Skip to main content
. 2017 Jul 27;8(47):82475–82490. doi: 10.18632/oncotarget.19628

Table 2. Univariate and multivariate analyses for OS and PFS.

N=108 Univariate Multivariate*
OS (HR, 95%CI) PFS (HR, 95%CI) OS (HR, 95%CI) PFS (HR, 95%CI)
Age (≥ 60 vs. < 60 year) 2.456 (1.340-4.504) 1.846 (1.094-3.115) 2.230 (1.192-4.170) 2.278 (1.320-3.931)
HSCT vs. non-HSCT 1.993 (1.021-3.888) - - - - - -
Progressive MDS vs.
non-progressive MDS)
- - 2.190 (1.127-4.258) - - 1.819 (1.418-2.332)
GATA1/GATA2 mutation 2.703 (1.048-6.970) - - 3.714 (1.341-10.287) - -
TP53 mutation 3.160 (1.117-59.320) - - - - - -
DNMT3A mutation 3.106 (1.211-7.971) - - 2.842 (1.070-7.547) - -
RUNX1 mutation - - 2.208 (1.136-4.293) - - - -
KRAS/NRAS mutation - - 2.678 (1.137-6.307) 3.525 (1.288-9.650) - -
SRSF2 mutation - - 2.870 (1.225-6.722) - - - -
TET2 mutation - - 2.248 (1.103-10.060) - - - -
IDH1/IDH2 mutation - - - - - - 0.273 (0.081-0.919)
≥ 2 mutations
vs. <2 mutations
- - - - - - 3.364 (1.428-7.925)

*Multiple variables were selected for the Cox proportional hazard model: age (≥ 60 vs. < 60 year), white blood cell count, platelet count, hemoglobin, bone marrow blast, IPSS-R, administered treatment therapy (HSCT vs. non-HSCT), diagnosis (progressive MDS vs. non-progressive MDS) and mutations (including mutations with frequency ≥ 5% and those involved in epigenetic modification: RUNX1, BCOR/BCORL1, U2AF1, KRAS/NRAS, GATA1/GATA2, ETV6, NOTCH1/NOTCH2, STAG2, SETBP1, SRSF2, TP53, CEBPA, ASXL1, TET2, DNMT3A, IDH1/IDH2, EZH2, SETD2, KDM6A, KMT2A. And when analyzing “ ≥ 2 mutations vs. <2 mutations”, the mutation variables were removed.

Abbreviations: OS, overall survival; PFS, progression-free survival.