Table 5.
Multivariable analysis of the PIK3CA mutation and p16 status association with overall survival, Cox regression model adjusted for HPV status, alcohol/tobacco consumption habits, and age at diagnosis.
| Factor | Coef (SE) | HR (95%CI) | p-Value |
|---|---|---|---|
| PIK3CA | 0.056 | ||
| Mutated | 1 | 1 | |
| Wild-type | 0.766 (0.401) | 2.15 (0.98–4.73) | |
| p16 overexpression | 0.670 | ||
| No | 1 | 1 | |
| Yes | −0.189 (0.439) | 0.83 (0.35–1.96) | |
| HPV status | 0.016 | ||
| Positive | 1 | 1 | |
| Not detected | 1.069 (0.443) | 2.91 (1.22–6.94) | |
| Consumption habits | 0.040 | ||
| Active | 1 | 1 | |
| No consumption | −1.041 (0.508) | 0.35 (0.13–0.96) | |
| Age at diagnosis * | |||
| Pspline, linear | 0.045 (0.017) | --- | 0.008 |
| Pspline, nonlinear | --- | --- | 0.016 |
Coef = estimated coefficient from the model; SE = standard error of coefficient; HR = Hazard ratio; 95% CI = 95% Confidence Interval. * Smoothing splines was used to adjust for age with a smoothed curve without categorisation or linearity assumption (theta = 0.973, degrees of freedom for age term = 2.66).