Table 3. Multivariate analysis of independent prognostic factors in NPC patients.
| TNM6th | TNM7th | ||||||||
| End point | Variable | Univariate analysis | Multivariate analysis | Univariate analysis | Multivariate analysis | ||||
| HR(95% CI) | P value | HR(95% CI) | P value | HR(95% CI) | P value | HR(95% CI) | P value | ||
| Distant failure | T Classification | <0.001 | 0.002 | <0.001 | 0.002 | ||||
| T1 | reference | reference | reference | reference | |||||
| T2 | 2.157(1.225–3.797) | 0.008 | 1.905(1.079–3.366) | 0.026 | 2.099(1.236–3.566) | 0.006 | 1.840(1.080–3.136) | 0.025 | |
| T3 | 2.839(1.675–4.813) | <0.001 | 2.280(1.333–3.901) | 0.003 | 2.693(1.657–4.377) | <0.001 | 2.148(1.310–3.521) | 0.002 | |
| T4 | 3.526(2.037–6.103) | <0.001 | 2.822(1.611–4.943) | <0.001 | 3.344(2.012–5.559) | <0.001 | 2.634(1.565–4.431) | <0.001 | |
| N Classification | <0.001 | <0.001 | <0.001 | <0.001 | |||||
| N0 | 0.464(0.318–0.677) | <0.001 | 0.502(0.342–0.736) | <0.001 | 0.491(0.329–0.734) | 0.001 | 4.679(2.947–7.431) | <0.001 | |
| N1 | reference | reference | reference | reference | |||||
| N2 | 1.305(1.004–1.698) | 0.047 | 1.275(0.978–1.663) | 0.073 | 1.404(1.086–1.814) | 0.009 | 1.358(1.047–1.760) | 0.021 | |
| N3 | 2.741(1.929–3.896) | <0.001 | 2.378(1.661–3.405) | <0.001 | 2.948(2.885–4.169) | <0.001 | 2.530(1.774–3.607) | <0.001 | |
| M Classification | |||||||||
| M0 | reference | reference | reference | reference | |||||
| M1 | 4.611(3.105–6.846) | <0.001 | 3.343(2.228–5.015) | <0.001 | 4.611(3.105–6.846) | <0.001 | 3.421(2.281–5.132) | <0.001 | |
| Death | T Classification | <0.001 | <0.001 | <0.001 | <0.001 | ||||
| T1 | reference | reference | reference | reference | |||||
| T2 | 1.656(0.986–2.780) | 0.056 | 1.641(0.974–2.764) | 0.063 | 1.797(1.088–2.967) | 0.022 | 1.737(1.048–2.877) | 0.032 | |
| T3 | 2.595(1.620–4.156) | <0.001 | 2.459(1.520–3.977) | <0.001 | 2.687(1.713–4.216) | <0.001 | 2.489(1.572–3.942) | <0.001 | |
| T4 | 4.135(2.549–6.706) | <0.001 | 4.017(2.442–6.608) | <0.001 | 4.283(2.695–6.806) | <0.001 | 4.038(2.503–6.513) | <0.001 | |
| N Classification | <0.001 | <0.001 | <0.001 | <0.001 | |||||
| N0 | 0.497(0.357–0.691) | <0.001 | 0.488(0.349–0.682) | <0.001 | 0.563(0.400–0.792) | 0.001 | 0.563(0.398–0.796) | 0.001 | |
| N1 | reference | reference | reference | reference | |||||
| N2 | 1.200(0.944–1.526) | 0.136 | 1.246(0.977–1.589) | 0.076 | 1.307(1.032–1.655) | 0.026 | 1.359(1.070–1.727) | 0.012 | |
| N3 | 2.761(1.986–3.836) | <0.001 | 2.478(1.766–3.478) | <0.001 | 3.007(2.170–4.166) | <0.001 | 2.696(1.925–3.776) | <0.001 | |
| M Classification | |||||||||
| M0 | reference | reference | reference | reference | |||||
| M1 | 5.098(3.524–7.374) | <0.001 | 3.576(2.443–5.234) | <0.001 | 5.098(3.524–7.374) | <0.001 | 3.653(2.496–5.345) | <0.001 | |
| Loco-regional relapse | T Classification | 0.114 | 0.135 | 0.071 | 0.079 | ||||
| T1 | reference | reference | reference | reference | |||||
| T2 | 2.053(1.138–3.705) | 0.017 | 2.026(1.118–3.674) | 0.02 | 2.110(1.202–3.706) | 0.009 | 2.100(1.189–3.707) | 0.011 | |
| T3 | 1.718(0.980–3.011) | 0.059 | 1.747(0.982–3.109) | 0.058 | 1.702(1.002–2.891) | 0.049 | 1.736(1.006–2.998) | 0.048 | |
| T4 | 1.590(0.855–2.955) | 0.143 | 1.629(0.857–3.098) | 0.137 | 1.575(0.871–2.846) | 0.133 | 1.604(0.866–2.972) | 0.133 | |
| N Classification | 0.035 | 0.056 | 0.087 | 0.169 | |||||
| N0 | 0.691(0.459–1.040) | 0.077 | 0.673(0.442–1.023) | 0.064 | 0.831(0.549–1.258) | 0.381 | 0.823(0.538–1.260) | 0.37 | |
| N1 | reference | reference | reference | reference | |||||
| N2 | 1.043(0.744–1.463) | 0.807 | 1.035(0.735–1.456) | 0.845 | 1.146(0.821–1.599) | 0.423 | 1.129(0.806–1.582) | 0.481 | |
| N3 | 1.547(0.898–2.664) | 0.116 | 1.479(0.849–2.578) | 0.167 | 1.699(0.990–2.917) | 0.054 | 1.606(0.924–2.792) | 0.093 | |
| M Classification | |||||||||
| M0 | reference | reference | reference | reference | |||||
| M1 | 2.084(0.978–4.441) | 0.057 | 1.695(0.783–3.667) | 0.18 | 2.084(0.978–4.441) | 0.057 | 1.714(0.793–3.706) | 0.171 | |
Abbreviation: HR = Hazard Ratio, derived from COX proportional hazard model.