Redefining Patient Care in Acute Lymphoblastic Leukemia: The Impact of Novel Antibody and Immune Technologies on Clinical Outcomes

Transcription

1 CME Redefining Patient Care in Acute Lymphoblastic Leukemia: The Impact of Novel Antibody and Immune Technologies on Clinical Outcomes Course Director and Moderator Daniel J. DeAngelo, MD, PhD Harvard Medical School Dana-Farber Cancer Institute Boston, Massachusetts Faculty Stephan A. Grupp, MD, PhD University of Pennsylvania Perelman School of Medicine The Children's Hospital of Philadelphia Philadelphia, Pennsylvania Faculty Mark R. Litzow, MD Mayo Clinic Rochester, Minnesota Faculty Charles Mullighan, MBBS, MD St. Jude Children s Research Hospital Memphis, Tennessee What s Inside Welcome and Introduction: Defining the Future for ALL Treatment The Intersection of MRD, Modern Prognostic Assessment, and Innovative Treatment in ALL The Arrival of Antibody Therapy in ALL: From Mechanisms to Foundational Data A Look at the Future of Antibody-Based Combinations in ALL Exploring the Role of CAR-T Cell Therapy in a New ALL Landscape Symposium Summary and Audience Q&A This CME activity is jointly provided by Penn State College of Medicine and PVI, PeerView Institute for Medical Education. Participate in interactive questions, download activity slides, and obtain your instant CME credit online.

2 Activity Information Activity Description and Educational Objectives In this activity based on a live symposium held at the 2017 American Society of Hematology Annual Meeting and Exposition, experts in the management of acute lymphoblastic leukemia (ALL) offer insights on recent prognostic and therapeutic innovations, including updated evidence on the use of novel antibody technologies and chimeric antigen receptor (CAR)-T cell therapy. The panel also discusses real-world experience with these innovative treatments, and assesses a future in which sequential and combination approaches may offer further improvements in patient care. Upon completion of this activity, participants should be better able to: Describe the mechanisms of novel antibody-based therapies in ALL, including antibody-drug conjugates, bispecific agents, and other options Summarize efficacy evidence on the use of antibody- and immune-based therapy in ALL, including in relapsed/refractory disease or patients with persistent minimal residual disease (MRD) Recommend patient-appropriate antibody therapy for individuals with ALL who present with relapsed disease or persistent MRD Cite updated evidence on novel strategies combining antibody-based therapy with other treatment modalities in the ALL setting, including in newly diagnosed patients Manage treatment-related toxicity in patients with ALL receiving novel therapeutics, including antibody-based treatment Target Audience This activity has been designed to meet the educational needs of hematologistoncologists, hematologists, oncologists, advanced practice oncology nurses, and other healthcare professionals involved in the care of patients with ALL. Requirements for Successful Completion In order to receive credit, participants must view the activity and complete the post-test and evaluation form. There are no pre-requisites and there is no fee to participate in this activity or to receive CME credit. Statements of Credit are awarded upon successful completion of the post-test and evaluation form. Media: Enduring Material Release and Expiration Dates: January 11, January 10, 2019 Time to Complete: 120 minutes Faculty & Disclosure / Conflict of Interest Policy In accordance with ACCME requirements, Penn State College of Medicine has a conflict of interest policy that requires faculty to disclose relevant financial relationships related to the content of their presentations/materials. Any potential conflicts are resolved so that presentations are evidence-based and scientifically balanced. Course Director and Moderator Daniel J. DeAngelo, MD, PhD Director of Clinical and Translational Research, Adult Leukemia Institute Physician Associate Professor of Medicine, Harvard Medical School Dana-Farber Cancer Institute Boston, Massachusetts Daniel J. DeAngelo, MD, PhD, has a financial interest/relationship or affiliation in the form of: Consultant for Amgen Inc.; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Shire; and Takeda Pharmaceuticals U.S.A., Inc. Faculty Stephan A. Grupp, MD, PhD Novotny Professor of Pediatrics University of Pennsylvania Perelman School of Medicine Director, Cellular Therapy and Transplant Section Director, Cancer Immunotherapy Frontier Program CCCR Director of Translational Research The Children's Hospital of Philadelphia Philadelphia, Pennsylvania Stephan A. Grupp, MD, PhD, has a financial interest/relationship or affiliation in the form of: Consultant for Jazz Pharmaceuticals, Inc. and Novartis Pharmaceuticals Corporation. Grant/Research Support from Novartis Pharmaceuticals Corporation. Charles Mullighan, MBBS, MD Member, Department of Pathology Co-Leader, Hematological Malignancies Program St. Jude Children s Research Hospital Memphis, Tennessee Charles Mullighan, MBBS, MD, has a financial interest/relationship or affiliation in the form of: Grant/Research Support from Loxo Oncology. Speakers Bureau participant with Amgen Inc. Medical Director Carmine DeLuca PVI, PeerView Institute for Medical Education Carmine DeLuca has no financial interests/relationships or affiliations in relation to this activity. Other PVI staff who may potentially review content for this activity have disclosed no relevant financial relationships. Penn State College of Medicine staff and faculty involved in the development and review of this activity have disclosed no relevant financial relationships. Providership, Credit & Support This activity has been planned and implemented in accordance with the accreditation requirements and policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint providership of Penn State College of Medicine and PVI, PeerView Institute for Medical Education. Penn State College of Medicine is accredited by the ACCME to provide continuing medical education for physicians. The Penn State College of Medicine designates this enduring material for a maximum of 2.0 AMA PRA Category 1 Credits TM. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Information about CME credit for this activity is available by contacting Penn State at or ContinuingEd@hmc.psu.edu. Reference course # G T. Providership This CME activity is jointly provided by Penn State College of Medicine and PVI, PeerView Institute for Medical Education. Support This activity is supported by an educational grant from Pfizer. Disclosure of Unlabeled Use The faculty of this educational activity may include discussions of products or devices that are not currently labeled for use by the FDA. Faculty members have been advised to disclose to the audience any reference to an unlabeled or investigational use. No endorsement of unapproved products or uses is made or implied by coverage of these products or uses in our reports. No responsibility is taken for errors or omissions in reports. For approved prescribing information, please consult the manufacturer s product labeling. The materials presented here are used with the permission of the authors and/or other sources. These materials do not necessarily reflect the views of PeerView or any of its partners, providers, and/or supporters. About This CME Activity PVI, PeerView Institute for Medical Education, and Penn State College of Medicine are responsible for the selection of this activity s topics, the preparation of editorial content, and the distribution of this activity. Our activities may contain references to unapproved products or uses of these products in certain jurisdictions. The preparation of PeerView activities is supported by educational grants subject to written agreements that clearly stipulate and enforce the editorial independence of PVI and Penn State College of Medicine. Mark R. Litzow, MD Professor of Medicine Division of Hematology Mayo Clinic Rochester, Minnesota Mark R. Litzow, MD, has a financial interest/relationship or affiliation in the form of: Consultant for Amgen. Grant/Research Support from Amgen; Astellas Pharma US, Inc.; and Novartis Pharmaceuticals Corporation. 2 Go online to complete the post-test and evaluation for CME credit

3 Redefining Patient Care in Acute Lymphoblastic Leukemia: The Impact of Novel Antibody and Immune Technologies on Clinical Outcomes Welcome and Introduction: Defining the Future for ALL Treatment Daniel J. DeAngelo, MD, PhD Harvard Medical School Dana-Farber Cancer Institute Boston, Massachusetts Dr. DeAngelo: Good evening. My name is Dan DeAngelo. I come from Boston. I m part of the adult leukemia program at Dana- Farber Cancer Institute, Harvard Medical School. Welcome and thank you for attending this education symposium on redefining patient care in acute lymphoblastic leukemia. We re trying to focus on the impact of novel antibody immune technologies on clinical outcomes. of the reasons maybe why. CR rates are still high, but a larger percentage of patients will have relapsed or refractory disease. And then the treatment of these patients with relapsed disease is somewhat difficult. There are higher rates of relapses, as I alluded to. There s more adverse cytogenetics and one of the important aspects of adult ALL is higher persistence of MRD following induction and consolidation. In addition, with older patients, the tolerance of chemotherapy is less as compared with our pediatric colleagues, and therefore patients have trouble not only tolerating high-dose chemotherapy, but also in terms of maintaining their remissions. Long-term survival in patients with relapsed disease is still a very, very difficult process, and I ll review some of those data with you, with longterm survival in the 5% to 10% range. Moving Beyond (and Augmenting) Chemotherapy in ALL Blinatumomab (r/r B-ALL, adults/children) Joining me tonight is Dr. Stephan Grupp from the University of Pennsylvania Perelman School of Medicine and the Children s Hospital of Philadelphia in Pennsylvania, and Dr. Charles Mullighan from St. Jude Children s Research Hospital in Memphis, Tennessee. Who is not sitting at the table is Dr. Mark Litzow, who will join us in a few minutes from the Mayo Clinic in Rochester, Minnesota. Several recent approvals, including antibody-based approaches, CAR-T cell therapy Combination studies with antibodies underway Approved Phase 3 Phase 1/2 Inotuzumab ozogamicin (r/r B-ALL, adults) Tisagenlecleucel (CTL019) (refractory B-ALL, second relapse or later, patients up to 25 years) Rituximab (+ chemotherapy; GRAALL-R 2005) INO, blinatumomab combinations (+chemotherapy), epratuzumab Combotox, ofatumumab, denintuzumab mafodotin Other CAR-T cell constructs PD-1 inhibitors Real progress in ALL Progress in ALL But New Therapies Are Needed In pediatric patients 1 CR rate of 95%; 5-year EFS: 80%-85% In adult patients 2-5 CR rates of 85%, but 3-year DFS and OS <45% Higher rate of relapse, more adverse cytogenetics and However persistent MRD Poor tolerance of prolonged chemotherapy, modest remissions in relapsed/refractory disease Long-term survival <5% in patients who relapse while on therapy or with CR duration <24 months 1. Pui CH et al. N Engl J Med. 2004;350: Larson RA et al. Blood. 1998;92: Kantarjian H et al. Cancer. 2004;101: Fielding AK et al. Blood. 2007;109: Thomas DA et al. Cancer. 1999;86: There s been a lot of progress in acute lymphoblastic leukemia with many new strategies. Those of you who take care of patients with leukemia know there were seven new agents approved in myeloid and lymphoid leukemias. But there s been a lot of progress in ALL, focusing on that. In pediatric patients, really the hallmark of chemotherapy approaches, remission rates approach 95% with 5-year event-free survival far north of 80%. The adult patients don t fare as well and Dr. Mullighan will go over some So as I ve alluded to, there are a lot of new agents in the leukemia world. Focusing on acute lymphoblastic leukemia, three approvals. Blinatumomab, which had early approval 2 years ago, got full approval for relapsed/refractory B cell ALL, both in adults and children, as well as in Philadelphia-positive ALL. Inotuzumab ozogamicin was approved just a couple of months ago for adult patients with relapsed/refractory B cell ALL, both Philadelphiapositive and -negative. And I m not going to even try and pronounce the word tisagenlecleucel Steve will teach us how to pronounce it properly the CTL019 CAR-T cell, the first CAR-T cell approved in the United States and the only CAR-T cell approved for patients with refractory B cell ALL, in pediatric patients as well as young adults up to age 25. There are some ongoing phase 3 trials that will be reviewed, using the CD20 antibody with rituximab. I ll review some of the data with inotuzumab, blinatumomab, and then there are a variety of combination therapies with some immunoconjugates and immunotherapeutics that will be reviewed today. 3

4 Tonight s Agenda Prognostic assessment in ALL and the latest thoughts on the increasing importance of MRD monitoring Evidence on antibody strategies in ALL from mechanisms of action to important phase 3 data Antibody combinations is this the next step forward in ALL management? CAR-T cell therapy from recent approvals to the next wave of treatments The Intersection of MRD, Modern Prognostic Assessment, and Innovative Treatment in ALL Charles Mullighan, MBBS, MD St. Jude Children s Research Hospital Memphis, Tennessee So tonight s agenda is Dr. Mullighan will teach us about the prognostic assessment in ALL, focusing on MRD monitoring specifically after induction and consolidation. Then I ll take the podium and review some of the antibody strategies, at least single-agent antibody strategies, in acute lymphoblastic leukemia. Dr. Litzow will then focus on some of the combination strategies with the new immunoconjugates, and then ending the evening will be Stephan Grupp on CAR-T cell therapy. So now, without further ado, I introduce Dr. Mullighan from St. Jude. Thank you. Minimal Residual Disease Concept first described 40 years ago Main methods are flow cytometric detection of leukemic immunophenotype (LIP), detection of ALL fusion transcripts, and detection of antigen receptor rearrangements Commonly to 10-4 (1:10000 cells) Timing of testing varies widely Important interaction with leukemic subtype and genomic alterations Role of more sensitive tests, and with newer treatment approaches less clear Dr. Mullighan: Thank you very much, Dr. DeAngelo, and thank you for the opportunity to present in this symposium, very exciting session. So we ve never seen a more active time in the use of MRD, its wider deployment, and the development of new MRD monitoring strategies in the management of ALL, nor have we seen a more exciting time in our understanding of the genetic basis of this disease. And the main thing I wanted to try and convey to you tonight is how the use of MRD is intimately related to how we diagnose patients from a genetic level and perhaps stratify them for new treatment approaches. So the notion that leukemia could be present in patients who no longer had morphologic evidence of leukemia in their bone marrow was described over 40 years ago, and then shown to be of prognostic relevance. This has become a standard of care in the management of pediatric leukemia and has now been shown in many adult studies to also be of prognostic significance. There are three main methods for monitoring levels of MRD: use of flow cytometry to detect leukemic immunophenotypes, the detection of ALL fusion transcripts, and also the detection of antigen receptor rearrangements that are often clonal events in both B- and T-lineage ALL. These are commonly performed to a level of sensitivity of 10-4, or 1 in 10,000 cells in the bone marrow, below which we know MRD still exists and I ll come back to this later in the talk. But this is often used as a threshold to deem patients MRD undetectable, if not MRD negative. 4 Go online to complete the post-test and evaluation for CME credit

5 Now that said, there is a great deal of variability in how MRD essays are used. The timing of testing, whether single or multiple tests are performed, the nature of the test, and how that s integrated into modern genetic and diagnostic testing and that s a key theme that I ll go through tonight. There are also several more sensitive tests that are based on nextgeneration sequencing approaches, and even some newer tests that are on the horizon that have the potential to further improve our management of patients with ALL, and I ll touch on all of these things in my talk tonight. Importance of MRD as a Prognostic Factor: Childhood ALL Event-Free Survival for Day 46 MRD in Provisional SR Patients With Day 19 MRD 1%: St. Jude Total XV 1 Event-Free Survival by End Induction MRD (Day 29): COG AALL0232 HR B-ALL 2 The same is true for adult ALL, and here these are data from the GMALL group, the German groups, showing that from several studies, measurement of molecular markers or flow-based measurements staging the level of MRD are very powerfully predictive of the risk of treatment failure or long-term diseasefree survival. So that is to say, if we look at patient cohorts overall without any genetic stratification, MRD remains a powerful prognostic marker. MRD Study Study Time Site AIEOP-BFM ALL 2009 COG AALL08B1 Day 15 Day 33 Day 78 Day 8 Day 29 MRD Measurement and Stratification in Childhood ALL 1 BM BM BM Blood BM Standard Risk No HR factors and d33 and d78 MRD- by PCR or d15 <0.1% by FCM Low risk NCI SR; favorable genetics, no unfavorable factors, d8 <0.01%, day 29 <0.01% Standard Risk DCOG Day 33 BM No unfavorable factors, d33 ALL-10 Day 79 BM and d79 undetectable by PCR NCRI UKALL Day 29 BM Low risk 2011 Week 14 BM d29 <0.005% BM Day 15 Low Risk SJCRH Total BM Day 42 Favorable factors, d15 <1%, Therapy XVI BM if d42 Week 15 d42 <0.01% 0.01% 1. Campana D, Pui CH. Blood. 2017;129: MRD Stratification High Risk HR factors, or day 15 10% by FCM Medium Risk or day % by PCR Others or for B-ALL only, d % and d78 positive <0.05% by PCR Average Risk High Risk NCI SR; favorable genetics, no NCI SR; favorable genetics, no unfavorable factors, d8 0.01%, d29 unfavorable factors, d % <0.01% or neutral genetics, no or neutral genetics, no unfavorable unfavorable factors, d8 <1%; d29 factors, d8 1%, d29 <0.01% <0.01% NCI HR; d29 <0.01% High Risk Medium Risk Unfavorable factors, day %, Others day % Intermediate Risk High Risk Unfavorable factors or d29 <0.005% Wk % and wk 14 <0.5% Standard Risk Favorable factors, d15 1%, d42 High Risk <1% or unfavorable factors, d42 d42 1% or wk % <1% 1. Pui CH et al. Lancet Oncol. 2015;16: Borowitz MJ et al. Blood. 2015;126: So, first I d like to show you some recent data from both the pediatric and adult context, showing the ongoing importance of measurement of MRD as a prognostic tool despite contemporary management. So, on the left we have data from St. Jude Total XV, the last published completed study of TOTAL Therapy of patients with newly-diagnosed ALL, showing in the two measurements that are performed day 19 and then end of induction, day 46 that for patients who have positive levels of MRD at the earlier time point, measurement at the second time point is still a powerful prognostic predictor of outcome. Now, the Children s Oncology Group study of high-risk ALL, the 0232 study, MRDs are measured somewhat differently, at day 29, but there is still a very powerful association between the level of MRD and ultimate outcome in B-lineage leukemia. Importance of MRD as a Prognostic Factor in Adult ALL (GMALL 06/99 and 07/03 trials) 1 How is this being incorporated into current clinical care? So this is a busy slide. The details are not important, but I show the detail to illustrate the complexity of how different groups are approaching the integration of MRD and other risk markers to stratify and riskadapt therapy in ALL. So, these are approaches from five of the major pediatric oncology groups around the world, from the United States and Europe, showing that they use different approaches for measuring MRD, they use different time points, they use bone marrow or blood. They use a different variety of genetic factors, and then have a different number of risk classifications based on these variables. And so this remains a very fluid field as to how these various aspects are integrated. It s only going to become more complex as we consider some of the new subtypes of leukemia, which also have a close relationship with MRD levels. Total XV ( ): Risk Classification Schema 1 Provisional low-risk B-ALL with WBC <50x10 9 /L and age 1 to 9.9 years, DNA index 1.16, or TEL-AML1 Low risk Provisional low risk Provisional standard-risk T-ALL and other B-ALL Day 19 MRD <0.01% % 1% Provisional standard risk Provisional standard risk Provisional high-risk Ph+ ALL Day 46 MRD <0.01% 0.01% <1% 1% 1% <1% Low risk Low risk Standard risk High risk Standard risk High risk 1. Pui CH et al. Lancet Oncol. 2015;16: Brüggemann M, Kotrova M. Blood Adv. 2017;1:

