Abstract: Introduction Allogeneic hematopoietic cell transplantation (HCT) remains the only potentially curative therapy for myelofibrosis (MF). However, despite improvements in donor availability, most patients receive non-HCT therapy in the form of conventional drugs (e.g. hydroxyurea), or more recently, JAK inhibitor therapy (JAKi). For a proportion of patients, JAKi offers durable clinical benefit in the form of symptom improvement, reduction in splenomegaly and improved quality of life. The role of HCT in the JAKi era has not been well studied, and despite recent advances in the understanding of the pathogenesis and refinement of prognostic scoring systems,real-world decision making remains challenging. The goal of this study was to compare the outcomes of patients who received upfront JAKi vs. HCT for MF in dynamic international prognostic scoring system (DIPSS)-stratified categories. Methods This multicentre study included adult patients up to age 70 years with primary or secondary MF in chronic phase who were first seen at one of the eight participating centres in Canada and the United States between January 1, 2012 and December 31, 2017. The primary outcome was overall survival (OS) in patients with DIPSS int-1 risk or higher who received JAKi vs. HCT. To compare the planned, upfront treatment strategy, patients who received a short-course of JAKi as bridging therapy prior to HCT ( & lt; 6 months or documented plan of care) were analysed in the HCT group. Similarly, patients who were treated with JAKi, but received a HCT following JAKi failure ( & gt;12 months or documented progression) were analysed in the JAKi group. To minimize selection and lead-time bias, OS was calculated from the start of JAKi and date of transplant, respectively. Patients who were transplanted for accelerated- or blast-phase disease were not included in the analysis. OS was calculated using the Kaplan-Meier method and differences were tested using the log-rank test. Results Between 2012 and 2017, 506 patients with MF were seen at the study centres and 311 received JAKi or HCT. Of these, 174 (56%) had PMF and 137 (44%) had post-ET or post-PV MF. An upfront HCT strategy was used in 86 patients and an upfront JAKi strategy was used in 225 patients. Of those, 53 patients went on to receive HCT following JAKi failure. The median duration of follow up of survivors was 32.8 (1.2 - 99.2) months. The median OS of MF patients with DIPSS int-1 or higher was 65.3 (95% CI: 55.7 - 76.4) months for patients treated with an upfront JAKi strategy and 89.4 (95% CI: 20.4 - not reached) months for those treated with an upfront HCT strategy (p=0.018, Figure). The survival of patients with int-1 risk disease was 0.78 (95% CI: 0.69 - 0.88) in the JAKi group vs. 0.60 (95% CI: 0.43 - 0.83) in the HCT group at 36 months and 0.68 (95% CI: 0.57-0.82) in the JAKi group vs. 0.60 (95% CI: 0.43-0.83) in the HCT group at 60 months. Given the small number of patients with DIPSS high risk, these patients were combined with the int-2 cohort for analysis. The survival of patients with int-2/high risk disease was 0.58 (95% CI: 0.49 - 0.69) in the JAKi group vs. 0.49 (95% CI: 0.36 - 0.65) in the HCT group at 36 months and 0.37 (95% CI: 0.24-0.55) in the JAKi group vs. 0.45 (95% CI: 0.32-0.62) in the HCT group at 60 months (Table). Conclusions Previous studies, which included many patients treated in the era before widespread availability of JAKi, supported an upfront HCT strategy in patients with higher risk MF. While these agents have not demonstrated consistent disease-modifying effects, many patients do experience durable clinical benefit in the form of symptom improvement and reduction in spleen size. In our study, there was no clear benefit of upfront HCT. The median OS of patients who received HCT upfront was longer than that of patients who were treated with upfront JAKi, but upfront HCT was associated with early mortality and the OS benefit was not apparent until after 5 years. An inherent limitation of this study is a lack of data on potentially important comorbid conditions which may have contributed to selection bias. However, to our knowledge this is the largest study to compare upfront HCT and JAKi strategies in patients with higher risk MF, making these findings relevant to modern clinical practice in the JAKi era. A delayed transplant approach may be appropriate for selected patients who are deriving clinical benefit from JAKi. Defining the optimal timing for HCT in higher risk MF remains a question for future research. Disclosures Maze: Pfizer: Consultancy; Novartis: Honoraria; Takeda: Research Funding. Arcasoy:CTI Biopharma: Research Funding; Samus Therapeutics: Research Funding; Gilead: Research Funding; Incyte: Research Funding; Janssen: Research Funding. Yacoub:Dynavax: Current equity holder in publicly-traded company; Ardelyx: Current equity holder in publicly-traded company; Cara Therapeutics: Current equity holder in publicly-traded company; Hylapharm: Current equity holder in private company; Incyte: Speakers Bureau; Agios: Honoraria, Speakers Bureau; Novartis: Speakers Bureau; Roche: Other: Support of parent study and funding of editorial support. McNamara:Novartis: Honoraria. Foltz:Celgene: Membership on an entity's Board of Directors or advisory committees; Constellation: Research Funding; Incyte: Research Funding; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees. Gupta:Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Bristol MyersSquibb: Honoraria, Membership on an entity's Board of Directors or advisory committees; Sierra Oncology: Consultancy, Membership on an entity's Board of Directors or advisory committees; Pfizer: Consultancy; Incyte: Honoraria, Research Funding.

