Ketterling, Rhett P.

Abstract Introduction: The therapy of relapsed/refractory AML remains poor and new treatment options are sorely needed. The Leukemia Committee of ECOG devised a randomized phase II trial of 3 novel regimens to determine overall response rates could be improved. Methods: Patients between the ages of 18 and 70 years with a prior diagnosis of AML who relapsed <1 year after initial achievement of complete remission (CR) or who were refractory to initial induction therapy (<2 courses) or to first reinduction after a relapse (<1 course) were eligible for this trial. Eligible patients were randomized equally to one of three regimens described in Table 1 below. The primary endpoint was achievement of a morphologic complete remission (CR) or a CR with incomplete blood count recovery (CRi). A 2 stage statistical design was planned with 18 patients (assuming 16 eligible) to be accrued in the first stage. If at least 3 CR+CRi were seen among the 16 eligible patients, an additional 19 patients (assuming 17 eligible) would be entered in the second stage. If 8 or more CR+CRi were seen in the 33 eligible patients, this would be considered evidence of a promising response rate. Results: This study activated October 16, 2008 and closed August 2, 2013 with a total of 92 patients accrued (91 eligible). The eligibility criteria were amended during the study to lower the upper age limit from 70 to 65 years because 6 treatment related deaths occurred in the initial 27 patients accrued to the FLAM regimen. All of these patients died of septic shock or multi-organ failure and 5 of the 6 died within 1 month of the start of therapy. Most of these patients were over age 65 ( 53, 62, 67, 68, 69 and 69 years). At the first stage analysis, of the first 16 eligible patients accrued to each arm there were 2 CR and 2 CRi in the CT arm, 2 CR and 2 CRi in the FLAM arm and 2 CR in the sirolimus-MEC arm. Thus, the sirolimus-MEC arm was closed to accrual with a total of 20 eligible patients and second stage accrual continued for the CT and FLAM arms. Among the 35, 36 and 20 eligible patients accrued to arms A, B and C, respectively, there were 2 CR and 3 CRi in the CT arm, 6 CR and 4 CRi in the FLAM arm, and 2 CR and 1 CRi in the sirolimus-MEC arm. All 10 CR or CRi of the FLAM arm were among the first 33 eligible patients, therefore FLAM was the only regimen to achieve our target. The overall response rates for arms A, B and C were 14% (90% CI 7.1% - 34.7%), 28% (90% CI 16.1% - 43.0%) and 15% (90% CI 4.2% - 34.4%), respectively. Among the 92 patients accrued, there were 5, 10 and 2 grade 5 events within 30 days of the end of therapy in arms A, B and C, respectively. The vast majority of these deaths were attributed to infection and multi-organ failure except two which were secondary to disease. Conclusion: In this high risk group of AML patients the FLAM regimen showed an encouraging response rate. A randomized phase II trial has also recently been completed exploring its use in the front-line setting with newly diagnosed high-risk AML. While the FLAM regimen was excessively toxic in elderly patients with relapsed or refractory disease, it is a suitable option for younger patients in this setting. Arm A: CT Carboplatin 150 mg/m2/day over 24 hours by continuous infusion IV for 5 days (days 1-5) Topotecan 1.6 mg/m2/day over 24 hours by continuous infusion IV for 5 days (days 1-5) Arm B: FLAM Flavopiridol 30 mg/m2 for ½ hour followed by 60 mg/m2 over 4 hours IV per day for 3 days Cytarabine 667 mg/m2 over 24 hours by continuous infusion IV for 3 days (days 6-8) Mitoxantrone 40 mg/m2 IV over 1-2 hours (day 9) Arm C: Sirolimus-MEC Sirolimus 12 mg by mouth on day 1, followed by 4 mg by mouth daily for 8 days (days 2-9) Mitoxantrone 8 mg/m2/day over 15 minutes IV for 5 days (days 4-8) Etoposide 100 mg/m2/day over 1 hour IV for 5 days (days 4-8) Cytarabine 1000 mg/m2/day over 3 hours IV for 5 days (days 4-8) Table 2. Responses Table 1. Chemotherapy Regimens Regimen n Median age (range), years Responses (%) Deaths (first 30 days) (%) CT 35 55 (23-67) 5 (14%) 5 (14%) FLAM 36 62 (19-69) 10 (28%) 10 (28%) Sirolimus-MEC 20 52 (22-68) 3 (15%) 2 (10%) Disclosures Off Label Use: Sirolimus-potentiate chemotherapy Carboplatin-chemotherapy for AML treatment Topotecan-chemotherapy for AML treatment.

