目的：評估手術後之合併化學與放射線治療於食道癌的效用，並尋找影響病患預後的因子。材料與方法：自1991年10月至2000年11月共有56位食道癌患者接受手術後放射線治療。其中23位病患接受單獨放射線治療，劑量約為50Gy，另外33位並接受合併化學與放射線治療，使用的藥物為cisplatin與5-FU。兩組病人除腫瘤長度外並無明顯差異。我們比較兩組病患之存活率以評估合併化學與放射線治療的效用，並分析年齡、腫瘤的期別、長度、侵犯情形與陽性淋巴的狀態等對預後的影響。結果：追蹤中值時間為16個月。56位病患的五年全部存活率為20.2%，無疾病存活率為19.1%。其中14.3%的患者發生局部復發，35.7%的病患有遠端轉移。接受單獨放射線治療病患的2及5年存活率為42.7%及23.7%，而合併治療的病患則為41.0%及16.0%，兩組於全部及無疾病存活率並無統計學上的差異（p＝0.694及0.875)。對於有淋巴轉移的患者而言，合併化學與放射線治療在全部存活率及無疾病存活率上並無優於單獨放射線治療（p＝0.273及0.486）。對合併治療的患者，接受三次或三次以上之化學治療其全部及無疾病存活率並無優於較少次數之病患（p＝0.968及0.406）。就單變數分析而言，神經侵犯、陽性淋巴的數目、範圍與比例是全部存活率與無疾病存活率的影響因子。以多因子分析而言，只有陽性淋巴的範圍影響存活率。討論：根據我們病患的數據，手術後合併化學與放射線治療並無增加存活率的效應。神經侵犯、陽性淋巴的數目、範圍與比例是影響術後病患預後的重要因子。

Purpose: To evaluate the effect of post-operative concurrent chemoradiotherapy (CCRT) and to determine the prognostic factors in esophageal cancer. Materials and Methods: From October 1991 to November 2000, fifty-six patients with esophageal cancer received post-operative radiotherapy at VGH-KS. Twenty-three patients were treated with radiotherapy (RT) alone at 50 Gy and others underwent RT combined with chemotherapy (C/T) consisting of cisplatin plus 5-flurouracil (5-FU). There was no significant difference in patient characteristics between two groups, except in tumor length. We compared overall, disease-free, loco-regional-free and distant-metastasis-free survival rates between two groups. We also analyzed various prognostic factors including age, tumor length, tumor stage, status of surgical margin, lymph node (LN) metastasis, vascular or neural invasion, and the time interval between operation and radiotherapy. Results: The median follow-up time was 16 months. For all patients, the overall survival (OS) at 5 years was 20.2%, and the disease-free survival (DFS) was 19.1 %. Loco regional failure occurred in 14.3% of patients and 35.7% of patients developed distant metastasis. The 2-year and 5-year overall survival rates were 42.7% and 23.7% for RT group and 41.0% and 16.0% for CCRT group. There was no significant difference in as and DFS between two groups (p= 0.694 and 0.875). In lymph node metastatic patients, overall and disease-free survivals of CCRT group were not superior to those of RT group (p= 0.273 and 0.486). In CCRT group, there was also no significant difference in as and DFS between the patients receiving less than 3 cycles of C/T and those receiving more (p= 0.968 and 0.406, respectively). By the uni-variate analysis, we found that neural invasion, numbers of LN metastases and involved LN levels, and the percentage of positive dissected LN affected the survival rates. The only prognostic factor of survivals in the multi-variate analysis was the involvement of more than one LN level. Conclusion: Based on our analysis, post-operative combination therapy is no better than adjuvant radiotherapy alone in disease control. The most important prognostic factors affecting the post-operative survival rate of patients are the number of positive lymph node and LN levels involved, the percentage of positive dissected LN and neural invasion.