Galardy, Paul J.

PURPOSE To improve the outcomes of patients with T-cell acute lymphoblastic leukemia (T-ALL) and lymphoblastic lymphoma (T-LL), the proteasome inhibitor bortezomib was examined in the Children's Oncology Group phase III clinical trial AALL1231, which also attempted to reduce the use of prophylactic cranial radiation (CRT) in newly diagnosed T-ALL. PATIENTS AND METHODS Children and young adults with T-ALL/T-LL were randomly assigned to a modified augmented Berlin-Frankfurt-Münster chemotherapy regimen with/without bortezomib during induction and delayed intensification. Multiple modifications were made to the augmented Berlin-Frankfurt-Münster backbone used in the predecessor trial, AALL0434, including using dexamethasone instead of prednisone and adding two extra doses of pegaspargase in an attempt to eliminate CRT in most patients. RESULTS AALL1231 accrued 824 eligible and evaluable patients from 2014 to 2017. The 4-year event-free survival (EFS) and overall survival (OS) for arm A (no bortezomib) versus arm B (bortezomib) were 80.1% ± 2.3% versus 83.8% ± 2.1% (EFS, P = .131) and 85.7% ± 2.0% versus 88.3% ± 1.8% (OS, P = .085). Patients with T-LL had improved EFS and OS with bortezomib: 4-year EFS (76.5% ± 5.1% v 86.4% ± 4.0%; P = .041); and 4-year OS (78.3% ± 4.9% v 89.5% ± 3.6%; P = .009). No excess toxicity was seen with bortezomib. In AALL0434, 90.8% of patients with T-ALL received CRT. In AALL1231, 9.5% of patients were scheduled to receive CRT. Evaluation of comparable AALL0434 patients who received CRT and AALL1231 patients who did not receive CRT demonstrated no statistical differences in EFS ( P = .412) and OS ( P = .600). CONCLUSION Patients with T-LL had significantly improved EFS and OS with bortezomib on the AALL1231 backbone. Systemic therapy intensification allowed elimination of CRT in more than 90% of patients with T-ALL without excess relapse.

Abstract Mature B-cell lymphomas are common in children and adults. While increasing biological knowledge has greatly improved the classification of these diseases, the ability to stratify therapy based on this knowledge is limited. As the outcomes for patients with relapsed/refractory disease is poor, it is important to optimize the upfront approach to these patients making enhanced biomarker discovery essential. We recently found that the de-ubiquitinase UCH-L1 is strongly induced in normal GC B-cells, is highly expressed in germinal center type diffuse large B-cell lymphoma, and is a potent prognostic factor associated with poor outcomes in this disease. Using transgenic mice (Uchl1Tg), we observed that deregulated UCH-L1 drives spontaneous B-cell lymphomagenesis and significantly accelerates disease when combined with the Iμ-HABCL6 or Eμ- myc models of GC derived lymphoma. At the molecular level, we showed that UCH-L1 regulates mTOR complex assembly and promotes AKT signaling. It has been alternatively hypothesized that stabilization of mono-ubiquitin is an essential function of UCH-L1 - a mechanism that does not require catalytic activity. To determine the impact of UCH-L1 catalytic activity on lymphoma development, we generated mice carrying a catalytically inactive transgene (Uchl1TgC90A). Consistent with our prior work, we observed that UCH-L1 promotes AKT phosphorylation and suppresses mTORC1 activity towards 4EBP1 and S6K in splenocytes in a manor dependent on catalytic activity. To examine the impact on lymphomagenesis, we crossed Uchl1Tg and Uchl1TgC90A mice with the Eμ- myc model. While we again found the wild-type transgene to accelerate Myc driven lymphoma development, there was no such effect of the C90A construct, indicating that the catalytic activity of UCH-L1 is required for its oncogenic activity. To further examine its role in MYC driven lymphomagenesis we crossed Eμ- myc mice with homozygous null Uchl1 mice and found that Uchl1 loss leads to a significant improvement in lymphoma free survival, indicating that UCH-L1 is a key cooperating factor with MYC in lymphoma. Consistent with this, gene expression profiling from lymphomas arising in Uchl1Tg mice share a striking similarity to those driven by MYC and constitutive PI3 kinase signaling (MYC/P110*) but are distinct from those seen in Iμ-HABCL6 or Lig4 null mice. These data indicate an mechanistic interplay between UCH-L1 and MYC in B-cell lymphoma and suggest that UCH-L1 inhibition may be an effective approach in B-cell lymphomas expressing MYC. Disclosures Galardy: Mission Therapeutics: Research Funding; Abvie: Equity Ownership; Abbot labs: Equity Ownership.

