Citations

Abstract Abstract 4637 Primary central nervous system lymphoma (PCNSL) is an aggressive form of extra-nodal, high-grade, non-Hodgkin B-cell neoplasm. It originates in the brain, leptomeninges, spinal cord or eyes and typically remains confined to the CNS. Further immunophenotypic characterization has revealed that these tumors have overlapping features of germinal center and activated B-cell differentiation. Although the cells of origin are lymphocytes, PCNSL should be considered a brain tumor because its therapeutic challenges resemble those of other brain tumors. Regarding the molecular pathogenesis of PCNSL, there is significantly less information in comparison with systemic lymphomas. The aim of this study was to perform a comprehensive high-resolution genomic analysis in a cohort of PCNSL cases to better characterize the disease at chromosomal and gene level. Copy-number abnormalities were analyzed in seven PCNSL by array-based comparative genomic hybridization (aCGH) using SurePrint G3 (1 million probes, ∼4 Kb resolution) microarray (Agilent). Overall, PCNSL cases were characterized by highly complex genomes with a median of 23 copy-number abnormalities per patient (range 16-49). The whole genomic regions affected in abnormalities correspond to areas between 3% and 29% of the human genome. Globally, 43 chromosomal regions were affected in a recurrent fashion (24 areas involved in recurrent deletions and 19 in copy-number gains). Gain of 12q12-q24.33 and losses of 9p21.3 and 6q were the most common abnormalities found in 71% patients each (5 out of 7). There was not a unique deleted region on 6q to all patients, with 6q14.1-q14.3, 6q16.3-q22.2 and 6q25.3-q26 losses being found in 4 out of 5 cases with 6q deletions. None of these regions include PTPRK, which has been previously proposed as a potential modulator of PCNSL progression. CDKN2A was the sole gene included in all cases with 9p21.3, being biallelically deleted in two out of 5 cases. Minimal deleted genes were also delineated in 3p21.1 (minimal affected area of 0.03Mb) and 3q26.32 (0.27Mb) in 44% of cases each. Those regions included PRKCD (inhibits B-cell growth through transcriptional regulation of IL-6) and TBL1XR1 (corepressor of nuclear hormone receptor), respectively. Overall, a total of 22 genes were included in homozygous deletions in at least one case. Besides CDKN2A (cell cycle), other genes of interest associated with the development and activation of T cells (TOX, CD58) were biallelically affected. Focal monoallelic deletions affecting negative regulators of the NFkB signaling pathway (MAP4K1, TANK, TAX1BP1, TRIB3), cell cycle (RB1) and immune-cell regulation (SIRPB1, CBLB, NFATC2) were also identified. Of interest, no copy-number abnormalities were identified affecting TP53 in any patient. Several chromosomal breakpoints were identified inside genes that encodes transcription factors previously identified as being part of fusion protein in other hematological malignancies (FOXP1, ETV6 and NCOA2) thus suggesting the potential existence of fusion proteins including these genes. Cytogenetic and genetic approaches are being currently used to validate this hypothesis. This is the first study characterizing a cohort of PCNSL cases by using this comprehensive high-resolution approach. Regarding the small number of cases analyzed so far, we can confirm the existence of a highly complex genome, mainly characterized by abnormalities affecting a plethora of genes involved in lymphocyte development, activation and NF-kB signaling pathways regulation. Molecular analyses are ongoing to validate these findings in a larger cohort of PCNSL patients. Disclosures: Fonseca: Otsuka: Consultancy; BMS: Consultancy; Amgen: Consultancy; Medtronic : Consultancy; Genzyme: Consultancy.

