Citations

Abstract Background: SGI-110 is a second generation HMA formulated as a dinucleotide of decitabine (DAC) and deoxyguanosine delivered as a small volume SC injection that yields longer half-life and more extended DAC exposure than DAC IV infusion. A phase 1 clinical trial of SGI-110 demonstrated a differentiated pharmacokinetic profile from DAC IV, potent hypomethylation, and clinical responses in previously treated patients with MDS and AML (Kantarjian et al. 2012). Phase 1 established 60 mg/m2 QDx5 as the biologically effective dose (BED), and 90 mg/m2QDx5 as the Maximum tolerated dose (MTD) in MDS patients given in 28-day cycles. Phase 2 is conducted to evaluate dose response between the BED and MTD in MDS patients. Methods: Int, or HR MDS, and CMML patients who were either treatment-naïve or previously treated were randomized to either 60 mg/m2 or 90 mg/m2QDx5 every 28 days. Efficacy was evaluated by the clinical responses of CR, PR, marrow CR (mCR), and Hematological Improvement (HI) based on the International Working Group Criteria 2006 as well as transfusion-independence. Adverse events (AEs) as graded by the CTCAE v4 criteria and pharmacodynamic biological activity as measured by Long Interspersed Nucleotide Element LINE-1 (an index of global DNA methylation) were also assessed. Patients are being followed for overall survival. Results: The study completed target enrolment with 102 patients: 53 who were refractory or have relapsed on prior treatment, and 49 previously untreated. Fifty three patients were randomized to 60 mg/m2 and 49 patients to 90 mg/m2 QDx5 with a median follow up of 8.2 months (range 1-21). Most baseline patient characteristics were well balanced between the 2 treatment arms with no significant differences in median age (71.7 vs. 72.5 y); male gender (70 vs. 61%); ECOG PS 2 (15 vs. 12%); HR MDS (28 vs. 37%); transfusion-dependence (57 vs. 55%), median baseline Hb (9.25 vs. 9.3 g/dL); platelets count (42.5 vs. 45 x109/L), and neutrophils count (1.19 vs. 1.16 x109/L) for the 60 and 90 mg/m2 QDx5 respectively. However, there were 15 CMML patients (28%) randomized to 60 mg/m2 vs. only 7 CMML patients (14%) randomized to 90 mg/m2(p=0.097). All but 2 patients in the previously treated group received one or more HMA. At the time of the data cutoff, 38 of 102 patients (37%) were still on treatment. Median number of treatment cycles was 4 for previously treated patients, and 5 for the treatment naïve group (range 1-22 cycles). CR+mCR were observed in 10/53 (19%), and 11/49 (22%) in the 60 and 90 mg/m2 arms respectively (p=0.8). CR was observed in 7/49 treatment naïve patients (14%) while CR+mCR were observed in 11/53 previously treated patients (21%) with no significant differences between 60 and 90 mg/m2 arms. Transfusion independence for at least 8 weeks was reported in 32% and 24% for platelets and RBCs respectively with no difference between the 2 treatment arms. Treatment naïve patients benefited from higher rate of transfusion-independence (58% and 46% for platelets and RBCs respectively). Potent demethylation was achieved in both treatment arms with a mean LINE-1 DNA demethylation of 27.3 and 30.5% for 60 and 90 mg/m2 respectively (p=0.12). Overall incidence of Grade ≥3 AEs regardless of relationship to treatment was reported in 81 vs. 88% for 60 and 90 mg/m2 treatment arms respectively (p=0.42).There was a slightly higher but non-significant difference in Grade ≥3 thrombocytopenia (51 vs 38%) and pneumonia (20 vs. 13%) for the 90 mg/m2 arm compared to 60 mg/m2. Early 8-week any-cause mortality occurred in 3/102 patients treated (3%), 2 at 60 and 1 at 90 mg/m2. Follow up for survival is still ongoing. Conclusions: SC SGI-110 is a well-tolerated novel HMA with biological and clinical activity in the treatment of Int and HR MDS, and CMML patients with particularly promising activity in patients previously treated with azacitidine or decitabine. Clinical responses, transfusion independence, DNA demethylation as assessed by LINE-1, and Safety did not significantly differ between the 2 treatment doses. Disclosures Tibes: Astex Pharmaceuticals, Inc.: Research Funding. Rosenblat:Astex Pharmaceuticals, Inc.: Research Funding. Yee:Astex Pharmaceuticals, Inc.: Research Funding. Griffiths:Astex Pharmaceuticals, Inc.: Research Funding. Issa:Astex Pharmaceuticals, Inc.: Consultancy, Research Funding. Naim:Astex Pharmaceuticals, Inc.: Employment. Taverna:Astex Pharmaceuticals, Inc.: Employment. Hao:Astex Pharmaceuticals, Inc.: Employment. Azab:Astex Pharmaceuticals, Inc.: Employment. Roboz:Astex Pharmaceuticals, Inc.: Consultancy.