Abstract: Background: Few studies have objectively assessed the value of routine clinical, laboratory and radiological evaluation to detect recurrence of Hodgkin lymphoma. The optimal follow up of patients (pts) in complete remission following initial therapy has not been defined. Methods: We identified 99 adult pts with Hodgkin lymphoma, who received treatment and follow up supervised by the British Columbia Cancer Agency and relapsed between Jan 1990 and April 2004. Pts who did not achieve complete remission or had a second hematological malignancy were excluded. Pts were followed with clinical assessment, chest radiograph, CBC and alk phos every 3 m for 2 y, then every 6 m for 3 y, then annually. Routine CT scans were recommended every 6 m for 3 y then annually for 2 y. Relapses were categorized as identified by pt (symptoms, new palpable disease) or by physician (routine physical examination or radiological or laboratory studies in asymptomatic pts). Results: Median age at original diagnosis was 28 y (range 14–73). 86 pts initally had advanced and 13 limited stage disease. Primary treatment was chemotherapy +/− radiation in 93 pts and radiation alone in 6 pts. 10 pts had autologous SCT for primary refractory disease. Median follow up from diagnosis was 82 months (range 12–241). Median time to first relapse from completion of treatment was 14 months (range 2–142). Of the 99 relapses, 75 (76%) were identified by the pt and 24 (24%) by the physician. Pt systemic symptoms of relapse were: fatigue 13 pts; alcohol induced pain 3; weight loss 11; pruritis 9; night sweats 19; fever 6. Local symptoms were shortness of breath 8; chest pain 8; back pain 9; abd pain 3. 29 pts had more than 1 symptom. 44 pts noted a new lymph node or mass and 1 pt had leg swelling. 24 relapses were detected by physician: 14 on CXR, 7 on CT scan, 1 on lymphangiogram and 2 on physical exam. No asymptomatic relapses were identified by laboratory abnormalities alone. 2 of the 13 relapses in pts with initially limited stage disease (18%) were detected by physician (1 CXR and 1 CT chest), vs 22 of 86 (26%) in advanced disease. 78% of relapses occurred within 36 months of completing initial treatment. Patient vs Physician Detected Relapses by Follow up Period Time from therapy completion Patient Detected Relapse Physician Detected Relapse Total # Relapses & lt; 12 months 37 (80%) 9 (20%) 46 12–35 months 19 (61%) 12 (39%) 31 ≥ 36 months 19 (86%) 3 (14%) 22 All Relapses 75 (76%) 24 (24%) 99 80% of relapses within 12 months of therapy completion were detected by pts despite more intensive physician surveillance in this period. The proportion of physician detected relapses was greatest 12–36 months after completion of treatment, possibly due to relapse with more slowly progressive disease amenable to detection on periodic routine testing while it is still being carried on relatively frequently. The 22% of relapses occurring in the period of less frequent surveillance, greater than 3 years after treatment, were primarily detected by pts. Conclusions: 76% of Hodgkin lymphoma relapses were detected by the pt and 78% of relapses occurred within 3 years of therapy completion. Asymptomatic relapse was detected on physical exam and radiological studies but not laboratory testing. The highest proportion of physician detected relapses occurred 12–35 months after treatment. Annual routine follow up beyond 36 months contributed minimally to relapse detection, identifying only 3% of total relapses.