6 So, here is one example. This is from my institution, St. Jude Children s Research Hospital, again TOTAL Therapy XV, to show that this is not just risk-stratified therapy based on initial features, such as the conventional prognostic markers of age, sex, white cell count, and favorable genetic features such as TEL-AML1, or highrisk features such as Ph-positive or MLL-rearranged leukemia. But it s risk-adapted therapy, meaning that after initial measurements of MRD early at day 19 and then subsequently at day 46, there s a provisional risk grouping and then a final risk grouping that s used to alter the intensity of therapy. ETV6-RUNX1 and Hyperdiploid ALL With D19 Negative MRD Candidates for Treatment Reduction 1 Leukemia Subtype No. a EFS, % Survival, % Relapse, % t(12;21)/(etv6-runx1) Hyperdiploidy > Other NCI standard-risk B-ALL t(1;19)/(tcf3-pbx1) T-ALL NCI high-risk B-ALL P =.023 P =.369 P =.017 a EFS rate, OS rate, and relapse risk reported as mean percentages with 95% confidence intervals; Total XV: Day 19 and Day 46 MRD and Outcome 1 10-Year EFS, % Group MRD 1% (n) MRD <1% (n) P Provisional low-risk 69.2 (36) 95.5 (244) <.001 Provisional standard risk 65.1 (56) 82.9 (142).008 Provisional SR Patients With Day 19 MRD <1% Provisional SR Patients With Day 19 MRD 1% 10 patients had missing MRD data on day 19 of remission induction. 1. Pui CH et al. Leukemia. 2017;34: If we then incorporate the measurement of MRD, the survival figures change. This identifies patients with negative levels of MRD in each of these subgroups and identifies groups of patients that have a very favorable outcome with at least a survival rate of 98% in some subgroups. And so, as we look forward, acknowledging the toxicity of the prolonged therapy of ALL, these groups then become candidates for treatment deintensification that may be curable with less intensive therapy that s given to standard-risk patients overall. 1. Pui CH et al. Lancet Oncol. 2015;16: There is a very strong relationship between these two measurements and outcome, as I showed before. So initially at day 19, these results are associated with outcome, but also the final risk group, based on both measurements of MRD, are strongly associated with outcome. Provisional Risk Total XV: Leukemia Subtypes Prognostic in MRD-Directed Treatment 1 Leukemia Subtype No. 10-Year EFS, % 10-Year Survival, % 10-Year Risk of Relapse, % A Patient Achieving MRD-Negative Remission (Total Therapy XV) 3-year-old boy presents with lethargy, bruising, and pancytopenia BMA 98% lymphoblasts, CD10+ CD19+ sig Cytogenetics: 27, X, +Y, +13, +18, +21 Total Therapy XV (prednisolone, vincristine, daunorubicin, L-asparaginase, HD-MTX) D19 MRD 3/10,000 cells (0.03%) D46 MRD undetectable Treated with intensive chemotherapy for 3 years, Remained MRD negative Remained in continued remission for >11 years t(12;21)/(etv6-runx1) Low risk Standard risk Hyperdiploidy > Other NCI standard-risk B-ALL t(1;19)/(tcf3-pbx1) T-ALL a 19.9 a NCI high-risk B-ALL a 85.8 a 15.3 a a Data contributing to the significant difference in the comparison in a given column. 1. Pui CH et al. Leukemia. 2017;34: There is also an interaction between genetic subtype and the level of MRD. So this lists some of the key subtypes we see in childhood leukemia. Low-risk subtypes, TEL-AML1 or ETV6-RUNX1, high hyperdiploidy and other risk groups, T-ALL and high-risk subtypes of leukemia, which overall have an inferior outcome. And this is showing the results irrespective of the level of measurement of MRD. You can see the favorable or less favorable survival rates associated with each of these groups. There is a strong interaction, as I ve alluded to, between MRD and genetic risk group, and this also applies to high-risk subtypes of leukemia. And so here I show my first case as one example. This is a 3-year-old boy that presented with classic features of leukemia, features of pancytopenia. He s diagnosed with otherwise conventional features of leukemia, with replacement of the bone marrow by lymphoblasts, of conventional pre-b immunophenotype, and is shown to have a grossly aneuploid karyotype with near-haploidy, with loss of most but not all chromosomes. So this patient was enrolled on TOTAL Therapy XV and was treated with conventional therapy initially in the high-risk arm, was found to have at day-19 low levels but detectable MRD, and at day 46 had undetectable levels of MRD. So he was treated with intensive chemotherapy for almost 3 years. He was not transplanted and has remained in continued remission for 11 years. 6 Go online to complete the post-test and evaluation for CME credit

7 Outcome of Hypodiploid ALL Depends on MRD 1,2 1. Holmfeldt L et al. Nat Genet. 2013;45: Mullighan CG et al. Blood. 2015;126: So those who are familiar with subtypes of leukemia would acknowledge that this is near-haploid ALL. This is otherwise a high-risk form of leukemia that comes in two main forms, near-haploid ALL with gross aneuploidy and losses of most chromosomes, and low hypodiploid ALL with less severe aneuploidy, which importantly is associated with germline p53 mutations, which is another important diagnostic consideration when this karyotype is identified. There is a strong interaction in St. Jude TOTAL Therapy studies between MRD response and outcome. So, hypodiploid ALL is classically considered a very high-risk subtype of leukemia. But in recently published data, we ve showed that patients who have a good MRD response with remission induction therapy do not have an inferior outcome, somewhat inferior to lowrisk ALL, but compared to other patients with this form of the disease, a remarkably favorable outcome. So the point here is that not all high-risk leukemias should be deemed high-risk, and measurement of MRD is important. Current Genomic Classification of B-ALL 1 >2500 ALL cases: SJCRH, COG/TARGET, MDACC, Alliance, ECOG, SWOG, NILG 100% 90% Ph-like 80% Ph+ 70% Other 60% 50% Hypodiploid 40% ZNF384 30% MEF2D 20% DUX4/ERG 10% MLL 0% 0 TCF3-PBX1 ETV6-RUNX1 Hyperdiploid Patients, % 1. Iacobucci I, Mullighan CG. J Clin Oncol. 2017;35: So now if we move to where we are at the moment and we consider the current genomic classification of B-ALL, where are we? Well, this remains a work in progress, but through collaborative efforts between St. Jude, the COG, and the adult cooperative groups, several thousand ALL cases have been studied and we ve extended the observation that genetic subtype is intimately related to age of diagnosis. So in standard-risk childhood ALL, we find over 50% of patients have good-risk subtypes such as hypodiploidy and ETV6-RUNX1. But with increasing age, we see two things: we see an increasing frequency of what we can term kinase-driven leukemia, Phpositive leukemia and Ph-like leukemia, comprising over 50% of adults. And we also see the emergence of new subtypes that were not previously recognized because they re genetically diverse. They have multiple fusions, not just a single rearrangement, and they re often cryptic on cytogenetic analysis and they d not been identified until we started doing agnostic genome sequencing approaches. A Patient With Ph-Negative B-ALL With Elevated MRD 10-year-old male with 3 weeks of fever and fatigue Pale, lymphadenopathy, and splenomegaly Pancytopenic, BMA 89%, lymphoblasts CD34+ CD10+ CD19+, sig, CSF Treatment: Prednisolone, vincristine, daunorubicin, PEG-asparaginase, IT AraC BMA d8, 22, 29 persisting leukemia Cytogenetics: del5(q33q33) FISH: rearrangement of PDGFRB RT-PCR: EBF1-PDGFRB1 Imatinib 260 mg/m 2 /d Rapidly achieved MRD- remission And I want to show you a couple of examples. I can t go through all of them, but a couple of examples that are clinically relevant and also their association with measurements of MRD. So, my second case is a patient, a 10-year-old boy who presented with a 3-week history of fever and fatigue, pale with lymphadenopathy and splenomegaly, again, with bone marrow replacement by lymphoblasts of conventional pre-b immunophenotype. He was treated with a COG-STAR regimen. He had persisting MRD right through remission induction therapy. And on cytogenetics was noted to have a focal deletion involving 5q, an interstitial deletion, and on FISH analysis was shown to have rearrangement of PDGF receptor-β, and was shown to have a fusion involving EBF1 to PDGF receptor-β, which we now understand to be one of the recurring fusions that is a hallmark of Ph-like or BCR-ABL-like acute lymphoblastic leukemia. So, this patient had the addition of imatinib to a fairly basic backbone of chemotherapy, steroids, and vincristine and achieved very rapidly an MRD-negative remission, which persisted for several years. 7

8 Y Y Y Y Y Redefining Patient Care in Acute Lymphoblastic Leukemia: The Impact of Novel Antibody and Immune Technologies on Clinical Outcomes Gene expression profile similar to Ph+ ALL Frequent alteration of IKZF1 (Ikaros) Ph-Like ALL 1-4 pathways that are also vulnerable to currently available targeted agents. Not only do these lesions activate the canonical signaling pathway, an important point is that they often activate parallel signaling pathway, and so additional targeted approaches are potentially useful in these forms of leukemia. Trials and New Agents in Ph-Like ALL 1. den Boer ML et al. Lancet Oncol. 2009;10: Mullighan CG et al. N Engl J Med. 2009;360: Roberts KG et al. N Engl J Med. 2014;371: Roberts KG et al. J Clin Oncol. 2017;35: So Ph-like leukemia has become a topic of great interest. Several groups, ours and the Dutch, have shown that this is a group of patients that lack a single lesion, but they have a gene expression profile very similar to that of Ph-positive leukemia, hence the name. And in large-scale studies that we ve now performed looking right across the age spectrum, looking at over 1,000 adults now, we show that the prevalence of this form of leukemia, based on its gene expression profile, rises to over 20% from adolescence and remains that prevalent right throughout age. And in all studies so far, this form of leukemia is associated with high-risk features, elevated MRD, and/or poor outcome, and in many studies, survival that is as poor or inferior to other high-risk subtypes of leukemia. It is perhaps the most genetically diverse form of any form of acute leukemia. We ve now found over 60 different rearrangements involving receptor tyrosine kinases and nonreceptor tyrosine kinases. These commonly deregulate kinases or receptors by juxtaposing to enhancers that result in chimeric fusions, which result in constitutively active kinases. Or there are non-fusion events like deletions and mutations that can also activate JAK-STAT signaling. 1. Deregulation of receptors/kinases by strong enhancers 2. Chimeric kinase fusions 3. Deletions/mutations activating signaling Signaling Pathways in Ph-Like ALL TSLP CRLF2 IL-7Rα EPO CRLF2 IL-7Rα ins F232C Truncated EPOR Y Y KD KD TYK2 JAK2 KD KD KD CSF1R ABL1/2 PDGFRB/A ITD ZMYM2 ETV6 TMEM2 DTT2 FLT3 KD KD FLT3 KD KD KD NTRK3 PTK2B BLNK COG AALL1131: Childhood Ph-like ALL with ABL class fusions treated with dasatinib COG AALL1521: Childhood CRLF2r/JAK-STAT mutated Ph-like ALL treated with ruxolitinib + chemotherapy 1. Tasian SK et al. Blood. 2017;129: Roberts KG et al. Blood Adv. 2017;1: St. Jude Total Therapy XVII: Dasatinib for ABL class lesions by day 15 Ruxolitinib for MRD+ JAK-STAT B-ALL and T-ALL by day 42 M (VENALL): Venetoclax + navitoclax in relapsed/refractory ALL Ruxolitinib + PI3K Inhibition 1 TKI + Venetoclax 2 So, where are we now with treating Ph-like leukemia? I'll show just a couple of examples. The COG has opened and actually had advanced enrollment on the two arms targeting Ph-like ALL with ABL fusions with dasatinib and JAK-STAT activated Ph-like leukemia with ruxolitinib and chemotherapy. St. Jude has opened TOTAL Therapy XVII, which is adding both dasatinib and ruxolitinib to these lesions. There are actually efforts as well to look at the potential for additional targeted agents. So, these are data from CHOP from Sarah Tasian and colleagues showing the potential for a synergistic therapy with PI3 kinase inhibition. And we and others have shown synergy between BCL2 inhibitors, such as venetoclax, with TKI therapy in Ph-like leukemia. So these agents are starting to emerge into clinical trials as well, such as a study using venetoclax and navitoclax in relapsed and refractory ALL. No Significant Difference in EFS Between Ph-Like and Other B-ALL With MRD-Directed Treatment 1 Total Therapy XV Study JAK-STAT Signaling ABL Signaling JAKi BCL2i STAT ABLi STAT PI3Ki PI3K/ mtori mtor Transcription X IKZF1/EBF1/ PAX5/CDKN2A/B Other Signaling MAPK/ ERK STAT PI3K/ mtor TRKi FLT3i MEKi JAKi FAKi PI3Ki mtori For reasons of time, I ll just summarize this by saying that despite the diversity, they converge on a relatively limited number of signaling pathways, most commonly the JAK-STAT signaling pathway, also relatively commonly ABL-class signaling, with ABL itself, PDGF receptor-β, and other kinases that are sensitive to imatinib and dasatinib and similar agents, and then less common rearrangements and mutations that activate additional signaling 1. Roberts KG et al. J Clin Oncol. 2014;32: An important point again is that if we have optimal MRD-adapted therapy with optimal supportive care, the outcome is not always inferior. These are data from TOTAL Therapy XV showing that the overall survival for Ph-like leukemia was not statistically inferior to 8 Go online to complete the post-test and evaluation for CME credit

9 non Ph-like B-lineage ALL. However, these cases were associated with elevated MRD levels, reclassification as high-risk leukemia, and often direction to intensive approaches such as allogeneic stem cell transplantation that could potentially at least have been avoided with the use of TKI therapy which we know many of these patients had lesions that were sensitive to TKI agents. Let me show you just two additional new subtypes of leukemia before moving on to some other data in newer agents. So here s one subtype of leukemia that s another example of a very homogeneous form of leukemia based on gene expression profile, but a diverse range of fusion partners that activate one downstream pathway. This is MEF2D-rearranged leukemia that s seen particularly in adolescents and adults. It s associated with intermediate to poor outcome. And it also has a therapeutic vulnerability because these fusions activate histone deacetylases and increase their expression, and the leukemic cells are then exquisitely sensitive to HDAC inhibitors, which are shown here for xenografted cells of leukemia, showing remarkable sensitivity to agents such as panobinostat. A third example I ll show is a subtype called ZNF384-rearranged B-ALL because it s really changing how we think about classification of leukemia overall. So, this comprises about 5% of B-ALL cases. There are now seven different partners that we know are rearranged to this gene, ZNF384. And interestingly, these cases, some of which had been described some years ago, often have aberrancy. They showed myeloid features occasionally. They ve been seen in case reports of mixedphenotype leukemia. And during therapy when these cases relapse, they can sometimes undergo lineage switch to AML or early T cell precursor leukemia. The gene expression profile is very distinctive and also distinctive because there's deregulation of a number of hematopoietic transcription factors, not just lymphoid transcription factors as well. So, at one level this is important because it s a high-risk leukemia in the B cell repertoire, and it does define a subset of B cell leukemia. Biphenotypic/Bilineal Leukemia 1 But it also defines a subset of mixed-phenotype acute leukemia. So, this is another study that is not yet published, and you d be familiar with cases that have multiple immunophenotypic clones within one patient or a single clone, a biphenotypic clone that expresses T and myeloid markers, or B and myeloid markers. We now know that the B myeloid cases, over half of these also have these rearrangements of ZNF384, exactly the same rearrangements that we see in B-ALL. Many of these patients have multiple immunophenotypic clones, but we ve looked in great detail and showed that the individual clones within a patient are basically genetically identical but they re not varying because of different mutations within the subclones. And if we look by gene expression, here we re looking at all cases of leukemia, not just B-ALL mixed phenotype leukemia, the ZNF384 rearranged cases within this red box show tight clustering of both the B-ALL cases and the mixed-phenotype acute leukemia cases with this same fusion. So, this really identifies a singular subtype of leukemia that can have lineage aberrancy. It can be diagnosed as B-ALL or mixed phenotype leukemia. And our suggestion would be that when you identify these cases, they should be considered one subtype, not divided as B-ALL or mixed phenotype leukemia depending on their initial immunophenotype. Secondary Events and MRD: IKZF1 1 Transcription factor required for lymphoid lineage commitment Deleted/mutated in ~15% of B-ALL, and >70% BCR-ABL1 ALL, Ph-like ALL Associated with poor outcome in era of TKI therapy Overall Survival 1. Slayton WB et al. Submitted. Event-Free Survival AALL0622 Dasatinib + chemotherapy So those are some comments about new subtypes and how they interact with MRD. What about secondary genetic events? Well, genetic profiling of ALL has identified a number of secondary lesions that we know are very important for leukemogenesis, and several of them are strongly associated with outcome also. CD3 CD19 1. Alexander and Mullighan, under review. MPO MPO The most widely studied and most powerful prognostic marker are alterations of IKZF1, or Ikaros, a lymphoid transcription factor, a very early-acting lymphoid transcription factor. These alterations, either deletions or mutations, are highly prevalent in Ph-positive leukemia and Ph-like leukemia. And looking at these data here from childhood Ph-positive leukemia from the COG study, which is using dasatinib plus chemotherapy, we still see a powerful association between the 9

10 presence of Ikaros alterations and poor outcome, even in the presence of optimal chemotherapy and targeted agent usage. MRD-L Combined Use of MRD and IKZF1 to Predict Outcome 1 MRD-M Favorable Outcome in DUX4/ERG ALL Despite IKZF1 Deletion Relapse-Free Survival 1 Event-Free Survival 2 MRD-H No. of Nonrelapse Relapse Variable P Patients n (%) n (%) All patients (81.7) 24 (18.3) MRD class <.001 MRD-L (29.9) 2 (8.3) MRD-M (65.4) 11(45.8) MRD-H 16 5 (4.7) 11 (45.8) Ikaros alteration status <.001 Alteration 20 7 (6.5) 13 (54.2) Wild type (93.5) 11 (45.8) 1. Slayton WB et al. Submitted. It s also been noted that if one combines measurements of MRD and Ikaros-genotyping, one can create a combined predictor that is more powerful at predicting relapse than measurement of either MRD or Ikaros status alone. And so these are data from the DCOG study showing the powerful association of poor outcome in patients that are both MRD positive and have positive Ikaros alterations. And for this reason, many groups are incorporating Ikaros mutation testing into upfront diagnostic testing at the time of initial diagnosis. EFS, % Benefit of Steroid Pulses in IKZF1-Mutated ALL EORTC DFS, % EFS Proportion i-bfm-95 2 IKZF1del EFS (5 y) 1. Harvey RC et al. Blood. 2010;116: Clappier E et al. Leukemia. 2014;28; An important note, though, is that not all Ikaros-mutated cases have poor outcome. So, yes, in the context of Ph-positive and Ph-like leukemia, Ikaros mutations are powerfully prognostic. There is one other subtype in which Ikaros alterations are prevalent and that is cases with rearrangements of a gene called DUX4. These cases also have deletion of gene called ERG. This is actually a subtype of leukemia that has very good outcome. This is previously known as R6 ALL in the COG studies, the most favorable prognostic group by gene expression profiling. Data from many groups here. The French have shown that in this group of patients, even if you have an Ikaros alteration, your outcome is good. So this is just, if you like, a sidebar to show that Ikaros alterations by themselves are not enough to predict poor outcome. One does need to consider other founding alterations to fully understand the prognostic effect of these lesions. New Agents: Rituximab (CD20) OS, % OS, % Cum. Incidence 1. Clappier E et al. Leukemia. 2015;29: Hinze L et al. Leukemia. 2017;31: IKZF1del CIR GMALL: 117 chemo+r vs 70 chemo-r (8 doses) SR: faster decrease in MRD at day 21 (60% vs 19%) and week 16 (89% vs 16%), higher CCR (64% vs 48% at 3 years) HR: 3-year CCR 64% vs 48%; post HSCT 75% vs 40% OS GRAALL-2005: Adults 18-59, 105 R+, 104 R (16 doses) No difference in MRD responses 1 Event-Free Survival Cumulative Incidence of Relapse There are some data to support that this approach might be useful. So data from the EORTC study in Europe have shown that the addition of steroid pulses in Ikaros-mutated cases can rescue the poor outcome of cases that have Ikaros alterations, shifting this curve for disease-free survival. But this has not been uniformly seen in the i-bfm-95 study. While Ikaros alterations were associated with poor outcome, the intensity of steroid therapy was not associated with poor response. However, many groups are still noting that these Ikaros alterations are associated with poor outcome and using them accordingly. 1. Maury S et al. N Engl J Med. 2016;375: So what about the use of new agents in the treatment of ALL, and you ll hear more about various of these approaches in the subsequent talks. But just a couple of comments about outcome and the use of MRD. So, firstly looking at rituximab, which is increasingly becoming a standard of care for two reasons, it can improve outcome and also potentially reduce the risk of allergy to asparaginase in patients that receive that agent. So, several studies have shown that this agent can improve outcome. 10 Go online to complete the post-test and evaluation for CME credit

