Citations

Abstract Vitreoretinal lymphoma (VRL) management remains a challenge. We present 72 patients with VRL, diagnosed at Mayo Clinic between 1990‐2018. Three nondiffuse large B‐cell lymphoma (DLBCL) histology cases were excluded. Among 69 DLBCL, 33 patients had primary VRL (PVRL), 18 concurrent intraocular and central nervous system (CNS) or systemic disease and 18 secondary VRL. Patients received intraocular chemotherapy (intraocular injections of rituximab or metothrexate or steroids or in combination), radiotherapy, systemic or combined systemic plus intraocular treatment in 9, 10, 35, and 15 cases, respectively. Among primary and concurrent VRL, median failure free survival (FFS), CNS relapse‐free survival (CNS‐RFS) and overall survival (OS) were: 1.8, 4.9, and 4.1 years, respectively; among PVRL, median FFS, CNS‐RFS, and OS were: 2.6 year, Not Reached and 9.3 year, respectively. No CNS relapse occurred beyond 4 years in PVRL. Median OS for patients diagnosed between 1990 and 1999 vs between 2000 and 2018 was 1.5 vs 9.4 years, respectively ( P = .0002). OS was significantly higher in PVRL, as compared with concurrent VRL ( P = .04). Previous immunosuppression and poor performance status were predictive of worse outcome. In PVRL, a combined systemic and intraocular therapy showed higher FFS ( P = .002) and CNS‐RFS ( P = .003), but no differences in OS. Among 18 secondary VRL, at a median follow‐up of 1.1 year after vitreoretinal relapse, median FFS and OS were 0.3 and 1.3 years. An improvement in survival of VRL has been observed over the decades. PVRL should undergo combined systemic and intraocular chemotherapy to prevent CNS progression.

11571 Background: Relapsed/refractory (R/R) non-Hodgkin lymphomas (NHL) have a poor prognosis with limited treatment options. Our expanding knowledge of molecular alterations seen in R/R NHL allows identification of patients that potentially may benefit from a precision medicine approach. However, experience in routine clinical implementation of precision medicine has been limited. Here, we summarize our clinical experience in molecular characterization of RR NHL targeted therapy (TT) using next-generation sequencing (NGS), and selection of targeted therapy (TT) based on molecular profile. Methods: We conducted a prospective study in RR NHL through the Center for Individualized Medicine at Mayo Clinic. Consenting patients underwent NGS using FoundationOne Heme panel from biopsies done at time of relapse. Results of NGS were discussed at the Genomic Tumor Board and recommendations for TT were given based on matching specific molecular alteration(s) with potential agent(s) predicted to be active based on NGS. The agents could include FDA-approved, off-label use and clinical trial therapies. Results: 28 cases were enrolled: 18 aggressive NHL, 10 follicular lymphoma (FL). Molecular alterations were present in all cases. In aggressive B-cell NHL, CDKN2A/B gene cluster alterations were seen in 73% (8/11), while seen in only 1/7 T-cell lymphomas (TCL), and 1/10 FL. TP53 deletions were second most common genomic alterations in DLBCL (57%) and seen in 40% FL. JAK-STAT and ERBB pathways were altered in TCL (2/7 each). IGH-BCl-2 gene rearrangement were common in FL (70%), followed by MLL gene alterations (50%). Targetable mutations were present in 86% (24/28) of cases. A TT was recommended in all 24 cases, but received by 2 patients only. Remaining patients did not due to benefit from current therapy (10/24), ineligibility or lack of clinical trial (7/24) or interim clinical deterioration (5/24). Conclusions: Targetable mutations were identified in most cases of RR NHL with TT recommended for all cases. However, access to TT limits potential clinical benefit of molecular-based matching strategy. More studies are needed to assess impact on clinical outcomes.

TPS7573 Background: The PD-1/PD-L1 pathway is an important immune checkpoint used by tumor cells and may provide an effective target for enhancing anticancer immune response. In DLBCL, high PD-L1 expression has been identified as a negative prognostic factor for overall survival. Durvalumab (MEDI4736) is a high-affinity human IgG1 monoclonal Ab that selectively blocks PD-L1 binding to PD-1 and CD80, and preliminary preclinical and clinical activity support the study of durvalumab in high-risk DLBCL subtypes. The primary study objective is to explore the clinical activity of durvalumab with R-CHOP in non-activated B-cell–like (non-ABC) and durvalumab with lenalidomide + R-CHOP (R2-CHOP) in ABC previously untreated DLBCL; secondary objectives are to evaluate safety and identify biomarkers predictive of clinical response. Methods: This is a phase II, two-arm, open-label, global, multicenter study of durvalumab combinations in patients with previously untreated, high-risk DLBCL (MEDI4736-DLBCL-001; EUDRA CT 2015-005173-20; NCT03003520). High risk was defined as Ann Arbor stage III/IV or II with bulky disease (≥7.0 cm), along with intermediate-high/high IPI ≥3 or NCCN-IPI ≥4; patients must also have CD20+ DLBCL, ECOG PS 0-2, and no prior antilymphoma therapy. All patients receive durvalumab + R-CHOP21 in induction cycle 1 simultaneous to cell-of-origin (COO) analysis by gene expression profiling with NanoString technology. Beginning with cycle 2, Arm A (non-ABC) receives durvalumab 1125 mg IV on day 1 with 6 or 8 cycles of R-CHOP21; Arm B (ABC) receives the same durvalumab + R-CHOP21 doses with oral lenalidomide 15 mg/day on days 1-14. Both arms receive durvalumab consolidation 1500 mg IV on day 1 q28d for ≤12 months from induction cycle 1, day 1. The primary endpoint is 2-y progression-free survival (PFS); secondary endpoints include clinical response to treatment in biomarker-defined subpopulations (tumor and peripheral blood) and safety as assessed per NCI CTCAE v4.03 criteria. Exploratory endpoints include PFS at 12 mo, complete response, and PK/PD. Recruitment is ongoing, with a target enrollment of 120 patients. Clinical trial information: NCT03003520.

