Abstract
6521
Background: MDS are hematologic neoplasms characterized by dysplasia and cytopenias. Two commonly used risk stratifications, IPSS (Greenberg et al Blood 1997) and IPSS-R are based on clinical factors (marrow blasts, number of cytopenias and genetic abnormalities determined by conventional karyotyping), with the IPSS based on 4 cytogenetic risk groups and the IPSS-R based on 5 cytogenetic risk groups defined by Schanz et al JCO 2012. A drawback to metaphase cytogenetics (MC) is its inability to detect small cryptic chromosomal defects and uniparental disomy (UPD). SNP-A can identify these small cytogenetic lesions/UPD and has been shown to be of prognostic value in MDS. The goal of this investigation was to determine if SNP-A detected defects could be used to refine the IPSS and IPSS-R. Methods: SNP-A karyotyping was performed as previously described (Tiu et al, Blood, 2011). We reviewed data from 327 patients idenfitying SNP-A results, MC, and the IPSS/IPSS-R-related clinical factors were identified. Overall survival (OS) measured from the date of sampling for SNP-A karyotyping was the primary endpoint. Data were analyzed using Cox proportional hazards models. Results: SNP-A lesions not detected by MC were grouped as 1) none or only gains/losses; 2) UPD only or gains+losses but no UPD; and 3) UPD+other defects. The frequencies/median os in months (mo) of the groups were: 66%/20.0, 23%/12.7 and 11%/5.8, respectively, p<.0001 for OS. Combining the SNP-A groups with MC resulted in revised definitions of cytogenetic risk that had better discrimination than the underlying models; and when combined with the relevant clinical factors resulted in refined IPSS (IPSS
SNP-A
) and IPSS-R (IPSS-R
SNP-A
)
risk stratifications (Table). Conclusions: Detection of chromosomal gains, losses, and UPD by SNP-A has prognostic value in MDS and can be used to refine current risk stratifications. [Table: see text]