Gupta, Vikas

Abstract Allogeneic hematopoietic cell transplantation (HCT) is the only curative therapy for myelofibrosis (MF). In this large multicenter retrospective study, overall survival (OS) in MF patients treated with allogeneic HCT (551 patients) and without HCT (non-HCT) (1377 patients) was analyzed with Cox proportional hazards model. Survival analysis stratified by the Dynamic International Prognostic Scoring System (DIPSS) revealed that the first year of treatment arm assignment, due to upfront risk of transplant-related mortality (TRM), HCT was associated with inferior OS compared with non-HCT (non-HCT vs HCT: DIPSS intermediate 1 [Int-1]: hazard ratio [HR] = 0.26, P < .0001; DIPSS-Int-2 and higher: HR, 0.39, P < .0001). Similarly, in the DIPSS low-risk MF group, due to upfront TRM risk, OS was superior with non-HCT therapies compared with HCT in the first-year post treatment arm assignment (HR, 0.16, P = .006). However, after 1 year, OS was not significantly different (HR, 1.38, P = .451). Beyond 1 year of treatment arm assignment, an OS advantage with HCT therapy in Int-1 and higher DIPSS score patients was observed (non-HCT vs HCT: DIPSS-Int-1: HR, 2.64, P < .0001; DIPSS-Int-2 and higher: HR, 2.55, P < .0001). In conclusion, long-term OS advantage with HCT was observed for patients with Int-1 or higher risk MF, but at the cost of early TRM. The magnitude of OS benefit with HCT increased as DIPSS risk score increased and became apparent with longer follow-up.

Abstract INTRODUCTION: Obesity is increasingly common in the US and is frequently associated with co-morbid medical conditions that may increase the risk of HCT, often the optimal treatment for AML. HCT risk and outcomes for AML on the basis of body mass index (BMI) have not been well-characterized. Using data from the Center for International Blood and Marrow Transplant Research (CIBMTR), we have previously shown no significant difference in outcomes for autologous HCT for lymphoma among normal weight, overweight, and obese patients (pts) but worse outcomes for underweight pts (Navarro et al, BBMT 2006, 12(5): 541–51). Here, we compare outcomes by weight groups for AML patients who underwent autologous, related, or unrelated HCT. METHODS: Our final population included patients age ≥ 18 who underwent myeloablative unpurged autologous or allogeneic HCT for AML in 1st or 2nd complete remission, primary induction failure, or 1st relapse reported to the CIBMTR from 1995 to 2004. Cord blood HCTs were excluded. Four weight groups were defined based on BMI (BMI=weight (kg)/ height (m2)): underweight <18; normal=18–25; overweight >25–30; and obese >30. Treatment-related mortality (TRM), relapse, leukemia-free survival (LFS), and overall survival (OS) were compared using multivariable proportional hazards regression analysis accounting for patient, disease and HCT-related variables. RESULTS: We included 373 autologous, 2041 related, and 1801 unrelated transplant recipients. Patient-, disease-, and transplant characteristics were well-matched across weight groups and transplant types. Multivariable analysis examining risks (95% confidence intervals) relative to the normal weight group are: HCT Type Normal Underweight Overweight Obese -- =not done due to insufficient number of pts; NS=not significant; treatment failure = death or recurrence of disease. Autologous n=164 n=5 n=112 n=81 Death -- NS NS Treatment failure -- NS NS Relapse -- NS NS TRM -- NS NS Related Allogeneic n=1161 n=31 n=543 n=268 Death 1.86 (1.24–2.78) NS 1.23 (1.04–1.47) Treatment failure 2.08 (1.37–3.15) NS 1.19 (1.00–1.42) Relapse 2.02 (1.18–3.47) NS NS TRM 2.22 (1.17–4.22) NS 1.32 (1.02–1.70) Unrelated Allogeneic n=846 n=31 n=523 n=368 Death NS NS NS Treatment failure NS NS NS Relapse NS 0.82 (0.68–0.99) 0.76 (0.60–0.96) TRM NS NS NS CONCLUSIONS: There were no significant differences in risk of TRM, LFS, relapse or OS for normal weight, overweight or obese patient groups who received autologous HCT. Obese recipients of related HCT for AML had increased risk of death, treatment failure, and TRM, though the magnitude was small, an effect was not seen in the unrelated HCT group. Underweight patients who received a related, but not unrelated HCT, fared substantially worse than normal weight patients for all outcomes. It may be that the higher risk of the unrelated HCT procedure masks important but less obvious risks associated with being underweight whereas in the related donor HCT setting, such risks become manifest. Small numbers of patients limit the ability to better characterize this finding in underweight patients. Disproportionately, fewer transplants have been reported in underweight patients which suggest they experience disease and patient-related factors that preclude transplantation. No differences were observed for incidence of acute or chronic GVHD for any weight group. Overweight and obesity should not be a barrier to HCT; however, caution should be exercised in selecting underweight patients for HCT.

