Table 2.
Covariate | Univariate analysis | Multivariate analysis | ||||
---|---|---|---|---|---|---|
HR | 95% CI | P-value | HR | 95% CI | P-value | |
Relapse | ||||||
Disease status* | 2.953 | 1.500–5.815 | 0.002 | 4.079 | 1.821–9.137 | 0.001 |
Pre-transplantation MRD (positive vs. negative) | 2.584 | 1.206–5.539 | 0.015 | |||
Donor-recipient sex-matched graft | ||||||
Female-male | 0.928 | 0.232–3.710 | 0.916 | 0.518 | 0.112–2.392 | 0.399 |
Female-female | 3.817 | 1.196–12.181 | 0.024 | 3.248 | 1.006–10.484 | 0.049 |
Male-male | 1.655 | 0.509–5.378 | 0.402 | 1.464 | 0.448–4.789 | 0.528 |
Male-female | 1 | 1 | ||||
Transplant–related mortality | ||||||
Platelet engraftment (yes vs. no) | 0.047 | 0.020–0.107 | <0.001 | 0.047 | 0.020–0.107 | <0.001 |
Leukemia-free survival | ||||||
Donor-recipient sex-matched graft | ||||||
Female-male | 0.774 | 0.306–1.962 | 0.590 | 1.344 | 0.499–3.568 | 0.566 |
Female-female | 2.363 | 1.072–5.209 | 0.033 | 3.166 | 1.414–7.086 | 0.005 |
Male-male | 1.505 | 0.708–3.196 | 0.288 | 2.426 | 1.088–5.408 | 0.030 |
Male-female | 1 | 1 | ||||
Platelet engraftment (yes vs. no) | 0.067 | 0.033–0.133 | <0.001 | 0.054 | 0.025–0.115 | <0.001 |
Overall survival | ||||||
Disease status | 1.956 | 0.980–3.905 | 0.057 | |||
Donor-recipient sex-matched graft | ||||||
Female-male | 0.492 | 0.168–1.439 | 0.195 | 0.898 | 0.290–2.780 | 0.851 |
Female-female | 1.823 | 0.799–4.158 | 0.154 | 2.471 | 1.067–5.723 | 0.035 |
Male-male | 1.352 | 0.629–2.909 | 0.440 | 2.290 | 1.006–5.215 | 0.048 |
Male-female | 1 | 1 | ||||
Platelet engraftment (yes vs. no) | 0.059 | 0.029–0.119 | <0.001 | 0.050 | 0.023–0.110 | <0.001 |
allo-SCT, allogeneic stem cell transplantation; MRD, measurable residual disease; HR, hazard ratio; CI, confidence interval.
All variables were first included in the univariate analysis, including sex, age, donor-recipient sex-matched grafts, donor-recipient relationship, ABO matched graft, pre-transplantation MRD status, disease status and hematopoietic engraftments; only variables with P < 0.1 were included in the Cox proportional hazards model with time-dependent variables.