Abstract Background: Persons of African ancestry (AA) have a 2-3-fold higher risk of multiple myeloma (MM) than persons of European ancestry (EA). Like other B-cell malignancies, genome-wide association scans (GWAS) have identified MM risk variants in the HLA region in persons of EA. We conducted a case-control analysis with data from the National Marrow Donor Program (NMDP)1comprising MM patients typed for bone marrow transplant to donor controls matched by race-ethnicity, and found associations between specific HLA alleles/haplotypes and MM risk that varied by race and ethnicity. To confirm our results and identify additional novel signals, we have now investigated associations between HLA alleles and haplotypes and MM risk in the African American Multiple Myeloma Study (AAMMS) Cohort. Methods: The source of subjects was the AAMMS, in which AA MM patients were identified from 10 cancer centers and 4 Surveillance, Epidemiology and End-Results (SEER) Program cancer registries in order to identify genetic risk factors for MM among AAs. A GWAS was conducted using the Illumina Human Core BeadChip array on DNA samples from 1,305 AA MM patients in the AAMMS comparing results to those from 7,078 AA controls with GWAS data generated from the Illumina 1MDuo2. The major histocompatibility complex (MHC) region single nucleotide polymorphisms (SNPs) were imputed to classical HLA variants using HIBAG. Unconditional logistic regression was used to estimate HLA associations, adjusting for sex, age and the first 2 principal components. P-values were adjusted for false discovery rate (FDR) for each locus group. Results: We did not identify any single HLA alleles associated with MM risk among AAs. However, several B*07:02-containing haplotypes were associated with MM risk (odds ratios [OR] ranging from 2.38 to 2.64 and FDR P-values ranging from 1.43 x 10-6 to 3.57 x 10-8). We found associations between MM risk and genotypes containing DRB3*02:02, including DRB3*02:02~DRB1*11:01+ DRB3*02:02~DRB1*11:01 (OR=1.93, PFDR= 9.36 x 10-5) similar to those observed in the NMDP study1. Novel findings included associations between MM risk and HLA Class I haplotypes B*53:01+ B*57:01 (OR=1.94, PFDR= 0.003) and C04:01~B*53:01+C*06:02~B*57:01 (OR=1.96, PFDR= 0.0050). Results from an ongoing meta-analysis between the two data sets (one based on an imputed GWAS and one based on NMDP HLA typing) will be presented. Conclusions: This study is the second to examine HLA alleles and risk of MM among AA's and is by far the largest. We confirmed a previously observed association between an HLA Class II DRB3 variant and MM risk and confirmed an association with B*07 haplotypes previously observed among EAs1. We also identified novel associations between other HLA Class I haplotypes and MM risk in AA's. Because HLA is highly polymorphic, many HLA alleles are rare variants for which genetic associations are difficult to detect without very large sample sizes. Further investigation with large sample sizes will be necessary to refine these associations in order to better identify the underlying causal alleles and determine the functional significance of these HLA associations. 1Beksac M, Gragert L, Fingerson S, et al.: HLA polymorphism and risk of multiple myeloma.Leukemia. 2016 Jul 27. doi: 10.1038/leu.2016.199. 2Rand KA, Song C, Hwang AE, et al. Genetic susceptibility markers of multiple myeloma in African-Americans. Abstract # 2030, 56th Annual American Society of Hematology Meeting, San Francisco, California, 2014. Disclosures Ailawadhi: Pharmacyclics: Consultancy; Novartis: Consultancy; Amgen Inc: Consultancy; Takeda Oncology: Consultancy. Nooka:Spectrum, Novartis, Onyx pharmaceuticals: Consultancy. Zonder:Pharmacyclics: Other: DSMC membership; Prothena: Consultancy, Honoraria; Celgene: Consultancy, Honoraria, Research Funding; Bristol Myers Squibb: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Takeda: Consultancy, Honoraria; Janssen: Consultancy, Honoraria. Lonial:BMS: Consultancy; Novartis: Consultancy; Millenium: Consultancy; Celgene: Consultancy; Janssen: Consultancy; Merck: Consultancy; Celgene: Consultancy; BMS: Consultancy; Novartis: Consultancy; Onyx: Consultancy; Janssen: Consultancy; Onyx: Consultancy.

