Habermann, Thomas M.

Abstract Abstract 2692 Background: Massive splenomegaly is occasionally encountered in patients with Non-Hodgkin's Lymphoma (NHL) and can be associated with abdominal symptoms, anemia and thrombocytopenia. The spleen may also be a primary focus of the disease itself. Splenectomy offers the benefit of localized treatment of the disease as well as relief of symptoms from splenomegaly and associated cytopenias. The safety and long term clinical outcomes of splenectomy for massive splenomegaly (>1500 gm) in patients with various histologies of NHL was analyzed. Methods: Data was obtained from clinical records of 90 patients with a diagnosis of NHL who underwent a splenectomy at Mayo Clinic Rochester from January 1998 to December 2007. Demographics, previous treatments, histology, post- surgical complications were collected and analyzed for the entire cohort. Additionally, median overall survival, improvement in symptoms, anemia and thrombocytopenia were analyzed by NHL type. Results: Median age at splenectomy was 67 years (range 42–84 yrs.) and 56 (62%) of patients were male. Average time from diagnosis to splenectomy was 3 months (range 0–217 months). Splenectomy was the first treatment in 56 (62%) patients; the other 34 patients were treated with chemotherapy with an average of two previous regimens. Most patients had involvement outside of the spleen, 65 (72%) with concurrent bone marrow involvement and 35 (50%) with lymph node involvement. Various different NHL histologies were represented: marginal zone (MZ) (34%), mantle cell (MC) (26%), diffuse large B-cell lymphoma (DLBCL) (10%), follicular (FL) (11%), lymphoplasmacytic (LP) (6%) and T cell NOS (4%). With a median follow-up time of 25 months, the median overall survival differed by histology type of NHL (Table 1). Indolent lymphomas such as FL, LP and MZ had more favorable survival compared with more aggressive lymphomas such as DLBCL and T cell NOS. Splenectomy was safe and well tolerated with only one treatment related death. There were no reported post-splenectomy septic events. Two patients did develop a portal vein thrombosis. 95% of patients had relief in pressure and abdominal symptoms post-splenectomy. With the exception of patients with DLBCL and T-cell NHL, a majority of patients had improvement in anemia and thrombocytopenia related to splenomegaly (Table 2). Conclusions: This represents the largest reported series of patients with lymphoma treated with splenectomy. Our results indicate that median overall survival after splenectomy is determined by the underlying type of NHL. Patients with LP, FL, MZ and MCL have a favorable outcomes following splenectomy. Splenectomy remains a safe and effective treatment option for massive splenomegaly in selected patients during the course of their disease management. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 4127 Introduction: Mantle cell lymphoma (MCL) is an aggressive lymphoma. NF-κB has been shown to be constitutively activated in MCL cell lines and patient biopsy samples and may play a key role in the growth and survival of MCL cells. We previously reported in a pilot study of 39 MCL patients from the NCI-SEER Survival Study that host genetic variation in candidate TNF and NF-κB genes was associated with overall survival after accounting for clinical variables (Blood 2007;110(11):472a). Here, we attempt to replicate the top 8 genes (NFKB1, IRF4, TNFSF13B, TNFRSF25, NFKBIA, LTA/TNF, TRAF5, RELB) and associated single nucleotide polymorphisms (SNPs) in an independent sample of MCL patients. Methods: We genotyped 71 SNPs from 8 genes in a prospective cohort of newly diagnosed MCL patients with germline DNA enrolled at the Mayo Clinic and University of Iowa from 2002–2008 as part of the Molecular Epidemiology Resource of the Iowa/Mayo Lymphoma SPORE. All patients were systematically followed through 2009 for overall survival (OS) and event-free survival (EFS, defined as disease progression, retreatment or death due to any cause). Genotyping was performed on an Illumina Golden-Gate platform. Cox regression was used to estimate Hazard Ratios (HRs) and 95% confidence intervals (CI) for individual SNPs with OS and EFS. Each SNP was modeled with the most prevalent homozygous genotype used as the reference group. An ordinal test was used to assess the trend across genotype. A principal components methodology was used for gene-level analyses. All Cox models were adjusted for MIPI and treatment. We compared our results to the previously reported NCI-SEER study that included MCL patients diagnosed from 1998–2000. The NCI-SEER study had a median age at diagnosis of 64 years (range 38–74), 27 deaths (69%), and a median follow-up for living patients of 47 months (23-85 months). Replication was declared based on a p-trend<0.10 in the gene-level test or p-trend<0.10 in the SNP-level ordinal trend