ALBERT

Description: B-cell receptor (BCR) signaling has been inferred as an important mechanism for disease progression in chronic lymphocytic leukemia (CLL) and other B-cell malignancies. In response to BCR activation, CLL cells secrete the chemokine CCL3, which fosters interactions between CLL cells and the leukemia microenvironment. CCL3 secretion correlates with expression of the 70-kDa ζ-associated protein (ZAP-70) and responsiveness of the CLL clone to BCR stimulation. Here, we measured CCL3 plasma levels by enzyme-linked immunosorbent assay (ELISA) in 351 CLL patients and examined CCL3 levels for associations with established prognostic markers and time from diagnosis to initial therapy. We found that CCL3 plasma concentrations were strongly associated with established prognostic markers. In a Cox proportional hazards regression model, CCL3 as well as established prognostic markers (immunoglobulin heavy chain variable-region mutation status, CD38 or ZAP-70 cytogenetics, clinical stage) were significantly associated with time to treatment. Multivariable analysis revealed that CCL3 (hazard ratio [HR] = 2.33, P 〈 .0001), advanced clinical stage (HR = 2.75, P = .0025), poor risk cytogenetics (del 17p, HR = 2.38; del11q, HR = 2.36, P = .001), and CD38 expression (HR = 1.43, P = .023) were independent prognostic markers. Collectively, CCL3 is a novel, robust, and independent prognostic marker in CLL that can easily and reliably be measured by ELISA. CCL3 therefore should become useful for risk assessment in patients with CLL.

Description: Abstract 560 Nuclear factor kappa B (NF-κB) encompasses a family of transcription factors involved in oncogenic processes including cellular proliferation and apoptotic inhibition. Constitutive activation of NF-κB has been observed in hematologic malignancies and is thought to confer resistance to chemotherapeutic agents. Here, we examine the role of the NF-κB pathway in chronic lymphocytic leukemia (CLL). Whole-exome sequencing was performed using tumor and matched germline DNA from 167 CLL patients. We identified 51 patients (30%) carrying 53 non-silent somatic variants in genes of the canonical NF-κB pathway, which consists of 272 genes as defined by the Ingenuity Pathway Analysis tool. Of the 99 patients whose germline sequences have been analyzed to date, 27 patients (27%) carry 34 non-silent germline variants in NF-κB pathway genes. A total of 67 patients (40%) have at least one non-silent somatic or germline variant. Variants in the NFKB1 gene, itself, were also observed: a somatic variant, H66R, found in two patients, and two germline variants, Y89F and R849W, each found in one patient. To evaluate the functional consequences of the NFKB1 variants, we performed site-directed mutagenesis to generate full-length NFKB1 cDNAs encoding these variants. We subsequently measured transcriptional activity of wild-type and mutant NFKB1 via luciferase assays in HEK293T cells using reporter cassettes containing the NFKB1 response element. Transcriptional activity of the three NFKB1 variants was found to be at least 2-fold higher than that of wild-type NFKB1 (p

Description: Key Points Trisomy 12 CLL cells exhibit upregulated integrin signaling and enhanced VLA-4-directed adhesion and motility. The increased expression of β2-integrins on trisomy 12 CLL cells is modulated by intercurrent NOTCH1 mutations.

Description: Abstract 3555 Introduction: Cytogenetic status represents a major prognostic factor for individuals with AML. However, approximately 50% of individuals have CN-AML and, within this cohort, genetic features have important prognostic and therapeutic implications. Mutations in the FLT3 receptor are found in 30–40% of patients with CN-AML. While the FLT3-internal tandem duplication (ITD) is associated with a poor DFS and OS when compared to CN-AML patients with a wtFLT3R, the prognostic significance of the FLT3-tyrosine kinase domain (TKD) mutation is less clear. There is considerable debate over the best postremission treatment for CN-AML patients who have a FLT3ITD. Our primary objective was to determine the relationship between different postremission therapies and outcome in CN-AML patients with FLT3 mutations. Methods: We retrospectively reviewed the outcomes of newly diagnosed AML patients at the Massachusetts General Hospital and Dana-Farber Cancer Institute/Brigham and Women Hospital between 2002–2008 who were younger than 60 years of age, had a normal karyotype, and had a FLT3ITD and/or TKD mutation at diagnosis. The method of Kaplan and Meier was used to estimate DFS and OS; groups were compared using the log-rank test. A p-value less than .05 was interpreted as statistically significant. Results: At diagnosis, there were 37 patients with an ITD mutation (7 also had a TKD mutation) and 19 patients with an isolated TKD mutation at diagnosis. Patients who received an allogeneic SCT were not included in the analysis. ITD positive patients with an allelic ratio (AR) of mutant to wt FLT3 〉.2 had a significantly worse DFS and OS than those with an AR

