Implications of Using the Revised International Prognostic Scoring System (IPSS-R) for Patients with Therapy-Related Neoplasms
Chi Young Ok, Zhuang Zuo, Guillermo Garcia-Manero, L Jeffrey Medeiros, Sa A Wang. University of Texas MD Anderson Cancer Center, Houston, TX
Background: Therapy-related myeloid neoplasms (t-MN) include cases of MDS and AML and these neoplasms are clinically more aggressive than their de novo counterparts. A risk model has not been established specifically for patients with t-MN. We studied patients with t-MDS who are often assigned a risk score according to the International Prognostic Scoring System (IPSS) for de novo MDS patients. The IPSS has been revised recently and risk categories are designated as very low, low, intermediate, poor, and very poor. The distribution for each category was 19%, 38%, 20%, 13%, 10%, and survival was 105.6, 63.6, 36.0, 19.2, and 9.6 months respectively. In this study, we applied the revised IPSS (IPSS-R) system to patients with t-MN.
Design: We searched all cases of t-MDS from 2000 to 2011. In order to apply the same criteria used in the IPSS-R, we included t-MN cases with 20-30% blasts. For each case we collected patient demographic data, complete blood counts (CBC), bone marrow blast percentage and karyotype. Cytogenetic risk stratification was based on the New Comprehensive Cytogenetic Scoring System adopted by IPSS-R. MLL gene rearrangement, infrequent in de novo MDS, but associated with Topo-II inhibitor therapy in t-MDS, was studied as a variant.
Results: A total of 451 patients were identified. These included 417 patients with <20% blasts and 34 with blasts 20-30%. There were 255 men and 196 women (1.3 to 1) with a median age of 65 years (range, 20-92). 432 (95.8%) cases had complete information. The IPSS-R in t-MDS patients were distributed as follows; very low 19 (4.4%), low 69 (16.0%), intermediate 93 (21.5%), poor 108 (25.0%) and very poor 143 (33.1%) with disease specific survival (DSS) of 36.9, 23.5, 19.5, 11.9, 8.8 months (p<0.001), respectively. In multivariate analysis, all components of the IPSS-R, except the absolute neutrophil count, were independent hazards for DSS. Additionally, MLL gene rearrangement independently predicted an inferior outcome (HR 3.16, p=0.03).
Conclusions: The IPSS-R can stratify t-MN patients into different risk groups. However, in the therapy-related setting, the distribution is skewed toward the higher risk categories (p<0.001). More importantly, these IPSS-R scores do not indicate a survival comparable to de novo MDS patients in the same categories, with the exception being the very poor risk category. MLL gene rearrangement is an independent hazard that is not captured by the IPSS-R score.
Monday, March 4, 2013 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 226, Monday Morning