Abstract Sorafenib is an oral multikinase inhibitor that blocks the autophosphorylation and activation of a number of kinases including the FLT3 tyrosine kinase with an internal tandem duplication mutation (FLT3-ITD). Sorafenib has activity against AML with FLT3-ITD mutations as a single agent in the setting of post-transplant relapse and in combination with chemotherapy for newly diagnosed AML. Historical Cancer and Leukemia Group B results using standard chemotherapy for older patients (pts) with FLT3-ITD have demonstrated a complete remission (CR) rate of 67%, median survival of 0.8 yrs, and 1-yr overall survival (OS) of 30% (Whitman et al., Blood 2010). More recently, sorafenib was demonstrated to significantly prolong event-free (EFS) and disease-free survival (DFS) of younger adults irrespective of FLT3 mutation status (Rollig, et al., ASH 2014). We hypothesized that the addition of sorafenib to induction and post-remission therapy would improve the overall survival of older pts with FLT3-mutated AML. The Alliance for Clinical Trials in Oncology conducted a multicenter, single-arm phase 2 study in pts ≥ 60 years old with AML and either a FLT3-ITD or a point mutation in the activation loop of the kinase domain (FLT3-TKD). Subjects with PML-RARA, core-binding factor AML or who had received prior treatment for AML were excluded. Induction chemotherapy consisted of cytarabine 100 mg/m2 CIVI on days 1-7 and daunorubicin 60 mg/m2 IV on days 1-3 (7+3) with oral sorafenib 400 mg bid on days 1-7. Those not achieving a hypoplastic bone marrow on day 14 were to receive a second cycle of cytarabine and daunorubicin (5+2) plus sorafenib 400 mg bid on days 1-7. Post-remission therapy consisted of intermediate-dose cytarabine 2 g/m2 on days 1-5 with sorafenib 400 mg bid on days 1-28 for 2 cycles followed by maintenance sorafenib 400 mg bid for 12 28-day cycles. The primary endpoint of the study was 1-yr OS. With a sample size of 39 FLT3-ITD patients and type I error rate of 10%, this design provided 90% power to distinguish between true 1-yr OS rate of 50% compared to our historical rate of 30% for the FLT3-ITD pts. A total of 474 pts were screened for FLT3 mutations though a central laboratory. FLT3 mutations were identified in 83 subjects (17.5%). Fifty-four pts with a median age of 67 years (60.3-82.7) were enrolled, including 39 with FLT3-ITD (71%) and 15 with FLT3-TKD (29%). All pts have now completed protocol treatment. Of the 54 pts, 37 achieved a CR or CRi (69%). The 30-day induction mortality was 9% with no additional deaths occurring between days 30-60. With a median follow up of 28.3 months, the observed 1-yr OS (95% CI) was 62% (45-78) for the FLT3-ITD pts and 71% (42-92) for the FLT3-TKD pts. The median DFS and OS was 12.5 mo (7.7-17.4) and 15.0 mo (10.4-20.1) respectively in the FLT3-ITD group and 9.0 (1.6-NA) and 16.2 mo (5.0-NA) for the FLT3-TKD group. For all subjects, the 2-yr OS was 28% (17-43) and DFS was 27% (15-46). There were no treatment-related deaths during consolidation or maintenance. For pts receiving sorafenib maintenance, the most commonly reported related adverse events were Grade 1 diarrhea, fatigue, transaminitis, and Grade 2 palmar-plantar erythodyesthesia. The mean plasma inhibitory activity (PIA) for FLT3 was 94% (n=25) on day 6 of induction, 93% (n=15) on day 15 of consolidation and 99.2% (n=6) on day 15 of maintenance demonstrating robust inhibition of FLT3 in vivo. This study represents the first prospective clinical trial for older adults with AML targeting a specific mutational profile within the US cooperative group setting. The study met the primary endpoint demonstrating that the addition of sorafenib to chemotherapy for FLT3-ITD AML improves the survival of older adults more than doubling the 1-yr OS compared to historical controls (62% vs 30%, p <0.0001). Disclosures Uy: Novartis: Research Funding. Off Label Use: Sorafenib for AML. DeAngelo:Novartis: Consultancy; Ariad: Consultancy; Bristol Myers Squibb: Consultancy; Pfizer: Consultancy; Amgen: Consultancy; Incyte: Consultancy; Agios: Consultancy; Celgene: Consultancy. Stock:Gilead: Membership on an entity's Board of Directors or advisory committees. Byrd:Acerta Pharma BV: Research Funding. Stone:Celator: Consultancy; Celgene: Consultancy; Pfizer: Consultancy; Roche/Genetech: Consultancy; Amgen: Consultancy; Karyopharm: Consultancy; Abbvie: Consultancy; AROG: Consultancy; Merck: Consultancy; Sunesis: Consultancy, Other: DSMB for clinical trial; Novartis: Research Funding; Juno: Consultancy; Agios: Consultancy. Larson:Bristol-Myers Squibb: Consultancy; Pfizer: Consultancy; Ariad: Consultancy, Research Funding; Novartis: Consultancy, Research Funding.