Abstract: Timely diagnosis of patients (pts) with polycythemia vera (PV) and essential thrombocythemia (ET) is important given the risks of thrombotic and hemorrhagic complications, disease progression and associated symptoms.  Pts often present initially to primary care physicians, who may have limited previous experience with PV/ET given the low prevalence.  Little is known about the timeliness of referral or diagnostic testing after identification of abnormal blood test results or if delays in diagnosis affect patient outcomes. Objectives To determine the time from initial lab abnormality to referral, diagnosis and treatment of pts with PV and ET. Methods Pts at a single Canadian academic institution newly diagnosed with PV or ET from Jan 2010 to May 2013 were identified.  Retrospective data was collected including demographics, lab values, diagnostic testing and treatments. Results Demographics: 26 pts with PV and 34 with ET were identified.  Median age was 67.5 (44-89) y for PV and 66.5 (34-92) y for ET. Delay in Referral and Diagnosis: 98% of pts were referred directly to a hematologist by their primary care physician.  69% of PV pts were referred within 30 days and 92% within 90 days of initial lab abnormality.  Median time from referral to diagnosis was 98 (0-221) days.  41% of ET pts were referred within 30 days and 56% within 90 days of initial lab abnormality.  Median time from referral to diagnosis was 121 (8-638) days.  PV pts were referred sooner, median 20 (0-187) days, than ET pts, median 67 (0-3743) days (p=0.01).  The median delay from referral until hematology assessment was 51 days for PV compared to 78 days for ET (p=0.08).  After assessment by the hematologist, it required a median of 35 days to make a diagnosis of PV and 25 days for a diagnosis of ET (p=0.31). Referrals by platelet (plt) count: There was a trend to earlier referral of ET pts with higher platelet (plt) counts.  15/20 (75%) ET pts with plt count 〉 600 were referred within 90 days of initial lab abnormality whereas only 4/14 (29%) of pts with plt count 450-600 were referred within 90 days (p=0.056). Treatment of PV pts: 22/26 (85%) pts received phlebotomy at or after referral at the direction of a hematologist.  Average delay in referral (and phlebotomy initiation) for patients treated with phlebotomy was 32 days.  13/26 (50%) pts were initiated on treatment with hydroxyurea within 2 months of diagnosis.  Average delay in diagnosis (and hydroxyurea initiation) in this subgroup was 142 days.   11/26 (42%) pts were receiving ASA prior to the initial hematological consultation.  12/26 (46%) were initiated on ASA at or shortly after hematological consultation.  Average delay to hematology consultation (and ASA initiation) was 90 days in this subgroup. Treatment of ET pts: 8/34 (24%) pts were initiated on treatment with hydroxyurea within 2 months of diagnosis.  Average delay in diagnosis (and hydroxyurea initiation) in this subgroup was 790 days. 17/34 (50%) pts were receiving ASA prior to the initial hematological consultation.  15/34 (44%) were initiated on ASA at or shortly after hematological consultation.  Average delay to hematology consultation (and ASA initiation) was 355 days in this subgroup. No thrombotic or major hemorrhagic complications occurred in any PV/ET pts between the time of initial lab abnormality and diagnosis. Discussion This study demonstrates the marked variability in time from lab abnormality to referral and diagnosis for PV/ET pts.  Primary care providers were more likely to promptly refer PV pts than ET pts, and particularly tended to overlook referral and investigation of pts with modestly elevated plt counts of 450-600.  This is a concern, as risk of thrombosis in ET pts is independent of plt count.  Delays were also apparent in wait times for hematology appointments and subsequent diagnostic tests.   The delay in diagnosis led to a delay in initiation of therapy to reduce risk of thrombosis in both PV and ET pts.  Possible strategies to expedite diagnosis include targeted education of primary care physicians focusing on identification of lab features of PV/ET.  Directive comments on lab reports by community hematopathologists may also facilitate prompt referral and investigation. Disclosures: No relevant conflicts of interest to declare.