Abstract Abstract 3655 Background Primary central nervous system lymphoma (PCNSL) is an aggressive lymphoma with devastating prognosis. High-dose methotrexate (HD-MTX) in combination with HD-cytarabine builds the backbone of current treatments. Elderly patients constitute 45% of cases, exhibit poor outcome and frequent iatrogenic toxicity. However, research efforts to optimize therapy in this subgroup have been neglected. On this background, we conducted a systematic review (SR) and individual patient data meta-analysis (IPDMA) to provide comprehensive evidence-based management strategy for elderly patients with PCNSL. Patients and Methods SR: We searched MEDLINE and EMBASE without language restriction. Eligibility criteria were prospective/retrospective observational studies or randomized trials (RCT) (all N>=10) that exclusively focused on first-line therapy/outcomes in immunocompetent PCNSL patients ≥60 years. Eligible studies were evaluated for methodological quality and reporting of the following baseline characteristics: Age, performance status (PS), involvement of deep brain structures (IDB), serum LDH at baseline, cerebro-spinal fluid (CSF) protein concentration elevation, neurotoxicity (as reported), specific co-morbidity indices, and functional status. For the IPDMA, investigators of eligible studies were asked to provide individual patient data. Minimal eligibility criteria: Age at baseline, details about first-line therapy, and follow-up information. If no data were available, studies were included in the SR, but not in the IPDMA. To maximize statistical power and generalizability, published/unpublished data from other international collaborators were included. Impact of different first-line treatments on overall survival (OS) was investigated using time dependent mixed effects multivariable Cox regression models (age and PS as fixed effects, study/database as random effect). Results SR: We identified 13 eligible studies including 583 patients in total, median age 68–76, published from 1996–2011. Design of studies: prospective (3 multicenter [1 RCT]; 2 single center), retrospective (4 multicenter and single-center, respectively). Accrual of the RCT was recently finished, but publication is pending. From published studies, information about age and therapy was given throughout, for clinical performance in 77%, for LDH and CSF protein in 15%, and IDB in 38%. Functional status was reported in only one study. From the identified 13 studies, 261 individual patient data were available for our IPDMA and pooled with 408 patients from other databases; altogether 669 patients diagnosed from 1977–2011. Preliminary results IPDMA: 50% were male, median age 68 (60–70 [N=431], 70–80 [N=211], >80 [N=22]); median KPS was 60% (10–100%). Therapy regimens widely varied. Overall response to first-line treatment was 65% (45% CR, 19% PR). After a median follow-up of 23 months (1–171), 44% were still alive, with a 3-year OS of 32% [95%CI, 29–37%]. Grouping by time of diagnosis revealed improvement for patients diagnosed after 1997 (N=462) (P<0.001). In multivariate analysis, MTX-based poly-chemotherapy (CT) (N=474) was associated with improved OS (Hazard ratio [HR] 0.69, 95% CI 0.52–0.91) compared to non-MTX regimens (N=195); this was consistent among patients who received consolidating WBRT (HR 0.33, 95% CI 0.01–0.66) and those who did not (HR 0.46, 95% CI 0.23–0.89). In patients who received any MTX-based poly-CT, addition of CHOP-like components (N=90) was not associated with improved OS (HR 0.98, 95% 0.65–1.48). Although any WBRT showed an overall trend for superior OS, it was associated with a 4-fold risk increase for neurotoxicity (Odds Ratio 4.21, 95% CI 2.23–8.21). Further results of treatment patterns and explorative comparisons will be presented at the meeting. Conclusion This international meta-analysis revealed widely varying treatment approaches and demonstrates that prognosis for elderly PCNSL patients is still poor. However, improvement over the last decades was observed. MTX-based poly-CT was associated with better outcome compared to non-MTX containing approaches. The addition of CHOP-like regimens to HD-MTX did not improve outcome. WBRT was associated with better outcome, but also clearly increased risk of neurotoxicity. Prospective trials designed ad hoc for elderly PCNSL patients are promptly needed. