Accurate FDG PET tumor segmentation using the peritumoral halo layer method: a study in patients with esophageal squamous cell carcinoma

Table 2 Comparison of endoscopic length, pathologic length, and frequency of metastasis according to FDG avidity of the primary ESCC

	FDG-avid (n = 85)	FDG-non-avid (n = 36)	P value
Endoscopic passage to distal portion of the tumor* (possible/impossible)	81/2	33/0	1.0000
Endoscopic length (cm)**			0.0042^†
Median, range	3.5, 1.0–9.0	2.0, 0.5–9.0
Pathologic length (cm)***			0.0001^†
Median, range	3.5, 0.9–7.6	1.7, 0.15–6.0
Metastatic LN on FDG PET/CT (yes/no)	40 / 45	5 / 31	0.0005^†
M1a metastatic LN or distant metastasis on FDG PET/CT (yes/no)	21 / 64	1 / 35	0.0037^†
Presence of metastatic LN on pathologic specimens**** (yes/no)	18 / 19	5 / 16	0.0941
Pathologic depth of invasion (T stage)***			< 0.0001^†
Basement membrane (Tis)	0	3
Mucosa (T1a)	2	9
Submucosa (T1b)	11	11
Muscularis propria (T2)	9	0
Adventitia (T3)	15	0
Invasion of adjacent structures (T4)	0	0

*Pretreatment endoscopic evaluation was performed in 116 patients at our hospital
**Measurement of endoscopic length was impossible in 2 patients with total esophageal obstruction
***Measurement of pathologic length and depth of invasion was possible for 60 patients who were treated by surgery (n = 58) or ESD (n = 2). There was no T4 case in these patients
****Frequency test of pathologically confirmed metastatic LNs was performed for 58 operative patients
LN = lymph node; ESD = endoscopic submucosal dissection
^†Statistically significant

ISSN: 1470-7330