![[title]](https://i.ytimg.com/vi/lM1rRYnWBKk/maxresdefault.jpg "[title]")
good afternoon, welcome to this week's lecture in honor of the first director of laboratories and clinics which i was told was the predecessor to my job as deputy for intramural research. before i introduce today's amazing speaker let me make a
couple announcements. we'll not have a wals next week. the next is march 11, the margaret pittman lecture at natcher, not masur. we expect a big audience but mostly because it's an rna conference and she's going to be also part of that conference.
it will be at 3:00, march 11th, pittman lecture. . so the lecture in honor of g burruogs mider, the speaker is crystal mackall. she has training at the children's hospital in internal medicine and pediatrics. i guess she decided to leave the
midwest and come east, and did a fellowship in pediatric oncology in the pediatric oncology branch and worked subsequently in the experimental immunology and returned to pediatric oncology branch where she became first a tenured track investigator and a dozen years ago was tenured as a
senior investigator at the nih. her research has focused on the interactions between the immune system and cancer, and initially i think the interest was more in the way in which cancer affect the the immune system, particularly how chemotherapy affected the immune system, it
was clear that patients particularly pediatric patients who were getting a lot of chemotherapy for leukemia had profound changes in their immune system, and so she began to study the homeostasis of t-cells in patients with cancer and actually developed an important
immunorestorative interleukin 7 in the treatment of children with pediatric cancer. more recently, i think, her work has morphed into the use of cells to treat cancer, particularly adoptive cell therapy of cancer, and you're probably all aware because it
deserves press, an amazing way of treating cancer, that her particular work is focused on the use of t-cells expressing the cd 19 chimeric antigen receptor to treat pediatric t-cell acute lymphocytic leukemia and she's published what i call amazing informatic
results, i suspect you'll hear more about that work today. her leadership in this field is recognized in a variety of ways including very recently being asked to head a cancer project to develop cell-based therapies for cancer. she has received many nih awards
including a distinguished clinical teacher award and nyi mentor of merit award, several nci and nih director awards for her scientific accomplishments. she's a wonderful nih citizen, crystal hardly ever says no when there's an important job to be done at the nih, and for many
years she's been an informal advisor to me on matters will clinical research. so let's welcome crystal mackall. her talk today is the immune system in childhood cancer, mobilizing the troops. [applause]
>> thank you, michael. thank, michael, for your kind introduction and support through the years and the nih leadership for providing me with this opportunity to speak as part of this distinguished lecture series. in preparing this talk i wanted
to find out more about g. burrougs mider and was happy to hear he was described as the first director of laboratories and clinics, because the ring of that was sort of right up my alley because as you will see, as i go through my clear a bit in this talk, my career here has
really been a celebration of the nih, both laboratories and clinics. and i hope that i can demonstrate to you that work such as this cannot be done without the scientific expertise that populates the bench laboratories around this cam put
and cannot be done without the world class clinical research program that is embodied in the nih clinical center. this work is really a synergy between the two amazing resources, that i've had the privilege to utilize over the last 2 1/2 decades.
the reason i'm here started on day in 1985 when i was an intern at this hospital in akron, ohio, had already been interested in oncology and immunology, and i read this article. it was a special report in the new england journal of medicine using isl 2 activated cells, i
never heard of anything like this before, but i was enamored of the notion of using the immune system to treat cancer and of course preliminary but there was some interesting results here, , i wanted to know who was doing this, the national cancer center in bethesda.
