Purpose: To determine the pathologic complete response rate, toxicity, and survival of patients with resectable squamous cell or adenocarcinoma of the esophagus treated with a 30-day preoperative chemoradiation regimen and surgical resection.
Patients and methods: Fifty patients (16 squamous, 33 adeno, one undifferentiated) who had carcinoma of the esophagus (limited to the primary tumor and regional or celiac nodes) were treated with cisplatin 26 mg/m2/day continuous infusion days 1 through 5 and 26 through 30, 5-fluorouracil (5-FU) 300 mg/m2/day continuous infusion days 1 through 30, and radiation 44 Gy, 2 Gy/fx in 22 daily fractions, days 1 through 30, followed by esophagectomy.
Results: Forty-seven patients underwent esophagectomy (94% operability rate), and 45 had total gross removal of disease and negative margins of resection (90% resectability rate). Nineteen patients (40%) had a pathologic complete response (CR). Forty (80%) received 100% of the planned cisplatin dose, 29 (58%) received 100% of the planned 5-FU dose, and 40 (80%) received > or = 80% of the planned 5-FU dose. Forty-five (90%) received the planned 44-Gy radiation dose. Grade 3 or 4 neutropenia occurred in 60% of patients. The incidence of febrile neutropenia was 34%. There was one septic death during chemoradiation and no operative deaths. Weight loss requiring nutritional support occurred in 50% of patients, secondary to anorexia, dysphagia, and/or esophagitis. The survival of all registered patients at a median follow-up of 43 months was 2-year survival 58%, median 31.3 months. Survival analysis by histology showed no difference between the two histologic types (squamous vs adenocarcinoma). However, survival by pathologic response was significantly different: pathologic CR, 19 patients, 2-year survival 78%, median survival 58 months; and pathology positive, 28 patients, 2-year survival 46%, median survival 22.4 months. A Cox proportional hazards model and logistic regression analysis demonstrated a significant survival advantage for pathologic CRs and stage I disease versus higher-stage disease and a correlation between chemotherapy dose received and pathologic staging.
Discussion: This 30-day chemoradiation regimen followed by surgery resulted in a high pathologic complete response rate, 40%, and apparent survival advantage for this group. The median survival rate of 31.3 months and 2-year survival rate of 58% suggest that this regimen may improve survival over surgical treatment alone. Randomized trials with large accrual and statistical power are necessary to confirm our results and to determine optimal treatment.