinical trials. Because of the crossover arm what we essentially we had were two groups of patients, one who got DCVax early, at new diagnosis, and one who got DCVax at first recurrence. Those were kind of the two populations were left with with this trial, and that's why it was controlled to external controls. So yes it's not a randomized controlled trial but that is the background of the methods of the study. The comparator trials were chosen based on predetermined criteria and a broad ??? was conducted using these criteria, which included contemporaneous study time period, a similar KPS and age groups, reported outcomes with overall survival, and you know various other criteria that are listed in the paper. So based on that there were five newly diagnosed studies that were chosen for the comparator population and ten recurrent glioblastoma studies that were chosen for the comparative population for that group, and then our statisticians did an internal validation using various specificity and sensitivity analysis. So overall that was the methodology of the study and you know I'm aware of the limitations of using external control populations, but given the conduct of the trial and the depletion of the placebo arm we felt that this was a comparison that made sense, and with its known limitations.
4:47
Dr. Nevel: I think that brings up an interesting topic that maybe is beyond the discussion for today about using these externally controlled cohort populations in cancer research specifically because of some of the the reasons that you mentioned, that the ethics behind not offering an intervention or ability to cross over for patients with cancer especially patients with incurable cancers.
5:13
Dr. Liau: Yeah, correct. And at the time of the trial initiation it was actually quite difficult to get patients to enroll because patients don't want to be on the placebo arm and so the crossover arm was placed to allow for that option, but retrospect the comparators were for this trial
Dr. Nevel: With that could you please review the most important results of your study
Dr. Liau: I think the most important results of the study are that patients with newly diagnosed glioblastoma receiving DCVax had a median overall survival of 22.4 months from the time of surgery, 19 months from the time of randomization, and the patients with recurrent GBM had a median overall survival of 13.2 months. And these survival statistics compared favorably with that of the external control populations from over 1300 newly diagnosed glioblastoma patients and 640 recurrent glioblastoma patients from various other published clinical trials. But what I think maybe even more interesting is the long tail end of the survival curve and some of the subgroup analysis results that came out of the trial. For instance the median survival of the MGMT methylated patients was over 30 months, with almost 20 percent five-year survival. It was interesting in that the majority of these patients surviving over five years are surviving without recurrence, which in the field of glioblastoma is actually quite unusual. So I think it does give us some information in aide in terms of how to design future clinical trials and perhaps how to stratify patients for immunotherapy trials and potential biomarkers.
6:54
Dr. Nevel: I found that particularly interesting to the different subgroup analyses and the MGMT group specifically, which we know in general is a subgroup of patients with glioblastoma who in general we expect to live longer and have longer progression-free survival than patients who do not have MGMT promoter methylation present or having what we call an unmethylated MGMT status. Sometimes but obviously more of a benefit you found in your study in that group of patients who received DCVax-L but I was curious if you had hypothesized a mechanism of or physiologic you know kind of pharmaceutical reason why the vaccine might work better for patients who had MGMT methylated tumors?
7:48
Dr. Liau: DCVax-L is an autologous dendritic cell-based vaccine that consists of a combination of a patient's bone dendritic cells which are antigen presenting cells that are cultured from blood cells from the patient's blood and then we put that in combination with a tumor lysate from the patient's tumor that is taken at the time of surgery and the concept is that the dendritic cell or antigen presenting cell is able to phagocytize and process tumor antigens from the tumor and then the cell combination is injected back into patients to activate endogenous T-cells and we have shown from previous pre-clinical and early clinical trials that the use of dendritic cell vaccination with autologous tumor lysates does induce proliferation or infiltration of T cells into brain tumors, which which we think is the initial first step for induction of an immune response in these brain cancer patients. One problem with immunotherapy for glioblastoma is that glioblastomas tend to be immunologically cold – there are no T cells in the tumor, so that's why a lot of these checkpoint Inhibitors and other immunomodulators don't work as a single agent. So by activating the immune response using a vaccine, we're hoping that getting the T cells into the tumor will allow for kind of that initiation of immunotherapeutic response
9:25
Dr. Nevel: Interesting. I guess this would probably be a good time to, if you don't mind, just in really kind of basic terms explain what exactly the DCVax-L is and how it works.
Dr. Liau: DCVax is a dendritic cell vaccine and the L stands for lysate, so the dendritic cells are antigen presenting cells. They're actually normal antigen presenting cells in the body but we've developed a way to actually grow them in large quantities outside of the body, taken from a progenitor blood cells, and these antigen presenting cells are then co-cultured with tumor lysate from the patients tumors; the concept behind that is because of the relative immune privilege of the brain, the dendritic cells circulating around do not typically encounter tumor antigens within the brain, so this is just a way to basically get the dendritic cells together with the tumor antigens and then that itself is a vaccine that's injected back into patients to activate the endogenous T cells. And it's the activated T cells that traffic and migrate into the tumors to initiate an immune response. We have found in our previous pre-clinical and early clinical studies that is something that DCVax can do is, it does induce the trafficking of T cells into brain tumors.