Abstract: BACKGROUND: Ruxolitinib (RUX) is a potent JAK1/JAK2 inhibitor that demonstrated improvements in splenomegaly and disease-related symptoms, as well as improved survival, in patients (pts) with intermediate (Int)-2- or high-risk myelofibrosis (MF), and has proved superior to placebo and best available therapy in the phase 3 COMFORT studies. JUMP is an expanded-access phase 3b trial designed to assess the safety and efficacy of RUX in pts with MF and includes patients with no access to RUX outside a clinical trial. As of Dec 2014, final enrollment was 2233 pts in 26 countries. METHODS: Eligible pts had Int-2- or high-risk MF with or without splenomegaly, or Int-1-risk MF with a palpable spleen (≥ 5 cm from the costal margin). Pts received starting doses of RUX based on platelet counts at baseline (5 mg twice daily [bid; ≥ 50 to 〈 100 × 109/L], 15 mg bid [100 to 200 × 109/L] , or 20 mg bid [ 〉 200 × 109/L]). The primary endpoint was assessment of safety and tolerability of RUX. Additional analyses included changes in palpable spleen length and symptom scores as measured by the FACT-Lymphoma total score (FACT-Lym TS). The final analysis will be performed after all pts have completed 24 months of treatment or discontinued the study. RESULTS: This analysis includes 1869 pts (primary MF, 59.1%; n = 1105) who started treatment ≥ 1 year before the data cutoff date (01 Jan 2015). At baseline, median age was 67 y (range, 18-89 y); 54.1% were male; median palpable spleen length was 12 cm below the costal margin; 87 pts did not have splenomegaly. Median hemoglobin (Hb) was 106 g/L, and 38.9% of pts had Hb levels ˂ 100 g/L; median platelet count was 257 × 109/L; mean FACT-Lym TS and FACIT-Fatigue score were 113.7 and 33.2, respectively. At data cutoff, 37.0% of pts remained on treatment; 26.1% had completed treatment per protocol. Primary reasons for discontinuation included adverse events (AEs; 17.4%), disease progression (8.2%), and death (3.4%). Median exposure was 13.6 months; the median average daily dose was 36.7 mg for pts starting at 20 mg bid (n = 1168; 62.5%) and 23.2 mg for pts starting at 15 mg bid (n = 559; 29.9%). The majority of pts (66.0%) had dose modifications, and 26.2% had a dose interruption. Grade 3/4 hematologic AEs included anemia (34.0%), thrombocytopenia (14.9%), and neutropenia (3.9%), which led to discontinuation in 2.2%, 3.3%, and 0.2% of pts, respectively. The most common nonhematologic AEs (≥ 10%) were pyrexia (14.5%), asthenia (13.8%), diarrhea (12.4%), and fatigue (10.3%), and were primarily grade 1/2; grade 3/4 AEs were low overall (≤ 2%), except pneumonia (3.9%), which led to discontinuation in 9 pts (0.5%). Rates of infections were low; all-grade infections ≥ 5% included pneumonia (6.2%), urinary tract infection (5.7%), and nasopharyngitis (5.3%). Tuberculosis was reported in 5 pts (0.3%; grade 3/4, 0.1%); hepatitis B was reported in 1 pt (0.1%; grade 3/4, 0.1%). At wk 24 and 48, 57.2% (742/1297) and 62.0% (588/949) of pts with baseline splenomegaly achieved a ≥ 50% reduction from baseline in palpable spleen length; 22.9% (297/1297) and 19.0% (180/949) had 25% to 50% reductions, respectively. Most pts (70.5%; 1208/1713) experienced a ≥ 50% reduction at any time; 23.3% (399/1713) had complete resolution of splenomegaly (Figure). At wk 24 and 48, 96.6% (57/59) and 91.5% (43/47) of evaluable pts without splenomegaly at baseline continued to have a nonpalpable spleen; 1.7% (1/59) and 4.3% (2/47) had a spleen that was 0-5 cm, and 1.7% (1/59) and 4.3% (2/47) had a spleen ≥ 5 cm. A large proportion of pts achieved a response (ie, a clinically significant improvement) on the FACT-Lym TS and FACIT-Fatigue at wk 24 (43.0% [525/1220] ; 47.1% [593/1258]) and wk 48 (43.2% [368/852] ; 45.7% [396/867]). Similar responses were seen in pts without a palpable spleen (FACT-Lym TS: wk 24, 44.0% [22/50] ; wk 48, 36.1% [13/36]; FACIT-Fatigue: wk 24, 49.1% [27/55] ; wk 48, 35.1% [13/37]). CONCLUSIONS: To date, JUMP includes the largest cohort of pts with MF treated with RUX. Consistent with previous findings, anemia and thrombocytopenia were the most common AEs but rarely led to discontinuation. As observed previously, most pts experienced reductions in splenomegaly and symptoms with RUX treatment. Clinically meaningful improvements in symptoms were also seen in pts with no palpable spleen, a pt group not included in the COMFORT studies. Overall, the safety and efficacy profile of RUX in JUMP is consistent with that in the phase 3 COMFORT studies. Disclosures Palumbo: Novartis: Honoraria, Other: Advisory Board. Le Coutre:Novartis: Honoraria. Al-Ali:Celgene: Honoraria, Research Funding; Novartis: Consultancy, Honoraria, Research Funding. Ullrich:Novartis: Honoraria. Brittain:Astellas: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pierre Fabre: Honoraria, Membership on an entity's Board of Directors or advisory committees; Roche: Honoraria, Research Funding; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees. Foltz:Promedior: Research Funding; Gilead: Research Funding; Novartis: Honoraria, Research Funding. Raanani:Bristol-Myers Squibb: Other: Advisory Board; Novartis: Other: Advisory Board, Research Funding; Ariad: Other: Advisory Board; Pfizer: Other: Advisory Board. Gupta:Incyte: Honoraria, Research Funding; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees. Ghosh:Novartis Pharmaceuticals Corporation: Employment. Tannir:Novartis Pharma AG: Employment. Perez Ronco:Novartis Pharma AG: Employment. Vannucchi:Novartis: Other: Research Funding paid to institution (University of Florence), Research Funding; Shire: Speakers Bureau; Baxalta: Membership on an entity's Board of Directors or advisory committees; Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau.

