Classification of Recurrent Esophageal Cancer after Radical Esophagectomy with Two- or Three-field Lymphadenectomy

Transcription

1 Classification of Recurrent Esophageal Cancer after Radical Esophagectomy with Two- or Three-field Lymphadenectomy HIROYUKI KATO, MINORU FUKUCHI, TATSUYA MIYAZAKI, MASANOBU NAKAJIMA, HITOSHI KIMURA, AHMAD FARIED, MAKOTO SOHDA, YASUYUKI FUKAI, NORIHIRO MASUDA, RYOKUHEI MANDA, HITOSHI OJIMA, KATSUHIKO TSUKADA and HIROYUKI KUWANO Department of General Surgical Science (Surgery I), Gunma University, Graduate School of Medicine, Maebashi, , Japan Abstract. Background: Although patients with recurrent diseases have a very poor prognosis, appropriate pretreatment classification for management of recurrent esophageal cancers has yet to be identified. The patterns of recurrence following radial esophagectomy were retrospectively assessed, and evaluated for possible adaptation as a novel classification of recurrent esophageal cancer. Patients and Methods: One hundred and sixty thoracic esophageal cancer patients (142 men; 18 women), who underwent radical esophagectomy without preoperative treatment, were studied. Results: Recurrence occurred in 59 (36.8%) patients. The relationship between recurrence and clinicopathological features revealed significant associations between recurrence and age at surgery (p<0.05), tumor (p<0.0001), lymph node (p<0.0001) and metastatic status (p<0.01), pathological stage (p<0.0001) and lymph node dissection (p<0.0001). Locoregional recurrence occurred in 13 (22%) patients, distant in 30 (51%) and mixed in 16 (27%). Mixed recurrence occurred the fastest and showed the poorest prognosis. A novel classification for recurrent esophageal cancer was proposed based on the clinical findings. Univariate analysis of the prognostic factors for post-recurrent survival revealed a significant association with distant organ recurrence (p<0.05). Conclusion: Mixed recurrence had a poorer outcome than other recurrences. The proposed classification of clinical findings for recurrent esophageal cancer was shown to be useful; however, further studies with a larger number of recurrent esophageal cancers are required for stage grouping of the proposed classification. Correspondence to: Hiroyuki Kato, Department of General Surgical Science (Surgery I), Gunma University, Graduate School of Medicine, , Showa-machi, Maebashi, , Japan. Tel: , Fax: , hiroyuki@ po.wind.ne.jp Key Words: Esophageal cancer classification, esophagectomy. Esophageal carcinoma has proven to be one of the most difficult malignancies to cure (1, 2), with cancer of the esophagus reportedly the ninth most common malignancy, ranking sixth in the most frequent causes of cancer death worldwide (1). Moreover, prognoses remain unsatisfactory, despite significant advances in surgical techniques and perioperative management (3, 4). Esophageal carcinoma is usually locally invasive, and lymphatic and hematogenous metastases are well documented. Accurate pre-treatment staging, particularly with regard to the depth of tumor invasion, lymph node involvement and distant metastases, is vital in determining the most appropriate procedure for the management of thoracic esophageal carcinomas (5). However, even after apparently curative surgery, a large number of patients experience recurrent diseases. Many investigators have examined recurrence rates after esophagectomy for esophageal carcinoma (6-15), with twothirds of patients showing recurrence within 1 year and nearly all patients showing recurrence within 2 years after the primary operation (4, 7). In approximately one-third of patients, recurrence is located in the operation field. Lesions are usually seen in the regional lymph nodes or as a mass originating in the mediastinum and infiltrating the gastric pull up from outside (7). However, the majority of recurrences are distant metastases, indicating the systemic character of this disease (14-18). Hematogenous deposits were also shown to accompany 59% of recurrences (17). Available therapeutic modalities for recurrent cancer are radical re-resection, palliative resection and bypass, laser thermocoagulation, stenting, chemotherapy, brachytherapy and radiotherapy, alone or in combination. The specific therapeutic modality of choice depends on the localization and extent of the recurrence. Raoul et al. (19) reported that early detection of recurrent disease is desirable because aggressive treatment might result in prolonged tumor-free survival. Although patients with recurrent disease have a very poor prognosis, an appropriate pre-treatment classification for the management of recurrent esophageal cancers has yet to be /2005 $