Abstract BACKGROUND. In primary myelofibrosis (PMF), survival from time of diagnosis is predicted by the international prognostic scoring system (IPSS), and from time of referral, by the dynamic IPSS (DIPSS) or DIPSS-plus. JAK2/CALR/MPL mutational status and presence and number of other prognostically-relevant mutations (ASXL1, SRSF2, EZH2, IDH1/2) provide IPSS/DIPSS-plus independent prognostic information. The objective in the current study was to revise IPSS, by including mutation-relevant prognostic information. METHODS. The study included 986 PMF patients divided into learning (n=588) and validation (n=398) cohorts. Previously published methods were used to screen for mutations involving JAK2,MPL, CALR, EZH2, ASXL1, IDH1/2 and SRSF2. The clinical variables assessed were those previously identified as being prognostically-relevant by IPSS or DIPSS-plus. The prognostic model (MIPSS) was developed through a stepwise selection process, based on a z-test of the regression coefficients, and its relative quality was measured by means of the Akaike information criterion. RESULTS. In the Italian learning cohort (n=588),median follow-up was 3.6years (95% CI, 0.03-30.8) for alive patients and 196 (33.3%) deaths and 67 leukemic transformations were documented. IPSS risk distribution was low in 26%, intermediate-1 in 30%, intermediate-2 in 24% and high in 20%. Mutational frequencies were approximately 63% for JAK2, 20% CALR, 6% MPL, 7% EZH2, 22% ASXL1, 2.5% IDH1/2 and 9% SRSF2. Constitutional symptoms were recorded in 28.6% of patients, splenomegaly 75% (18% >10cm from LCM) and grade 3 fibrosis 21%. Overall 252 patients were evaluable for karyotype and abnormalities were detected in 35%, including 9.5% unfavorable. Univariate analysis identified the following risk factors for inferior survival: age >60 years, constitutional symptoms, hemoglobin <100g/L, leukocytes >25x109/L, platelets <200x109/L, circulating blasts 1% or greater, splenomegaly >10cm, grade 3 fibrosis, unfavorable karyotype, triple-negativity for JAK2, MPL and CALR (TN), JAK2 or MPL mutation, and mutations in ASXL1, SRSF2, EZH2 or IDH1/2. In multivariable analysis, age >60 years, constitutional symptoms, hemoglobin <100g/L, platelets <200x109/L, TN, JAK2 or MPL mutation, ASXL1 and SRSF2 mutation remained significant; these variables were subsequently assigned the following HR-weighted adverse points: 1.5, 0.5, 0.5, 1.0, 1.5, 0.5, 0.5 and 0.5, respectively. Subsequently, four risk groups with no overlap in the 95% CI of their survival curves were delineated (Figure 1A): low (score 0-0.5); Intermediate-1 (score 1-1.5); Intermediate-2 (score 2-3.5); and high (score 4 or greater). The number of patients in each risk category, median survivals and HR and 95% CI values are included in figure 1A. Akaike information criterion indicated that MIPSS performed better than IPSS in predicting survival (1611.6 vs 1649.0). Leukemia-free survival was also predicted by high (p<0.01) and intermediate 1/2 (p<0.01) risk MIPSS. In the Mayo Clinic validation cohort (n=398), median age 63 years; males 64%,median follow-up for alive patients was 7.5 years and 271 (68%) deaths and 51 (13%) leukemic transformations were documented. DIPSS-plus risk distribution was low in 13%, intermediate-1 18%, intermediate-2 37% and high 32%. Mutational frequencies were approximately 51% for JAK2, 30% CALR, 7.5% MPL, 5% EZH2, 36% ASXL1, 5.6% IDH1/2 and 13% SRSF2. MIPSS performed equally well in this cohort, as illustrated in figure 1B, which includes number of patients in each risk category with their median survival and HR and 95% CI values. Leukemia-free survival in the Mayo cohort was also predicted by high (p<0.01) and intermediate-2 (p=0.04) risk MIPSS. Patients in the Mayo cohort were fully annotated for karyotype, which was MIPSS-independent in its prognostic value for both overall (p<0.01) and leukemia-free (p<0.01) survival. CONCLUSIONS. MIPSS addresses the mutation-relevant prognostic information gap in IPSS and performs equally well in the setting of patients seen at time of diagnosis or referral. Karyotype carries MIPSS-independent prognostic value that is recognized in an alternative GIPSS prognostic model (see accompanying ASH 2014 abstract), which complements MIPSS. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Abstract Abstract 851 Clinical studies have demonstrated that acute myeloid leukemia (AML) is heterogeneous with respect to presentation and to clinical outcome, and many studies have shown that cytogenetics can be used to improve prognostication. More recently, it has been shown that FLT3, NPM1, and CEBPA mutations can be used to further risk stratify AML patients. Although recent studies have identified novel, recurrent somatic mutations in AML in genes including TET2, ASXL1, IDH1, IDH2, and PHF6, their prognostic value has not yet been evaluated among patients treated on a large phase III clinical trial. Here we report full-length DNA resequencing of FLT3, NPM1, CEBPA, H/K/NRAS, KIT, WT1, TET2, ASXL1, IDH1/2, TP53, RUNX1, PTEN, and PHF6 in pre-treatment genomic DNA from 398 patients with de novo AML younger than 60 years of age enrolled in the ECOG E1900 protocol. We could identify a clonal alteration, defined as a somatic mutation or cytogenetic abnormality, in 91.2% of all patients; 42% had 1 somatic alteration, 36.4% had 2 alterations, 11.3% had 3 alterations and 1.5% had 4 or more somatic alterations. Mutational Analysis: We identified somatic mutations in FLT3 (37% total; 30% ITD, 7% TKD), NPM1 (14%), CEBPA (10%), TET2 (10%), NRAS (10%), WT1 (10%), KIT (9%), IDH2 (8%), IDH1 (6%), RUNX1 (6%), ASXL1 (4%), PHF6 (3%), KRAS (2.5%), TP53 (2%), and PTEN (1.5%) (Figure 1A). We then performed correlation analysis to determine the frequency of different mutational combinations in AML patients. In addition to previously identified mutational correlations (FLT3 and NPM1, KIT and core binding factor leukemia), we found that FLT3 and ASXL1 mutations were mutually exclusive in this large cohort (p =0.0008). In addition, we found that IDH1/IDH2 mutations were mutually exclusive of both TET2 (p=0.02), and WT1 (p=0.01) mutations consistent with overlapping roles in AML pathogenesis (Figure 1B). Integration of mutational data with response data and outcome: Mutations in ASXL1 (CR rate 33.3% versus 66.5% for ASXL1-WT patients; p= 0.009) were enriched in patients who did not respond to induction chemotherapy. We found that mutations in FLT3 (p=0.0005), ASXL1 (p=0.005; Figure 1C), and PHF6 (p=0.02) were associated with reduced OS. Notably, the adverse effects of ASXL1 and PHF6 mutations on outcome were restricted