Abstract: Abstract 1903 Poster Board I-926 Introduction: The discovery of the JAK2 V617F gene mutation has significantly altered the clinical diagnostic approach to the myeloproliferative neoplasms, reflected in the revised 2008 WHO diagnostic criteria. Both the 2001 and 2008 diagnostic criteria for essential thrombocythemia (ET) require a bone marrow biopsy showing megakaryocytic proliferation to make a diagnosis of ET. Published expert opinion based on clinical studies suggests that ET patients 〉 60 years old or with history of thrombosis should be characterized as high risk and treated with hydroxyurea. It is unclear whether physicians in clinical practice utilize the WHO diagnostic criteria or follow expert treatment recommendations for ET. Methods: We conducted a retrospective chart review of all patients with a clinical diagnosis of ET made by a hematologist who were seen in clinic from 2006 to 2008 at two university teaching hospitals in Vancouver, Canada. Data collected included demographic information, thrombosis history, diagnostic tests performed and treatment administered. Testing for JAK2 V617F became locally available in 2006, so for assessment of diagnostic tests performed, patients were divided into cohorts of diagnosis pre-2006 and 2006–2008. Patients were characterized as high risk if 〉 60 y or history of thrombosis at the time of diagnosis. All other patients were considered low risk. Results: Diagnostic information was available for 116 patients diagnosed prior to the availability of testing for JAK2 V617F. 65% (75/116) of patients in this cohort had a bone marrow biopsy performed (table 1). 44 patients received a new diagnosis of ET from 2006–2008. Only 48% (21/44) patients in this cohort had a bone marrow biopsy performed, significantly less than in the historical cohort (p = 0.019). 41/44 had JAK2 V617F testing performed: 41% (17/41) were JAK2 V617F negative, 56% (23/41) positive and 1 equivocal. Bone marrow biopsy was performed in 59% (10/17) of JAK2 V617F negative patients and 39% (9/23) of JAK2 V617F positive patients (p = 0.055) (table 1). Bone marrow biopsy was also performed in 1 patient with equivocal JAK2 V617F testing and 1 patient not tested for JAK2 V617F. 170 patients diagnosed with ET were seen in follow up 2006–2008. 64% (109/170) were high risk due to age 〉 60 y or history of thrombosis. The remaining 36% (61/170) were considered low risk. Hydroxyurea was used preferentially over anagrelide for treatment of ET (table 2). Only 76% of high risk patients were receiving cytoreductive treatment. 23% of low risk patients received cytoreductive treatment. ASA was prescribed to 89% of high risk and 79% of low risk patients. Conclusion: Despite the requirement for a bone marrow biopsy to meet the WHO criteria for ET, hematologists performed a bone marrow biopsy in less than half of patients they diagnosed with ET since 2006. Hematologists performed bone marrow biopsy less frequently after JAK2 V617F testing became available, particularly in JAK2 V617F positive patients. A substantial portion of high risk patients (24%) were receiving no cytoreductive therapy, contrary to expert recommendation. Further study is required to understand the barriers to implementing treatment recommendations in clinical practice. This study highlights the challenges in translating published diagnostic criteria and treatment guidelines into changes in patient care. Disclosures: No relevant conflicts of interest to declare.

