Cause-specific Cox proportional hazard regression models using CIMP subgroups and D29 (NOPHO)/33 MRD (DCOG) to predict the risk of relapse (accounting for competing events) and death in pediatric patients with T-ALL
| . | Risk of relapse . | Risk of death . | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NOPHO (n = 172, n = 23 relapses) . | DCOG (n = 137, n = 18 relapses)† . | NOPHO (n = 172, n = 28 deaths) . | DCOG (n = 137, n = 24 deaths)† . | |||||||||
| Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | |
| Univariable analysis | ||||||||||||
| CIMP low | 5.27 (1.95-14.2) | .001∗∗ | 213.5 | 3.43 (1.29-9.15) | .014∗ | 167.9 | 3.06 (1.38-6.77) | .006∗∗ | 272.6 | 2.31 (1.03-5.20) | .044∗ | 229.4 |
| D29/33 MRD ≥ 0.1% | 3.10 (1.27-7.53) | .013∗ | 220.3 | 2.55 (1.00-6.46) | .049∗ | 170.6 | 7.84 (2.72-22.6) | <.001∗∗∗ | 259.3 | 3.89 (1.66-9.10) | .002∗∗ | 223.0 |
| log10(WBC) | 1.69 (1.02-2.82) | .044∗ | 222.4 | 1.78 (0.72-4.38) | .212 | 162.5 | 1.19 (0.80-1.79) | .386 | 280.2 | 1.43 (0.90-2.28) | .131 | 220.5 |
| CNS3 | 1.13 (0.38-3.33) | .827 | 227.1 | 1.70 (0.39-7.44) | .481 | 161.9 | 0.67 (0.20-2.22) | .509 | 280.5 | 1.85 (0.55-6.27) | .321 | 210.1 |
| Male sex | 1.38 (0.47-4.05) | .560 | 226.7 | 0.70 (0.26-1.88) | .483 | 174.0 | 0.50 (0.23-1.09) | .081 | 278.1 | 0.57 (0.25-1.30) | .183 | 231.9 |
| Age | 0.56 (0.34-0.92) | .022∗ | 221.0 | 1.99 (1.21-3.25) | .006∗∗ | 166.4 | 1.12 (0.77-1.63) | .552 | 280.6 | 1.09 (0.73-1.64) | .669 | 233.4 |
| Multivariate analysis | ||||||||||||
| CIMP low | 4.92 (1.82-13.28) | .002∗∗ | 209.9 | 2.98 (1.10-8.09) | .032∗ | 167.7 | 2.72 (1.23-6.03) | .013∗ | 254.7 | 1.83 (0.80-4.19) | .150 | 222.9 |
| D29/33 MRD ≥ 0.1% | 2.80 (1.15-6.84) | .023∗ | 2.05 (0.79-5.30) | .138 | 7.32 (2.53-21.13) | <.001∗∗∗ | 3.47 (1.46-8.23) | .005∗∗ | ||||
| Multivariate analysis | ||||||||||||
| CIMP low | 4.06 (1.46-11.3) | .007∗∗ | 208.1 | 2.53 (0.90-7.07) | .078 | 156.0 | 2.71 (1.20-6.12) | .017∗ | 258.1 | 1.65 (0.70-3.88) | .254 | 216.4 |
| D29/33 MRD ≥ 0.1% | 2.58 (1.05-6.37) | .039∗ | 1.94 (0.74-5.07) | .176 | 7.40 (2.56-21.3) | <.001∗∗∗ | 3.33 (1.40-7.94) | .006∗∗ | ||||
| Age | 0.65 (0.41-1.01) | .055 | 1.81 (1.12-2.92) | .016∗ | 1.14 (0.80-1.64) | .473 | 1.06 (0.71-1.59) | .781 | ||||
| log10(WBC) | 1.30 (0.76-2.20) | .337 | 1.34 (0.70-2.55) | .378 | 1.07 (0.74-1.55) | .723 | 1.26 (0.78-2.05) | .345 | ||||
| CIMP-MRD interaction analysis | ||||||||||||
| D29/33 MRD < 0.1%/CIMP high | Ref. | Ref. | Ref. | Ref. | ||||||||
| D29/33 MRD < 0.1%/CIMP low | 13.21 (2.42-245) | .001∗∗ | 6.89 (1.39-34.13) | .018∗ | 3.30 (0.64-23.8) | .152 | 3.63 (0.87-15.18) | .078 | ||||
| D29/33 MRD ≥ 0.1%/CIMP high | 9.11 (1.56-172) | .012∗ | 5.68 (1.04-31.03) | .045∗ | 6.63 (1.71-43.5) | .005∗∗ | 6.22 (1.61-24.05) | .008∗∗ | ||||
| D29/33 MRD ≥ 0.1%/CIMP low | 22.76 (4.48-415) | <.001∗∗∗ | 8.53 (1.72-42.29) | .009∗∗ | 15.48 (4.43-97.7) | <.001∗∗∗ | 8.01 (2.17-29.61) | .002∗∗ | ||||
