Prediction of serosal invasion in gastric cancer: development and validation of multivariate models integrating preoperative clinicopathological features and radiographic findings based on late arterial phase CT images

Table 1 Univariate analysis of demographic data, endoscopic biopsy, and morphological characteristics in the primary and validation cohorts

Characteristics	Primary cohort		P value	Validation cohort		P value
Characteristics	T1–3 (n = 78)	T4 (n = 25)	P value	T1–3 (n = 40)	T4 (n = 11)	P value
Demographic data
Gender			0.154			0.214
Male	58	22		23	4
Female	20	3		17	7
Age (y)			0.543			1.000
< 60	29	11		13	3
≥ 60	49	14		27	8
Endoscopic biopsy
Differentiation degree			0.004*			0.141
Poor	48	23		26	10
Moderate & Well	30	2		14	1
Morphological characteristics
Major location			0.884			0.690
Cardia	21	6		14	2
Body	27	8		14	5
Antrum	30	11		12	4
Tumor range			0.121			0.009*
1 location	51	12		32	4
≥ 2 locations	27	13		8	7
Major orientation			0.637			0.591
Lesser curvature	35	11		19	7
Greater curvature	2	2		3	0
Anterior wall	7	2		2	1
Posterior wall	34	10		16	3
Circumferential range			0.428			0.010*
1/4	34	7		24	1
2/4	26	11		10	7
3/4	11	3		3	1
4/4	7	4		3	2
Infiltrative growth			0.009*			0.001*
Absent	54	10		30	2
Present	24	15		10	9
Ulceration			0.036*			0.305
Absent	20	12		20	8
Present	58	13		20	3
Adjacent adipose tissue stains			0.267			0.009*
Absent	56	15		25	2
Present	22	10		15	9
Mucosal line status			0.280			0.002*
Interruption	28	12		11	9
Thickening	50	13		29	2
Morphological type			0.026*			0.037*
Thickening type	12	9		11	7
Mass type	66	16		29	4
Linitis plastica			0.091			0.292
Absent	76	22		37	9
Present	2	3		3	2
Lymphadenectasis			1.000			0.598
Absent	66	21		36	9
Present	12	4		4	2

ISSN: 1471-2407