Description: Background: CCL3 and CCL4, previously called macrophage inflammatory protein-1 alpha (MIP-1α) and MIP-1β, are chemokines of the CC subfamily, which are secreted by CLL cells in response to B cell receptor (BCR) stimulation and by co-culture with nurse-like cells (NLC), presumably to attract T lymphocytes and monocytes that can deliver pro-survival signals. We previously reported, that CLL patients have elevated plasma levels of CCL3 and CCL4, which rapidly normalize after treatment with kinase inhibitors that target BCR signaling, such as the BTK inhibitor ibrutinib and the PI3K inhibitor idelalisib (Blood 119:1182-9, 2012)(Blood 118:3603-12, 2011). Moreover, in untreated CLL patients, high CCL3 levels are an independent prognostic marker for short time to disease progression (Blood 117:1662-9, 2011), emphasizing the clinical relevance of these chemokines in CLL. Aim: To better understand the dynamic changes, we analyzed plasma levels of CCL3 and CCL4 in a large cohort of serial samples from untreated CLL patients. Additionally, we analyzed separately from the entire cohort, CCL3 and CCL4 changes overtime in patients who subsequently received therapy and patients who were still on observation at time of analysis. Methods: CCL3 and CCL4 plasma levels were measured by ELISA in serial samples collected over time in untreated CLL patients. A total of 193 individual patients with 603 plasma samples were included in the analysis (entire cohort), of the 193 patients, 81 of them required treatment at a later time, although all the samples analyzed were collected prior any treatment, 112 patients did not need any intervention over the period of sample collection and were managed with observation. The duration between first and subsequent samples ranged between 5 months to 12 years. The duration between the date of the last sample and the date of the initiation of treatment ranged between less than a month to 88 months with a median of 6 months. Mixed linear models with repeated measure were used to determine whether there was a difference in mean levels of CCL3 and CCL4 over time. Results: The median and range of CCL3 and CCL4 levels for the initial sample over all 193 patients were 11.8 pg/mL (2.5 – 1977.1 pg/mL) and 88.8 pg/mL (6.9 – 2601.3 pg/mL) respectively. There were significant associations of time between first and subsequent serial samples (p

