Pardanani, Animesh

Abstract Background: CD123 is the α-subunit of the high-affinity receptor for interleukin-3; its expression in normal bone marrow (BM) is highly restricted, however CD123 is highly expressed in certain hematolymphoid malignancies. In a pilot study of 11 systemic mastocytosis (SM) patients (Lasho et al., Leukemia, 2015), we made the following novel observations (i) CD123 was aberrantly expressed on neoplastic mast cells (MC); and (ii) there was focal proliferation of plasmacytoid dendritic cells (PDC) around MC aggregates, suggesting a tumor promoting role for PDC. Objectives: To study MC CD123 expression by IHC in a larger cohort of SM patients to ascertain: (i) whether there is a significant difference in CD123 expression between SM subtypes; (ii) whether focal PDC proliferation around MC aggregates is correlated with CD123 expression; and (iii) whether CD123 expression has prognostic value for overall survival (OS). Methods: The current study was approved by the Mayo Clinic institutional review board. SM was classified as per standard criteria. Archived BM slides were stained for tryptase and CD123 as previously described (Lasho et al., Leukemia, 2015). Slides were reviewed independently by 2 hematopathologists who were blinded with regards to SM subgroups and scored on the following pre-defined parameters: (i) CD123 expression yes/no; (ii) CD123 expression strong/weak; (iii) focal PDC proliferation yes/no; and (iv) BM MC%. Results: We studied 58 SM patients and 3 controls. Of the former, 23 had indolent SM (ISM), 10 aggressive SM (ASM), 23 SM with associated hematological disease (SM-AHD), and 2 had mast cell leukemia (MCL). Of the SM-AHD patients, 10 (43%) had CMML, 5 (22%) myeloproliferative neoplasm-unclassified (MPN-u), 3 (13%) PV, 2 (9%) acute myeloid leukemia, and 1 each had lymphoma, myeloma and myelodysplastic syndrome. The median follow up from diagnosis was 32 months (range 0.1 to 162). Over this period, 3 patients (5%) had leukemic transformation and 30 patients (52%) died. Concordance rates between hematopathologists for CD123 positive/negative, CD123 strong/weak, focal PDC proliferation yes/no and MC% were 97%, 94%, 98% and 100%, respectively. MC CD123 expression was absent in control cases but demonstrable in 37 (64%) SM patients; expression rates were 100%, 61%, 57%, and 0% in ASM, ISM, SM-AHD and MCL, respectively (p=0.02). Similar to MCL, both cases with SM-AML showed absent MC CD123 expression. Strong MC CD123 expression (comparable to background interstitial PDC) was observed in 23 patients (62%), including 85% with SM-AHD, 60% with ASM, and 43% with ISM (p=0.08). Focal PDC proliferation around MC aggregates, regardless of CD123 expression status, was noted in 44 patients (76%), including 90% with ASM, 87% with ISM, 65% with SM-AHD and 0% with MCL (p=0.02). Focal PDC proliferation was significantly higher in CD123-positive versus -negative patients (87% versus 50%, p=0.005). OS was predicted by SM subtype; the median OS for ISM, ASM, SM-AHD and MCL patients was not reached, 30 months, 25 months and 3 months, respectively (p=0.005). CD123 expression status did not have prognostic value for OS overall, however, in SM-AHD patients, CD123 expression was associated with significantly inferior survival (p=0.04), independent of the associated disease (ie, CMML vs AML vs MPN vs other). Survival outcome of SM-CMML patients expressing MC CD123 (n=3, median OS=11 months) was significantly inferior as compared to patients who did not (n=7, median OS=44 months), independent of age and absolute monocyte count (p=0.01). CD123 staining intensity had borderline significance for OS in the overall cohort (p=0.08). When restricted to non-indolent SM cases, strong CD123 expression was associated with significantly worse overall survival (15 versus 88 months, p=0.04). Conclusions: In this cohort, aberrant MC CD123 expression was observed in the majority of SM patients with highest prevalence in ASM patients and uniformly absent expression in leukemic phase SM (MCL or SM-AML). Focal PDC proliferation around MC aggregates was significantly higher in CD123-positive cases, indicating a potential functional interaction. CD123 expression identified SM-AHD patients with inferior OS, especially those with SM-CMML. Strong CD123 staining intensity may be associated with inferior OS in non-indolent SM patients. CD123 expression on MC and PDC constitutes a novel therapeutic target in non-leukemic SM. Disclosures Pardanani: Stemline: Research Funding. Brooks:Stemline Therapeutics, Inc.: Employment, Equity Ownership, Patents & Royalties.