7059 Background: Patients with chronic myelomonocytic leukemia (CMML) have historically poor outcomes, with ~6-7 mos median OS in relapsed/refractory (r/r) setting. Splenomegaly is a poor prognostic factor and potential target in CMML. CD123 is detected on blasts, monocytes, and neoplastic microenvironmental plasmacytoid dendritic cell (pDC) infiltrates part of the CMML malignant clone (Solary, et al). Tagraxofusp, a novel CD123 targeted therapy, demonstrated high levels of activity in BPDCN, an aggressive hematologic malignancy derived from CD123-expressing pDCs, and is FDA approved in BPDCN. As such, tagraxofusp may offer a novel approach in CMML. Methods: Multicenter, 2-stage Ph1/2 enrolling patients (pts) with r/r CMML. Objectives: determine optimal dose, evaluate safety and efficacy. Stage 1 dose escalation: IV tagraxofusp (7, 9, and 12 mcg/kg/day) dosed on days 1-3 every 21 days (C1-4), 28 days (C5-7), and 42 days (C8+). Stage 2 (ongoing), pts receive optimal S1 dose (12 mcg/kg/day; no MTD). Results: 20 pts (12 CMML-1; 8 CMML-2) enrolled. 18 pts 2nd-line (2 pts in 1L), HMAs most common prior therapy. Median age 69 (43-80); 81% male. 11 (52%) had baseline splenomegaly (spleen palpable below left costal margin by physical exam) of 2-27 cm. Most common TRAEs: hypoalbuminaemia (35%), thrombocytopenia (35%), nausea (30%), vomiting (30%), and fatigue (20%). Most common ≥Gr3 TRAEs were thrombocytopenia (35%) and nausea (5%). Capillary leak syndrome in 3 pts (15%; all Gr1&2). 100% (10/10) of pts with splenomegaly had spleen response: 80% (8/10) had reductions ≥50% and 67% (4/6) with spleen size ≥5 cm had reductions ≥50%. 3 pts achieved bone marrow complete responses, including 1 pt bridged to SCT in remission. Conclusions: Tagraxofusp demonstrated single agent activity in CMML as 80% of pts showed ≥50% reduction in splenomegaly by palpation. Splenomegaly in CMML, as in myelofibrosis, is a major cause of morbidity and associated with poor prognosis and impaired QoL. Targeting splenomegaly in myeloid neoplasms with proliferative features may be an important therapeutic goal. Given CD123 expression on CMML blasts, monocytes, and malignant pDCs, tagraxofusp may offer a targeted approach, especially in pts with splenomegaly. Updated safety and efficacy data to be presented. Registrational trial planned. Clinical trial information: NCT02268253.

7058 Background: Patients with myelofibrosis (MF) who fail or are intolerant to JAK inhibitors (JAKi) have no standard treatment options. CD123 is expressed on a variety of malignancies, including MF. CD123+ plasmacytoid dendritic cells (pDCs), in the MPN microenvironment, including chronic myelomonocytic leukemia and MF, may be tumor-promoting. Monocytosis in MF associated with rapid disease progression and short survival, suggesting an accelerated disease phase. Notably, monocytes share a common precursor with CD123+ pDCs. Tagraxofusp, a novel CD123 targeted therapy, demonstrated high activity in patients with BPDCN, an aggressive hematologic malignancy derived from CD123+ pDCs, and is FDA approved in BPDCN. As such, tagraxofusp may offer a novel therapeutic approach in MF. Methods: Multicenter, 2-stage Ph 1/2 trial enrolling patients (pts) with MF relapsed, refractory, or intolerant to JAKi. Objectives: determine optimal dose, evaluate safety and efficacy. Stage 1 dose escalation: IV tagraxofusp (7, 9, and 12 mcg/kg/day) dosed daily days 1-3 every 21 days (C1-4), 28 days (C5-7), and 42 days (C8+). Stage 2 (ongoing): pts receive optimal S1 dose (12 mcg/kg/day; no MTD). Results: 23 r/r pts treated. Median age 69 (55-81); 57% female. DIPSS Plus: 4% INT-1, 55% INT-2, 41% high. Baseline platelets: median 59 K/uL (15-579); 70% (16/23) <100 K/uL, 8 pts <50 K/uL. 87% (20/23) baseline splenomegaly (palpable ≥5 cm below left costal margin by physical exam). Most common TRAEs: headache (22%), hypoalbuminaemia (22%), ALT incr. (17%) and thrombocytopenia (17%). Most common ≥Gr3 TRAE thrombocytopenia (2%). Capillary leak syndrome in 1 pt (4%; Gr3). 57% (8/14) of pts with baseline spleen ≥5cm BCM spleen responses: 43% (6/14) had ≥29% and 21% (3/14) had ≥45% reduction. 100% of pts with baseline spleen ≥5cm and monocytosis splenomegaly reductions: 80% (4/5) had ≥29% and 40% (2/5) had ≥45%. 6 pts (3 monocytosis pts and 5 pts platelets <100 K/uL) had 6 mos+ duration, 9 pts ongoing. Conclusions: Tagraxofusp demonstrated single agent activity (reduction in splenomegaly) and manageable safety in R/R MF, including pts with monocytosis, an unmet medical need. Given the presence of CD123+ pDCs, tagraxofusp may offer a novel targeted approach in MF. Updated data to be presented. Registrational designs are being evaluated. Clinical trial information: NCT02268253.

6624 Background: RUX is a JAK1/JAK2 inhibitor that demonstrated significant clinical benefit in patients (pts) with MF in COMFORT-I (NCT00952289), a phase 3, randomized (1:1), double-blind, placebo (PBO)-controlled study (N=309). Methods: Pts assessed MF-related symptoms daily using the modified Myelofibrosis Symptom Assessment Form v2.0; Total Symptom Score (TSS) was calculated from individual scores for abdominal discomfort, pain under left ribs, early satiety, night sweats, itching, and bone/muscle pain. Therapy was interrupted if platelet or absolute neutrophil count fell below 50,000/µL or 500/µL, respectively. AEs were evaluated during treatment interruption or discontinuation. The protocol suggested an optional tapering strategy and possible addition of steroids if RUX therapy was discontinued for reasons other than thrombocytopenia. Results: In RUX-treated patients with treatment interruption, TSS gradually returned to baseline levels over approximately 1 week. The most common AE leading to treatment interruption or discontinuation in each group was thrombocytopenia (n=11, RUX) and abdominal pain (n=4, PBO). Of these, only 1 RUX pt discontinued for thrombocytopenia. A summary of new onset/worsening AE data after treatment interruption or discontinuation is presented (Table). Of 58 pts who discontinued study treatment, 23 (n=10, RUX; n=13, PBO) experienced a total of 43 SAEs (19, RUX; 24, PBO) that showed no specific pattern or difference between treatment groups. Four of 21 RUX-treated pts had dose tapering following study drug discontinuation. Conclusions: Apart from the expected return of MF-related symptoms, there was no pattern of AEs to suggest that RUX interruption or discontinuation is associated with a specific withdrawal syndrome. [Table: see text]