to patients treated with standard dose daunorubicin (p<0.01 for each mutational cohort); ASXL1 and PHF6 mutational status did not affect outcome in patients treated with high-dose daunorubicin, demonstrating that increased intensity induction chemotherapy can improve outcomes in genetically defined poor risk AML subsets. We also found that mutations in CEBPA (p= 0.04) and in IDH2 (p=0.003) were associated with improved OS; the favorable impact of IDH2 mutations on outcome was exclusive to patients with IDH2 R140Q mutations (Figure 1D). Within the subset of patients with cytogenetically defined intermediate risk AML, we found, as expected, that FLT3-ITD mutations were associated with reduced OS and CEBPA mutations were associated with improved OS. Importantly, we found that IDH2 R140Q mutations were associated with improved OS (p=0.003) in patients with intermediate risk AML regardless of FLT3 mutational status. In contrast, TET2 mutations (P=0.0002) and PHF6 (p<0.0001) were associated with reduced OS in intermediate risk, FLT3-WT, AML. Summary: These data provide important clinical implications of genetic alterations in AML and provides insight into the multistep pathogenesis of adult AML. We identified novel mutations that impact response to induction chemotherapy and predict overall outcome, including ASXL1, PHF6, TET2 and IDH2, suggesting these mutations can be used to improve AML risk stratification and therapeutic decisions in AML patients 18–60 years of age. Most importantly, we identify TET2, IDH1/2, and WT1 mutations as mutually exclusive somatic alterations that form a novel mutational complementation group that contributes to AML pathogenesis. Figure 1. A) Circos diagram depicting relative frequency and pairwise co-occurrence of mutations in de novo AML patients enrolled in the ECOG protocol E1900. B) Circos diagram showing mutational events in IDH1/2, TET2, and WT1 mutant AML patient. C) Overall survival in AML patients with and without ASXL1 mutations (p=0.005). D) Overall survival in AML patients with and without IDH2 R172 and R140 mutations (p=0.003). Figure 1. A) Circos diagram depicting relative frequency and pairwise co-occurrence of mutations in de novo AML patients enrolled in the ECOG protocol E1900. B) Circos diagram showing mutational events in IDH1/2, TET2, and WT1 mutant AML patient. C) Overall survival in AML patients with and without ASXL1 mutations (p=0.005). D) Overall survival in AML patients with and without IDH2 R172 and R140 mutations (p=0.003). Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 3803 Background: Ring sideroblasts (RS) represent abnormal mitochondrial iron deposits that are commonly present in myelodysplastic syndromes (MDS). The presence of ≥15% RS, by definition (World Health Organization-WHO) is necessary for a diagnosis of a MDS-RS; with refractory anemia with ring sideroblasts (RARS) being a specific morphologic category. RS can also be seen with other morphologic categories such as; refractory cytopenias with multilineage dysplasia (RCMD), MDS-unclassifiable (MDS-U) and refractory anemia with excess blasts (RAEB-1/2). Amongst these categories, RARS is generally believed to have the best survival rate with the lowest risk of leukemic transformation and this has been attributed to the absence of associated dysplasia in more than one lineage. However, it has not been systematically shown whether or not the exact percentage of RS provides additional prognostic information in the context of an accurate morphologic categorization and risk stratification by karyotype or the revised IPSS (IPSS-R). Methods: 200 patients with primary MDS without excess blasts and ≥1% RS were seen at the Mayo Clinic from 1997 through 2007. All patients underwent bone marrow (BM) examination and cytogenetic evaluation at diagnosis and the pathology slides, including iron stains, were centrally re-reviewed to accurately quantify bone marrow (BM) RS percentage and to confirm WHO morphologic categories. Molecular profiling included analysis for IDH1, IDH2, JAK2 and MPL mutations. For the purposes of this study patients were divided into 4 categories based on the RS percentage; <5%, 5–14%, 15–50% and >50%. In addition, each patient was assigned IPSS and IPSS-R prognostic scores at diagnosis and risk stratified by karyotype according to the IPSS-R cytogenetic risk categories. Cox proportional regression method was used for multivariable analysis. Results: Of the overall 200 study patients, 140 (70%) were male and median age was 71 years (range, 17–90 years). At presentation, 34 (17%) patients were red cell transfusion-dependent. There were 56 (28%) patients with <5% RS (RCMD-52, MDS-U-14), 32 (16%) with 5–14% RS (RCMD-29, MDS-U-3), 79 (39%) with 15–50% RS (RARS-43, RCMD-29, MDS-U-2) and 33 (16%) with >50% RS (RARS-13, RCMD-17, MDS-U-3). Three patients (2%) were IDH1 mutant (RARS-1, RCMD-2), 17 (8%) IDH2 mutant (RARS-1, RCMD-14, MDS-U-2), 5 JAK2 V617F (3%) mutant (RARS-2, RCMD-3) and one patient with RCMD had the MPL W515L mutation. At a median follow-up of 33 months, 156 (73%) deaths and 24 (12%) leukemic transformations were documented. Median survivals were 63 months for MDS with >50% RS, 43 months for MDS with 15–50% RS, 35 months for MDS with 5–14% RS, and 14 months for MDS with <5% RS (p=0.005). In univariate analysis, additional risk factors included decreased hemoglobin, decreased platelet count, increased circulating blasts, WHO morphologic categories, red cell transfusion need, IPSS, IPSS-R and cytogenetic risk categories per IPSS-R. On a multivariable analysis only the IPSS-R prognostic score (p<0.0001) and the WHO morphologic categories (p=0.02 for RARS) retained significance. Univariate analysis disclosed significantly inferior leukemia-free survival (LFS) in patients with RCMD, thrombocytopenia, increased circulating blasts, and poor IPSS and IPSS-R prognostic scores. On multivariable analysis once again only the IPSS-R prognostic scores (p=0.02) and WHO morphologic categories (p=0.02) retained their significance. In other words, RS percentage, independent of WHO classification, as a continuous or categorical variable (i.e.: <5%, 5–14%, 15–50% & >50%) did not affect either overall survival or LFS. Conclusions: In MDS without excess blasts, once an accurate WHO morphologic categorization is made based on the presence or absence of multi-lineage dysplasia, there is no additional prognostic value for quantifying bone marrow ring sideroblasts. WHO classification and IPSS-R prognostic scores remain the most important factors in assessing patients with MDS. Disclosures: No relevant conflicts of interest to declare.

