The significance of the characteristics of intra-thoracic lymph node metastasis for radiotherapy range in small cell lung cancer

Objective To discuss the relationship between the incidence of lymph node metastasis and tumor size in patients with non small cell lung cancer, and to provide evidence for clinical operation and follow-up treatment. Methods 180 patients with non small cell lung cancer who were hospitalized and recieived Lobectomy operation were selected, the clinical data of patients and resection of tumor size, lymph node metastasis were retrospectively analyzed, and then the relationship between the lymph node metastasis and tumor volume was investigated. Results 180 cases of patients with lymph node metastasis rate was 37.20%, in which hilar lymph node metastasis rate was 11.02 percent and mediastinal lymph node metastasis rate was 10.95%.Lung lymph node metastasis rates in lymph node metastasis, hilar and mediastinal lymph node metastasis of T1a were 13.21%, 15.09% and 13.21%; those of T1b period were 38.09%, 35.71% and 28.57%; of T2a period were 44.44%, 37.78% and 31.11%; of T2b period were 64.26%, 57.14% and 53.57%; and those of T3 period were 58.33%, 66.67% and41.67%.Lymph node metastasis, hilar and mediastinal lymph node metastasis in lung lymph nodes transfer rate with patients of T1a had significantly differences compared with patients of T1b, T2a, T2b and T3 stage.(χ2=5.363, 6.175, 4.372, 8.778, 5.363, 9.631, 7.936, 9.631, 20.463, 7.521, 6.175, 4.387, all P<0.05). Hilar lymph node metastasis and lung metastases of T1b and T2b had a significant difference (χ2=12.353, 4.585, all P<0.05). Conclusion For non-small cell lung cancer, tumor size is positively correlated with lymph node metastasis and the greater the tumor size, the greater the likelihood of lymph node metastasis. Key words: Tumor volume; Tumor metastasis; Lymph nodes; Carcinoma, non small cell lung

To explore the reasonable radiotherapy range by analyzing the characteristics of supraclavicular lymph node metastasis in limited-stage small cell lung cancer (LS-SCLC). From January 2005 to December 2011, patients of LS-SCLC were reviewed. Supraclavicular zone was further divided into five subgroups including para-recurrent laryngeal nerve (region I and region II ), para-internal jugular vein (region III ), supraclavicular region (region IV), as well as the other regions except for the mentioned above (region V). The characteristics of the lymph nodes in each region were analyzed. The supraclavicular lymph node metastasis was found in 60 patients, with a positive rate of 34.5%. In multivariate Logistic regression analysis,intra-thoracic lymph node metastasis in the lymph node stations of level 2 and 3 were found to be the risk factors of supraclavicular lymph node metastasis (P = 0.006,P = 0.000). Our data suggests that the frequencies of metastasis in region I and III were much higher than those in the other areas.Among the sixty patients with supraclavicular lymph node metastasis, 95.0% were found at region I or III while the incidence of skip metastasis was only 5.0%. It is advisable to contain the bilateral supraclavicular nodes in patients with mediastinal lymph nodes metastasis to the level 2 or 3 for elective radiation target volume.The clinical target volume (CTV) exterior margin containing the outer margin of internal jugular vein may be suitable.