Abstract Multiple myeloma (MM) is 2-3 times more common among African-Americans compared to non-Hispanic whites. The 2-3-fold increased risk among family members of cases suggests a genetic contribution to risk. Genome-wide association studies (GWAS) in populations of European ancestry have identified seven novel risk loci at 2p23.3 (rs6746082), 3q26.2 (rs10936599), 3p22.1 (rs1052501), 6p21.32 (rs2285803), 7p15.3 (rs4487645), 17p11.2 (rs4273077) and 22q13.1 (rs877529) (Broderick, et al. Nat Genet, 2011, Chubb, et al. Nat Genet, 2013), three of which were replicated in another European series (Martino et al., Br J Haematol, 2012). Here we examined the index signals and conducted fine-mapping for each locus in a case-control study of 1,049 multiple myeloma cases and 7,084 controls of African ancestry to identify better markers of risk and novel independent loci in seven previously reported regions in this high risk population. Incident cases were recruited from 10 clinical centers and SEER cancer registries from 2011 to 2013 and genotyped using the Illumina HumanCore GWAS array. Control data were obtained from previous genome-wide studies of breast and prostate cancer, genotyped using the Illumina 1M-Duo in 4425 male controls from the African Ancestry Prostate Cancer Consortium (consisting of 14 independent studies) and 2632 female controls from a breast cancer GWAS of African-American women (consisting of 9 independent studies). Imputation to 1000 Genomes (March 2012 release) was conducted for regions around six of the previously identified single nucleotide polymorphisms [SNPs] (the HLA region harboring rs2285803 is still being imputed, results will be presented). A case-control analysis of SNPs/indels >1% frequency within 250 kb of each index variant was conducted using unconditional multivariable logistic regression adjusting for age, sex and five leading principal components. Region-specific alpha levels were determined through permutation tests. The minimum alpha level across the six regions was α=0.002. All previously reported risk variants were common in African-Americans (minor allele frequency [MAF] >0.05). For five of the six SNPs, we had ≥94% power to detect the same effect observed in non-Hispanic whites, and 64% power for the less common variant rs10936599 (MAF=0.07). We observed directionally consistent effects (odds ratio [OR]>1) for the six risk variants tested, with three replicating at p≤0.05 (7p15.3, p=1.4x10-7; 17p11.2, p=0.05; 22q13.1, p=0.02). For three of the six regions, we observed better markers of risk in African-Americans that were correlated with the index SNP in Europeans (7p15.3, rs56333627, p=1.5x10-5, r2=0.89; 17p11.2, rs34562254, p=2.9x10-3, r2=0.90; 22q13.1, rs2092410, p=1.1x10-4 r2=.71). The missense variant identified in the 17p11.2 region (rs34562254, Pro251Leu) is located in TNFRSF13B, which encodes the protein TACI, a B cell surface receptor which plays a role in B cell maturation, apoptosis and antibody production by inducing activation of transcription factors including NFAT and NFκβ. In addition, there is evidence suggesting that TACI is involved in MM pathogenesis. Our results demonstrate that many of the risk loci for MM found in European ancestry populations are also risk loci in men and women of African ancestry and that by fine-mapping, we are able to identify variants that better capture risk in populations of African ancestry. Disclosures Terebelo: Celgene Corp: Membership on an entity's Board of Directors or advisory committees. Lonial:Millennium: The Takeda Oncology Company: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Bristol-Myers Squibb: Consultancy, Research Funding; Onyx Pharmaceuticals: Consultancy, Research Funding.

Abstract Purpose: Clinical trials in APL have generally reported similar outcomes for children and adults. However, sub-group analysis of pediatric patients treated on 2 consecutive trials of the European APL group demonstrated that children <5 years (yr) had an increased risk of relapse (Bally et al., JCO, 2012). Here we report the effect of age < 5 yr on outcome for the North American Intergroup Study C9710. Methods: Outcomes were compared among young children (age <5 yr, n=16) to those of older age (5-12 yr, n=25 and 13-18 yr, n=45) enrolled on C9710. Treatment on C9710 included induction with daunorubicin, cytarabine and ATRA followed by two consolidation cycles with daunorubicin and ATRA and maintenance for one year with ATRA +/- mercaptopurine and methotrexate. Patients of age ≥15 yr were randomized to receive or not receive two 5-week cycles of arsenic trioxide (ATO) before the chemotherapy consolidation. However, analysis in this study of young children compared to those of older age was restricted to patients who did not receive ATO. Results: Clinical remission (CR) rates were similar between the 3 age groups (age <5 yr= 88%, age 5-12 yr= 80%, age 13-18 yr= 82%; P=0.93). Overall survival (OS) Kaplan-Meier estimates were not significantly different (log rank P=0.84), and 5 yr OS rates were 77% (age <5 yr), 80% (age 5-12 yr), and 84% (age 13-18 yr). Similarly, event-free survival (EFS) Kaplan-Meier estimates were not significantly different (log rank P=0.48), and 5 yr EFS rates were 56% (age <5 yr), 47% (age 5-12 yr), and 59% (age 13-18 yr). Young children did not have a higher relapse rate (5-yr cumulative incidence of relapse by age: <5 yr=29%, 5-12 yr=49%, 13-18 yr=29%). Conclusions: Results of this intergroup trial demonstrate that ATRA administration during induction, consolidation and maintenance leads to CR rates >80% and 5 yr OS rates >75%. These results are superior to recently published results of pediatric non-APL acute myeloid leukemia and confirm results from the prior APL intergroup trial (INT0129) that demonstrated that ATRA in induction and/or maintenance significantly improved outcomes for pediatric APL (Gregory et al., Ped Blood and Cancer, 2009). Our sub-group analysis showed no association of age with outcomes in children treated with this regimen. Disclosures Off Label Use: Daunorubicin- labeled for use in AML for adults and for pediatric ALL (not pediatric AML) Arsenic Trioxide- labeled for use in relapsed/refractory APL (not de novo APL) Mercaptopurine and Methotrexate- labeled for use in pediatric ALL (not APL).