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 140 Background: Peripheral T-cell lymphomas (PTCLs) have poor outcomes, and there is a lack of prognostic biomarkers and therapeutic targets to guide treatment. We recently proposed that the transcription factor, interferon regulatory factor-4 (IRF4, also called multiple myeloma oncogene-1 [MUM1]), might be oncogenic in PTCLs based on its expression in association with translocations between IRF4 and the T-cell receptor gene, TRA@. IRF4 is a therapeutic target in multiple myeloma, where high expression is a poor prognostic factor. In addition, at least two germline IRF4 single nucleotide polymorphisms (SNPs), rs12203592 and rs872071, are associated with IRF4 expression and disease risk and progression in various lymphoid neoplasms. However, the prognostic effects of IRF4/MUM1 expression and IRF4 SNPs in PTCLs are unknown. Methods: Forty seven newly diagnosed PTCL patients with available tissue were identified from the University of Iowa/Mayo Clinic Lymphoma SPORE Molecular Epidemiology Resource. There were 5 anaplastic large cell lymphomas, 14 cutaneous T-cell lymphomas (CTCLs), 6 cytotoxic T-cell lymphomas (cytTCLs: 4 extranodal NK/T-cell and 2 enteropathy type), and 22 PTCLs, not otherwise specified (NOS). Patients were diagnosed between September 2002 and February 2008 and systematically followed through March 2010 for overall survival (OS). Tumor cell IRF4/MUM1 expression was examined by immunohistochemistry on paraffin tissue sections (MUM1p clone; Dako). Positivity was defined as >30% of tumor cells with nuclear staining, as in previous studies. rs12203592 and rs872071 were genotyped in peripheral blood DNA. Cox regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for both IRF4/MUM1 expression and SNPs with OS. Chi-squared tests were used to assess the relationship between IRF4/MUM1 expression and IRF4 SNPs. Results: The median age at diagnosis was 60 years (range, 24–88). At a median follow-up of 62 months (range, 29–89), 22 (47%) of the patients had died. Fifteen (32%) of the 47 T-cell patients had IRF4/MUM1 positive tumors. IRF4/MUM1 positivity was associated with poorer overall survival (HR=4.3; 95% CI: 1.8–10.2; p=0.0008). This association was seen across PTCL subtypes, including PTCL, NOS (HR= 6.5; 95% CI: 1.5–27.7; p=0.01), CTCL (HR=13.4; 95% CI: 1.2–150.1; p=0.03), and cytTCL (HR=5.8; CI: 0.5–65.9; p=0.15). The minor allele (T) in SNP rs12203592 was positively associated with IRF4/MUM1 positivity in a dominant model, with IRF4/MUM1 expression in 60% of patients with CT or TT genotypes compared to 19% of patients with CC genotype (p=0.01). Patients with the CT/TT genotype at rs12203592 also had inferior overall survival (HR=3.7; 95% CI: 1.4–9.5; p=0.007). The rs872071 SNP showed no significant association with either IRF4/MUM1 expression (p=0.38) or overall survival (p=0.71). Conclusions: This study is the first to demonstrate that IRF4/MUM1 expression is poor prognostic factor in PTCLs. This association was observed across PTCL subtypes, including the most common subtype, PTCL, NOS. IRF4/MUM1 expression and poor survival in our patients also were associated with the minor allele (T) in the IRF4 SNP rs12203592; these findings are consistent with previous in vitro data showing the major allele (C) represses IRF4 promoter activity. Interestingly, rs872071 was not associated with IRF4/MUM1 expression or prognosis, similar to findings in multiple myeloma. In contrast, rs872071 is a risk and prognostic factor in chronic lymphocytic leukemia and classical Hodgkin lymphoma, diseases in which IRF4/MUM1 expression has been associated with favorable prognosis. IRF4/MUM1 expression is a poor prognostic factor in PTCLs. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 2319 Poster Board II-296 Background: Data from two phase 3 clinical trials indicate that plerixafor plus G-CSF is safe and more effective than G-CSF alone in mobilizing CD34+ hematopoietic stem cells (HSC) for autologous HSC transplantation (auto-HSCT) in patients with non-Hodgkin's lymphoma (NHL) or multiple myeloma (MM). Herein we report 1-year overall survival in patients with NHL or MM who have undergone auto-HSCT with cells mobilized with plerixafor plus G-CSF compared to placebo plus G-CSF. Furthermore, we also report 1-year overall survival in patients with NHL who failed mobilization on either arm and entered a rescue protocol. Methods: Data were obtained from two prospective, randomized, double-blind, placebo-controlled, phase 