. 2016 Jul 21;16:514. doi: 10.1186/s12885-016-2555-z

Table 3.

Relationship between expression of CXCL12/CXCR4 and clinicopathological factors

	All patients (n = 79) (%)
	CXCR4(+)	CXCR4(−)	p	CXCL12(+)	CXCL12(−)	p
	n = 48 (61)	n = 31 (39)	p	n = 62 (78)	n = 17 (22)	p
Age, median (range)	58.5 (44–79)	61 (44–74)	0.84^c	59 (44–79)	57 (50–70)	0.84^c
Gender
Male	44	27	0.71^a	54	17	0.19^a
Female	4	4	0.71^a	8	0	0.19^a
Histological grade
1	11	5	0.69^b	12	4	0.90^b
2	34	23		45	12
3	3	3		5	1
Pathological T
Tis	5	5	0.63^b	8	2	0.39^b
T1	24	11		28	7
T2	4	3		7	0
T3	15	12		19	8
Pathological N
Positive	28	18	0.98^b	36	10	0.96^b
Negative	20	13	0.98^b	26	7	0.96^b
Lymphatic invasion
Absent	13	8	0.90^b	17	4	1.00^a
Present	35	23	0.90^b	45	13	1.00^a
Vessel invasion
Absent	27	17	0.90^b	36	8	0.42^b
Present	21	14	0.90^b	26	9	0.42^b
Recurrence of lymph node metastasis
Positive	13	11	0.43^b	22	2	0.08^a
Negative	35	20	0.43^b	40	15	0.08^a
Recurrence of distant organ metastasis
Positive	13	7	0.65^b	17	3	0.54^a
Negative	35	24	0.65^b	45	14	0.54^a

Recurrence of lymph node metastasis or distant organ metastasis means that the initial recurrence region after ESCC resection was lymph node or distant organ

^aFisher’s exact probability test

^bPearson’s χ-square test

^cMann-Whitney U test