Resection of a Posterior Mediastinal Metastasis of Colon Cancer

Objective To investigate the status of intrapulmonary lymph node metastasis and its impact on pathological staging in patients with non-smallcell lung cancer (NSCLC). Methods A total of 177 patients undergoing lobectomy or pneumonectomy and systemic lymph node dissection for lung cancer inDepartment of Thoracic Surgery, Tianjin Chest Hospitalbetween January 2015 and March2016 were selected. A special pathological examination protocol on lobar lymph nodes (No. 12) and segmental lymph nodes (No. 13) surgical specimens was conducted after routine pathological examination. The N stage retrieved by special pathological examination protocol was compared with thatobtainedby routine pathological examination. In addition, the risk factors of the intrathoracic lymph node metastasis were explored. Results A total of 1268 N1 lymph nodes were examined, among which 736 were detected during routine pathological examination protocol and 532 were retrieved byspecial pathological examination protocol. The combination of routine pathological examination and special pathological examination provided a median of 7 N1 lymph nodes(range, 2 to 24), a significant increase from the number of N1 lymph nodes examined by routine pathological examination alone(4; range, 0 to 18)(P 0.05). The pathological stage of 15 patients(8.4%) changed from N0 to N1 after specialpathological examination. Conclusions Routine pathological practice frequently leaves a large number of N1 lymph nodes unexamined, a clinically significant proportion of which harbor metastasis. The special pathological examination protocol is suggested to improve the accuracy of pathological staging of NSCLC. Key words: Non-small-cell lung cancer; Lymph node; Metastasis; Pathological examination

Objective To investigate the characteristics and risk factors of abdominal lymph node metastasis in thoracic esophageal squamous cell cancer. Methods The clinical data of 586 patients with thoracic esophageal cancer who underwent surgery via transabdominal and transthoracic approaches between June 2009 and June 2014 at the Sichuan Cancer Hospital were retrospectively analyzed. All the patients received resection of esophageal cancer and lymph node dissection, and the transabdominal right thoracic approach or cervico-thoracic-abdominal triple incision was selected according to the condition of patients. No.18, 19, 20 lymph nodes were dissected seperately and No.16, 17 and lesser curvature lymph nodes were separated. All the specimens of lymph nodes were detected by regular pathological examination. Measurement data with normal distribution were presented as ±s and count data were described as rate. Comparisons of rate between 2 specimens and among the multiple specimens were respectively analyzed using the chi-square test and partition of chi-squared. The multivariate analysis was done using the logistic regression. Results The number of lymph node dissected in 586 patients was 12 524 with an average number of 20±11 per case, and the rate of lymph node metastasis was 55.63%(326/586). The number of mediastinal lymph node dissected was 7 012 with an average number of 12±5 per case, and a rate of mediastinal lymph node metastasis was 40.96%(240/586). The number of abdominal lymph node dissected was 5 512 with an average number of 9±8 per case, and a metastasis rate was 31.74% (186/586). The abdominal lymph node metastasis rate of the upper, middle and lower thoracic esophageal cancer were 13.73%(14/102), 31.51%(92/292) and 41.67%(80/192), respectively, showing a significant difference among the above 3 indexes (χ2=25.91, P<0.05). The lymph node metastasis rate in No.16, 17, 18, 19, 20 and lesser curvature lymph nodes were 12.80%(75/586), 16.89%(99/586), 1.71%(10/586), 0.68%(4/586), 1.71%(10/586) and 2.05%(12/586), respectively, with