The deubiquitinating enzyme (DUB) UCHL1 is implicated in various disease states including neurodegenerative disease and cancer. However, there is a lack of quality probe molecules to gain a better understanding on UCHL1 biology. To this end a study was carried out to fully characterize and optimize the irreversible covalent UCHL1 inhibitor VAEFMK. Structure-activity relationship studies identified modifications to improve activity versus the target and a full cellular characterization was carried out for the first time with this scaffold. The studies produced a new inhibitor, 34, with an IC50 value of 7.7 µM against UCHL1 and no observable activity versus the closest related DUB UCHL3. The molecule was also capable of selectively inhibiting UCHL1 in cells and did not demonstrate any discernible off-target toxicity. Finally, the molecule was used for initial probe studies to assess the role of UCHL1 role in proliferation of myeloma cells and migration behavior in small cell lung cancer cells making 34 a new tool to be used in the biological evaluation of UCHL1.

Abstract Deubiquitinating enzymes are increasingly recognized to play important roles in cancer, with many acting as oncogenes or tumor suppressors. In this study, we employed a bioinformatics approach to screen for enzymes from this family involved in cancer and found USP24 as a potent predictor of poor outcomes in neuroblastoma, an aggressive childhood cancer. USP24 resides in a region commonly deleted in neuroblastoma, yet was independently associated with poor outcomes in this disease. Deletion of Usp24 in a murine model resulted in degradation of collapsin response mediator protein 2 (CRMP2), a regulator of axon growth, guidance, and neuronal polarity. Cells lacking USP24 had significant increases in spindle defects, chromosome missegregation, and aneuploidy, phenotypes that were rescued by the restoration of CRMP2. USP24 prevented aneuploidy by maintaining spindle-associated CRMP2, which is required for mitotic accuracy. Our findings further indicate that USP24 is a tumor suppressor that may play an important role in the pathogenesis of neuroblastoma. Significance: This study identifies the chromosome instability gene USP24 as frequently deleted in neuroblastoma and provides important insight into the pathogenesis of this aggressive childhood cancer.

Background: The prognosis for patients (pts) with relapsed T-ALL and T-LL is dismal; the primary goal of T-ALL/T-LL treatment is to prevent relapse. AALL1231 was a COG phase 3 clinical trial that randomized children and young adults (age 1-30 years) to a modified augmented BFM (aBFM) backbone +/- the proteasome inhibitor bortezomib during induction and delayed intensification (DI) (1.3mg/m2 x 4 doses per block). Bortezomib was tested in frontline therapy based on strong preclinical data and data in relapse on COG AALL07P1. Pts were stratified as standard (SR), intermediate (IR), or very high risk (VHR), primarily based on disease response: morphologic and minimal residual disease (MRD) at end induction and end consolidation (T-ALL) and radiographic response (T-LL). To eliminate cranial radiation (CXRT) in all pts, (except VHR: Day 29 M3 marrow or EOC MRD >0.1% or pts with overt CNS leukemia at diagnosis, CNS3), the aBFM backbone was modified to use dexamethasone (dex) as the sole corticosteroid and an extra pegaspargase dose was added in both induction and DI, following the MRC strategy. IR pts received a second interim maintenance (IM) phase (one Capizzi MTX; one HD-MTX). Following consolidation, VHR pts received 3 BFM high-risk intensification blocks in lieu of IM. Results: AALL1231 accrued 847 patients (824 eligible and evaluable) of 1400 anticipated from 2014 until early closure in 2017 when COG AALL0434 established that nelarabine (NEL) improved DFS in T-ALL (AALL1231 did not include NEL). The 3-year EFS for Arm A (no bortezomib) vs Arm B (bortezomib) were 81.7±2.4% and 85.1±2.2 % (HR=0.782, p=0.074) (3/31/20 data cut-off; see Table 1 for additional outcomes). SR and IR pts, who account for 95% of pts, had significantly improved EFS on Arm B as compared with Arm A. Yet, VHR patients had improved EFS on Arm A. Patients with T-LL had improved EFS and OS with bortezomib: 3-year EFS (76.5±5.9% vs 88.3±4.5%; p = 0.01); 3-year OS (78.0±5.8% vs 89.5±4.2%, p = 0.007). A similar improvement in EFS and OS was not seen in T-ALL; however, with longer follow-up this may change. No excess toxicity was seen on Arm B. A dex-based Induction did result in lower MRD rates; more T-ALL pts on AALL1231 had Day 29 MRD <0.1% as compared with AALL0434 which used a prednisone-based Induction (AALL1231 Arm A: 69.6%; Arm B: 72.2%; AALL0434: 64.6%; p = 0.02). However, this did not translate into improved survival. Indeed OS, but not EFS was worse on AALL1231 than AALL0434. On-going analyses are investigating the increased mortality on AALL1231, but preliminary analyses suggest a combination of increased toxic deaths and overall poor outcome in the VHR group. On AALL0434, 90.8% of T-ALL pts received CXRT. On AALL1231, 9.5% of subjects were scheduled to receive CXRT (CNS3 T-ALL/T-LL: 5.7%; VHR T-ALL: 4.1%). A comparison of AALL0434 pts that received CXRT with similar AALL1231 pts not receiving CXRT on AALL1231 demonstrated similar EFS (p = 0.14) and OS (p = 0.42) (Table 2). CNS relapse rates were higher in these pts on AALL1231 (4.5%) as compared with AALL0434 (1.7%), but overall relapse rates were the same (6.5% vs 6.4%). Notably the benefit of NEL in AALL0434 was due to reduction of CNS relapses. 128 AALL1231 pts came off protocol therapy after the study was closed for physician or patient/parent choice. Data collection is underway to understand the reasons for removal, including if it was to receive NEL. Conclusions: Outcomes for SR and IR pts with T-ALL and T-LL treated with bortezomib were excellent despite the elimination of prophylactic CXRT. Bortezomib significantly improved 3-year EFS for these groups, comprising ~95% of pts. Outcomes for VHR pts were dismal and worse on the bortezomib arm. T-LL pts had significantly improved EFS and OS with bortezomib on the AALL1231 backbone. This is the first trial to demonstrate an OS benefit for de novo pediatric T-LL with a new agent; however, longer follow-up is needed. Therapy intensification allowed elimination of CXRT in the majority of pts without excessive relapse. These results should be interpreted cautiously as the 3-yr OS on AALL1231 was inferior to AALL0434. Nevertheless, incorporating bortezomib into standard therapy for de novo T-LL appears advantageous. Future COG T-ALL/T-LLy trials will build on the positive findings from AALL0434 and AALL1231, balancing intensity while mitigating toxicity to maintain high cure rates without routine cranial radiation. (MLL, SPH, EAR contributed equally) Disclosures Teachey: Amgen: Consultancy; Janssen: Consultancy; La Roche: Consultancy; Sobi: Consultancy. Dunsmore:Dexcom: Current equity holder in publicly-traded company. Galardy:Abbott: Current equity holder in publicly-traded company; Abbvie: Current equity holder in publicly-traded company. Harker-Murray:Regerenon Pharmaceuticals: Consultancy. Hermiston:Sobi: Membership on an entity's Board of Directors or advisory committees; Novartis: Membership on an entity's Board of Directors or advisory committees. Shimano:Novartis: Research Funding; Daiichi Sankyo: Research Funding; Pfizer: Research Funding; Dova Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees. McKay:Immunogen: Current Employment. Bollard:Mana Therapeutics: Other: IP. Loh:Medisix Therapeutics: Membership on an entity's Board of Directors or advisory committees; Pfizer: Other: Institutional Research Funding. Hunger:Novartis: Consultancy; Amgen Inc.: Current equity holder in publicly-traded company, Honoraria. Raetz:Celgene/BMS: Other; Pfizer: Research Funding. OffLabel Disclosure: Bortezomib for the treatment of acute lymphoblastic leukemia under an IND

Pediatric aggressive mature B-cell lymphomas are the most common types of non-Hodgkin lymphoma in children, and they include Burkitt lymphoma (BL) and diffuse large B-cell lymphoma (DLBCL). These diseases are highly aggressive but curable, the treatment is complex, and patients may have many complicated supportive care issues. The NCCN Guidelines for Pediatric Aggressive Mature B-Cell Lymphomas provide guidance regarding pathology and diagnosis, staging, initial treatment, disease reassessment, surveillance, therapy for relapsed/refractory disease, and supportive care for clinicians who treat sporadic pediatric BL and DLBCL.