Abstract Purpose: Recent studies have demonstrated differential outcomes for adolescents and young adults (AYAs) vs. children treated for acute myeloid leukemia. Here we analyze outcomes by age group for APL patients treated on the North American Intergroup Study C9710. Methods: Patient characteristics and outcomes were compared between children (<15 yr, n=55) vs. AYAs (15-31 yr, n=42). To control for differences in treatment center, only young adults registered at Children’s Oncology Group (COG) treatment sites were included. Treatment included induction with daunorubicin, cytarabine and ATRA. All patients (pts) received two consolidation cycles with daunorubicin and ATRA and maintenance for one year with ATRA +/- mercaptopurine and methotrexate. Pts ≥15 yr were randomized to an additional two cycles of arsenic trioxide (ATO). Sub-group analysis was also performed to exclude pts ≥15 yr who received ATO and thus compare only patients <15 yr vs. ≥15 yr who had not received ATO. Results: Children and AYAs had similar characteristics of gender, race/ethnicity, Sanz risk grouping and M3v histology. Among all pts enrolled at COG sites on the study, complete response (CR) rates were similar (<15yr =85% vs. ≥15yr =81%, P=0.59). AYAs had improved 5 yr disease-free survival (DFS) from time of CR compared to younger pts (<15 yr =47% vs. ≥15 yr =70%, log rank P=0.05), and there was a trend toward difference in 5 yr event-free survival (EFS) (<15 yr =46% vs. ≥15 yr =63%, log rank P=0.13). Overall survival (OS) at 5 years was not significantly different (<15 yr =82% vs. ≥15 yr =75%, log rank P=0.47). Twenty-three of 42 pts ≥15 yr received ATO consolidation cycles, which may have contributed to their improved DFS and EFS. Thus, we analyzed outcomes by removing from the analysis patients ≥15 yr who received ATO consolidation cycles, and compared pts <15 yr who did not receive ATO (n=55) to pts ≥15 yr who were randomized to no ATO (n=19). CR rates were again similar (<15 yr =85% vs. ≥15 yr no ATO =84%, P=1.0). There was no significant difference in other outcomes including 5-yr DFS from time of CR (<15 yr =47% vs. ≥15 yr no ATO =63%, log rank P=0.31), 5-yr EFS (<15 yr =46% vs. ≥15 yr no ATO=53%, log rank P=0.89), or 5-yr OS (<15 yr =82% vs. ≥15 yr no ATO=79%, log rank P=0.78). Conclusions: The C9710 study demonstrated improved EFS, DFS and OS among adult patients receiving ATO consolidation compared to those who did not receive ATO (Powell et al., Blood, 2010). The current analysis of results for AYAs versus younger pediatric pts showed improved DFS among the whole AYA cohort (of which 55% received ATO consolidation) compared to pts <15 yr (of which 0% received ATO). When the analysis was restricted to AYA pts not receiving ATO, however, the results of this intergroup APL trial showed similar CR, DFS, EFS and OS for AYAs and younger pts. Thus, unlike other subtypes of AML, APL appears to have a consistent response across the pediatric and AYA age groups. The most recent COG clinical trial in APL is now evaluating whether pediatric patients can also benefit from ATO consolidation. Disclosures Off Label Use: Daunorubicin- labeled for use in AML for adults and for pediatric ALL (not pediatric AML) Arsenic Trioxide- labeled for use in relapsed/refractory APL (not de novo APL) Mercaptopurine and Methotrexate- labeled for use in pediatric ALL (not APL).

