Diagnostic accuracy of tumor staging and treatment outcomes in patients with superficial esophageal cancer

Background We examined the current status and diagnostic accuracy of currently available techniques for tumor staging and assessed treatment outcomes in patients with superficial esophageal cancer who received esophaguspreserving therapy, such as endoscopic mucosal resection (EMR) alone or combined with chemoradiotherapy (CRT). Methods In 274 patients with superficial esophageal cancer, we examined the depth of tumor invasion and the degree of lymph node metastasis by means of endoscopy, magnifying endoscopy, endoscopic ultrasonography (EUS), computed tomography (CT), and cervical and abdominal ultrasonography (US). We compared treatment outcomes among treatment groups according to the depth of tumor invasion. Results The rates of correctly diagnosing the depth of tumor invasion were 89.6% on conventional endoscopy, 90.1% on magnifying endoscopy, and 85% on scanning with a high-frequency miniature ultrasonic probe (miniature US probe). Diagnostic accuracy for the m3 or sm1 cancers was poor. Magnifying endoscopy allowed invasion to be more precisely estimated, thereby improving diagnostic accuracy. However, lesions that maintained their surface structure despite deep invasion were misdiagnosed on magnifying endoscopy. A miniature US probe was useful for the assessment of such lesions. The diagnostic accuracy of EUS for lymph node metastasis was 83%, with a sensitivity of 76%. The sensitivity of CT was 29%, and that of cervical and abdominal US was 17%. Patients with m1 or m2 cancer had good outcomes after esophagus-preserving therapy. Although there were no significant differences in survival rates, many patients with sm2 or sm3 cancer who received CRT died of their disease. Nodal recurrence was diagnosed by EUS. In patients who received CRT, the time to the detection of recurrence was slightly prolonged. Conclusions Long-term follow-up at regular intervals is essential in patients with m3 or sm esophageal cancers who receive esophagus-preserving treatment. At present, EUS is the most reliable technique for the diagnosis of lymph node metastasis and is therefore essential for pretreatment evaluation as well as for follow-up. Earlier detection of recurrence at a level that would potentially salvage treatment remains a topic for future research. Review articles on this topic also appeared in the previous issue (Volume 4 Number 3). An editorial related to this article is available at .