Abstract We recently reported that the induction of polyploidization of malignant megakaryocytes shows great promise as a new therapy for acute leukemia. Polyploidization inducers such as dimethylfasudil (diMF) and MLN8237, both of which target Aurora A kinase (AURKA), induce proliferation arrest, polyploidization, expression of megakaryocyte differentiation markers and apoptosis of leukemic megakaryocytes in vitro and in vivo. Since megakaryocytes in primary myelofibrosis (PMF) show impaired polyploidization and maturation, and likely directly contribute to the disease, we predicted that polyploidization inducers would provide a new therapeutic strategy. To determine the effect of these compounds on the growth of MPN cells, we first treated the JAK2 V617F mutant megakaryocytic SET2 cell line with varying doses of MLN8237 and diMF. Both compounds effectively and dose dependently inhibited proliferation, induced polyploidization and upregulation of lineage specific markers CD41 and CD42, and increased apoptosis. Furthermore, MLN8237 synergized with ruxolitinib to induce apoptosis of the SET2 cells and also potently induced growth arrest of JAK2 inhibitor persistent SET2 cells. We observed a similar polyploidization and differentiating activity of MLN8237 and diMF on megakaryocytes derived from primary human PMF progenitors. The ability of these agents to induce polyploidization was specific, as the non-megakaryocyte fractions of the cultures were not affected. Next, we assayed the activity of polyploidization inducers on progression of MPNs in two mouse models: JAK2V617F conditional knockin mice and mice engrafted with MPLW515L expressing bone marrow progenitors. Of note, spleens from both mouse models displayed a robust increase in both total and phosphorylated forms of AURKA relative to control animals, further suggesting that AURKA is a rational target in this disease. We first assayed the activities of MLN8237 and diMF in the MPLW515L bone marrow transplantation model. Recipient mice develop a rapid MPN characterized by leukocytosis, thrombocytosis and bone marrow fibrosis. Both MLN8237 and diMF reduced the disease burden, as evidenced by significant reductions in the liver and spleen weights, white cell counts and platelet counts. Both compounds also led to a significant decrease of fibrosis in the bone marrow, diminished infiltration of megakaryocytes and granulocytes in the liver, and a profound reduction in the numbers of megakaryocytes within the spleen. Moreover, plasma levels of TGF-β a known myelofibrogenic cytokine, were decreased by more than 3-fold by the drug treatment. Both diMF and MLN8237 led to selective polyploidization of megakaryocytes in the spleen as well as marked reductions in the levels of p-AURKA. Of note, neither agent affected the extent of phosphorylation of STAT3 or STAT5. Therefore, we tested whether the combined use of a JAK inhibitor and a polyploidy inducer would show enhanced activity in vivo. Indeed, the combination of MLN8237 and ruxolitinib led to greater reductions in tumor burden in the MPLW515L mouse model than either agent alone. Similar results were obtained using the JAK2V617F knock-in model. To further validate our conclusion that AURKA is a target in PMF, we infected Aurkafl/fl floxed bone marrow progenitors with MPLW515L and transplanted the cells to irradiated recipients. Excision of both alleles of Aurka by Cre mediated recombination completely resolved the disease, while heterozygous deletion of Aurka significantly reduced the disease burden, in a manner similar to treatment with MLN8237. Given that heterozygous deletion of Aurka does not alter normal hematopoiesis in mice, the fact that a 50% reduction in kinase expression was associated with a significant decrease in disease burden suggests that there is an effective therapeutic window in which AURKA inhibitors will be effective against MPN while sparing normal hematopoiesis. Although JAK inhibitors provide symptomatic relief, it is becoming clear that they are not curative. Thus, there is an urgent need to develop new agents to use in combination with JAK inhibitors. Our data reveal that inducing polyploidization and differentiation of dysplastic megakaryocytes in PMF ameliorates features of the disease both in vitro and in vivo. These results support the initiation of clinical studies that combine a JAK inhibitor with an AURKA inhibitor. Disclosures: Crispino: Sanofi: Research Funding.