test and a HR of similar direction and magnitude to the NCI-SEER study result. Results: The median age at diagnosis of the 101 patients in the SPORE was 64 years (range 42–88), and by simplified MIPI score there were 38% low risk, 37% intermediate risk, and 26% high risk patients. The most common initial therapy was anthracycline-based chemotherapy ± rituximab (59%) with or without stem cell transplantation. Through 2009, there were 61 events (60%) and 31 (31%) deaths, with a median follow-up for living patients of 59 months (range 29–90). At the gene-level, none of the 8 genes replicated at p<0.10. However, at the SNP level, the genes with SNPs that replicated for OS included TRAF5 (rs3738199, rs6684874, rs12569232), TNFRSF25 (rs3138156), and RELB (rs10424046, rs1560725). The strongest TRAF5 SNP was rs3738199 which had a minor allele frequency (MAF) of 0.35. Compared to patients with the AA genotype, those with the AG (HR=2.14, 95% CI 0.91–5.01) or GG (HR=3.72, 95% CI 1.34–10.3) genotypes had inferior survival (p-trend=0.0092). The TNFRSF25 SNP rs3138156 had a MAF of 0.054; compared to patients with the AA genotype, those with the AG or GG genotype (HR=2.46, 95% CI 0.88–6.89) had inferior survival (p=0.087). Finally, the strongest RELB SNP was rs10424046, and had a MAF of 0.54. Compared to patients with the GG genotype, those with the GC (HR=0.57, 95% CI 0.25–1.28) or CC (HR=0.39, 95% CI 0.14–1.12) genotypes had superior survival (p-trend=0.065). Similar results for EFS were observed for the TRAF5 and RELB SNPs. Discussion: Germline genetic variation in the TNF and NF-κB pathway genes TRAF5, RELB, and TNFRSF25 were associated with prognosis in MCL after adjustment for clinical and treatment factors, and this result replicates findings from an independent dataset. TRAF5 is one of the components of a multiple protein complex which binds to TNF receptor cytoplasmic domains and mediates TNF-induced activation; RELB is part of the NF-κB complex; and TNF super family receptor member 25 stimulates NF-κB activity and regulates apoptosis through signal transduction that is mediated by various death domain containing adaptor proteins. In summary, genetic variation in TNF and NF-κB pathway genes may play a role in disease progression and overall survival in MCL, supporting further targeting of this pathway for therapy. Support: Lymphoma Research Foundation, P50 CA97274, R01 CA129539. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 3088 Background: Despite the use of modern immunochemotherapy (R-CHOP) regimens, almost 50% of patients with diffuse large-B-cell lymphoma (DLBCL) will relapse. Current prognostic models, most notably the International Prognostic Index, are comprised of patient and tumor characteristics and are unable to identify patients with less than a 50% chance of long-term survival. However, recent observations demonstrate that factors related to host adaptive immunity and the tumor microenvironment are powerful prognostic variables in non-Hodgkin lymphoma Methods: We retrospectively examined the absolute neutrophil count (ANC), monocyte count (AMC) and lymphocyte count (ALC), obtained from an automated complete blood count with differential, as prognostic variables in a cohort of 255 consecutive DLBCL patients that were uniformly treated with R-CHOP between 2000 and 2007 at a single institution. The primary study objective was to assess if ANC, AMC, and ALC at diagnosis were predictors of overall survival (OS) in DLBCL. Results: At diagnosis, the median ANC was 4720/uL (range 1190–17690), the median AMC was 610/uL (range 30–4040), and the median ALC was 1220/uL (range 140–5410). The median follow-up for these patients was 48 months. In the univariate analysis, each of these variables predicted OS as continuous variables. As dichotomized variables, an elevated ANC (≥5500/μL; hazard ratio 1.75, 95% confidence interval 1.14–2.60, p=0.01) and AMC (≥610/μL; hazard ratio 3.36, 95% confidence interval 2.10–5.59, p<0.0001) were each associated with inferior OS. In contrast, the presence of lymphopenia, defined as an ALC ≤1000/uL, was associated with inferior OS (hazard ratio 2.21, 95% confidence interval 1.43–3.39, p=0.0004). When components of the IPI were included on multivariate analysis only the AMC and ALC were independently significant prognostic factors for OS, with hazard ratios of 3.37 (95% confidence interval 2.05–5.74, p<0.0001) and 2.19 (95% confidence interval 1.38–3.44, p=0.0009), respectively. The dichotomized AMC and ALC generated the AMC/ALC prognostic index (PI) and stratified patients into 3 risk groups: very good (AMC <610/uL and ALC >1000/uL), good (AMC ≥610/uL or ALC ≤1000/uL), and poor-risk (AMC ≥610/uL and ALC ≤1000/uL) populations. For both the very good (n=79) and good-risk (n=134) groups median OS has not been reached with estimated 5-year overall survival of 88% and 69%, respectively. Median