Next, Gautam Mehta is going to talk about his outcomes with adoptive self-transfer with TILs for patients with melanoma brain metastases. All right, so thank you for having me out to speak today. I'll be talking about our clinical outcomes using a brand of immunotherapy that we use at the NIH called adoptive self-transfer for patients with melanoma brain metastases. So we have no disclosures. Melanoma brain metastases, as you all know, affect over a third of patients with metastatic melanoma. The survival after a diagnosis of a brain metastasis is quite poor and often less than a year, and these lesions are frequently hemorrhagic. The type of immunotherapy that's typically used at the surgery branch in the NCI is something called adoptive self-transfer, and what this is is we take a tumor from a patient with metastatic melanoma and grow it out ex vivo in the presence of T-cell proliferating factors like IL-2, and then take the T-cells and transfuse them back into the patient who has metastatic melanoma. And this can have stunning effects at the primary site, at distant sites, and can lead to curative, durable cures in a subset of patients with metastatic disease. So just before we get into this, just a little bit of vocabulary. So the terms we use for metastatic melanoma are a little different than what we use for intrinsic brain tumors. So metastatic melanoma has three stages, M1a, M1b, M1c, and we're going to focus on the most severe stage, M1c. And typically, the responses are measured using what's called the RISC criteria, which most of you are probably familiar with. So as we heard before, there's been building enthusiasm for immunotherapy for melanoma, specifically with checkpoint inhibition over the past decade. But one thing that's separated out a lot of these trials is that they've excluded patients who have untreated or active melanoma brain metastases. Despite this, there have been a number of studies that have looked at the use of different immunotherapeutic options for patients with brain metastases, including IL-2 alone, which was effective in about 5% of patients, checkpoint inhibition with anti-CTLA-4, which is effective in about 5% to 16% of patients, and PD-1 blockade and PD-L1 blockade, which is effective in about 20% of patients, but typically produces a partial response. Our group at the NIH has previously reported the results of using adoptive cell transfer with these tumor-infiltrating lymphocytes for patients with melanoma brain metastases and shown a response rate of as high as 40% in these patients in the brain metastases. However, all these patients did poorly, so they all progressed systemically. So now that we have these different immunotherapeutic options, how do we actually treat these patients who present with small melanoma brain metastases? And so we wanted to use our model of immunotherapy, adoptive cell transfer, as a way to try to answer this question, first by studying the overall response rates in a more modern cohort of patients treated with this therapy. But then to kind of answer this question, whether or not upfront immunotherapy makes sense in these patients with brain metastases, we wanted to compare the outcomes of patients with brain metastases that were untreated with those who did not have untreated brain metastases. And then finally, when these patients progressed, where did they progress first after treatment? So we included patients with metastatic melanoma who received adoptive cell transfer, and this included a subset of patients who had untreated brain metastases that were less than one centimeter. Patients were treated with lymphodepleting chemotherapy regimen and received the T-cell infusion plus a high-dose IL-2. Patients have pretty significant thrombocytopenia during this treatment, and we tried to keep platelets above 20,000. Responses were measured using the Reneau criteria, which follow Rhesus guidelines, and we also looked for hemorrhage after treatment and overall survival. So this broke down into three groups. We only included M1c patients, the highest stage, and so we had one group with no brain metastases, a second group with treated brain metastases, and then a third group with untreated brain metastases. And the baseline characteristics of these groups were relatively similar. However, the cell dose in the group without brain metastases was slightly higher. Among the patients with untreated brain metastases, the ones we were most interested in, we had 18 patients. Three of them had had prior treatment for some brain metastases, some other brain metastases. None had whole brain radiotherapy. And in these 18 patients, there were 32 tumors with a median size of 6 millimeters. These were small tumors. 44% had multiple brain metastases. We first looked at the in-brain responses, so just what happened to the brain metastases, and found an objective response rate of 28%, a little bit lower than what we'd reported previously. There was progressive disease in 56% of patients, and this included one patient who had a mixed response. So some tumors grew, some tumors shrank after treatment, and then one patient who had a symptomatic hemorrhage after treatment. We had an additional three patients who had stable disease at last follow-up, and this was probably due to very aggressive systemic disease progression. And then finally, all our objective responses were complete and durable. So again, these tumors are very small, so you guys will have to squint or come up to the front of the room to see them. But these are two examples of patients with growing melanoma brain metastases before treatment that were completely eradicated after treatment. So one on the top, and then one on the bottom. The first two are before treatment, and then the second two are after treatment. And then this is our example of a patient with a mixed response, showing that maybe this is not an all-or-none response to immunotherapy. We had two hemorrhages of which one was symptomatic, and there didn't seem to be a relationship with platelet nadir in this group, and we didn't have any other neurotoxicities. So then, changing gears, we looked at the overall outcomes in these patients. So we first looked at overall progression-free survival, and that's looking at brain and systemic progression, and found that the progression in the patients who had untreated brain metastases was significantly worse than the progression-free survival in patients without brain metastases. And as you saw in that prior slide, almost everybody progresses with this metastatic melanoma and the severe stage disease, as high as 94% of the group with untreated brain metastases. When we look specifically at the site of initial progression, this tended to be very low, 13% in the group without brain metastases, and it was as high as 71% in the group with untreated brain metastases. And then when we broke it down further, the rate of progression at that untreated site in the brain was almost 60%. Overall survival mirrored the progression-free survival. And so in conclusion, we found that 28% of patients with small subcentimeter untreated melanoma brain metastases had in-brain durable objective responses. There were also mixed in-brain responses. Neurotoxicity was mostly limited to hemorrhage, which was present in 11% of patients, and there was no other neurotoxicity. And then more importantly, these patients who had untreated melanoma brain metastases did worse than the patients who had no brain metastases or treated brain metastases. And significantly, the initial progression was most often in the brain in these patients and most often at the untreated site. So this suggested to us that in these patients with untreated active melanoma brain metastases, they should have treatment directed at those untreated sites first, either surgery or radiosurgery, before they have immunotherapy with adoptive cell transfer. And I'd just like to thank the following people for their help with this work, specifically Dr. Rosenberg and Rick Sherry. Thank you.

Gautam Mehta

clinical outcomes

adoptive self-transfer

TILs

melanoma brain metastases