Background Patients with myelofibrosis (MF) who fail or are intolerant to JAK inhibitors (JAKi) have limited treatment options. Tagraxofusp is a targeted therapy directed to CD123 that was approved by the FDA for the treatment of patients with blastic plasmacytoid dendritic cell neoplasm (BPDCN). CD123 is expressed on a variety of malignancies including BPDCN, acute myeloid leukemia (AML), and certain myeloproliferative neoplasms (MPN), including MF. High-expressing CD123+ plasmacytoid dendritic cells (pDCs), the cell of origin of BPDCN, have been detected in the microenvironment of certain myeloid neoplasms, including MF, where they may play a tumor-promoting role. Notably, pDCs share a common precursor cell with monocytes, and monocytosis has been reported as a poor prognostic factor in patients with MF, associated with shortened overall survival. We hypothesized that tagraxofusp may offer a novel and rational therapeutic approach in patients with relapsed/refractory MF, and in particular in patients with monocytosis. Aims To assess safety and evaluate efficacy of tagraxofusp in patients with MF who were relapsed, refractory, or unable to tolerate JAKi. Methods This multicenter, open-label Phase 1/2 trial is enrolling patients with MF. In the Stage 1 (dose escalation), tagraxofusp was administered as a daily IV infusion at 7, 9, or 12 mcg/kg on days 1-3 every 21 days (cycle 1-4), 28 days (cycles 5-7), and 42 days (cycles 8+). In Stage 2 (expansion), patients are receiving the recommended Phase 2 dose (12 mcg/kg). Results 27 patients with MF have been treated, including 14 patients who had previously received ≥ 3 lines of therapy. Median age was 69 years (range 54-81), 52% were female, and 26% had baseline monocytosis (≥ 1x109/L). Baseline risk assessment based on the DIPSS Plus risk group assessment showed 1 patient (4%) with intermediate-1, 16 patients (59%) with intermediate-2, and 9 patients (33%) with high-risk. At study entry, median platelet count was 59 K/uL; 67% of patients had baseline platelets &lt;100 K/uL, of which 10 patients had platelets &lt;50 K/uL. 74% of patients had baseline splenomegaly (spleen palpable ≥ 5 cm below the left costal margin). In terms of molecular mutations: JAK2 (n=18); CALR (n=3); and MPL (n=1), and additional molecular analysis showed ASXL1 (n=5). In terms of cytogenetics, 30% (8/27) of patients had an abnormal karyotype, mainly 20q- (n=3), -5/5q-, -7/7q-, and 12p- (each n=2). Most common treatment-related adverse events (TRAEs, incidence ≥ 15%) include headache, hypoalbuminemia, increased levels of alanine aminotransferase, thrombocytopenia, and anemia. Most common ≥ grade 3 TRAE's were thrombocytopenia (8%) and anemia (15%). Capillary leak syndrome was reported in 1 patient (grade 3). Among the 17 evaluable patients with baseline splenomegaly, 53% had spleen size reductions, of which 4 patients had reductions of &gt; 45%. In 5 patients with baseline splenomegaly and monocytosis (&gt;1x109/L monocytes), 80% (4/5) had spleen size reductions, of which 2 had reductions of &gt;45%. Five patients had treatment duration of 12 months or more, with two of those patients ongoing (13+,and 24+ months, respectively). Six patients with platelets &lt;100 K/uL had treatment duration of 6 months or more, including 3 patients with baseline monocytosis who had treatment duration of 12 months or more. 7 patients had objective responses (IWG-MRT 2013 responses), including Clinical improvement (CI) (n=5) and CI (Symptom) (n=2). 