a significant difference among the above 6 indexes (χ2=287.95, P<0.05). The results of univariate analysis showed that the tumor location, surgical procedure, T stage, N stage, G stage, pathological stage and mediastinal lymph node metastasis were risk factors affecting abdominal lymph node metastasis of thoracic esophageal cancer (χ2=24.02, 23.97, 37.87, 136.85, 38.79, 7.70, 154.27, P<0.05). The tumor in the lower thoracic portion, N3 stage and stage Ⅳ were independent risk factors affecting abdominal lymph node metastasis of thoracic esophageal cancer in the multivariate analysis (RR=5.80, 2.36, 2.76, 95% confidence interval: 1.022-1.813, 1.317-3.950, 1.652-12.351, P<0.05). Conclusions Abdominal lymph node metastasis is common in thoracic esophageal cancer in which No.16 and 17 lymph nodes predominate, and it is easy to occur in patients with lower thoracic esophageal cancer, and advanced N stage and pathological type. Key words: Esophageal neoplasms; Abdominal lymph node; Metastasis

Objective To explore the reasonable clinical target volumes by analyzing the characteristic of mediastinal lymph node metastases in non-small-cell lung cacer (NSCLC).Methods 291 NSCLC patients was performed pre-therapy CT scans,and the incidence of mediastinal lymph node metastases was analysed.Results Among the 152 patients with right lung NSCLC,the incidence of involvement of the ipsilateral supraclavicular lymph nodes was 15.8 ％ (24/152),and the incidence of involvement of the contralateral lymph nodes was 6.6 ％ (10/152).The highest incidence of mediastinal nodal involvement occurred in the ipsilateral hilar nodes (59.2 ％),followed by area 4R (56.6 ％),area 1-2R (36.2 ％),area 7 (33.6 ％),area 4L(20.4 ％),area 10-11L (5.9 ％),area 6 (3.9 ％),area 5 (2.0 ％),area 1-2L (2.0 ％),respectively.Among the 139 patients left lung NSCLC,the incidence of involvement of the ipsilateral supraclavicular lymph nodes was 15.8 ％ (23/139),and the incidence of involvement of the contralateral lymph nodes was 5.8 ％ (8/139).The highest incidence of mediastinal nodal involvement occurred also in the ipsilateral hilar nodes (54.0 ％),followed by area 7 (33.8 ％),area 4R (26.6 ％),area 4L (24.5 ％),area 1-2R (15.8 ％),area 5 (10.8 ％),area 6 (9.4 ％),area 1-2L (5.8 ％),area 10-11R (5.0 ％) respectively.Conclusion The right side primaries or left side primaries of NSCLC have different high risk lymph node areas for metastasis,and selective irradiation to these lymph node areas maybe increase the tumor control rate and reduce the recurrence rate. Key words: Carcinoma, non-small-cell lung; Radiotherapy; Neoplasm metastasis, mediastinal lymph node; Target volumes

Objective To investigate the patterns of intrathoracic lymph node metastasis in nonsmall cell lung cancer (NSCLC) and to provide a theoretical basis for the delineation of radiotherapy target volume.Methods A retrospective analysis was performed on the clinical data of 314 NSCLC patients after operation.Our focus was to investigate the distribution characteristics and metastatic extent of intrathoracic lymph nodes and their relationship with tumor size,pathological type,and primary site.Comparisons between groups were made by one-way analysis of variance.Results The frequencies of metastases to lymph nodes at stations 4,5,7,10,and 11 were all above 12％,while those at stations 1,2,3,6,8,and 9 were all below 12％.The lymph node metastasis rate was similar on the primary tumor site (P =0.102).The patients with T3 and T4 NSCLC had a significantly higher frequency of N2 lymph node metastasis than those with T1 and T2 NSCLC (17.0％ vs.11.6％,P =0.002) ;the patients with adenocarcinoma had a significantly higher frequency of N2 lymph node metastasis than those with squamous cell carcinoma (34.5 ％ vs.23.2％,P =0.008).Conclusions For patients with T3 and T4 NSCLC of adenocarcinoma subtype,we should highlight the lymph nodes at stations 4,5,7,10,and 11 when delineating the radiotherapy target volume after operation. Key words: Non-small cell lung cancer; Lung adenocarcinoma; Lung squamous carcinoma; Lymph node metastasis

