Skip to main content

Table 4 Cox proportional hazards regression model for overall survival in hypopharyngeal and laryngeal squamous cell carcinoma

From: Prognostic implications of Fibroblast growth factor receptor 1 (FGFR1) gene amplification and protein overexpression in hypopharyngeal and laryngeal squamous cell carcinoma

Variables Univariate Multivariate
P value HR (95% CI) P value HR (95% CI)
≥ 65 years 0.009 2.102 (1.207–3.663) 0.035 1.835 (1.042–3.231)
Male 0.503 1.493 (0.463–4.816)   
Primary sites
 Larynx   1 (reference)   1 (reference)
 Hypopharynx 0.042 1.812 (1.022–3.213) 0.702 1.142 (0.580–2.248)
Smoking
 Never smoker   1 (reference)   
 Smoker 0.621 1.263 (0.500–3.190)   
Alcohol
 Non- or social- drinker   1 (reference)   
 Heavy drinker 0.160 1.772 (0.798–3.937)   
Differentiation
 Well   1 (reference)   1 (reference)
 Moderate 0.007 2.885 (1.343–6.199) 0.033 2.422 (1.076–5.452)
 Poor 0.026 3.180 (1.147–8.819) 0.422 1.398 (0.532–4.513)
Lymphovascular invasion 0.040 1.842 (1.029–3.296) 0.569 0.819 (0.411–1.632)
Perineural invasion 0.002 2.608 (1.411–4.821) 0.098 1.825 (0.896–3.720)
Positive resection margin 0.006 2.236 (1.265–3.951) 0.002 2.712 (1.425–5.161)
pT-classification
 pT1   1 (reference)   
 pT2 0.358 1.464 (0.649–3.301)   
 pT3 0.032 1.715 (1.076–5.171)   
 pT4a 0.011 2.224 (1.286–6.928)   
pN-classification
 pN0   1 (reference)   
 pN1 0.085 2.385 (0.888–6.409)   
 pN2 0.005 2.982 (1.382–6.431)   
 pN3 0.001 3.003 (1.578–5.714)   
Pathological stage
 Stage I   1 (reference)   1 (reference)
 Stage II 0.482 1.561 (0.451–5.402) 0.457 1.625 (0.452–5.847)
 Stage III 0.011 3.890 (1.361–11.117) 0.048 3.018 (1.010–9.018)
 Stage IV 0.004 4.102 (1.591–10.581) 0.046 2.967 (1.018–8.647)
Adjuvant treatment 0.615 1.154 (0.661–2.013)   
p16 positivity 0.127 0.331 (0.080–1.368)   
FGFR1 amplification 0.252 2.128 (0.585–7.738)   
FGFR1 high expression 0.319 1.504 (0.674–3.354)   
  1. aAbbreviations: CI Confidence interval, HR Hazard ratio
  2. bVariables of the pT and pN classification were not included in the multivariate analysis, because they were included in the pathologic stage