Description: Abstract 3426 Poster Board III-314 Lenalidomide is an immunomodulatory drug recently reported to have an objective response rate (ORR) of 35-53% in relapsed/refractory CLL, even with poor risk features. Initial therapy of CLL currently includes chemoimmunotherapy combinations like FCR, which have high response rates and improved PFS, but at the cost of myelosuppression and infection. Given the high ORR reported with lenalidomide and its potential to spare immune function, we undertook this Phase I study to explore the safety and tolerability of lenalidomide in combination with fludarabine and rituximab. Eligibility criteria included a confirmed diagnosis of CLL/SLL, previously untreated with systemic therapy with an indication for therapy by NCI-WG 1996 criteria; ANC 〉 1000, platelets 〉 50K, and adequate organ function. Six dose levels were planned, beginning with fludarabine 25 mg/m2 days 1-3, rituximab 375 mg/m2 day 1, and lenalidomide 2.5 mg administered on days 1-21 of a 28-day cycle. The study used a standard 3+3 dose escalation design, with dose limiting toxicity (DLT) assessed in the first 28 days and defined as grade 3 or greater non-hematologic toxicity, grade 4 neutropenia or thrombocytopenia, grade 3 febrile neutropenia, or a 〉2-week delay in initiating the second cycle. Nine patients were enrolled on this study, 7 males and 2 females, with a median age of 59 yrs (range 37-66). The median time from diagnosis to study entry was 66.1 months (range 11.7-82.8 months). The median WBC at study entry was 99.6 (range 11.1-325.7). Six patients had Rai 3-4 disease, and three patients had bulky lymphadenopathy (〉5 cm) on physical exam. Median beta-2 microglobulin level at study entry was 3.8 mg/L (range 2.5-7.7). The first cohort enrolled four patients, of whom two developed DLTs (grade 4 neutropenia persisting through day 50; febrile syndrome with grade 3 rash, myalgias and grade 4 CPK elevation). The study then proceeded to enroll five patients at dose level -1, with the same dosing of rituximab and fludarabine, but reduced lenalidomide dosing of 2.5 mg every other day for 21 of 28 days. Only two of five patients on dose level -1 completed the planned six cycles of therapy. The other three patients discontinued after 3 cycles due to toxicity: persistent grade 2 thrombocytopenia preventing further therapy; recurrent grade 3-4 neutropenia and thrombocytopenia, despite growth factor support, dose reduction and holding lenalidomide; and intermittent recurrent grade 3 tumor flare, rash and hand-foot syndrome, along with recurrent grade 3 neutropenia despite similar measures as above. At that point the study was closed to enrollment due to the significant myelotoxicity and idiosyncratic tumor flare, resulting in only two of nine patients completing the planned therapy. The median number of cycles completed for all patients was three. Six of the nine enrolled patients were evaluable for response, with five of nine having objective responses (56%, 90% CI 25-83%). No significant difference was observed in IgG and IgM levels between baseline and two months on study, although a trend toward a decrease in IgA levels (p=0.05) was observed. Baseline CD4 counts were variable (median 1139, range 529-2687), but showed significant declines of 574 cells by week 4 (p=0.003) and 707 cells by post-treatment (p=0.002). Analysis of alterations in serum cytokines is ongoing. We conclude that administering lenalidomide concurrently with fludarabine and rituximab is difficult, does not appear to preserve immune function, and limits the ability to deliver adequate therapy with either drug. Other trials currently in progress are exploring alternative schedules of lenalidomide administration with fludarabine and other standard CLL chemotherapy regimens; our data would favor a sequential schedule. Disclosures Brown: Celgene: Honoraria, Provided funding for this investigator-initiated study; Genzyme: Research Funding; Genentech: Consultancy; Calistoga: Consultancy. Off Label Use: Lenalidomide for CLL. Hochberg:Enzon: Speakers Bureau; Biogen-Idec: Speakers Bureau; Genentech: Speakers Bureau; Amgen: Speakers Bureau.

Description: Abstract 2689 Introduction: Acute promyelocytic leukemia (APL) is a distinct subtype of acute myeloid leukemia (AML) characterized by promyelocytes in the blood and bone marrow, coagulopathy, and characteristic translocation between chromosomes 15q22 and 17q21. Internal tandem duplication (ITD) mutations within the FMS-like tyrosine kinase receptor (FLT3) are found in approximately 30% of patients with APL. While cytogenetically normal AML patients with ITD mutations (especially with high ratio of mutant: WT allele) have a poorer prognosis due to high relapse rate, the prognostic value of ITD in APL is debated. We analyzed the clinical outcome of patients with APL to determine the relationship between FLT3 receptor status and prognosis. Materials and Methods: We conducted a retrospective analysis of all adult APL patients 〈 age 60 who were newly diagnosed and tested for FLT3R mutations at Massachusetts General Hospital and Dana-Farber Cancer Institute/Brigham and Women's Hospital between 2002 and 2008. Patient characteristics, complete remission (CR) rates, disease free survival (DFS), and overall survival (OS) were assessed by medical record review under an IRB-approved protocol. CR, DFS and OS were defined according to standard criteria. Kaplan-Meier method was used to estimate median DFS and OS and log-rank p-values were presented. We assessed the associations of clinical characteristics and the type of mutation patients had using Kruskal-Wallis tests. A p-value less than 0.05 was interpreted as statistically significant. Results and Discussion: We identified a total of 26 patients with APL for whom FLT3 mutation status was known. There were 13 patients with wtFLT3, 9 with ITD (1 also had TKD), and the remaining 4 had an isolated TKD mutation. Of the 13 patients with wt FLT3APL, 4 had prior chemotherapy and/or radiation. No patients in the FLT3ITD APL or FLT3TKD APL groups had prior chemotherapy or radiation. As we had only 4 patients had an isolated FLT3TKD mutation, we restricted our analyses to patients who had either wt FLT3 or FLT3ITD+/− TKD. Of the wtFLT3 APL patients, 9/13 were enrolled and treated on CALGB 9710. Of the remaining 4 patients, two were treated according to the PETHEMA regimen, one with a history of breast cancer and significant anthracycline exposure was treated with ATRA and AsO3, and the last patient died before therapy could be initiated. Of the 9 patients with FLT3ITD APL, 5 were treated on CALGB 9710, 2 were treated according to 9710 but off protocol, and 2 with the PETHEMA regimen. Baseline clinical characteristics were similar between the two groups including CR rates [92 (12/13) versus 100% (9/9) for wt and ITD, respectively]. Patients with the ITD had a lower fibrinogen at presentation than those with wt (103 vs. 235 mg/dl, p=.04). While patients with ITD appeared to have a higher WBC, it was not statistically significant (p=.1). However, patients with ITD had an inferior DFS compared to wt (Figure 1) (p=.01) while there was no significant difference in OS between the two groups (p=.39). Of the 5 patients with ITD who relapsed, 3 received AsO3 reinduction and autologous SCT, 1 with CNS recurrence received myeloablative allogeneic SCT, and the fifth died in relapse without treatment. The observation that FLT3ITD APL patients have a reduced DFS raises the possibility that APL therapy may be improved for this group of patients, possibly by incorporating FLT3 inhibitors. Disclosures: DeAngelo: Deminimus: Consultancy. Stone:Novartis: Consultancy.