10:53
Dr. Nevel: So we were talking earlier obviously about the methods and design of the study and the externally controlled cohort, and we talked about that as one of the potential limitations of the study. In addition to that, any other limitations or challenges that you encountered conducting this study?
Dr. Liau: The patients treated with DCVax in the study were compared with the control arms of contemporaneous matched external controls who receive standard of care alone. To mitigate some of the limitations using external controls we did use a novel methodology to neural oncology called a MAIC analysis, which stands for Matching Adjusted Indirect Comparison Analysis, which matched patient characteristics between the trials as best we could at the population level. However, we were unable to obtain individual patient level data from the previously published comparator trials so we're not able to do individual patient matching which we are all aware is a current limitation with the use of external controls, so I'm actually hoping that our publication and the discussions around, it will spur the sponsors of these clinical trials to publicly release their individual patient data so that we as a field can further innovate and get regulatory interest in the use of external control arms in neurology and oncology trials. Because for patients with a deadly disease I think there's less and less interest in being the placebo control arm for large randomized controlled trials, especially since the data of standard of care is actually pretty well known from trials that we've done over the past 10 to 15 years.
12:43
Dr. Bevel: On that note for the individual patient data how detailed would the data be?
Dr. Liau: I think even things that contain the typical prognostic indicators – age, extent
of resection, MT methylation, doesn't necessarily need to be terribly complex because quite frankly we actually don't have very many prognostic indicators in the glioblastoma field, but I think it'll open up a way to really get patient level data, and I think the FDA is becoming more open to the use of external controls as the comparator arm for large clinical trials.
13:19
Dr. Bevel: So how do you foresee these results potentially impacting patient care in the future?
Dr. Liau: I think for next steps for the DCVax-L, based on what we've learned about the mechanisms of action and the potential biomarkers of response, we are currently conducting additional studies to look at the combinations of DC vaccines with various immune checkpoint
inhibitors and immune modulators so it can help to further enhance the efficacy, because even though there does seem to be a signal of efficacy with almost 20 percent five-year survival rate, that still means 75 to 80 percent of people aren't living five years, so I think we as a field need to do better. I think there's not going to be a one magic bullet that cures glioblastoma, but having different options available to patients for real world studies of rationalized personalized immunotherapy I think would be useful for the field. So given the very favorable toxicity profile for DCVax and its potential efficacy, I do hope that it can be FDA approved someday and incorporated into testing various combinations, if not the standard of care in the future if these subsequent trials pan out.
Dr: Bevel: So on that note, do you foresee DCVax-L being in a combination randomized controlled trial study looking at DCVax-L with a different agent or another study with temozolomide in a randomized controlled trial setting?
14:53
Dr. Liau: You know one problem with a lot of randomized controlled trials is that it itself is also artificial compared to the real world use of a lot of our therapy. So whether I foresee another randomized controlled trial, I think it depends on the sponsor of the trial, but I think if we are able to get these options out to the community for testing, I'm hoping that there will be greater data that we can generate, because I think one of the problems we have with a lot of these clinical trials of single agents is that without access to the agents it's hard for people to do these combination studies. So I'm hoping that there will be innovation in the field in terms of how we can best move some treatments forward, given the current regulatory as well as clinical trial environment, that kind of limits how we do larger trials with combination therapies.
Dr. Bevel: So patients sometimes, especially since the publication of your paper, they call in, and this is not an FDA approved therapy, but how can they access DCVax-L right now? Is that information you can share with us? Could you talk to that a little bit?
16:11
Dr. Liau: DCVax is currently available for compassionate use in the UK; it was approved by the MHRA for that purpose. It's not yet FDA approved and I think in terms of access to treatments there's the balance between access and also getting more data and validation of efficacy, and I hope there will be a way in the future to be able to do both – to be able to allow patients access to treatments and then I have the neuro-oncology community continue to test agents and come up with innovative trials to prove or disprove efficacy. One problem is that unless you can get access to the agent it's very difficult to do that, so perhaps a way to open access to DCVax before continued testing would be beneficial to the field.
17:09
Dr. Bevel: Absolutely, and speaking just on a broader scope, patient access to medications is such a big issue in many fields of medicine, cancer care included. With insurance and approvals let alone access to things that are not approved for compassionate use, or experimental trial use. It's a challenge. But I appreciate your comments and you being so candid about it.
Dr. Liau: A lot of patients now in the neuro-oncology community, they're just taking a lot of different off-label drugs that have even less proof of efficacy for glioblastoma, but it's FDA approved for another indication and they want to try it, and neuro oncologists are giving it. I guess my hope for DCVax is that maybe it could be accessible so that we can do these studies and allow patients the right to try.
18:05
Dr. Bevel: Thank you so much Linda for joining me and answering my many questions and the dialogue about your study of DCVax-L and the challenges surrounding cancer clinical trial design. I really appreciated our conversation today.
Dr. Liau: Thank you thank you for having me.
