Prognostic value and interobserver agreement of endoscopic ultrasonography for superficial squamous cell carcinoma of the esophagus: a prospective study

Int J Gastrointest Cancer. 2003;34(1):1-8. doi: 10.1385/IJGC:34:1:01.

Abstract

Background and aims: Submucosal invasion of superficial esophageal cancer (SEC) is related to the prognosis. We prospectively analyzed outcomes of SEC in patients staged by endoscopic ultrasonography (EUS).

Patients and methods: We staged 31 endoscopically diagnosed SEC cases using a 20/15-MHz thin probe. The EUS tumor stage was classified as EUSM (limited within mucosa), EUS-SM (with submucosal invasion), or EUS-MP over (invading the muscularis propria or deeper). Lymph node metastasis and 2-yr survival were analyzed according to the EUS tumor stage in 29 squamous cell carcinoma cases. Interobserver agreement of the EUS stage was tested between the examiner and a blind reviewer.

Results: Lymph node metastasis was significantly frequent in the EUS-SM group (8 of 18 cases [44.4%]) compared with the EUS-M group (1 of 10 cases [10%]) (p = 0.03). Patient survival at 2 yr after initial therapy was 72.2% in the EUS-SM group and 90% in the EUS-M group. Death from cancer was noted only in the EUS-SM group (three cases). The accuracy rates of EUS tumor staging were 74.1% by the examiner and 66.7% by the blind reviewer, with moderate interobserver agreement (kappa = 0.46).

Conclusions: Thin-probe EUS can classify SEC into two groups: the EUS-M group with excellent outcome and the EUS-SM group with a significant risk of lymph node metastasis.

Publication types

  • Clinical Trial

MeSH terms

  • Aged
  • Aged, 80 and over
  • Carcinoma, Squamous Cell / diagnostic imaging*
  • Carcinoma, Squamous Cell / pathology*
  • Endosonography*
  • Esophageal Neoplasms / diagnostic imaging*
  • Esophageal Neoplasms / pathology*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Neoplasm Invasiveness
  • Neoplasm Staging / methods*
  • Observer Variation
  • Prognosis
  • Prospective Studies
  • Sensitivity and Specificity
  • Survival Analysis