Background: Despite growing understanding of clinical and genetic basis of myelodysplastic syndromes (MDS) and increased use of allo-HCT, the disease remains incurable in approx. half of cases. Availability of next generation gene sequencing is an important tool that helps further prognosticate outcomes.

Pts, materials and methods: We performed targeted amplicon-based next generation sequencing (NGS) of the 26 most frequently mutated genes in MDS. DNA was extracted from paraffin embedded bone marrow (FFPE) samples obtained during pre-transplant work-up at a median of 30 (9-53) days prior to allo-HCT. Sequencing was performed on the Illumina platform with an average read depth of 1500x. The Qiagen GeneRead software suite was used for alignment and variant calling. Frameshift, nonsense, and missense variants that were not present in germline databases at >1% frequency and those that were predicted to be functionally significant by SIFT and Polyphen were deemed true mutations. Cutoff for variant allele frequency was set at 20%. Kaplan-Meier estimates were used for overall survival (OS) and Cox regression for multivariable analysis.

Results: We identified 139 pts who received an allo-HCT between 01/2005 and 06/2012 for MDS (89%), AML (4%) or CMML (7%). However, 38 samples were excluded for inability to isolate DNA due to poor sample quality. Patient, disease, and transplant characteristics are summarized (Table 1). Median age of pts was 58 (22-74) yrs and the majority (77%) had received prior azacitidine. Somatic mutations (≥1) at time of allo-HCT were identified in 39% of cases. The most common mutations in decreasing frequency were: ASXL1 (11%), DNMT3A (6%), IDH2 (5%), KRAS (4%), RUNX1 (4%), TP53 (4%), TET2 (3%), SRSF2 (3%), SF3B1 (2%), EZH2 (1%), BCL (1%), MLL (1%), and WT1 (1%). FLU-BU was the most common preparative regimen (92%). Median F/U for all surviving pts was 36 months. Median OS for all pts was 29 (95% CI=10-48) mos. There was no difference in OS between pts harboring ≥1 mutation vs. those with none (34 (95%CI=10-57) mos vs. 29 (95%CI=16-42) mos, p=0.7), or in the presence of >5% vs. ≤ 5% BM myeloblasts at time of allo-HCT (20 (95%CI=10-31) mos vs. 34 (95%CI=13-55) mos, p=0.5), or if pts were ≥ 60 yrs of age vs. <60 (34 (95%CI=15-34) mos vs. 29 (7-52) mos, p=0.7). Pts with R-IPSS (poor and very poor risk) at time of allo-HCT had over 2-fold lower OS vs. R-IPSS (very low, low, and int) (15 vs. 41 mos, p=0.06). When analyzing OS based on specific mutations, presence of mutated TP53 (6 (95%CI=0-17) mos vs. 34 (95%CI=14-53) mos, p=0.02) or IDH2 (11 (95%CI=6-16) mos vs. 34 (95%CI=12-56) mos, p=0.006) predicted for significantly worse OS after allo-HCT. No difference in OS was observed when mutated ASXL-1, TET-2, DNMT3A, or RUNX-1, were present. Multivariable analysis (adjusted for %BM blasts and karyotype) identified TP53 (HR=3.3 (95% CI 1.1-10), p 0.03) and IDH2 (HR=3.9 (95% CI 1.5-11), p 0.006) as independent predictors of poor outcomes after allo-HCT.

Conclusion: Our study confirms the adverse prognostic significance of TP53 mutation in MDS pts undergoing allo-HCT and identifies mutated IDH2 as an independent predictor of inferior OS in allo-HCT. Identifying these extremely high-risk populations (TP53 and IDH 2 mutations) is useful as standard allo-HCT seems ineffective (OS< 12 mos); and there is an unmet need to improve outcomes by integrating novel agents as post-transplant consolidation or maintenance. The lower frequency of somatic mutations in our study sample compared to previously reported might be explained by our more stringent criteria to screen out FFPE artifacts, among others. Given the relatively small sample size, our findings merit validation in a larger data set.

Table 1.
VariablesN=101
Recipient median age (range), yrs 58 (22-74) 
Recipient sex M=54% 
WHO at Dx RA=2%
RARS=3%
RCMD=16%
RAEB-I=31%
RAEB-II=26%
CMML=8%
MDS-unclas=5%
MDS/MPN=1%
MDS-RS=5%
AML=4% 
Azacitidine before allo-HCT No=22%
Yes=77%
Unk=1% 
Best response to Azacitidine pre-HCT CR=8%
CR marrow=4%
PR=12%
HI=5%
SD=43%
PD=7%
Missing=22% 
Donor source MRD=35%
MUD=46%
MMUD=17%
Cord=3% 
Cell source PBSC=96%
BM=1%
Cord=3% 
Preparative regimen FLU-BU=92%
FLU-MEL=3%
Other=5% 
GVHD prophylaxis MTX=66%
SIR=19%
MMF=15% 
R-IPSS at allo-HCT Very low=15%
Low=24%
Int=18%
High=20%
Very high=20%
Missing=4% 
Cytogenetic scoring 2=56%
3=16%
4=22%
5= 2%
Missing= 4% 
Mutation present No=61%
Yes=39% 
VariablesN=101
Recipient median age (range), yrs 58 (22-74) 
Recipient sex M=54% 
WHO at Dx RA=2%
RARS=3%
RCMD=16%
RAEB-I=31%
RAEB-II=26%
CMML=8%
MDS-unclas=5%
MDS/MPN=1%
MDS-RS=5%
AML=4% 
Azacitidine before allo-HCT No=22%
Yes=77%
Unk=1% 
Best response to Azacitidine pre-HCT CR=8%
CR marrow=4%
PR=12%
HI=5%
SD=43%
PD=7%
Missing=22% 
Donor source MRD=35%
MUD=46%
MMUD=17%
Cord=3% 
Cell source PBSC=96%
BM=1%
Cord=3% 
Preparative regimen FLU-BU=92%
FLU-MEL=3%
Other=5% 
GVHD prophylaxis MTX=66%
SIR=19%
MMF=15% 
R-IPSS at allo-HCT Very low=15%
Low=24%
Int=18%
High=20%
Very high=20%
Missing=4% 
Cytogenetic scoring 2=56%
3=16%
4=22%
5= 2%
Missing= 4% 
Mutation present No=61%
Yes=39% 

Disclosures

Komrokji:Novartis: Research Funding, Speakers Bureau; Celgene: Consultancy, Research Funding; Pharmacylics: Speakers Bureau; Incyte: Consultancy. Field:PDL Biopharma: Research Funding. Perkins:PDL Biopharma: Research Funding. Lancet:Seattle Genetics: Consultancy; Pfizer: Research Funding; Boehringer-Ingelheim: Consultancy; Kalo-Bios: Consultancy; Amgen: Consultancy; Celgene: Consultancy, Research Funding. List:Celgene Corporation: Honoraria, Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution