Chromosomal 3 Abnormalities in Myelodysplastic Syndromes: Implications on Prognosis and Response to Treatment

Background: Cytogenetic abnormalities are the most powerful co-variates in myelodysplastic syndromes (MDS) prognostic models. In the revised international prognostic scoring system (R-IPSS), abnormalities of chromosome 3 (i.e. inv(3), t(3q), del(3q)) represent a poor risk karyotype and this derivation alone translates to higher risk disease for risk of acute myeloid leukemia (AML) transformation and decreased overall survival (OS). In the recent World Health Organization (WHO) MDS pathology classification meeting, a proposal was entertained to classify MDS with chromosome 3 abnormalities as a separate entity. There is only minimal data to support such classification and chromosome 3 karyotype data regarding response to hypomethylating agents (HMA) or allogeneic transplantation is limited. In response to this proposal, we examined outcomes of MDS patients and chromosome 3 abnormalities.

Patients and Methods: MDS patients with chromosome 3 abnormalities were identified through the Moffitt Cancer Center MDS database (n=1900). Patients were included if they had a WHO diagnosis of MDS, MDS/myeloproliferative neoplasm (MPN), therapy related MDS (t-MDS), or AML (20-30% myeloblasts) and had any karyotypic abnormality in chromosome 3. MDS data included baseline demographics, disease characteristics, treatment and outcome. Response was evaluated by IWG 2006 criteria. Kaplan-Meier estimates were used for overall survival

Results: A total of 129 patients were identified in our cohort with a median age of 70 years. WHO classification was as follows: 8% RA/RARS, 18% RCMD, 40% RAEB-1, 29% RAEB-2, 4% MDS/MPN, 1% AML. One third of the patients had t-MDS. Overall, 92% of patients were higher risk by R-IPSS (intermediate to very high risk) with a mean blast % in bone marrow of 7.2%. Distribution of cytogenetic abnormalities were inv(3) (7%), del(3q) (18%), t(3) (18%), monosomy 3 (19%), 3p abnormalities (15%), and other chromosome 3 changes (23%). Median OS for the cohort was 17.4 months (95% C.I. 14.6 to 20.2 months) and 51% of patients transformed to AML. R-IPSS was predictive for median OS across subgroups (p=0.04). Of interest, cytogenetic abnormality was predictive for survival (p=.001) with median OS as follows: inv(3) (30 months), t(3) (26 months), 3p (22 months), del(3q) (17 months), other (17 months) and monosomy 3 (13 months). Complex karyotype (>/= 3 abnormalities) was observed in 68% of patients and correlated with decreased OS (14 months versus 30 months, p < 0.005). Frontline HMA treatments were utilized in 66% of patients (92% azacitidine). Of patients who received azacitidine as therapy (73%), the overall response rate was 28% (12% hematological improvement (HI), 6% PR, 10% CR) with stable disease in 29%. Median OS with and without HMA was 19.3 months versus 10.3 months (p=0.034). Allogeneic transplantation was performed in 16% (n=20) of cases with median OS of 40 months versus 16 months in non-transplanted patients (p=0.001).

Conclusion: To our knowledge, this is the largest cohort reported to date describing MDS patients with chromosome 3 abnormalities. These patients have a poor prognosis and very high rate of leukemic transformation. However, survival was heterogeneous based on the specific cytogenetic abnormality with inv(3) having more favorable overall survival. Azacitidine therapy is effective in this population. Although the median OS is similar to patients with poor karyotype on the AZA-001 study, the ORR is less than expected. Allogeneic transplantation has significant impact on improved OS but is only an option for a subset of patients.