3 clinical trials that compared the safety and efficacy of plerixafor (0.24 mg/kg/day SQ) plus G-CSF (10 mg/kg/day) with placebo plus G-CSF for mobilization of HSC for autologous HSCT in patients with NHL or MM. In the NHL study, patients in either study arm who failed mobilization (&0.8 ×106 CD34+ cells/kg in 2 days or &2 × 106 CD34+ cells/kg in 4 days) were eligible to enter the rescue protocol and received plerixafor + G-CSF mobilization. Overall survival in patients was estimated by the Kaplan-Meier method. Results: In the NHL study, 135/150 (90.0%) patients in the plerixafor group compared with 82/148 (55.4%) patients in the placebo group, underwent transplantation (p&0.001) with a median (range) of 5.41 (1.95-17.6) and 3.85 (1.98-8.74) × 106 CD34+ cells/kg, respectively (p&0.001). A total of 62 patients (10 from the plerixafor arm and 52 from the placebo arm) failed mobilization and proceeded to rescue mobilization with plerixafor + G-CSF. Of these, 52 patients (84%) proceeded to transplantation with a median (range) of 3.8 (1.1-10.5) x 106 CD34+ cells/kg. Kaplan-Meier estimates of 1-year overall survival for patients in the plerixafor and placebo groups in the primary ITT population (rescue patients were censored as of consent date for rescue) were 76.0% and 64.7%, respectively. Kaplan-Meier estimate of 1-year overall survival in the rescue group was 83.7%. The Log rank p-value comparing all 3 survival estimates (plerixafor, placebo and rescue) was 0.765 (Figure 1A). In the MM study, 142/148 (95.9%) patients in the plerixafor group and 136/154 (88.3%) patients in the placebo group underwent transplantation (p=0.014) with a median (range) of 5.40 (1.20-16.74) and 3.96 (1.76-16.88) × 106 CD34+ cells/kg, respectively (p&0.001). In this trial, Kaplan-Meier estimates of 1-year overall survival in the plerixafor and placebo groups were 93.6% and 95.8%, respectively (Log-rank p-value=0.771; Figure 1B). Conclusions: These data suggest that overall survival at 1 year is similar between patients with MM or NHL mobilized with plerixafor plus G-CSF compared to placebo plus G-CSF. Furthermore, in the NHL study, patient survival in the rescue protocol was similar to that seen in the phase III trial. Collection of follow-up data on overall survival is ongoing. Disclosures: Micallef: Genzyme Corporation: Membership on an entity's Board of Directors or advisory committees, Research Funding. Stiff:Genzyme Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Stadtmauer:Genzyme Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Bolwell:Genzyme Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Marulkar:Genzyme Corporation: Employment, Equity Ownership. Hsu:Genzyme Corporation: Employment, Equity Ownership. DiPersio:Genzyme: Honoraria.

INTRODUCTION: We present a case of idiopathic varices in the small bowel and colon discovered after extensive work up for iron-deficiency anemia and portal hypertension. CASE DESCRIPTION/METHODS: 41-year-old with no known medical history underwent esophagogastroduodenoscopy (EGD) and colonoscopy for evaluation of iron-deficiency anemia. EGD was normal. Colonoscopy revealed diffuse colonic varices extending from the cecum to the rectum (Figure 1). Computed tomography enterography showed numerous varices of the mid jejunum. Magnetic resonance imaging of the abdomen showed patent portal circulation and no other abnormalities. Echocardiogram did not show any evidence of right heart failure. Transjugular liver biopsy showed a portosytemic pressure gradient of 4 mm of Hg indicating the absence of portal hypertension. Liver biopsy only showed mild steatosis and no fibrosis. A direct measurement of portal pressure was 3 mmof Hg again indicating the absence of portal hypertension. DISCUSSION: Given the extensive workup with no evidence of portal hypertension, patient is thought have idiopathic varices in the small bowel and colon. Ectopic varices, especially in the colon are rare and are usually associated with portal hypertension. Idiopathic colonic varices in association with small bowel varices are even more infrequently described in existing literature. Clinical presentation has been reported with massive hemorrhage, recurrent bleeds and even as incidental findings in endoscopies and imaging performed for other reasons. Management options include conservative frequent follow up or surgery if there is evidence of recurrent bleeds or massive hemorrhage.

