Cytogenetic Features and Prognosis in Argentinean Patients with Myelodysplastic Syndrome: a Multicenter Study

BELLI, CAROLINA; BENGIÓ, RAQUEL; NEGRI ARANGUREN, PEDRO; SAKAMOTO, FRANCICO; FLORES, GABRIELA; WATMAN, NORA; NUCIFORA, ELSA; PRATES, M VIRGINIA; ARBELBIDE, JORGE; LARRIPA, IRENE

Orlando

Congreso; 52nd American Society of Hematology Annual Meeting and Exposition; 2010

Sociedad Americana de Hematología

Myelodysplastic Syndrome (MDS) comprises a group of heterogeneous hematological disorders with variable risk of leukemic evolution (LE) and short survival (SV). Around 40-50% of patients show abnormal karyotype at diagnosis and cytogenetic findings are an independent prognostic factor in MDS. Although the International Prognostic Scoring System (IPSS) differentiated 3 cytogenetic categories of risk (CCR), the Intermediate one is heterogeneous. The aim of this study was to characterize the cytogenetic profile, to test its prognostic value and to evaluate cytogenetic groups of risk in the Argentinean MDS population. Also, we tried to ascertain whether some abnormalities could be segregated from their respective CCR. This is a multicenter retrospective analysis of 488 primary Argentinean patients with MDS evaluated from 1984 to 2008 (including 183 patients from the Pilot Study for MDS Registry organized by the Argentinean Society of Hematology). Patients’ distribution according to French-American-British classification (FAB) was: 235 Refractory Anemia (RA), 50 RA with Ringed Sideroblasts (RARS), 121 RA with excess of Blast (RAEB), 27 RAEB in transformation (RAEBt) and 55 Chronic Myelomonocytic Leukemia (CMML). The median age was 69 (17-92) years with a gender ratio (M/F) of 1.3. During the follow-up (mean: 25 months (m), range: 1-266 m), 110 (22.5%) underwent LE and 217 (44.5%) patients died. Age, sex, percentage of bone marrow blast, hemoglobin level, platelets count, number of cytopenias, LDH level and red blood cell transfusion requirements were significant predictive variables for prognosis (Kaplan-Meier and Long-Rank test, p<0.05). FAB and World Health Organization (WHO) classifications and scoring systems (Lille, Lausanne-Bournemouth, IPSS, GCECGH and WPSS) allowed us to differentiate groups of risk for SV and LE. Cytogenetic results were available in 421 patients and 176 (42%) showed abnormal karyotype. Cytogenetic profile showed that all chromosomes were involved and different cytogenetic alterations were found (total or partial chromosome losses were predominant). The most common cytogenetic aberrations were: -5/5q- (20% among cases with abnormal karyotype), -7/7q- (16%), +8 (20%), 20q- (9%) and –Y (8%). No particular aberration was associated to any FAB subtype though the frequency of abnormal karyotypes increased from 36% for RA, 39% RARS, 50% RAEB to 74% RAEBt and 39% for LMMC. Karyotypes were further divided according to IPSS CCR into 68% Good, 21% Intermediate and 12% Poor with median SV of 48, 34 and 17m and a LE cumulative risk to 1-year: 13%, 25% and 38%, respectively, p<0.0001. CCR were also predictive in the WHO classified population (p<0.0001 for SV and p=0.0021 for LE). Patients with normal karyotype had better outcome than those with cytogenetic alterations (median SV of 51 vs. 21 m, p=0.0012, and LE cumulative risk to 1-year: 13% vs. 26%, p=0.0047). When we tried to ascertain whether some alterations could be segregated from their respective Good and Poor CCR, no significant differences were observed both for SV and for LE. However, the outcome of the Intermediate CCR was heterogeneous ranging between patients with 12p- (median SV: 65 m and a LE cumulative risk to 1-year: 0%) and those with rearrangements 3q/ del(17p)/ +19/ t(11)(q23) (median SV: 15 m and a LE cumulative risk to 1-year: 48%, p=0.0220). Cytogenetic findings had a clear impact on SV and LE in our population and results in the present series, the largest in Latin America, are coincident with published data. However, the wide spectrum of low frequency aberrations stresses the importance of large study groups where the impact of such aberrations could be statistically evaluated to properly segregate them from their original CCR.