Abstract Abstract 3069 Background: Monosomal karyotype (MK) is defined as the presence of two or more distinct autosomal chromosome monosomies or a single autosomal monosomy associated with at least one structural abnormality (Breems DA et al. J Clin Oncol 2008; 26: 4791). In acute myeloid leukemia (AML), MK has been shown to be prognostically worse than complex or other unfavorable karyotype (Breems DA et al. J Clin Oncol 2008; 26: 4791). In primary myelofibrosis (PMF), complex karyotype or isolated trisomy 8 predicts inferior survival (Hussein K et al. Blood 2010; 115: 496). Objective: To determine if MK in PMF is prognostically distinct from previously defined poor cytogenetic risk categories including complex karyotype and isolated trisomy 8. Methods: The Mayo Clinic database for PMF was used to identify consecutive patients with unfavorable karyotype including complex karyotype and sole trisomy 8. WHO criteria were used for PMF diagnosis and leukemic transformation (Vardiman JW et al. Blood 2009; 114: 937). Results: Among 793 PMF patients with cytogenetic information at the time of their first time referral to the Mayo Clinic, 452 displayed a normal karyotype and 341 (43%) an abnormal karyotype. Of the latter, 41 (12%) displayed complex karyotype and 21 (6%) sole trisomy 8. Among the 41 patients with complex karyotype, 17 (42%) met the criteria for MK and 24 (58%) displayed complex karyotype without monosomies. Overall survival was significantly inferior in patients with MK compared to those with either complex karyotype without monosomies (p=0.02; HR 2.3, 95% CI 1.1–4.8) or trisomy 8 (p=0.02; HR 2.4, 95% CI 1.2–5.1) (Fig. 1). Prognosis among all three groups was significantly worse than patients with normal karyotype (Fig. 1). Leukemia-free survival was also significantly inferior in patients with MK compared to those with either complex karyotype without monosomies (p=0.02; HR 6.9, 95% CI 1.3–37.3) or trisomy 8 (p=0.02; HR 14.8, 95% CI 1.7–130.8) (Fig. 2). LFS in patients with normal karyotype was similar to those with either complex karyotype without monosomies (p=0.31) or trisomy 8 (p=0.86) (Figure 2). Conclusions: Monosomal karyotype in PMF is distinctly associated with extremely poor overall and leukemia-free survivals that are significantly worse than those seen in PMF patients with other unfavorable karyotype including complex karyotype without monosomies and sole trisomy 8 abnormalities. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 3068 Background: Abnormal cytokine expression accompanies the disease process in primary myelofibrosis (PMF) and the levels of proinflammatory cytokines have been shown to be down-regulated by JAK inhibitor therapy. The prognostic relevance of increased cytokine levels in PMF has not been systematically studied. The objectives of the current study were to describe the spectrum of abnormal cytokine expression in PMF and determine its prognostic relevance. Methods: A multiplex biometric sandwich immunoassay (Invitrogen multiplex assays, Invitrogen Corporation-Carlsbad CA, USA) was used for simultaneous analysis of 30 cytokine levels. Diluted plasma was mixed with fluorescently addressed microbeads bound with anti-cytokine antibodies. Biotin anticytokine secondary antibodies were then added and allowed to bind to cytokine-bead complexes. Total surface fluorescence for each bead was then measured using a Luminex (Luminex Corporation, Austin, TX, USA) fluorescent flow-based detection system. Results: Cytokine levels at diagnosis were analyzed for a total of 127 treatment naïve patients with PMF, who had stored sera at the time of diagnosis or first referral to the Mayo Clinic. The circulating levels of 30 cytokines in the multiplex assay were compared between PMF patients (n=127) and normal controls (n=31). Significant differences (p<0.05) were evident for 20 of the 30 cytokines: IL-1b, IL-1RA, IL-2R,IL-6, IL-8, IL-10, IL-12, IL-13, IL- 15, TNF-α, G-CSF, INF-α, INF-γ, MIP-1α, MIP-1β, HGF, CXCL10, MIG, MCP-1 and VEGF. On univariate analysis, increased levels of 9 of these 20 cytokines were associated with shortened survival (p<0.05): IL-1RA, IL-2R, IL-6, IL-8, IL-12, IL-15, MIP-1α, CXCL10 and MIG. On multivariable analysis, 4 of these 9 cytokines maintained their prognostic value for survival: IL-2R (p=0.0011), IL-8 (p<0.0001), IL-15 (p=0.0124) and CXCL10 (p=0.0031). The independent prognostic value of these 4 cytokines was sustained during additional multivariable analysis that included the Dynamic International Prognostic System Score and cytogenetic risk category. Next, we divided the 127 patients into three cytokine risk categories based on the serum levels of IL-2R, IL-8, IL-15 and CXCL10: none elevated, one or two elevated and three or four elevated. The respective median survivals were 69, 37 and 19 months (p=0.01; Figure 1) and such cytokine risk categorization was both DIPSS and cytogenetic risk independent; p values for multivariable analysis including cytokine, DIPSS, cytogenetic risk groups were 0.02, 0.0001 and 0.004, respectively. Conclusions: Elevated serum levels of IL-2R, IL-8, IL- 15 and CXCL10 independently predict shortened survival in PMF and such prognostication appears to be independent of DIPSS and cytogenetic risk category. The current study suggests clinical relevance of multiple cytokines in PMF that could be therapeutically targeted. Disclosures: No relevant conflicts of interest to declare.