Abstract BACKGROUND: Burkitt Lymphoma (BL) represents approximately 40% of all childhood and adolescent non-Hodgkin lymphoma (Miles/Cairo, BJH 2012). Children with relapsed or progressive BL develop chemoimmunotherapy-resistant disease and can rarely be cured with salvage therapy (Cairo et al, JCO 2012). Bruton's tyrosine kinase (BTK) is a regulator of normal B-cell development and is activated upon B-cell receptor (BCR) stimulation. Chronic active BCR signaling through BTK activation can be inhibited by the selective and covalent BTK inhibitor, ibrutinib (Young/Staudt et al, Nat Rev Drug Dicov 2013). Preclinical studies of ibrutinib in chronic lymphocytic leukemia (CLL) and mantle cell lymphoma (MCL) suggest that its inhibitory effect on cell proliferation is in the range of 1.0uM to 25.0uM (Herman et al, Blood 2011; Cinar et al, Leuk Res 2013), which is higher than the typical plasma concentration range in patients at the recommended dose of 560mg (<0.6uM) (Advani et al, JCO 2013). Despite these relatively high IC50 values in vitro, ibrutinib has been highly effective in the treatment of patients with refractory CLL and MCL (Byrd et al, NEJM 2013 and Wang et al, NEJM 2013). Ibrutinib was initially approved by the FDA (IMBRUVICA, USPI) for patients with MCL in November 2013 and CLL in February 2014, who have received at least one prior therapy. BL, however, is associated with tonic vs chronic active BCR signaling; while, both CLL and MCL have chronic active BCR signaling. These findings suggest that the antitumor activity of ibrutinib in BL, which is associated with tonic BCR signaling, may be less pronounced than in Activated B cell-like diffuse large B-cell lymphoma (ABC-DLBCL), CLL or MCL. OBJECTIVES: We hypothesize that ibrutinib may be a potential adjuvant agent in the treatment of BL. Therefore, we investigated the in vivo anti-tumor activity of ibrutinib in BL xenografted NSG mice. METHODS: Raji Burkitt lymphoma (BL) cells (ATCC) were exposed to ibrutinib (0-10uM, generously provided by Janssen R&D LLC) alone and in combination with dexamethasone (1uM) for 5 days and then MTS cell proliferation assay (Promega) was performed. The IC50 values were determined with CompuSynTM software (Chou and Martin, ComboSyn, 2005) based on data derived from MTS assay. Raji BL cells were stably transfected with a firefly luciferase expression plasmid (ffluc-zeo), kindly provided by Laurence Cooper MD, PhD. NSG mice (NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ, 6-8 weeks old) from Jackson laboratory were irradiated (2.5 Gy) and then mice were subcutaneously injected with one million ffluc-zeo tumor cells. Tumor progression and tumor burden were monitored by Bioluminescent Imaging system using the Xenogen IVIS-200 (Caliper Life Sciences). Mice were orally gavaged with either vehicle or ibrutinib (12.5mg/kg) for 10 days (once a day) at post-tumor cells injection. Analysis of survival rates were determined by the Kaplan-Meier method and differences evaluated by log-rank test using the Prism Version 6.0 software. RESULTS: Ibrutinib compared to control induced a significant dose-dependent decrease in cell proliferation in Raji BL cells (5uM, 0.561 ± 0.170, p<0.03, IC50=5.20uM) following 120 hours (5 days) of ibrutinib treatment. Interestingly, we observed a significant decrease in cell proliferation in Raji BL cells (5uM, 0.213 ± 0.078, p<0.002, IC=0.50 ± 0.374uM) following ibrutinib and 1uM dexamethasone combination treatment following 120 hours (5 days) of treatment. We observed a significant decrease of tumor progression by tumor luminescence signal intensity following ibrutinib treated Raji BL xenografted NSG mice at day 20 (12.5mg/kg, p<0.001 and 25mg/kg, p<0.05) and day 25 (12.5mg/kg, p<0.05 and 25mg/kg, p<0.05) compared to control (Figure 1A).Ibrutinib (12.5 mg/kg) treated mice (n=39) significantly prolonged survival compared to vehicle controls (n=28) (p<0.0001) (Figure 1B). CONCLUSIONS: Ibrutinib (12.5mg/kg) significantly decreased tumor progression and significantly increased survival in Raji BL xenografted NSG mice, suggesting that ibrutinib may be a potential adjuvant agent for therapy in the treatment of BL. Future studies will investigate the efficacy of combination therapy with ibrutinib on survival in BL xenografted NOD/SCID mice. Disclosures Galardy: Mission Therapeutics: Research Funding.