TPS7571 Background: Patients ineligible for stem cell transplantation (SCT) or who relapse after SCT, and those who fail to respond to second-line or salvage chemotherapy, represent an unmet medical need for which new therapeutic strategies are required. MOR208 is a novel Fc-enhanced, humanized, monoclonal antibody directed against CD19. Significant single-agent activity of MOR208 in patients with R-R DLBCL (Jurczak et al., J Clin Oncol 34, 2016 [suppl; abstr 7545]) and enhancement of MOR208-mediated cytotoxicity by BEN in preclinical studies, provide a strong rationale to study MOR208 + BEN in patients with R-R DLBCL. Methods: B-MIND is a randomized (1:1), two-arm, multicenter, open-label, adaptive design, phase II/III study of MOR208 + BEN vs RTX + BEN in adult patients with histologically confirmed DLBCL who have relapsed after or are refractory to 1 to 3 prior lines of therapy and who are not candidates for high-dose chemotherapy and autologous SCT. At least 1 prior therapy line must have included a CD20-targeted therapy. Other key inclusion criteria: age ≥18 years; measurable disease; availability of tumor tissue for central pathology review; ECOG 0–2, and adequate major organ systems function. Key exclusion criteria: primary refractory DLBCL; central nervous system involvement, and known double/triple hit DLBCL genetics. The safety of the combination will be assessed in an initial phase II evaluation. Treatment will comprise 6 cycles of MOR208 (12 mg/kg IV) + BEN (90 mg/m2 IV) or RTX 375 mg/m2 IV + BEN. Patients achieving a response after cycle 6 will continue to receive antibody treatment for up to 18 additional cycles. Primary endpoint: progression-free survival (PFS); secondary endpoints include: best overall response, overall survival, safety, quality of life, immunogenicity and pharmacokinetics. Enrollment of 330 patients is anticipated in Europe, US and Asia-Pacific countries. Fourteen patients have been randomized to date. Clinical trial information: NCT02763319.

7547 Background: E1412 statistical design is based on results of lenalidomide/RCHOP (R2CHOP) vs RCHOP in ABC-DLBCL as determined by NanoString gene expression profiling (GEP). Central pathology review (CPR) was conducted to confirm diagnosis and to ensure adequate tissue for GEP. Initially, CPR occurred after randomization and treatment initiation. Due to high ineligibility rate (IR), the protocol was amended to include real-time CPR to determine patient eligibility prior to study enrollment. We describe the revision and how it affected the IR. Methods: Pre-amendment, CPR was done retrospectively. Post-amendment, CPR was done prior to enrollment, in real time, on a submitted tissue block, H&E, and CD20 slides. Based on CPR, if a diagnosis of DLBCL was confirmed, and sufficient tissue remained for GEP, a patient was deemed eligible and the submitter was notified by fax. Protocol goal was notification within 2 working days (WD) of receipt of materials at the CPR site. Results: Pre-amendment, 219 patients were enrolled. Material was typically received at the CPR site 6-9 months after patient registration. The IR with CPR was 36% for all those enrolled, and 26% for patients with adequate tissue for GEP. Post-amendment, 218 patients were submitted for CPR: 145 (67%) were eligible; 73 (33%) were ineligible. Reasons for ineligibility included insufficient tissue (n=27) or a diagnosis other than de novo DLBCL (n=46). Notification of eligibility occurred in a median of 2 WD (Mean 2 WD; Range 1-5 WD). 90% were notified within the protocol goal of 2 WD. GEP for all enrolled was completed within 6 weeks of CPR. Conclusions: The success of this novel, real-time CPR serves as a model for the future of NCTN DLBCL trials. When CPR is performed rapidly prior to enrollment, study slots may more accurately reflect the target population and eliminate excess costs. In the precision medicine era, rapid collection of relevant pathology and biomarker data is essential to trial success. Study Coordinated by ECOG-ACRIN Cancer Research Group (Robert L. Comis, MD and Mitchell D. Schnall, MD, PhD, Group Co-Chairs), supported by NCI grant # CA180820, CA180794, CA180790, CA180799, CA180833.