in 1989 i finished my residency and was very happy to have the opportunity to enroll in the fellowship program in pediatric oncology at the nih pediatric branch. at the time, the branch was being led by a man named phil pizzo, who really was quite a
mentor for me. not a scientific mentor per se, phil was mostly a clinical researchers, but a very charismatic leader and demonstrated when leading sign activive groups to goals that aren't always difficult how important it is to have the
charismatic leader in your midst, the and the message phil left me with, phil resonates with me today, as a pediatric physician scientist, it's notjust an option to try to create new therapies for your children but it's an ethical imperative. there are people, and people
even today, who have trouble thinking about doing investigative studies for children with dread disease. sometimes it's easier to say, well, they are a vulnerable population, and maybe you'll be pushing an ethical boundary, but phil really taught me, and i
agree, it's our ethical interactive to take the advances to our children. we had a lot of kids with hiv on the ward and i got interested scientifically as michael alluded to in the problem of t-cell lymphopenia, and at the time we had very little
understanding of how the immune system, how the body, is able to respond to an insult with the t-cells. and so i needed to find a scientific mentor, and i was lucky enough, and i mean this absolutely lucky enough to find ron gress who was leading the
transplantation immunology session in the experimental immunology branch at that time. and of course, interested in immunotherapy of cancer, and at that time certainly bone marrow transplant was and still was the first effective immunotherapy for cancer, and so i went to
work with ron, and ron taught me so much, not only about science but the culture of science, he was also a physician who understood how to make that link between the bench and bedside. we got really interested in this problem of t-cell depletion. this was the clinical vignette i
remember. i'm sure ron remembers as well. this was a 15-year-old with high grade brain tumor treated in the pediatric branch. at the time the myeloid growth factors had just come out. if we could give chemotherapy
and higher dose, it was limited by neutripenia, if we could give more, we could cure more patients of cancer, so using gmss, if we started to give high doses every two weeks of cyclophosphamidi. one patient went home, he needed
a platelet transfusion, he received them at the local hospital that weren't irradiated and came back with a devastation from which he ultimately died. this vignette, i think to a physician, you know, it teaches you the importance of irradiated blood products but to the
scientist in us it was really a red flag that something is happening with these patients that we don't understand. we had no idea how their immune system could be so weakened and why this would be the case when the neutrophils were recovering so well.
and this was of course associated with other complications we were observing, so we began a bench to bedside to bench project focused on the problem of impaired immunoconstitution in humans and this is just a little bit of raw data from our early
publications. again you see how well with each cycle of chemotherapy most of the cell populations recover quickly but look here in red. the lymphocytes get worse and worse and there really wasn't a good pathway to make these lymphocytes come back but we had
no idea what the rate limiting factor was. i went into the lab, using mouse models, and began to dissect how the immune system can -- the t-cell immune system in particular can regenerate itself after it's been depleted by disease or therapy.
some of this, there was sort of a general understanding of it, but it really was not very well fleshed out. there was the recapitulation of ontogeny going through the thymus and regenerating but there was this other pathway that wasn't well understood
called homeostatic peripheral expansion, but could become dominant. there was this kind of vestigial pathway that people liked to talk about and we ended up demonstrating these were the two major pathways and that you could distinguish progeny using
phenotype with flow cytometry. using patient treated in the pediatric branch for lymphoma we were able to identify the profound cd4 depletion and we noted that while the patient started with naive and memory cells you lose your naive
population and upon t-cell recovery in the 3-year-old we saw a recovery of t-cells and we predicted this represented a biomarker for thigh thymic activity. in older populations, even teenagers we did not see recovery of naive or t-cells.
we identified age was the single most important factor that impact the the i ability of the human to regenerate the immune system and early in life humans lose this capacity to regenerate all the other thymic dependent pathways. we were subjecting patients to
reliance on other pathways, and we were able to image the thymus and turned to the problem of, well, it wasn't going to be easy to turn the thymus on. i turned my focus to the homeostatic expansion pathway, the other by which we can make and as imperfect as it is it
turned out to be something that's an important tool now in the treatment of cancer with immunotherapy. and to that end i was luck any you've to recruit a young investigator terry fry as one of the first fellows in my independent laboratory and terry
made an important discovery that when our patients become profoundly lymphopenic, this is associated with a response, and that response is an elevated in available interleukin 7, it was really the kind of data that was almost too good to be true, mirror image of one another.
terry got to where he could predict whether a patient was lymphopenic and you felt confident uncovered something fundamental. we made this observation in people long before we made it in mice, because of the technology that's not there in mice to be
able to measure it. subsequently, we able to demonstrate the reason the il-7 rises is the body utilized less, not produces more. this became a new paradigm, elevation in homeostatic cytokines responsible for the change in homeostasis.