Abstract Background Cytogenetic abnormalities have significant prognostic role in AML at the time of diagnosis. Several prognostic groupings have been recommended including the NCCN and European LeukemiaNet classification. It is still unknown whether cytogenetic abnormalities have any prognostic role at the time of relapse for predicting clinical outcome. In addition, clonal evolution (CE) has been seen in many AML cases, but its prognostic value remains to be identified. Aim To study cytogenetic abnormalities and clonal evolution impact on clinical outcome for patients diagnosed with rAML Method A retrospective, single institution study through chart review of all cases diagnosed with rAML at Mayo Clinic Rochester between 2003 to 2011 was performed. Pts who achieved complete remission and relapsed at our institution were included. Pts with PML-RARA and BCR-ABL rearrangements were excluded from this study. Pts with available cytogenetics data, both at presentation and relapse, were included. Response to treatment was defined as per international working group (IWG) criteria (Cheson JCO 2003). Appropriate IRB approval was obtained in accordance with Helsinki declaration. Comparison between groups’ medians was done using Wilcoxon test, while survival estimates were calculated using Kaplan-Meier curves using JMP V9. Results Out of 192 pts, 74 pts matched our inclusion criteria. The median age at diagnosis is 60 years (range 18-81), 53% were males. At relapse, median white blood cell (WBC) was 3.2 x109, hemoglobin 10.4 g/dL, platelet 59 x109, peripheral blood (PB) blast 2% (range, 0-96), bone marrow (BM) blasts 23% (range, 1-95) and cellularity 60%. Relapse site was mainly in the BM (95%). Flt3 mutation was found in 12/43 (28%), while NPM1 mutation in 2/7 (29%). Upon treatment with salvage therapy, 27 (36%) pts achieved complete remission (CR) and 16 (22%) pts partial remission (PR) for a total overall response (OR) of 58%. Twenty one (28%) pts were able to proceed to allogeneic stem cell transplant (SCT) after achieving remission. Median overall survival (mOS) was 198 days. Pts had a longer mOS if they proceeded to SCT (386 days vs 121 days, p =0.0003). Cytogenetic analysis (CG) were diploid in 32 (43%) of 74 pts at relapse. Five pts had favorable cytogenetics (3 pts with t(8,21) and 2 pts with inv (16)). CG was divided into favorable, intermediate and poor prognosis as per NCCN classification in 5 (7%), 45 (61%), 24 (32%), respectively. As per European LeukemiaNet, there were 4 CG groups divided into: favorable, Intermediate-1, intermediate-2, and adverse in 5 (7%), 33 (45%), 12 (16%), 24 (32%), respectively. Upon comparison of CG at diagnosis and relapse; clonal evolution (a change in CG due to gain or loss of chromosomal abnormalities) was found in 42/74 (57%) of pts. CG reverted back to diploid in 8/42 (19%) or gained additional CG abnormalities at relapse in 32 (76%) pts. CR2 was achieved in 80% of favorable CG (NCCN grouping), 40% (intermediate), and 21% in poor risk (p=0.01). Using NCCN CG grouping mOS was: 435 days (favorable), 266 (intermediate), and 104 (poor) (p value = 0.01). However, using European LeukemiaNet grouping mOS was: 435 days (favorable), 174 (intermeidtae-1), 363 (intermediate-2), and 104 (poor) (p=0.004). CE did not affect mOS (p =0.7). On multivariate analysis, response to salvage therapy (p<0.0001), NCCN CG grouping (p =0.03), wbc (p=0.045), and PB blasts (p=0.01) were significant but not age, BM blast, platelet and CE. Conclusion CG were diploid in 43% of relapsed pts. CE (gain or loss of cytogenetic abnormalities) was seen in 57% of CG at relapse compared to diagnosis but did not affect mOS. NCCN (and not European LeukemiaNet) cytogenetic grouping did affect clinical outcome through both achieving second CR and mOS (both on univariate and multivariate analysis). Our results confirm the prognostic value of CG on relapse, and need to be confirmed by larger studies. Disclosures: No relevant conflicts of interest to declare.