To investigate the regularity of lymph node metastasis in cardiac carcinoma and its risk factors. Complete clinicopathological data of 768 cardiac carcinoma patients undergoing radical resection and lymph node dissection were collected. A retrospective cohort study was performed to analyze the distribution of lymph node metastasis (lymph node metastasis rate=number of patients with lymph node metastasis/number of patients with lymph node dissection; lymph node metastasis frequency=number of metastatic lymph node/number of total resected lymph node) and the influence of clinicopathological factors on lymph node metastasis. Of the 768 patients, 599 were male and 169 were female, with mean age of 61(28 to 85) years. According to gastric cancer staging criteria from the American Joint Cancer Association (AJCC) 7th edition in 2010, there was 256 cases in N0 stage, 171 cases in N1 stage, 181 cases in N2 stage, 160 cases in N3 phase; 18 cases in T1 stage, 30 cases in T2 stage, 9 cases in T3 stage, 711 cases in T4 stage. Borrmann type I( was found in 61 cases, type II( in 306 cases, type III( in 358 cases, type IIII( in 43 cases. The histological type was adenocarcinoma in 738 cases and signet ring cell carcinoma in 30 cases. A total of 9 183 lymph nodes were resected during operation for 768 patients with mean 12(0 to 57) nodes per case, while 510 patients were found to have 2 889 metastatic nodes; the lymph node metastasis rate was 66.4%(510/768), and lymph node metastasis frequency was 31.5%(2 889/9 183). Besides, 483 patients were found to have 2 759 metastatic lymph nodes and 8 246 resected lymph nodes in abdominal cavity with lymph node metastasis rate of 62.9%(483/768) and lymph node metastasis frequency of 33.5% (2 759/8 246); 57 patients were found to have 130 metastatic lymph nodes and 937 resected lymph nodes in thoracic cavity with lymph node metastasis rate of 7.4%(57/768) and lymph node metastasis frequency of 13.9%(130/937). Stations with the higher lymph node metastasis rate included paracardiac (left cardia: 38.8%, right cardia: 39.9%), lesser curvature of stomach(41.9%), left gastric artery (46%) and posterior pancreatic (38.5%). A total of 361 patients had resected lymph node number ≥12 during operation, while other 407 patients had number <12. Univariate analysis showed that Borrmann type, depth of tumor invasion and resected lymph node number were associated with lymph node metastasis. Lymph node metastasis rates of Borrmann type I(, II(, III( and IIII( patients were 55.7% (34/61), 62.7% (192/306), 73.7% (264/358) and 51.2%(22/43) respectively, and the difference was statistically significant (χ2=18.115, P=0.000). Lymph node metastasis rates of T1, T2, T3, T4 stage patients were 0%(0/18), 30%(9/30), 100%(9/9) and 69.5%(494/711) respectively, and the difference was statistically significant (χ2=63.971, P=0.000). Lymph node metastasis rate of patients with resected lymph node number ≥12 was 79.5%(287/361), which was significantly higher than 55.3%(225/407) of those with resected lymph node number <12(χ2=50.496, P=0.000). Multivariate analysis revealed that higher T stage (OR=2.326, 95%CI: 1.758 to 3.078, P=0.000) and resected lymph node number ≥12(OR=2.998, 95%CI: 2.142 to 4.195, P=0.000) were independent risk factors of lymph node metastasis. The lymph node metastasis rate of cardiac carcinoma is quite high. The metastasis occurs mainly in the surrounding of cardia, the small curvature of the stomach, the left artery of stomach and posterior pancreatic. The depth of tumor invasion and the number of lymph node dissection are independent risk factors of lymph node metastasis.

