MDS-Specific Comorbidity Index (MDS-CI) Identifies Overall Survival Differences in Myelodysplastic Syndrome Patients,

Abstract 3793

MDS-specific comorbidity index (MDS-CI) is a score reported by the Pavia group for myelodysplastic syndromes (MDS) patients. This score is a time-dependent index developed for predicting the effect of comorbidities on outcome of these subjects. We applied this score on a total of 450 MDS patients: comorbidities were recorded at the time of diagnosis and considered by medical staff for the analysis. All patients were consecutively diagnosed and followed at our institute in a period between January 1992 and December 2006. Statistical analysis was carried out using SPSS software; survival was defined as time from diagnosis to last contact or death for any cause. Median age of the whole population was 69 years (range 21–88), with a prevalence of male sex (ratio m/f 1.6). Overall, we found the presence of one or more comorbidities in 94% of the examined patients. The most common comorbidities were cardiac disorders observed in 40% of patients, followed by diabetes with organ damage recorded in 22 cases. Application of MDS-CI score identified 300 patients with score 0, 55 patients with score 1, 80 patients with score 2 and 15 patients with score >2. We found significant differences in OS according to MDS-CI stratification: from 38 months for low risk patients (score 0) to 22 months for high-risk patients (score >2, p=0.02). We then evaluated the prognostic effect of MDS-CI on patients stratified according to WPSS prognostic index. WPSS application was possible in 330 of 450 patients who entered the analysis (73%), due to cytogenetic availability of data: we identified 112 patients (34%) with very low/low risk, 137 patients (41.5%) with intermediate risk and 81 patients (24.5%) with high/very high risk. As reported by the Pavia group, we assessed prognostic relevance of comorbidities in very low/low risk WPSS patients, intermediate and high/very high-risk patients. We found in the first category a significant difference in OS stratification: from 48.5 months for patients with score 0 to 20.4 months for patients with score >2 (p=0.002). In WPSS intermediate risk we found similar significant OS difference: from 32.3 months for patients with score 0 to 18.3 months for patients with score 2 (p=0.001). Conversely, we did not find significant differences in WPSS high/very high-risk patients. We also found significant correlations between presence of comorbidities at baseline, as stratified with MDS-CI, and risk of non-leukemic death: from 22% in patients without comorbidities to 75% in patients with score >2 according to MDS-CI (p=0.002). Our analysis showed the efficacy of MDS-CI on a large series of MDS consecutively seen, diagnosed and followed in a single center while strengthening the results reported by the Pavia group: in fact, the value of comorbidities was confirmed in a cohort of patients similar to that observed in the original paper, in terms of median age (69 years in our series vs 66 years in the Pavia series), but with higher percentage of patients with intermediate WPSS risk (41.5% in our series vs 18% in the Pavia population). MDS-CI is a very valid tool capable to differentiate MDS patients with very low/low and intermediate WPSS risk in terms of OS and non-leukemic death risk.