11 In the GMALL study, the German study, for example, they used eight doses of rituximab. It was shown that a rapid decrease in MRD was associated with a more favorable outcome. MRD Sequencing Analysis and Comparison of Results 1 In the GRAALL-2005, the French-led study, this also showed an improved outcome with use of rituximab, but no association with MRD, so more studies need to be done to understand the relationship of MRD and the use of this agent. New Targets: CD19 and CD22 Blinatumomab Morphologic response in 43%-69%, 76%-88% MRD negative MRD negativity predicts longer survival CAR-T cell therapy CR in 60%-90% with high rates of MRD negativity CD22: inotuzumab ozogamicin CR 58%-81%, MRD <0.01% in 72%-78% 1. Faham M et al. Blood. 2012;120: Now, what about new approaches to measure MRD? In just the last couple of slides I want to refer to these. So, the field is increasingly turning to sequencing-based approaches to detect MRD, and this is built on the principle that most patients with ALL have one or more clones. With a clonal antigen receptor rearrangement, B cell or T cell, one can design PCR primers specific for a patient and then use next-generation sequencing approaches to monitor the depth of response. What about other new agents? So Dr. DeAngelo has mentioned some of these, blinatumomab, CAR-T cell therapy and inotuzumab ozogamicin, each of which have had profound benefit to patients with relapsed and refractory disease. There is some evidence that the use of MRD can identify patients that will have a more salutary response to the use of these agents. But again, more work needs to be done; there are not so extensive data as in the context of chemotherapy. Role of MRD With Novel Therapy So these are some of the first data showing that there was generally good concordance between flow-based measurement of MRD and sequencing-based measurement of MRD. But one could detect patients that had levels of MRD detectable by sequencing that were not detectable by flow, and that s not surprising given the potential for these agents to go deeper than conventional MRD-based monitoring approaches. Discordant Flow and Sequencing MRD Has Intermediate Outcome 1 Relapse rates higher than with conventional chemotherapy MRD negativity necessary but not sufficient for durable response Long term outcomes influenced by Rapidity and depth of MRD response Subsequent therapy (eg, HSCT) Salvage status 1. Wood B et al. Unpublished. These novel agents, we typically see at least at the moment that some of them are associated with higher relapse rates than with conventional chemotherapy for frontline-treated ALL. And it may be that MRD alone is not enough and that more subtle measurements, such as the rapidity of MRD response, the depth of MRD response, as well as the type of subsequent therapy, such as stem cell transplantation, and the stage of the patient at which these therapies were used are very important. What is not clear is the prognostic power of detecting flownegative, sequencing-positive MRD. So, these are some unpublished data from Brent Wood in the Children s Oncology Group, suggesting that flow-negative, sequencing-positive cases have an intermediate outcome. So, the top line here are cases that are concordant, negative for both. These are cases that are positive for both. But these are cases that are flow-negative and sequencing-positive, at least to the level of 1 in 10-4 for sequencing. So, this detection may predict a group of patients that have intermediate outcome. 11

12 But from these data, it indicates that the level of detection is important. If one uses a more sensitive threshold, this predictive power is lost. So, the bottom line here is that we will see these approaches being used more extensively, but more data are required. Clonal Evolution in ALL Total Therapy XVII: Precision Medicine of Childhood ALL at St. Jude Gene sequencing of somatic (leukemia) samples Subgroups of ALL Identification of lesions and targeting therapy eg, Ph-positive, Ph-like ALL Sequence-based sensitive MRD: identify high risk Gene sequencing of germline (skin or blood) samples Identify leukemia predisposition Pharmacogenomics-based dose adjustment 6-MP: TPMT, NUDT15 VCR: CEP72 So finally, what about a term that I like to call precision MRD? So now that we ve done a lot of sequencing, not only at diagnosis and at relapse, we ve been able to map out the clonal structure of why patients relapse from a genetic viewpoint and the mutations that we see emerging over time. And I conclude with just showing some of the approaches that we re using now at St. Jude to guide the use of genomics, MRD, and precision medicine. We re sequencing all patients at diagnosis. We re targeting patients, not only with TKIs, but other approaches as well when these lesions are present. We re also using sequencing to identify patients with germline predisposition, and also using genomics to identify patients that might have toxicities of therapy or poor drug responsiveness. And I think, increasingly, this is the way the field will be moving both in children and adults. We now know that some of these mutations that are enriched at treatment failure are associated with resistance that s highly drug selective, CREBBP mutations associated with steroid resistance, NT5C2 mutations associated with thiopurine resistance. So the question is, like the sequencing-based approaches, if we could come in early or during therapy and detect these low-level clones with these point mutations before they emerge, would that allow us to adjust our therapy and perhaps adjust the nature of the agents or the conventional targeted approaches that we might use? MRD-Level Detection of Resistance Mutations in ALL 1 Precision resistance : CREBBP and glucocorticoids, NT5C2 and thiopurines Can we sensitively detect resistance mutations? 1. Faham M et al. Blood. 2012;120: Well, these are some unpublished data soon to be published showing that we can; we can actually, using focus mutation specific approaches, get to MRD-appropriate levels of detection of specific relapse-associated point mutations. And so, this is an approach that we and others are pursuing as a research aim in our front-line studies of ALL. 12 Go online to complete the post-test and evaluation for CME credit

13 The Arrival of Antibody Therapy in ALL: From Mechanisms to Foundational Data Age-Related Survival in ALL 1 Children: 83% 5-year EFS 1. Larson RA et al. Blood. 1998;92: Daniel J. DeAngelo, MD, PhD Harvard Medical School Dana-Farber Cancer Institute Boston, Massachusetts Adults: 40% 3-year DFS Dr. DeAngelo: That was a fantastic review. So now my focus over the next discussion is going to be on some of the immunotherapy strategies, really the newer antibodies that have just been approved over the last year or two. So this just reviews some of the outcomes Charles touched on this a little bit. Our pediatric colleagues have been really heroic in improving the outcomes of acute lymphoblastic leukemia, with 5-year event-free survivals that are approaching 90%, whereas our adults unfortunately have really lagged behind. Outcome for 609 Adults With Relapsed ALL: MRC UKALL2/ECOG2993 Study 1 Outcome of Patients After First Relapse 5-year OS: 7% So this was the large ECOG 2993/UKALL2 study that randomized patients in an intent-to-treat fashion between transplant and chemotherapy in first remission. And then the question that was asked that was a whole separate issue, but the question that was asked and the focus of this slide is what happened to those 600 patients or so that relapsed on that study. Well, these are the data. Many of those patients were able to get into remission, but regardless of how they were consolidated after remission, with either an allogeneic approach, an autologous approach, or just more chemotherapy, the 5-year survival was really dismal, less than 10%. So clearly, new strategies are absolutely required. Monoclonal Antibodies and Their Targets in ALL Antigen Target CD19 CD20 CD22 CD52 Antibodies Blinatumomab SGN-CD19a SAR3419 Combotox Rituximab Ofatumumab Epratuxumab Inotuzumab Combotox BL22, HA22 Alemtuzumab This leads us into the monoclonal antibodies, and there are a variety of different classifications of different targets. We re not going to have time to review all. There's CD19-based, CD20-based, CD22, and CD52. For time purposes, I m going to focus on some of the CD19 strategies, monoclonal antibody SGN19, as well as the bispecific antibody. Then I ll focus on inotuzumab in the CD22 venue. Antibody-Drug Conjugate: Denintuzumab Mafodotin (SGN-CD19a) 1 SGN-CD19A Proposed Mechanism of Action 1. Fathi A et al. 57th Annual Meeting of the American Society of Hematology (ASH 2015). Abstract Fielding AK et al. Blood. 2007;109: One of the issues I always like to show this slide before I discuss the relapsed/refractory, because it really sets the stage for how important novel therapeutics are in the relapsed/refractory setting, at least in adult patients. 13

14 So let s start with CD19. Denintuzumab mafodotin, or SGN- CD19a as it has also been called, is a monoclonal antibody drug conjugate, where the monoclonal antibody binds to CD19 and has an ability to target and traffic the chemotherapeutic agent to the target tumor cell. And then through this mechanism of internalization and then acid lysosome release and degradation of the antibody-drug conjugate, we re able to get the drug into the cell and hopefully cause cell-cycle arrest and apoptosis. Antibody-Drug Conjugate: Denintuzumab Mafodotin (SGN-CD19a) (Cont d) 1 Overall Survival: B-ALL Patients Median OS = 18 weeks (95% CI, 12-28) Antibody-Drug Conjugate: Denintuzumab Mafodotin (SGN-CD19a) (Cont d) 1 Parallel ALL and NHL study in both adults and pediatric patients The study is evaluating 2 schedules of IV SGN-CD19a administration Weekly (days 1 and 8 of 21-day cycles; mg/kg), or Every 3 weeks (0.5-6 mg/kg) 1. Fathi A et al. ASH Abstract And here s the overall survival of this antibody drug conjugate. It s not really clear what Seattle Genetics has in mind for this agent but this is an isolated monoclonal antibody drug conjugate. Bispecific Antibodies Work in Several Different Ways 1 1. Fathi A et al. ASH Abstract So, this agent was done in both pediatrics and adults, both ALL and non-hodgkin s lymphoma on both a weekly phase 1 study as well as an every-3-week schedule. BiTE antibodies are furthest along in clinical development in patients with ALL Antibody-Drug Conjugate: Denintuzumab Mafodotin (SGN-CD19a) (Cont d) 1 Composite complete remission (CRc) rate (CR + CRp + CRi) Schedule A: 19% (95% CI, 7-36) Schedule B: 38% (95% CI, 16-55) In the subset of patients with Ph+ B-ALL, 50% (5/10 patients) achieved CR and 1 patient had PR In patients with CRc and available local MRD assessments, 62% (8/13 patients) were MRD negative MTD not established Ocular toxicity: may be dose limiting Steroid eye drops for prophylaxis 1. Kontermann RE, Brinkmann U. Drug Discov Today. 2015;20: Then moving onto the bispecific antibodies, most of you know about blinatumomab. But the idea is to really bring the patient s own T cells, or if the patient had a prior transplant, the donor s T cells to the tumor site, that is the lymphoblast. And these are furthest along in the development in patients with ALL, although there s a lot of strategies using this technology in other diseases. But obviously, we re going to focus on lymphoblastic leukemia. 1. Fathi A et al. ASH Abstract Complete remissions were seen in about 20% on the weekly schedule and 38% on the 3-week schedule, including patients with Philadelphia-positive B cell ALL. Actually, the response rate was a little bit higher there, 50% achieved a first remission. In patients with cytogenetic remission, available local MRD assessments were that these patients had an MRD-negative rate of 62%. That was mostly assayed by flow, but it was based on local MRD. The maximum tolerated dose was not established, but there was a particular peculiar toxicity with some keratitis that was seen early on that was really mitigated with steroid eye drops. Mode of Action of BiTE Antibody Blinatumomab 1 1. Bargou R et al. Science. 2008:321; Blinatumomab: A BiTE antibody designed to direct cytotoxic T cells to CD19-expressing cancer cells Approved for use in relapsed/refractory Ph B-ALL 14 Go online to complete the post-test and evaluation for CME credit

15 So, this cartoon shows the characterization of this agent. It took about a decade or so with a German company, Micromet, to really establish the linker phenomenon of taking the antibody part of CD3 against the T cell and the antibody part of CD19 against the tumor cell, and then bridging it together, kind of the Goldilocks phenomenon. If the linker is too big, you re not going to get activation of the T cell. If it s too short, you re going to get steric hindrance, so you need to find the right fit. And what happens is these T cells are activated, they re allowed to proliferate, and then there s redirected cell lysis against the tumor cell. Screening and enrolment Phase 2 Study of Blinatumomab in Adults With Ph-Negative R/R ALL 1 Blinatumomab civ 28 mcg/day a N = weeks on, 2 weeks off Up to 2 cycles Primary endpoint assessment Blinatumomab civ 28 mcg/day 4 weeks on, 2 weeks off Up to 3 cycles Consolidation in case of CR/CRh AlloHSCT 100-day TRM assessment Prephase dexamethasone in patients with >50% peripheral blasts or elevated LDH; prophylactic DEX given to all patients within 1 hour of treatment start Follow-up Primary endpoint: CR/CRh b within 2 cycles Secondary endpoints: RFS and OS AlloHSCT realization 100-day TRM after allohsct Incidence and severity of AEs Exploratory endpoint: MRD response within 2 cycles Patients with relapsed/refractory ALL N = 405 Induction (2 cycles) If 5% blasts Consolidation (3 cycles) If 5% blasts Maintenance (up to 12 mo) Follow-up 1. Kantarjian H et al. N Engl J Med. 2017;376: Phase 3 TOWER Study: Randomization and Dosing 1 Randomization 2:1 (blinatumomab:soc) Stratified by age, prior salvage, and prior allohsct Blinatumomab Continuous infusion 4 wk on, 2 wk off; 9 mcg/d for 7 d, then 28 mcg/d wk 2-4 Continuous infusion 4 wk on, 8 wk off; 28 mcg/d SOC chemotherapy Investigator s choice: FLAG ± anthracycline, HiDAC-based, high-dose MTX based, or clofarabine-based But moving onto the TOWER study, which was led to full approval; it was published in The New England Journal of Medicine this year by Dr. Hagop Kantarjian. This took patients with morphologic relapse, building off the phase 2 study that led to early approval a couple of years ago. Four hundred patients with relapsed/ refractory B cell ALL were enrolled on a two-to-one randomization between blinatumomab versus investigator s choice of high-dose AraC-containing regimen. a 9 mcg/day on days 1-7 of cycle 1. b Defined as <10-4 detectable blasts by central laboratory Q-PCR of Ig/TCR rearrangements. 1. Topp MS et al. Lancet Oncol. 2015;16: So, there are a lot of studies that have been developed. So, this is one of the approaches and really leading on the segue from Charles, looking at the use of blinatumomab in patients who are MRD positive. Now, MRD has not, unfortunately, taken hold in the adult population or the adult physicians as it has in our pediatric colleagues, as Charles has nicely demonstrated. So, this was done in Europe focusing on patients with MRD-positive disease, and patients could receive up to two cycles. Blinatumomab in Relapsed/Refractory ALL: Response patients treated with blinatumomab 28 mcg/d civ x 4 wk every 6 wk Response n (%) CR 63 (33) CRh 18 (10) CR + CRh 81 (43) No marrow blasts 17 (9) Median OS: 5.9 mo; median RFS: 6.1 mo Toxicities: CNS 1. Topp MS et al. Lancet Oncol. 2015;16: And you were able to demonstrate that about 60% of patients were able to get a complete remission, 81% if you add incomplete hematologic recovery. Patients could get up to two cycles of chemotherapy, followed by up to three cycles of consolidation. And then there was an option to get maintenance therapy. And this was in lieu of the other randomization which would be standard chemotherapy. Proportion of Patients (With Upper 95% CI) 60% 50% 40% 30% 20% 10% 0% Hematologic Response in Induction 1 44% 25% 1. Kantarjian H et al. N Engl J Med. 2017;376: Blinatumomab (N = 271) SOC (N = 134) P <.001 OR (CR/CRh/CRi) 34% P <.001 P = % 30% 16% N = N = 9% 1% 4% 4% CR CRh CRi OR (as treated) Hazard ratio for EFS 0.55 (0.43, 0.71); P <.001 So, if you look at the hematologic responses, very simply, those in orange are the blinatumomab, and the overall response rate looking at complete remission, complete remission with hematologic recovery or incomplete recovery fared better in the blinatumomab arm, 44% versus 25%. And even if you parceled it out to complete remission or complete remission with incomplete hematologic recovery, it still favored the blinatumomab. 15

16 Proportion of Patients (With Upper 95% CI) Molecular Remission Among Responders 1 60% 50% 40% 30% 20% 10% 0% 74/97 (76%) OR (CR/CRh/CRi) 1. Kantarjian H et al. N Engl J Med. 2017;376: Blinatumomab (N = 271) SOC (N = 134) 16/33 (48%) 57/74 (77%) 11/21 (52%) CR Molecular remission was defined as <10-4 blasts in the first 12 weeks Overall Survival by Subgroup 1 Subgroup Blinatumomab SOC Hazard Ratio (95% CI) Age Events/Subjects <35 years 68/123 (55.3) 34/60 (56.7) 0.70 ( ) 35 years 96/148 (64.9) 53/74 (71.6) 0.77 ( ) Prior salvage therapy S0 53/114 (46.5) 39/65 (60.0) 0.60 ( ) S1 61/91 (67.0) 32/43 (74.4) 0.59 ( ) S2+ 50/66 (75.8) 16/26 (61.5) 1.13 ( ) Prior allohsct Yes 58/94 (61.7) 26/46 (56.5) 0.81 ( ) No 106/177 (59.9) 61/88 (69.3) 0.70 ( ) Baseline BM blasts <50% 31/69 (44.9) 16/30 (53.3) 0.66 ( ) 50% 132/201 (65.7) 71/104 (68.3) 0.78 ( ) Unknown 1/1 (100.0) 0/0 (0.0) NE (NE) Overall 164/271 (60.5) 87/134 (64.9) 0.71 ( ) Blinatumomab Better SOC Better 1. Kantarjian H et al. N Engl J Med. 2017;376: If you focus on those patients who achieved a remission and then asked the question: well, how many of those patients actually achieved a remission that was deep enough to be established MRD negative? And we ve learned, at least in the upfront setting from Charles, that the establishment of MRD negative is important. This is asking a slightly different question of MRD negative in secondline or third-line, is that important? And here we re seeing that about three-quarters of patients who were treated with blinatumomab were able to achieve an MRDnegative status compared to less than half those who were treated on standard care chemotherapy. Overall Survival (ITT) Median OS (95% CI) 0.9 Blinatumomab: 7.7 mo ( ) 0.8 SOC: 4.0 mo ( ) Stratified log-rank P = HR = 0.71 ( ) No. at Risk 1: : Months At 76% of events, stratified log-rank test surpassed O Brien-Fleming boundary (P <.0194) to stop the study for benefit 1. Kantarjian H et al. N Engl J Med. 2017;376: Survival Probability And if you look at the overall survival, this fared better on the blinatumomab randomized arm, to 7.7 months versus 4 months. There does seem to be a plateau on the curve. I would argue that this is a very, very good strategy to get patients to a stem cell transplant. That s still the approach that we re taking. Not everybody has an allogeneic donor, although it s becoming more fashionable as we do haploid transplants, but you can see here that blinatumomab not only had higher remission rates, but led to an improved overall survival, and this led to the full approval, at least in the United States, this summer for blinatumomab for Philadelphia-negative ALL. If you parcel it out based on different subgroups, those patients who were younger than 35 or older than 35, all in favor of the blinatumomab. The earlier you gave the blinatumomab, the immunoconjugate, the better patients did, as compared to second or greater salvage. I think that s the theme and we ll come back to that when we think about inotuzumab as well. I think this is very important, at least in the community that I practice in. Many adult patients, different from our pediatric colleagues, are treated in the community, and so many don t have access to these agents and will upon first relapse use chemotherapy approaches, and only when those fail to work, be sent into the academic center. And you can see from here that I think that s just not an appropriate approach. Whether a patient has had relapse post allogeneic transplant, again, there was no difference. Blinatumomab seemed to work whether a patient had a prior allogeneic transplant or not, as well as a baseline blast count of less than 50 or greater than 50. Adverse Events (Regardless of Causality) 1 Any AE, n (%) Any grade 3 AE Any grade 4 AE Any grade 5/fatal AE Grade 5 infection Grade 3 AE of interest, n (%) Neutropenia Infection Neurologic event Cytokine-release syndrome Blinatumomab Treated (n = 267) 263 (99) 98 (37) 82 (31) 51 (19) 30 (11) 101 (38) 91 (34) 25 (9) 13 (5) SOC Treated (n = 109) 108 (99) 33 (30) 48 (44) 19 (17) 13 (12) 63 (58) 57 (52) 9 (8) 0 (0) Events occurring up to 30 days after last dose of protocol-specified therapy, or before allohsct 1. Kantarjian H et al. N Engl J Med. 2017;376: Adverse events were common. There was some hematologic toxicity that actually was higher in the chemotherapy arm. 16 Go online to complete the post-test and evaluation for CME credit