Abstract Abstract 757 Background: Chronic lymphocytic leukemia (CLL) is the most common leukemia seen in adults. Previous population-based research in CLL has not represented the current ethnic makeup and diversity within the US population. We undertook a large Surveillance Epidemiology and End Results (SEER) based analysis to assess disparities in outcome among different ethnic subgroups of CLL patients, so that therapeutic resources can be utilized effectively. Methods: The SEER-18 Registry (1973–2009) was used to retrieve a population of 60,293 patients with established diagnosis of CLL (Histology code: 9823). The cohort was restricted to adults (>18 yrs) diagnosed after January 1, 1992, since ethnic minorities were underrepresented in SEER prior to that. Cases reported in death certificate/autopsy, those without a confirmatory diagnosis, those with ethnicity recorded as unknown/other or those that did not have follow up post diagnosis were excluded. Using Cox proportional hazards models, adjusted for confounding variables, association between patient characteristics and survival (overall; OS and CLL-specific; CLLSS) was evaluated. Analysis was performed utilizing the SAS software (v9.2). All statistical tests were two-sided with a significance level of 5%. Results: A total of 37,834 patients (22,581 males; 60%, 15,253 females; 40%) were included in the final analysis. Age group (yr) cohorts included: 18–44 (861; 2%) 45–54 (3,587; 9%), 55–64 (7,684; 20%), 65–74 (10,966; 29%), and ≥75 (14,736; 39%). For survival trends over time, stratification by year of diagnosis was used: 1992–1999 (9,837; 26%), 2000–2004 (14,129; 37%), 2005–2009 (13,868; 37%). Stratification by race/ethnicity included: White (32,980; 87%), African-American (AA: 2,486; 7%), Hispanic (1,536; 4%), Asian (737; 2%), and Native American (NA: 95; 0.25%). Median age ranged from 68 yr (AA and Asian) to 71 yr (Whites) (Figure 1). For survival analysis, NA ethnic group was excluded due to their small numbers. Females had a significantly better median OS (8.3 yr vs. 6.4 yr; p<0.001) as well as CLLSS (p<0.001) than males, when evaluated for the whole population and across different age groups. Both, OS and CLLSS progressively and significantly (p<0.001) declined with advancing age. Progressive significant improvement was seen in OS over time across cohorts by year of diagnosis (6.3 yr, 7.3 yr, 7.4 yr; p<0.001). This improvement in OS persisted over time for age groups <75 yr, but for ≥75 yr, the improvement was lost in the most recent time period cohort (2005–2009). The percentage of patients dying of CLL has decreased progressively over time suggesting an improvement in therapy. Among various ethnicities, Asians had the best (7.5 yr) and AA had the worst (4.8 yr) OS (p=0.006) (Figure 2) and CLLSS (p<0.001). In a multivariate model, AA continued to have the worst OS across all age groups (p<0.001), while Whites had the best OS in age groups <75 yr and Asians had the best among patients ≥75 yrs. Percentage of patients dying from CLL was highest in AA across age groups, but lowest for Whites in age groups <75 yr, while in Hispanics in patients ≥75 yr. Trends in CLLSS were noted to be more heterogeneous among ethnicities than trends in OS. Conclusions: This report presents the largest population-based analysis of patient characteristics including ethnicity and survival in CLL in the era of novel therapeutic agents. Older patients had worse OS among all ethnic subgroups. Progressive improvement in CLLSS was noted over time suggesting a possible impact of new therapeutics, however this was not seen in ≥75 yr patients. AA patients had the worst OS among all ethnicities and age groups. This survival difference was most significant and pronounced in younger patients, which has important implications. For example, this survival disadvantage may represent poor access to optimal healthcare warranting development of new initiatives directed towards this ethnic minority. On the other hand discrete biological variants may underlie the more aggressive disease biology in younger AA patients underscoring the need to further investigate and delineate the molecular biology of CLL in ethnic minorities. Studies of outcome disparities are important not only for evaluating disease characteristics but also management needs of specific patient populations and are critical for designing programs to deliver optimal healthcare to ethnically diverse populations. Disclosures: No relevant conflicts of interest to declare.

