Table 5.
Analyses of Recurrence-Free Survival (RFS) for Patients with Localized Disease
| Multivariate Analyses | |||||||
|---|---|---|---|---|---|---|---|
| Univariate Analyses | Full Model | Reduced Model | |||||
| Patient Characteristic | Hazard Ratio (95% CI) | P-value | Hazard Ratio (95% CI) | P-value | Hazard Ratio (95% CI) | P-value | |
| Tumor Stage | T3 vs T1/T2 | 1.0 (0.5, 2.0) | 0.96 | 1.7 (1.1, 2.5) | 0.01 | 1.7 (1.2, 2.5) | 0.005 |
| T4 vs T1/T2 | 3.1 (1.5, 6.8) | 0.003 | 2.3 (1.3, 3.9) | 0.002 | 2.5 (1.5, 4.0) | <0.001 | |
| LN Positive | Yes vs No | 2.5 (1.4, 4.6) | 0.003 | 2.0 (1.4, 2.8) | <0.001 | 2.1 (1.5, 2.8) | <0.001 |
| Year of Nephrectomy* | 1996–2005 vs 1986–1995 | 1.1 (0.5, 2.5) | 0.74 | 0.8 (0.5, 1.3) | 0.30 | 0.8 (0.5, 1.3) | 0.30 |
| 2006–2011 vs 1986–1995 | 1.3 (0.5, 3.1) | 0.61 | 1.2 (0.8, 1.9) | 0.31 | 1.2 (0.8, 1.9) | 0.29 | |
| Positive Margin | Yes vs No | 1.8 (1.0, 3.5) | 0.07 | 1.1 (0.7, 1.6) | 0.66 | NI | |
| LVI | Yes vs No | 2.3 (1.2, 4.5) | 0.01 | 1.1 (0.8, 1.6) | 0.48 | NI | |
|
| |||||||
| Age | >55 vs <=55 | 0.8 (0.4, 1.3) | 0.34 | ||||
| ECOG Status | 1 vs 0 | 1.0 (0.6, 1.6) | 0.89 | ||||
| >=2 vs 0 | 1.1 (0.1, 8.1) | 0.94 | |||||
| Clear Cell Histology | Yes vs No | 0.7 (0.4, 1.3) | 0.24 | ||||
| Percent Sarcomatoid | 25–49 vs 1–24 | 1.7 (0.7, 3.8) | 0.23 | ||||
| 50–74 vs 1–24 | 1.9 (0.8, 4.3) | 0.14 | |||||
| 75–99 vs 1–24 | 1.1 (0.5, 2.2) | 0.88 | |||||
| Necrosis | Necrosis Yes vs No | 1.2 (0.7, 2.2) | 0.45 | ||||
RFS is only calculated for patients who both had localized disease and were NED after nephrectomy. With only 56 RFS events among 73 applicable patients, only 5 characteristics could be included in a multivariable model and maintain model stability. Tumor stage, LN positivity, year of nephrectomy, positive margins, and LVI were selected based on known or suspected clinical importance combined with new interest based on univariable analyses. The remaining variables were explored in univariable analyses only.
NI=Not included. After step-wise backward selection, variables that did not remain in the full model due to lack of statistical significance were not included in the final reduced model.
Even though each individual comparison for year of nephrectomy is not significant in the multivariate model, it remains in the model because the overall p-value for any difference among the 3 time groups is significant (p=0.04).