Abstract: Introduction: JAK2V617F mutation is one of the major criteria in the diagnosis of myeloproliferative neoplasms (MPN). The disease phenotype and outcomes are dependent on variant allele fraction (VAF) of JAK2V617F. Recently, a new entity termed clonal hematopoiesis of indeterminate potential (CHIP) defines patients with normal cell counts and VAF of at least 2%. Outcomes of patients with 〈 2% VAF are scarce and we aimed to retrospectively study characteristics and outcomes of patients with JAK2V617F VAF 〈 2% compared to patients with VAF 2-10%. Methods: The study population included all patients in the province of British Columbia with JAK2V617F testing performed during 2010-2015. We compared the patient characteristics, disease phenotypes, overall survival (OS), thrombosis-free survival (TFS) and cumulative incidence of thrombotic events between patients with VAF 〈 2% and 2-10%. Parallel real-time quantitative polymerase chain reaction (RQ-PCR) for wild type JAK2 and JAK2V617F was used as detection method. MPN diagnoses were based on the treating physicians' assessment. Results: We identified 216 patients with JAK2V617F VAF 〈 10%. Twenty-seven patients were excluded due to missing follow-up data. A total of 189 patients were included for final analysis (89 patients with VAF 〈 2% and 100 patients with VAF 2-10%). Patient characteristics, diagnoses and outcomes are shown in the Table. Patients with JAK2V617F 〉 2% have significantly higher rate of splenomegaly, higher platelet counts and higher MPN diagnoses. Ten patients (10.0%) with VAF 2-10% had no hematologic diagnoses, consistent with CHIP, while 24 patients (27.0%) with VAF 〈 2% had no hematologic diagnoses. There were no differences in all outcomes measured including thrombotic complications, progression to hematologic or solid cancers and death. The median follow-up time for the whole cohort was 5.2 years with interquartile range (IQR) 3.5-6.6 years. The 5-year OS were 81.0% for VAF 〈 2% and 81.7% for VAF 2-10%, log-rank P = 0.922. TFS at 5 years were 71.2% and 69.5%, respectively (P = 0.982). The 5-year cumulative incidences of thrombotic complications (considering death as a competing event) were 8.8% and 11.3%, respectively (Pepe-Mori P = 0.574). Further analysis by clinical diagnoses classified patients into polycythemia vera (PV) 40 (21.2%), essential thrombocythemia (ET) 99 (52.3%), primary myelofibrosis or MPN, NOS (PMF/MPN) 16 (8.5%) and clonal hematopoiesis of indeterminate potential (CHIP) or no hematologic diagnosis 34 (18.0%). Patients with PMF/MPN were significantly older than patients with other diagnoses (median age PV 64.2, ET 64.3, PMF/MPN 80.7 and CHIP 54.1 years, P=0.019). The 5-year OS were: PV 91.4%, ET 90.0%, PMF/MPN 31.3% and CHIP/no hematologic diagnoses 58.7%, P 〈 0.001. TFS at 5 years were 83.1%, 74.7%, 25.0%and 57.4%, respectively, P 〈 0.001. Conclusion: Patients with JAK2V617F VAF 〈 2% have less splenomegaly and are less likely to have a diagnosis of MPN compared to patients with VAF 2-10%. However, the incidence of thrombotic events was similar between patients with VAF 〈 2% and 2-10%. In the combined VAF 〈 10% cohort, PMF/MPN patients were older and had the worst survival outcomes. The mortality in this PMF/MPN group was mostly unrelated to MPN diagnoses. Interestingly, patients with CHIP/no hematologic diagnoses in this study have the next worse OS and TFS. This could be explained by selection bias for performing JAK2 testing in acute or chronically ill patients with reactive changes in the peripheral blood. Table. Table. Disclosures Foltz: Gilead: Research Funding; Novartis: Consultancy, Honoraria, Research Funding; Promedior: Research Funding; Incyte: Research Funding.

Abstract: BACKGROUND: Ruxolitinib is a potent JAK1/JAK2 inhibitor that has proved superior to placebo and best available therapy in the phase 3 COMFORT studies for patients (pts) with intermediate (Int)-2- or high-risk myelofibrosis (MF). Ruxolitinib-treated pts demonstrated improvements in splenomegaly and MF-related symptoms as well as improved overall survival. JUMP is an expanded-access phase 3b trial designed to assess the safety and efficacy of ruxolitinib in pts with MF and includes those with no access to ruxolitinib outside of a clinical trial. Here, we report updated safety and efficacy findings for the full enrollment of JUMP, which includes 2233 pts in 26 countries. METHODS: Eligible pts had Int-2- or high-risk MF, with or without splenomegaly, or Int-1-risk MF with a palpable spleen (≥ 5 cm from the costal margin). Pts received starting doses