17 a Completed suicide. 1. Kantarjian H et al. N Engl J Med. 2017;376: Neurologic Events 1 Any central neuropsychiatric event due to direct neurotoxicity, n (%) Any grade 3 Any grade 4 Any grade 5/fatal AE Central neuropsychiatric events of grade 3 occurring in 2 or more patients, n (%) Cytokine-release syndrome (grade 3) Hematophagic histiocytosis Headache Syncope Blinatumomab (n = 267) 163 (61) 22 (8) 2 (1) 1 (<1) a 9 (3) 3 (1) 1 (<1) 0 (0) SOC (n = 109) 54 (50) 7 (6) 2 (2) 0 (0) 0 (0) 0 (0) 3 (3) 2 (2) Events occurring up to 30 days after last dose of protocol-specified therapy or before HSCT And just to call out a couple side effects that were more common in the blinatumomab arm, and Stephan will touch on this when he talks about CAR-T cell. We saw very similar toxicities, but muted a bit with blinatumomab that is central neurologic toxicity, seizures, headaches, syncope, tremors, that were seen with blinatumomab. Of course, you can treat with steroids or turn off the infusion, a little bit harder to do that with the CAR-T cell. ALL: Probability of Survival by Molecular MRD 1 So, the use of blinatumomab, which has really become the fashion, at least in our institute as well as others, is you use this agent in the setting of minimal residual disease. And what the BLAST trial is trying to describe, at least in Germany, was these are patients who were MRD positive at the end of induction or consolidation, or had early relapse with MRD-positive disease. 116 patients (median age 45 years; range 18-76) with ALL in CR but MRD 0.1% after 3 intensive courses; 35% in CRD2 Blinatumomab 15 mcg/m 2 /d x 4 weeks Q6W x 4 88 patients (78%) MRD-negative post cycle 1 Parameter N % MRD-negative CRD CRD2/3 39/2 71/50 MRD10-1 to MRD10-3 to Gökbuget N et al. ASH Abstract 379. Blinatumomab in MRD-Positive ALL 1 And what Nicola Gökbuget in the German group did was add blinatumomab to treat MRD-positive disease. These are patients in morphologic remission. Probability of survival for patients in the SR and HR groups according to molecular response status in week 16, (A) overall (P <.0001) and (B) excluding HSCT in first CR (P <.0001) 1. Gökbuget N et al. Blood. 2012;120: A B Open-Label, Multicenter, Confirmatory Phase 2 Study in MRD-Positive B-Precursor ALL 1 Ph-negative patients in hematologic CR Overall Survival by Complete MRD Response And here s the molecular MRD. These are the Kaplan-Meier curves showing that those patients who achieved an MRD complete remission had a very good prognosis versus those who did not. Screening and enrolment N=116 Phase 2 Study of Blinatumomab in Adults With MRD-Positive B-ALL 1 Blinatumomab civ 15 mcg/m 2 /day 4 weeks on, 2 weeks off 1 cycle Induction Primary endpoint assessment Blinatumomab civ 15 mcg/m 2 /day 4 weeks on, 2 weeks off Up to 3 cycles Consolidation in responders a Defined as <10-4 by Q-PCR of Ig/TCR gene rearrangements. b MRD 10-3 in an assay with minimum sensitivity of 10-4 ; either through inability to achieve MRD negativity after prior therapy, or through MRD reappearance after a period of MRD negativity. 1. Gökbuget N et al. ASH Abstract 680. Allogeneic HSCT 100-day TRM assessment Follow-up Major differences compared with exploratory study Inclusion restricted to patients with MRD 10-3a Inclusion of patients in 2nd remission Primary endpoint assessed after 1 cycle of treatment Primary endpoint: Complete MRD response in cycle 1 b Secondary endpoints: Hematologic RFS at 18 months following initiation of blinatumomab OS, MRD response b Incidence and severity of AEs; 100-day TRM after allohsct Duration of complete MRD response; time to hematologic relapse Complete MRD response (primary endpoint): MRD negative, no amplification in PCR (minimum sensitivity 10-4 ). The landmark analysis by MRD response included patients with OS of 45 days. a Log-rank P value for association between OS and MRD response; causality not implied. Underlying baseline characteristics may also influence both outcomes. 1. Gökbuget N et al. ASH Abstract 680. And what she was able to demonstrate is that a large percentage of patients were able to get into remission and then stay in remission. And some of these patients actually did not go into transplant. The goal of the study from the BLAST study was to get patients into remission, MRD-negative remission, in order to transplant those patients. But regardless of whether they had a transplant donor or not, they seemed to do very, very well. 17

18 Open-Label, Multicenter, Confirmatory Phase 2 Study in MRD-Positive B-Precursor ALL (Cont d) 1 Role of HSCT in patients with complete MRD response; transplant realization rate: 72% Simon-Marukh Plot of RFS (Landmark 45 Days) Cox Model Analysis of HSCT as Time-Dependent Covariate for RFS First Remission Second or Later Remission HR (95% CI) 2.26 ( ) 0.33 ( ) P Gökbuget N et al. ASH Abstract 680. First Remission (n = 60) Second or Later Remission (n = 60) And if you look at those patients who were censored at transplant, you can see that establishing an MRD-negative status with blinatumomab, even in a patient who was not able to get to a transplant, seemed to do well. Actually, there was an increment that was doing better than transplant. So, I think that was an exciting thing, although this wasn t the focus of this particular study. So, this has at least captivated our approach in terms of thinking about MRD negativity and focusing on blinatumomab. Blinatumomab in Pediatric Patients With Relapsed/Refractory ALL 1 Eligible patients (>28 d to <18 y) had 5% blasts and r/r ALL (refractory, 2 relapses or relapse after HSCT) Screening / Prephase 14 days Blinatumomab civ, 4 wk on, 2 wk off Safety follow-up up to 30 days after last dose HSCT offered to patients in CR 1. Locatelli F et al American Society of Clinical Oncology Annual Meeting (ASCO 2017). Abstract Long-term follow-up up to 18 months after first dose The data were recently presented at ASCO on 40 pediatric patients, establishing a remission rate of 60%, including a high remission rate in patients with a blast burden of greater than 50%. The vast majority of those patients who went into remission established an MRD-negative status, about three-quarters. And the toxicity was well tolerated. Blinatumomab in Ph-Positive ALL (ALCANTARA) 1 R/R Ph+ ALL to 2+ generation TKI (N = 45): T315I (n = 10) 3 TKI (n = 17) Prior ponatinib (n = 23) Primary endpoint (CR/CRh during first 2 cycles): CR/CRh: 16 (36%) T315I: 4 (40%) 3 TKI: 8 (47%) Prior ponatinib: 8 (35%) 1. Martinelli G et al. J Clin Oncol. 2017;35: Single-Agent Blinatumomab Secondary endpoints: Complete MRD response: 14 (88%) Proceed to allohsct: 7 (44%) Median RFS: 6.7 mo (median follow-up: 9 mo) Median OS: 7.1 mo (median follow-up: 8.8 mo) Updated FDA approval: treatment of relapsed or refractory B-ALL in adults and children Blinatumomab also got a separate indication in the Philadelphiapositive setting, based on the data from Martinelli and colleagues that was published in the Journal of Clinical Oncology this summer. Forty-five patients with relapsed/refractory Philadelphia-positive ALL that were defined as failing two or more TKIs or had a 315I; many of these patients had three or more TKIs and 23 patients had failed ponatinib. The remission rate was 36%, including 40% remission in those that were 315I-positive, and 47% of patients achieved a CR or CRh if they had three or more TKIs. So, very refractory patients achieving a remission, including an MRD-negative status in 88%. So, this was the last update of blinatumomab, so you went from an accelerated approval to full approval, approval in pediatrics, and now including Philadelphia-positive ALL. What about blinatumomab in pediatric patients? Well, this has gotten a recent approval, given the phase 4 study using a very similar 4-week-on/2-week-off continuous infusion. Blinatumomab in Pediatric Patients With Relapsed/Refractory ALL: Summary 1 Response, n/n (%) N = 40 CR during first 2 cycles 25/40 (63) <50% blasts15 (68) 15/22 (68) 50% blasts10 (56) 15/22 (68) t(17;19) 2/2 (100) MRD response during first 2 cycles 19/25 (76) <50% blasts15 (68) 12/15 (80) 50% blasts10 (56) 7/10 (70) t(17;19) 2/2 (100) 40% of responders subsequently received allohsct Safety profile (CRS, neurologic events) consistent with prior blinatumomab experience CD22 Is an Attractive Therapeutic Target CD22 is expressed on the malignant cells in >90% of B-lymphoid malignancies CD22 is internalized upon antibody binding CD22 is not shed into the extracellular environment 1. Locatelli F et al. ASCO Abstract Go online to complete the post-test and evaluation for CME credit

19 So, let s move on to the other target, CD22-targeted agents. CD22 is expressed on the majority of B cells. Those of you who historically were doing flow patterns for B cell ALL CD22, at least in our institution, was not part of the platform that we were using. It now is because this is an algorithm where we want to know whether a patient is CD22-positive or not, both at diagnosis as well as relapse. It is internalized upon antibody binding. It s an important target then. And it s not shed into the extracellular environment, again making it a useful target. Inotuzumab Ozogamicin (INO) Inotuzumab: Clinical Data in Relapsed/Refractory ALL 1-3 Study Institution/ Center Phase 2 1 MDACC 49 Phase 2 2 MDACC 41 Phase1/2 3 Multicenter N Dosing ORR 37 (1/2) + 35 (2) S2+ 1. Kantarjian H et al. Lancet Oncol. 2012;13: Kantarjian H et al. Cancer. 2013;119: DeAngelo DJ et al. Blood Adv. 2017;1: Every 4 weeks 1.8 mg/m 2 57% Weekly 1.8 mg/m 2 59% ( ) Weekly 68% mg/m 2 (total) 69% Humanized IgG4 anti-cd22 AcBut linker: 4-(4 -acetylphenoxy) butanoic acid dimethyl hydrazide So, there were several phase 2 trials done at MD Anderson, then a phase 1/2 multicenter trial that I led that was recently published. MD Anderson first looked at a lymphoma-like schedule, every 4 weeks and then, due to some hepatotoxicity, switched that to a weekly schedule, showing an overall response rate of almost 60%. N-acetyl gamma calicheamicin Average loading of calicheamicin derivative on mab is 5-6 moles of calicheamicin/mole of mab (range, 3-9) for INO ~100% of mabs conjugated Inotuzumab ozogamicin is a humanized IgG4 anti-cd22 antibody that is linked to the chemotherapy conjugate, calicheamicin, a very potent chemotherapy. Very similar to that from gemtuzumab, which just got approval in the United States after a long hiatus. We took a different approach, did a standard classic phase 1, established an MTD, and then went into a phase 2 or expansion cohort in second or greater salvage, and showed about a 68% overall response rate with the weekly regimen in a multicenter approach. MRD in Patients With CR/CRi 1,2 The concentration, interestingly, of calicheamicin for the inotuzumab is a little bit higher, about 5 to 6 moles of calicheamicin per monoclonal antibody, versus gemtuzumab, which is about 1 to 2 moles. Inotuzumab Ozogamicin: MOA Retains Activity Against Tumor Cells With Slow Cycling Times Response MRD-negative, a n (%) Median (range) time to MRD negativity, d 1.2 mg/m 2 (n = 2) 1.6 mg/m 2 (n = 9) 1.8 mg/m 2 (n = 8) 1.8 mg/m 2 Exp (n = 6) 1.8 mg/m 2 S2+ (n = 24) 2 Total (n = 49) 2 (100) 8 (89) 8 (100) 6 (100) 18 (75) 42 (86) 99 (98-99) 32 (22-64) 30 (22-141) 38 (21-134) 26 (21-80) 34 (21-141) A E C B B C D a <1 abnormal cell out of 10 4 mononuclear cells in bone marrow by 6-color multiparameter flow cytometry per central lab analysis performed at the University of Washington, Seattle. 1. DeAngelo DJ et al. ASH Abstract Advani AS et al. ASH Abstract Not only did these patients get into remission, but the vast majority of these patients achieved an MRD-negative status. Eighty-six percent of the 49 patients who got into remission were MRD negative, and most of those patients went MRD negative after one or two cycles, so very rapid responses, very rapid MRDnegative status. And they re very similar to any antibody-drug conjugate. It binds to the surface. It s internalized due to the acid ph of the lysosome. The antibody-drug conjugate is degraded, releasing calicheamicin into the cytosol, trafficking into the nucleus to cause DNA damage and in a sensitive lymphoblast or a sensitive disease, causing apoptosis. 19

20 INO-VATE: Inotuzumab vs Chemotherapy in Relapsed/Refractory CD22+ ALL 1,2 An ongoing phase 3 study: 326 patients randomized at 117 sites in 19 countries (INO-VATE ALL) Relapsed/ refractory CD22+ ALL Due for salvage 1 or 2 therapy Ph or Ph+ 1:1 Randomization (N = 326) Stratifications Duration of first remission 12 vs <12 mo Salvage 2 vs 1 Aged 55 y vs <55 y INO dose was reduced to 1.5 mg/m 2 /cycle once the patient achieved CR/CRi Primary endpoints: response, OS 1. Accessed December 6, Kantarjian HM et al. N Engl J Med. 2016;375: Inotuzumab ozogamicin Starting dose 1.8 mg/m 2 /cycle 0.8 mg/m 2 on day 1; 0.5 mg/m 2 on days 8 and 15 of a day cycle ( 6 cycles) Investigator s Choice CT ( 4 cycles) FLAG HiDAC Cytarabine/mitoxantrone And this will form the basis of the randomized phase 3 study of INO-VATE, inotuzumab versus standard chemotherapy. This was a one-to-one randomization, including patients who were CD22- positive, had failed standard chemotherapy, and were in first or second salvage. The difference between this and the TOWER study is it did include Philadelphia-positive patients, which has a little bit of an Achilles heel, if you will, because some of these patients were in Europe where there s a chemo-light approach to the therapy of the Philadelphia-positive. Many of those patients were treated with TKI and steroids alone and were rather chemotherapy naïve. Patients were stratified to duration of first remission, salvage one or two, and age 55. Inotuzumab was given on a weekly schedule and once patients got into remission, the dose was reduced to 1.5 mg/m ² from 1.8 mg/m 2. And again, the control arm was investigator s choice, a high-dose AraC-containing regimen. And again, recapitulating the role of these monoclonals, just like blinatumomab, the vast majority of patients who went into remission were able to achieve an MRD-negative status, 78% versus 28%. And again, most of the patients achieved a CR/CRi after one cycle of therapy. CR/CRi by Stratification Factors 1 CR/CRi, % a INO (n = 109) SOC (n = 96) 1-Sided P Value In Favor of INO CR/CRi Rate Difference, % % Rate Difference (97.5% CI) All patients < (34-61) Duration of First Remission <12 mo < (34-67) 12 mo (18-64) Salvage < (41-72) Salvage (2-56) Age < < (27-62) Age < (31-75) a Analysis of CR/CRi based on a modified ITT population excluding 13 untreated patients from the SOC arm (ie, n = 96 for the SOC arm rather than n = 109). 1. Kantarjian HM et al. N Engl J Med. 2016;375: If you look at the stratification factors, they all favored inotuzumab based on duration of remission, less than 12 months/greater than 12 months, first or second salvage, as well as age, 55 under or older. INO-VATE: INO vs SOC in R/R ALL: Duration of Remission Among Responders 1 INO-VATE: Treatment Response 1 INO SOC 1-Sided P Value N a CR/CRi,% (95% CI) 80.7 (72-88) 33.3 (24-44) <.0001 CR 35.8 (27-46) 19.8 (12-29).0056 CRi 45.0 (35-55) 13.5 (7-22) <.0001 MRD negativity among responders, n (%) [95% CI] CR/CRi 69/88 (78.4) [68-87] 9/32 (28.1) [14-47] <.0001 CR 35/39 (89.7) [76-97] 6/19 (31.6) [13-57] <.0001 CRi 34/49 (69.4) [55-82] 3/13 (23.1) [5-54].0034 In both arms, most patients achieved CR/CRi in cycle 1 (INO, 73%; SOC, 91%) a Analysis of CR/CRi based on a modified ITT population excluding 13 untreated patients from SOC arm (ie, n = 96 for SOC arm rather than n = 109). 1. DeAngelo DJ et al. 20th Congress of the European Hematology Association 2015 (EHA 2015). Abstract Kantarjian HM et al. N Engl J Med. 2016;375: This was the duration of remission amongst responders, 4.6 months versus 3.1 months. Again, we re moving the curve forward but just like blinatumomab, I view this as a bridge to an allogeneic transplant. INO-VATE: INO vs SOC in R/R ALL: Progression-Free Survival 1 So, there were two primary endpoints for this study, a little bit unusual. But the first 218 patients were enrolled with the point of having a primary endpoint of response, which is shown here. Those patients randomized to inotuzumab had an 80% response rate compared to 33% in the standard chemotherapy arm. 1. Kantarjian HM et al. N Engl J Med. 2016;375: Go online to complete the post-test and evaluation for CME credit

21 If one looks at progression-free survival, again favoring the inotuzumab arm of 5 months versus about 2 months, it was statistically significant. INO-VATE: INO vs SOC in R/R ALL: Overall Survival 1 Data appeared to depart from proportional hazards assumption 2-y survival probability higher with INO (23% [95% CI, 16-30] vs 10% [5-16]) Inotuzumab Ozogamicin in Pediatric Patients With Relapsed/Refractory Acute Lymphoblastic Leukemia 1 Deepa Bhojwani, Richard Sposto, Nirali Shah, Vilmarie Rodriguez, Maureen Megan O'Brien, Jennifer Lynn McNeer, Mignon Loh, Susan Rheingold 1. Kantarjian HM et al. N Engl J Med. 2016;375: FDA approved in 2017 for the treatment of adults with relapsed or refractory B-cell precursor ALL 1. Bhojwani D et al. ASCO Abstract So what about inotuzumab in pediatric patients? So, just like blinatumomab, this has been tested and was presented at ASCO this past summer. And this is the overall survival curve, which has just a very slight statistical significance, with the median survival of 7.7 versus 6.7 months. But if you look at the 2-year survival probability, it was much higher in inotuzumab at 23% versus 10%. The caveat is those patients are the patients who ended up getting transplanted. About two and a half times more patients were transplanted on the inotuzumab arm as compared to the standard of care. And this is what led to the FDA approval in the United States later this past summer for patients with relapsed/refractory B cell ALL, including both Ph-negative as well as Ph-positive. Response to Inotuzumab Therapy 1 Baseline Disease Status CR CR MRD PR Mixed M2/M3 (n = 40) 24 (60%) 15 (38%) 3 (8%) 2 (5%) MRD only (n = 7) NA 4 (57%) NA 1 Of patients who achieved CR 15 (62%) were MRD negative 20 (83%) in CR after 1 cycle 1. Bhojwani D et al. ASCO Abstract VOD/SOS Among Inotuzumab-Treated Patients 1 VOD incidence: INO, 13% (n = 22) vs SOC, 1% (n = 1) 5 (3%) patients had VOD during study treatment (2 with pre-study HSCT) 77/164 (47%) on INO had post-study HSCT vs 33/162 (20%) in the SOC arm 17/77 (22%) on INO had VOD post-hsct (5/17 also had pre-study HSCT) Median (range) time to VOD after HSCT: 15 (3-57) days Multivariate Analysis of Factors Associated With Post-HSCT VOD 2 Factor OR (95% CI) P Alkylator conditioning (dual vs single) 7.6 ( ).008 Age ( 55 vs <55 y) 4.8 ( ) Kantarjian HM et al. N Engl J Med. 2016;375: Marks D et al. EBMT Presidential symposium. The response to single-agent inotuzumab in the pediatric setting, 40 patients with histologic relapse showed a CR rate of about 60%, including 40% of those patients who achieved an MRD-negative status. Survival Post-Inotuzumab Therapy 1 Median follow-up: 4 months (range 1-24) 20/48 (42%) of patients proceeded to HSCT post-inotuzumab Event-Free Survival Overall Survival So, one of the problems with inotuzumab, similar to gemtuzumab is veno-occlusive disease. In the inotuzumab arm it was 13% of all patients as compared to 1% in the standard of care. Some of these patients did develop VOD during inotuzumab treatment without post-inotuzumab transplant. Many of these patients had had a transplant before and then got inotuzumab. The factors that were associated with veno-occlusive disease in this study were patients who were older and patients who received a double-alkylating regimen, conditioning regimen, for allogeneic transplant. And you can see the odds ratio of developing VOD if you had one of these or both of these factors. 1. Bhojwani D et al. ASCO Abstract And here s the post-inotuzumab therapy in this pediatric population, with median follow-up short, only 4 months. Fortytwo percent of patients were able to proceed to transplant postinotuzumab. 21