Objective To investigate the diagnostic value of EBUS imaging features for metastatic mediastinal/hilar lymph node enlargement in lung cancer.Methods The lung cancer patients with a pathological diagnosis and without preoperative anti-tumor treatment who got the EBUS-TBNA examination from October 2009 to September 2011 were retrospectively analysis.422 lung cancer patients with 683 mediastinal / hilar lymph nodes were enrolled in this study,including 335 males and 87 females; the median age is 61 years old (range 24-82),EBUS lymph node ultrasound image and the final pathological or follow-up results were compared by the statistical analysis.Homogeneity in the lymph node EBUS image feature was defined as:uniform echo in the ultrasound images,cortex existed in the peripheral areas,medulla existed in lymph central with a slightly stronger echo and represent as a small strip.Heterogeneity was defined as: the ultrasound image was defined as uneven echo involved with coagulation necrosis sign,which was the hypoechoic areas without blood flow in the lymph nodes and represent no blood flow in the CDPI mode.The coagulation necrosis was associated with necrosis within the lymph node.In addition,if the CNS region occupied more than 11％ of the entire lymph node in a complete lymph node or just a part of huge lymph nodes in the EBUS imaging window frame,we also regard it as heterogeneity.If a complete lymph node was seen in the EBUS imaging window frame,we measured the longest diameter to the long axis and its vertical maximum diameter to the short axis.If the lymph node was huge and extended the EBUS imaging window frame,we measured the longest diameter in the frame as the long axis of its vertical maximum diameter to be the short axis.As to the EBUS-TBNA negative lymph nodes,we regarded it was malignant lymph node if the diameter of the lymph node increased by 20％ in the patients who did not received any chemotherapy or radiotherapy or the diameter of the lymph node increased or decreased by 20％ in the patients who received any chemotherapy or radiotherapy six month later in the chest enhanced CT scan,otherwise,it was identified as benign lymph node.We used the RECIST 1.1 solid tumors criteria to evaluate the efficacy of the chemotherapy.Results 422 patients were enrolled this study including 93 squamous carcinomas,137 adenocarcinomas,97 small cell lung cancer,42 poorly differentiated non-small cell lung cancer,29 adenosquamous carcinoma and 24 other malignant tumors (including large cell carcinoma,sarcomatoid carcinoma,carcinoid tumors,etc).The sensitivity of the EBUS-TBNA was 93.8％ (396/422).The diagnostic methods and results in the 683 lymph nodes were as the following: 506/683 (74.1％) was confirmed as cancer by the EBUS-TBNA while 177/683 (25.9％) was diagnosed as benign disease.Among these,32/683 (4.7％) was confirmed as cancer and 57/683 (8.3％)was confirmed as benign disease by surgery,9/683 (1.3％) was confirmed as cancer and 79/683 (11.6％) was confirmed as benign disease by the method of follow-up.the sensitivity for the EBUS-TBNA to be malignant was 506/547 (92.5 ％),specificity was 136/136 (100％),positive predictive value was 506/506 (100％),negative predictive value was 136/177(76.8％) and accuracy was 642/683 (94.0％).The short axis diameter in the 683 lymph nodes ranged from 0.40cm to 4.60cm with an average diameter of (1.58 ± 0.68) cm.Among them,the short axis diameter in the malignant lymph node was (1.75 ± 0.63) cm,and in the benign lymph nodes was (0.92 ± 0.40) cm.527 lymph nodes presented heterogeneity under the ultrasound imaging,in which,519/527 (98.5 ％)were malignant lymph nodes.While,156 lymph nodes presented homogeneity and 28/156 (17.9％) were malignant lymph nodes (x2 =489.5,P ＜0.01).In the heterogeneous lymph node with a short axis diameter more than 1.0cm,the sensitivity to be malignant was 89.4％,specificity was 100％ and accuracy was 89.6％.In the homogeneous lymph node with a short axis diameter less than 0.8cm,the sensitivity to be benign was 43.8％,specificity was 67.8％ and accuracy was 48.1％.Conclusion EBUS-TBNA is new biopsy method for the mediastinal / hilar lymph node.The classification based on EBUS imaging-based lymph node ultrasound image features was helpful to identify the procedure for the diagnostic purposes and could help to distinguish the benign or malignant mediastinal / hilar lymph node in lung cancer patients. Key words: Lung neoplasms; Diagnosis; Biopsy,needle; Ultrasonography,interventional

