S840 impact of 1st line treatment on overall survival in 1.633 patients with chronic myelomonocytic leukemia – a multinational collaborative effort coordinated by the agmt study group

Background: Chronic myelomonocytic leukemia (CMML) is a rare myeloid stem cell disorder, defined by the presence of absolute monocytosis (≥1.0x10 9 /μl), with dysplastic (MD) and proliferative (MP) features (Orazi A, IARC Press Lyon 2017, pp82–86). The only curative treatment option is allogeneic stem cell transplantation (allo‐SCT), from which most patients are excluded (Symeonidis A, BJH 2015, pp239–246). Other options including hypomethylating agents (HMA), are not curative. Only one matched‐pair analyses (Pleyer L, Leuk Res 2014, pp475–843) and one randomized clinical trial exclusively recruiting CMML patients (Wattel E, Blood 1996, pp2480–2487) have been published, with another one being underway (NCT02214407). We do not currently know whether active treatments provide overall survival (OS) benefit compared to best supportive care (BSC). Aims: To evaluate the impact of 1 st line and cumulative treatment on OS. Methods: Data was collected from 7 European study groups and 2 US MDS Centers of Excellence. Data from 1633 patients was received; 67 patients were excluded (missing data); data cut‐off 10.02.19; ML and LP performed data cleaning. Assign Data Management and Statistics GmbH performed statistical analyses with SAS 9.3. Results: Median age (range) was 73 (18–104) years, male (68%), splenomegaly (25%), B‐symptoms (17%), elevated serum LDH (55%), ECOG PS 0–1 (49%), t‐CMML (7%), MP‐CMML (49%), transfusion dependent (TD) (33%), IPSS cytogenetic good risk (57%), CMML‐0 (39%), CMML‐1 (29%), CMML‐2 (19%); 33% were diagnosed in the pre‐HMA era; 71% had died at data cut‐off. In total, 42, 32, 16 and 10% of patients received 0, 1, 2, or ≥3 treatment lines, respectively; 24, 23, 6, 5 and <1% received hydroxyurea (HU), HMA, low‐dose AraC (LDAC)/ImiDs/other drugs, intensive chemotherapy (IC) or allo‐SCT as 1 st line treatment, respectively. Patients receiving BSC only had significantly less adverse risk factors, as compared to patients receiving treatment: Splenomegaly (8 vs 38%, p < 0.001), B‐symptoms (4 vs 27%, p < 0.001), elevated LDH (30 vs 70%, p < 0.001), MP‐CMML (31 vs 56%, p < 0.001), ECOG PS 3 2 (11 vs 16%, p = 0.008), RBC‐TD (12 vs 45%, p < 0.001), PLT‐TD (3 vs 12%, p < 0.001), poor IPSS cytogenetic risk (4 vs 9%, p < 0.001), trisomy 8 (0 vs 11%, p < 0.001), complex karyotype 0 vs 5%, p < 0.001), CMML‐2 (14 vs 22%, p < 0.001). Median OS [95% CI] was 27.9 [25.6, 26.9] months as of initial diagnosis for the total cohort, and 18.9 [17.1, 20.8] months for all patients with 1 st line treatment, respectively. The impact of 1 st line treatment, number of treatment lines, FAB subtype and initial diagnosis timepoint on OS are summarized in Fig1. Median OS [95% CI] for patients receiving BSC only was 29.3 [24.7, 34.5] months, and median OS from diagnosis (treatment start) was 24.1 (17.6), 32.2 (21.1), 31.8 (26.5), 19.1 (11.5) and 42.0 (38.2) months for patients receiving HU, HMA, others, IC or Allo‐SCT as 1 st line treatment, respectively. Patients with MP‐CMML had significantly worse OS as compared to MD‐CMML (median OS 18.7 vs 37.0 (p < 0.001) and 16.1 vs 24.5 months (p < 0.001) from diagnosis and treatment start, respectively). Patients diagnosed after approval of HMA had significantly longer OS from diagnosis as compared to those diagnosed in the pre‐HMA era (29.9 vs 24.5 months, p < 0.001). Summary/Conclusion: Allo‐SCT as 1 st line treatment remains the most effective treatment approach. Patients treated with HU or IC 1 st line seem to fare worst. Multiple therapy lines are not indicative of poor outcome but may help maintain OS. Biologic selection may play a role. image