OS for poor-risk (n=42) patients was 1.7 years (95% confidence interval 1.1–2.7 years) with an estimated 5-year overall survival of 28% (p<0.0001). By comparison, the R-IPI was unable to identify a group of patients with a median survival less than 8 years. The estimated 5-year OS was 93%, 71% and 53% for very good, good and poor-risk patients, respectively. We sought to determine whether the AMC/ALC PI may provide additional prognostic information when combined with the R-IPI. To test this possibility, the 171 very good/good risk and 84 poor risk patients identified by the R-IPI were subsequently risk stratified using the AMC/ALC PI. Among R-IPI very good/good risk patients a subset of poor risk patients (n=21) with a median OS of 2.2 years (95% confidence interval 1.1–6.6 years) and 35% 5-year OS could be identified with the AMC/ALC PI. In contrast, 5-year OS ranged from 75%-88% among very good and good risk patients. Similarly, stratification of R-IPI poor risk patients by the AMC/ALC PI identified subsets of very good (n=19) and good risk (n=44) patients with median OS that had not been reached and 86% and 55% 5-year OS, respectively. High risk (n=21) patients had a median OS of 1.4 years (95% confidence interval 0.9–2.2 years) and an estimated 5-year OS of less than 25%. Conclusions: Measurement of AMC and ALC at diagnosis is widely applicable, cost effective, predicts OS, and identifies high-risk patients with DLBCL. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 589 Background: The initial results of this intergroup trial with median follow-up of 3.5 years (yrs) were previously reported (J Clin Oncol 24:3121, 2006). We present updated results with median follow-up of 9.4 yrs from induction therapy randomization, and 9.0 yrs from maintenance randomization. Methods: 632 patients (pts), age >60 yrs, with DLBCL were randomized to CHOP+rituximab 375 mg/m2 IV, administered Day -7, -3, and two days before cycles 3/5/7 (if given) (R-CHOP), versus CHOP, for two cycles beyond best response for 6–8 cycles total. Pts were stratified by the International Prognostic Index (IPI) (<3 vs >3). 415 pts responding to R-CHOP or CHOP were then randomized to maintenance rituximab 375 mg/m2 weekly times 4, every 6 months for 2 yrs starting 4 weeks after the last chemotherapy (MR, n=207), or observation (OBS, n=208). Results are presented for the 546 (267 R-CHOP; 279 CHOP) pts for induction, and 352 (174 MR; 178 observation) evaluable, centrally reviewed, maintenance pts. Failure-free survival (FFS) was the primary endpoint. The stratified weighted Cox regression was used to remove the effect of MR in comparing induction treatment, and stratified log-rank test used to assess maintenance effect. Results: Baseline characteristics and response to induction were balanced. 9-yr FFS and OS) are 35% and 44% for R-CHOP, and 25% and 37% for CHOP. Compared with CHOP, R-CHOP significantly prolonged FFS (p=.008), but not OS (p=.11). Pts were categorized into low-risk (LR) and high-risk (HR) groups according to their IPI (<3 or 33) (LR, n=217; HR, n=327). A significant difference in the effect of induction therapy was observed for high-risk patients only for FFS (p=0.02, HR=0.61, 95% CI, 0.51 to 0.93) but not for OS. MR significantly prolonged FFS (log-rank p=0.018, HR=0.71, 95%CI, 0.54 to 0.94), but not OS (p=0.44, HR=0.89, 95%CI, 0.65 to 1.20). A significant interaction between maintenance and induction therapies was observed in that MR significantly prolonged FFS after CHOP (p=0.003, HR=0.56, 95%CI, 0.38 to 0.82), but not after R-CHOP (p=0.89, HR=0.97, 95% CI, 0.64 to 1.47). There was no OS difference with MR after CHOP (p=0.19, HR=0.76, 95% CI. 0.50 to 1.15) or R-CHOP (p=0.77, HR=1.07, 95% CI, 0.68 to 1.68). Median time to failure after maintenance randomization following CHOP+MR was 9.5 yrs vs 2.0 yrs for CHOP+OBS (p=0.003) and following R-CHOP+MR was 8.5 yrs vs 7.5 yrs for R-CHOP+OBS (p=NS). Proportionately more treatment failures occurred within 2 yrs after CHOP+OBS (73%) compared to CHOP+MR (47%), p=0.01. In contrast, the proportion of failures within 2 yrs was similar for R-CHOP+OBS (38%) and R-CHOP+MR (36%), p=NS. Conclusions: Initial R-CHOP therapy, as compared with CHOP, resulted in improved DFS and FFS for older DLBCL pts. MR after CHOP, but not after R-CHOP, significantly prolonged time to failure but did not prolong OS. However, FFS was 42% at 9 yrs among R-CHOP responders, with or without MR. Future treatment strategies should build upon these findings in this older patient population, often with significant co-morbidities. Disclosures: Morrison: merck: Consultancy, Membership on an entity's Board of Directors or advisory committees; celgene: Consultancy, Speakers Bureau; amgen: Consultancy, Speakers Bureau; genentech: Speakers Bureau; pfizer: Speakers Bureau. Fisher:roche: Consultancy. Horning:Genentech: Employment.