20 patients were evaluable for symptomatic assessment (MPN-SAF TSS). Symptom response rate was 45% (defined as ³50% reduction in TSS score). Reductions in both symptom score and spleen size was noted in 6/6 patients who had splenomegaly at baseline. Conclusions Tagraxofusp demonstrated single agent clinical activity, with a predictable and manageable safety profile, in patients with relapsed/refractory MF, including in patients with monocytosis, a poor prognostic factor and unmet medical need. Tagraxofusp may offer MF patients, and MF patients with monocytosis in particular, a novel treatment option. Enrollment continues, and updated trial data will be presented. This clinical trial is registered at clinicaltrials.gov: NCT02268253. Disclosures Pemmaraju: affymetrix: Research Funding; sagerstrong: Research Funding; Daiichi-Sankyo: Research Funding; plexxikon: Research Funding; novartis: Consultancy, Research Funding; Stemline Therapeutics: Consultancy, Honoraria, Research Funding; cellectis: Research Funding; celgene: Consultancy, Honoraria; samus: Research Funding; mustangbio: Consultancy, Research Funding; abbvie: Consultancy, Honoraria, Research Funding; incyte: Consultancy, Research Funding. Gupta:Sierra Oncology: Honoraria, Membership on an entity's Board of Directors or advisory committees; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Incyte: Honoraria, Research Funding. Yacoub:Dynavax: Equity Ownership; Cara: Equity Ownership; Ardelyx: Equity Ownership; Incyte: Consultancy, Honoraria, Speakers Bureau; Seattle Genetics: Honoraria, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Agios: Speakers Bureau; Hylapharm: Equity Ownership. Wang:Agios: Other: Advisory role; Amgen: Other: Advisory role; Daiichi: Other: Advisory role; Stemline: Other: Advisory role, Speakers Bureau; Pfizer: Other: Advisory role, Speakers Bureau; celyad: Other: Advisory role; Abbvie: Other: Advisory role; Kite: Other: Advisory role; Jazz: Other: Advisory role; Astellas: Other: Advisory role, Speakers Bureau. Lee:AstraZeneca Pharmaceuticals: Consultancy; Ai Therapeutics: Research Funding; Karyopharm Therapeutics: Consultancy; Roche Molecular Systems: Consultancy; Jazz Pharmaceuticals, Inc: Consultancy; Helsinn: Consultancy. Schiller:Agios: Research Funding, Speakers Bureau; Karyopharm: Research Funding; Onconova: Research Funding; Novartis: Research Funding; Pfizer Pharmaceuticals: Equity Ownership, Research Funding; Sangamo Therapeutics: Research Funding; Amgen: Other, Research Funding; Astellas: Research Funding; Biomed Valley Discoveries: Research Funding; Bristol Myer Squibb: Research Funding; Celgene: Research Funding, Speakers Bureau; Constellation Pharmaceutical: Research Funding; Daiichi Sankyo: Research Funding; Eli Lilly and Company: Research Funding; FujiFilm: Research Funding; Genzyme: Research Funding; Gilead: Research Funding; Incyte: Research Funding; J&J: Research Funding; Jazz Pharmaceuticals: Honoraria, Research Funding. Sardone:Stemline Therapeutics: Employment, Equity Ownership. Wysowskyj:Stemline Therapeutics: Employment, Equity Ownership. Chen:Stemline Therapeutics: Employment, Equity Ownership. Brooks:Stemline Therapeutics: Employment, Equity Ownership, Patents & Royalties. Poradosu:Stemline Therapeutics: Employment, Equity Ownership, Patents & Royalties. McDonald:Stemline Therapeutics: Employment, Equity Ownership. Rupprecht:Stemline Therapeutics: Employment, Equity Ownership. Talpaz:Imago BioSciences: Consultancy, Research Funding; Celgene: Consultancy, Research Funding; CTI BioPharma: Research Funding; Constellation: Research Funding; Incyte: Research Funding; Novartis: Research Funding; Samus Therapeutics: Research Funding. Taparia:Novartis: Consultancy. Verstovsek:Protaganist Therapeutics: Research Funding; Ital Pharma: Research Funding; Astrazeneca: Research Funding; Pharma Essentia: Research Funding; Sierra Oncology: Research Funding; Novartis: Consultancy, Research Funding; Blueprint Medicines Corp: Research Funding; Incyte: Research Funding; Roche: Research Funding; NS Pharma: Research Funding; Celgene: Consultancy, Research Funding; Gilead: Research Funding; Promedior: Research Funding; CTI BioPharma Corp: Research Funding; Genetech: Research Funding; Constellation: Consultancy; Pragmatist: Consultancy. Khoury:Kiromic: Research Funding; Angle: Research Funding; Stemline Therapeutics: Research Funding. Patnaik:Stem Line Pharmaceuticals.: Membership on an entity's Board of Directors or advisory committees.