Abstract Background: KIT D816V is considered the driver mutation in systemic mastocytosis (SM) however this mutation alone does not explain the diverse clinical manifestations of SM. While mutations in other myeloid-relevant genes have been reported in SM, their pattern(s) of co-segregation and prognostic value, independent of age and SM subtype, have not been fully defined in a mature dataset. Objectives: In this pilot study of advanced SM patients, we sequenced a panel of myeloid-relevant genes for known pathogenic mutations, to ascertain: (i) the frequency of individual mutations; (ii) the pattern(s) of mutation co-segregation; and (iii) the prognostic value of mutations, both individually and in combination, on overall- (OS) and leukemia-free survival (LFS). Methods: The current study was approved by the Mayo Clinic institutional review board. SM was classified as per World Health Organization criteria. DNA was isolated from archived bone marrow (BM) cell pellets. Next-generation deep sequencing by HiSeq SBS chemistry (Illumina) was performed to investigate the coding regions of 22 genes: KIT, JAK2, CALR, SRSF2, U2AF1, SF3B1, ASXL1, TET2, EZH2, SETBP1, DNMT3A, IDH-1/2, SUZ12, CBL, NRAS, FLT3, PTPN11, RUNX1, CEBPA, TP53 and NPM1. Results: We studied 19 patients with advanced SM; 6 had aggressive SM (ASM) and 13 had SM with an associated hematological disease (SM-AHD). Of the latter, 5 patients (38%) had myeloproliferative neoplasm-unclassified (MPN-u), 3 (23%) chronic myelomonocytic leukemia (CMML), 2 (15%) acute myeloid leukemia (AML), and 1 (8%) each polycythemia vera, myelodysplastic syndrome, and multiple myeloma. The median follow up from diagnosis was 20 months (range 2 to 84). Over this period, 5 patients (26%) developed leukemic transformation and 12 patients (63%) have died. KIT D816V was identified in 9 (47%) patients; 9 of 10 patients who tested negative by deep sequencing were screened by allele-specific PCR (sensitivity ~0.01%), with 8 additional patients testing positive. Thus, overall KIT D816V frequency was 89%. Fourteen (74%) patients harbored at least one additional non-KIT mutation; of these 7 patients had one additional mutation (ASXL1/TET2/SF3B1 = 2 each and JAK2 =1), 3 patients each had two and three additional mutations, and 1 patient had 4 additional mutations. The most frequently mutated genes (other than KIT) were ASXL1/TET2 (5 patients each, 26%), CBL (4 patients, 21%), SETBP1/SF3B1/DNMT3A (2 patients each, 11%) and JAK2/CALR/FLT3/IDH-1/RUNX1/TP53 (1 patient each, 5%). There was no overlap between patients harboring TET2 and ASXL1 mutations however 2 and 1 CBL-mutated patients concurrently harbored TET2 and ASXL1 mutations, respectively. SETBP1 and SF3B1 but not DNMT3A mutations occurred independently of TET2 and ASXL1 mutations. The distribution of mutations was similar between ASM and SM-AHD subgroups. When considered individually, the following non-KIT mutated genes (present in at least 2 patients) were associated with significantly inferior OS: ASXL1, CBL and SETBP1 (p<0.05). When the 3 mutated genes were considered together, only ASXL1 and CBL maintained their significance, independent of age and SM subtype, for inferior OS (Hazard ratio (95% confidence interval)=34.8 (3.1-393.9) and 59.7 (4.1-861.5), respectively). Patients harboring ASXL1 and/or CBL mutations (n=8, 8 deaths, median OS=11 months) had a significantly worse OS as compared to those without either mutation (n=11, 4 deaths, median OS=84 months) (p=0.0002) (Fig 1). When considering the number of non-KIT mutations (zero, one, two, three and four mutations in 5, 7, 3, 3 and 1 patients, respectively), the group with ≥3 mutations (n=4, 4 deaths, median OS=7 months) had a significantly inferior OS as compared to those with <3 mutations (n=15, 8 deaths, median OS=48 months) (p=0.01) (Fig 2). Presence of mutations was not correlated with LFS in this analysis. Conclusions: In this cohort of advanced SM patients, non-KIT mutations were frequent (prevalence 74%). TET2, ASXL1, SETBP1 and SF3B1, but not CBL or DNMT3A, mutations were mutually exclusive. Presence of ASXL1 and CBL mutations was associated with significantly inferior OS (HR=35 and 60, respectively), independent of known prognostic factors. An increasing number of non-KIT mutations (≥3, HR=5.9) was also associated with a significantly worse survival outcome. Sequencing results from a larger cohort will be presented at the meeting. Figure 1. Figure 1. Figure 2. Figure 2. Disclosures Pardanani: Stemline: Research Funding.