we took this further, this is in collaboration with ron's group undertaking first in human studies with interleukin 7 and the agent behaved beautifully, not toxic, and it provided, in and of itself, a single maneuver, a dramatic augmentation in circulating and
we believe even total body t-cell numbers, that invoked homeostatic peripheral. we were able to image this using pet imaging which demonstrates increased metabolic activity of populations and when i gave il-7 you saw that activity in the lymphoid tissue, quite
remarkably, demonstrating this wasn't just some change in the bloodstream. at the end of these efforts, we really had contributed a new paradigm to thinking about t-cell homeostasis, and that is that when an individual is rendered lymphopenic by whatever
maneuver, this is accompanied by profound changes in nearly every aspect of t-cell biology. and the bottom line is that t-cells become much more reactive to almost any antigen that they are presented with, and in some cases for high affinity antigen, you end up
with just a larger response let's say to a vaccine. in some cases to low affinity antigen you change the nature of the antigens you're capable of responding to that you would have normally ignored become sufficient for antigen expansion and the memory pool
is largely tcr dependentent. i was recruited back by lee helman, involved with stalled progress for tumors. things haven't changed much in the last ten years, this is showing the decline in mortality rates for both solid tumors and
this is leukemia, lymphoma, solid tumors. the point is that the rate of decline has been greater for humannology make lig nancy, we have a ways to go but we've not so much progress against solid tumors, there was some really flattening of the curve for at
least a decade in the '90s. this is an example of the kind of data, this is generated here in our clinical trials at the pediatric oncology branch involved in for years, frankly using standard therapy today it doesn't look much different. if you have a patient with a
sarcoma of childhood with non-metastatic disease this is the survival, not so good. if they present with metastatic disease, it's just dismal. so the immunology section was established to ask the question, can we improve outcomes with immunotherapy?
and of course we went back to what we knew. we knew that this lymphopenia was likely to be a problem, we were inducing in all of those patients, this is an example of a patient who comes to you with a new diagnosis of sarcoma, here is the cbc, you can see a normal
lymphocyte count when we get them but eight months or six months later when we're done with treatment a profound lymphopenia, and there's a variety of publications out there that demonstrate relationships between outcome and lymphopenia, these are
correlations, they aren't causations, but it raises the question that maybe, maybe, just maybe, lymphopenia is diminishing our patient's capacity to eradicate residual neoplastic disease. so as an immunotherapist i wanted to try to develop ways to
take what we've learned about homeostasis and apply it to the problem of pediatric cancer and at the time all we really had available was to undertake approaches that could augment existing immunity, okay? and so these are the kind of immunotherapies you heard about
for many decades. let's give a patient a tumor vaccine. more recently, let's use checkpoint inhibitors which relieve suppression on t-cells. let's take tumor filtrating lymphocytes being induced by cause of the natural response
and expand them and administer them, in order for this time of approach to work you have to have a tumor with an antigenic target immune system has to be intact and you have to deal with suppression. we went about this by going after the period of minimal
residual disease often induced in the worst childhood tumors. this is a cat scan through the pelvis, a large tumor, administered dose intensive chemotherapy, you tumor burden goes down dramatically after chemotherapy and radiation, you've lost logs, but this
patient had full pulmonary nodule but the horse is out of the barn and it's hard to get back in control but we reason there's a window of opportunity at the period of low tumor burden where we might use immunotherapy when the patients were profoundly lymphopenic.
we developed an approach that might be able to enhance reconstitution. it was elementary, a first generation trial. why it was important, and why we had these notable publications, were not so much because the clinical outcomes were good but
because we did important and intense biologic endpoints. the dendritic cells we were using were not up to snuff. and that we had a problem of expansion of regulatory t-cells made much worse by administration il-2. and so subsequently now ten
years later we started a second generation trial which was more complex, more involved, i would maybe even call it herculean, the kind of trial you can only do i think at the nih clinical center. but we were asking fundamental questions, and basically
targeting the same population with a new diagnosis of sarcoma or recurrent sarcoma, asking them to come for a tumor biopsy and autologous lymphocyte harvest, they go home in a low burden state and come back for immunotherapy, again very complex and reuse all kind of
cocktails to make a better dendritic cell, depleted regulatory cells. i'm showing you how hard we tried. and in fact, we did a good job in terms of immune reconstitution, il-7 was quite effective at reconstituting the
patients' immune systems compared who patients who didn't get ii-7, and patients had good survival and were treated on this trial. so that's very encouraging. the problem is, i don't know, i don't know what the tumor antigens might have been in my
tumor lysate, those that are a response to the tumor lysate are doing better but it's hard to take this forward when it's so complex and you don't know what is your mediator of response. furthermore, when you measure the response to tumor lysate, when i say what's causative,
we're talking a very, very low number of tumor reactive t-cells, maximum level was .014% in the best patient we had. remember that number because you'll see it later. i'll refer to it later. of course, you know, during the '90s, what the tumor antigens
might be, a lot of individuals including steve's group had done work on this, but it became clear it could be almost anything. and so we really had a difficult time trying to pin this down. and in fact, more recently, there's been sort of a
groundswell, i don't know whether this will eventually end up being dogma or not, but many people believe now that the immunogenicity relates to the degree to which it's mutated, and the more mutations a tumor has, the more likely it is to be immunogenic and there's plenty
of evidence the mutated genes can be immunogenic driven by melanoma happens to be one of the most mutated tumors and one of the most immunogenic. look what's here. if you look along the span of cancer, what are the tumors that are the least mutated?