INTRODUCTION: Crohn's Disease (CD) is a chronic inflammatory condition, primarily of the gastrointestinal tract with an increasing incidence and prevalence worldwide. In clinical practice, gastroenterologists are often faced with challenges in evaluating disease activity, severity, and prognosis to guide appropriate management of CD. We studied the association between serum albumin level and the rate of active clinical disease in CD patients seen at a tertiary care Inflammatory Bowel Disease (IBD) referral center. METHODS: We designed a retrospective cohort study using adult patients (>18 years) with CD followed for at least one year at University of Alabama at Birmingham (UAB) IBD center by a single trained IBD specialist between 2014-2018. Serum albumin levels were divided into: low mean serum albumin (≤3.2 mg/dL) vs. appropriate mean serum albumin (>3.2 mg/dL). Clinical disease activity was measured using the Harvey-Bradshaw index (HBI). Active disease was defined as HBI score of 8 or higher. Poisson Regression Models (GPR) for Rate Data were used to estimate partially adjusted and fully adjusted incidence rate ratios (IRR) of active clinical disease among CD patients. We also examined IRRs for serum albumin level as a continuous variable. RESULTS: In this single IBD practice at UAB, 269 patients with CD had albumin drawn at initial visit and serially at each subsequent visit. Poisson regression model adjusted for age, gender, race, duration of disease, steroid use, smoking and Crohns location demonstrated that CD patients with a low mean serum albumin during clinical observation were approximately 1.5 times more likely to have active clinical disease during observation (IRR 1.49, 95% CI 1.04-2.15, P = 0.032) compared to those with a serum albumin level of >3.2 mg/dL. The adjusted model with serum albumin as a continuous variable demonstrated that with every unit (0.1 mg/dL) reduction in mean serum albumin, there was 30% higher likelihood of active Crohns in clinic (IRR 1.30, 95% CI 1.02-1.66, P = 0.037). CONCLUSION: Lower mean serum albumin during clinical observation is associated with a higher likelihood of active clinical disease among CD patients. Whether interventions designed to correct or increase serum albumin stores/levels in these patients would be beneficial remains to be examined.

INTRODUCTION: Functional Dyspepsia (FD) is a common functional GI disorder in the upper abdomen (above the navel). The estimated prevalence is 15-30% worldwide. There are currently no FDA-approved medications for the treatment of FD. Proton pump inhibitors (PPIs) are commonly prescribed off-label. However, long-term use of these drugs has been associated with serious health concerns. Prokinetic agents are also often used off-label. However, many are associated with significant side effects (e.g. metoclopramide). A safe, effective, and readily available treatment option for FD is greatly needed. This study describes an important 24-month post-marketing point on an option for FD that has already demonstrated good efficacy and low side effects in RCTs. These RCTs have demonstrated that the combination of caraway oil and peppermint oil improves FD symptoms. Caraway oil plus L-menthol (the major component of peppermint oil) with site-specific targeting (COLM-SST) is a novel delivery system of triple-coated, solid-state microspheres designed to target the duodenum, the main center of disturbance in FD. This study was conducted to report on the overall safety profile of COLM-SST and to determine if any unexpected pattern of side effects emerged. METHODS: A call-in number for reporting adverse events (AEs) was provided on all boxes containing COLM-SST. An independent call center with pharmacovigilance-trained health care personnel, in accordance with the FDA and global regulatory guidelines on properly reporting AEs, was retained to receive and record customer AEs. The AEs for this study were collected and processed from July 8, 2016 to July 8, 2018. RESULTS: An estimated 558,345 individual patients used COLM-SST during the surveillance. An analysis of the data showed that there were no serious AEs reported. Additionally, the self-reporting rate of non-serious AEs was low, with only 148 events received (from 127 patients), a 0.023% event rate for COLM-SST. The top reported, non-serious AEs were abdominal pain/discomfort/distension (18; 0.0032%), which were consistent with the pattern of commonly reported symptoms of FD, and headache (7; 0.0013%). CONCLUSION: 24 months of monitoring in-market, real-world use of caraway oil and L-menthol with site-specific targeting, for the treatment of FD demonstrates an extremely low rate of self-reported AEs. These data, combined with earlier RCT data, indicates that COLM-SST is a safe and effective therapy for FD, and is a viable alternative to PPIs and prokinetics.