Abstract Abstract 3840 Background: Patients with myeloproliferative neoplasms (MPN) harbor a multitude of somatic mutations involving JAK2, MPL, LNK, CBL, TET2, ASXL1, IDH1, IDH2, EZH2, DNMT3A, IKZF1 and TP53. None of these mutations have been consistently traced back to the ancestral clone and most of them are currently believed to represent secondary genetic events. The higher prevalence of IDH, LNK, IKZF1 and TP53 mutations in blast-phase MPN, as opposed to chronic-phase disease, suggests their pathogenetic contribution to leukemic transformation (Harutyunyan A, et al. N Engl J Med. 2011;364:488). We examined this possibility for TP53 mutations or its SNPs in chronic-phase PMF; TP53 SNPs involving codon 72 have been associated with increased risk in the development of several cancers. Methods: Mayo Clinic MPN database was utilized to identify adult patients (>18 years) with chronic-phase PMF. Study eligibility criteria included availability of bone marrow histology and cytogenetic information at time of referral. In addition, all patients were annotated for IDH, MPL and JAK2 mutational status. Patients with blast-phase disease at the time of their referral were excluded from the study. DNA from bone marrow or peripheral blood was extracted using conventional methods and amplified with one set of primers covering TP53 exons 4 through 9. The amplification products were then sequenced with eight primers using the Big Dye Terminator sequencing kit v1.1 (Applied Biosystems, Foster City, CA) and the ABI 3130xl Genetic Analyzer. Sequences were examined for mutations by the Sequencher software (Gene Codes, Ann Arbor, MI). Any mutation found was confirmed with a second PCR and sequencing reaction. A mutation was considered to be any sequence other than genomic reference sequence NC_000017.10 or a well documented polymorphism such as the rs1042522 in exon 4. Mutations were compared with the IARC p53 Database (http://www-p53.iarc.fr/) and/or the TP53 Mutation Database (http://p53.free.fr/index.html) to determine if the mutation had been reported before. Results I: Baseline characteristics and clinical course A total of 107 patients with chronic-phase PMF were studied. Median age at time of referral was 64 years (range 32–81) and 71% were males. DIPPS-plus risk assignment was low in 13%, intermediate-1 in 15%, intermediate-2 in 38% and high in 34%. Karyotype was normal in 70 (65%) patients, favorable in 26 (24%) and unfavorable in 11 (10%). JAK2, MPL and IDH mutations were detected in 62 (58%), 7 (7%) and 6 (6%) cases, respectively. To date, 74% of patients had died and 15% had documented leukemic transformation. Results II: TP53 mutations/polymorphisms and their clinical relevance TP53 mutations were discovered in only 4 (4%) cases: i) E204E; GAG>GAA (silent exon 6); ii) G245D; GGC>GAC (exon 7); iii) R175H; CGC>CAC (exon 5); and iv) six base insert (GGCGAG) after bp13767 (exon 6). There was no significant difference in presenting features or survival between TP53 mutated and unmutated cases. None of the four patients with TP53 mutation incurred leukemic transformation. TP53 exon 4 SNPs data for codon 72 was available on 104 patients and included 58 (56%) patients with homozygous Arg72Arg, 34 (33%) patients with heterozygous Pro72Arg and 12 (11%) patients with homozygous Pro72Pro. Allele frequencies were 0.7 for G (corresponding to arginine) and 0.3 for C (corresponding to proline), which is similar to the general population (0.65 and 0.35 respectively). There were no significant differences among the three codon 72 genotypes in terms of presenting characteristics or survival. Conclusions: TP53 mutations are infrequent in chronic phase PMF. A higher number of informative cases are needed to accurately define their phenotypic or prognostic effects. We are currently analyzing more cases in this regard. On the other hand, TP53 polymorphisms at codon 72 do not appear to influence phenotype or prognosis in chronic-phase PMF. Disclosures: No relevant conflicts of interest to declare.