Abstract Acute lymphoblastic leukemia (ALL) in adolescents and young adults (AYA) is characterized by distinct presenting features and inferior prognosis compared to pediatric ALL. Age, as a continuous variable, is negatively correlated with prognosis, in spite of risk-adapted combination chemotherapy. To better understand the etiology of ALL in this age group, we performed the first genome-wide association study (GWAS) to comprehensively examine germline single nucleotide polymorphisms (SNPs) for their association with susceptibility to B-ALL in AYAs. In the discovery GWAS, we compared genotype frequency at 635,297 SNPs between 308 AYA ALL cases (age 16-39 years, treated on the Children’s Oncology Group [COG], the Alliance-Cancer and Leukemia Group B, Eastern Cooperative Oncology Group, MD Anderson Cancer Center, and St. Jude Children’s Research Hospital trials) vs. 6,661 non-ALL controls. The association between genotypes at each SNP and ALL susceptibility was tested by using a logistic regression model after adjusting for genetic ancestries to control for population stratification. SNPs that reached P≤ 5×10-8 in the discovery GWAS were tested in an independent cohort of 82 AYA ALL cases from the COG protocols and 5,755 non-ALL controls. We identified a single genome-wide significant susceptibility locus on 10p14 signified by two SNPs within the GATA3 gene: rs3824662, P=2.8x10-10, odds ratio (OR)=1.77; rs3781093, P=3.2x10-9, OR=1.73, both of which were validated in the replication cohort (P=1.9x10-8 and P=4.3x10-5, respectively). We also examined the association signals in AYAs for susceptibility loci previously identified in pediatric ALL: ARID5B, IKZF1 and PIP4K2A variants were nominally significant in AYAs in the discovery GWAS and/or in the replication analysis, whereas CEBPE or CDKN2A/CDKN2B were not significant. These results imply both similarities and differences in genetic predisposition to ALL between children and AYAs. At the GATA3 locus, rs3824662 risk variant was over-represented in Philadelphia chromosome (Ph)-like ALL in AYAs (P=0.02), confirming our previous report of Ph-like ALL susceptibility variants in GATA3 (Nat Genet 45:1494). Importantly, even after excluding Ph-like cases, rs3824662 remained associated with the risk of developing ALL in AYAs, suggesting that the influence of the GATA3risk variant on ALL susceptibility in AYAs extends beyond the predisposition to Ph-like subtype. We next examined the relationship between GATA3 risk allele frequency and age at diagnosis in a cohort of unselected childhood and adolescent ALL cases enrolled in the COG P9900 protocols (N=1,827). Dividing patients into four consecutives age groups (<5, 5-10, 10-15 and >15 years), we observed a clear progressive increase in the risk allele frequency at rs3824662 (P=6.29×10-11) with increasing allelic odds ratio (i.e. relative risk of ALL conferred by each copy of the risk allele). This correlation between genotype and age was evident regardless of genetic ancestry, although the risk variant was more common among individuals with higher Native American ancestry. In contrast, the frequency of ALL susceptibility variant in ARID5B decreased progressively with increasing age at diagnosis (P=0.006), whereas PIP4K2A, CDKN2A/CDKN2B, IKZF1 and CEBPE variants were not related to age. Finally, we compared rs3824662 risk variant frequency by age in the COG P9900 protocols after stratifying the ALL cases into TCF3-PBX1, ETV6-RUNX1, hyperdiploid, MLL-rearranged and B-other. There was a trend that the risk allele was more frequent in cases older than 16 years compared to those below 16 in the five subtypes examined. In conclusion, we have identified inherited GATA3 genetic variants that strongly influence ALL susceptibility in adolescent and young adults, indicating potential age-related differences in ALL biology. Disclosures No relevant conflicts of interest to declare.

Abstract Abstract 1038 The relationships of inherent disease-related characteristics and of treatment to the nature of APL cell clones that emerge at disease relapse are poorly understood. We studied these relationships in 45 patients who relapsed after ATRA/chemotherapy treatment on the non-arsenic trixoxide arm of intergroup phase III trial C9710 (Powell, et al. Blood, Epub). Four variably-expressed APL cell clonal markers were assessed: PML-RARA ligand binding domain mutations (LBD-M), FLT3 internal tandem duplication mutations (ITD), FLT3 receptor tyrosine kinase mutations (D835; RTK-M) and additional chromosome abnormalities (ACA), i.e., in addition to the hallmark t(15;17). The methods for mutation analysis of PML-RARA and FLT3 have been reported and were RNA transcript-initiated supplemented by selected quantitative FLT3 mutation analyses initiated from DNA. Karyotype data were derived by standard cytogenetic methods. Four patients were excluded from analysis because the sample level of PML-RARA, as determined by quantitative RT-PCR, was insufficient to exclude possible false-negative mutation results. In 41 evaluable patients, the marker incidences were: LBD-M, 44%; ITD, 37%; RTK-M, 12%; ACA (22 tested), 59%. At presentation, the corresponding incidences were: LBD-M, 0% (by high-sensitivity testing a minor subclone of the relapse mutation was found in 2/7 patients tested); ITD, 43% (37 tested); RTK-M, 22% (37 tested); ACA, 24% (34 tested). The low frequency of RTK-M impeded further analysis. The other markers, using relapse determinations, were assessed for potential associations between the markers and with the following parameters: age, sex, presenting WBC count, PML-RARA type, time to relapse, relapse on or off ATRA treatment (on = taking or within 30 days of discontinuing) and