Abstract Background: Peripheral blood (PB) JAK2V617F testing is ordered in routine clinical practice in several different clinical scenarios. Interpretation of the mutation level/allele burden information is however challenging given lack of a standardized assay and lack of adequate information to permit data interpretation in a non-research setting. Objectives: To describe: (i) disease distribution in unselected JAK2V617F-positive patients seen in routine clinical practice: (ii) disease-specific dispersion of JAK2V617F allele burden in these patients; and (iii) correlation of JAK2V617F allele burden with patient clinical characteristics both between and within specific diseases. Methods: The current study was approved by the Mayo Clinic institutional review board. Myeloid neoplasms were classified as per World Health Organization criteria. Clinical, laboratory and pathology results were centrally reviewed to confirm the reported diagnoses. JAK2V617F testing was done as part of routine clinical care; testing included allele burden measurement by a quantitative allele-specific PCR assay (sensitivity 0.01%) that is in clinical use at our institution (http://www.mayomedicallaboratories.com/test-catalog/Overview/31155); reporting of allele-burden data started in January 2012. The current analysis was restricted to PB samples from adult patients seen at our institution who consented to use of their medical records for research use; results of consecutive samples tested from January 2012 to April 2015 were collected. The PCR assay calibrator is a mixture of DNA from a JAK2V617F-positive cell line (HEL) and a -negative cell line (HL60). The final result is reported as % JAK2 V617F of total JAK2 (ie, [mutated/(mutated + wild-type)] x 100%). Results: We studied 349 patients (166 female; median age 65 years, range 18-92) who met eligibility criteria. The disease distribution was: polycythemia vera (PV) 146 patients (42%), primary myelofibrosis (PMF) 63 (18%), essential thrombocythemia (ET) 55 (16%) , post-PV MF 34 (10%), post-ET MF 12 (3%), myeloproliferative neoplasm-unclassified (MPN-u) 10 (3%), refractory anemia with ringed sideroblasts and thrombocytosis (RARS-T) 8 (2%), post-MPN acute myeloid leukemia (AML) 6 (2%), MPN-myelodysplastic syndrome-unclassified (MPN/MDS-u) 4 (1%), no identified hematological disease 2 (0.5%), and 'other' 9 (2.5%). The median (range) %JAK2V617F was as follows: PV 10.5% (0.5-86%), PMF 8% (0.5-88%), ET 2% (0.5-21%), post-PV MF 24% (0.8-95%), post-ET MF 14.5% (1-88%), MPN-u 4% (0.7-11%), RARS-T 2.5% (0.5-9%), post-MPN AML 31.5% (9-92%), MPN/MDS-u 14.5% (7-74%), no hematological disease 0.7% (n/a), and other 0.9% (0.5-19%) (p<0.0001). Distribution of patients in the following 7 JAK2V617Fallele burden strata: <1%, 1-2% , 3-5%, 6-10%, 11-20%, 21-50%, and 51-100% was: 42 patients (12%), 44 (13%), 61 (17%), 61 (17%), 49 (14%), 50 (15%) and 42 (12%), respectively. PV patients were identified across the entire allele burden spectrum, namely 3%, 12%, 16%, 19%, 14%, 20% and 16% in the aforementioned strata, respectively. Disease distribution in the 42 patients with <1% allele burden was: ET 14 patients, PMF 12, PV and 'other' 5 each, RARS-T and no hematological disease 2 each, and post-PV MF and MPN-u 1 each. JAK2V617F allele burden in ET (peak 21%) was significantly lower than PMF or PV (peak 86-88%) (p<0.0001). Mutation levels in post-ET (p<0.0001) and post-PV (p=0.004) MF were significantly higher than in ET and PV, respectively. Levels in post-PV/post-ET MF were also significantly higher than in PMF (p<0.0001). In contrast, JAK2V617F levels in ET and RARS-T were comparable (p=0.8). Finally, levels in post-MPN AML were significantly higher than in ET (p=0.0001) and PMF (p=0.007), marginally higher relative to PV (p=0.05), but not significantly higher relative to post-ET/post-PV MF. Conclusions: The current study provides guidance for the interpretation of peripheral blood JAK2 V617F allele burden, in routine clinical practice: i) levels below 1% are not infrequently reported, even in phenotypically classic PV, and do not imply low likelihood of a myeloid neoplasm, ii) allele burden per se is not diagnostically specific, but a threshold of greater than 20% might be used to question a diagnosis of ET or RARS-T, and iii) the highest allele burdens are seen in fibrotic and/or leukemic transformation of chronic-phase myeloproliferative neoplasm. Disclosures Pardanani: Stemline: Research Funding.