Abstract Background : TET2, located on chromosome 4q24, is frequently mutated in the majority of patients with chronic myelomonocytic leukemia (CMML). TET2 has 11 exons, and variations, especially in exon 3 have been described as a part of age related clonal hematopoiesis (Jaiswal NEJM 2015). In CMML, somatic TET2 mutations in the absence of ASXL1 mutations (ASXL1wt/TET2mt) were previously shown to impart a favorable outcome on survival (Patnaik BCJ 2015). In the current larger CMML patient cohort, we describe the number and type of TET2 mutations and examine their phenotypic and prognostic effects. Methods : 261 patients with CMML were included in the study. All patients had bone marrow (BM) biopsies and cytogenetics performed at diagnosis. Targeted capture assays were carried out on BM DNA specimens obtained at diagnosis for the following genes; TET2, DNMT3A, IDH1, IDH2, ASXL1, EZH2, SUZ12, SRSF2, SF3B1, ZRSR2, U2AF1, PTPN11, Tp53, SH2B3, RUNX1, CBL, NRAS, KRAS, JAK2, CSF3R, FLT3, KIT, CALR, MPL, NPM1, CEBPA, IKZF, and SETBP1. TET2 (NM_001127208.2) coverage extended from exons 3-11, with frame shift, non-sense, and missense variations considered pathogenic. Previously annotated single nucleotide polymorphisms (http//www.hapmap.org) were considered non-pathogenic. Results: Among the 261 study patients, 65% were males and median age was 70 years (range, 28-91). 154 (59%), 64 (25%) and 43 (16%) patients were classified as CMML-0, 1 and 2, respectively. At a median follow-up of 23 months, 174 (67%) deaths and 37 (14%) leukemic transformations were documented. Mutational frequencies ≥4% were encountered in; ASXL1 45%, TET2 42%, SRSF2 40%, NRAS 14%, SETBP1 13%, CBL 10%, JAK2 7%, RUNX1 6%, DNMT3A 6%, U2AF1 6%, SF3B1 5%, ZRSR2 4%, Tp53 4% and IDH2 4%. i) TET2 mutations type, number and phenotypic correlates: TET2 mutations were seen in 109 (42%) patients; these included frameshift 33 (30%), nonsense 29 (27%) and missense 11 (10%) variants whereas 36 (33%) had more than one type of mutation. Overall, 57 (52%) patients had more than 1 TET2 mutations: 52 (48%) patients had 1, 46 (42%) 2 and 11 (10%) ≥3 TET2 mutations. Among all 109 TET2 mutated patients, 65% were male, and median age was 71 years with no significant difference in age and gender distribution between mutated and un-mutated cases, or type of TET2 mutations; however, older patients were more likely to carry multiple TET2 mutations (p=0.01). Compared to their un-mutated counterparts, TET2 mutated cases were less likely to have a low hemoglobin (p<0.001), circulating immature myeloid cells (IMC) (p=0.001), peripheral blood (PB) (p=0.009) and BM blasts (p=0.009), and have high-risk stratification per Mayo Molecular Model (p<0.001); these differences were not affected by the type or number of TET2 mutations. TET2 mutated cases were more likely to have a higher frequency of SRSF2 (p=0.004) and lower frequency of ASXL1 (p=0.03), Tp53 (p=0.04) and IDH1/2 mutations (p<0.001); these associations were also not affected by the type or number of TET2 mutations. ii) Impact on survival: Median survival for the entire cohort (n=261) was 24 months. In univariate analysis, survival was superior in TET2 mutated (median 33 months) versus wild-type (median 21 months) patients [p=0.03; HR 1.3 95% CI 1.12-1.86] (figure 1A). This survival difference remained significant after adjustment for age (p=0.04), leukocyte count (p=0.017), absolute monocyte count (p=0.02), absolute lymphocyte count (p=0.02), platelet count (p=0.015), circulating IMC (p=0.03), DNMT3A (p=0.02) and ASXL1 (p=0.045) mutations; however, significance was lost after adjustment for abnormal karyotype (p=0.32) and the Mayo Molecular Model (p=0.0003). These observations were not affected by the type or number of TET2 mutations. Finally, our previous observation regarding the survival advantage of ASXL1wt/TET2mt versus other genotypes was most apparent for patients with multiple TET2 mutations (p=0.02) (figure 1B). Conclusions: TET2 mutations are common in CMML and constitute approximately equal proportions of frameshift and non-sense mutations, while missense mutations are less frequent. Majority of TET2 mutated CMML cases harbor more than one mutant variant. Regardless, the relevance of type and number of TET2 mutations in CMML was limited to an association between older age and number of mutations and the latter with possibly improved survival in the absence of clonal ASXL1 mutations. Disclosures No relevant conflicts of interest to declare.