22 Inotuzumab Ozogamicin in Relapsed/Refractory Pediatric ALL: VOD/SOS 1 Only observed in pediatric patients who underwent transplant post-inotuzumab (N = 20) 10/20 (50%) developed VOD/SOS 6/10 with prior HSCT; 3/10 with no prior HSCT I presented to you two phase 3 studies, blinatumomab versus standard of care and inotuzumab versus standard of care, showing better improved CR, MRD rates, as well as both progression and overall survival. There are ongoing studies that Dr. Litzow will take us through combining blinatumomab in an upfront fashion and inotuzumab also in an upfront fashion, at least in the AYA population. VOD/SOS: 3 mild, 2 moderate, 3 severe, 1 fatal 1. Bhojwani D et al. ASCO Abstract And very similar to the adult patient population, we did see some veno-occlusive disease in this patient population; 10 out of 20 developed veno-occlusive disease, with 6 out of 10 prior transplant. So, most of these patients had relapsed after transplant and then proceeded onto transplant and got veno-occlusive disease, but not all. Conclusions Monoclonal antibodies: high ORR for single-agent blinatumomab and inotuzumab High single-agent MRD-negative rates Toxicities Inotuzumab ozogamicin: thrombocytopenia, LFT, VOD Blinatumomab: cytokine release and neurotoxicity Randomized phase 3 studies: Blinatumomab vs SOC: improved CR and MRD rates as well as PFS and OS Inotuzumab vs SOC: improved CR and MRD rates as well as PFS and OS Future Studies: Blinatumomab: ECOG 1910 upfront phase 3 study in patients (enrolling) Inotuzumab: Alliance led AYA phase 3 study in patients (awaiting CTEP approval) So in conclusion, the monoclonal antibodies do represent a very high single-agent response rate. One of the things that we ll learn from our next speaker, Dr. Litzow, is how these agents are going to be combined to really get more bang for our buck. I mean, what we ve learned since Sidney Farber s days in the 40s is that the treatment for ALL is multiagent therapy. We don t rely on singleagent therapy. But what I ve shown you is single-agent activity with the bispecific antibody, blinatumomab, as well as inotuzumab, high single-agent remission rates and very, very high MRD-negative status. They do have toxicities. Inotuzumab, thrombocytopenia, liver function abnormalities, including veno-occlusive disease, specifically those patients moving on to second transplant or getting doublealkylator conditioning regimen. And the blinatumomab, patients were developing cytokine release or neurotoxicity. 22 Go online to complete the post-test and evaluation for CME credit

23 A Look at the Future of Antibody- Based Combinations in ALL Mark R. Litzow, MD Mayo Clinic Rochester, Minnesota So, just a reminder that CD19 and CD22 it s no accident that these targets have been focused on for drug development since they re on the whole spectrum of B cells from early development to more mature development. Mechanisms of Action of Monoclonal Antibody Conjugates 1 Presentation Objectives Review the established and emerging role of monoclonal antibody-based therapy, focusing on combination approaches with chemotherapy Rituximab Inotuzumab ozogamicin Blinatumomab Others Update the use of antibody-based therapy, TKIs, and BMT in the treatment of Ph+ ALL 1. Parikh S, Litzow M. Future Oncology. 2014;10: A. Naked (unconjugated) antibodies B. Bi-specific T cell engaging antibody C. Antibodies linked to toxins D. Antibodies linked to drugs E. Chimeric antigen receptor T cells So, we re going to talk about some of these different constructs, and this is just an overview of some of the different mechanisms: the naked or unconjugated antibodies; the bispecific antibody, blinatumomab; inotuzumab ozogamicin, which is antibody linked to a drug; and then Dr. Grupp is going to be talking about the CAR-T cells. Dr. DeAngelo: The third speaker is my colleague and friend, Dr. Mark Litzow from Mayo Clinic, who will review the combination of antibody-based therapies in acute lymphoblastic leukemia. Dr. Litzow: Thank you very much, Dan, and I m grateful to PeerView for inviting me to speak today. It s a great honor. So, I m tasked with discussing how we ll combine some of the agents that Dr. DeAngelo talked about into front-line therapy. We ll talk about an established agent, rituximab, and then talk about inotuzumab and blinatumomab, and then the others I have there, as I ll briefly mention epratuzumab. Then I want to talk about how we re looking at combining some of these antibody constructs into the treatment of Ph-positive ALL. Target Antigens in ALL Pro-B Pre-B1 Immature B Mature B Plasma IgM IgM IgG Overall Survival With Hyper-CVAD ± Rituximab 1 OS Probability Total CR Die Relapse Die in CR 0.2 Hyper-CVAD rituximab Hyper-CVAD Thomas DA et al. Blood. 2009;113: Months Hyper-CVAD + rituximab Hyper-CVAD <60 years old P =.002 So, rituximab has the longest history here. This goes back to the MD Anderson studies back almost a decade ago, combining rituximab with hyper-cvad. And in these retrospective studies, showed an advantage for combining it with hyper-cvad, and this was primarily in younger adult patients. D-J VDJ V-J light CD19 CD22 CD20 CD10 Igα/β Cytoplasmic Surface 23

24 Immunochemotherapy With Rituximab 1,2 GMALL: standard risk CD20+ patient Retrospective comparison of 2 cohorts Outcome, % Chemo Alone n = 181 Chemo + Rituximab n = 82 CR Early death 3 5 MRD <10-4 wk P = OS at 3 y Hoelzer DH et al. Haematologica. 2009;94:195(suppl; abstr 0481). 2. Hoelzer DH et al. Blood. 2010;116:78 (abstr 170). The Germans confirmed similar findings with their BFM-type regimen. You can see that although the CR rates were similar with chemotherapy alone versus chemotherapy plus rituximab, the MRD rate was much better combining rituximab, and this led to a survival advantage. Again, these were retrospective studies. Addition of Rituximab Improves the Outcome of Adult Patients With CD20-Positive, Ph-Negative, B-Cell Precursor Acute Lymphoblastic Leukemia 1,2 Results of the Randomized GRAALL-R 2005 Study (NCT ) Sébastien Maury, Sylvie Chevret, Xavier Thomas, Dominik Heim, Thibaut Leguay, Françoise Huguet, Patrice Chevallier, Mathilde Hunault, Nicolas Boissel, Martine Escoffre-Barbe, Urs Hess, Norbert Vey, Thorsten Braun, Jean-Pierre Marolleau, Yves Chalandon, Véronique Lhéritier, Kheira Beldjord, Marie-Christine Béné, Norbert Ifrah, Hervé Dombret 1. Maury S et al. ASH Abstract Maury S et al. N Engl J Med. 2016;375: This was finally studied prospectively, and this is the study that many of you are likely aware of that was presented at a plenary session at ASH a couple of years ago and was published in The New England Journal of Medicine last year. Addition of rituximab improves the outcome of adult patients with CD20-positive/Phnegative B-ALL. Eligibility: GRAALL-R ,2 Patients with newly-diagnosed previously untreated Ph B-cell precursor (BCP) ALL included in the GRAALL 2005 trial Aged years old With 20% bone marrow non-burkitt ALL blasts No concomitant or recent (<6 months) other malignancy No contra-indication to intensive chemotherapy Signed informed consent So, in the GRAALL-R 2005 trial, these were newly-diagnosed, untreated, Philadelphia-negative, B cell precursor ALL, aged 18 to 59. They had to have greater than 20% blasts, no prior therapy. And they defined the CD20 expression, as many do, as 20% or more. Rituximab, 375 mg/m 2 for 16 (to 18) infusions Prephase and induction Consolidation 1 and 2 Late intensification Consolidation 3 CNS irradiation 24-hour maintenance Treatments 1,2 PRED DNR-VCR-PRED-CPM/hyperC-ASP IT IT (x2) HiDAC DXM-ASP HD-MTX VCR-ASP 6-MP HD-CPM VP16-MTX IT DNR-VCR-PRED-CPM/hyperC-ASP IT (x2) HiDAC DXM-ASP 6-MP HD-MTX VCR-ASP 6-MP VCR-PRED (x12, monthly) MTX 6-MP (24 mo) HD-CPM VP16-MTX IT 1. Maury S et al. ASH Abstract Maury S et al. N Engl J Med. 2016;375: MRD1 (6 wk) MRD2 (12 wk) HiDAC DXM-ASP HD-MTX VCR-ASP 6MP HD-CPM VP16-MTX IT Ig/TCR MRD at 6 and 12 weeks Sustained G-CSF support Allogenic HSCT after consolidation 1 or 2 in high-risk ALL patients younger than 56 years This is the schema of the study, the typical multiagent therapy. And the arrows represent where rituximab was added. It was a total of 16 doses, with the possibility of two additional if there was some additional induction given. And you can see that six of those doses were in maintenance. Probability of EFS % (55-73) 1. Maury S et al. N Engl J Med. 2016;375: Event-Free Survival % (42-61) 0.25 Control HR = 0.66 ( ) Rituximab P = Months No. at Risk Control Rituximab After censoring at HSCT in CR1: HR = 0.59 ( ) P =.021 So, the endpoint of this study was event-free survival, and you can see that that was improved with the addition of rituximab. And up in the right after censoring for transplant, this advantage was still seen. There was not an overall survival advantage overall, but again, when censoring was done for transplant, there was a survival advantage seen. Positive for CD20 expression Defined as 20% or more CD20+ leukemic blasts 1. Maury S et al. ASH Abstract Maury S et al. N Engl J Med. 2016;375: Go online to complete the post-test and evaluation for CME credit

25 Multivariable Analysis (EFS) 1 Hazard Ratio (range) Treatment with rituximab 0.61 ( ).018 Age 1.02 ( ).022 WBC 30 x 10 9 /L 1.87 ( ).005 CNS involvement 2.38 ( ).012 Allogeneic HSCT in first remission (time-dependent variable) P 1.14 ( ).62 I just wanted to point out that our colleagues in the United Kingdom are conducting this trial, adding rituximab to their standard chemotherapy regimen that s what the red arrow here shows. They re also asking some transplant questions and also a T cell question with nelarabine. This study has completed the accrual to the B cell arm, not quite to the T cell arm, so we re anxious to see these results. Prednisolone Up-Regulates CD20 Expression Maury S et al. N Engl J Med. 2016;375: The multivariate analysis treatment with rituximab, age, high white counts, CNS involvement were all significant factors in outcome. Conclusion 1 Approximately 32% of adults with Ph BCP-ALL have CD20+ ALL In these patients, the addition of rituximab to standard intensive chemotherapy is well tolerated and Significantly improves EFS Also prolongs OS of patients not receiving allogeneic HSCT in first CR Adding rituximab to standard therapy should thus become a standard of care for these patients The optimal dose and schedule of rituximab administration remains to be determined 1. Maury S et al. N Engl J Med. 2016;375: CD20 MFI Dworzak MN et al. Blood. 2008;112: mcg/ml Day 0 Day 8 In vitro In vivo Suggests CD20 cutoff of 20% is arbitrary and of uncertain significance One difference in this study is based on these data from the Austrians showing that the addition of corticosteroids upregulates CD20 expression. So, my flow cytometry colleague in our ECOG- ACRIN group, Elizabeth Paietta, likes to point out that the CD20 cutoff of 20% is really arbitrary and of uncertain significance. And actually, in the UK study, they are giving rituximab to all-comers in the B cell arm and not only those who have more than 20% CD20 expression. So, on this study, only about 30% of patients were CD20-positive. The addition of rituximab significantly improved event-free survival and improved overall survival in those who did not go on to transplant. Really established this as the standard of care. The authors acknowledge that we don t really know the optimal way to give this agent in the therapy of ALL, but certainly a number of the different programs have shown an advantage. Novel Antibody Combination Strategies UKALLXIV Trial Schema Phase 1 induction (4 weeks): Precursor B- and T cell patients receive PEG-ASP + standard phase 1 induction therapy Pre-B-cell ALL T cell ALL B1: Standard phase 1 therapy only B2: Standard phase 1 therapy + CD20 (rituximab) Standard phase 1 therapy only Phase 2 induction (4 weeks): Precursor B- and T cell patients receive standard phase 2 induction therapy Pre-B-cell ALL T cell ALL Standard phase 2 therapy only T1: Standard phase 2 therapy T2: Standard phase 2 therapy + nelarabine only Patients in complete remission: Risk assessment performed on all patients at this time-point Sibling donor No sibling donor High risk Standard risk >40 years old: 40 years old: Intensification with Myeloablative: >40 years old: 40 years old: Continue MTX HD-MTX + PEG-ASP Conditioning with Intensification with Myeloablative: intensification, Conditioning regimen with fludarabinemelphalan and alemtuzumab To receive allohsct Conditioning: fludarabine- etopside-tbi maintenance etopside-tbi HD-MTX + PEG-ASP Conditioning with consolidation, To receive allohsct (sibling) (sibling) melphalan and alemtuzumab To receive allohsct To receive allohsct (MUD) (MUD) So let me turn now and talk about sort of the studies that are in development or ongoing, combining some of these antibody constructs with a chemotherapy. Randomization to P1: Standard P2: Collapsed palifermin dose P1: Standard P2: Collapsed 1. Accessed November 21,

26 Phase 3 Alliance Trial (AYA ALL): INO in Young Adults With Pre B-Cell ALL 1 Ph CD years Stratification: Age, CD20 status LDA-card (Ph-like signature) Primary endpoint: 3-year EFS C10403 Induction R for CD Accessed November 21, R 2 cycles INO post induction No INO C10403 consolidation maintenance Eligibility Previously untreated B-cell ALL Patients ages years Presence of surface CD22+ lymphoblasts Philadelphia negative cytogenetics Initial run-in safety phase in 6-12 patients before randomized trial opens A041501: Objectives Primary objective: To confirm tolerability of the combination regimen with the addition of inotuzumab ozogamicin to the pediatric-inspired regimen of CALGB Safety run in in 6-12 patients To determine whether the addition of inotuzumab ozogamicin significantly improves the event-free survival (EFS) achieved with the pediatric-inspired regimen of CALGB % power to detect an improvement in 3-year EFS 65% to 80% (randomization occurs after induction) 157 per arm (140 responses per arm) And this is a trial that was just launched this summer that Dr. DeAngelo is leading through the Alliance on behalf of the NCTN, the Intergroup mechanism here in the United States. This is , and this builds on the successful study that was done through the Intergroup that Dr. Wendy Stock led, the C10403 trial that showed that we could take a pediatric-intensive regimen and apply that to adults up to the age of 40 here in the United States. So, this trial is using the C10403 backbone, adding rituximab, and then randomizing patients after induction. They either get two cycles of inotuzumab, three doses of 0.5 mg/m ² on days 1, 8, and 15, or to not get the inotuzumab and then to go on, as all the patients will, to consolidation and maintenance. You can see the stratification here. We re going to be looking at the Ph-like signature. Here is the primary endpoint, and here are some of the eligibility criteria. Currently this run-in safety phase is being conducted at some of the Alliance centers because this adding inotuzumab here has never been done previously. So once that safety signal is shown, this is going to open up more widely throughout the United States. I DNR VCR DEX PEG-ASP IT-MTX IT-AraC Alliance US Intergroup Study for AYA: Randomized Phase 3 trial (CD22+; Ph ) 1 Inotuzumab cycles 1 and 2 C Cyclo VCR DEX PEG-ASP AraC 6-MP IT-MTX (R-CD20+) IM MTX VCR PEG-ASP IT-MTX (R-CD20+) DI DOX Cyclo DEX PEG-ASP AraC 6-TG IT-MTX (R-CD20+) M DEX VCR 6-MP MTX IT-MTX Maintenance therapy continues for 2 (F) to 3 (M) years CD22 positive B-cell ALL Accessed November 21, Stock W et al. ASH Abstract 796. This just shows where the inotuzumab is going to be added in the overall schema of this regimen. The primary objective is to confirm the tolerability and then to see if it improves event-free survival. And I think the key here is to see if we can deepen the responses and get more patients to MRD negativity. This is going to need to accrue a little over 300 patients, and you can see the improvement in 3-year event-free survival that the study is looking for. Inotuzumab Ozogamicin (INO) in Combination With Low-Intensity Chemotherapy (mini-hypercvd) as Frontline Therapy for Elderly Patients With Acute Lymphoblastic Leukemia (ALL): Interim Results of a Phase II Clinical Trial 1 Sasaki K, Jabbour E, O Brien S, Ravandi F, Thomas D, Garcia-Manero G, Daver N, Borthakur G, Jain N, Konopleva M, Short N, Pemmaraju N, Alvarado Y, Jacob J, Garris R, Thompson P, Cortes J, Kantarjian H 1. Sasaki K et al. ASH Abstract 588. So inotuzumab has been combined with chemotherapy. Our colleagues at MD Anderson have done this in elderly patients with ALL, combining it with mini-hypercvd. And this was presented last year at this meeting, and I ll just go over some of these data. Dose-reduced hyper-cvd for 8 courses Cyclophosphamide (150 mg/m 2 x 6) 50% dose reduction Dexamethasone (20 mg) 50% dose reduction No anthracycline Methotrexate (250 mg/m 2 ) 75% dose reduction Cytarabine (0.5 g/m 2 x 4) 83% dose reduction Inotuzumab on day 3 (first 4 courses) 1.8 mg/m 2 course mg/m 2 courses 2-4 Rituximab days 2 and 8 (first 4 courses) for CD20+ Intrathecal chemotherapy days 2 and 8 (first 4 courses) POMP maintenance for 3 years 1. Jabbour E et al. ASH Abstract 794. MiniHCVD-INO in ALL: Design 1 26 Go online to complete the post-test and evaluation for CME credit

27 So, this is dose-reduced hyper-cvad. The doses of chemotherapy are reduced. No anthracycline is given. And this is a schedule of inotuzumab for the first four courses of therapy. Rituximab is allowed. And then, of course, intrathecal therapy and then maintenance therapy. 1. Sasaki K et al. ASH Abstract 588. MiniHCVD-INO in ALL: Design (Cont d) 1 Intensive phase D3 D3 D3 D3 Maintenance phase 36 months MiniHCVD Mini MTX-cytarabine POMP Maintenance Inotuzumab Inotuzumab First 6 Patients Patients 7-33 Patients 34 and Beyond First cycle (mg/m 2 ) C2-4 (mg/m 2 ) MiniHCVD-INO in ALL-VOD (n = 4, 8%) 1 Parameter #1 #2 #3 a #4 b Age, y MRD status Neg Neg Neg Neg No. cycles C3D57 C2D40 4 cycles C2D70 Grade Transplant No No Yes No Status Alive Died Died Died a VOD day 43 post MRD allohsct with fludarabine/busulfan. b Multiple co-morbidities, early manifestations of VOD. 1. Sasaki K et al. ASH Abstract 588. Dr. DeAngelo mentioned veno-occlusive disease or sinusoidal obstruction syndrome is a risk with inotuzumab, and there were four patients on this trial that developed veno-occlusive disease. You can see that unfortunately two of them led to deaths, and the other two were grade 2, but certainly a risk factor to keep in mind. And this just shows a schematic of this. The arrows show where the inotuzumab is given. This is a typical alternating schedule still of hyper-cvad but with the reduced doses. And you can see the doses of the inotuzumab that are given. They did some dose adjustments and have settled on the 1.3 mg/m ² dose with a reduction in cycles two to four. MiniHCVD-INO in ALL Response (n = 43) 1 4 patients enrolled with CR No early deaths on study 19 pts received POMP maintenance Median follow-up: 24 mo (range: 1-55) Response, % INO + MiniHCVD (N = 43) ORR 98 CR 84 CRp 12 CRi 2 Cytogenetic CR a 100 MRD d21 b 76 MRD overall c 96 MiniHCVD-INO vs HCVAD in ALL: Overall Survival 1 Survival in Patients >60 Years of Age 1. Sasaki K et al. ASH Abstract Short NJ et al. ASH Abstract Total Fail 3-Year OS Median % NR MiniHCVD-INO ± Ritux HCVAD ± Ritux % 16 mo P =.007 ASH 2017: similar efficacy with longer follow-up for minihcvd-ino 2 And this shows a retrospective comparison of outcome in patients over the age of 60 adding inotuzumab ozogamicin to this regimen versus their hyper-cvad regimen in this age group with rituximab, and showing a superior outcome with the combination. a n = 22 evaluable pts with abnormal karyotype. b n = 41 evaluable pts. c n = 46 evaluable pts. 1. Sasaki K et al. ASH Abstract 588. So, last year at this meeting, they presented the data on 43 patients. Four patients had had some prior chemotherapy, I think one cycle, and then went on the study. So, of the evaluable patients, the overall response rate was extraordinary, 98%. The CR rate [was 84%], with a cytogenetic complete response of 100%, and a very high MRD response as well. A few more patients have been added to this study and there are going to be updated data on this regimen presented on Sunday in a poster session. Salvage MiniHCVD+INO in Ph-Negative Relapsed/Refractory ALL 1 n = 59, median age 35 y (range18-87) 78% response rate 59% CR (82% of these MRD ) 26 patients (44%) went to allohsct 9 patients with VOD With median follow-up 24 mo, RFS/OS = 8/11 mo 1. Jabbour E et al. JAMA Oncol Aug 31 [Epub ahead of print]. 27