| . | Risk of relapse . | Risk of death . | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NOPHO (n = 172, n = 23 relapses) . | DCOG (n = 137, n = 18 relapses)† . | NOPHO (n = 172, n = 28 deaths) . | DCOG (n = 137, n = 24 deaths)† . | |||||||||
| Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | Hazard ratio (95% CI) . | P value . | AIC . | |
| Univariable analysis | ||||||||||||
| CIMP low | 5.27 (1.95-14.2) | .001∗∗ | 213.5 | 3.43 (1.29-9.15) | .014∗ | 167.9 | 3.06 (1.38-6.77) | .006∗∗ | 272.6 | 2.31 (1.03-5.20) | .044∗ | 229.4 |
| D29/33 MRD ≥ 0.1% | 3.10 (1.27-7.53) | .013∗ | 220.3 | 2.55 (1.00-6.46) | .049∗ | 170.6 | 7.84 (2.72-22.6) | <.001∗∗∗ | 259.3 | 3.89 (1.66-9.10) | .002∗∗ | 223.0 |
| log10(WBC) | 1.69 (1.02-2.82) | .044∗ | 222.4 | 1.78 (0.72-4.38) | .212 | 162.5 | 1.19 (0.80-1.79) | .386 | 280.2 | 1.43 (0.90-2.28) | .131 | 220.5 |
| CNS3 | 1.13 (0.38-3.33) | .827 | 227.1 | 1.70 (0.39-7.44) | .481 | 161.9 | 0.67 (0.20-2.22) | .509 | 280.5 | 1.85 (0.55-6.27) | .321 | 210.1 |
| Male sex | 1.38 (0.47-4.05) | .560 | 226.7 | 0.70 (0.26-1.88) | .483 | 174.0 | 0.50 (0.23-1.09) | .081 | 278.1 | 0.57 (0.25-1.30) | .183 | 231.9 |
| Age | 0.56 (0.34-0.92) | .022∗ | 221.0 | 1.99 (1.21-3.25) | .006∗∗ | 166.4 | 1.12 (0.77-1.63) | .552 | 280.6 | 1.09 (0.73-1.64) | .669 | 233.4 |
| Multivariate analysis | ||||||||||||
| CIMP low | 4.92 (1.82-13.28) | .002∗∗ | 209.9 | 2.98 (1.10-8.09) | .032∗ | 167.7 | 2.72 (1.23-6.03) | .013∗ | 254.7 | 1.83 (0.80-4.19) | .150 | 222.9 |
| D29/33 MRD ≥ 0.1% | 2.80 (1.15-6.84) | .023∗ | 2.05 (0.79-5.30) | .138 | 7.32 (2.53-21.13) | <.001∗∗∗ | 3.47 (1.46-8.23) | .005∗∗ | ||||
| Multivariate analysis | ||||||||||||
| CIMP low | 4.06 (1.46-11.3) | .007∗∗ | 208.1 | 2.53 (0.90-7.07) | .078 | 156.0 | 2.71 (1.20-6.12) | .017∗ | 258.1 | 1.65 (0.70-3.88) | .254 | 216.4 |
| D29/33 MRD ≥ 0.1% | 2.58 (1.05-6.37) | .039∗ | 1.94 (0.74-5.07) | .176 | 7.40 (2.56-21.3) | <.001∗∗∗ | 3.33 (1.40-7.94) | .006∗∗ | ||||
| Age | 0.65 (0.41-1.01) | .055 | 1.81 (1.12-2.92) | .016∗ | 1.14 (0.80-1.64) | .473 | 1.06 (0.71-1.59) | .781 | ||||
| log10(WBC) | 1.30 (0.76-2.20) | .337 | 1.34 (0.70-2.55) | .378 | 1.07 (0.74-1.55) | .723 | 1.26 (0.78-2.05) | .345 | ||||
| CIMP-MRD interaction analysis | ||||||||||||
| D29/33 MRD < 0.1%/CIMP high | Ref. | Ref. | Ref. | Ref. | ||||||||
| D29/33 MRD < 0.1%/CIMP low | 13.21 (2.42-245) | .001∗∗ | 6.89 (1.39-34.13) | .018∗ | 3.30 (0.64-23.8) | .152 | 3.63 (0.87-15.18) | .078 | ||||
| D29/33 MRD ≥ 0.1%/CIMP high | 9.11 (1.56-172) | .012∗ | 5.68 (1.04-31.03) | .045∗ | 6.63 (1.71-43.5) | .005∗∗ | 6.22 (1.61-24.05) | .008∗∗ | ||||
| D29/33 MRD ≥ 0.1%/CIMP low | 22.76 (4.48-415) | <.001∗∗∗ | 8.53 (1.72-42.29) | .009∗∗ | 15.48 (4.43-97.7) | <.001∗∗∗ | 8.01 (2.17-29.61) | .002∗∗ | ||||
AIC, Akaike information criterion, in which the fittest model reveals the lowest value and vice versa; CI, confidence interval; CNS3, central nervous system 3 involvement status; Ref., reference group.
∗P < .05, ∗∗P < .01, ∗∗∗P < .001.
Due to WBC count missings, DCOG models including WBC have n = 134 with 17 relapses and 23 deaths.