of ruxolitinib based on platelet counts at baseline (5 mg twice daily [bid; ≥ 50 to 〈 100 × 109/L], 15 mg bid [100 to 200 × 109/L] , 20 mg bid [ 〉 200 × 109/L]). The primary endpoint was assessment of safety and tolerability of ruxolitinib. Additional analyses included changes in palpable spleen length and symptom scores. The final analysis will be performed after all pts have completed 24 months of treatment or discontinued the study. RESULTS: This analysis includes 2233 pts (primary MF, 59.4% [n = 1326]) who started treatment ≥ 1 year before the data cutoff date (01 Jan 2016). At baseline, median age was 67 y (range, 18-89 y); 54.5% were male; median palpable spleen length was 12 cm below the costal margin; median time since diagnosis was 25.8 months. Median hemoglobin (Hb) was 106 g/L, and 38.3% of pts had Hb levels ˂ 100 g/L; median platelet count was 254 × 109/L; mean FACT-Lym TS and FACIT-Fatigue score were 113.9 and 33.4, respectively. At data cutoff, 15.8% of pts (353/2233) remained on treatment and 45.1% (1006/2233) completed treatment per protocol (ie, transitioned to commercial ruxolitinib). The primary reasons for treatment discontinuation were adverse events (AEs; 17.7%), disease progression (8.6%), and death (4.1%). Median exposure was 12.4 months; the median average daily dose of ruxolitinib was 23.1 mg for pts starting at 15 mg bid (n = 647 [29.0%]) and 36.5 mg for pts starting at 20 mg bid (n = 1384 [61.9%] ). Overall 66.7% of pts had dose modifications and 26.3% had a dose interruption. Grade (G) 3/4 hematologic AEs included anemia (34.1%), thrombocytopenia (16.3%), and neutropenia (4.5%), which led to discontinuation in 1.5%, 2.7%, and 0.1% of pts, respectively. Nonhematologic AEs were primarily G1/2, and the most common (≥ 10%) were pyrexia (15.6%; G3/4, 2.3%), asthenia (14.9%; G3/4, 2.2%), and diarrhea (12.0%; G3/4, 1.1%). Rates of other G3/4 AEs were low (≤ 2%), except pneumonia (4.3%), which led to discontinuation in 10 pts (0.5%). Rates of infections were generally low; all-grade infections in ≥ 5% of pts included pneumonia (6.8%), urinary tract infection (5.6%), and nasopharyngitis (5.0%). Herpes zoster was reported in 4.6% of pts (G3/4, 0.5%), tuberculosis in 0.2% (G3/4, 0.04%) and hepatitis B in 1 pt (G3/4, 0.04%). At wk 24 and 48, 56.6% (874/1545) and 61.6% (658/1069) of pts with baseline splenomegaly achieved a ≥ 50% reduction from baseline in palpable spleen length; 23.3% (360/1545) and 18.9% (202/1069) had 25% to 50% reductions, respectively. Most pts (70.2% [1441/2054]) experienced a ≥ 50% reduction at any time; 23.7% had complete resolution of splenomegaly (Figure). At wk 24 and 48, 97.1% (67/69) and 92.3% (48/52) of evaluable pts without splenomegaly at baseline continued to have a nonpalpable spleen. A large proportion of pts achieved a response (ie, a clinically significant improvement) on the FACT-Lym TS and FACIT-Fatigue at wk 24 (42.4% [596/1406] ; 46.6% [675/1447]) and wk 48 (42.9% [404/941] ; 45.4% [434/957]). Overall, 221 patients received ESAs to manage anemia (G1, 7.2%; G2, 49.8%; G3, 38.9%; G4, 4.1%), and the majority had improved (32.6%) or resolved (31.7%) anemia (worsened, 16.3%; no change, 19.5%). CONCLUSIONS: This study includes the largest cohort of pts with MF treated with ruxolitinib to date. Consistent with findings from other studies, anemia and thrombocytopenia were the most common AEs but rarely led to discontinuation. As observed previously, most pts experienced reductions in splenomegaly and symptoms with ruxolitinib treatment. Overall, the safety and efficacy profile of ruxolitinib in JUMP is consistent with that in the COMFORT studies. Disclosures Foltz: Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Palumbo:Celgene: Speakers Bureau; Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Shire: Honoraria; Roche: Honoraria. Al-Ali:Novartis: Consultancy, Honoraria, Research Funding; Celgene: Consultancy, Honoraria, Research Funding. Le Coutre:Pfizer: Consultancy, Honoraria; BMS: Consultancy, Honoraria, Other: travel, accomodations, expenses; Ariad: Consultancy, Honoraria; Novartis: Consultancy, Honoraria, Other: travel, accomocations, expenses, Research Funding. Vannucchi:Novartis: Consultancy, Research Funding, Speakers Bureau; Baxalta: Speakers Bureau; Shire: Speakers Bureau. Gupta:Novartis: Consultancy, Honoraria, Research Funding; Incyte: Consultancy, Research Funding. Bouard:Novartis Pharma AG: Employment. Perez Ronco:Novartis Pharma AG: Employment. Khanna:Novartis Healthcare Pvt. Ltd: Employment. Zaritskey:Janssen: Consultancy; Novartis: Consultancy.

