Aim: Purpose of this study was to assess prognostic value of endoscopic ultrasound in patients with inoperable squamous cell carcinoma of oesophagus treated by radio-chemotherapy.
Patients and methods: Between January 1993 and March 1996, 89 patients (77 males and 12 females, mean age 60.3 years) with squamous cell carcinoma of the oesophagus were treated exclusively by radio-chemotherapy consisting of 3 courses of chemotherapy using 5FU-Cis-platyl and 3 courses of radiation therapy (3 x 15 Gy). Endoscopy and endoscopic ultrasound (Pentax FG 32 UA) were performed before beginning treatment and two weeks after last cycle of radio-chemotherapy. Classical criteria for endoscopic ultrasound lymph node metastases were used after irradiation; response was considered as complete only if endoscopic ultrasound indicated that integrity of oesophageal wall was fully restored.
Results: Complete endoscopic ultrasound assessment was achieved in 73 cases (84.9%). Tumours were classified as T1N1 in 1 case, T2N1 in 7, T3N0 in 4, T3N1 in 24, T4N0 in 1 and T4N1 in 49. For patients with a non-invasive tumour (usT1 or T2), malignancy of lymph nodes was proved by endoscopic ultrasound guided biopsy. Eighty-two patients presented one or more suspicious lymph nodes. Metastatic lymph nodes were located in posterior mediastinum in 43 cases, at distant sites in 27 (laterotracheal in 16 and coeliac in 11) and in 16 lymph nodes were located simultaneously in mediastinum and at distant sites. Median overall survival in these 89 patients was 16 months. There was no significant difference in median survival between patients in stage T3 and T4. Conversely, there was a significant difference between patients with more or less than 4 metastatic lymph nodes (9 vs 36 months, respectively, p = 0.005). Site of lymph node metastasis was also a prognostic factor with better survival in patients presenting mediastinal nodes than those presenting coeliac nodes (30 vs 9 months, respectively, p < 0.0001). Median survival was also significantly better in patients considered as having achieved a complete response by both gastroduodenal fibrescopy and endoscopic ultrasound than in those considered to have a complete response by gastroduodenal fibrescopy but not by endoscopic ultrasound (49 vs 10 months). Conversely, there was no difference in survival in function of treatment response assessment by thoracic chemotherapy-scan.
Conclusion: Endoscopic ultrasound findings regarding number and site of suspicious lymph nodes and degree of treatment response are significant prognostic factors in patients with squamous cell carcinoma of oesophagus treated exclusively by radio-chemotherapy.