2 identified. Furthermore, no optimal treatment for recurrent esophageal cancers has been uniformly established. Therefore, to accurately classify recurrent esophageal cancer, we retrospectively assessed the patterns of recurrence following radial esophagectomy, and evaluated whether they could be adapted for the novel classification of recurrent esophageal cancer. Patients and Methods Patients. Between January 1988 and December 2003, 273 patients with thoracic esophageal squamous cell carcinoma (SCC) were identified at the Department of General Surgical Science, Gunma University, Graduate School of Medicine, Japan. One hundred and sixty patients (142 men and 18 women) with thoracic esophageal cancer, who underwent radical esophagectomy without preoperative treatment, were included in this study. Written informed consent was obtained from all patients prior to surgery and those without residual tumors (R0) at operation were enrolled. Patients with palliative resection and apparent residual tumors were excluded. The median age of the patients was 62.5 years (range: 36 to 79). The tumor stage and disease grade were classified according to the sixth edition of the TNM classification of the International Union Against Cancer (UICC). Resectability was determined by conventional staging, which included computed tomography (CT) of the neck, chest and abdomen, a bone scan, endoscopic ultrasound, esophagography, endoscopy and 18 F-fluorodeoxyglucose positron emission tomography (FDG-PET). Patients with limited local metastasis, considered by thoracic surgeons to have resectable diseases, were included. The mean follow-up period for all patients was 39.2±37.2 months (range: 6 to 119). Surgery and clinical outcomes. Two different procedures were used. In suitable cases, a standard esophagectomy was performed according to the McKeown method (right thoracotomy followed by laparotomy and neck incision with a cervical anastomosis) with three-field (thoracoabdominal and cervical; n=114) lymph node dissection. In other cases, an Ivor Lewis esophagectomy was used (right thoracotomy and laparotomy with anastomosis in the chest) with two-field (thoracoabdominal; n=46) lymph node dissection if indicated (20). All patients received curative thoracic esophagectomy including the esophagogastric junction. According to the TNM staging, distant lymph node metastasis was classified as M1 with no distant organ metastasis at the time of surgery. The histological grade included 147 squamous cell carcinoma, 7 adenocarcinoma, 2 adenosquamous carcinoma, 1 basaloid carcinoma, 2 small cell carcinoma and 1 mucoepidermoid carcinoma. After surgery, the lymph nodes were separated from the resected esophagus and adjacent tissue and assigned specific numbers indicating their localization according to the guidelines of the Japanese Society for Esophageal Diseases (JSED) (21). The surgical specimens were fixed, embedded, stained with hematoxylin and eosin (HE) and examined microscopically by two pathologists. In routine pathological examination of the dissected lymph nodes, the pathologists examined one maximum cross-section. The extent of lymph node metastasis and localization of the lymph nodes were documented. Definition and classification of recurrence. Recurrent disease was assessed by physical examination, histological findings, clinical follow-up and imaging; i.e., CT, magnetic resonance imaging (MRI), ultrasonography, bone scintigraphy, FDG-PET and specific X-ray studies. If no recurrent disease was diagnosed histologically, in clinical follow-up or radiological imaging, investigations were repeated within 6 months. The organ with recurrence was defined as the first site of recurrence and additional recurrence found within 1 month was considered to have occurred simultaneously. The time to recurrence was defined as the period from the date of surgery until detection of the first recurrence. The locoregional lymph nodes comprise intrathoracic and perigastric lymph nodes. Cervical, celiac axis and paraaortic node metastasis were classified as distant recurrence, and simultaneous locoregional and distant