post-relapse survival. The most essential positive findings are summarized below:Association of Marker with Parameter or Other MarkerCases with ParameterCases without Parameterp-valuePretreatment parameters: LBD-M with WBC count < 5,000/uL12/18 (67%)7/23 (30%)0.030 ITD with WBC count >10,000/uL10/15 (67%)5/26 (19%)0.006 ITD with S-form PML-RARA15/15 (100%)10/26 (38%)<0.0001Relapse parameters: ITD with lack of ACA1/7 (14%)12/15 (80%)0.007 LBD with relapse On-ATRA11/18 (61%)7/23 (30%)0.064 Off-ATRA only: LBD with lack of ITD0/7 (0%)9/16 (56%)0.019 The results confirm reports, which tested pretreatment samples, of a strong association between FLT3ITD mutations present at relapse and high WBC count and S-form PML-RARA. At relapse, there was, also, divergent selection of ITD-harboring vs ACA-harboring clones. LBD-M occurred more frequently in patients with low presenting WBC counts. They, also, occurred in patients who relapsed while on ATRA, consistent with an ATRA selective role for clones harboring LBD mutations. A significant segregation of patients with LBD-M or ITD was observed after relapse off ATRA treatment (p = 0.019), while no segregation was observed in overall relapse patients (p = 0.346) or after relapse on ATRA (p = 0.316). In 2 patients who relapsed on ATRA, quantitative analysis indicated that LBD-M and ITD mutations must be present in the same APL cell clone. There were no significant survival differences related to any clonal marker (post-relapse survival = 66%). Although the small number of cases is cautionary, the overall data suggest that LBD-M may have an overriding, dominant APL clone selection role in relapse that occurs on ATRA therapy. Other inherent clinicopathologic characteristics of APL may predispose to divergent molecular pathways of disease progression and relapse in patients alternatively harboring clones with LBD-M or ITD after the termination of ATRA selection pressure. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 464 Myeloid-derived suppressor cells (MDSC) are important, albeit incompletely characterized, immune regulatory cells which suppress host anti-tumor immunity in malignancies like melanoma. We sought to address 3 outstanding questions in the present study: i) does tumor burden drive the expansion of MDSC; ii) do MDSC express immunosuppressive B7-H family members,;iii) are MDSC depleted following chemotherapy administration. As MDSC expand in response to tumor-derived factors, including those elevated in melanoma patients, we hypothesized that the frequency of MDSC present in the peripheral blood of melanoma patients may be dependent on tumor stage. Therefore, PBMC were obtained from both normal donors and melanoma patients with all stages of disease (stage I, n=10; stage II, n=10; stage III, n=10; stage IV, n=47). A discreet population of CD14+HLA-DR-/lo MDSC was appreciated in samples obtained from melanoma patients, but was largely absent in normal donors. Approximately 0.5% of PBMC were CD14+HLA-DR-/lo in normal donors. In contrast, the frequency of these cells progressively increased with tumor stage in melanoma. While a two-fold increase in CD14+HLA-DR-/lo MDSC was observed in stage I patients, this difference did not reach statistical significance. In contrast, a significant increase in CD14+HLA-DR-/lo MDSC was observed in patients with stage II-IV disease (2.2% Stage II, p<.01; 3.4% stage III, p<.001; 13.2% stage IV, p<.001). While the observed range (range: 0.8-66.3%) in frequency of CD14+HLA-DR-/lo MDSC was broad in stage IV melanoma, these cells comprised more than 10% of total PBMC in the majority of patients. Various B7 homologues (e.g. B7-H1, B7-H2, B7-H3 and B7-H4) may inhibit T-cell activation, proliferation and cytokine production or promote the induction of T-cell apoptosis and unresponsiveness. Due to their immunosuppressive affects, and expression by tumor cells or immune regulatory cells (e.g. Treg, MDSC), these B7-H members are associated with adverse outcomes in multiple human malignancies. As the expression of these important immunoregulatory molecules on MDSC is unknown, we sought to determine whether CD14+HLA-DR-/lo MDSC in melanoma may express B7-H family members. Therefore, PBMC from patients with metastatic melanoma were stained with either an isotype control or antibodies against B7-H1, B7-H2, B7-H3 or B7-H4. In all cases, B7-H family members were more highly expressed on CD14+HLA-DR-/lo MDSC when compared with the expression observed on CD14+ cells from normal donors. While B7-H1 (p<.01), B7-H2 (p<.001) and B7-H4 (p<.01) were preferentially expressed on HLA-DR-/lo cells in melanoma patients, B7-H3 was more highly expressed on both HLA-DR+ (p<.05) and HLA-DR-/lo (p<.001) CD14+ cells. Serial blood samples, both at baseline and at least 12 weeks after therapy, were availabel for 10 patients with metastatic melanoma undergoing therapy wither either carboplatin and nanoparticle albumin bound paclitaxel (nab paclitaxel; n=6) or carboplatin, paclitaxel and bevacizumab (n=4). The frequency of both CD14+HLA-DR-/lo MDSC and conventional CD14+HLA-DR+ monocytes was determined both before and after chemotherapy. An approximately 50% reduction in the frequency of CD14+HLA-DR-/lo MDSC was observed following chemotherapy (p=.004), with no significant reduction in the frequency of conventional HLA-DR+ monocytes (p=.08). Collectively, the data presented implicates multiple B7-H family members in MDSC-mediated immune suppression and demonstrates that the expansion of B7-H+CD14+HLA-DR-/lo MDSC in melanoma is directly associated with the tumor burden, but may be abrogated by the administration of taxane-based chemotherapy. Disclosures: No relevant conflicts of interest to declare.