28 MD Anderson has also utilized this regimen in patients with relapsed and refractory disease. This was presented last year at this meeting and then just recently published in JAMA Oncology by Dr. Jabbour. Fifty-nine patients these were all-comers with relapsed and refractory disease, so from young adults to elderly patients. Fifty-nine percent complete remission rate, 78% response rate overall, with 82% of these leading to MRD negativity and allowing 26%, or a little less than half, of these patients to go to transplant. Again, veno-occlusive disease was seen, and this is the relapse-free and overall survival. Salvage MiniHCVD + INO: Survival Outcomes 1 General OS and Responder RFS OS by Salvage Status CD22 Epratuzumab (E) 1,2 Investigational humanized IgG1 monoclonal antibody directed against CD22 More than 300 adults with B-NHL have received E ~85 patients with autoimmune disease have received E Distinct mechanism of action, modulating B-cell activation 1. Carnahan J et al. Mol Immunol. 2007;44: Goldenberg Future Drugs C-C 6 C-C 5 C-C 4 C-C 3 C-C 2 C-C 1 C-C Epratuzumab Human IgG1 Murine CDRs A few brief comments about epratuzumab, another CD22 antibody, a humanized IgG1 monoclonal antibody. This has been used extensively in non-hodgkin s lymphoma, in some patients with autoimmune diseases. It has a distinct mechanism of action and modulates B cell activation. Epratuzumab in Pediatric ALL Jabbour E et al. JAMA Oncol Aug 31 [Epub ahead of print]. And here are the curves from that study. The overall survival and relapse-free survival pretty much paralleling each other. And then looking at outcomes based on whether this was first salvage, second salvage, or third or greater salvage, and you can see that, not surprisingly, introducing this regimen earlier on led to a somewhat better outcome. Concept: Introduce new agents at first relapse along with VCR/PEG/ PRED/DOX backbone COG: ADVL04P2: Add 1x or 2x/week E, compare with historical CR rate of 67% with either schedule No better than historical CR rate = 66% More patients were MRD-negative than expected IntReALL (International Study for Treatment of Childhood Relapsed ALL): phase 3 trial in children with relapsed ALL 1. Raetz EA et al. ASH Abstract Raetz EA et al. J Clin Oncol ; Raetz EA et al. Pediatr Blood Cancer. 2015;62: Salvage MiniHCVD + INO: Survival Outcomes (Cont d) 1 OS by MRD Status OS by Therapy The pediatricians have used this most extensively as a single agent and then combining it with chemotherapy. They used it in this COG study that s shown here. The CR rate did not appear to be significantly better than the historical rate, although they saw more MRD negativity and felt this was encouraging enough data to pursue a phase 3 trial, which is currently ongoing. SWOG0910: Cytarabine and Clofarabine + Epratuzumab for Relapsed/Refractory ALL 1 1. Jabbour E et al. JAMA Oncol Aug 31 [Epub ahead of print]. Then this looks by MRD status. And then a comparison of inotuzumab with the mini-hyper-cvd regimen compared to historical data with inotuzumab alone, again in the relapsed/ refractory setting, and showing that the addition of chemotherapy to inotuzumab was beneficial. Phase 2 Study Clofarabine 40 mg/m 2 /day IV days 2-6 and cytarabine 1 g/m 2 /day on days 1-5 Epratuzumab 360 mg/m 2 IV days 4, 11, 18, and 25 CR/CRi of 52%: compared to 17% in a prior trial of cytarabine plus clofarabine alone Encouraging results, but a randomized trial is needed to answer this question 1. Advani AS et al. Br J Haematol. 2014;165: Go online to complete the post-test and evaluation for CME credit

29 In the adult world, Anjali Advani led this study for SWOG, combining epratuzumab with cytarabine and clofarabine, and you can see the schedule of the chemotherapy and epratuzumab here. CR rate was encouraging compared to cytarabine and clofarabine alone, but obviously, a randomized trial would be required to answer this question. I m not honestly sure that s going to be pursued, but would be a consideration. S1318: BCR-ABL Negative Cohort-Age 65 Years, Newly Diagnosed ALL 1 Yes Induction with Blinatumomab for 1 cycle CR or CRi No E1910: Randomized Phase 3 Trial Combining Blinatumomab With CT in Adult Frontline ALL 1 Post-remission therapy with Blinatumomab 3 cycles POMP chemotherapy 18 months CR or CRi Cycle 1 (induction phase 1) MRD Arm D Arm A Cycle 2 (induction phase 2) Consolidation cycle 1 Allogeneic BMT MRD Arm B CR No CR Off treatment Consolidation cycle 2 Consolidation cycle 3 MRD Intensification MRD Consolidation cycle 4 MRD Arm C Randomization Blinatumomab cycle #1 Consolidation cycle 1 Study Design US intergroup study Blinatumomab cycle #2 Consolidation cycle 2 n = 265/360 (509) patients MRD USA, Canada, Israel Consolidation cycle 3 1:1 randomization Allogeneic BMT MRD Consolidation cycle 4: Blinatumomab Consolidation cycle 5 Arm E Consolidation cycle 6: Blinatumomab Maintenance therapy 1. Accessed November 21, Let me turn now and talk about blinatumomab, bringing that to the upfront setting. This is the Intergroup trial that s led by ECOG-ACRIN group. I have the privilege of being a PI on this study. What we re doing in this study is giving two cycles of induction chemotherapy to patients between the ages of 30 and 70. If they achieve a complete remission, they go on to CNS intensification. If they don t achieve a complete remission, they re off treatment. Then at this point they re in remission, they get randomized to either get two cycles of blinatumomab, and following those they can go on and get consolidation therapy with two more cycles of blinatumomab later in their course. And this is the typical 4 weeks of continuous infusion of blinatumomab or at investigator discretion they can be taken to an allogeneic transplant. If they re randomized to not get blinatumomab, then they continue chemotherapy with the same four consolidation cycles and can also go off to allogeneic transplant. This is the first study in the United States in adults where we re trying to standardize the MRD assessment. Dr. Elizabeth Paietta is doing this in New York, and you can see the time points where we re looking at this. And we re trying to also look at patients that are MRD positive or MRD negative, those two subsets. There s a lot of correlative studies, and Dr. Mullighan is helping us extensively with those. This study is about halfway accrued. We actually are going to have to increase the accrual goal somewhat because the randomization rate is a little lower than we had anticipated. So, this study is ongoing and will be ongoing for another 2 to 3 years Accessed November 21, Dr. Advani, again, has developed this innovative study using blinatumomab essentially as a single agent in older patients, over the age of 65 with newly-diagnosed ALL. There are two cohorts, one a Philadelphia-negative group that gets blinatumomab for one cycle. If they get a complete remission or incomplete CR then they can go on and get three more cycles and then go to maintenance. There is CNS prophylaxis built into this. If they don t achieve that, they don t get a CR then they can still go on and get more blinatumomab to see if we can get them into remission, and then go on to maintenance. S1318: BCR-ABL Positive Cohort, Age 65 Years, Newly Diagnosed ALL 1 Yes Continue DAS to d84, PRED d32 Post-remission therapy with blinatumomab/das 3 cycles DAS/PRED chemotherapy to intolerance/disease progression Induction with dasatinib (DAS)/ prednisone (PRED) Day 28 CR or CRi Yes Yes 1. Accessed November 21, No Day 56 CR or CRi Blinatumomab 1-2 cycles CR or CRi No No Off protocol Rx And then there is a Ph-positive cohort in this trial where patients, à la the Italians, get induction with dasatinib and prednisone. If they achieve remission, then they can continue dasatinib and then get blinatumomab for post-remission therapy with dasatinib. And if they don t, then they get blinatumomab and if they achieve a remission then can cross over. So, this trial is nearing the end of its accrual. It was a relatively small study and we re awaiting these results. 29

30 Proposed Trial A041703: Antibody Sequencing in ALL A phase 2, open-label, multicenter clinical trial: Inotuzumab ozogamicin blinatumomab for older adults with untreated CD22+ Ph B-lineage ALL, or adults with relapsed or refractory CD22+ Ph B-lineage ALL Matthew Wieduwilt, MD, PhD 11/4/2017 Here s the eligibility. They get one cycle of inotuzumab. If they don t respond to that, then they d switch over to blinatumomab. If they get at least a PR, then they get another cycle of inotuzumab and then go on and get more blinatumomab, and then go into follow-up. Cohort 2: Relapsed/Refractory CD22+ Ph B-cell ALL In the interim, Dr. Matt Wieduwilt, on behalf of the Alliance and the Intergroup is proposing a trial to give inotuzumab and then potentially add blinatumomab, again for older adults, either untreated patients or relapsed/refractory patients. Eligibility R/R CD22+ Ph B-ALL Age 18 1st or 2nd salvage No CNS disease Inotuzumab ozogamicin x 1 cycle No cytoreduction CR/CRi or PR Blinatumomab x 5 cycles Inotuzumab ozogamicin x 1 cycle Follow-up Proposed Trial A041703: Antibody Sequencing in ALL (Cont d) Study Design Phase 2 single-arm, open-label, multicenter, 2 cohort clinical trial Cohort 1: Untreated, older, transplant-ineligible CD22+ Ph B-cell ALL Cohort 2: Relapsed/refractory CD22+ Ph B-cell ALL Primary Objectives Cohort 1: To estimate the 1-year EFS of older, transplant-ineligible patients with Ph B-cell ALL treated with inotuzumab ozogamicin induction blinatumomab consolidation Cohort 2: To estimate the 1-year EFS of patients with relapsed or refractory Ph B-cell ALL treated with inotuzumab ozogamicin induction blinatumomab consolidation Simon 2 stage design, 29 patients in Cohort 1, 58 in Cohort 2 Here are some further details on the design. Cohort one is the untreated, older transplant-ineligible patients. Cohort two is the relapsed/refractory patients. The goals are similar for both these cohorts, to look at 1-year event-free survival. This is a stage 2 study so it has an assignment two-stage design. This just shows the simplified schema for the study. Cohort 1: Untreated, Older, Transplant-Ineligible CD22+ Ph B-cell ALL Eligibility Untreated CD22+ Ph B-ALL Age 60 Transplant ineligible No CNS disease Inotuzumab ozogamicin x 1 cycle No cytoreduction CR/CRi or PR Blinatumomab x 5 cycles Inotuzumab ozogamicin x 1 cycle Follow-up And essentially the same schema for the relapsed/refractory patients. This study just got CTEP approval, is still relatively early in the process. But this will be sort of a replacement for the SWOG trial that I just showed you and may well activate next year. Experimental Design Schema: AALL 1331 Yes Treatment failure Prior blinatumomab or CNS disease Off protocol therapy No Blinatumomab-S (Blinatumomab Salvage) up to 2 cycles Arm A Block 2 HR/IR HR/IR randomization Evaluation 2 Block 3 Blinatumomab (2nd cycle) Evaluation pre-hsct Eligible HSCT Arm B Blinatumomab (1st cycle) Ineligible 1st relapse B-ALL (on study) Relapse Off protocol therapy Pre-treatment evaluation Block 1 Evaluation 1 LR LR randomization Arm C Arm D Block 2 Evaluation 2 Blinatumomab Block 3 (1st cycle) Continuation 1 Continuation 1 Blinatumomab Continuation 2 (2nd cycle) Continuation 2 Maintenance Blinatumomab (3rd cycle) HR/IR: 75 patients per arm over 2.5 years; increase DFS from 45% to 63% LR: 85 patients per arm over 4 years; increase DFS from 73% to 84% Maintenance The pediatricians have been active looking at blinatumomab. Dr. DeAngelo showed some of the early studies. Dr. Pat Brown is leading this study on behalf of COG. So these are B-lineage ALL patients, obviously. There s a high-risk, intermediate-risk cohort, where there s a randomization to where patients get chemotherapy and then get blinatumomab added or not. And then also a low-risk randomization with a somewhat different backbone based on the lower risk but with blinatumomab added in. If patients fail therapy and they have not had prior blinatumomab, they can get that. If they have, then they get offprotocol therapy. So, this study is ongoing. I want to conclude with just a few comments about Ph-positive ALL and the role of some of these antibody combinations. 30 Go online to complete the post-test and evaluation for CME credit

31 Relapse-Free Survival Overall Survival Outcome With Blinatumomab in Patients With Relapsed/Refractory ALL 1 Median RFS, mo n (95% CI) No censoring of allohsct (4.4-NE) Censoring of allohsct (3.8-NE) Median RFS, mo n (95% CI) 7.1 (5.6-NE) No censoring of allohsct 45 Censoring of allohsct (5.6-NE) n = 45 patients, relapsed/refractory to 1 TKI n = 20 had prior allohsct CR/CRh in 36% (4/10 T315I+) 88% CR/CRh patients MRD neg 7 responders to allohsct Median RFS/OS 6.7/7.1 months Survival Probability Long-Term Follow-Up: Imatinib in Pediatric Ph+ ALL (AALL0031) Chemo only, n = Chemo only, n = URD BMT, n = URD BMT, n = MRD BMT, n = 21 P = MRD BMT, n = 13 P = Years Years 1. Martinelli G et al. J Clin Oncol. 2017;35: Schultz KR et al. Leukemia. 2014;28: Dr. DeAngelo showed this study, the ALCANTARA study that treated 45 patients with relapsed/refractory, Ph-positive ALL, and these are some of the characteristics. And again, I won t go into a lot of detail because Dr. DeAngelo showed this. These were the relapse-free and overall survival curves from that study, but certainly demonstrating/confirming the feasibility and efficacy of blinatumomab in this setting. Blinatumomab-TKI Combination Therapy 1 Our pediatric colleagues have primarily used imatinib combined with chemotherapy for their Ph-positive ALL patients. Kirk Schultz has led these efforts. When I say chemo-only, this is chemo with TKI. They re getting results that are equivalent to what they re seeing with either unrelated donor or match-related donors, and showing that there s really not a difference in outcome, and are starting to move away from transplant. And that s one of the big questions in Ph-positive ALL is if you can get a patient to MRD negativity, do they need to have a transplant in the TKI era? 12 patients 9 relapsed/refractory Ph+ ALL 3 CML-lymphoid blast crisis Blinatumomab + Ponatinib (n = 8) Dasatinib (n = 3) Bosutinib (n = 1) Responses Hematologic 50% Cytogenetic 71% Molecular 75% 1. Assi R et al. Clin Lymphoma Myeloma Leuk Aug 18 [Epub ahead of print]. Many of us have asked, what about combining blinatumomab with a TKI? In the ALCANTARA study, no TKI was given during the blinatumomab treatment. The MD Anderson center has done this and published this small study, 12 patients, nine with relapsed/ refractory, Ph-positive ALL, three with lymphoid blast crisis of CML, combining blinatumomab with these TKIs. The ponatinib is given at a dose of 30 mg and then subsequently reduced. You can see that here were the responses, and you can see the survival curves for these small cohorts of patients all patients and the ponatinib patients. Schema Overview: Proposed US Intergroup Trial Ph+ ALL Randomization Induction TKI + steroids + blinatumomab TKI + steroids + chemotherapy MRD eradication MRD neg Randomization AlloHSCT? TKI maintenance AlloHSCT followed by TKI maintenance In the US Intergroup, we re going to be proposing this study that s not been approved yet. But to randomize patients with Ph-positive ALL to get a tyrosine kinase inhibitor we re targeting at this time dasatinib with blinatumomab and comparing that to TKI, both getting steroids. And the chemotherapy here would be hyper- CVAD based on the MD Anderson data. Patients can cross over to the other arm if they don t achieve MRD negativity, we re proposing. If they achieve MRD negativity, then we re going to ask patients to consider being randomized to either get a transplant or to go to TKI maintenance with close monitoring and doing a delayed transplant if they relapse or develop MRD positivity. So, we ll see where this study goes. 31

32 Molecular Formats of Bispecific Antibodies Exploring the Role of CAR-T Cell Therapy in a New ALL Landscape Quadroma Knobs-into-holes Knobs-into-holes and CrossMAb Dual-variable domains Stephan A. Grupp, MD, PhD University of Pennsylvania Perelman School of Medicine The Children's Hospital of Philadelphia Philadelphia, Pennsylvania IgG-scFv K body Half-molecule exchange DART-Fc scfv-hsa-scfv Nanobyd Dock-and-lock Diabody Single chain diabody Tandem scfv Tandem diabody DART I just wanted to comment that the area of immunotherapy in B-lineage diseases is really exploding, and there are many different constructs that are being developed. So I m not going to go over this slide, but just to show that there are a lot of interesting agents out there and I think this area is really going to explode in the next few years. Chimeric Antigen Receptor (CAR) 1,2 Conclusions New immunotherapeutic approaches (eg, INO, blinatumomab) plus measurement of MRD bring hope for improved outcomes in adult ALL These agents are being tested in the front-line setting with encouraging results that will be hopefully confirmed in phase 3 randomized trials Addition of immunotherapeutic approaches to TKIs shows promise in Ph+ ALL and may lead to minimal use of chemotherapy New antibody constructs hold the hope for continued advances in the treatment of ALL and other diseases Just the last thing I want to say is that I think we re moving to an era I sort of compare ALL to the analogy of APL, where we re not giving chemotherapy to low-risk APL patients anymore. I think the time could come in ALL where that same thing might be happening. So, I think it s a very exciting era. Thank you. 1. Maude SL et al. Blood. 2015;125: Image copyright Sue Seif. Dr. Grupp: We re running a bit behind, but I m going to spend the next 25 minutes talking about CAR-T cell therapy. We ve had an admirable introduction to the concept of various kinds of immunotherapy, and this is sort of kicking it up a notch, taking the antibody concept, employing it in a T cell context, directly forcing interactions between T cells and cancer cells. And all of you are familiar with this concept in a general sort of way. So, the CAR we have a tumor cell that expresses a target. We have a T cell that can t see it. We force that T cell to see it with a CAR. You can use a binder of various sorts, but everything up until now has been a fragment of a monoclonal antibody. So, anything you have a monoclonal antibody to, we can build a CAR from. That part is actually pretty easy. There are a variety of signaling aspects that are important to the T cell, the primary signal always delivered by CD3 zeta. And then a costimulatory signal that these days, at least, is being delivered by either CD28 or 4-1BB, in the case of the CAR that I ll be spending more of the time talking about today. And there are advantages and disadvantages to either of those. 32 Go online to complete the post-test and evaluation for CME credit

33 Redirecting T Cell Specificity in CTL019 Cells Goals for modern, highly active cell therapy Proliferation: high level of in vivo proliferation correlates with high response rates Persistence: longer term persistence may allow longer term disease control Length of persistence needed for long-term disease control is unknown Lentiviral vector CD19 CTL019 cell Native TCR Anti-CD19 CAR construct How you make these cells really matters, and so a lot of the studies that have shown really significant clinical results to date, not all, have used some variation of this immobilized CD3/CD28 that provides both signal 1 and signal 2 to T cells. This is a much better way to grow T cells, and any other way to grow T cells has to recapitulate this aspect, which is very significant expansion but with maintenance of early T cell memory states, younger type T cells that are really the ones that are allowing for this great proliferation and potential persistence. Dead tumor cell I think the key point here is that a T cell has got to grow for patients to respond, so proliferation is absolutely required, at least in the hematologic malignancies, to see significant clinical responses. Before 2010 we didn t see proliferation after we put CAR-T cells into patients, and we did not see responses. And once we started seeing proliferation, we started seeing very high response rates. So, that is sine qua non, you have to have proliferation. And then what we argue about a lot is does it matter if the T cells last? Is 4 to 6 weeks enough? Is 6 months enough? Is 6 years enough? And so there is a lot of discussion back and forth about the characteristic of T cells that we admire, which is that they can stick around for a long time, and we don t really know how long you need to have a reasonable likelihood of actually curing these patients or at least achieving a long-term disease control. Process of making these cells is complex but we ve learned how to do it, which is kind of cool. So, we ve learned how to do this at scale. We ve learned how to do this across multiple countries, and we ve learned how to do it in an industry context. So what we re doing is we re taking T cells this is an autologous product, this is not an allogeneic product we are taking T cells from the patient. We re exposing them to a lentivirus, which is a modified form of the HIV virus that only has left in it the property of being able to modify T cell genomes, which is something that lentiviruses, a.k.a. HIV, are awesome at. These viruses don t grow. They re not in the patient. They can t cause HIV; it s impossible. This then drives the expression of a protein on the cell surface, which is the CAR protein, and that drives two things subsequently. One is hopefully an interaction with the cancer cell that kills it. It s certainly what we re looking for. But without T cells growing, again, you don t get the second thing that really matters, which is proliferation. So, you put in one cell, you get a thousand cells and that s why patients really respond. So, you want longer telomeres and you want to minimize the phenomenon of replicative senescence. If you make 10 billion awesome short-lived T cells, it s not enough to get a really good therapeutic effect. So this idea has been around for a while. There are papers in the late 80s, early 90s that were talking about this CAR concept without necessarily calling it a CAR per se. It was just this idea of whether we can assemble all the pieces needed for T cell activation and force an antibody-based recognition, which is the whole CAR idea. In 2010, the group at U Penn, Dave Porter, Carl June started looking at this in CLL and saw dramatic responses in three adults with CLL, two of whom remain in remission to this day. This then led to the first pediatric study where we treated the first couple of patients with ALL, showing ability to get resistant disease in remission, so that was exciting, followed up by the observation that there was a 90% complete response rate and that a number of these patients were having long-term remissions without further therapy beyond the CAR-T cell. So, that was all sort of what really got the ball rolling. (Not Your Typical) Treatment Process Screening Leukapheresis Enrollment Manufacture Stabilization Infusion CTL019 Now, if you guys do this, you re very familiar with it. If you re not doing it, you may be doing in the future. It is a little bit complicated to sort this all out. First thing is we need cells, right, and so if you give a patient a bunch of chemotherapy and you don t give them any window for count recovery, you don t have any T cells. And there s nothing magic about this; you have to get some T cells out of the body to create the T cells you need for the therapeutic product. 33

