Table 2.
Multivariate analysis for overall survival, non-relapse mortality, and relapse incidence with traditional patient, disease and transplantation-related variables (model 1) extended with RBC-transfusions (model 2), and comorbidity (model 3), respectively
| Overall survival | Non-relapse mortality | Relapse incidence | ||||
|---|---|---|---|---|---|---|
| HR | p value | HR | p value | HR | p value | |
| Model 1 (n = 199) | ||||||
| Donor type (HLA matched unrelated* vs sibling) | 1.03 | 0.89 | 1.20 | 0.43 | 0.28 | 0.002 |
| Cytogenetic abnormalities** | 0.54 | 0.91 | 0.02 | |||
| Good and intermediate | 1 | 1 | 1 | |||
| Poor | 1.31 | 0.29 | 0.98 | 0.94 | 3.01 | 0.009 |
| Time Dx-Tx (>6 vs ≤6 months) | 1.31 | 0.20 | 1.33 | 0.24 | 1.26 | 0.54 |
| WHO classification | 0.01 | .03 | 0.006 | |||
| RA/RARS | 1 | 1 | 1 | |||
| RCMD | 3.12 | 0.001 | 2.80 | 0.006 | 3.58 | 0.15 |
| RAEB-1 | 1.15 | 0.65 | 0.77 | 0.48 | 5.95 | 0.001 |
| RAEB-2 | 1.57 | 0.13 | 1.41 | 0.29 | 1.81 | 0.35 |
| MDS/AML | 1.85 | 0.03 | 1.40 | 0.28 | 5.66 | 0.003 |
| Age (≤50 vs >50 years) | 1.28 | 0.24 | 1.12 | 0.64 | 1.85 | 0.10 |
| Sex-match donor-recipient (m-f vs other) | 1.05 | 0.84 | 0.99 | 0.98 | 1.11 | 0.80 |
| Model 2 (n = 125) | ||||||
| RBC-transfusion (>20 RBS-units vs ≤20 RBC-units) | 1.99 | 0.006 | 1.89 | 0.027 | 2.67 | 0.031 |
| Model 3 (n = 143) | ||||||
| Comorbidity (yes vs no) | 1.43 | 0.15 | 1.67 | 0.08 | 0.71 | 0.44 |
Cox regression models for the (cause-specific) hazards were fitted on different restricted datasets with full information on the main variables of interest. The entries in the table are hazard ratios and their associated p values and group size. HRs (with their p values) in italics have a statistically significant impact on OS, NRM, or RI, at the 5 %-level. (*) Mismatched siblings and mismatched unrelated donors are included in the HLA matched unrelated donor category. (**) patients with missing information for this variable were kept in the analysis with variable level “missing” (HRs not shown)