Multivariate Cox regression analyses of TTT with interactions between biological prognostic markers
| Parameter . | HR (95% CI)* . | P . |
|---|---|---|
| ZAP-70 more than 20% and CD49d more than 30% | 3.12 (1.17-8.32) | .023 |
| IGVH UM and CD49d more than 30% | 2.95 (1.46-5.96) | .002 |
| Mod-Rai II-III | 3.36 (1.59-7.13) | .002 |
| β2M more than 2.2 g/L | 2.08 (1.20-3.60) | .009 |
| Parameter . | HR (95% CI)* . | P . |
|---|---|---|
| ZAP-70 more than 20% and CD49d more than 30% | 3.12 (1.17-8.32) | .023 |
| IGVH UM and CD49d more than 30% | 2.95 (1.46-5.96) | .002 |
| Mod-Rai II-III | 3.36 (1.59-7.13) | .002 |
| β2M more than 2.2 g/L | 2.08 (1.20-3.60) | .009 |
Multivariate Cox regression analyses of TTT were performed by including the following covariates: CD49d, CD38, ZAP-70, IGHV, β2M, sCD23, mod-Rai, and interactions between ZAP-70/CD49d and IGHV/CD49d (131 patients). For each variable, the selected cutoff points or the compared categories are indicated.
RA indicates hazard ratio; CI, confidence interval; other abbreviations are as in Tables 1,2–3.
Based on the final model after stepwise selection of covariates.