Abstract Using an activity-based approach, we previously described the over-expression of the deubiquitinase ubiquitin carboxy terminal hydroxylase 1 (UCH-L1) in several B-cell malignancies including multiple myeloma, Burkitt lymphoma, non-Hodgkin lymphoma and chronic lymphocytic leukemia. UCH-L1 over expression also correlates with a poor prognosis in non-lymphoid tumors including non-small cell lung, esophageal, colorectal and pancreatic cancers. Whether UCH-L1 expression contributes to the malignant phenotype in these conditions, however, has not been established. We therefore sought to determine the role of UCH-L1 in lymphoid malignancies through the use of lentivirusencoded short hairpin RNA (shRNA) targeting UCH-L1. When compared with a control non-silencing shRNA, reduced UCH-L1 levels are associated with a profound block in proliferation and loss of viability (>90%) in the UCH-L1 expressing myeloma cell line KMS11. The introduction of these shRNAs into the UCH-L1 negative myeloma cell line KMS12 does not affect proliferation or viability. The association of cell death with reduced UCH-L1 levels was further reinforced using a doxycycline inducible shRNA construct. KMS11 and KMS12 cells grow equally well following transduction and selection of stable viral integrants. However, upon induction of shRNA production with doxycycline, greater than 90% of KMS11 cells die within 48-hours timed with the depletion of UCH-L1 while KMS12 cells continue to proliferate. These results strongly suggest that UCH-L1 is essential for survival in UCH-L1 expressing myeloma cells. Given the critical role of NF-kB signaling in many lymphoid malignancies, we examined the impact of UCH-L1 depletion on NF-kB activity. Using a secreted luciferase reporter assay, we determined NF-kB activity in control and UCH-L1 depleted KMS11 cells. We observed a 5-fold reduction of NF-kB activity in UCH-L1 knockdown cells compared to control. In order to establish a correlation between loss of UCH-L1 expression and NF-kB down-regulation, cells transduced with an inducible shRNA were treated with doxycycline and monitored daily for UCH-L1 and NF-kB activation. We observed a direct relationship between the level of UCH-L1 and NF-kB activity in these cells with a 5-fold reduction in NF-kB activity by 48-hours accompanied by reduced viability. These effects are entirely reversible with the removal of doxycycline from the growth medium. Doxycycline induction had no effect on the NF-kB activity of the UCH-L1 negative cell line KMS12. Knockdown of UCH-L1 does not affect TNF-alpha-induced NF-kB activity in these cells suggesting that UCH-L1 modulates the non-canonical NF-kB pathway. A relationship between UCH-L1 and NF-kB is further suggested by observing a 3.8-fold increase in baseline NF-kB activity in HeLa cells over-expressing UCH-L1. These data strongly indicate that UCH-L1 enhances NF-kB through the non-canonical pathway. In order to determine whether UCH-L1 over-expression can by itself promote tumorigenesis in vivo, we have engineered a UCH-L1 transgenic mouse. These mice will provide us the opportunity to study the effect of UCH-L1 expression on spontaneous as well as induced (carcinogen or oncogene) tumorigenesis. Further studies are examining the effect of forced UCH-L1 expression on B-cell development. This study will provide a better understanding of the role of UCH-L1 in hematological malignancies and may validate UCH-L1 as an important drug target for B-cell malignancies.