INTRODUCTION: Colorectal cancer (CRC) screening is broadly endorsed for individuals 50–75 years of age. Given recent increases in CRC incidence and mortality for patients <50 years old, the American Cancer Society now recommends initiation of average-risk CRC screening at age 45 (Wolf AMD CA Cancer J Clin 2018). The performance of recommended tests, including the multi-target stool DNA (mt-sDNA; marketed as Cologuard™) test, has been minimally described in populations under 50 years old. To address this knowledge gap, we conducted a retrospective study of DNA markers (KRAS, BMP3, NDRG4) included in the mt-sDNA assay to quantify and compare tissue marker levels in sporadic CRC cases and normal colon controls from patients in 45–49 and 50–64 year-old age groups. Our primary hypothesis was that mt-sDNA marker levels would be statistically similar between cases in both age groups. METHODS: Patient samples were identified from a pathology tissue archive at Mayo Clinic, excluding those with inflammatory bowel disease or familial cancer syndromes. Formalin-fixed, paraffin-embedded tissues were reviewed by a GI pathologist for 45–49 year-old (n = 90) and 50–64 year-old (n = 96) CRC cases; normal colon samples (45–49, n = 76; 50–64, n = 92) were also included for comparison. DNA extracted from core punches was blinded and processed in a randomized order. Quantitative Allele specific Realtime Target and Signal amplification (QuARTS) assays were used to quantify KRAS mutations, and BMP3 and NDRG4 methylation. Marker level in each sample was reported as the QuARTS product of mutant KRAS , methylated BMP3 or NDRG4 relative to the ACTB product in each sample, reported as a percentage. At least 60 patients per group provided sufficient power to demonstrate differences in marker levels of ±25% between case groups and between control groups using the Wilcoxon Rank Sum test. RESULTS: CRC cases in the 45–49 vs 50–64 year-old groups did not differ significantly with respect to sex ( P = 0.77) or race/ethnicity ( P = 0.06); left-sided CRC was more common in the 45–49 vs 50–64 year-old group (87% vs. 72%; P = 0.02). None of the marker levels were statistically different between cases of the younger vs older groups or between controls of these groups (Figure 1). CONCLUSION: These data support the expectation of no age-related differences in CRC detection by mt-sDNA, as the relevant molecular markers, when they occur, are present at similar levels across age groups 45–49 and 50–64 years.

INTRODUCTION: Most Americans consume only half of the daily fiber recommended by the National Academy of Medicine, which is 25 grams per day of fiber for women and 38 grams per day for men. Fiber has been shown to promote bowel regularity, lower cholesterol, control blood sugar, and maintain body weight. To fill the “fiber gap” many patients use fiber supplements. There is a perception that fiber supplementation can lead to bloating and excess gas production, among other adverse events. This study aimed to examine whether or not a pattern of adverse events was associated with chewable inulin fiber (CIF). METHODS: A call-in number for reporting adverse events (AEs) was provided on bottles of a specific brand of CIF. An independent call center with pharmacovigilance-trained health care personnel in accordance with the FDA and global regulatory guidelines on properly reporting AEs was retained to receive and record customer AEs. The AEs for this study were collected and processed from January 1, 2018 to December 31, 2018. RESULTS: An estimated 809,259 individual patients consumed CIF during the surveillance. An analysis of the data showed that there were no serious AEs reported. Additionally, the self-reporting rate of non-serious AEs was low, with only 23 non-serious events reported by 21 individuals; a rate of 2.8 per 100,000 users. The top reported, non-serious AEs were flatulence, diarrhea, and constipation; however, the rates for these were extremely low (4.9, 3.7, and 1.2 per million, respectively). Two events were reported that were classified as musculoskeletal and connective tissue disorders; however, these were considered to not be related, having no basis for biological plausibility. CONCLUSION: 12 months of monitoring real-world use of CIF in a population of over 800,000 patients demonstrates an extremely low rate of self-reported AEs. Contrary to conventional thinking, the rate of flatulence was remarkably low (4.9 per million users), and can most likely be considered statistical background noise. These data, combined with previous RCTs, indicates that CIF is a safe and effective option for fiber supplementation.