Abstract 357


Molecular predictors for treatment outcome after allogeneic hematopoietic stem cell transplantation (HSCT) of MDS and AML patients are limited. Recently, mutations in the splicing gene machinery have been described as frequent aberrations in MDS. The aim of this study was to investigate the prognostic impact of mutations in the splicing genes U2AF1, SRSF2 and SF3B1 in a large cohort of patients with high risk MDS or secondary AML following MDS (sAML) undergoing allogeneic HSCT.

Patients and Methods:

Patients (n=339) with a diagnosis of MDS (50.1%) or sAML (49.9%) who received allogeneic HSCT at four German university medical centers (Dresden, Düsseldorf, Hamburg and Hannover) between 1996 and 2011 and for whom genomic DNA was available from a time when the disease was active, were evaluated for the presence of mutations in the splicing genes U2AF1, SRSF2, and SF3B1 by direct sequencing.


Median follow up from time of transplantation was 3.27 years. Median patient age at time of HSCT was 58 years (range 19–74). 74 patients (21.8%) were in complete remission and 265 patients (78.2%) had active disease before transplantation. Low, intermediate, and high risk cytogenetics according to IPSS were found in 204 (60.2%), 42 (12.4%), and 76 (22.4%) patients, respectively (in 5% cytogenetic information was not available). Related donor HSCT was performed in 82 patients (24.2%), and unrelated donor HSCT in 257 patients (75.8%). Myeloablative preparative regimens were used in 51 patients (15%), and a non-myeloablative regimen was given to 288 patients (85%).

Mutations in U2AF1, SRSF2 and SF3B1 were detected in 14 (4.1%), 32 (9.4%) and 18 (5.3%) patients, respectively. SRSF2 and SF3B1 mutations co-occured in two patients, while the other patients had not more than one mutation in the investigated genes. Baseline characteristics were similarly distributed between U2AF1, SRSF2, or SF3B1 mutated and wildtype patients, respectively (sex, MDS vs sAML, cytogenetics, CMV status of patient, type of previous treatment, and remission status prior to transplantation), except a higher median age of U2AF1 mutated compared to wildtype patients (P=.02). There were no differences regarding transplant-related characteristics between patients with mutated or wildtype U2AF1, SRSF2, and SF3B1 (reduced intensity vs standard conditioning, GvHD prophylaxis, donor age, donor sex, CMV and HLA compatibility between recipient and donor).

U2AF1 mutations were associated with a significantly shorter overall survival (OS, median 0.58 vs 3.6 years in mutated vs wildtype patients, respectively, HR 2.54; 95%CI 1.41–4.58; P=.002). The cumulative incidence of relapse (CIR) was higher in U2AF1 mutated compared to wildtype patients (3-year CIR 50% vs 22%, P=.002), while non-relapse mortality (NRM) was similar between mutated and wildtype patients (3-year NRM 43% vs 29%, P=.11). Mutations in SRSF2 and SF3B1 were not associated with OS (P=.98 and P=.44, respectively), CIR (P=.19 and P=.19, respectively), and NRM (P=.49 and P=.44, respectively). In multivariate analysis, when considering variables with P<.15 in univariate analysis (age above or below 55 years, karyotype, stage [MDS vs sAML], CMV serostatus of patient, donor type [related vs unrelated], donor sex), U2AF1 mutations independently predicted shorter OS (HR 2.6; 95%CI 1.39–4.85; P=.011) besides karyotype, stage, CMV serostatus, and donor sex. Mutations in U2AF1 independently predicted higher CIR in multivariate analysis (HR 3.02, 95% CI 1.36–6.7, P=.007). The rates of acute and chronic GvHD were similar in U2AF1 mutated and wildtype patients.


U2AF1 mutations independently predicted worse patient outcome after allogeneic HSCT in MDS and sAML patients in our study due to a higher incidence of relapse. The U2AF1 mutation status may become useful to identify patients at high-risk for relapse after transplantation independent of established prognostic factors.


Platzbecker:GlaxoSmithKline: Honoraria, Research Funding.

#Nikolaus Kröger and Michael Heuser share the senior authorship

Author notes


FT and CK contributed equally to this work.

This icon denotes a clinically relevant abstract