it's our pediatric solid tumors, maybe our children are not going to be able to generate the kind of immune responses that an older patient with a more mutated tumor would be. melinda merchant in our group who led many clinical trials i'll be discussing today led a
trial of the first checkpoint inhibitor in pediatric cancer, this is another way of getting at this question, can we uncover some kind of existent immune response. if you treat with this you might be able to turn the t-cell on. 30 children were enrolled, the
children got plenty of immune side effects, some severe no objective responses. there was a little bit of a hint of increased survival, if you had an immune related adverse event, but this is hardly anything to get excited about. and so we sort of went back to
the drawing board and said to ourselves, augmenting existing immunity might not be good enough, it might not be of high enough potency and pediatric tumors may have lowering antigenicity. from my point of view the field is exciting, we can now turn
this paradigm on its head and say, wait a minute, maybe we don't have to rely upon natural immunity or upon naturally immunogenic antigens, maybe we can train the immune system to recognize a molecule as antigen but wouldn't otherwise do so, moving yourself into the realm
of synthetic biology. monoclonal antibodies are a perfect example of synthetic biology having an impact on treatment of cancer. we've been focusing on genetic engineering of t-cell populations, chimeric antigen receptors and so i'm going to
talk about efforts in this area. for those who haven't followed this field, of course a t-cell receptor recognizes a peptide in the context of mhc, requires signal 1, especially in the case of naive t-cells requires a second signal, prototype is cd-28 to get actually activated.
on the other hand a chimeric antigen receptor that is currently being used essentially by all groups now in 2015 incorporated the second signal so once the antigen is recognized, you get signal 1 and signal 2. and the other difference with
the chimeric enzyme receptor, it utilized usually a fragment of an antibody to recognize cell surface molecule, it's not mhc restricted, you get signal 1 and 2 when it fires. we're working on several cars in the lab, i'll talk about 19 and 22 and gd 2 and two domains.
i want to give credit where credit is due, this area has gotten a lot of press. this is the first person to come up with the idea, on sabbatical here, it was an out of the box idea when he did this work and he wanted to somehow get t-cells out from under the mhc
restriction problem. this was 1989. it took as you can see essentially 20 years to take this idea and move it to something that had any kind of activity in the clinic. this was investigator initiated work in academia, not done by
industry, and frankly wasn't all that well thought of for the first about 18 years. so this really shows that sometimes plugging away pays off. and the first payoff really came from this institution in 2010, in this publication in blood was
the first report of the activity of a cd-19 car, again a lymphoma. this car was generated by jim and steve's group and targeted cd-19 and use the stimulatory domain. we saw the potential for this in childhood tumors and anybody it
initiated a study with the same car getting the this in children, let by trey lee. it sounds difficult to do this, it sounds like this must be so tedious and involved but i will tell you from my point of view, i've been most impressed by how
feasible it is to generate these products. now, we have an outstanding department of transfusion medicine in the nih clinical center, sophisticated folks who have been doing cell therapy for some time. but i still think that the
notion that this is just too difficult is indeed rapidly fading away. so the way we make these cells is we do an enrichment step to pull the t-cells away, let's say from contaminating non-t-cells, they are exposed to a viral vector and undergo usually a
second exposure, two transductions, and then are expands ex vivo and very importantly back to our early days on the homeostasis issue, if you don't give chemotherapy to these patients it really doesn't work. there's been several groups that
show not necessarily in randomized trials but i think there really isn't much in the way of this field that have not bought this idea that lymphophonic patient is more receptive. it's one of the major points that allowed the field to
progress to where it is today. so we published these results last fall in "the lancet," these are the only intent to treat results. everybody is reported in the report that signed a consent, not patients that were told to go home three months and came
back and treated on another trial. so it's really important to point out that this really is a proper intent to treat trial. all patients were treated with the same chemotherapy as well, a modest chemotherapy. i don't think anybody in the
field would -- - they were all treated with the same chemotherapy. six were never in remission. the fundamental question you you want to answer is, is your therapy doing something that chemotherapy can't do? are you able to eradicate the
chemo resistant cells and the results were quite gratifying, all rendered completely free of disease. 90% of patients were feasible, infused, two were not feasible because the numbers of cells generated didn't meet our bar, one of those patients did have a
good remission and remains alive and well. median transduction efficiency you can see was high. this is our waterfall plot for the 20 patients with all, the patient with lymphoma did not respond to therapy. this shows the change in marrow