34 If they have a white count of 0.1, it s going to be pretty hard, so you need some window to actually collect these cells. And that s why thinking of apheresis early is really important, because by the time patients have had multiple cycles of intensive chemotherapy it gets harder and harder to collect cells in these patients. If you then have collected the patient, you re setting them up either for a clinical trial or you re thinking about now a commercial product and you re setting up manufacturing, which is going to take 3 to 4 weeks. While you re setting this up, you need some kind of bridging strategy to get them to this point. Because there is now commercial manufacturing available, this period of waiting to actually make the product not how long it takes to make it but how long it takes to get there, which is really the main source of delay has gotten much, much better. And then once you have a manufactured product, you have to set them up for infusion. Patient Stabilization Time from screening to treatment weeks to months Goals Prevent rapid progression Avoid organ toxicity and infectious complications NOT to induce remission or reduce disease burden Options Maintenance chemotherapy (steroids, monthly or weekly VCR, +/- oral 6-MP and MTX) Hydroxyurea No chemotherapy if low white count and asymptomatic Now the stabilization this is counterintuitive to anybody who is used to the ALL world, to taking patients and beating the stuff out of them to try to get them into an MRD-negative remission or below that magic 0.1% so that they can be a good transplant candidate, at least in the pediatric world. We ain t looking for this. Keep the lid on, keep the disease from exploding, keep from hurting your patient, okay. If you give three cycles of high-dose AraC and clofarabine and then they come in with an active Aspergillus infection, we can t give them CAR-T cells. If on the other hand, they come in with 5% blasts from the bone marrow, that s catastrophic from a transplant standpoint, but it s no problem in CAR-T cells. That s a perfectly great patient to treat. Lymphodepleting Chemotherapy Given 1 week prior to infusion Purpose Disease control Induce lymphopenia to facilitate engraftment and homeostatic expansion of CTL019 T cells Agents Cyclophosphamide 500 mg/m 2 IV daily days 1-2, fludarabine 30 mg/m 2 IV daily days 1-4 We do give lymphodepleting chemotherapy. It s almost always the combination of cyclophosphamide and fludarabine. This is moderate dose IV-based chemotherapy that can be given in an outpatient setting. It s not the more intensive cyclophosphamide that you might be used to from some of the NCI type studies. So, this is 500/m ² for 2 days of the cyclophosphamide. And this is decent ALL therapy, so it s going to control the disease for a period of time and that s definitely a welcome feature. But the main intent is to induce lymphopenia, to allow the T cells to grow the best, and so having less T cells in the patient allows for more room and more cytokines for the infused T cells. And that s why this seems to be important. And this issue of hematologic malignancies has been rehashed over and over again, and a lot of the time we end up back in the same ballpark of flu-cy being a pretty decent lymphodepletion regimen. It s not magic but it seems to work pretty well. Proliferation Is Key to Highly Active CAR-T Cell Therapy 1 d+1 d+15 d+21 Persistence of CTL019 and B-cell aplasia out to 5 years in responding patients 1. Grupp SA et al. Best Pract Res Clin Haematol. 2014;27: And sometimes we find ourselves talking to our referring colleagues and say, Hey, it s cool. You don t need to give highdose AraC. We re not trying to get a deep remission, we just want to maintain that patient. And often oral chemotherapy is enough for some of these kids, but it s so individualized and this is why the referring doc needs to know what the goal is. But they know their patient and they know how to treat them best. So this is what the proliferation actually looks like. This is very typical for the patients that we treat, not the best one we ever saw. So what we re looking at is T cells here. And back in the day when you weren t seeing highly proliferative cells, you wouldn t see these cells by flow, you d be looking for them by PCR. You d see 2 or 3 copies/mcg genomic DNA, and that is just above the noise, and it s not really indicating CAR-T cell proliferation. 34 Go online to complete the post-test and evaluation for CME credit

35 This is pretty easy to convince people that this is proliferation. So, these are CD3 cells, T cells. These are the CAR-T cells, which are not there. Two weeks later though, you have significant T cell proliferation in the tissues. They re starting to spill out into the peripheral blood, and now we re seeing two-thirds or more of the circulating T cells being these CAR-T cells. So, this is enormous, 1,000, 10,000-fold proliferation of these cells. And this is really, at least in hematologic malignancies solid tumors may be totally different but at least in hematologic malignancies it s really what s required for a therapeutic effect. If you see significant proliferation, at least with ALL, that patient will almost certainly go into remission. If you don t, they won t. Remission Rate by Disease Burden 1,2 a <0.01% MRD by flow cytometry. 1. Maude SL et al. N Engl J Med. 2014;371: Maude SL et al. ASCO Abstract Patient population 2nd relapse or refractory Majority refractory to multiple prior therapies ~2/3 relapsed post-allohsct Outcomes 93% (56/60) CR at day % CNS remission Now all of you realize if you take care of ALL patients, and even if you don t, that CNS disease control is absolutely required for longterm disease control. We didn t start curing patients with ALL with chemotherapy until we addressed the CNS. So, we do have some experience now. We have patients who were CNS3 in the past, we have patients who were CNS2 at the time of infusion, and actually we have patients who are CNS3 high-level CNS disease at the time of infusion, and we get those patients into remission. We have not seen any CNS relapses in these patients, which we find very interesting. I can give you an example of a patient that I saw in clinic a couple of weeks ago. This young man had Ph-positive disease, had relapsed, undergone allogeneic transplant, had relapsed in the CNS. Had relapsed in the CNS six more times after a variety of different therapies. Had received definitive radiation, had relapsed. Got CAR-T cells 3.5 years ago, remains in remission with no further therapy. That means no intrathecal methotrexate, no further radiation, just CAR-T cells. So, these do go to the spinal fluid. We see that in 98% of the patients. We see patients with CNS3 disease and lesions in the brain, where the lesions in the brain go away. Do I know the CAR-T cell went there? Well, I ain t doing biopsies, so I don t know that. Now, there is a lot of discussion about how effective these immunotherapies are in higher disease burdens, and we know that, as an example, blinatumomab in patients with high-level disease, over 50% blasts, only works less than 30% of the time. So we ve done an analysis, this is actually work that Shannon Maude did looking at patients we treated with a variety of disease burdens after lymphodepleting chemotherapy. So at the moment the T cells went in, how much disease did they have? And so we have patients who entered as MRD-negative CR just with the chemo, so they re obviously in remission to start with, so that s not much of a challenge. But for patients who have MRDlevel disease, less than 5% blasts, 88% CR rate, and with patients with more than 50% blasts this is really refractory disease 83% CR rates. So, we re able to get on top of very significant levels of disease. CTL019 Outcomes in ALL With CNS Involvement 1,2 Although it s very interesting that you can MRI these patients during the height of the T cell proliferation and the inflammatory response, and you get classic immunological pseudo-progression, where the lesion looks worse and the doc on the floor is like, Okay, time to radiate the kid, and we re like, Well, can you just wait a week? And a week later their symptoms are gone, and 2 weeks later you MRI them and the lesion is almost gone, and a month later the lesion is gone. So, clearly the T cells do cause more inflammation. They do cause more symptoms temporarily, and then the patient then responds. Cytokine Release Syndrome (CRS) CRS is related to T cell expansion and may be necessary for efficacy Symptoms typically occur 1-14 days after CTL019 cell infusion in ALL Fever Myalgias Nausea/vomiting Hypotension Respiratory insufficiency Renal insufficiency Coagulopathy CNS cohort 16 patients CNS3 within 12 months of infusion Ranged from 1st to 6th relapse pre-ctl019 4 patients CNS2 and 2 patients CNS3 on day -1 5 achieved CNS remission, 1 not evaluable 11/15 in continuous CR as long as 2.5 years (4 BM relapses) Severity scales with disease burden Entire cohort 98% of all patients treated have CTL019 detectable in CSF LBs out to 1 year demonstrate CTL019 in CSF 0 CNS relapses 1. Maude SL et al. N Engl J Med. 2014;371: Maude SL et al. ASCO Abstract

36 Some degree of CRS is probably required for a therapeutic response, especially if the patient has any significant disease burden. The patients who aren t getting a fever actually, their parents, I mean, I m a pediatrician so I m talking to the parents most of the time the parents are saying to me, I m so worried because Johnny looks great. And you re like, Oh, God. Really? Is that really where we are because people expect you to get sick. But the reality is that at least some degree of fever is really expected in patients who have any kind of leukemia that we want to control. This happens typically within about 3 or 4 days but it can be late, day 10, day 11. By day 14, really almost everybody. We admit them for their fever. We treat their myalgias. They sometimes need narcotics. You can be pretty achy with this stuff. This is the mild form of CRS. They re on antibiotics. But the ones who are getting sicker, they require blood products. They require oxygen. They may require a short period of mechanical ventilation, and they almost universally require pressors. That really is the dividing line between not-so-severe CRS and severe CRS. Severe CRS: High Ferritin Levels Suggest Macrophage Activation Syndrome 1 Now, this is a spectrum, and it goes from patients who are admitted to the hospital with fever, they get myalgias, they get nausea, vomiting. Admitted to the hospital, okay, what does that mean? It means when we give these cells, we give the patients their cells in the clinic. So, 200 kids at CHOP, 95% of which have been infused in the clinic. It is perfectly safe to give the cells in the clinic. 1. Maude SL et al. N Engl J Med. 2014;371: And the reason why it s safe is that the first thing that happens to these patients is they get a fever, and they get admitted for their fever. So this is why we re able to do this. If they get sick, which some of them may, they get sick after their fever and they re already in the hospital. But in patients who get really sick, those are patients with high disease burden, now we re seeing ICU-type problems: hypotension, unstable hypotension requiring one or two or three pressors; respiratory insufficiency, which is usually the result of the fluid resuscitation that occurs in the first 12 hours this is temporary and it goes away. We do see renal insufficiency in some of these patients, and significant coagulopathy associated with very low fibrinogens in about 10% or 15% of them, sometimes higher if you have a higher disease burden population. CRS Management Interesting technical point, I find this fascinating. This is work that David Teachey did at our hospital where these patients in addition to and probably completely overlapping with their cytokine release syndrome have a macrophage activation syndrome. They re two sides of the same coin. And so what we see is that in an analysis we did in a number of different settings, we looked at the patients who didn t go to the ICU, "No," versus the patients who did go to the ICU and have severe CRS, "Yes." And what you re seeing is ferritin as a marker of macrophage activation syndrome. And so this is elevated in a lot of patients, so this is 5,000 here, this is a log scale. Here we go 100,000, 500,000. These patients with the severe CRS have a severe macrophage activation syndrome. These are not numbers that you typically see. Cytokine Release Syndrome Associated With IFN-γ and IL-6 1 Symptom Management Antipyretics Analgesics (often narcotics required) Antiemetics, TPN Blood products (FFP, cryo) O 2 Vasopressors CPAP, ventilation IL-6 IFN-γ IFN-γ 1. Maude SL et al. N Engl J Med. 2014;371: Again, how sick you get is really dependent on how many cells you have how many leukemia cells you have and how much the T cells have to grow to get on top of your disease. So what do we do for this? 36 Go online to complete the post-test and evaluation for CME credit

37 The key point, and the only reason we re talking about this this evening, is that there are very characteristic cytokine elevations that go along with this cytokine release syndrome. And so one of the main ones is interferon gamma, and you expect that s an effector cytokine that s being produced by the T cells. Okay, no problem. And you see again, the patients who go to the ICU, they have 100-fold higher interferon gamma level, which goes with higher level T cell activation, which you totally see in these patients. High disease level, high CRS, high T cell proliferation, high interferon gamma, okay, I get it. But what s this? Well, interleukin-6 is not made by T cells and that is the big surprise here. So, interleukin-6 is something we saw on our initial cytokine analyses, also 100-fold higher in patients who get sick, part and parcel of the macrophage activation syndrome and, fortunately, druggable. IL-6 receptor antagonist Blocks IL-6 mediated effects Indicated in Juvenile idiopathic arthritis (JIA) Rheumatoid arthritis (RA) In Japan, indication for Castleman s Disease Given once or twice Rare side effects of transaminitis and neutropenia Now indicated for CRS treatment (8/30/17) 1. Grupp SA et al. N Engl J Med. 2013;368: Tocilizumab 1 So when this works, it works very well. It works in almost all patients, but sometimes it s a hectic 2 or 3 days while the patients are getting better. You get a patient with high spiking fevers, then the fevers level out at a high level. So, in this example, 104, I m more used to centigrade so 40, 40.5, 41. Some of these patients have really impressive fevers. In kids, that s no big deal. You know, my adult colleagues remind that in a 70-year-old, that s actually a medical emergency. You give them tocilizumab when they start getting a soft blood pressure, and you see often within hours the fever is gone. That s a great sign. The patient is weaning on their pressors if they re already on them, and you get the patient looking much better in a very short period of time. Typically, the protocol we re using and what s recommended in the package insert is dose the tocilizumab, not steroids first, but you guys already knew that. If 12 hours later you re still febrile and still on significant pressor support, throw in steroids for a couple of days; that s not going to harm the CAR-T cells. And if that doesn t work, 12 hours later a second dose of tocilizumab and that takes care of 85%, 90% of the patients. Beyond that, it gets a little improvisational. And actually, that drug siltuximab is used in sort of a fourth-line setting. Correlates With Severe CRS: Disease Burden Highly Predictive 1 Baseline Disease Burden So, it just so happens there s this drug called tocilizumab, which blocks the receptor, and that s actually a key point; it blocks the receptor so you have a vast excess of ligand, you have a limited supply of receptors and we re blocking the receptors, so a very high IL-6 level can still be controlled. This is obviously used by the rheumatology folks for rheumatoid arthritis, for juvenile idiopathic arthritis. We used it in the first patient we ever treated, who had grade 4 and absolutely lifethreatening macrophage activation syndrome and CRS, with dramatic turnaround within hours. Cytokine Release Syndrome: Treatment With Tocilizumab Anticytokine Therapy 1. Maude SL et al. N Engl J Med. 2014;371: These are the data that actually demonstrate what I said about disease burden. So this is a rough estimate of disease burden based on bone marrow blasts. And what we see is that over 50%, most of these patients go to the ICU. Under 50%, a lot of them don t. 37

38 Toxicity Summary 1 Cytokine release syndrome Correlates with T cell proliferation and efficacy Severity related to disease burden Reversed with tocilizumab (siltuximab is 4th line therapy) Severe CRS mirrors HLH/MAS Neurotoxicity Seen in several CD19 immunotherapy trials: CAR-T cells (NCI, CHOP/UPenn, MSKCC, Seattle) and blinatumomab In our experience: generally untreated, fully resolves Chronic B-cell aplasia requiring IgG replacement 1. Maude SL et al. N Engl J Med. 2014;371: ELIANA: Primary Efficacy Analysis 1 Parameter Efficacy Analysis Set a (n = 63) Primary endpoint % (n/n) 95% CI P Overall remission rate (CR + CRi) within 3 months 83 (52/63) (71-91) <.001 d Best overall response, % b CR 63 CRi 19 Secondary endpoint Best overall response of CR or CRi within 3 months with MRD-negative c BM 83 (71-91) <.001 d Primary efficacy analysis consistent with interim analysis where primary endpoint was met a Patients infused with CTL019 3 months prior to data cutoff. b The response was unknown in 6 patients. c MRD negative = MRD <0.01%. d Nominal P value presented to test null hypothesis of overall remission rate <20% for comparison with historical control. 1. Buechner J at al. EHA Abstract S476. Neurotoxicity, interesting. So, CRS is CRS across these different platforms and it s managed in a very similar way. But neurotoxicity can really differ between platforms. And what we see with the 4-1BB CARs is a confusional syndrome in 15%, 20%, 25% of patients, rare seizures that are treated by anti-seizure medicines that go away. What we don t see is cerebral edema and obviously, that s a big concern in some of the CD28 CARs. Seeing very similar results, 83% CR rate, all MRD negative, well above the threshold in this single-arm, phase 2 trial to convince the Food and Drug Administration that we might be on to something. CTL019: Efficacy Outcomes in ALL Studies 1 However, these CARs are persistent. They do last and while they last, these patients are in complete B cell aplasia. They don t have a single CD19 cell and those patients do require IVIG replacement, at least the pediatric patients do. The adults are really arguing about whether that s actually necessary or not, and they may have a better CD19-negative plasma cell burden and don t need that. Global, Multicenter ELIANA Trial: ALL Registration Study Duration of Response in B2202 and B2101J 1. Buechner J at al. EHA Abstract S Maude S et al. ASCO Abstract RFS 12 mo: 60% (95% CI, 48-75) 24 mo: 53% (95% CI, 39-70) 24 patients in continuous remission 1 year, 19 without further therapy 7 patients proceeded to HSCT, 1 to DLI 2 relapses after HSCT ELIANA is a single arm global study with centralized manufacturing of tisagenlecleucel 25 sites in 11 countries across North America, Europe, Australia, and Asia Interesting to me is, you know, we had a single-institution trial at CHOP where we have longer follow-up data, the sense that there is a plateau. There are not many later events. There is one this is in this data set, subsequently there has been a later event so I don t want to misrepresent that. FPFV=8 APR 2015 Data cutoff: 23 NOV 2016 This led to a registration trial, which we presented actually at ASH last year. And this was 25 centers in 11 countries, EU, Canada, Japan, Australia. So, we really were able to do this all over the world with global logistics, which was very exciting. But we re seeing a lot of patients with long-term disease control, actually 90% of whom don t go on to transplant. But if you compare that to the curve with the registration trial, visually they look the same. Is this entirely statistically illegitimate? Yes, it is but it just, you know, gives you a sense that these are roughly the same. And clearly, we need to see longer follow-up in the registration trial data to see if it s actually going to look like the single-center data. 38 Go online to complete the post-test and evaluation for CME credit

39 Mechanisms of Relapse CD19+ relapse due to short persistence CD19: relapse due to antigen escape T cell intrinsic? Immune-mediated rejection? Is CD19 deleted/mutated/no longer expressed? Relapse happens. Not all these patients stay in remission. The probability of duration of remission is somewhere in the 55%, 60% range at a year. And when patients lose their T cells quickly, which most do not, then they re at risk for CD19-positive relapse. But if they maintain their CARs, they still are at risk for CD19-negative relapse, and of the relapses that we get, two-thirds or more are CD19-negative and not CD19-positive, and that s how the ALL learns how to avoid the T cells. And so we obviously may need another approach to this by using a second antigen to try to deal with that. That s under active development at a number of centers. But the FDA then approved a second cell therapy, as all of you know, which is axicabtagene ciloleucel for adult diffuse, large B cell lymphoma, so now there are two approved products. What Are the Current Labeled Indications for CTL019 (Tisagenlecleucel)? 1 ALL up to age 25 Refractory or second relapse Other key points Patients do not need to be in complete remission to receive treatment No donor is required There is a plan for a registry No requirement for RCL testing 1. Kymriah (tisagenlecleucel) Prescribing Information. Accessed November 17, The indications are ALL up to age 25. Now, here s the key point. You re used to prescribing off-label or at least if you re a pediatrician, almost all of your practice is off-label. That ain t happening, all right. This is a unique circumstance where the drug company knows about your patient when they provide the drug back to the wholesaler to send the drug to you. So, if you send them a 26-yearold s T cell, they will not take it. So, this notion of being able to prescribe off-label, the patient doesn t have ALL, the patient s 30 not possible. So, important this is not transplant. Patients do not need to be in remission. They obviously don t need a donor. And there s no requirement for 15-year testing, which, you know, thank goodness that actually came out of the discussions. Other CAR-T Cell Constructs in ALL: KTE-C19 and JCAR017 So, this resulted in the FDA approval of tisagenlecleucel at the end of the summer this year, August 30. So, that was the first approved cell therapy. KTE-C19: anti-cd19 CAR-T cell therapy; tested in phase 1 ZUMA-3 and ZUMA-4 trials in adult and pediatric relapsed/refractory ALL 1 82% (9/11) achieved CR/CRi 100% responders tested negative for MRD 38% (5/13) of patients experienced grade 3 CRS JCAR017 explored in a phase 1 study for pediatric and young adult patients with relapsed/refractory CD19-positive ALL (N = 45) 2 MRD-negative CR rate was 93% Severe CRS: 23% 1. Shah BD et al. ASH Abstract Gardner R et al. ASH Abstract 219. KTE-C19 (axicabtagene ciloleucel) is the second gene therapy approved in the US There are other CAR-T cells constructs that are being tested. Axicabtagene ciloleucel is being tested in ALL with very similar induction/remission rates but less persistence. And if you want to keep patients out of transplant, maybe that s a useful factor. There is a defined composition product called JCAR017, where they do use a 4-1BB signaling domain and may have potentially a little bit lower CRS and decent persistence. 39