Objective The aim of this study was to evaluate the prognosis of resection followed by chemotherapy compared with chemoradiotherapy for limited-stage small cell lung cancer. Methods The clinical data of 230 limited-stage small cell lung cancer patients with curative treatment between January 2006 and December 2011 were retrospectively analyzed. All patients divided to two group: the resection plus chemotherapy (S+ C) and chemoradiotherapy (R+ C). And the prognostic factors were further analyzed with limited stage small cell lung cancer. The Kaplan-Meier method was used for the survival analysis. Results The overall survival rates of 1-year, 3-year and 5-year were 87.0%, 38.9%, 25.4%, respectively and the media survival time (MST) 26.0 months. When patients were stratified by clinical stageⅠ+ Ⅱ, the 1-year, 3-year and 5-year overall survival rates of S+ C group and R+ C group were 92.6%, 63.2%, 47.3% and 76.2%, 42.9%, 30.6%, respectively (χ2=7.851, P 0.05). In univariate analysis, tumor location, tumor stage, lymph node metastasis, TNM stage, the cycle of chemotherapy, treatment modalities were significantly associated with survival (RR=1.735, P<0.05). The multivariate analysis only showed TNM stage were independent factors of prognosis. Conclusions The results suggested that resection plus chemotherapy could improve the prognosis of early-stage(stageⅠ+ Ⅱ) small cell lung cancer, but patients in ⅢA stage should received the definitive chemoradiotherapy. The TNM stage was still the independent factor of prognosis. Key words: Limited-disease; Small cell lung cancer; Resection plus chemotherapy; Chemoradiotherapy; Prognostic

We analyzed the lymph node (MLNs) metastasis of thoracic esophageal squamous cell carcinoma (ESCC) to explore the patterns of lymphatic spread and the rational surgical procedure and extent of lymph node dissection for ESCC. We retrospectively evaluated 313 consecutive patients treated in our hospital between January 2010 and May 2014 who underwent minimally invasive esophagectomy (MIE) for ESCC. The information of lymph node status was obtained and the features of lymph node metastasis were analyzed. Of the 313 cases, 122 (39.0%) were found to have lymph node metastasis. In the 4461 dissected lymph nodes, metastasis was identified in 294 (6.6%) lymph nodes. The recurrent laryngeal nerve lymph nodes were the most frequent metastatic nodes with a metastasis rate of 25.2%, followed by the paracardiac and left gastric artery lymph nodes (18.2%). Chi-square test showed that the lymph node metastasis is associated with tumor invasion and tumor differentiation (P<0.001 for both). Metastases were more frequently found in the recurrent laryngeal nerve lymph nodes in patients with tumors in the upper third esophagus and with histologically poor differentiation (P<0.05 for both). The metastasis rate of para-cardiac and left gastric artery lymph nodes was associated with tumor in the lower third of esophagus, T stage and differentiation (all P<0.05). Logistic regression analysis showed that tumor differentiation and location are independent factors affecting the metastasis of recurrent laryngeal nerve lymph nodes (P<0.05 for all). T stage, tumor differentiation and location were independent factors associated with metastasis of para-cardiac and left gastric artery lymph nodes (P<0.05 for all). (1) Metastases of thoracic esophageal carcinoma are often found in the recurrent laryngeal nerve lymph nodes, para-cardiac and left gastric artery lymph nodes. (2) Extensive lymph node dissection should be performed for ESCC with poor differentiation and deep tumor invasion.