Abstract We conducted a two-stage Simon phase II, multi-center study in two groups: related and unrelated donor transplants. Based on the results of a previously conducted prospective study (MPD-RC 101, Rondelli et al, Blood, 2014) demonstrating relatively poor outcome in the unrelated donor arm due to high non-relapse mortality and graft failure (GF), we sought to improve this noted disparate outcome. The hypothesis driving this trial was that pre-transplant JAK1/2 inhibition would result in an increase in performance status, reduction in splenomegaly and inflammatory cytokine expression leading to improved HCT outcome. In the unrelated arm in the first 11 patients, the trial was to be terminated if 6 or more failures (GF or death due to any cause by 100 days post-transplant) were observed. In the related arm in the first 11 patients, the trial was to be terminated if 3 or more failures were observed. Key eligibility criteria included diagnosis of primary MF or post-PV MF or post-ET MF, age 18-70 years, and DIPSS int-2/high risk disease or int-1 risk disease with at least one additional risk factor. Patients could have received up to 6 months of ruxolitinib prior to enrollment provided that there was no evidence of loss of response to ruxolitinib during this period. Ruxolitinib was given for 56 days followed by a 4-day taper ending day -6. The conditioning regimen utilized was IV fludarabine 40 mg/m2 IV and IV busulfan 2.0 mg/KBW IV once daily for 4 days from days -5 to -2. GVHD prophylaxis consisted of cyclosporine or tacrolimus and methotrexate. Patients undergoing unrelated donor HCT also received low dose thymoglobulin (0.5 mg/KBW on day -3/2.0 mg/KBW on day -2 & day -1). The study has enrolled 21 patients (related, 7; unrelated, 14) with a median age of 59 years (range 39-70). 17 (81%) patients had palpable splenomegaly with median spleen size of 11 cm, and 16 patients had int-2/high risk disease. Of the 21 patients, 19 patients received transplantation (related 5/7 [71%], unrelated 14/14 [100%] underwent transplant). Two patients in the related arm did not proceed with transplantation as one progressed to acute myeloid leukemia (baseline blast count 14%) and another experienced sudden death of unknown cause the day after starting ruxolitinib. In addition, there was one GF in the related arm. Due to these 3 failures in the related arm, further enrollment was stopped in this arm per protocol criteria. In the unrelated arm, there were 4 failures (Table 1), and the study met the protocol criteria for further enrollment. Mean percent reduction from baseline to day -9 in palpable spleen length below the left costal margin was 30% (p=0.04) in 12 patients with palpable spleen. 5/12 (42%) had any reduction, and 4/12 (33%) had ≥50% reduction. From baseline to best score on day -9 or day -5, 11/19 (58%) patients with data had any improvement in MPN-SAF Total Symptom Score, and 7/19 (37%) had ≥50% reduction. Of the 19 patients undergoing transplant, ruxolitinib was tapered off successfully in all patients with no significant withdrawal syndrome or delay in transplant noted. In the pre-transplant phase, most common Grade 3 or higher toxicity considered to be related to ruxolitinib was anemia (6/21, 29%), and one patient had legionella pneumonia deemed not related to ruxolitinib. Median time to neutrophil recovery (> 0.5 x 109/L) and platelet recovery (>20 x 109/L) was 19 and 19.5 days. By day 100, 7/19 (37%) experienced Grade 2-4 acute GVHD (G2 in 5, G3 in 2). In the first 30 days post HCT, most common Grade 3 or higher toxicity was febrile neutropenia seen in 5/19 (26%) patients. In all 21 patients from time of registration with median follow-up of 5.8 months, 6-mo overall survival was 75% (95% CI 44-90%). For the 19 transplant recipients, 6-mo OS was 83% (95% CI 55-94%) from date of transplant. Conclusions: Stage 1 trial results highlight important observations. First, ruxolitinib may not be suitable for all myelofibrosis patients pre-transplant, eg, alternative approaches might be more appropriate in the context of accelerating disease. Second, we demonstrated a safe approach for tapering ruxolitinib prior to transplant which allowed patients to commence conditioning with reduced spleen size and symptoms. Third, pre-transplant ruxolitinib does not appear to have any adverse impact on early post-HCT outcomes following unrelated donor transplant, although Stage 2 enrollment will be required to better assess the potential benefits of this approach. Disclosures Gupta: Novartis: Consultancy, Honoraria, Research Funding; Incyte: Consultancy, Research Funding. Yacoub:Incyte: Consultancy, Honoraria, Speakers Bureau; Seattle Genetics: Consultancy, Honoraria, Speakers Bureau; Alexion: Honoraria. Mesa:Promedior: Research Funding; Celgene: Research Funding; CTI: Research Funding; Gilead: Research Funding; Incyte: Research Funding; Galena: Consultancy; Ariad: Consultancy; Novartis: Consultancy. Mead:Novartis: Honoraria, Research Funding, Speakers Bureau. Mascarenhas:Novartis: Other: DSMB , Research Funding; Roche: Research Funding; Promedior: Research Funding; Janssen: Research Funding; CTI Biopharma: Research Funding; Incyte: Other: Clinical Trial Steereing Committee, Research Funding.