Description: Abstract 465 CLL is among the most heritable of all cancers. To understand the genetic basis of this heritability, we have undertaken a comprehensive genomic analysis of familial CLLs including copy number analysis, gene expression profiling (GEP) and whole exome sequencing (WES). First, we examined whether familial and sporadic cases differ in the spectrum of acquired somatic mutations by WES of tumor and germline DNA of 36 familial CLLs (from 31 affected families). Compared to 55 sporadic CLLs, we observed that the somatic mutation rate in the familial CLLs was similar (mean 0.89 mutations/Mb (range 0.29–3.06) for the sporadics vs mean 0.97/Mb (range 0.11–3.78) for the familials, p=0.40). We also examined the spectrum of somatic mutations by testing for enrichment of 9 recently identified putative tumor drivers in our large CLL sequencing study (reported elsewhere in this meeting). We observed a similar distribution of these recurrent CLL mutations among the 36 familial CLLs as the 55 sporadic CLLs. These results were further confirmed by genotyping of the CLL driver mutations in an additional 32 familial and 67 sporadic CLLs. Collectively, these studies suggest that while the predisposing germline events may differ between familial and sporadic CLL, the spectrum of mutations and pattern of mutagenesis appear similar in the established CLL tumors. We therefore proceeded to examine the genetic characterization of germline DNA to identify predisposing loci, which we hypothesized might be enriched in a familial disease context. We first examined germline copy number variations (CNVs), which have not been previously characterized in this disease. We used high resolution Affymetrix 6.0 SNP arrays to study both tumor and germline DNA of 58 individuals representing 44 different families with CLL and lymphoproliferative disorders (LPDs). We identified two families (A and B) with autosomal dominant inheritance of CLL who carried distinct germline CNVs that affect genes previously implicated in CLL. Members of Family A carried a 525 kb germline deletion targeting DLEU7 at 13q14, but not affecting DLEU2, miR-15a, or miR-16–1. Importantly, by examining the tumor genome from these family members, we observed a uniform loss of the second allele of DLEU7 in 2/2 available CLLs from this family, suggesting an acquired “second hit” of a tumor suppressor gene. These findings underline the complexity of the most common somatically acquired copy number aberration (CNA) in CLL, 13q14 deletion, by demonstrating the role of additional regions other than the heavily investigated miRNA cluster. Members of Family B carried a 720 kb germline gain of 6p25 affecting the IRF4 gene, previously implicated in CLL through the identification of a GWAS risk allele located in the 3' UTR of IRF4, as well as the recent description of a recurrent somatic mutation affecting 1.5% of CLL cases. In Family B, the coding regions of the four genes located in this 6p gain, namely IRF4, DUSP22, EXOC2 and HUS1B, were sequenced, and no somatic mutations or novel SNPs were identified. However, the 6p gain in Family B represents an allele-specific enrichment of the haplotype carrying the GWAS risk SNP and, as previously described for that allele, results in lower expression of IRF4 in the two CLLs tested in this family. GEP further identified a signature associated with 6p gain that preserved low expression of IRF4 and showed high expression of KLF6. These results demonstrate that germline CNVs may facilitate the “path to cancer” by providing either an allelic deletion of a tumor suppressor or an amplification of a risk allele. As most familial CLL cases have not been accounted for by known SNPs or germline CNVs, we have initiated an in depth analysis of the WES germline results from familial cases compared to both sporadic CLL patients and normal individuals. Candidate variants have been filtered to exclude all SNPs described in the 1000 Genomes project and to focus on highly conserved sites. Thus far we have found that rare germline variants in patients with familial CLL contain a rich source of loci with relevance to B cell biology. Studies in progress are focused on further analysis of informative families and functional analyses of candidate variants. These comprehensive genomic analyses are expected to identify multiple cooperating genetic mechanisms that contribute to CLL pathogenesis, including CNVs and somatic and germline mutations. Disclosures: No relevant conflicts of interest to declare.

