Table 3.

Multivariable Cox regression analysis of clinical, cytogenetic, and molecular predictors of adverse outcome in patients with AA

VariableHR (95% CI)P value
DNMT3A mutation 2.01 (0.04-104.74) .73 
Age at diagnosis 1.02 (1.00-1.05) .094 
PIGA mutation 0.56 (0.20-1.59) .274 
CUX1 mutation 8.18 (0.74-89.99) .086 
Total number of mutations 1.02 (0.67-1.56) .916 
U2AF1 mutation 0.58 (0.04-7.68) .682 
RUNX1 mutation 1.06 (0.06-19.70) .969 
TP53 mutation 1.60 (0.06-46.32) .785 
PNH clone size at diagnosis (%) 1.00 (0.97-1.04) .948 
TET2 mutation 0.81 (0.08-7.81) .854 
ZRSR2 mutation 0.58 (0.02-20.84) .768 
ASXL1 mutation 1.03 (0.10-10.20) .981 
BCOR/L1 mutation 0.77 (0.08-7.26) .82 
EZH2 mutation 0.93 (0.06-15.49) .959 
SETBP1 mutation 1.42 (0.06-35.24) .832 
PNH clone size of >0% 0.86 (0.38-1.96) .727 
Abnormal karyotype at AA diagnosis 0.59 (0.02-14.74) .747 
Karyotype del(13q) 0.45 (0.01-17.95) .674 
SAA or VSAA diagnosis 1.71 (0.64-4.52) .284 
Male sex 1.21 (0.54-2.70) .637 
AA-PNH overlap syndrome diagnosis 1.50 (0.29-7.92) .63 
VariableHR (95% CI)P value
DNMT3A mutation 2.01 (0.04-104.74) .73 
Age at diagnosis 1.02 (1.00-1.05) .094 
PIGA mutation 0.56 (0.20-1.59) .274 
CUX1 mutation 8.18 (0.74-89.99) .086 
Total number of mutations 1.02 (0.67-1.56) .916 
U2AF1 mutation 0.58 (0.04-7.68) .682 
RUNX1 mutation 1.06 (0.06-19.70) .969 
TP53 mutation 1.60 (0.06-46.32) .785 
PNH clone size at diagnosis (%) 1.00 (0.97-1.04) .948 
TET2 mutation 0.81 (0.08-7.81) .854 
ZRSR2 mutation 0.58 (0.02-20.84) .768 
ASXL1 mutation 1.03 (0.10-10.20) .981 
BCOR/L1 mutation 0.77 (0.08-7.26) .82 
EZH2 mutation 0.93 (0.06-15.49) .959 
SETBP1 mutation 1.42 (0.06-35.24) .832 
PNH clone size of >0% 0.86 (0.38-1.96) .727 
Abnormal karyotype at AA diagnosis 0.59 (0.02-14.74) .747 
Karyotype del(13q) 0.45 (0.01-17.95) .674 
SAA or VSAA diagnosis 1.71 (0.64-4.52) .284 
Male sex 1.21 (0.54-2.70) .637 
AA-PNH overlap syndrome diagnosis 1.50 (0.29-7.92) .63 

A multivariable Cox proportional hazards model was constructed to assess the association between baseline clinical characteristics, somatic mutations, karyotypic abnormalities, and PNH clone size with the risk of adverse outcomes in patients with aplastic anemia. HRs with 95% CIs and P values are shown for each variable. Among the top-ranked features by model-derived impact scores, CUX1 and DNMT3A mutations demonstrated the highest HRs (HR of 8.18 and 2.01, respectively), although neither reached statistical significance. Age at diagnosis trended toward significance (HR, 1.02 per year; P = .094). Other mutations, including TP53, RUNX1, TET2, and ASXL1, as well as clinical features such as PNH clone size, abnormal karyotype, and AA-PNH overlap, were not significantly associated with outcome in this cohort.

CI, confidence interval; HR, hazard ratio.

or Create an Account

Close Modal
Close Modal