Prognostic Factors and Response to the Hypomethylating Agents in Hypocellular Myelodysplastic Syndromes: A Retrospective Study from the Korean Society of Hematology AML/MDS Working Party

Introduction: Hypocellular myelodysplastic syndrome (h-MDS) comprised approximately 10-20% of total MDS cases. The difference in clinical features and prognosis between h-MDS and normo/hypercellular MDS (NH-MDS) remains uncertain and the treatment strategy of h-MDS patients has not been well-defined in epigenetic therapy era. Therefore, we conducted this study to define the prognostic factors and response to the hypomethylating agents (HMA) for Korean h-MDS patients.

Methods: We collected clinical and laboratory data from web-based Korean MDS registry (www.mdsregistry.or.kr). A total of 1976 evaluable MDS patients, who were diagnosed between the years 1993 and 2014, form 23 Korean hospitals were registered in Korean MDS Registry. Hypocellularity was defined as less than 30% of cellularity in patients <70 years, and <20% in patients 70 years or older in the bone marrow biopsy specimens.

Results: We identified 358 (18.1%) h-MDS patients among 1976 evaluable MDS patients. Gender, performance status, Hemoglobin, platelet counts, prior transfusion and bone marrow blast percentage did not differ between h-MDS and NH-MDS patients. However, h-MDS patients were younger (median 59.8 years vs 63.4 years, p=0.002), more neutropenic (median ANC 0.9x10⁹/L vs. 1.14x10⁹/L, p=0.000), with lower level of LDH (median 357 unit/L vs. 386 unit/L, p=0.041).There was no difference in overall survival (OS) between the h-MDS and the NH-MDS patients (110.9 months vs. 71.8 months, P=0.174) and 9% of h-MDS patients and 15.1% NH-MDS patients transformed to acute leukemia (p=0.075).

The IPSS and IPSS-R were able to identify different risk groups within each group and were not significantly different between the groups. Based on IPSS-R, the overall survival for h-MDS patients were very low, low, intermediate, high and very high in 115.4, 110.9, Not reached, 30.5 and 29.8 months, respectively. The IPSS-R was a lesser ability to predict OS for h-MDS patients. Of the 358 h-MDS patients, 141 h-MDS patients were treated with HMA (60.3% azacitidine and 39.7% decitabine). The median OS after the treatment of HMA was 31.4 months and 32 months for h-MDS and NH-MDS patients (p=0.909) and leukemic transformation rate at 3 years was 24.1% and 27.6%, respectively (p=0.224).

At univariate analysis to define prognostic factors in h-MDS, the factors associated with a poor overall survival were age (≥70), poor performance status, low albumin (<4.0g/dL), anemia (<10g/dL), neutropenia (<0.8x10⁹/L), prior transfusion and high LDH (≥600IU/L). Using the Cox proportional hazard regression model, older age, neutropenia, low albumin and prior transfusion were independent prognostic indicators. Based on those independent factors, h-MDS patients were divided into 3 risk groups and it discriminated 3 discreet groups with distinct survival rates (median OS, 150.2 vs. 46.9 vs.21.5 months, p<0.0001).

Conclusions: h-MDS had similar clinical features and survival compared to NH-MDS and response to HMA treatment was not different. However, IPSS-R was less powerful in detecting difference for h-MDS and h-MDS patients had a different pattern from those of NH-MDS regarding prognostic factors. Further study and validation are warranted to stratify the risk of h-MDS patients.