Comparison of Two MDS Prognostic Scoring Systems in Patients Undergoing Allogeneic Transplant

Background: To retrospectively compare the predictive value for overall survival (OS) of the IPSS (Blood 1997;89:2079-2088) to the system published by Schaffer et al. (S3) (JCO 2016;34(16):1864-1871) in patients with MDS who underwent allogeneic PBSCT at our institution between June 2010 and March 2017. Methods: We identified 48 patients with MDS who underwent allogeneic PBSCT. They received myeloablative conditioning (FBT, n = 13) with fludarabine 50 mg/m2/day d−6 to d−2, busulfan 3.2 mg/kg/day d−5 to d−2 and TBI 200 cGy/day on d−1 and d0 or reduced intensity conditioning (FB2, n = 35), fludarabine 30 mg/m2/day d−6 to d−2, busulfan 3.2 mg/kg/day d−3 and d−2. We compared the OS at three years as predicted by the IPSS with the S3. One point was accrued for each of the following criteria used in the S3: blood blasts >3%, platelets ≤50 × 109/L at transplantation, age 30 to 49 years, cytogenetic risk score of poor or very poor, and KPS < 90%. Monosomal karyotype and age ≥50 years were assigned 2 points each. Results: The published 3-year overall survival from transplantation using S3 was 71% in low risk patients (0 to 1 points), 49% intermediate (2 to 3), 41% high (4 to 5) and 25% very high (>6). Using S3 to stratify our patients yielded OS of 56% (intermediate, n = 9), 32% (high, n = 28) and 36% (very high, n = 11). No patients met criteria for low risk with S3. Applying the IPSS resulted in survival of 67% (Low, n = 3), 32% (Int-1, n = 19), 32% (Int-2, n = 19), and 43% (High, n = 7). The authors of S3 suggested that the HCT comorbidity score might add prognostic power. We modified the S3 values by adding 1 point for CMS > 2. The OS at 3 years was 83% (S3Int, n = 6), 31% (S3High, n = 26) and 31% (S3Very High, n = 16). Conclusion: Given the limitation of low numbers, no scoring system appeared superior in predicting survival. The addition of one point for a CMS > 2 to the S3 might identify patients with a better prognosis in the intermediate risk category. Prospective confirmation of this finding in larger cohort is planned.Tabled 1S3 (published)S3S3 + CMSIPSSLow71% (n = 47)NA (n = 0)NA (n = 0)67% (n = 3)LowIntermediate49%(n = 258)56% (n = 9)83% (n = 6)32% (n = 19)Int-1High41%(n = 104)32% (n = 28)31% (n = 26)32% (n = 19)Int-2Very High25%(n = 18)36% (n = 11)31% (n = 16)43% (n = 7)HighS3 published validation cohort (3-year OS published by Schaffer), S3 (applied to our patients), S3 + CMS (applied to our patients + 1 point for CMS > 2), IPSS (applied to our patients). Open table in a new tab S3 published validation cohort (3-year OS published by Schaffer), S3 (applied to our patients), S3 + CMS (applied to our patients + 1 point for CMS > 2), IPSS (applied to our patients).