Description: Introduction: The Bruton tyrosine kinase (BTK) inhibitor ibrutinib (PCI-32765) has proved to be an effective targeted therapy in patients with chronic lymphocytic leukemia (CLL). A multicenter phase Ib/II study of ibrutinib in 85 patients with relapsed CLL or small lymphocytic lymphoma (SLL) showed the overall response rate of 71% with an additional 15 to 20% of patients having a partial response with lymphocytosis. A phase II trial of ibrutinib in combination with rituximab in 40 high-risk CLL patients conducted at MD Anderson Cancer Center demonstrated an overall response rate of 95%. However efficacy of ibrutinib in combination with rituximab vs. ibrutinib alone has not been directly compared. A randomized study of ibrutinib vs. ibrutinib+rituximab in patients with relapsed CLL was designed to compare long-term progression free survival rate in two groups (NCT02007044). Here we test a hypothesis that addition of rituximab may abrogate ibrutinib-induced lymphocytosis and accelerate response of other biomarkers, including plasma levels of chemokines CCL3 (MIP-1a) and CCL4 (MIP-1b), which are known to reflect the activation status of CLL cells. Methods: Patients were randomized to ibrutinib alone (arm A) or ibrutinib plus rituximab arm (arm B). Patients on arm A were treated with ibrutinib 420 mg orally once daily continuously throughout the study. Patients randomized to arm B received ibrutinib starting at day 1 or day 2. Rituximab (375 mg/m2) was administered intravenously on days 1, 8, 15, and 22 (cycle 1), then monthly during cycles 2-6. Study inclusion required previously treated CLL/SLL or treated or untreated high-risk disease (del17p or TP53 mutation). Serial blood samples were drawn at pre-treatment, then weekly during the first month, and monthly until 5/6 months. Plasma levels of CCL3 and CCL4 were measured at pre-treatment, 1 week, 1 month, 3 months and 5 months of follow-up by ELISA. A linear mixed effects model with randomly varying intercepts and slopes was constructed for each absolute lymphocyte count (ALC) measured repeatedly over time using SAS PROC MIXED. Models included an unstructured covariance and the Kenward-Rogers degrees of freedom adjustment to take into account unequally spaced unbalanced data. Models over at least 4 months of measurements included both linear and quadratic components. Results: Accrual began in December 2013; currently 93 of 208 patients are enrolled. 45 patients (48%) were randomized to arm A and 48 (52%) to arm B. Median ALC at pre-treatment was similar on both arms (26.28 and 25.14 K/µl). A mixed model with quadratic trend was used to assess the dynamic changes in ALC during the first month of treatment. The model confirmed a curvature pattern for arm A, consistent with lymphocytosis: ALC rose immediately after initiation of treatment and then began to stabilize. In arm B, however, no statistically significant changes in ALC were observed during this initial period. To appreciate the difference between two treatment arms over a longer time period, changes in ALC were analyzed in the 52 patients who had 4 or more months of follow-up. A mixed model with a quadratic trend demonstrated ALC downward trend for both arms, with a more rapid decrease in arm B (p=0.03 for the linear component; p=0.004 for the quadratic). The curvature in ALC reflecting initial lymphocytosis was clear in arm A, but not arm B. Descriptive analysis suggested that arm B had higher percentage of patients with normal (