recurrences were classified as mixed recurrence. Table I shows the proposed novel classification of clinical findings for recurrent esophageal cancer. This classification was based on a modification of the TNM classification of the International Union Against Cancer (UICC). In this classification, recurrent disease findings should be recorded using the small letter r before the symbol for clinical findings. Recurrent diseases in recurrent tumors result from the primary tumor (rt), lymph node recurrence (rn), or distant organ recurrence (rm). Definitive treatment should be recorded with the following prefixes: Ra (radiotherapy), Ch (chemotherapy), Ra+Ch (combined radiotherapy and chemotherapy), EMR (endoscopic mucosal resection), EMR+Ra (endoscopic mucosal resection followed by radiotherapy), ESD (endoscopic submucosal dissection) and Es (esophagectomy). Statistical analysis. The relationships between each recurrence and the clinicopathological features were determined using the 2 method and Fisher s exact test. The relationship between recurrence and time to recurrence was determined using analysis of variance (ANOVA) or the Student s t-test. Survival rates after recurrence were calculated using the Kaplan-Meier method for analysis of censored data. The significance of differences in survival was analyzed with the log-rank test in univariate analysis. The relationships between recurrence patterns and other parameters were determined using univariate and multivariate logistic regression analyses. Results Occurrence of recurrence and clinicopathological features. Recurrence was observed in 59 (36.8%) out of 160 patients. The follow-up period for non-recurrent patients was at least 24 months. The overall 5-year survival was 52%, and the cause-specific survival was 57% at the time of analysis. The relationship between recurrence and clinicopathological features at surgery are shown in Table II. Significant associations between recurrence and age at surgery (p<0.05), tumor status (p<0.0001), lymph node status (p<0.0001), metastatic status (p<0.01), pathological stage (p<0.0001) and lymph node dissection (p<0.0001) were observed, but there were no significant associations between recurrence and gender, location and histological grade. Pattern of recurrence and time to recurrence. Locoregional recurrences occurred in 13 (22%) patients, with a mean time to recurrence of 18.8 months between post-operative months 4.5 and Distant recurrence occurred in

3 Kato et al: Classification of Recurrence in Esophageal Cancer Table I. Classification of recurrent esophageal cancer based on clinical findings. 1 Recurrent disease consisting of a recurrent tumor from the primary cancer, lymph node recurrence, or distant organ recurrence. 2 Recurrent disease findings should be recorded using the small letter r before the appropriate symbol for clinical findings. 3 Definitive treatment should be recorded with the following prefixes: Ra Radiotherapy Ch Chemotherapy Ra+Ch Combined radiotherapy and chemotherapy EMR Endoscopic mucosal resection EMR+Ra Endoscopic mucosal resection followed by radiotherapy ESD Endoscopic submucosal dissection Es Esophagectomy 4 Initial fingings of the recurrent disease should be recorded as follows: 1) Recurrent tumor invasion from primary cancer (rt): TX A recurrent tumor that cannot be assessed. T0 No evidence of a recurrent tumor. T1 A recurrent tumor within the submucosa. T2 A recurrent tumor beyond the muscularis propria. T3 A recurrent tumor in the esophageal adventitia. T4 A recurrent tumor in adjacent organs. Tex A recurrent tumor in the extramural space (post-esophagectomy). Note: If the esophagus is not resected, the recurrent disease should be indicated by T1-4. 2) Lymph node recurrence (rn): NX The lymph node recurrence cannot be assessed. N0 No lymph node recurrence. N1 Lymph node recurrence in locoregional lymph nodes. Note: Locoregional lymph nodes consist of intrathoracic and perigastric lymph node groups. If necessary, the number of lymph node recurrences should be recorded in parentheses; Ex. N1(3) 3) Distant organ recurrence (rm): MX Distant organ recurrence cannot be assessed. M0 No distant organ recurrence. M1 Distant recurrence in one organ. M2 Distant recurrence in two organs. M3 Distant recurrence in three or more organs. Note: The organ or structure with recurrence should be recorded in parenthesis; Ex. M1 (lung), M2 (liver, bone). (51%) patients, with a mean time to recurrence of 13.9 months between post-operative months 2.6 and Mixed recurrence occurred in 16 (27%) patients, with a mean time to recurrence of 8.4 months between post-operative months 2.4 and The time for mixed recurrence to occur was therefore shorter than for locoregional and distant recurrences. Organ of recurrence and time to recurrence. Fifty-nine patients had recurrent diseases in a total of 89 organs (Table III). Locoregional recurrent diseases occurred in 5 local and 26 regional lymph nodes. Distant organ recurrences occurred in a total of 58 organs; 16 patients had recurrence in the liver, 13 in the lung, 9 in the bone, 2 in the brain, 1 in the skin, 1 in the pleura and 16 in the distant lymph nodes. The mean time to recurrence in the liver was 6.9 months, which was shorter than in the distant lymph nodes (p<0.05). Sixteen patients had recurrence in distant lymph nodes within 15.2 months between post-operative months 2.6 and Treatment of recurrence and prognosis. Of the 59 patients with recurrence, 19 were not treated because their general condition, age and/or social background were not appropriate. Of the remaining 40 patients, 12 received chemotherapy, 21 received radiotherapy or chemoradiotherapy and 7 received reduction surgery. The surgical procedures were 6 lymphadenectomy (4 cervical, 1 mediastinal and 1 abdominal) and 1 partial lung resection. Post-recurrence survival curves of patients with locoregional recurrence were higher than those of patients with distant or mixed recurrence (log-rank test, p=0.0575, Figure 1). The 1- and 2-year survival rates of patients with locoregional recurrence were 50.0 and 26.7%, respectively, with a mean survival time (MST) of 15.7 months (Figure 1); those of patients with distant recurrence were 24.1 and 17.2%, respectively, with a MST of 9.8 months; and those of patients with mixed recurrence were 13.9 and 0%, respectively, with a MST of 5.5 months. Post-recurrence survival curves of patients who underwent surgery were higher than those of patients who received non- 3463

4 Table II. Recurrence and clinicopathological characteristics in 160 patients with esophageal carcinomas. Parameters Recurrence Positive Negative p-value n=59 n=101 Gender Male Female 6 12 Age at surgery (years) > Location upper mid lower Tumor status T T T T4 5 3 Lymph node status N N Metastatic status M M Pathological stage Stage I Stage II Stage III Stage IV 16 9 Histological grade G G G Others 5 8 Lymph node dissection 2 Field Field G1: well-differentiated carcinoma G2: moderately-differentiated carcinoma G3: poorly-differentiated carcinoma surgical treatment (log-rank p=0.0018, Figure 2). The 1- and 2-year survival rates of patients who underwent surgery were 87.5 and 70.0%, respectively, with a MST of 24.6 months (Figure 2); those of patients who received radiotherapy or chemoradiotherapy were 33.3 and 14.3%, respectively, with a MST of 12.3 months; and those of patients who received chemotherapy were 12.0 and 0%, respectively, with a MST of 5.2 months. The survival rates of patients who received no treatment were 0 and 0%, respectively, with a MST of 2.9 months. Novel classification of recurrence according to clinical findings. Our proposed classification of recurrent esophageal cancer Table III. Organ of recurrence and time to recurrence. Organ of recurrence n Time to recurrence (mo) Locoregional Local recurrence ( ) Regional lymph node recurrence ( ) Distant organ Liver ( ) Lung ( ) Bone ( ) Brain ( ) Skin Pleura Distant lymph node ( ) is summarized in Table IV. Eight patterns of recurrence were observed. Twenty-one patients showed recurrence according to the classification of Es-rT0N0M1, while 9 were Es-rT0N0M2. These recurrences are equal to conventional distant recurrence, while conventional locoregional recurrence is equal to both Es-rT0N1M0 and Es-rTexN0M0; the remaining types are equal to mixed recurrence. The results of univariate analysis of the prognostic factors for post-recurrent survival are shown in Table V. A significant association between post-recurrent survival and distant organ recurrence was observed (p<0.05). Discussion Numerous reports have documented recurrence rates after esophagectomy for esophageal carcinoma (6-15) (Table VI). In the present study, recurrence had occurred in 59 (36.8%) out of 160 patients at the time of