Table 2. Univariate and multivariate analyses of associations (A) between EZH2 expression and clinicopathological characteristics and (B) between Vimentin expression and clinicopathological characteristics.

(A)		Univariate analysis		Multivariate analysis		(B)		Univariate analysis		Multivariate analysis
Factors		OR	P value	OR	P value	Factors		OR	P value	OR	P value
Age	<69	1.11	0.625			Age	<69	1.00
(years)	≥69	1.00				(years)	≥69	1.30	0.315
Sex	Female	1.00				Sex	Female	1.27	0.362
	Male	1.58	0.033				Male	1.00
Smoking	Never	1.00				Smoking	Never	1.00
status	Smoker	2.06	<0.001			Status	Smoker	1.08	0.766
Pathological	I	1.00				Pathological	I	1.00		1.00
stage	≥II	3.41	<0.001			stage	≥II	2.64	<0.001	2.15	0.007
pl	Absent	1.00				pl	Absent	1.00
	Present	2.52	<0.001				Present	2.49	0.002
ly	Absent	1.00				ly	Absent	1.00
	Present	3.76	<0.001				Present	2.59	0.0053
v	Absent	1.00		1.00		v	Absent	1.00
	Present	6.66	<0.001	6.18	<0.001		Present	2.32	0.002
EGFRa	Wildtype	1.49	0.150			EGFRa	Wildtype	1.00
	Mutant	1.00					Mutant	1.12	0.745
Vimentin	Negative	1.00		1.00		E-cadherin	Negative	1.77	0.030
	Positive	2.73	<0.001	2.28	0.006		Positive	1.00
E-cadherin	Negative	1.25	0.302			EZH2	Negative	1.00		1.00
	Positive	1.00					Positive	2.23	0.004	2.23	0.004

^acases in which data were available.

CI, confidence interval; EGFR, epidermal growth factor receptor; EMT, epithelial-mesenchymal transition; EZH2, Enhancer of Zeste Homolog 2; ly/v, lymphovascular invasion; OR, odds ratio; pl, pleural invasion