Discrepancies between preoperative assessment and final pathological criteria in early gastric cancer

Table 1 Discrepancy between the pre-endoscopic submucosal dissection indications and the post-endoscopic submucosal dissection pathological criteria, n (%)

Variable	Total	Post-ESD criteria
		Absolute criteria	Expand criteria	Surgery criteria
Pre-ESD indication
Absolute indication	407	256 (62.9)	40 (9.9)	109 (26.8)
Surgery indication	25	2 (8.0)	2 (8.0)	21 (84.0)
Total	430	258 (59.7)	44 (9.8)	130 (30.1)

ESD: Endoscopic submucosal dissection.

Table 2 Discrepancy between the pre-surgery indications and the post-surgery pathological criteria, n (%)

Variable	Total	Post-surgery criteria
		ESD criteria	Surgery criteria
Pre-surgery indication
Surgery indication	234	38 (10.40)	196 (53.60)
ESD indication	132	82 (22.40)	50 (3.60)
Total	366	120 (32.80)	246 (67.20)

ESD: Endoscopic submucosal dissection.

Table 3 Comparison of baseline characteristics according to assessment discrepancy status, n (%)

Variable	Same (n = 592)	Under (n = 162)	Over (n = 42)	χ2 value	P value
Age				2.097	0.350
≤ 60	228 (38.5)	52 (32.7)	14 (33.3)
> 60	364 (61.5)	107 (67.3)	28 (66.7)
Sex				0.332	0.847
Male	411 (69.4)	116 (71.6)	30 (71.4)
Female	181 (30.6)	46 (28.4)	12 (28.6)
History of smoking				6.215	0.044
No	391 (66.0)	90 (55.6)	27 (64.3)
Yes	201 (34.0)	72 (44.4)	15 (35.7)
History of alcohol consumption				2.732	0.255
No	432 (73.0)	110 (67.9)	26 (65.0)
Yes	160 (27.0)	52 (32.1)	14 (35.0)
Family history				4.436	0.109
No	510 (86.1)	148 (91.4)	34 (81.0)
Yes	82 (13.9)	14 (8.6)	8 (19.0)
Helicobacter pylori infection				3.220	0.200
No	421 (71.1)	104 (64.2)	31 (73.8)
Yes	171 (28.9)	58 (35.8)	11 (26.2)

For same-estimated cases, postoperative management aligned with preoperative assessment, underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively. Same: Same-estimated; Under: Under-estimated; Over: Over-estimated.

Table 4 Univariate analysis of factors associated with preoperative-postoperative assessment discrepancies, n (%)

Variable	Same (n = 592)	Under (n = 162)	Over (n = 42)	χ2 value	P value
Pre-endoscopic findings
Location				8.132	< 0.001
Upper	111 (18.8)	65 (38.3)	12 (30.0)
Middle	232 (39.2)	51 (31.5)	14 (32.5)
Lower	249 (42.1)	49 (30.2)	16 (37.5)
Size (cm)				8.590	0.072
≤ 2 cm	401 (67.7)	112 (69.1)	23 (57.5)
> 2 cm and ≤ 3 cm	122 (20.6)	37 (22.8)	9 (22.5)
> 3 cm	69 (11.7)	12 (8.0)	10 (20.0)
Gross type				2.204	0.698
Elevated	282 (47.6)	74 (45.7)	16 (38.1)
Flat	220 (37.2)	65 (40.1)	20 (47.6)
Depressed	90 (15.2)	23 (14.2)	6 (14.3)
Presence of ulcers				1.761	0.415
No	423 (71.5)	117 (72.2)	34 (81.0)
Yes	169 (28.5)	45 (27.8)	8 (19.0)
Pathology				53.839	< 0.001
Differentiated	411 (69.9)	156 (96.3)	24 (57.1)
Undifferentiated	178 (30.1)	6 (3.7)	18 (42.9)

For same-estimated cases, postoperative management aligned with preoperative assessment, underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively. Same: Same-estimated; Under: Under-estimated; Over: Over-estimated.

Table 5 Multivariate analysis of factors associated with preoperative-postoperative assessment discrepancies

Variable		P value	Exp(B)	95% confidence interval
				Lower	Upper
Under-estimated
History of smoking	Yes	0.108	1.357	0.935	1.969
	No	Reference
Location	Upper	< 0.001	2.158	1.373	3.390
	Middle	0.536	1.151	0.738	1.794
	Lower	Reference
Pathology	Undifferentiated	< 0.001	0.104	0.045	0.241
	Differentiated	Reference
Over-estimated
History of smoking	Yes	0.871	0.945	0.483	1.846
	No	Reference
Location	Upper	0.790	1.984	0.888	4.435
	Middle	0.851	0.916	0.436	1.923
	Lower	Reference
Pathology	Undifferentiated	0.416	2.005	1.036	3.879
	Differentiated	Reference

The variable reference category is the same-estimated. For same-estimated cases, postoperative management aligned with preoperative assessment, underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively.