analysis. The relationship between recurrence and clinicopathological features at surgery revealed significant associations between recurrence and tumor status, lymph node status, metastatic status, pathological stage and lymph node dissection. However, it is expected that patients with more advanced cancer will have a higher possibility of recurrence. The association with lymph node dissection was possibly observed because esophagectomy with 2-field dissection was performed in patients with less advanced cancer. With regard to the pattern of recurrence, half the patients showed distant recurrence, while half of the remainder showed locoregional recurrence and the other half mixed recurrence. The mean time to mixed recurrence was shorter than that to the other recurrences, and post-recurrence survival curves of patients with mixed recurrence were lower than those of patients with locoregional and distant recurrences. From these findings, we concluded that mixed recurrence occurs the fastest, presenting the poorest prognosis after recurrence. Further, these results were probably caused by the 3464

5 Kato et al: Classification of Recurrence in Esophageal Cancer Figure 1. Post-recurrence survival curves according to patterns of recurrence. Post-recurrence survival curves of patients with locoregional recurrence were higher than those of patients with distant or mixed recurrence (log-rank test, p=0.0575). Their 1- and 2-year survival rates were 50.0 and 26.7%, respectively, with a mean survival time (MST) of 15.7 months. Figure 2. Post-recurrence survival curves according to treatment of recurrence. Post-recurrence survival curves of patients who received surgery were higher than those of patients who received non-surgical treatment (log-rank p=0.0018). Their 1- and 2-year survival rates were 87.5 and 70.0%, respectively, with a mean survival time (MST) of 24.6 months. Table IV. Classification of recurrence and survival. Classification No. of Time to Survival after patients recurrence recurrence rt rn rm (n=59) (mo) (mo) T0 N0 M M T0 N1 M M M Tex N0 M M Tex N1 M association between post-recurrence survival and distant organ recurrence. The recurrence rates after esophagectomy reported in the literature range from 27 to 53% (Table VI). Locoregional recurrence rates range from 22 to 55%, distant recurrence from 22 to 75% and mixed recurrence from 2 to 50%. The mean time to recurrence was from 10 to 11.7 months. In recent studies, the mean time to distant or hematogenous recurrence was shorter than that to locoregional or non-hematogenous recurrence (17, 18). In addition, Kyriazanos et al. (17) reported Table V. Univariate analysis of prognotic factors for post-recurrent survival. Variate Risk ratio 95% CI p-value rt (T0 & Tex) rn (N0 & N1) rm (M0 & M1, M2) CI: confidence interval that hematogenous deposits accompany 58.5% of recurrences. Univariate analysis revealed adjuvant chemoradiation, tumor location in the lower esophagus and tumor dedifferentiation not to be promoting factors for hematogenous recurrence. Moreover, using multivariate analysis, Nakagawa et al. (18) reported that hematogenous recurrence was associated with depth of invasion and number of lymph node metastases; locoregional recurrence was also associated with depth of invasion and distant lymph node metastasis. It is possible that locoregional recurrence is mainly influenced by the extent of the local tumor and lymph node metastasis; hematogenous recurrence is not only influenced by the stage, but also the oncological behavior of the tumor. Matsubara et al. previously reported that recurrence could be divided into three patterns, and that patients with mixed recurrence had a poorer outcome 3465

6 Table VI. Previous reports of recurrence after esophagectomy. Pattern of recurrence Author Surgical procedure No. of patients Recurrence Locoregional Distant Mixed Mean time to Year rate recurrence recurrence recurrence recurrence (%) (%) (%) (%) (M) Clark (6) Two-field lymphadectomy 38 (ADE) (Lymph node) (Local) Law (7) Transthoracic resection 108 (SCC) Matsubara (8) Three-field lymphadenectomy Kato (9) Three-field lymphadenectomy (Lymph node) (Local) Bhansali (10) Three-field lymphadenectomy Dresner (11) Two-field lymphadenectomy 176 (ADE 113, SCC 63) Hulscher (12) Transhiatal resection 137 (95 ADE, SCC) Kyriazanos (17) Three-field lymphadenectomy 151 (SCC) (Non hematogenous) 8.4 (Hematogenous) Nakagawa (18) Three-field lymphadectomy 171 (SCC) (Locoregional) (Distant) Kato Two or three-field (Locoregional) lymphadectomy 13.9 (Distant) 8.4 (Mixed) SCC: squamous cell carcinoma ADE: adenocarcinoma than all other patients (8). Our results were similar to theirs. Regarding time to recurrence, two-thirds of recurrences occurred within 1 year, and about 90% within 2 years after the primary operation, which was similar to previous reports (4). Post-operative adjuvant therapies, such as radiotherapy and chemotherapy, had no influence on the incidence of any recurrence (10, 18). A number of investigators have reported on the pattern of recurrence after esophagectomy, but differences in follow-up evaluation hinder the comparison of results. Recurrent lesions spread with time and, therefore, the observed pattern or organ of recurrence is greatly affected by the length of the follow-up studies (8). In addition, the diagnostic accuracy of follow-up studies in detecting recurrences also has a significant influence on the results (22). In this study, recurrence was classified as locoregional, distant and mixed recurrence, as in the reports of Matsubara et al. (8). However, Nakagawa et al. classified both simultaneous locoregional and hematogenous recurrence as hematogenous recurrence, because they occur soon after surgery (18). Here, cervical, celiac axis and paraaortic node recurrence were classified as locoregional because of the manner of tumor spread through the lymphatic system. On the other hand, Kyriazanos et al. suggested the classification of recurrence as hematogenous or non- hematogenous recurrence (17), and they considered lymphatic recurrence (defined as M1 by TNM classification) as non-hematogenous. They possibly considered whether recurrence occurs in the operative or non-operative field, and through lymphatic or hematogenous metastasis. For the above reasons, it is difficult to compare the findings of different reports. We, therefore, established an appropriate pre-treatment classification for the management of recurrent esophageal cancers, proposing a simplified and understandable version which is similar to the common TNM classification with regard to the primary tumor and regional lymph node recurrence. However, with our version, the degree of distant recurrence is indicated as the number of the distant recurrent organs. To include salvage surgery after EMR or definitive chemoradiotherapy within the classification, we proposed the inclusion of definitive treatment as a prefix. Conventional recurrence patterns were also adapted for use in the proposed classification system. However, it was difficult to determine stage grouping because survival is affected by the site and size of distant recurrence. Further studies with a larger number of recurrent esophageal cancers are, therefore, required for the stage grouping of our classification. In this study, univariate analysis revealed an association between post-recurrence survival and distant organ recurrence. Furthermore, the post-recurrence survival curves of patients who received surgery were higher than those of patients who received non-surgical treatment. Some 3466

7 Kato et al: Classification of Recurrence in Esophageal Cancer reports have suggested that FDG-PET is a useful tool for detecting the recurrence of esophageal cancer (22-24), with early detection of recurrent lesions offering the possibility of additional therapy (8). In particular, when patients with recurrent lesions macroscopically confined to a limited area undergo re-resection, they show significantly longer survival rates after recurrence (8). However, distant recurrence cannot be controlled by surgery alone and, therefore, further studies using a larger sample size are needed to establish the optimal treatment of recurrence. In conclusion, mixed recurrence was shown to have a poorer outcome than other recurrences, due to the association between post-recurrence survival and distant organ recurrence. Moreover, the novel classification for recurrent esophageal cancer presented in this study was shown to be useful; however, further studies with a larger number of recurrent esophageal cancers are required in the future for stage grouping of the proposed classification. Acknowledgements We would like to thank Akie Nakabayashi, Hideko Emura, Tomoko Ogasawara, Sachiko Ueno and Yukie Saitoh for their