5-fluorouracil and cisplatin as palliative treatment of advanced oesophageal squamous cell carcinoma: A multicentre randomised controlled trial

Objective: To compare chemotherapy with no chemotherapy as palliative treatment for oesophageal squamous cell carcinoma. Design: Randomised study. Setting: Multicentre trial in France. Subjects: Of 161 patients with histologically confirmed oesophageal squamous cell carcinoma located more than 5 cm from the mouth of the oesophagus, five were withdrawn because of protocol violation. The remaining 156 patients, 149 men and 7 women, mean (SD) age 58 (9) years range 36 to 77, were randomly allocated to either a control group without chemotherapy (n = 84) or a group treated by chemotherapy (n = 72). Patients were divided into four strata: I = complete resection of the tumour but with lymph node involvement (n = 62); II = incomplete resection of tumour leaving gross tumour behind (n = 58); III = no resection because of local or regional invasion (n = 22); and IV = no resection because of distant metastasis (n = 14). Exclusion criteria were histologically confirmed tracheobronchial involvement, oesophagotracheal fistula, Karnosky score < 50, cerebral metastases, or hepatic metastases occupying more than 30% of the liver, peritoneal carcinomatosis, associated or previously treated ear-nose-throat carcinoma, or complete resection of tumour without lymph node involvement. Interventions: 5 fluorouracil (5FU) and cisplatin (CDDP) were given in 5-day courses, once every 28 days, for a maximum of eight cycles. 5 FU, 1 g/m 2, was infused for 24 hours after a water overload, during five days. Cisplatin was given either in one dose of 100 mg/m 2 at the beginning of the cycle or 20 mg/m 2/day over three hours for five days. Duration of treatment ranged from 6-8 months. Outcome measures: Median and actuarial survival. The subsidiary endpoint was quality of survival judged by complications of treatment, swallowing disorders, and the duration of ability to feed normally. Results: There was no difference in survival, either overall (median = 12 months) or in any of the strata. There were however significantly more patients with neurological (p < 0.003), haematological (p < 0.0001), and renal (p < 0.0002) complications in the treated group compared with the control group. Four patients (6%) died of complications of chemotherapy. The course of swallowing disorders did not differ between the two groups. The duration of autonomous oral feeding was exactly the same in both groups (median = 10.5 months). Conclusion: The results suggest that 5FU and CDDP do not help in patients with squamous cell carcinoma of the oesophagus whether or not the tumour has been resected.