Multivariate prognostic analyses
| (Tested category vs reference category) . | Overall survival . | Freedom from progression . | Freedom from treatment failure . | ||||
|---|---|---|---|---|---|---|---|
| Variables . | No. of patients . | Relative risk (95% CI) . | P . | Relative risk (95% CI) . | P . | Relative risk (95% CI) . | P . |
| A. UGT1A1 polymorphisms vs clinical parameters | |||||||
| Prognostic parameter | |||||||
| Age (≥ 45 vs < 45) 64 vs 209 | 4.07 (1.80-9.22) | .0007 | 1.93 (1.02-3.67) | .04 | 2.69 (1.51-4.82) | .0008 | |
| ESR (≥ 30 vs < 30) | 167 vs 106 | > .05 | 4.82 (1.88-12.3) | .001 | 4.58 (1.94-10.8) | .0005 | |
| B symptoms (yes vs no) | 127 vs 146 | 3.40 (1.35-8.60) | .01 | > .05 | > .05 | ||
| Polymorphisms | |||||||
| UGT1A1 (66 vs 67 and 77) | 142 vs 131 | 2.54 (1.05-6.14) | .04 | 2.70 (1.37-5.31) | .004 | 2.37 (1.28-4.40) | .006 |
| B. UGT1A1 polymorphisms vs IPS | |||||||
| Prognostic parameter | |||||||
| IPS (≥ 2 vs < 2) | 151 vs 127 | > .05 | 1.99 (0.94-4.21) | .07 | 2.21 (1.06-4.61) | .03 | |
| IPS (≥ 3 vs < 3) | 83 vs 195 | 4.91 (2.09-11.5) | .0003 | > .05 | > .05 | ||
| IPS (≥ 4 vs < 4) | 39 vs 239 | > .05 | 2.12 (1.02-4.41) | .04 | 2.60 (1.34-5.03) | .005 | |
| UGT1A1 (TA6/TA6 vs 67 and 77) | 144 vs 134 | — | 2.40 (1.24-4.65) | .01 | 2.13 (1.16-3.90) | .01 | |
| (Tested category vs reference category) . | Overall survival . | Freedom from progression . | Freedom from treatment failure . | ||||
|---|---|---|---|---|---|---|---|
| Variables . | No. of patients . | Relative risk (95% CI) . | P . | Relative risk (95% CI) . | P . | Relative risk (95% CI) . | P . |
| A. UGT1A1 polymorphisms vs clinical parameters | |||||||
| Prognostic parameter | |||||||
| Age (≥ 45 vs < 45) 64 vs 209 | 4.07 (1.80-9.22) | .0007 | 1.93 (1.02-3.67) | .04 | 2.69 (1.51-4.82) | .0008 | |
| ESR (≥ 30 vs < 30) | 167 vs 106 | > .05 | 4.82 (1.88-12.3) | .001 | 4.58 (1.94-10.8) | .0005 | |
| B symptoms (yes vs no) | 127 vs 146 | 3.40 (1.35-8.60) | .01 | > .05 | > .05 | ||
| Polymorphisms | |||||||
| UGT1A1 (66 vs 67 and 77) | 142 vs 131 | 2.54 (1.05-6.14) | .04 | 2.70 (1.37-5.31) | .004 | 2.37 (1.28-4.40) | .006 |
| B. UGT1A1 polymorphisms vs IPS | |||||||
| Prognostic parameter | |||||||
| IPS (≥ 2 vs < 2) | 151 vs 127 | > .05 | 1.99 (0.94-4.21) | .07 | 2.21 (1.06-4.61) | .03 | |
| IPS (≥ 3 vs < 3) | 83 vs 195 | 4.91 (2.09-11.5) | .0003 | > .05 | > .05 | ||
| IPS (≥ 4 vs < 4) | 39 vs 239 | > .05 | 2.12 (1.02-4.41) | .04 | 2.60 (1.34-5.03) | .005 | |
| UGT1A1 (TA6/TA6 vs 67 and 77) | 144 vs 134 | — | 2.40 (1.24-4.65) | .01 | 2.13 (1.16-3.90) | .01 | |
Patient characteristics and prognostic parameters used in Table 2 and all the polymorphisms were entered in the model. A total of 273 patients without missing values are included in the analysis. The IPS (as 4 dummy covariates, see Table 4) and all the polymorphisms were entered in the model. Missing values were not replaced for the estimation of the IPS, and only 278 patients without missing values were included in the analysis. The covariates significantly related to the studied endpoint at the end of the backward selection process are displayed with their relative risk and 95% confidence interval and P value. No other covariate is retained in the model (P < .05). UGTA is not selected (P = .08) in the multivariate analysis of overall survival.