Abstract Abstract 2702 Introduction/Background Tumor Lysis Syndrome (TLS) is a potentially life-threatening metabolic syndrome that frequently complicates the management of newly diagnosed mature B-NHL in children and adults (Cairo et al., BJH 2010). We have demonstrated rasburicase to be superior to allopurinol in the rapid reduction of uric acid (UA). (Goldman/Cairo et al, Blood 2001) In the previous FAB/LMB/96 international mature B-NHL trial, variables associated with a significant inferior EFS included advanced stage, elevated LDH, primary mediastinal involvement (PMBL) and combined BM/CNS disease (Cairo et al, JCO, 2011). We have previously reported the pharmacokinetics and safety of the addition of rituximab to a modified FAB/LMB/96 backbone in children and adolescents with advanced stage mature de-novo B-NHL. (Cairo et al., ASCO, 2010) Here we focus on outcome in those specific high-risk sub-categories of advanced childhood mature B-NHL following the incorporation of a recombinant urate oxidase (Rasburicase) and targeted immunotherapy (Rituximab) to the FAB/LMB/96 chemotherapy backbones. Material and Methods Children with newly diagnosed stage III/IV or high-risk (BM+±CNS+) de-novo mature B-NHL were prospectively treated with cyclophosphamide, vincristine, and prednisone (COP) reduction therapy preceded by 4–24 hours of non-alkaline hydration and rasburicase (0.2mg/kg) (generously supplied by Sanofi-Aventis). Laboratory and clinical TLS (LTLS and CTLS) was prospectively monitored as previously defined (Cairo & Bishop, BJH, 2004). Immunochemotherapy consisted of FAB B4 or C1 therapy (adriamycin 1 hr infusion) as we have previously described (Patte et al and Cairo et al, Blood, 2007) with the addition of rituximab 375 mg/m2/dose (generously supplied by Genentech) with 2 doses administered in each induction cycle and 1 dose in each consolidation cycle. Results With respect to LTLS and CTLS, 78 patients were evaluable: group B (stage III/IV) with NL-LDH (<2X ULN, n=23), Group B with HI-LDH (≥ 2X ULN, n=18), and Group C (n=37). The mean initial UA levels (mg/dl +/− SD) were 4.7±2, 8.7±4.6, and 7.1±4.6, respectively. At presentation, 31% of patients (all groups) had an elevated (> 8mg/dl) UA, and 47% received elective doses of rasburicase in the days prior to the first day of chemotherapy. The incidence of LTLS prior to rasburicase in Group B NL-LDH, Group B HI-LDH, and Group C, was 4.3%, 5.5%, and 16.6%, respectively. The incidence of CTLS prior to rasburicase was 0%, 11%, and 8.6%, respectively. Following the administration of rasburicase, the incidence of new onset LTLS and CTLS for the entire cohort was 13% (2.6% with hyperuricemia) and 4% (1% with hyperuricemia), respectively. Furthermore, there was a significant improvement in the estimated GFR in those with renal impairment from day -1 prior to rasbursicase to day+7 post R-COP (55 ± 29 ml/min1.73 m2 to 136 ± 29 respectively; p<0.0007; Figure 1). Following combined FAB 96 chemotherapy (with 1 hour Adriamycin) and rituximab, the 3- year EFS (95% CI) for pilot patients (6 doses of rituximab) with Stage III/IV and Group C BM/CNS patients were 95% (80–99%) (Figure 2) and 89% (73–98%), respectively. Interestingly, two of four patients with centrally reviewed mediastinal primaries developed recurrent disease, while none of the 19 stage III/IV patients with LDH < 2X ULN developed recurrent disease or toxic deaths. Conclusions Rasburicase and non-alkaline hydration largely prevented (94% of the time) new onset TLS with hyperuricemia safely during COP reduction therapy in children with advanced stages of de-novo mature B-NHL. For patients with mediastinal primaries, the high failure rate with group B4 therapy + rituximab suggests it would be appropriate to investigate new treatment strategies. Excluding this histology, combined modified FAB/LMB 96 chemotherapy and rituximab resulted in only 5 deaths (2 toxic deaths, 3 disease related deaths) among our cohort of 62 with advanced mature B-NHL. In comparing these current results from the previous FAB/LMB 96 trial without rituximab, the stage III/IV patients had a 95% vs. 84% and patients with CNS+ disease a 93% vs. 72% 3 year EFS. (Patte et al., Cairo et al. Blood, 2007) Future efforts to intensify chemotherapy in patients with CNS involvement should proceed only with careful toxicity monitoring, as preliminarily these patients seem to have excellent outcome with this immunochemotherapy approach. Disclosures: Gross: Pfizer Oncology: Consultancy; Roche/Genentech: Consultancy; Eli Lilly: Consultancy.