Abstract Background Tunneled central venous catheters have traditionally been used for administration of chemotherapy (chemo) for acute leukemia treatment. Peripherally inserted central catheters (PICC) are increasingly being used, but there is an increased risk of PICC-associated thrombosis (PAT). There is limited information on management of PAT in this subgroup of patients (pt). Methods In this retrospective cohort study we investigated the primary management of PICC related upper extremity venous thrombosis (TB) in pt with hematologic malignancy undergoing chemo at Mayo Clinic Rochester. Eligible pt were 18 years and older, had documentation of PICC placement at our institution, initiation of chemo within a week of PICC placement, and the TB was objectively documented by compression venous ultrasound (CVU). Superficial venous TB (SVT) was defined as TB of subcutaneous veins (basilic and cephalic) and deep venous TB (DVT) defined as TB of brachial, axillary, subclavian, or innominate superior vena cava and internal jugular veins. We retrospectively stratified initial management (IM) into 1) PICC removal alone (PR), 2) PR and therapeutic anticoagulation (AC), 3) AC alone, and 4) conservative management (CM) consisting of symptomatic care and observation. We also investigated factors influencing IM of PAT. Long term outcomes including incidence of pulmonary embolism (PE), post phlebitic syndrome and recurrence of TB were noted. Results Between 2006 and 2012, 190 pt with AML (n=160), MDS or MDS/MF (n=10), or ALL (n=20) met our study criteria. Overall, 40/190 (21.6%) developed PAT: AML n=38, ALL n=1, MDS/MF n=1. SVT occurred in 16/40 (40%) pt and DVT occurred in 24/40 (60%) pt. IM is shown in Table 1. One pt with SVT received AC for 6 weeks as well as PR. In this pt, IM included AC because of proximity of TB to deep veins. 7 pt had a follow up (f/u) CVU within 3 months demonstrating: TB progression (prog, n=1, IM with PR alone but with prog AC was started); TB stability (stab, n=2, both IM with PR alone) and TB resolution/improvement (R/I, n= 4, 2 IM with PR and 2 IM with CM) In the 5 pt with CM, 2 pt had f/u CVU demonstrating improvement, one pt had charted clinical improvement , and 2 pt had no charted clinical f/u. Of DVT pt on AC (n=14), 5 pt completed at least 2 months of AC. Reasons for early cessation (n=9) of AC included thrombocytopenia (tcp, n=4), hematoma (n=1), hematuria (n=1), subarachnoid hemorrhage (n=1), and unknown (n=1). One pt died 2 days after presence of DVT was noted. Two pt initially treated with PR alone developed a second TB with placement of a new PICC on contralateral arm and AC was subsequently initiated (duration: 3 months (n=1), and 8 days (n=1)). 14 pt with DVT had f/u CVU within 3 months which demonstrated prog (n=2), stab (n= 8), or R/I ( n=4) of TB (Table 2). The 2 pt with TB prog subsequently received longer term AC guided by platelet count. In 5 pt with PICC not initially removed, 3 pt had PR after completion of chemo. (1-4 days), one pt had PR at discharge (27 days from TB), and one pt had PICC in place on discharge. One pt with DVT experienced a PE within one year f/u. This pt had initially completed 3 days of AC which was stopped because of tcp. Conclusion In this cohort of patients with PAT, the most common IM consisted of PR, which appeared to result in a low recurrence rate among pt with SVT. In pt with DVT, the role of AC remains to be defined given the abbreviated course of AC in most pt. Though, the lack of CVU followup in all pt limits conclusions, there appeared to be a low rate of symptomatic prog. Initiation of AC was largely guided by platelet count and degree of occlusive symptoms. Thrombocytopenia was also the most commonly cited reason for discontinuing AC. Close clinical follow up aided by the use of CVU was integral to ensure that PAT did not lead to further adverse events. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 1909 Poster Board I-932 Background: The International Prognostic Scoring System (IPSS) for primary myelofibrosis (PMF) utilizes five independent predictors of inferior survival; of these, a hemoglobin level <10 g/dL has the highest impact on survival (Cervantes et al. Blood 2009;113:2895). In the current study, we examined the additional prognostic impact of transfusion need at diagnosis or becoming transfusion-dependent in the first year of diagnosis. These events were also correlated with JAK2 or TET2 mutational status. Methods: Patients were selected from the Mayo Clinic PMF database based on availability of bone marrow histology and IPSS-relevant information at diagnosis and follow-up transfusion history at one year from diagnosis. WHO criteria were used for PMF diagnosis. Patients who underwent allogeneic hematopoietic cell transplantation were censored at time of transplant. Patient records were updated in July, 2009. Survival curves were constructed using the Kaplan-Meier method and compared by the log-rank test. Multivariable survival was analysed using Cox regression model. Results: A consecutive cohort of 254 patients was studied (median age 59 years; range 28-87; 159 males). IPSS risk category was low, intermediate (int)-1, int-2 and high in 75, 71, 62 and 46 patients, respectively. JAK2V617F was present in 118 (62%) of 192 patients and TET2 mutations in 6 (13%) of 45 patients evaluated. Transfusion need at diagnosis was documented in 62 patients whereas an additional 22 patients became transfusion-dependent during the first year of their diagnosis. The remaining 170 patients remained transfusion-independent for at least one year post-diagnosis. To date, 139 patients have died. In patients who are alive, median follow-up was 5.3 years. Median survivals in IPSS high, Int-2, Int-1 and low risk patients were, 3, 3.9, 6.8 and 12.8 years, respectively (p<0.0001). Median survival for patients requiring transfusions at diagnosis was similar to that of patients who became transfusion-dependent in their first year of diagnosis, and both were significantly shorter than the median survival seen in patients who remained transfusion-free during the first year post-diagnosis: 2.9, 2.2 and 9.7 years, respectively (p<0.0001; figure). Multivariable analysis confirmed the IPSS-independent prognostic value of transfusion status-based risk stratification. Neither JAK2 nor TET2 mutational status correlated with transfusion need. Conclusions: In PMF, becoming transfusion-dependent in the first year of diagnosis is prognostically as detrimental as requiring transfusions at initial presentation. These events are not affected by JAK2 or TET2 mutational status and confer an IPSS-independent adverse prognosis. The ability to identify Int-1 risk patients with shortened survival (Figure) holds major treatment implications. Disclosures: No relevant conflicts of interest to declare.