Abstract Abstract 2634 The peripheral blood absolute lymphocyte/monocyte ratio (ALC/AMC) at diagnosis has been recently reported [Porrata et al, Hematologica 2012; 97(2):262-9] and confirmed [Koh et al, Oncologist 2012; 17(6):871-80] to be a predictor of clinical outcomes in classical Hodgkin lymphoma (HL). In addition, an inversed correlation has been reported between the ALC/AMC ratio and tumor associated macrophages in classical HL suggesting an association between the host biological response in the macro-environment and micro-environment. However, a limitation of the ALC/AMC ratio at diagnosis, the International Prognostic Score at diagnosis, the Positron-Emission Tomography (PET-scan) at diagnosis, or interim PET-scan during treatment is their inability to sequentially assess the host/tumor interaction during treatment. Therefore, we studied the ALC/AMC ratio recovery, as a surrogate marker of host immunity (i.e., ALC) and tumor micro-environment (i.e., AMC), at each treatment cycle phase to assess its predictive ability for clinical outcomes in classical HL. To participate in the study, patients were required to be diagnosed, treated, and followed at Mayo Clinic, Rochester, Minnesota and treated with adriamycin, bleomycin, vinblastine, and dacarbazine (ABVD) with or without radiation therapy. An ALC/AMC ratio ≥ 1.1 cut-off was used in the study based on our previous publication [Porrata et al, Hematologica 2012; 97(2):262-9]. From 1990 until 2008, 190 classical HL patients qualified for the study. The cohort was evenly split with 50% males and 50% females. Fifty-eight percent of the patients had an interim PET-scan performed. The median age at diagnosis was 36 years (range: 18–83 years). Seventeen patients received 2 cycles; 5 patients 3 cycles; 51 patients 4 cycles; 5 patients 5 cycles and 112 patients 6 cycles. The median follow-up was 3.7 years (range: 0.2–20 years). The ALC/AMC ratio ≥ 1.1 at each treatment cycle phase of ABVD was a predictor for overall survival (OS), lymphoma-specific survival (LSS), progression-free survival (PFS) and time to progression (TTP). Patients were stratified based on how many cycles the ALC/AMC ratio did not reach the cut-off of 1.1. The more cycles with an ALC/AMC < 1.1 the worse survival was observed. The 23 patients who did not achieve an ALC/AMC ≥ 1.1 during the treatment cycles phase experienced the worst survival rates: 5 year OS of 33%; 5 year LSS of 43%; 5 year PFS of 16%; and 5 year TTP of 27%. Patients were then dichotomized into patients that did not achieve an ALC/AMC ratio ≥ 1.1 for all the cycles versus patients with an ALC/AMC ratio ≥ 1.1 in any cycle during treatment. Patients with an ALC/AMC ratio ≥ 1.1 in any cycle (N=167) experienced better survival compared with patients that in all cycles the ALC/AMC ratio was < 1.1 [ OS, the hazard ratio(HR) was 0.09, 95% CI (0.04–0.19), p < 0.0001; for LSS, HR =0.07, 95% CI (0.02–0.17), p <0.0001; for PFS, HR = 0.08, 95%CI (0.04–0.14), p < 0.0001; and for TTP, HR = 0.05, 95% CI (0.02–0.11), p < 0.0001]. In the multivariate analysis, an ALC/AMC ratio ≥ 1.1 in any cycle versus none was an independent predictor for clinical outcome when compared to ALC/AMC ratio at diagnsosis, PET-scan, International Prognostic Score, limited verus advanced stage, and radiation therapy [OS, HR = 0.14, 95% CI (0.04–0.40), p< 0.0002; for LSS, HR = 0.09, 95% CI (0.01–0.37), p <0.0006); for PFS, HR = 0.19, 95% CI (0.05–0.82), p < 0.03); and for TTP, HR = 0.15, 95% CI (0.02–0.89), p < 0.02]. In conclusion, the ALC/AMC ratio during ABVD treatment cycles in classical HL is a predictor for survival and provides an opportunity to develop therapeutic modalities to manipulate the ALC/AMC ratio during ABVD chemotherapy to improve clinical outcomes in classical HL. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 287 Purpose: Farnesyltransferases are key cellular enzymes involved in the prenylation of proteins. Prenylated proteins are important for malignant cell growth. The purpose of this phase II study (NCT00082888) was to assess tumor response and toxicity associated with the oral farnesyltransferase inhibitor tipifarnib (R115777, Johnson and Johnson) in patients with relapsed aggressive, indolent, or uncommon lymphoma. The uncommon group included low incidence lymphoma types such as Hodgkin lymphoma (HL) and T-cell non-Hodgkin lymphoma- cutaneous T-cell lymphoma (CTCL) and peripheral T-cell lymphoma (PTCL). Patients and Methods: Patients with relapsed or refractory non-Hodgkin (NHL) or Hodgkin (HL) were eligible if they had measurable disease, performance status ≤2, platelet count ≥75,000 × 10(6)/L, an absolute neutrophil count ≥1000 × 10(6)/L and a creatinine ≤2 × upper limit of normal. Patients initially received tipifarnib 300 mg twice daily, days 1–21 every 28 days. The trial was conducted by the University of Iowa/Mayo Clinic Lymphoma SPORE. Results: Ninety-three patients with lymphoma (42 aggressive, 15 indolent, and 36 uncommon) enrolled in the trial between March 2004 and November 2008. The median age was 62 years (range, 18–91). The median number of prior therapies was 5 (range, 1–17). The overall response rate (ORR) was 20% (19/93) with 6% (6/93) complete responses (CR) and 14% (13/93) partial responses (PR). The ORR was 17%, 7%, and 31% for the aggressive, indolent, and uncommon groups, respectively (Table). Among the 19 responders were 7 diffuse large B-cell lymphoma (DLBCL), 7 T-cell NHL, 1 follicular grade II, and 4 HL. The overall median response duration for the 19 responders was 7.5 months (mean, 12.8; range, 1.8 – 48.1 months). The median response duration was 11.3 months (mean, 10.1; range, 1.8 – 18.5 months), 2 months, and 7.5 months (mean, 15.5; range, 1.9 – 48.1 months) for the aggressive (7), indolent (1), and uncommon (11) groups, respectively. Five patients in the uncommon group are still receiving treatment with a range of treatment duration between 25 – 60+ months. Among the 5 still receiving treatment 4 responded and 1 is stable. The range of response duration is 15.4, 17.7, 43.1, and 48.1 months. These five patients (2 with HL, 1 anaplastic large cell cutaneous, and 2 PTCL) are currently receiving doses of 400 (1), 200 (2), and 100 (2) mg twice daily. The median time to progression for all patients was 3.6 months (95% CI: 2.1 – 4.5 months). The grade 3,4 toxicities observed were primarily fatigue (10%) and reversible myelosuppression with 11% anemia, 37% neutropenia, and 32% thrombocytopenia. Correlative studies demonstrated tipifarnib-induced upregulation of Bim, a marker of tipifarnib action in lymphoma cell lines, in 7 of 9 paired lymphoma biopsies. Conclusions: Tipifarnib has activity in lymphoma, with responses in patients with heavily pretreated DLBCL, HL and T-cell types, but little activity in follicular NHL. This agent has an excellent toxicity profile and targets a novel pathway. Further studies in combination with other agents are warranted for patients with DLBCL, HL or T-cell NHL. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 2859 Background: PTLD is the most common malignancy, other than melanomatous skin cancer complicating solid organ transplantation and has been one of the most commonly observed fatal consequences. The histologies are diverse and include B-cell, T-cell and Hodgkin lymphomas. Rituximab has affected the outcome in CD20+ patients with PTLD. This single institution study sought to identity the outcomes in patients of all histologies and presentations managed in the post-rituximab era. Methods: Patients with solid organ transplantation who were diagnosed with PTLD at Mayo Clinic (Rochester, MN) between December 1970 and October 2009 were identified through the Mayo Clinic Lymphoma Data Base. The rituximab era was defined as patients diagnosed after December 26, 2000, the date of diagnosis for the first patient treated with rituximab in our series. Cox proportional hazards models were used to assess the association of clinical factors and overall survival (OS). Results: 143 patients diagnosed with PTLD between December 1970 and October 2009 were identified. The median age at the time of diagnosis of PTLD was 50 years (range 15–84) with 41 patients (29%) over the age of 60. 93 were male (65%). At a median follow-up of 84 months (range 1–295), 95 patients (65%) have died. The types of organ transplant included renal (39%), liver (31%), cardiac (11%), lung (6%), renal/pancreas (6%), pancreas (4%), and multiorgan (4%). 83% of patients presented with extranodal disease. 28% had involvement of the engrafted organ. 76% of the patients developed PTLD that was EBV+ by in situ hybridization. 