Abstract Background U2AF1 mutations occur in approximately 16% of patients with primary myelofibrosis (PMF) and were significantly associated with anemia, thrombocytopenia, older age, JAK2V617F, ASXL1 mutations and normal karyotype (Leukemia 2014;28:431); furthermore, U2AF1 mutations were associated with inferior survival in univariate but not multivariable analysis. In the current study, we looked for the possibility of a differential effect from U2AF1 mutation variants on these observations in PMF. Methods Study patients fulfilled the 2016 WHO criteria for the diagnosis of PMF (Blood. 2016;127:2391). Additional selection criteria included the availability of U2AF1 mutational status. Previously published methods (Leukemia 2014;28:431), including targeted next generation sequencing (Blood 2015 126:354), were used to screen for U2AF1 and other prognostically-relevant mutations. Statistical analyses considered clinical and laboratory parameters obtained at time of initial referral to the Mayo Clinic. Results Patient characteristics: U2AF1 mutational status was available for 457 patients and 72 (16%) harbored one of several mutation variants: these mutations affected residue Q157 in 44 (10%) patients, S34 in 24 (5%) and the remaining 1% displayed other variants. The 44 patients with U2AF1Q157 mutations included 23 with Q157P, 18 with Q157R and 3 with Q157-Y158insYE. The 24 patients with U2AF1S34 included 16 with S34F and 8 with S34Y. Only one patient harbored both Q157 and S34 mutations (Q157R, S34Y). Among all 457 study patients, median age was 63 years, 64% were males, dynamic international prognostic scoring system (DIPSS)-plus (JCO 2011;29:392) risk distributions were 13% low, 18% intermediate-1, 38% intermediate-2 and 32% high and driver mutation distributions were 54% JAK2, 22% CALR type 1/ type1-like, 4% CALR type 2/type 2-like, 7% MPL and 13% triple-negative. Cytogenetic studies were available in 449 patients: 39% abnormal and 10% unfavorable. All 457 patients were screened for ASXL1 mutations with 37% mutated, 450 for SRSF2 mutations with 15% mutated, 403 for SF3B1 with 8% mutated, 366 for IDH1/2 with 5% mutated and 369 for EZH2 with 4% mutated. Median follow-up was 4.4 years and during this period, 318 (70%) deaths and 53 (12%) leukemic transformations were documented. Phenotypic correlates of U2AF1 mutation variants: Because of the relatively small number of informative cases with specific mutations, we classified all patients into Q157 (n=44) and S34 (n=24) mutation variants and compared them with the 385 U2AF1 un-mutated cases. First, we confirmed our previous observations regarding the association between U2AF1 mutations in general and older age (p=0.0003), JAK2V617F (p<0.0001), ASXL1 mutations (p=0.0002), normal karyotype (p=0.03), hemoglobin <10 g/dL (p<0.0001) and platelet count <100 x 10(9)/L (p<0.0001); when the two U2AF1 mutation categories were analyzed separately, the corresponding p values for Q157 were 0.0005, 0.001, <0.0001, 0.04, <0.0001 and <0.0001 and for S34 were 0.12, 0.001, 0.41, 0.41, <0.0001 and 0.67. Phenotypic correlates of U2AF1 mutation variants: In univariate analysis, survival was adversely affected by U2AF1Q157 (p<0.0001; HR 2.2, 95% CI 1.6-3.1) but not by U2AF1S34 (p=0.8; HR 1.1, 95% CI 0.6-1.8) mutations (Figure 1a). Furthermore, the negative survival effect of U2AF1Q157 mutations was independent of anemia, thrombocytopenia, ASXL1, SRSF2, IDH1/2 and EZH2 mutations, as well as driver mutational status; multivariable analysis that included all molecular alterations identified U2AF1Q157 (HR 1.6, 95% CI 1.1-2.3), ASXL1 (HR 2.3, 95% CI 1.8-3.5), SRSF2 (HR 1.6, 95% CI 1.2-2.2) and absence of CALR type-1/like (HR 2.6, 95% CI 1.8-3.5) mutations as independent risk factors for survival. Finally, the survival effect of U2AF1Q157 mutations was independent of DIPSS-plus in patients with hemoglobin ≥10 g/dL (HR 2.6, 95% CI 1.3-5.3; p=0.007) (Figure 1c) but not in those with hemoglobin <10 g/dL (p=0.98) (Figure 1b). Conclusion Both U2AF1Q157 and U2AF1S34 mutation variants in PMF are associated with JAK2V617F and severe anemia whereas only the former is associated with ASXL1 mutations and thrombocytopenia. More importantly, U2AF1Q157, but not U2AF1S34, mutation variants in PMF are predictive of inferior survival, independent of other adverse mutations, and, in the absence of severe anemia, independent of DIPSS-plus. Disclosures No relevant conflicts of interest to declare.

Abstract Background:Recurrent mutations in ATM, JAK1/3 and STAT5B genes have been identified in T-cell prolymphocytic leukemia (T-PLL), a mature T cell leukemia with poor survival. However, much remains to be understood regarding the dysregulated pathways involved in the pathogenesis of this aggressive disease. Here, we integrated the analyses of whole exome sequencing (WES) and transcriptomes of a cohort of 12 T-PLL samples to gain further insights into molecular mechanisms of this malignancy. Methods: Leukemic cells were isolated from peripheral blood or bone marrow samples of T-PLL patients. Tumor DNA and RNA, and germline DNA were extracted from purified leukemic T cells and bone marrow derived stromal cells, respectively. WES (n=12) and transcriptome analysis (n=10) were performed and data analyzed with published standard methods (Robinson et al. 2010 and DePristo et al. 2011) Results: 7 paired tumor and germline DNA and 5 tumor-only DNA samples were used for WES analysis. Mutations in ATM and TET2 were detected in 4 (33%) and 1 (8%) patients, respectively. JAK3 (p.M511I, p.Q501H and p.Q503H) or STAT5B (2 p.T628S and 1 p.N642H) mutations each occurred in 3 (25%) cases and were mutually exclusive, consistent with prior reports. Importantly, we found several novel mutations in genes with regulatory roles in the NOTCH signaling pathway. Specifically, 2 somatic mutations p.A3V and p.A3S in NOTCH2 and 2 other mutations (p.S244L and p.A1414V) in NOTCH4 were identified in 4 (33%) patients. One mutation was in the EGF- like and another one in the NODP domain critical for NOTCH-mediated cellular differentiation. Two (16.7%) novel missense mutations (p.G174D and p.V836I) were found in MAML2. Three missense mutations (2 p.S583N and 1 p.R190H) occurred in MAML1. Both MAML2 and MAML1 are essential cofactors of the NOTCH signaling pathway. In addition, a somatic mutation in the NOTCH signal regulator DTX2 (p.G268D) was detected. Altogether 7 (58.3%) patients harbored mutations in genes with a regulatory role in the NOTCH pathway. Among the epigenetic modifiers, we found deletions and point mutations in KDM6B and KDM6A in 4 (30%) patients including 2 codon deletions in KDM6B (p.X751Val) and 1 missense mutation each in KDM6B (p.R1016Q) and KDM6A (p.V510G). KDM6A and KDM6B are histone H3 lysine 27 demethylases known to be involved in regulation of DNA repair and play an important role in embryonic differentiation. Transcriptome analyses of 10 T-PLL cases versus normal CD3+ T cells from 5 healthy donors revealed 706 up-regulated and 655 down-regulated genes. Consistent with prior data, TCL1A and TCL1B were significantly overexpressed (23.7 and 13.1 fold, respectively). Using Ingenuity Pathway Analysis, we found PTEN Signaling, glutamate receptor, and axonal guidance signaling to be among the top 3 upregulated pathways. The top upstream regulators implicated for the up-regulated signals were FGF, Wnt/β-catenin and TNF. The top 3 down-regulated pathways were granulocyte adhesion and diapedesis, IL-17A in psoriasis and IL-10 signaling. The top regulators for the down-regulated pathways were LPS, TNF and IFNg. Wnt/β-catenin and Human embryonic stem cell pluripotency signaling were among the top 20 up-regulated pathways. Many genes involved in the regulation of Wnt signal pathway were included in the above pathways, including Wnt ligands (Wnt9A, Wnt4), transmembrane receptors (CDH1, LRP5, FZD6, FZD7, SMO), cytoplasmic factor (AXIN2), and nuclear factors (MYC, SOX12, SOX8, SOX18, CCND1). Summary: Our study identified recurrent mutations in the NOTCH signaling pathway as well as dysregulated embryonic developmental pathway, e.g. Wnt/NOTCH as newly implicated molecular mechanisms in T-PLL pathogenesis for further functional studies. Additionally, novel (KDM6A and KDM6B) and known mutations (ATM, TET2) in chromatin regulatory genes underscore the importance of epigenetic dysregulation in the tumorigenesis of T-PLL. Disclosures Parikh: Pharmacyclics: Honoraria, Research Funding. Mustjoki:Ariad: Research Funding; Bristol-Myers Squibb: Honoraria, Research Funding; Pfizer: Honoraria, Research Funding; Novartis: Honoraria, Research Funding. Ding:Merck: Research Funding.

Abstract Abstract 3619 Extramedullary (EM) acute myeloid leukemia (AML) is generally associated with a poor prognosis. We evaluated the baseline characteristics, response and survival characteristics in a subset of patients age ≤ 60 years with de novo AML and EM manifestations at presentation, treated on the Eastern Cooperative Oncology Group E1900 protocol: daunorubicin 45 or 90 mg /m2 x3 days in conjunction with cytarabine 100 mg/m2 × 7 days followed by risk-stratified consolidation. (Fernandez et al New Engl J Med 2009) Methods: Patient baseline characteristics were compared using Fisher's exact test if they were categories and Wilcoxon rank sum tests if they were continuous. Logistic regression models were used for complete remission (CR) rates and Cox regression models were used for overall survival (OS) and disease-free survival (DFS). The univariate models included indication of whether patient had extramedullary disease involvement, adjusted by induction treatment. The multivariate models were further adjusted for age, gender, cytogenetic risk, ECOG PS, WBC, platelets, marrow blast, hemoglobin, peripheral blasts, and secondary AML. All P values were based on 2-sided tests. Results: 58 (8.9%) of 651 patients in the study presented with evidence of EM at diagnosis. Patients with EM disease were younger and presented with higher peripheral WBC counts and circulating blasts than those without. Immunophenotyping did not demonstrate a difference in baseline CD56 expression (p=0.77). Sites of EM involvement included gingival hypertrophy (n= 33), skin (n= 7), central nervous system (n= 6), lymphadenopathy (n=6), mediastinum (n=2) and other (n=6). 3 patients had multiple sites. Cytogenetic risk in the EM subgroup was favorable (8.6%), indeterminate (inadequate/missing) (34.5%), intermediate (37.9%) and unfavorable (19%) and did not differ statistically from those without EM disease. The complete remission (CR) rates were similar for patients with EM disease compared to those patient's without, regardless of the anthracycline dose (univariate model p= 0.93; multivariate model p=0.88. DFS (patients in CR) and OS of patients with EM disease was not significantly different in the univariate (p= 0.99, p=0.49) or multivariate models (p= 0.92, p=0.37), respectively. On multivariate modeling patients with EM AML who received the standard-dose of daunorubicin had a shorter OS (HR 1.86, 95% CI 1.20 to 2.91, p=0.006). OS was not improved in EM patients treated with high-dose daunorubicin (HR 0.62, 95% CI 0.34 to 1.14, p=0.12) (Table). There was a significant difference in OS for patients with EM treated with the 90 mg/m2 dose of daunorubicin compared to patients with EM treated with the 45 mg/m2 dose (multivariate model: HR 0.30, 95% CI 0.12 to 0.76, p=0.01). Conclusion: Younger patients with de novo EM AML are sensitive to higher doses of anthracycline in induction therapy and should be treated in a similar fashion as patients without EM involvement. For both groups the higher doses of anthracycline are recommended as CR rates, DFS and OS are improved. Further evaluation of immunophenotypic and molecular/ genetic characteristics should be undertaken in this subset of patients. Disclosures: No relevant conflicts of interest to declare.