Abstract Background : Current risk stratification for thrombosis in essential thrombocythemia (ET) utilizes age >60 years, history of thrombosis, JAK2V617F mutation and cardiovascular risk factors (Barbui, Blood 2012;26:5128). Thrombosis history is also predictive of inferior overall survival (OS) in ET (Passamonti, Blood 2012;6:1197). However, whether or not the prognostic impact of thrombosis in ET is contributed by arterial versus venous thrombosis is unclear. Methods: More than 45,000 patient charts with "thrombocytosis" were reviewed to identify those who met either PVSG (older patient cohort) or WHO criteria for diagnosis of ET. Details of type of thrombotic events were obtained by careful review of the medical record. A subset of patients was molecularly annotated for the JAK2, CALR and MPL mutations. OS was calculated from time of initial diagnosis to the time of last follow-up or death. Thrombosis-free survival (TFS) was calculated from the time of diagnosis of ET to the time of development of thrombosis or last follow up or death, in patients censored for thrombosis. Conventional statistics was utilized for all analyses. Results :Patient characteristics: A total of 610 patients met the above-stipulated criteria (median age 57 years; 61% females). 76 patients (12%) had thrombosis documented either prior to or at diagnosis of which 50 patients (66%) experienced arterial events. 302 patients were screened for JAK2/CALR/MPL mutations with a frequency of 53.5%, 31%, and 3% respectively, while12.5% were "triple-negative".Impact of thrombosis history on OS: At a median follow-up of 8.9 years (range: 0-43.6 years), 185 (30%) deaths were documented. Themedian OS for the entire cohort was 20 years. On univariate survival analysis, the predictors of inferior OS included: age ≥60 years (p<0.001), leukocyte count ≥11 x 10(9) (p<0.001), male sex (p=0.0002), thrombosis before or at diagnosis (p=0.01), and presence of the JAK2V617F mutation (p=0.03). All but JAK2 V617F (p=0.64) remained significant on multivariable analysis. When arterial and venous events were considered separately, in univariate analysis, only arterial (p<0.001) but not venous (p=0.387) thrombotic events retained significance. The significant prognostic contribution, to survival, of arterial thrombosis history was sustained (p=0.02) during multivariable analysis, that included age ≥ 60 years (p=<0.001), leukocyte count ≥11 x 10(9) (p=0.001), and male sex (p=0.001) on OS.Impact of thrombosis history on TFS: A total of 84 (14%) patients experienced thrombotic events during follow-up amongst which 60 (71%) experienced arterial thrombosis. On univariate analysis the predictors of inferior TFS included: age ≥ 60 years (p=0.016), leukocyte count ≥11 x 10(9) (p=0.001), thrombosis before or at diagnosis (p<0.001), and presence of the JAK2V617F mutation (p=0.002). On multivariable analysis, thrombosis before or at diagnosis (p<0.001), along with leukocyte count ≥11 x 10(9) (p=0.03), and presence of the JAK2V617F mutation (p=0.01) remained significant; the significance of age ≥60 years became borderline (p=0.07). When arterial and venous thrombosis were analyzed separately, only arterial (p<0.001), but not venous (p=0.48), thrombosis history was significantly associated with inferior TFS. The factors that predicted worse outcome with TFS included arterial thrombosis at or before diagnosis (p<0.001), leukocyte count ≥11 x 10(9) (p=0.03) and presence of the JAK2V617F mutation (p=0.01); the significance of age ≥60 years became borderline (p=0.13). Conclusions : The prognostic impact of thrombosis in ET, in terms of both overall and thrombosis-free survival, might be attributed to only arterial but not venous thrombotic events. The current study also confirms the independent prothrombotic role of JAK2V617F mutation in ET and suggests an additional role for leukocytosis. Disclosures Barbui: Novartis: Speakers Bureau.