Objective To study the application of two-field dissection and three-field dissection in esophageal dissection.Methods From January 2013 to January 2014,100 patients with thoracic esophageal carcinoma in our hospital were randomly divided into two-field group and three-field group,with 50 cases in each group.The patients in two-field group received two-field lymphadenectomy joint with Ivor-Lewis surgery,and the patients in three-field group received three-field lymphadenectomy combined with Akiyama surgery,the clinical effects were compared.Results Two wild group statistics show the total number of lymph node dissection was 950,ranged 8-39 months,the number of metastatic lymph nodes was observed,a total of 86 ; three wild groups in the total number of lymph node dissection was shown by the statistics of 1500,the range of 16-62 months to observe the lymph nodes turn more than the number of 80 ; statistical incidence of lymph node metastasis in two wild group was 48％,three wild was 70 ％,there was significant difference between the two groups (P ＜ 0.05).Plus two wild groups of mediastinal lymph node metastasis of gastric rate of 50％,the three groups was 44％ ; two wild celiac lymph node metastasis group was 4％,6％ for the three wild group,there was no significant difference(P ＞ 0.05).Two wild group under the mediastinal lymph node metastasis was 32 ％,three wild group was 24％,there was significant difference(P ＜0.05).Two wild group of cervical lymph node metastasis can not be assessed,three wild group was 42％.Between groups and mediastinal lymph node metastasis rate of major differences.Two-field lymph node metastases 3.3,the statistical showing three wild group 2.8,are in the range of 1 to 10,there was no significant difference(P ＞0.05).Two-field lymph node dissection group was 17.0,after statistics show three wild group 32,there was significant difference (P ＜0.05).Two wild group of lymph node degrees is 10％,significantly higher than the three wild groups of 6％ in the second transfer roughness wild group was 20％,significantly higher than the 10percent three wild groups were statistically significant (P ＜ 0.05).Conclusions Line two field lymph node dissection when required to be better left recurrent laryngeal nerve lymph node dissection,threefield lymphadenectomy prognosis may be classified as an ideal,turn on the neck and mediastinum can remove multiple lymph nodes,the abdominal lymph nodes,neck dissection and mediastinal lymph nodes,the prognosis of esophageal chest cancer can be significantly improved. Key words: Chest segment; Esophageal cancer; Lymph nodes ; Two-field dissection ; Three-field dissection ; Contrast

Objective To investigate the pattern of lymph node metastasis and analyze prognostic factors of Siewert type Ⅱ adenocarcinoma of esophagogastric junction (AEG). Methods The retrospective case-control study was conducted. The clinicopathological data of 368 patients with Siewert type Ⅱ AEG who were admitted to Tianjin Medical University Cancer Institute and Hospital from June 2010 and November 2015 were collected. There were 323 males and 45 females, aged from 35 to 80 years, with an average age of 64 years. Of 368 patients, 209 underwent left transthoracic surgery, 112 underwent thoracoabdominal surgery, and 47 underwent Ivor-Lewis surgery. Observation indicators: (1) total lymph node metastasis and metastasis of various lymph node stations; (2)follow-up and survival; (3)prognostic factors analysis; (4) influencing factors affecting thoracic lymph node metastasis. Follow-up using outpatient examination and telephone interview was performed to detect survival of patients up to November 2018. Measurement data with skewed distribution were represented as M (range). Count data were represented as absolute number or percentage. The survival time and rate were calculated using the Kaplan-Meier method. The univariate and multivariate analyses were done by the COX proportional hazard model. Results (1) Total lymph node metastasis and metastasis of various lymph node stations: the total lymph node metastasis rate was 66.58%(245/368) in 368 patients. The metastasis rates of abdominal lymph nodes, thoracic lymph nodes, lower mediastinal lymph nodes, and upper mediastinal lymph nodes were 65.49%(241/368), 12.77%(47/368), 12.23%(45/368), and 1.09%(4/368), respectively. The order of metastasis rate of various lymph node stations from high to low was 51.99%(170/237) of