Table 2. Multiple logistic regression models assessed the correlation between risk factors and ILN metastasis.

Exposure		Non-adjusted	Adjust I	Adjust II
		HR,95%CI P value	HR,95%CI P value	HR,95%CI P value
Maxium_ILN
	<1.5cm	1	1	1
	≥1.5cm	2.3 (1.0, 5.1) 0.047	2.4 (1.1, 5.6) 0.035	10.7 (2.1, 53.3) 0.004
T_stage
	Ta/T1	1	1	1
	T2 and above	3.0 (1.3, 7.1) 0.011	3.1 (1.3, 7.5) 0.010	7.1 (1.7, 28.9) 0.006
Differentiation
	high	1	1	1
	low-middle	3.6 (1.4, 8.8) 0.006	4.0 (1.5, 10.4) 0.004	6.2 (1.9, 20.2) 0.003
LVI
	no	1	1	1
	yes	3.6 (1.1, 11.6) 0.032	3.6 (1.1, 11.8) 0.033	7.4 (1.3, 40.8) 0.022

The data in the table: β (95%CI) P value/OR (95%CI) P value

Outcome variable: metastasis Exposed variables:

1) Maximum_ILN

Non-adjusted model adjusted for: None

Adjust I model adjust for: age; HBP; diabetes; CDV

Adjust II model adjusted for: model1+onset_time; T_stage; differentiation; LVI; umber_ILN

2) T_stage

Non-adjusted model adjusted for: None

Adjust I model adjusted for: age; HBP; diabetes; CDV

Adjust II model adjusted for: model1+onset_time; Number_ILN; Maximum_ILN; differentiation; LVI; nerve infiltration

3) Differentiation

Non-adjusted model adjusted for: None

Adjust I model adjusted for: age; HBP; diabetes; CDV

Adjust II model adjusted for: model1+onset_time; Number_ILN; Nerve_infiltration

4) LVI

Non-adjusted model adjusted for: None

Adjust I model adjusted for: age; HBP; diabetes; CDV

Adjust II model adjusted for: model1+onset_time, Number_ILN, Maximum_ILN, T_stage, Nerve_infiltration, differentiation

ILN, inguinal lymph nodes; LVI, lymphatic vascular infiltration.