blasts in these patients, everybody here in the light blue was rendered into a complete response that was a completely negative by our most sensitive measures of measuring, of flow cytometry, two were flow cytometry positive. i don't have time to talk about
cytokine release today but would be happy to in the future off line. it was remarkable, remember i showed you the .014% and everything we had to do to get to that number? this is a patient who is probably one of the highest
levels but not necessarily out of the ordinary. these are the car cells looking at the car cells with an anti-idiotype on day ten. orders of magnitude beyond anything we could accomplish with the other approaches that are simply vaccines or
checkpoint inhibition. and coincident with expansion of these cells we see eradication of cd19 cells. the expansion was dramatic, so was the contraction. this was bothering us. we were scratching our heads and i'll talk about that in the last
half year. the cells migrate through the cerebrospinal fluid. the car cells go in, the disease disappears. so patients were rendered into an mrd negative remission, outcomes were good, most patients go on to receive
transplant because we just don't know whether the remissions are going to be durable. and that is considered standard of care for patients rendered into remission. to summarize, with single infusion, one cell product, small cell dose, we've got a 70%
complete response rate and refractory b-cell all, we've seen complete responses in all six patients who are primary refractory all, confirming this is doing something chemotherapy can't do. this is an effective bridge to transplant in this population,
whether it's curative is not entirely clear. they traffic for tissues, we had a 14% grade 4 site cytokine release syndrome, it's an important area i don't have time to talk about today but b-cell recovery in all patients, again back to this question, why
didn't the cd-19 car cells persist? people involved in the immunotherapy of cancer have been sort of stating the mantra for the last decade or so, is that persistence is required. now, persistence can be defined differently and different
trials, sometimes it might be persistent several years, sometimes several years, but i think if you want to eradicate specially large bulky tumors, tumors living in a small niche, we believe we can do better if our cells persist. and so one possibility is
there's an anti-car immune response, in this case a murine, and we look for human anti-mouse antibodies, didn't find them, the single chain of c is not the area where they are directed. we looked at t-cell proliferation in response to car transduced versus mock in 11 and
found t-cell responses to the car. in general, the proliferation index to car transduced cells was higher than non-transduced cells, and in some patients that actually preceded the administration of the car, but we found it more often in
patients after administration of the car but even in patients who had the responses we often had good response rate so we couldn't correlate this immune response with the immune response to the car, the t-cell response to the car, with the eradication of the leukemia and
conclude t-cell mediated car responses with commonly found, and might contribute to diminished persistence and the field is moving towards thinking about using fully human single gfbs. as michael mentioned, i'm leading this multi-institution
consortium that has really brought in investigators from around north america who are engaged in this type of therapy, and the goal of the grant is not only to identify new targets using genomics for solid tumors but also to increase collaboration among those of us
conducting car therapy so we can learn more about how best to optimize these and iteraactively improve these. the penn group and seattle group each had a similar cd-19 car in the clinic, using a similar -- the same signal chain, similar expansion platform, and the
notable thing is whereas penn noted prolonged c well aplasia, as the seattle group, our cd-19 car not associated. we see the cars go in, expand and then disappear. and so it was the first kind of clue that maybe the co-stimulatory domain was
playing a role in impacting persistence of these cells and of course cd-28, something we all know, we think we know a lot about, remember the imglobulin super family, the prototypical signal 2 for naive t-cells signals to a variety of pathways, the counterpoint was
4-1-bb, a different family, a super family, but it goes through similar pathways so we didn't have a good feeling for why these should be necessarily different. but we began to hypothesize that maybe the difference in the persistence relates to the
difference in the incidence of t-cell exhaustion. and t-cell exhaustions is something immunologists have been working diligently on, mouse and human, for the last several years, t-cell exhaustion is something that happens in disease state so under normal
circumstances t-cell is presented withen an antigen, the antigen disappears, the t-cell becomes a memory cell. if the antigen stays around and keeps signaling t-cell receptor you see a progressive loss in functionality of the t-cells, this is associated with
diminished cytokine production and capacity and increase in expression of checkpoint receptors. checkpoint receptors are those receptors now being targeted with things like anti--td-1, presumably there to prevent overproliferation.
