Abstract
Cytogenetics (CG) represent one of the most important factors in determining survival for pts with MDS, being therefore required by the IPSS to predict life-expectancy. However, the impact of CG on outcome of pts undergoing allo-HSCT for MDS is unknown. The aim of this EBMT-CLWP study was to carry out a retrospective analysis of the impact of CG on overall survival (OS), relapse-free survival (RFS), relapse probability (REL), and transplant related mortality (TRM) in pts with MDS/sAML undergoing allo-HSCT from HLA-identical siblings.
Data from 1506 pts with MDS/sAML who underwent a first allograft from HLA-identical siblings from 1984 to 2004 reported to EBMT were assessed. The following covariates were included: CG (good- vs standard- vs high-risk according to the IPSS); stage at HSCT (untreated vs treated in CR vs treated not in CR); FAB classification (RA/RARS vs RAEB/CMML vs RAEB-t/sAML); age (as a continuous covariate in the COX models); time from dx to HSCT (<5 vs 5 to 8 vs > 8 months); calendar year in which HSCT was performed; type of conditioning (standard vs RIC); source of stem cells (BM vs PB). A complete information dataset was available in 692 pts, who are the subject of this analysis.
20% of pts had RA or RARS, 28% had RAEB or CMML, and 52% had RAEB-t or sAML. Age was ≥50 years in 30%. CG classified 55% of pts as good-, 24% as intermediate-, and 21% as high-risk. At the time of HSCT, 38% were untreated; among treated pts, 222 (52%) were in CR. A RIC regimen was administered to 93 pts (13%). Source of stem cell was PB in 38%. By univariate analysis, subdivision of pts in the IPSS risk-categories for CG associated with significantly different OS (at 60 mos, alive pts in good-, interm.-, and high-risk groups were 47%, 40% and 31%, respectively), RFS (alive pts at 60 mos 40%, 35% and 21%, respectively), relapse probability (34%, 35% and 57% at 60 mos, respectively) and TRM (33%, 42% and 46% at 60 mos, respectively).
By multivariate COX analysis, age and CG associated with all the outcome variables (for OS, high- vs good-risk HR 1.4; interm.- vs good-risk HR 1.2; p<0.02). OS, RFS and REL were also determined by FAB category (for OS, RAEB-t/sAML vs RA/RARS HR 1.5, p<0.01) as well as by stage at HSCT (for OS, CR vs untreated HR 0.65, p<0.005). Conditioning intensity associated both with TRM (reduced vs standard HR 0.5, p<0.01) and REL (reduced vs standard HR 1.63, p<0.02). Concerning REL, the interval between dx and HSCT was also significant, whereas the source of stem cells associated only with OS (PB vs BM HR 0.78, p<0.05). In summary, this study provides evidence that CG have strong prognostic impact on outcome of pts undergoing allo-HSCT from HLA-identical siblings for MDS/sAML and should be taken into consideration when selecting candidates for this treatment strategy. Detailed analyses will be presented.
Disclosure: No relevant conflicts of interest to declare.
Author notes
Corresponding author