62% were stage 3–4. The IPI was intermediate-high in 39% of patients. Initial approaches to management included reduction in immunosuppression (45%), chemotherapy (12%), reduction in immunosuppression with rituximab (12%), rituximab (8%), and radiation therapy (6%). The overall response to therapeutic approaches was 46% in the pre-rituximab era and 56% in the post-rituximab era (chi-squared p=0.25). Patients treated with rituximab had an improved overall survival (HR = 1.66, 95% CI: 0.99–2.79). Patients treated in the rituximab era had improved 1 year survival (76%) compared to patients in the pre-rituximab era (59%) (chi-squared p=0.03). However, the survival advantage in the rituximab era was not maintained in longer follow-up. The median overall survival in the pre-rituximab era was 31 months compared to 56 months in the post-rituximab era (logrank p=0.63). Conclusion: The overall response rates and overall survival are improved with rituximab treatment in CD20+ PTLD, however, the natural history of all PTLD has not improved significantly in the post-rituximab era. New approaches are needed for the prevention and management of all the histologies of PTLD. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 1952 Poster Board I-975 Background: Vitamin D is a naturally occurring steroid hormone. In addition to its well-established role in maintaining serum calcium homeostasis, vitamin D has effects on cellular differentiation, proliferation, apoptosis, and angiogenesis. Vitamin D3 is naturally produced in skin in response ultraviolet-B (UVB) radiation from the sun. Several recent studies, however, have shown that a high proportion of community-dwelling subjects in both tropical and temperate climates are deficient in vitamin D, and that subjects in northern latitudes often require dietary supplementation to maintain vitamin D sufficiency. Further, several reports now suggest that vitamin D sufficiency is protective against the development of several cancers, including non-Hodgkin lymphoma (NHL). However, it is not known whether vitamin D impacts prognosis in diffuse large B-cell lymphoma (DLBCL), the most common NHL subtype. Methods: We evaluated serum vitamin D concentrations in two cohorts of DLBCL patients. The first cohort consisted of 374 patients newly diagnosed from September 2002-February 2008 who were prospectively enrolled in the University of Iowa/Mayo Clinic Lymphoma SPORE Molecular Epidemiology Resource, an observational study in which neither therapeutic nor diagnostic management is prescribed. All serum was obtained within 120 days of diagnosis, and was prior to treatment in 221 (59%). In the second study, 62 patients newly diagnosed from February 2006-August 2007 were enrolled on NCCTG clinical trial N0489, and all serum was obtained before the first course of therapy. Central pathology review was performed on all patients to confirm the diagnosis of DLBCL. Serum 25-hydroxyvitamin (25-OH-VitD) levels were measured by the gold standard method for vitamin D determination: liquid chromatography tandem mass spectroscopy (LC-MS/MS). Vitamin D deficiency was defined as total serum 25-OH-VitD < 25 ng/mL (62.5 nmol/L). SPORE patients were followed per a standard protocol for event-free (progression, retreatment, or death due to any cause) and overall survival (EFS and OS, respectively); N0489 patients were followed in a similar way per the clinical trial protocol. We used Cox regression to estimate hazard ratios (HRs) and 95% confidence intervals (CI). Results: SPORE patients were primarily treated with immunochemotherapy (83%); all N0489 patients were treated with epratuzumab + R-CHOP21. The median age at diagnosis was 62 years (range, 19-93) for SPORE patients and 61 years (range, 21-82) for N0489 patients. Median follow-up for SPORE patients was 36 months (range, 1-77) with 95 deaths and 131 events; median follow-up for N0489 patients was 24 months (range, 7-39) with 13 deaths and 18 events. Vitamin D deficiency was identified in 188 SPORE patients (50%) and 15 N0489 patients (24%). There were no differences in the prevalence of vitamin D deficiency by age, IPI, or whether serum was obtained prior to starting treatment (all p > 0.28). Vitamin D deficiency was associated with inferior overall (HR=2.33, 95% CI 1.53-3.57, logrank p = 0.0001) and event-free survival (HR=1.71, 95% CI 