Abstract: The prognostic significance of CD38 expression and the cut off value has not been fully investigated. As CD38 is readily available test in patients with chronic lymphocytic leukemia (CLL), we investigated its role in prediction of disease progression when a cut off value of 20% is used. Progression free survival (PFS) was defined as the time from diagnosis to first treatment or last follow up. An electronic database search of pts with CLL who presented at St Paul’s Hospital between 1969 and 2007 was performed. Among 465 pts with CLL, 161 pts (35%) had their CD38 expression measured by flow cytometry. CD38 expression and its association with other prognostic factors such as age, Rai stage, lymphocyte count at diagnosis, gender and other immunophenotypic makers were analyzed. Out of 161 pts, positive CD38 expression ( & gt;20%) was found in 36 patients (22%). Comparing the baseline characteristics of the CLL pts with CD38+ and negative disease, we found CD38 positivity more common in male pts than in female pts (p=0.03). Also patients with CD38 positive disease tend to present with more advanced stage disease (p=0.056). Progression free survival at 2, 5 and 10y for the CD38+ CLL pts was 89%, 61% and 41% respectively compared with 95%, 81% and 62% for the CD38 negative group (p=0.03). Univariate analysis revealed the following factors as significant or marginally significant for disease progression: CD38+ (p=0.03), male gender (p=0.07), Rai stage (p & lt;0.0001), lymphocyte count above 20 ×109/l at diagnosis (p & lt;0.0001), CD5 expression & lt;10% (p=0.01). On multivariate analysis, only disease stage at diagnosis (p & lt;0.0001) and CD38 expression above 20% (p=0.04) retained significant and were predictive for disease progression. We conclude that CD38 expression above 20% at the time of diagnosis can be prognostically useful and predicts for disease progression and along with Rai staging can provide inexpensive tool to follow and monitor patients with CLL. Table: Characteristics of patients with CLL based on CD38 ≥20% and & lt;20% Parameter CD38 ≥20% (%) CD38 & lt;20% (%) p value* *for differences between the groups. Number 36 125 Sex: M/F (ratio) 22/14 (1.6:1) 66/60 (1.1:1) 0.03 Age above 60 y 21 (58) 81 (65) 0.2 Rai stage: 0, 1+2, 3+4 20, 11, 5 (55, 31, 14) 100, 17, 1 (80, 14, 0) 0.056 Lymphocyte count above 20×109/l 9 (25) 23 (18) 0.3 CD5 & lt;10% 6 (16) 29 (23) 0.3 Fig: Progression free survival for patients with CLL based on CD38 expression, cut off level 20%. Fig:. Progression free survival for patients with CLL based on CD38 expression, cut off level 20%.

Abstract: Richter transformation (RT) is a rare complication of chronic lymphocytic leukemia (CLL). There is little information in the literature about its risk and outcome. In this study we assessed the incidence, presenting characteristics and outcomes of patients (pts) with CLL who developed RT. An electronic database search of pts with CLL who presented at St Paul’s Hospital between 1969 and 2007 was performed. Among 465 pts with CLL, 24 pts (5%) developed RT. Presenting features included B-symptoms (17%), lymph node enlargement (58%), progressive cytopenia (29%), hypercalcemia (4%), and spleno/hematomegaly (13%). The median age at diagnosis of CLL and RT were 64 y (range 33–80 y) and 67 y (range 48–81 y) respectively. The median time to transformation from CLL diagnosis was 61 months (range 1–257 m). Twenty one patients (88%) had been previously treated for CLL. Seventeen patients (71%) had received & gt;1 prior therapy. The median lymphocyte count at diagnosis was 12 ×109/L (range 4–120 ×109/L). Six patients (25%) are still alive with a median follow up of 38 m (range: 3–66 m). The only predictive factor for better survival post-transformation on univariate analysis was age of less than 60 y at CLL diagnosis (p=0.01). Other factors such as CLL Rai stage, lymphocyte count at diagnosis were not predictive for survival. This group of patients was compared with randomly selected group of patients with CLL but did not have RT. The baseline characteristics of the groups are presented in the table. No significant differences were found between the two groups in terms of gender, age at diagnosis, Rai stage or median lymphocyte count at diagnosis. The 5 and 10 year OS for the RT group were 76% and 39% compared to 93% and 84% for the CLL group (p= 0.002), respectively. In summary, RT significantly shortens the survival of CLL patients. There were no obvious predictive factors for RT in CLL pts at diagnosis. Table: Baseline characteristics of Richter and CLL groups. Parameter Richter group (%) CLL group (%) P value Number 24 37 Sex: M/F (ratio) 16/8 (2:1) 22/15 (1.5:1) 0.052 Age at Diagnosis: Median (range) 64 (33–80) 60 (37–85) 0.6 Rai stage at diagnosis: 0, 1+2, 3+4 7, 16,1 (29, 67, 4) 23, 14, 0 (62, 38, 0) 0.3 Median lymphocyte count at diagnosis (range) 12 ×109/1 (4–120) 8 × 109/1 (5–394) 0.056 Fig: OS of pts with RT compared with CLL pts and no RT Fig:. OS of pts with RT compared with CLL pts and no RT