Validation of New Prognostic Model Including Comorbidities in Patients with Myleodysplastic Syndrome Receiving Hypomethylating Therapy

A new prognostic model including comorbidities, which are assessed using the Adult Comorbidity Evaluation-27 (ACE-27), was proposed for the patients with MDS by researchers from MD Anderson Cancer Center, Houston, TX (J Clin Oncol 2011;29: 2240). The model was developed after analysis of 600 patients who presented to the center. We aimed to validate the new prognostic model for the Korean patients with MDS who were treated with hypomethylating agents.

Between September 2006 and October 2010, 149 patients were treated with either azacitidine or decitabine for MDS defined by the WHO classification and chronic myelomonocytic leukemia (CMML) in 3 Korean institutes. The new prognostic model included age (> 65 years, 2 score points), comorbidity assessed by ACE-27 (mild/moderate, 1 point; severe, 3 points), and IPSS (intermediate-2, 2 points; high, 3 points). Patients were divided into one of 3 risk groups: low (score 0–1), intermediate (scores 2–4), and high (scored 5–8). Risk group by the new prognostic model could not be assigned in 10 patients, thus a total of 139 patients were included in this analysis. Azacitidine 75 mg/m²/day was administered as a subcutaneous injection for 7 consecutive days (n=68) and decitabine 20 mg/m²/day as a 1-hour intravenous infusion for 5 consecutive days (n=71). Both agents were repeated every 4 weeks. Clinico-patholoical data including comorbidities were collected at time of hypomethylating therapy. Treatment response was evaluated using modified International Working Group response criteria.

Median age was 61 years (range, 23–83); 47 patients were over 65 years old. Overall comorbidity score assessed by ACE-27 was as follows: none (n=65), mild (n=53), moderate (n=15), and severe (n=6). IPSS risk category was low/intermediate-1 in 72, intermediate-2 in 55, and high in 12. Risk group measured by the prognostic model was low in 42 (30.2%), intermediate in 79 (56.8%), and high in 18 (12.9%). Hypomethylating therapy induced hematologic responses (CR/PR/mCR) in 28 patients (20.1%) and rate for overall responses (CR/CR/mCR/HI) was 57.6% (80/139). The treatment responses were not significantly different between 3 risk groups. At a median follow-up time of 27.0 months (range, 3.3–55.5) among surviving patients, 70 patients died and overall survival (OS) probability was 49.5% at 2 years; median OS was 24.1 months (95% CI, 11.7–36.5). OS was significantly different according to the presence of comorbidities (OS at 2 years, none vs. mild/moderate vs. severe, 63.4% vs. 37.5% vs. 33.3%, P=0.006) or risk groups by the prognostic model (OS at 2 years, low vs. intermediate vs. high, 71.1% vs. 46.1% vs. 18.5%, P<0.001). The survival differences by the prognostic model were maintained after adjustment for other variables (low vs. intermediate, HR, 2.810, 95% CI, 1.444–5.468, P=0.002; low vs. high, HR, 6.037, 95% CI, 2.708–13.459, P<0.001).

The new prognostic model including comorbidities assessed by ACE-27 was useful to predict overall survival in patients with MDS receiving azacitidine or decitabine, although the model could not predict response to the hypomethylating agents.

No relevant conflicts of interest to declare.