[Predict response to decitabine in patients with myelodysplastic syndrome and related neoplasms]

Objective: To identify clinical and molecular signatures for predicting response to decitabine (DAC) in patients with myelodysplastic syndrome (MDS) and related neoplasms. Methods: The clinical characteristics of 109 patients with MDS and related neoplasms who were treated with DAC were analyzed retrospectively and the next target sequencing was performed to define recurrently mutated genes in these disease samples, to examine the association of the clinical and molecular signatures with response to DAC treatment. Results: Of 109 MDS and related neoplasms patients, there were 70 males and 39 females, the median age was 61 years old (ranges: 17-85 years old) . According to the international prognostic scoring system (IPSS) , 46 cases were included in the relatively low risk group (low risk and intermediate-1 risk) , 63 in the relative high risk group (intermediate-2 and high risk) . There were 21 cases with complex karyotype, 17 chromosome 7 abnormality and 17 monosomal karyotype. The median courses of DAC treatment was 4 (2-11) . A total of 74 patients achieved response (67.9%) and 30 (27.5%) achieved complete response (CR) . Univariate analysis found that CR was higher in patients with high risk of IPSS, complex karyotypes, monosomal karyotypes, chromosome 7 abnormality, and platelet doubling after one cycle of DAC treatment. Patients with TP53 gene mutation were more likely to receive CR, 10 of 15 patients with TP53 mutations achieved CR. (66.7%) , which was significantly higher than that of the patients without TP53 gene mutation (21.3%) (P=0.001) . Multivariate analysis showed that TP53 gene mutation, platelet doubling after one cycle of DAC treatment and the complex karyotype were independent prognostic factors for CR. Of them, TP53 gene mutation is the strongest predictor (OR=4.39, 95%CI, 1.20-16.06, P=0.026) . Conclusion: TP53 mutation, platelet doubling after one cycle of DAC treatment and complex karyotypes could predict CR to DAC.