i'm going to change gears and share some data we generated using a different car that ultimately came back and shed light we think on why we observe the short persistence. gd2 had already been vetted as a tumor antigen among monoclonal antibodies and pediatric
neuroblastoma. you can see here in the rhabdomyosarcoma, and osteosarcoma, getting back to the pediatric solid tumors if we could make a car targeting dd2 perhaps we could improve outcome. we did this in the laboratory,
led by an incredible stay tuned in my lab, adrienne long. what adrienne observed was that the gd2 car functioned pretty well in the test tube but when you put it into mice it really didn't do much of anything against the osteosarcoma, in contrast to the cd-19 car which
also had reasonable activity in vitro in terms of peeling but using a bioluminescence model you can see how well you were able to eradicate leukemia in the mice. we said, yeah, but that's leukemia, we're trying to cure solid tumors, they are more
difficult. it's probably not that the car is different, rather the tumor is different. ed adrienne rose to the occasion to get to the bottom of the disparity and she created an osteosarcoma cell line that expressed gds had and cd-19 and
we asked how good is the cd-19 car and expected to find the solid tumor was more difficult to kill but we were surprised. in fact, even though the osteosarcoma was equivalentry killed in the test tube only the 19 car was able to shrink the established osteosarcoma.
the problem was the car, the rest of the signaling domains are the same, and adrienne began to study these really prior to the time they went into the mice so you take a donor's t-cells, activate it, add your corrector and this is now before you even challenge this with antigen, and
when you look on day 7 the cars look different. they showed signs of overactivation, mock transduced cells had already settled back down by day 7 and their activation marker expression tended to normal ice, gd2 car sells remained activated, as
evidenced. it was exactly what you would expect of the cells continuing to get antigens. in fact, when you look at the data, phosphorylation, you can distinguish the zeta associated with the car from the native cd-3 and look at
phosphorylation, you had the gd2 car already phosphorylated without antigen in the system. and then that followed by the acquisition of checkpoints, you can define exhaustion any one of a number of ways but pd1 is a good place to start. we certainly wanted to rule out
the presence of antigen and i don't have time to go through all of this data. oh, there you go. so we did rule out that there was antigen in the system, gd2 was not present in the culture system. adrienne made point mutations
that ablated, and the exhaustion phenotype remained. fusion protein did not bind to the cells in the system rolling out cross-reactive antigen and we ended up with a working model that this car was signaling without antigen. how could that happen?
you think about well, what's on the surface of the cell? oh, it's single change fbs, usually because of interactionings in the framework region so we began to reason maybe there was spontaneous dimerization. an undergraduate in the lab,
kelsey, tagged our car with a fluorescent protein, and indeed we were able to visualize this. i think quite remarkably, better than we ever expected. the cd-19 car, and we estimate there's about a million of these on the surface of the t-cells, we're talking really high level
of these proteins, but the cd-19 car distributed beautifully, homogenously around the surface of the cell. the gd-2 car is hunting, it's aggregating clearly fretting with itself. we tried to localize the problem and were told it was likely to
be the framework region and so we began to do kind of a series of swapping experiments where we said the cdr-3 region, 1, 2, 3 are antigen binding domains so if it's the framework of the gd-2 car, if we put that in the cd-19 car we should reproduce the phenotype and vice versa fix
the car. we were able to reproduce the phenotype. el when we inserted we saw the presence of exhaustion indicated by the high pd-1. we were not able to get expression on the surface,
difficult for us to do to localize the area on the single chain and mutate the area, likely to involve multiple amino acids. we don't think this is unique to the gd-2 car because when we looked across a variety of cars the gd-2 car is the most obvious
but looking at a variety of cars, we see this phenotype of early exhaustion and, you know, the whole shebang, and shown here is the piechart where you can see mock and cd-19 are the most pristine and you get more and more of the tonic signaling. to summarize, early cart cell
exhaustion, good criminology ex vivo, but for persistence and pore activity in vivo. car t-cell exhaustion is associated with tonic antigen independent signaling. most car show some degree. gd-2 is the worse's cd-19 is the most pristine raising the notion
maybe that's why the cd-19 is so good. here the scientist in you comes out and says, look, we have exhaustion in the dish. immunologists have been wanting to study exhaustion for years but you can only find it in a patient with a chronic
infection. here we start with homogenous cell populations and in 7 to 9 days induced exhausted t-cell. what is the biology that surrounds this? basically what adrian did was see what necessary to induce exhausting.