excellent secretarial assistance, and Midori Ohno for assistance with data management and biostatistical analysis during the preparation of this report. References 1 Pisani P, Parkin DM and Ferlay J: Estimates of the worldwide mortality from eighteen major cancers in 1985: implications for prevention and projections of future burden. Int J Cancer 55: , Levine MS: Esophageal cancer: radiologic diagnosis. Radiol Clin North Am 35: , Baba M, Aikou T, Yoshinaka H, Natsugoe S, Fukumoto T, Shimazu H et al: Long-term results of subtotal esophagectomy with three-field lymphadenectomy for carcinoma of the thoracic esophagus. Ann Surg 219: , Nigro JJ, DeMeester SR, Hagen JA, DeMeester TR, Peters JH, Kiyabu M et al: Node status in transmural esophageal adenocarcinoma and outcome after en bloc esophagectomy. J Thorac Cardiovasc Surg 117: , Kato H, Kuwano H, Nakajima M, Miyazaki T, Yoshikawa M, Ojima H et al: Comparison between positron emission tomography and computed tomography in the use of the assessment of esophageal carcinoma. Cancer 94: , Clark GW, Peters JH, Ireland AP et al: Nodal metastasis and sites of recurrence after en bloc esophagectomy for adenocarcinoma. Ann Thorac Surg 58(3): , Law SY, Fok M and Wong J: Pattern of recurrence after oesophageal resection for cancer: clinical implications. Br J Surg 83: , Matsubara T, Ueda M, Takahashi T, Nakajima T and Nishi M: Localization of recurrent disease after extended lymph node dissection for carcinoma of the thoracic esophagus. J Am Coll Surg. 182: , Kato H, Tachimori Y, Watanabe H, Igaki H, Nakanishi Y and Ochiai A: Recurrent esophageal carcinoma after esophagectomy with three-field lymph node dissection. J Surg Oncol 61: , Bhansali MS, Fujita H, Kakegawa T, Yamana H, Ono T, Hikita S et al: Pattern of recurrence after extended radical esophagectomy with three-field lymph node dissection for squamous cell carcinoma in the thoracic esophagus. World J Surg 21: , Dresner SM and Griffin SM: Pattern of recurrence following radical oesophagectomy with two-field lymphadenectomy. Br J Surg 87: , Hulscher JB, van Sandick JW, Tijssen JG et al: The recurrence pattern of esophageal carcinoma after transhiatal resection. J Am Coll Surg 191: , van Lanschot JJ, Tilanus HW, Voormolen MH and van Deelen RA: Recurrence pattern of esophageal carcinoma after limited resection does not support wide local excision with extensive lymph node dissection. Br J Surg 81: , Fahn HJ, Wang LS, Huang BS, Huang MH and Chien KY: Tumour recurrence in long-term survivors after treatment of carcinoma of the esophagus. Ann Thorac Surg 57: , Fujita H, Kakegawa T, Yamana H, Shima I, Tanaka H, Ikeda S et al: Lymph node metastasis and recurrence in patients with a carcinoma of the thoracic esophagus who underwent three-field dissection. World J Surg 18: , Mantravardi RV, Lad T, Briele H and Liebner-EJ: Carcinoma of the esophagus: sites of failure. Int J Radiat Oncol Biol Phys 8: , Kyriazanos ID, Tachibana M, Shibakita M et al: Pattern of recurrence after extended esophagectomy for squamous cell carcinoma of the esophagus. Hepatogastroenterology 50: , Nakagawa S, Kanda T, Kosugi S et al: Recurrence pattern of squamous cell carcinoma of the thoracic esophagus after extended radical esophagectomy with three-field lymphadenectomy. J Am Coll Surg 198(2): , Raoul JL, Le Prise E, Meunier B, Julienne V, Etienne PL, Gosselin M et al: Combined radiochemotherapy for postoperative recurrence of oesophageal cancer. Gut 37: , Kuwano H, Miyazaki T, Masuda N, Tsukada K, Kato H and Maekawa S: Appraisal of the simultaneous right-thoracic and abdominal approach with intrathoracic reconstruction after esophageal resection for patients with intrathoracic esophageal cancer. Hepatogastroenterology 51(60): , Japanese Society for Esophageal Diseases: Guidelines for the Clinical and Pathological Studies on Carcinoma of the Esophagus. The 9th edition. Tokyo: Kanehara, Kato H, Miyazaki T, Nakajima M et al: Value of positron emission tomography in the diagnosis of recurrent oesophageal carcinoma. Br J Surg 91(8): , Flamen P, Lerut A, Van Cutsem E et al: The utility of positron emission tomography for the diagnosis and staging of recurrent esophageal cancer. J Thorac Cardiovasc Surg 120(6): , Skehan SJ, Brown AL, Thompson M et al: Imaging features of primary and recurrent esophageal cancer at FDG PET. Radiographics 20(3): , Received June 2, 2005 Accepted June 21,