Abstract Introduction: Allogeneic hematopoietic stem cell transplantation (HCT) is the only curative therapy for myelofibrosis (MF). Consideration of HCT is recommended by international working groups and national guidelines for MF patients (pts) age <70 with intermediate-1 with adverse indicators, intermediate-2 or high-risk disease by the Dynamic International Prognostic Scoring System (DIPSS) for MF- a recommendation made in the absence of clear data indicating the optimal timing of HCT for MF. In this large multicenter retrospective study, we analyze overall survival in MF pts treated with and without HCT. Methods: Disease characteristics, treatments, and outcome data from MF pts receiving non-transplant therapy at 14 US academic medical centers between 2000-2014 were retrospectively collected. MF pts who underwent HCT were identified from the Center for International Blood and Marrow Transplant Research (CIBMTR). The Cox proportional hazards model was used. The reference time point (time zero) was time of referral for the non-transplant (non-HCT) arm and the time of transplant for the HCT arm. The main effect variable (HCT vs. non-HCT) violated the proportionality assumption where comparing to non-HCT, mortality was higher with HCT in early time period from time zero but then was lower in late time period; therefore, the comparison is presented as early time period and late time period. The Cox model identified 14 months from time zero as the ideal cut point to define early and late time periods. The proportionality assumption is satisfied within each of these two periods. Results: A total of 1377 and 551 pts were included in the non-HCT and HCT arms, respectively (Table 1). In the overall cohort, survival was higher with non-HCT vs. HCT in early time period (relative risk [RR]: 0.34, P< .0001, Figure 1D), but in late time period survival was lower with non-HCT vs. HCT (RR: 2.37, P< 0.001) (Table 2). In the DIPSS low-risk MF group, while survival was higher with non-HCT vs. HCT in the early time period (RR: 0.19, P=0.007, Figure 1A), survival was lower with non-HCT in the late time period, but the latter did not reach statistical significance (RR: 1.45, P=0.39). In the DIPSS intermediate-1 risk group, a survival advantage was present with non-HCT treatments vs. HCT in the early time period (RR: 0.27, P < .0001, Figure 1B), however survival was lower with non-HCT in the late time period (RR: 3.13, P < .0001). Similarly, in those with DIPSS intermediate-2 and high-risk MF, survival advantage was observed with non-HCT in the early time period (RR: 0.41, P< .0001, Figure IC), but survival was lower with non-HCT in the late time period (RR: 2.82, P < .0001). Conclusion: A long-term survival advantage with transplant was observed for pts with intermediate-1 or higher risk MF, but at the cost of potential early mortality. The magnitude of benefit increased as DIPSS risk score increased. Although this retrospective study has limitations, the results have an impact on clinical practice by suggesting that transplantation could be considered earlier in the disease course and supports the recommendation for consideration of HCT in the setting of intermediate-1 risk MF. Disclosures Gowin: Incyte: Consultancy, Other: Scientific Advisory Board, Speakers Bureau. Verstovsek:Celgene: Membership on an entity's Board of Directors or advisory committees; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Incyte: Consultancy; Italfarmaco: Membership on an entity's Board of Directors or advisory committees. Ali:Incyte Corporation: Membership on an entity's Board of Directors or advisory committees. Gupta:Novartis: Consultancy, Honoraria, Research Funding; Incyte: Research Funding. Gerds:Celgene: Consultancy; Apexx Oncology: Consultancy; CTI Biopharma: Consultancy; Incyte: Consultancy.