cd 28 makes it worse, signaling awrong with zeta makes a cell more likely to become exhausted, you can look at this using transcription factors or using your checkpoints. on the other hand, 41bb is protective. co-stimulation leads to much
less impression of these checkpoint molecules, much better cytokine production when the cell is challenged with antigen, and improved anti-tumor activity in vivo. finally we were able to get an anti-tumor effect using the dg-2 car by preventing exhausting of
the t-cells. now you say, okay, that's unique to the gd-2 car, might be the only car that acts like this, how important is it to other cars? well, in fact when we go back and look at the cd-19 car, because you don't have to get
exhausted by tonic signaling, you can get exhausted by high antigen burden, we see that indeed the 41bb co-stimulation leads to more persistence in the blood and diminished expression of checkpoint molecules. in the cd-22 car generated in my laboratory and now in the clinic
targeting patients with cd-19 negative leukemia, again this is work generated by terry fry, that indeed the 41bb version of the cd-22 car is more effective and more persistent and that's the trial that is open now in our branch being led by terry fry.
so, again, it's exciting to make a better mouse trap. it's exciting to make a new therapeutic that works but provides an tune opportunity to learn about basic biology and thought we now have the cell in a dish where we had either -- very similarly treated except
exhausted and then sort of rescued exhaustion, and so we began to map gene expression in these different populations and we saw that indeed they localize with the 28 zeta car, very different, and the bm modified gene expression but not back to these so there was a unique
subset that the bb provided that modulated exhaustion that gave us a clue about the biology and some were the usual candidates, so exhaustion associated genes were quantitatively diminished, and that was gratifying but perhaps not surprising but what was more surprising is when we
used unbiased analysis of the gene expression profile, don't look for something you already know has been reported in exhaustion but what is the most important distinction between the bb expressing rescued and 28 expressing exhausted, and we began to find gene families not
what you would expect, response to hypoxia, apoptosis didn't surprise me, but response to hypoxia as a mediator of exhaustion is not something that's been reported so this opens new opportunities to study biology of exhaustion, the metabolism is one we've been
quite interested in. mostly because early on we noticed that these cells looked i told you that even by day 7 when they were highly activated before that was antigen what adrienne noted, they always acidifided the media. they had a higher glycolytic
rate, and a significantly higher mitochondrial mass. so it was clear that these cells were metabolically deregulated and whether this was a system or cause of exhaustion is an area of important question to answer. we've gotten clues. we're thinking of ways similar
to the cancer biologist to modulate and to see whether it might repair some functionality of the cells and i can tell you that diminishing by providing lactose in this case can substantially improve cytokine production of this exhausted car so simply modulating metabolic
pathway may be sufficient to increase functionality of the t-cell and that same effect is not seen on the 19 car which is exhaustion suggesting this is an exhaustion-specific phenomena. cd-28 signaling exacerbates exhaustion and in contrast 41bb induces anti-exhaustion program
in chronically activated c tells explaining differential persistence for 19bb zeta cars compared to 19.28, we propose they are likely to be more effective against cancer and going forward we'll probably focus our efforts on bb cars, despite the vast amount of work
being done in human immunology i think we still have a pretty superficial understanding of t-cell exhaustion and this data provided us with novel insight about pathways. just to conclude, i hope i demonstrated to you that the bench to bedside to bench
approach is an efficient one for illuminating new biology, developing new therapy, and i think there's no better place to do this type of work than in the nih intramural research program with its laboratories and technological advances in synthetic biology are yielding
novel therapeutics, crafted in academia, you can make these things in your lab, but it's not just there are new therapeutics. these are also providing novel and elegant models for the study of biology. and i hope i've convinced you that we've come a long way since
i read that article in 1985, it's really an exciting time for the immunotherapy of cancer. i want to thank the people who contributed to this, clinical people in the pediatric oncology branch, an incredibly dedicated clinical team, melinda merchant, trey lee, terry fry and nerali
shaw who provides forth for patients and our research nurse, donna, amanda, cindy and bonnie who really make this stuff happen. sharon is our regulatory expert and joanne is our nurse practitioner. i mentioned adrienne's work.
karen, alec, jillian has done recent work on metabolism in the lab and waleed, and the support of the nih clinical center, medicine, these people absolutely work tirelessly to turn out these products with relatively limited resources, this program is led by dave and
jo and i am indebted to them. jim made the cd-19 car, steve provided the vector that we've used in our trials, and mary alice provides important monitoring for patients. i'll stop there and take questions. thank you.
