P1323: PREDICTION OF NON-RELAPSE MORTALITY BY EASIX AND HCT-CI SCORES IN PATIENTS WITH CHRONIC MYELOMONOCYTIC LEUKEMIA UNDERGOING ALLOGENEIC STEM CELL TRANSPLANT

Abstract

Background: The endothelial activation and stress index (EASIX) score has been shown as a predictor of non-relapse morality (NRM) after allogeneic stem cell transplantation (alloSCT) [Luft, et al. Bone Marrow Transplant. 2020;55(3):553–561]. However, data in patients with chronic myelomonocytic leukemia (CMML) is limited. Aims: Our goal was to study the risk of NRM and the impact on overall survival (OS) by EASIX and HCT-CI scores in patients with CMML undergoing alloSCT. Methods: We retrospectively evaluated patients with CMML who underwent alloSCT. We reviewed patient demographics, disease characteristics, and post-transplant outcomes. EASIX score was calculated using the formula: LDH (U/L) × Creatinine (mg/dL) / platelet count (10^9/L) and analyzed based on log2 transformed values. EASIX score was calculated within 45 days of alloSCT, but prior to starting conditioning therapy. The optimal log2 EASIX score cutoff was determined using maximally selected log-rank statistics. Patients with an HCT-CI score ≥ 3 were considered to have a high HCT-CI score. Results: Between, January 1992 to February 2019, a total of 70 CMML patients underwent alloSCT, 13 patients were excluded due to incomplete records and 57 patients (39 males and 18 females) were evaluated for the purposes of the study. CMML-0 was the most common subtype (21 patients, 36.8%), followed by CMML-2 (19 patients, 33.3%)) and CMML-1 (17 patients, 29.8%). Seventeen (29.8%) patients had progressed to CMML blast phase before alloSCT. Of the entire cohort, 18 (31.6%) patients were in complete remission at the time of alloSCT. Median age at transplant for the entire cohort was 58 years (range: 18 – 73 years). Median time of follow-up was 2 years (range: 11 days – 16 years). Median OS was 2 years, with a 3-year OS rate of 49.1%. Eighteen (31.6%) patients were alive at last follow-up. The median log2 EASIX score was 0.88 (range -2.29 to 5.91). The optimal log2 EASIX score cutoff using the maximally selected LogRank statistics was found to be 2.33. Patients with log2 EASIX score > 2.33 were considered to have a high EASIX score. Thirteen (22.8%) patients had a high EASIX score and 44 (77.2%) had a low EASIX score. The cumulative incidence of NRM among patients with high EASIX score was significantly higher compared to those with low EASIX score at 100 days (30.8% vs 0%, p < 0.001), at 1-year (46.2% vs 9.1%, p = 0.002) and at 3-years (53.8% vs 15.9%, p = 0.003) after alloSCT (Figure 1A). A total of 55 patients were evaluated for HCT-CI score, of whom 30 (54.54%) were found to have a low HCT-CI score, i.e., HCT-CI < 3, and 25 (45.45%) were found to have a high HCT-CI score. The median HCT-CI score was 2 (range: 0-7). NRM after alloSCT among patients with high HCT-CI score was not significantly different compared to those with low HCT-CI scores at 100 days (8% vs 6.7%, p = 0.85), at 1 year (24% vs 13.3%, p = 0.33), and at 3 years (24% vs 26.7%, p = 0.92) after alloSCT (Figure 1B). In multivariate competing risk regression analysis, a high EASIX score significantly increased NRM at 3 year (HR 4.14, 95% CI: 1.42 – 12.01, p = 0.009). Patients with a high log2 EASIX score had significantly worse OS at 3 years (median 8 months vs. NA, p = 0.007). OS at 3 years did not differ among patients with high vs. low HCT-CI scores (median 1.02 years vs. NA, p = 0.38). Image:Summary/Conclusion: In CMML, EASIX predicted NRM following alloSCT. EASIX was a better predictor of NRM compared to HCT-CI in this patient population. Larger patient cohort is required to confirm our findings.