1.20-2.44, logrank p = 0.003) in the SPORE cohort. These associations remained significant after adjusting for IPI and treatment: OS HR=1.97, 95% CI 1.27-3.07; EFS HR=1.47, 95% CI 1.02-2.21. These results were similar for the subset of SPORE patients who were treated with R-CHOP. The HRs for vitamin D deficiency and event-free survival were consistent in the N0489 patients (HR=1.63, 95% CI 0.61-4.35, IPI adjusted HR=1.43, 95% CI 0.53-3.85), although these results lacked precision due to a smaller sample size. We were unable to evaluate OS in N0489 due to the small number of deaths. Conclusions: Approximately 50% of all DLBCL patients in this northern US latitude population are vitamin D deficient at the time of diagnosis and treatment. Vitamin D deficient patients have an inferior event-free and overall survival compared to patients with vitamin D levels within the normal range. Vitamin D supplementation during treatment of DLBCL patients with vitamin D deficiency should be evaluated in a clinical trial setting. Disclosures: No relevant conflicts of interest to declare.

Abstract Introduction Relapse in the central nervous system is an uncommon complication of diffuse large B cell lymphoma (DLBCL) associated with a poor prognosis. The addition of rituximab to chemotherapy has improved outcomes in patients with DLBCL but its effect on the outcome of patients with CNS relapse is not well characterized. Here we present the natural history of patients with CNS relapse in a large cohort of patients with DLBCL who were treated in the immunochemotherapy (IC) era. Methods Newly diagnosed patients with DLBCL or primary mediastinal B-cell lymphoma (PMBCL) and treated with primary anthracycline based IC were prospectively enrolled on the University of Iowa/Mayo Clinic SPORE Molecular Epidemiology Resource (MER) within 9 months of diagnosis and followed for relapse, retreatment, and death. Clinical management at diagnosis and subsequent therapies were per treating physician. Patients with documented CNS involvement at diagnosis, primary CNS lymphoma, or post-transplant lymphoproliferative disorder at diagnosis were excluded. All CNS relapse and retreatment details were verified by medical record review. Results 1017 patients with newly diagnosed DLBCL or PMBCL, no documented CNS disease at diagnosis, and treated with IC were enrolled in the MER between 2002 and 2012. At a median follow-up of 59 months (range 1-148), 36 patients had CNS relapse. The cumulative incidence of CNS relapse was 3.1% (95% CI: 2.2%-4.4%) at 2 years and 3.7% (95% CI: 2.7%-5.0%) at 5 years after diagnosis. CNS involvement was identified after first line IC in 25 patients, and after salvage therapy in 11 patients. 22 patients had an isolated CNS relapse, while 14 had both systemic and CNS disease at relapse. The incidence of isolated CNS relapse at two years from diagnosis was 1.9% (95% CI: 1.2%-3.0%, figure 1). CNS involvement was parenchymal in 22 patients, leptomeningeal in 11, and parenchymal and leptomeningeal in 3 patients. At diagnosis, this subset of 36 patients had a median age of 61 years (range 20-86); 25 (69%) were male. IPI was 0-1 in 6 patients, 2 in 13 patients, 3 in 11 patients and 4-5 in 6 patients. Cell of origin per Hans algorithm was available in 17 patients, 8 were GCB and 9 were non-GCB. Extranodal sites of disease at diagnosis included testicular (2), renal (5), bone (10) and bone marrow (6). 7 patients had no sites of extranodal disease. LDH was elevated in 71%. The German High-Grade Non-Hodgkin Lymphoma Study Group CNS risk score was low risk in 6 patients, intermediate risk in 22 patients, and high risk in 8 patients. 6 patients received CNS prophylaxis with their initial immunochemotherapy, including 3 of the 21 patients who subsequently developed isolated CNS disease. First line therapy after CNS relapse included high dose methotrexate based chemotherapy in 22 patients and other systemic regimens in 6, with intrathecal (IT) chemotherapy in 5. Overall survival in patients after CNS relapse was poor with a median survival of 6.3 months (95% CI: 2.9-15.5) and 12-month survival rate of 29% (95% CI: 17%-49%). Patients who had CNS relapse following initial IC had better subsequent survival (median OS = 7.6 months) compared to patients with first presentation of CNS relapse following salvage therapy (OS = 2.1 months, p=0.005, see figure 2). 