Relapsed/Refractory Diffuse Large B-Cell Lymphoma John Kuruvilla, MD Princess Margaret Cancer Centre University of Toronto
Disclosures Research Support Employee Leukemia and Lymphoma Society US, Rasch Foundation Roche N/A Consultant Abbvie, BMS, Gilead, Janssen, Roche, Seattle Genetics Major Stockholder N/A Speakers Bureau Honoraria Scientific Advisory Board N/A Abbvie, Amgen, BMS, Celgene, Gilead, Roche, Janssen, Lundbeck, Merck, Seattle Genetics, Lymphoma Canada (Chair) I will likely discuss all sorts of off-label use of agents as well as investigational agents
The older literature defined the role of ASCT Autologous stem cell transplantation (ASCT) was the standard of care following the randomized Parma trial DHAP salvage chemotherapy BEAC HDCT Hovon-44 demonstrated improved outcome with the addition of rituximab to salvage chemotherapy DHAP-VIM-DHAP Essentially no rituximab re-treatment Philip NEJM 1995; Vellenga Blood 2008
Question #1 In a healthy ASCT-eligible patient with relapsed DLBCL treated with R-CHOP that does not have CNS disease, I would use: 1. GDP because they ve had rituximab before 2. R-GDP because the rituximab can t hurt 3. R-GDP because there are data supporting the use of rituximab in R treated patients 4. R-DHAP since Health Canada doesn t believe R-GDP is a standard of care regimen 5. R-ICE (since you re going to the NYC Lymphoma/Myeloma meeting)
Question #2 Which of the following features about a DLBCL patient in the relapse setting worries you the most? 1. Age 60-70 2. High IPI 3. ABC subtype 4. MYC+ 5. Prior rituximab
Problems and Questions (2000) ½ of patients respond to therapy prior to planned ASCT Non-responders do not do well with SCT Can we improve salvage chemotherapy? Particularly if you had rituximab with primary therapy ½ of patients who undergo ASCT relapse It s not hard to identify high risk patients Can we improve the transplant regimen or consider post-sct maintenance?
Answers 2014: the salvage chemo benchmark Arm (n) ORR (%) ASCT (%) EFS/PFS (%) comment CORAL R-ICE (202) 63 51 3Y: 31 R-DHAP (194) 64 55 42, p=0.4 LY12 (R)-GDP (310) 46 46 4Y: 26 Non-inferior (R)-DHAP (309) 45 45 26, p=0.95 Improved QOL, PE ORCHARRD O-DHAP (222) 38 32 2Y:21 R-DHAP (223) 42 36 26, p=0.27 Over 1400 patients randomized, and no benefits in efficacy Gisselbrecht JCO 2010, Crump JCO 2014, van Imhoff ASH Abs 928, 2014
Answers 2014: Maintenance Rituximab Arm (n) EFS (%) OS (%) comment CORAL Rituximab (122) 4Y: 52 4Y: 52 Observation (120) 53 556, p=0.3 LY12 Rituximab (115) 2Y: 64 HR:0.82 Includes Observation (115) 51 TRFL
Impact of Rituximab Salvage in Rituximab treated patients R-Salvage Salvage P ORR 46 25 0.0003 4Y EFS 27 22 0.0954 4Y OS 43 31 0.045 ASCT rate 52 31 0.0004 This is the only prospective (non-randomized) data regarding the role of rituximab in R-pretreated patients with relapsed DLBCL. It appears justified to use rituximab in these patients. Baetz. Leukemia and Lymphoma 2016
Problem Populations: Early relapse, Prior rituximab and MYC+ FISH Relapse < 12 months MYC FISH Gisselbrecht JCO 2010, Cuccuini Blood 2012
Summary Salvage chemo and maintenance era CORAL and LY12 (chemotherapy) did not improve outcome LY12 was a positive trial due to favourable toxicity, QOL and pharmacoeconomics ORCHARRD (ofatumumab) did not improve outcome Is this a generalizable finding? Rituximab-based primary chemotherapy has made the salvage therapy arena more challenging than 10 years ago High risk patients are easy to identify but hard to treat We are in the targeted era of therapeutics in heme-onc Are we doing biomarker driven trials?
Allo-SCT in DLBCL What did we learn from DSHNHL R3? Feasibility: You can successfully transplant a younger subset of aggressive NHL with an aggressive myeloablative regimen 45% OS/PFS But only enrolled 84 patients in 5 years! But avoid mismatches, use ATG and hope for GVHD Interesting to see proof of concept of GVLY in DLBCL This contradicts some older (and registry-based) data This strategy is not applicable to the majority with this disease But new techniques will change this (haploidentical etc.) If this is of interest, somebody should do a proper RCT Glass Lancet Oncol 2014
RR-DLBCL in 2016 We have probably reached limits of conventional chemotherapy SCT (Auto or Allo) benefits the minority Novel strategies are clearly needed Therapeutics Trial design Translational research A fundamental question: do we build on the SCT platform, or do we abandon it? This is the (potentially) curative setting
The problems for development in DLBCL There are no home run drugs in DLBCL HL has brentuximab and nivolumab Multiple (moderately) active agents Drivers (mutations or otherwise) unclear Targets and activity may not be clear How do you run trials in the curative setting that are: Efficient Potentially effective (minimize ineffective)
CCTG: Design Concepts Build on R-GDP (less heme tox) Need to evaluate multiple drugs quickly Multiple arms Discard ineffective drugs early Use an early endpoint as surrogate Need control arm Minimize patients wasted with control treatment Modified version of Pick a winner from Hills and Burnett in MRC UK AML trials Hills and Burnett Blood 2011 15
Statistical Design Primary endpoint is overall response rate Will prospectively evaluate PET-CR as endpoint Designed to detect a 20% improvement in ORR for Secondary Endpoints include: ASCT rate Tolerability/toxicity Stem cell collection rate EFS and OS A Treatment Arm will warrant PIII investigation if: One-sided 90% CI for RR difference is >20% One-sided 80% CI of transplantation rate difference is > 10% Treatment arm has acceptable safety and tolerability profile
LY17 Statistical Design Assuming RR of control arm is 50%, 64 patients / arm are required to detect a 20% improvement 80% power, one-sided 90% alpha Accrual planned for 2 years Interim futility analyses are planned when 16 and 32 pts/arm are accrued Hills pick the winner design to be used 1 st IA: futility is met if RR (tx arm) < RR (control arm) 2 nd IA: futility is met if RR (tx arm) < RR (control arm + 10%) 17
CCTG LY17: Pick a winner study design CONTROL: R-GDP All patients R Stratified by: c-myc status IPI at relapse prior response duration prior PET scan R-GDP + A R-GDP + B R-GDP + C X Y Which experimental agents and why?