40 Autologous vs Allogeneic CAR-T Cell Therapy Various approaches to CAR-T Cell Therapy Symposium Summary and Audience Q&A Daniel J. DeAngelo, MD, PhD Stephan A. Grupp, MD, PhD Mark R. Litzow, MD Charles Mullighan, MBBS, MD Primary results (phase 1): UCART19 active, no unexpected toxicities in adults with CD19+ R/R B-ALL 1 1. Benjamin R et al. ASH Abstract 887. Question 1 I ll mention that several companies are now trying to develop off-the-shelf approaches that use allogeneic T cells from a third party using gene editing, getting rid of the T cell receptor plus other molecules. That s not at the point where we really know if it s going to be comparable to autologous T cell therapies, but they exist and it is a possibility. So, with that, I just want to thank you all for your attention on this and also for the opportunity to be with these great folks talking about the variety of approaches to ALL therapy. And, of course, need to mention the many people involved like Carl June and the other group at the University of Pennsylvania and Children s Hospital Philadelphia. So, thank you. Are there any data suggesting that tocilizumab affects the response to therapy? What dose of tocilizumab do you use, both for the initial and/or the second dose? Dr. DeAngelo: We do have a lot of questions that have been coming through, and since Steve just gave a fantastic review of the current state-of-the-art of CAR-T cell, we ll start with Steve. Two questions that came through both on tocilizumab: Are there any data that tocilizumab affects the response to therapy? What dose of tocilizumab do you use, both for the initial and/or the second? Dr. Grupp: Yeah, great questions. And so the answers are the use of tocilizumab for high-grade CRS doesn t seem to interfere with proliferation at all. What we worry about is giving it earlier, although it s probably fine there too, by the way. We have fewer data to prove that. Patients, at least in our CHOP study, who get tocilizumab for true grade 4 CRS have 100% CR rate, because their T cells are growing. That s why they re getting sick. So, the answer is it doesn t interfere. What we worry about is long courses of steroids. We definitely had a patient early on with CLL who got 2 weeks of steroids and that knocked out their CAR-T cell, knocked out their clinical response. So, a lot of steroids are not a great idea, but a day or two of steroids around peak proliferation, again, no problem. The dose is the labeled dose, for the first dose, for the second dose. So, use 8 per kilo or 12 per kilo for patients under 30 kg of weight, and that is the listed dose. 40 Go online to complete the post-test and evaluation for CME credit

41 Question 2 Is there a cutoff for when to start tocilizumab that you're using? Also, when do you add corticosteroids and how much? Does it affect response? Question 3 At what time points do you think MRD should be tested in adult patients, both with Ph-positive and Ph-negative disease? When would you skew a patient toward transplant? Dr. DeAngelo: So, just to follow up, so you said you re not giving early and you kind of threw that in that you re starting it with the first fever. So, is there a cutoff that you re using? Then the second question, if I may, is corticosteroids. When? How much? Does it affect response? Dr. Grupp: Yeah, so okay, so first off, our current trigger and what s recommended in the [tisagenlecleucel] REMS is somewhere around the time you re starting pressors on this patient. So, what happens is they develop unstable hypotension. They get their first bolus; they re still hypotensive. They get their second bolus; they re still hypotensive. At that point in our hospital, they get placed on pressors and that s typically where we re intervening with tocilizumab. We re not giving it at the time of fever. We are doing an experimental trial where we re giving it early. It s in a preemptive strategy with that first fever that you mentioned. We don t have data yet to say whether that works or not. Steroids are second-line, so if you don t see a response in 12 hours, 12 to 18 hours, you re not going in the right direction, especially if the fever is not gone. And then you put that patient on typically everybody does something different, some people like dexamethasone, some people like hydrocortisone. What we ve used is 2 per kilo of methylprednisolone per day. And again, we have that for a day or two. If that doesn t work in 12 to 18 hours, that s when we give another dose of tocilizumab at the same dose. And by then we ve sorted out almost everybody, not 100% but almost everybody at this point, even with really bad CRS is sorted out at that level. Dr. DeAngelo: All right, thank you. Charles, a question for you about MRD analysis. So, it s kind of a multi-phased question, but at what time points do you think MRD should be tested in adult patients, both with Philadelphia-positive as well as Philadelphia-negative, and when would you skew a patient toward transplant? Dr. Mullighan: I don t think there s a clear answer to that, the first question, because if you look at the practices of the major groups, they varied MRD measurements from as early as 6 weeks, even earlier measurements, out to several months. Ideally, the MRD measurements will be in the context of a trial so that there will be enough patients accrued with a similar approach to make that sort of judgment. I m not sure if, you know, Mark, you want to add anything to this as well. In terms of MRD predicting transplantation outcome, there are different results from the adult setting and the pediatric setting. In the adult setting, again there is some diversity in the results that have been seen. I mean, there is evidence, for example, from the French in a large study from the GRAALL group that there is no benefit to allogeneic transplantations in people with Phnegative ALL that achieve an MRD-negative response because any improvement in outcome is counterbalanced by an increase in toxicity. However, patients that have an MRD-positive outcome do have an improvement in outcome if they re then directed toward transplantation. There is also some evidence that the rapidity and depth of response, and I mentioned this in the context of the novel therapies but also there are some data both from Ph-positive and Ph-negative leukemia that some studies have found that there is a subset of patients that achieve MRD negativity very early, you know, within a couple of weeks, that may have a particularly good outcome. So I m sorry; I don t have a simple answer to this because practices are still so divergent and there s not been really a head-to-head comparison. But the bottom line is that really the main factor is the achievement of an MRD-negative response of at least that level of 10-4 as most predictive of subsequent good outcome. 41

42 Dr. DeAngelo: Thank you very much, Charles. Question 4 What about sequencing novel antibodies after transplant for high-risk patients? Question 6 Do you recommend defibrotide prophylaxis for patients receiving inotuzumab therapy who are going to transplant? So a question for Mark: What about sequencing blinatumomab or inotuzumab or blinatumomab specifically after transplant for high-risk patients? We didn t really talk about that but it s an interesting question. So, the idea of trying to, you know, minimize relapse following transplant, as these bridge to therapies, we re using these monoclonals. Any thoughts on that? Dr. Litzow: Well, I don t think there s much data to guide us. I think the only situation in which I would consider doing that now would be if someone was MRD positive post-transplant, I would consider adding it, blinatumomab at that point. Obviously, that patient is at high risk to relapse and needs some additional therapy. One could consider doing a donor lymphocyte infusion in that setting as well. If a patient has achieved MRD negativity after transplant, I think there s less rationale to consider doing blinatumomab in that setting. And to follow up with a second question and I ll ask Mark as well, are you recommending defibrotide prophylaxis for patients receiving inotuzumab therapy, specifically going to transplant? The answer is no, we re not recommending it prophylactically. We re not using that. It s really not indicated. But at the first sign, when the patient smells like they re developing a VOD, we will initiate, at least at my institution, defibrotide. I don t know, Mark, what your thoughts are. Dr. Litzow: I would agree. I would agree with you. Question 7 Could transplant improve the prognosis of patients with Ikaros deletion or are there other modalities? Dr. DeAngelo: Fair enough. Question 5 What about inotuzumab in terms of relapse following inotuzumab or their CD22-negative blasts? One of the questions that came to me is what about inotuzumab in terms of relapse following inotuzumab or their CD22-negative blasts. Interestingly, in the monoclonal conjugate inotuzumab, we really did not see a CD22 negativity emerging as a reason for relapse. It was probably due to chemotherapy, the classic chemotherapy resistance. Dr. DeAngelo: So a question for Charles. There s a lot of questions in terms of some of the Philadelphia-like questions but one specific question is on the role of Ikaros deletion. You showed some data suggesting that these patients have a poorer outcome. Maybe you can augment it with MRD approach strategies, but could transplant improve the prognosis of these patients? Is that the approach that everybody with an Ikaros deletion needs a transplant or are there other modalities? Dr. Mullighan: Yes, transplantation is expected to improve the outcome of at least a subset of these patients, but I think the way the field is trying to move, both in the pediatric and adult setting, is to try and use targeted or innovative approaches to avoid transplantation and to avoid the toxicity of transplant-based approaches. 42 Go online to complete the post-test and evaluation for CME credit

43 So, the data are still very early. Most of the data that we have for TKIs are based in relapsed and refractory disease. Much of the data are anecdotal. They ve often shown spectacular responses in refractory disease, particularly for cases with the ABL class alterations that we consider as being like a phenocopy of Phpositive leukemia. The jury is still out on JAK-STAT inhibition in JAK-STAT active Ph-like leukemia as well. So, I think for the moment, most groups are trying to implement TKI therapy, particularly upfront so we re not waiting for relapse, to get that signal and see if this is a viable approach. Or there are other approaches as well. I know MD Anderson, for example, looking at using onartuzumab in conjunction with hyper-cvad, particularly in the JAK-STAT group because they have such a dismal outcome, particularly in the adult group. Dr. DeAngelo: All right, thank you. Question 8 How would you use CAR-T cells for specific CNS relapse, and would you consider an isolated CNS relapse an indication for CAR-T cell therapy? Steve, there are a lot of CAR-T cell questions and so I m trying to focus them a little bit. You mentioned a couple of cases, and there were a couple of questions. And so I d like you to elaborate on CNS relapse, so the use of CAR-T cells for specific CNS relapse. And then a second question, which is similarly related. Would you consider an isolated CNS relapse an indication for CAR-T cell therapy? Dr. Grupp: So, the second question first: if it s a second relapse? The label says second relapse. So, would I? Yes, I absolutely would, and we do use it for that purpose Dr. DeAngelo: Even if there s no systemic disease? Dr. Grupp: isolated CNS relapse, but second. The first relapse is only available in a clinical trial because that doesn t fit the label. Dr. DeAngelo: Well, let s go back to that. So the issue is isolated CNS relapse second, third relapse, you know, so well within the label, and no marrow disease, because at least in the adult studies, you were required to have marrow disease. What s the efficacy in that situation? Dr. Grupp: One hundred percent remission induction, 0 relapses. But I don t want to oversell this because we don t treat patients with more than 50 blasts in their CSF, so we want some degree of control. And if you have an overt parenchymal lesion in your brain, detectable by MRI, you will have symptom progression attributable to inflammation in that area that will last for about a week. And that has to be okay. But it has been okay and we have several examples this is all, you know, less than 10 patients in any of these categories where we had overt parenchymal disease. I had one patient who had frank parenchymal disease and sugarcoating of cranial nerves that was only responding 22 days at a time to liposomal AraC, who got a CAR-T cell product and is now 2 years out. So, I mean there s certainly the prospect for we know nothing is guaranteed, nothing is 100%, even though I just said that. But it is a perfectly reasonable thing to do in patients with no marrow disease. Dr. DeAngelo: I would assume that you need to have antigen presentation, stimulation, in order to get proliferation. So, you know, what s going on here? Dr. Grupp: Well, I think first off, you got plenty of normal B cells and that s more than enough to drive this. And secondly, we know that from a much larger group of patients who were temporarily MRD negative from their chemotherapy coming into CAR-T cell treatment. So they had plenty of disease when they were enrolled on the study. They were still not refractory. They got benefit from the lymphodepleting chemotherapy. They got CAR-T cells, they have perfectly good proliferation and just as good persistence, and then long-term disease control in the same 55% to 60%. So the reality is that even if you don t have detectable marrow disease, you get CAR-T cell proliferation. Question 9 Who do you think would be an ideal candidate for the AYA phase 3 trial using inotuzumab? How are you using the monoclonals blinatumomab and inotuzumab in your practice? Dr. DeAngelo: Okay. Dr. Grupp: Would I regard first relapse as a reasonable patient for CAR-T cell therapy? I think there it s absolutely a clinical trial question and especially in kids. The question is can you avoid radiation and transplant in those patients. But you know, I would regard that as not being a current indication. 43

44 Dr. DeAngelo: Interesting. Mark, so a question is who do you think would be an ideal candidate for the AYA phase 3 trial that you showed using inotuzumab? And to follow up the question, how are you using the new agents, the monoclonal blinatumomab and inotuzumab in your practice? Dr. Litzow: Well, I think any adolescent or young adult with newly diagnosed B-lineage ALL would be a candidate for the AYA trial. Dr. DeAngelo: I guess are you concerned about inducing VOD in this young age population? Dr. Litzow: Well, I am. I think your trial is going to show us whether that s an issue or not. Hopefully, we don t see that in the run-in phase. The vast majority of the adolescents and young adults with the excellent outcomes that we re seeing aren t going to need a transplant, so that lessens my concern with the VOD question or setting them up for the development of VOD after transplant. So I think that s less of a concern. So for blinatumomab in my practice, certainly in patients that are eligible, I m enrolling them on the E1910 trial. I m a PI for the trial, so certainly supporting that. Outside of that, certainly we re using it in patients who relapse if they haven t gone to transplant. So let s say we think they re in a favorable-risk category, they relapse, we use blinatumomab. And I think you showed that bringing these agents and then someone asked me earlier today about, whether in someone who relapses with B-lineage ALL, would you go to one of these monoclonal antibody-based therapies or would you give more chemotherapy and reserve it for later? And I think I would utilize them as the initial therapy for most relapses. You commented about patients that are MRD positive and using blinatumomab, and while I think that the German studies showed the high efficacy in that setting, we truly don t know whether that improves outcomes prior to transplant. So is somebody who is MRD negative after chemotherapy the same as someone who s MRD negative after blinatumomab? In my trial, we re trying to answer that question, so we re looking at the subgroups of patients who are MRD positive and those who are MRD negative, but it does require investigators being willing to randomize them. And not everyone, honestly, has been willing to do that. I think it is a question an uncertain question in that setting. Dr. DeAngelo: I think you raise a good point. I mean, we know there s a difference between prognostic markers and predictive markers, although they re not semantically the same. And I think the point you raise, we know that MRD positive going into transplant has a poor outcome. Charles showed that data, we ve all shown that data. The question is if you convert somebody to MRD-negative status and then transplant them, have you changed their prognosis? It s a different question that hopefully you ll address in your study as well as in the AYA study as we re looking at MRD pre- and postthese immunoconjugate therapies. Question 10 What about next-generation sequencing platforms? Is there a way to standardize NGS and PCR and flow? Charles, there are two questions on molecular analysis and I ll ask them both in a sequential fashion. The first was MRD assessment. You showed flow-based as well as PCR-based. What about next-generation sequencing platforms? And is there a way to standardize NGS and PCR and flow? Where do we go into? Dr. Mullighan: So, there are a couple of considerations here. I mean, standardization is clearly important. The methodology itself has become standardized. There are commercial approaches now that have, you know, defined sets of primers to identify the rearrangement and then approaches to do the sequencing. But then you have the benchmarking across sites and between centers. So, in one respect, this is similar to flow-based MRD or prior molecular methods of measuring MRD, that there are working groups that are looking at methods to standardize and exchange reagents, and to standardize the levels of MRD detection as has been done before. There are additional considerations, though, that one can get some surprises from next-generation sequencing-based MRD measurement. Firstly, you can detect multiple clones, you know, more clones than you might detect by flow-based measurement. You can sometimes get gross discordance. You can have someone that looks like they re doing very well in a complete flow-based MRD remission, but you can get one large clone based on the NGS measurement that might just represent a normal regenerating hematogone. So, there is this additional consideration of multiclonality and how to manage these unusual situations that people are just starting to get broader experience with, and that s just going to take time and more data. Dr. DeAngelo: What are you guys doing at your institution? 44 Go online to complete the post-test and evaluation for CME credit

45 Dr. Mullighan: So, we re having a run-in period in TOTAL XVII where we re doing both. So, we did a pre TOTAL XVII period of about 30 cases where we were confident the approach was working well. But then we wanted to roll it out in real time and show that we could deliver both clinical sequencing in a timely fashion, but also the next-generation sequencing based MRD measurement, and also compare it to flow on protocol before making a decision about whether we stop flow-based measurement in a subset of patients or maybe a subset of time points, or even set it aside completely. But the ultimate goal is we hope we can then switch completely to NGS measurement. Question 11 How are you determining who is Philadelphia-like at baseline in your center, and is this extrapolated to all pediatric centers? Dr. DeAngelo: So, I have a follow-up question: How are you determining who is Philadelphia-like at baseline in your center and is this extrapolated to all pediatric centers? Dr. Mullighan: This is a very good question, a very important question. So I ll tell you how we approach it and give you some other illustrations as well. So at St. Jude we wanted to take not just Ph-like leukemia but eventually supplant most conventional diagnostics with sequencing-based approaches. So our approach is whole-genome sequencing and RNA sequencing at diagnosis. The RNA sequencing is returned for fusions in 14 days or less, and the rest of the genomic information by day 28 to 42 or indeed less. And that window is shortening as sequencing gets faster and analysis gets better. We re not actually reporting the Ph-like phenotype. That s a gene expression profile based phenotype. We report the lesions, so we report the fusions and the other alterations, and we say this is a fusion that s associated with Ph-like leukemia or this is a JAK-STAT mutation that s seen in ETP-like leukemia as well. And we have our criteria for then implementing TKI therapy in those groups. So that s, I guess, one extreme where we have the luxury of being able to do that and not all places do. And there are other approaches. So for example, in the COG, they re currently using a tiered approach where because of the sheer number of patients they need to test, they start with detecting Ph-like leukemia from a low-density array TaqMan array, 8 to 15 genes, and then follow that up with a range of targeted testing approaches, and then finally sequencing-based approaches at our center for the minority of patients that lack a lesion in one of those approaches. So, the question really of what testing is appropriate depends on two things: their capability and the desired output. So if for example, you want to target your patients with a TKI for ABL-class fusions and you re not so interested in RUX, for instance, you can achieve that with a relatively parsimonious set of FISH-based assays or, indeed, RNA sequencing or focused RNA sequencing. If in part of a larger protocol, such as the E1910 study that Mark referred to, you want to identify all patients with the phenotype and have all of that information for research correlative objective as well, then you do need an approach, whether it s the screening approach or RNA sequencing to identify the patients upfront. Now, for the clinician that might not be associated with a trial or in a center that can do sequencing, there are some other options. There are some labs for children like Nationwide that will do this testing or there are some centers like Foundation that will do, you know, kind of hybrid-focused but broader sequencing of point mutations and fusions that would detect most Ph-like fusions as well. And as time goes on, we re seeing more and more centers starting to flip their labs toward next-generation sequencing based testing. Dr. DeAngelo: Well thank you. I know the hour is late and I want to thank everybody for their attention and for participating tonight and all the wonderful questions. I hope you enjoy the rest of ASH and the rest of the snowstorm. So I thank my colleagues and thank you. Narrator: This activity has been jointly provided by Penn State College of Medicine and PVI, PeerView Institute for Medical Education. 45

46 CME Redefining Patient Care in Acute Lymphoblastic Leukemia: The Impact of Novel Antibody and Immune Technologies on Clinical Outcomes Based on a panel discussion and on data from recent medical literature. The materials presented here are used with the permission of the authors and/or other sources. These materials do not necessarily reflect the views of PeerView or any of its partners, providers, and/or supporters. This CME activity is jointly provided by Penn State College of Medicine and PVI, PeerView Institute for Medical Education. This activity is supported by an educational grant from Pfizer. Copyright , PeerView Sign up for alerts on new clinical advances and educational activities in your specialty: New and improved source for free CME/CE