10 patients (28%) proceeded to SCT after CNS relapse, of which 6 remain alive at a median follow-up of 39 months. There were no differences in post-CNS relapse survival by age, sex, CNS risk score, cell of origin, extranodal site involvement at diagnosis, or site of CNS relapse (all p>0.10), though power was limited due to small numbers. Conclusions The incidence of CNS relapse in DLBCL is approximately 4% in the immunochemotherapy era. The vast majority of CNS relapses occur within 2 years of diagnosis. Isolated relapse in patients without CNS involvement at diagnosis is a relatively rare event with an incidence of approximately 2% at 2 years of diagnosis. Outcomes remain poor, and standard clinical variables do not predict survival after CNS relapse. Novel therapeutic approaches are needed in this population, with consideration to autologous SCT, which produces durable remission in a subset of patients. Disclosures Maurer: Kite Pharma: Research Funding. Farooq:Kite Pharma: Research Funding. Thompson:Kite Pharma: Research Funding. Ansell:Bristol-Myers Squibb: Research Funding; Celldex: Research Funding. Cerhan:Kite Pharma: Research Funding. Link:Genentech: Consultancy, Research Funding; Kite Pharma: Research Funding.

Abstract Background Immune editing is a major mechanism used by tumors to promote their survival. We have previously reported the presence of immunosuppressive monocytes (CD14+HLA-DR low/neg) in a number of cancers. Increased presence of these cells was associated with decreased treatment response and OS. Here we report an in vitro model to examine tumor monocyte cross talk and demonstrate that these monocytes may directly promote tumor cells’ resistance to chemotherapy. Method Monocytes from healthy donors were co-cultured with lymphoma cell lines (OCI-Ly1, OCI-Ly3, OCI-Ly7, OCI-Ly10, Su-DHL2, Su-DHL6, Jeko-1, Granta-519) with or without doxorubicin (DOX). Cultured cells were assessed for phenotype, viability, proliferation, and induced protein expression. Changes in vitro were confirmed in primary tissue by gene expression profile from 48 primary diffuse large B cell (DLBCL) lymphoma tumors. Results Monocytes converted CD14+HLA-DR+ to the CD14+HLA-DRlow/neg phenotype by co-culture with 5 of the 8 B cell lymphoma cell lines. The HLA-DR suppression was not due to the known regulatory effects of IL-10 nor was the effect seen in co-culture using B cells from healthy donors. Thus, the magnitude of DR loss is mediated by a yet undetermined mechanism. DOX incubation induced apoptosis in all of the cell lines. Co-culture with monocytes improved lymphoma cell survival in 3 of the 4 lymphoma cell lines. For example, untreated OCI-Ly3 had a 1.3±0.2 fold expansion that was reduced to 0.6±0.1 when treated with DOX (p<0.0001, n=13). However, when the cells were treated with DOX and co-cultured with monocytes the survival improved to 0.8±0.2 (p<0.0001, n=13). Having previously shown that CD14+HLA-DRlow/neg were capable of suppressing patient immunity, partially through arginase I, we determined if co-culture of monocytes with tumor cells increased arginase I expression in either monocytes or tumor cells. Not only did we see arginase I expression increases during co-culture(n=10, p<0.01), but surprisingly, the greatest increases were seen during co-culture and DOX treatment (n=10, p=0.01). To determine if our in vitro observations reflect human tumors, we used gene expression data obtained from 48 DLBCL tumors and found that increased CD14 gene expression was positively correlated with increased arginase I expression (n=48, p=0.02). Conclusions We have demonstrated that certain tumor cells induced phenotype changes in monocytes in an IL-10 independent and a tumor specific way. In those cells where tumors factors mediate monocyte phenotype, the monocytes can promote tumor cell resistance to cytotoxic killing by DOX, thus demonstrating a cross-talk between monocyte and tumor function. We found that in an environment of monocyte/tumor cross talk, arginase I expression is further increased in a cytotoxic environment. Together, this data demonstrates an active cross talk between monocytes and tumors resulting in multiple mechanisms of tumor resistance to chemotherapy. Disclosures: No relevant conflicts of interest to declare.