Chronic Active BCR Signaling Activates NF-κB in ABC DLBCL Upstream mutation of BTK suggests pathway active Can activate NFKB downstream of BTK Wilson. Nat Med 2015 Can activate NFKB independent of BTK
Would you develop ibrutinib in an unselected population? P=0.007 Wilson Nat Med 2015
Select for real ABC that is CD79B Mutant and/or CARD11 mutant Wiilson Nat Med 2015
So Arm #1 in LY17 is: R-GDP + Ibrutinib R-GDP given at standard dose and schedule with ibrutinib 560 mg PO daily Accrued (almost) first stage Some toxicity (infection, hematologic) Awaiting efficacy assessment regarding proceeding
Arm#2 in LY17 is: R-DICEP Institutional data from Calgary in 113 patients from 1995-2009 highlight: DICEP (n=93) or R-DICEP (n=20) mlos: 21days (7-51), FN 55%, infection 10% median number of CD34+ cells collected:19x10 6 /kg (0.3-142) 83.5% ORR 90% (102) proceeded to ASCT 5yr PFS rate 42% [mfu=94 mo] 32% if relapse aaipi=2-3 35% if initial TTP <1year 56% if failed initial Rituximab-induction. Early TRM 3 pts (2.7%) Late MDS/AML=2 Vijay Leukemia and Lymphoma 2013
R-DICEP then HDCT/ASCT Rituximab 375 mg/m 2 /d d1, 5 Cyclophosphamide 1.75g/m 2 /d d2-4 Etoposide 350mg/m 2 /d d2-4 Cisplatin 35mg/m 2 /d d2-4 G-CSF 5mcg/kg/d d15-20 Apheresis HDCT/ASCT Day 1 2 3 4 5 15 20 21 35-56 Courtesy Doug Stewart
What else is going on? A lot of translational biology is written into LY17 Tumour specimens Peripheral Blood Some interesting compounds to consider Venetoclax Idelalisib Selinexor Checkpoint inhibitors Lenalidomide (or CC122) Building Phase 1 to facilitate quick evaluation of these regimens
Question #3 Which of the following novel therapies are you the most excited about in RR-DLBCL? 1. Obinutuzumab 2. Ibrutinib 3. Idelalisib 4. Venetoclax 5. IMiDs 6. Checkpoint inhibitors 7. CAR-T cells
NCI Trial CAR-T Trial in REL/REF B-cell lymphoma Tumor Type (n evaluable) Overall Response Rate Complete Response Rate Any (29) 76% 38% DLBCL/PMBCL (17) 65% 35% CLL (7) 86% 57% Indolent NHL (5) 100% 25% 16 patients still in response; 12 ongoing > 1 year 3 patients were re-treated after progression; all in ongoing response (17+ - 52+ months) Kochendorfer JCO 2014, ASH 2014
NCI Trial CAR-T: Summary of Adverse Events Prominent toxicities were related to transient cytokine release syndrome, neurotoxicity and B-cell aplasia Fever, hypotension, hypoxia Generally resolved within 1-3 weeks Reversible neurotoxicity aphasia/dysphasia, confusion, somnolence, motor (tremor) Chemo-induced cytopenias No deaths on study deemed related to anti-cd19 CAR Kochenderfer Blood 2012; Kochenderfer et al, JCO 2014, Kochenderfer et al. ASH 2014
ZUMA-1 KTE-C19-101: Phase 1-2 Trial in Aggressive NHL Key Eligibility Criteria DLBCL, PMBCL or TFL Chemotherapy refractory disease SD or PD to last therapy or Relapsed post transplant within 1 year Adequate prior therapy At minimum, anthracycline-containing regimen ECOG 0 or 1 Endpoints and anti-cd20 mab Incidence of DLT (primary phase 1) Objective response rate (primary phase 2) Duration of response, PFS, OS and safety Phase 1 Refractory Aggressive NHL Phase 2 Cohort 1: DLBCL (n=72) Cohort 2: PMBCL/TFL (n=40) Conduct Proceed to phase 2 after phase 1 safety evaluation Courtesy Kite Pharma 29
Summary REL/REF DLBCL in 2016 In NHL, RCTs (Parma, CORAL and LY12) have set the stage GDP is the standard (favourable toxicity) Rituximab maintenance did not work Path forward will require well designed prospective trials Incremental benefits are likely to be modest and come with increased toxicity Some promising agents available Integration of novel agents earlier in the disease course needs to be tested
Thank You Enjoy Toronto and CHC 2016!