Table 2.
Multivariate analyses
| RFS | OS | |||
|---|---|---|---|---|
| Variable | HR (95% CI) | p | HR (95% CI) | p |
| ≥1 persisting mutation | 2.34 (1.31–4.17) | 0.0039 | 2.14 (1.05–4.35) | 0.036 |
| Age (continuous, hazard ratio per 10 y increase) | 0.98 (0.80–1.20) | 0.62 | 1.03 (0.80–1.32) | 0.85 |
| Leukocyte count (continuous, hazard ratio per 10 y increase) | 1.04 (1.01–1.08) | 0.0042 | 1.00 (1.00–1.00) | 0.44 |
| ELN 2017 genetic risk group | ||||
| Favorable | 1.07 (0.51–2.24) | 0.85 | 1.42 (0.55–3.68) | 0.47 |
| Intermediate | 1 (reference) | – | 1 (reference) | |
| Adverse | 3.23 (1.58–6.61) | 0.0013 | 3.80 (1.61–8.94) | 0.0023 |
| CRi vs. CR | 1.01 (0.53–1.93) | 0.97 | 0.97 (0.46–2.03) | 0.94 |
| AlloSCT in first CR/CRia | 0.28 (0.14–0.57) | 0.0004 | 0.69 (0.33–1.45) | 0.32 |
The Akaike Information Criterion (AIC) was 412.00 for the RFS model shown here, compared to 418.47 for the same model but without persisting mutations, with a lower value indicating a better fit of the model. The AIC was 308.54 for the OS model, compared to 311.01 for the same model but without persisting mutations. Both multivariate regression models were stratified according to trial arm.
aAllogeneic stem cell transplantation was included as a time-dependent covariate. In either model, there was no statistically significant interaction between alloSCT and mutation persistence.
RFS relapse-free survival, OS overall survival, HR hazard ratio, CI confidence interval, ELN European LeukemiaNet, CR complete remission, CRi complete remission with incomplete blood count recovery, alloSCT allogeneic stem cell transplantation.