Table 6 Multivariate analysis of factors associated with preoperative-postoperative assessment discrepancies

Variable		P value	Exp(B)	95% confidence interval
				Lower	Upper
Under-estimated
History of smoking	Yes	0.331	1.436	0.692	2.980
	No	Reference
Location	Upper	0.851	1.087	0.454	2.602
	Middle	0.592	1.257	0.545	2.896
	Lower	Reference
Pathology	Undifferentiated	< 0.001	0.052	0.018	0.146
	Differentiated	Reference

The variable reference category is the over-estimated. Underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively.

Table 7 Comparison of preoperative biopsy and postoperative pathological findings, n (%)

Postoperative pathology	Pretreatment diagnosis	n = 796	Misdiagnosis rate (%)
Tumor diameter
≤ 3 cm	≤ 3 cm (accurate)	626 (78.6)
	> 3 cm (misdiagnosis)	33 (4.2)	14.0
> 3 cm	≤ 3 cm (misdiagnosis)	78 (9.8)
	> 3 cm (accurate)	59 (7.4)
Depth
Mucosal invasion	Mucosal invasion or submucosal invasion: < 500 μm (accurate)	577 (72.2)
	Submucosal invasion: ≥ 500 μm (misdiagnosis)	37 (4.7)	25.3
Submucosal invasion	Mucosal invasion or submucosal invasion: < 500 μm (misdiagnosis)	164 (20.6)
	Submucosal invasion: ≥ 500 μm (accurate)	20 (2.5)
Ulcerative findings
Absence	Absence (accurate)	217 (27.1)
	Presence (misdiagnosis)	33 (4.1)	4.9
Presence	Absence (misdiagnosis)	6 (0.8)
	Presence (accurate)	542 (68.0)
Main histology
Differentiated type	Differentiated type (accurate)	499 (62.7)	17.1
	Undifferentiated type (misdiagnosis)	41 (5.2)
Undifferentiated type	Differentiated type (misdiagnosis)	95 (11.9)
	Undifferentiated type (accurate)	161 (20.2)

Table 8 Univariate analysis of pathological features among three preoperative assessment groups, n (%)

Variable	Same (n = 592)	Under (n = 162)	Over (n = 42)	χ2 value	P value
Diagnosis of tumor diameter				15.067	< 0.001
Accurate diagnosis	526 (88.9)	126 (77.8)	33 (82.5)
Misdiagnosis	66 (11.1)	36 (22.2)	9 (21.4)
Diagnosis of depth				76.933	< 0.001
Accurate diagnosis	485 (81.9)	78 (48.1)	32 (76.2)
Misdiagnosis	107 (18.1)	84 (51.9)	10 (23.8)
Diagnosis of ulcerative findings				16.047	< 0.001
Accurate diagnosis	571 (96.5)	151 (93.2)	35 (82.5)
Misdiagnosis	21 (3.5)	11 (6.8)	7 (17.5)
Diagnosis of main histology
Accurate diagnosis	548 (92.6)	85 (52.5)	27 (64.3)	155.227	< 0.001
Misdiagnosis	44 (7.4)	77 (47.5)	15 (35.7)

Same: Same-estimated; Under: Under-estimated; Over: Over-estimated. For same-estimated cases, postoperative management aligned with preoperative assessment, underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively.

Table 9 Multivariate analysis of pathological discrepancies among three preoperative assessment groups

Variable		P value	Exp(B)	95% confidence interval
				Lower	Upper
Under-estimated
Diagnosis of tumor diameter	Accurate diagnosis	Reference
	Misdiagnosis	< 0.001	2.555	1.515	4.309
Diagnosis of depth	Accurate diagnosis	Reference
	Misdiagnosis	< 0.001	4.641	3.049	7.064
Diagnosis of ulcerative findings	Accurate diagnosis	Reference
	Misdiagnosis	0.008	3.065	1.346	6.979
Diagnosis of main histology	Accurate diagnosis	Reference
	Misdiagnosis	< 0.001	11.220	7.030	17.908
Over-estimated
Diagnosis of tumor diameter	Accurate diagnosis	Reference
	Misdiagnosis	0.039	2.339	1.044	5.242
Diagnosis of depth	Accurate diagnosis	Reference
	Misdiagnosis	0.372	1.414	0.660	3.030
Diagnosis of ulcerative findings	Accurate diagnosis	Reference
	Misdiagnosis	< 0.001	6.805	2.612	17.729
Diagnosis of main histology	Accurate diagnosis	Reference
	Misdiagnosis	< 0.001	7.691	3.774	15.797

Table 10 Multivariate analysis of pathological discrepancies among three preoperative assessment groups

Variable		P value	Exp(B)	95% confidence interval
				Lower	Upper
Under-estimated
Diagnosis of tumor diameter	Accurate diagnosis	Reference
	Misdiagnosis	0.838	1.092	0.469	2.544
Diagnosis of depth	Accurate diagnosis	Reference
	Misdiagnosis	0.003	3.281	1.504	7.156
Diagnosis of ulcerative findings	Accurate diagnosis	Reference
	Misdiagnosis	0.135	0.450	0.158	1.282
Diagnosis of main histology	Accurate diagnosis	Reference
	Misdiagnosis	0.307	1.459	0.707	3.010

The variable reference category is the over-estimated. Underestimated and overestimated groups represented cases requiring therapeutic escalation or de-escalation, respectively.