Multivariable analyses of variables impacting on LFS
| Gene variants | Univariate analysis | Covariate analysis (NKG2A and NKG2D SNPs in one model) | Multivariate analysis | ||||||
| HR | 95% CI | P value | HR | 95% CI | P value | HR | 95% CI | P value | |
| NKG2D SNP LFS | 1.58 | 0.87 to 2.87 | 0.1 | 1.09 | 0.57 to 2.07 | 0.8 | 1.39 | 0.76 to 2.56 | 0.3 |
| NKG2A SNP LFS | 2.70 | 1.32 to 5.49 | 0.006 | 2.6 | 1.21 to 5.57 | 0.01 | 2.19 | 1.05 to 4.59 | 0.01 |
Cox regression analysis with both indicated SNPs as numerical predictors. In the covariate analysis, both SNPs were included in the same model. For the multivariate analysis, previously reported predictive variables affecting the clinical outcome of AML patients including age, risk group (ELN-2010), number of induction chemotherapy cycles to attain complete remission (1 or >1), number of consolidation courses (0–2 or>2), CMV serostatus and HLA-B-21 (rs1050458) genotype were analyzed by univariable Cox regression analysis. Covariates with p values of below 0.1 (age, number of induction chemotherapies cycles, CMV status and HLA-B-21 (rs1050458) genotype status, NKG2A SNP (rs1983526), NKG2D SNP (rs1049174)) were included in the multivariate analysis. The ELN-2010 risk stratification used in this study included various cytogenetic and molecular aberrations which stratify patients into favorable, intermediate I, intermediate II and adverse groups.54
AML, acute myeloid leukemia; CMV, cytomegalovirus; LFS, leukemia-free survival; SNP, single nucleotide polymoprhism.