Abstract Background The Dynamic International Prognostic Scoring System-plus (DIPSS-plus) uses eight adverse factors to predict survival in primary myelofibrosis (PMF): age >65 years, hemoglobin <10 g/dL, leukocytes >25 x 109/L, circulating blasts 31%, constitutional symptoms, red cell transfusion need, platelets <100 x 109/L and unfavorable karyotype (J Clin Oncol. 2011;29:392). The latter two also predicted leukemic transformation. Most recently, an international study of 879 PMF patients identified ASXL1, SRSF2, EZH2 and IDH1/2 mutations as inter-independent predictors of shortened overall or leukemia-free survival (Leukemia 2013Apr 26 doi: 101038/leu2013119. 2013). Molecular prognostication is now well established for acute myeloid leukemia with normal karyotype (NK). There is currently little information on clinical or molecular prognostication in NK-PMF. Methods Diagnosis of PMF was according to WHO criteria. The study population included all referrals to the Mayo Clinic, Rochester, MN, USA or the University of Florence, Florence, Italy. Information on clinical and laboratory parameters and karyotype was available in all study patients, at time of referral, which coincided with time of sample collection for mutation screening. The current study considered only those patients with NK-PMF. ASXL1, EZH2, SRSF2, IDH1/2 and JAK2 mutations were performed on variable number of patients depending on bone marrow or granulocyte DNA availability. Clinical parameters examined for prognostic relevance included those listed for DIPSS-plus, less unfavorable karyotype. Results A total of 690 patients with NK-PMF patients were studied. Median age was 65 years (14-89). The percentage of patients with age >65 years was 48%, red cell transfusion dependent 35%, hemoglobin <10 g/dL 49%, platelets <100 x 10(9)/L 19%, leukocyte count >25 x 10(9)/L 15%, circulating blasts ≥1% 49% and constitutional symptoms 33%. At a median follow-up of 29 months, 351 (51%) deaths and 39 (6%) leukemic transformations were recorded. Mutational frequencies and clinical correlates The respective frequencies (mutated/number of patients studied) of JAK2V617F, ASXL1, EZH2, SRSF2 and IDH1/2 mutations were 60% (284/473), 34% (72/214), 8% (14/179), 15% (37/249) and 5% (14/262). There was no significant association between JAK2V617F and any one of the aforementioned mutations. Inter-mutation association was evident only between SRSF2 and IDH1/2 (p=0.0005). Each one of the DIPSS-plus risk parameters was examined for correlation with a specific mutation: ASXL1 was associated with leukocyte count >25 x 10(9)/L (p<0.0001) and circulating blasts ≥1% (p=0.0005); EZH2 with leukocyte count >25 x 10(9)/L (p=0.008); and SRSF2 with age >65 years (p=0.0007), transfusion need (p=0.04) and hemoglobin <10 g/dL (p=0.03). JAK2V617F was associated with age>65 years (p=0.003) and leukocyte count >25 x 10(9)/L (p=0.02). Predictors of overall and leukemia-free survival In univariate analysis, all 7 DIPSS-plus parameters, as well as ASXL1, EZH2, SRSF2 and IDH1/2 mutations showed significant association with shortened survival: When each of these analyses was adjusted for age, all except constitutional symptoms and SRSF2 mutations retained their significance. Multivariable analysis of the six age-independent DIPSS-plus variables identified all but transfusion-need as independent predictors of inferior survival. A similar analysis restricted to mutations identified ASXL1 and EZH2 as being independently adverse. When both mutations and DIPSS-plus variables were included in the multivariable model, only four parameters remained significant: age >65 years (HR 4.2; p<0.0001), platelets <100 x 10(9)/L (HR 3.4; p<0.0001), ASXL1 mutations (HR 2.2; p=0.0001) and EZH2 mutations (HR 2.7; p=0.001). A similar analysis identified SRSF2 mutations (HR 5.9; p=0.0003) and platelets <100 x 10(9)/L (HR 4.3; p=0.005) as the only predictors of leukemia-free survival. Conclusions In NK-PMF, molecular markers might be prognostically more useful than conventional models that rely on clinical features. In the current study, thrombocytopenia was the only clinical variable, other than age, with additional value to molecular prediction of survival and leukemic transformation. Disclosures: No relevant conflicts of interest to declare.