[applause] >> yes, in pediatric pathology it's well known neural blastoma can spontaneously regress. how does that relate to any of this and wouldn't that shine light on how to best attack what we used to call the small blue cell group of tumors in general?
>> yeah, you know, that provocative observation plagued us. many investigators hypothesize that was immune mediated. i believe that most investigators now feel that that's a differentiation program.
when they regress what they usually do is differentiate, there's also differentiation to more mature elements, retinoids and epigenetics can cause them, but i would suggest the prevailing thinking is this is more of a cell intrinsic differential and not in the
interaction. >> thank you. >> do you know if the bb versus cd-28 is changing the pattern of expression on the cell surface? >> no, it absolutely does not. in fact, the tonic activation is the same, and the activation expression is the same.
and the warburg makes the media turn yellow but they are not as exhausted so it's providing some kind of protection against the exhaustion. >> the second thing, if you're willing to work on myeloid malignancies your situation of a hit and run is perfect.
>> you're absolutely right. >> the people that are trying to develop cars and il-3 receptor are worrisome in terms of even a few of these t-cells, if they stick around it will be for like. allo or auto. is there a way to predict this?
>> i don't know, cindy. you know, you're going to incorporate a lot of protections in it there, incorporate a vector. >> but you're never a hundred percent. >> i agree, hit and run of the 28 could be useful for an aml
car, absolutely. >> brilliant data. a lot of these patients with hemo malignancies coming here having received depleting chemotherapy, have you looked it the phoresis product, the t-cells in the phoresis product coming out more exhausted in a
patient who hasn't gotten prior chemo? >> a really good question. i got to believe that's playing out here somehow. it's actually remarkable to me we're even able to grow product from those of these patients and the fact that they are able to
go into the patient and expand, i'm flabberglasted because, you know, i spent the first years of my life showing how messed up their immune system is and yet put a gene in there and this thing becomes an army. i think there's probably a situation where you're beyond
the capacity to be able to rescue the cells. you know, it's surprising how effective they have been. >> the second question, i hate to monopolize, how would one sort of decide whether or not car therapy is a destination rather than a bridge to
transplant, what kind of trial would have to be designed to have that as an endpoint? >> oh, that's a really good question. my own feeling, it will be hard to answer that in children. certainly we're going to learn in the multi-transplanted
population first, each translant is less effective. it's already questionable whether we should be transplanting patients a second time that are put into remission. and certainly a third transplant, you know, come on,
it's the toxicity that's high. so i think we'll learn from those patients and the young adult population where the current way of treating young adults with all is not great and toxicity is less tolerable. it's a big challenge, probably will be done by clinical
researchers rather than people like me. >> thank you for the wonderful contribution that you've made to pediatric oncology, first of all, and your wonderful talk. i was wondering if you had any idea how the patients with refractory all did post
transplant, and then secondly in terms of the t-cell exhaustion whether that has anything at all to do with thymic involution. >> post transplant course of patients we believe to be -- the numbers are small but we haven't seen anything to make us think it's different than the usual.
one patient died. again, that's the kind of things that happens, this is the second transplant. i don't think it change the the outcome of the transplant. as far as exhaustion, this gets to what we were talking about that the more these cells have
been driven, the thymic cells are new again, they are the antithesis of exhaustion, so when we're relying on cells that haven't been through the thymus in the recent past, i'm talking number of divisions, they are more likely to be exhausted. if you come in already on the
verge of exhaustion, then you stick that car in there and you're tonically signaling, you'll get to the point of no return more quickly, so i think it is all wrapped up in the same package but the complexity we haven't been able to -- this is just an idea.
we haven't shown data. >> i have one additional question related to the earlier question about using the cd-19 car, primary treatment of leukemia in children. one of the points that you made is these are chemo resistant patients that had multiple
rounds of chemotherapy. given what you know about the homeostasis is that a necessary component or do you think that's unrelated? >> no, it's interesting. it could. it could. i don't know.
the primary refractories i guess are reassuring because those are -- have often only been guided, they have multiple rounds of chemo but they hadn't had years of chemo. i don't know, michael. it's a good question. >> and of course we're
interested in drug resistance, the cells themselves are changing because the resistance is recent, they have physiological changes that may be -- >> absolutely, yeah. they may be less fit. >> right.
absolutely. >> all right. if there are no further questions, let's thank dr. mackall for an amazing talk. and there will be a reception in the library, and you can ask more questions there.
0 comments:
Post a Comment