Table 4.
Univariate | Multivariate | |||||
---|---|---|---|---|---|---|
Variables | HR | 95% CI | P‐value | aHR | 95% CI | P‐value |
Extent of resection (SPSR) | 1.026 | 0.383–2.748 | 0.958 | 0.852 | 0.276–2.897 | 0.852 |
Inappropriate margin of SPSR | 3.736 | 1.296–10.766 | 0.015 | 2.554 | 0.601–10.854 | 0.204* |
Age | 1.035 | 0.981–1.091 | 0.211 | – | – | – |
Gender (male) | 2.209 | 0.737–5.590 | 0.171 | – | – | – |
Smoking history (smoker) | 3.556 | 1.331–9.502 | 0.011 | 2.127 | 0.602–7.511 | 0.241 |
Preoperative CEA level (abnormal) | 1.043 | 1.011–1.076 | 0.007 | 1.003 | 0.964–1.042 | 0.898 |
Charlson comorbidity score | 1.275 | 0.969–1.677 | 0.082 | 1.091 | 0.735–1.619 | 0.666 |
Maximum tumor dimension | 2.408 | 0.996–5.825 | 0.051 | 0.234 | 0.021–2.620 | 0.239 |
Radiologic appearance (solid‐dominant) | 3.582 | 1.239–10.354 | 0.018 | 2.125 | 0.536–8.423 | 0.283 |
Preoperative PET/CT (without) | 1.172 | 0.435–3.160 | 0.753 | – | – | – |
Surgical method (VATS) | 0.821 | 0.107–6.295 | 0.849 | – | – | – |
Clinical stage | 2.143 | 1.261–2.143 | 0.005 | 2.213 | 0.555–8.823 | 0.260 |
Clinical characteristics | ||||||
Pathological stage | 1.299 | 0.842–2.004 | 0.237 | – | – | – |
Pleural invasion (PL2) | 9.320 | 3.333–26.067 | <0.001 | 1.916 | 0.422–8.710 | 0.400 |
Histology differentiation (poorly) | 5.794 | 2.089–16.066 | 0.001 | 1.504 | 0.299–7.572 | 0.621 |
Angiolymphatic invasion (yes) | 16.236 | 5.878–44.849 | <0.001 | 7.114 | 1.295–39.076 | 0.024 |
Predominate pattern group (high grade) | 7.506 | 2.781–20.254 | <0.001 | 1.130 | 0.219–5.832 | 0.884 |
Lymph node sampling (<3) | 2.901 | 0.656–12.821 | 0.160 | – | – | – |
Lymph node station (<15) | 1.383 | 0.499–3.833 | 0.533 | – | – | – |
Adjuvant chemotherapy (yes) | 1.867 | 0.602–5.793 | 0.280 | – | – | – |
Adjusted with smoking history, preoperative CEA level, Charlson comorbidity score, maximum tumor dimension, radiologic appearance, pleural invasion (PL2), histology differentiation, angiolymphatic invasion and predominate pattern.
Calculated by Cox regression method; only variables with P ≤ 0.1 after the univariate analyses were entered into the multivariate model;
SPSR, surgeons' preference sublobar resection.