Background: Conflicting results of preoperative radiochemotherapy in patients with esophageal cancer have been obtained; only patients with a complete pathological response seem to benefit from this therapy. However, there is evidence that preoperative radiochemotherapy leads to considerable postoperative morbidity. Therefore, postoperative morbidity was retrospectively investigated in 82 patients with an esophageal cancer who received preoperative radiochemotherapy.
Methods: One hundred twenty-two consecutively operated on patients were included (1991 to 2001). Preoperative radiochemotherapy was initiated in 1994 for cT >1, cNx, cM0 regardless of histology (n = 82); 36 Gy was applied (1.8 Gy daily, days 1 to 5, weeks 1 to 4), concurrently 5-fluorouracil (500 mg/m(2) days 1 to 5, weeks 1 to 4), and cisplatin (20 mg/m(2) days 1 to 5, weeks 1 and 4). Postoperative morbidity was categorized as surgery- and nonsurgery-related morbidity. Survival was calculated by the Kaplan-Meier method. Results were stratified into histology and compared with patients who were operated on only (n = 40).
Results: Complete pathological response after preoperative radiochemotherapy was achieved in 22%. An increase in surgery-related morbidity was observed after preoperative radiochemotherapy due to lesion of recurrent nerve (38% versus 12.5%, P = 0.009), as well as a marked difference in pulmonary morbidity (57% versus 37.5%, P = 0.05). The proportion of combined morbidity was increased after preoperative radiochemotherapy (49.4% versus 15%, P = 0.02), which led to a considerable prolongation of postoperative hospital stay (33 versus 21 days median, P = 0.0022). Patients with a longer postoperative hospital stay (>30 days; 43.2%) lived significantly shorter than patients with a shorter postoperative hospital stay (56.8%, P = 0.001). There was no statistical survival benefit in the neoadjuvant treated group. However, calculation of long-term survival revealed a significant survival advantage in patients with squamous cell cancer and a complete pathological response compared with patients without response (median 642 days versus 302, P = 0.026).
Conclusions: Perioperative morbidity was significantly increased after preoperative radiochemotherapy. Long-term survival was clearly affected by the length of postoperative stay. Therefore, we need better patient selection for application of preoperative radiochemotherapy.