No.7 left gastric artery, 34.23%(89/260) of No.1 right paracardial region, 33.88%(83/245) of No.2 left paracardial region, 28.91%(85/294) of No.3 lesser curvature, 27.10%(29/107) of No.11 splenic artery, 19.75%(16/81) of No.9 celiac trunk, 15.25%(36/236) of No.E8Lo lower paraesophageal region, 11.94%(16/134) of No.4 greater curvature, 11.76%(6/51) of No.E8M middle paraesophageal region, 11.11%(10/90) of No.8 common hepatic artery, 4.65%(4/86) of No.E9L left inferior pulmonary ligament and 3.39%(2/59) of No.E7 subcarinal region. (2) Follow-up and survival: of the 368 patients, 309 were followed up for 1-103 months, with a median follow-up time of 38 months. The survival time of 309 patients was 0.7-101.9 months, and the median survival time was 35.9 months. During the follow-up, the postoperative 1-, 2-, 3-year overall survival rates were 85.9%, 68.6%, and 58.7%, respectively. (3) Prognostic factors analysis. Results of univariate analysis showed that tumor differentiation degree, presence of thoracic lymph node metastasis, number of metastatic lymph nodes, T staging, tumor diameter, and length of esophageal invasion were associated factors affecting prognosis of patients (χ2=8.776, 26.582, 46.057, 18.679, 22.460, 9.158, P<0.05). Results of multivariate analysis showed that presence of thoracic lymph node metastasis, number of metastatic lymph nodes, T staging, and tumor diameter were independent influencing factors for prognosis of patients [odds ratio (OR)=1.699, 1.271, 1.422, 1.238, 95% confidence interval: 1.102-2.621, 1.019-1.481, 1.090-1.856, 0.971-1.481, P<0.05]. (4) Influencing factors affecting thoracic lymph node metastasis: results of univariate analysis showed that tumor diameter, length of esophageal invasion, number of lymph lodes harvested in thorax were related factors for thoracic lymph node metastasis (χ2=5.129, 43.140, 10.605, P<0.05). Results of multivariate analysis showed that length of esophageal invasion ≥2 cm, number of lymph lodes harvested in thorax ≥4 were independent risk factors for thoracic lymph node metastasis (OR=6.321, 1.097, 95% confidence interval: 2.982-13.398, 1.026-1.173, P<0.05). Conclusion Lymph node metastasis of Siewert type Ⅱ AEG spreads two regions, mainly at abdominal lymph nodes, followed by the thoracic lymph nodes. Presence of thoracic lymph node metastasis, number of metastatic lymph nodes, T staging, and tumor diameter are independent influencing factors for prognosis of patients. Presence of thoracic lymph node metastasis indicates poor prognosis of patients. Length of esophageal invasion ≥2 cm and number of lymph lodes harvested in thorax ≥4 are independent risk factors for thoracic lymph node metastasis. Key words: Esophagogastric junction neoplasms; Adenocarcinoma of esophagogastric junction; Siewert Ⅱ type; Thoracic lymph node metastasis; Esophageal invasion; Prognosis

Objective To investigate the local-regional recurrence in thoracic esophageal cancer after radical surgery including two-field lymph node dissection and provide evidence for postoperative radiotherapy. Methods We reviewed local-regional recurrence for 134 cases with esophageal squamous cell carcinoma after radical surgery from 2004 to 2009. Results In 134 cases, lymph node metastasis rate,anastomosis recurrence rate and tumor bed recurrence rate was 94. 0%, 9. 7% and 3.7%, respectively. As to the 126 cases with lymph node metastasis, significant difference was detected between mediastinal metastasis, supraclavicular metastasis and abdominal lymph node metastasis (80. 2%, 43.7% and 13.5%,respectively, χ2= 113. 15, P = 0. 000). Furthermore, the relative metastasis rate in upper mediastinum,middle mediastinum and the lower mediastinum was 73.8%, 39.7% and 1.6%, respectively, the difference was statistically significant ( χ2 = 139. 11, P = 0. 000 ). Significant difference was identified between right and left supraclavicular lymph node metastasis (31.7% vs 16. 7%, χ2= 7. 81, P = 0. 005 ).To confirm the analysis above,lymph node metastasis rate of left recurrent laryngeal nerve nodes, (including region 1L, 2L, 4L and 5) ,right recurrent laryngeal nerve nodes, azygos nodes, subcarinal nodes, and 2R region was 38.9%, 43.7%, 15.1%, 34.1% and 25.4%, respectively. Conclusions The main characteristics of local-regional recurrence may be lymph node metastasis for esophageal squamous cell carcinoma after radical surgery. On the contrary, tumor bed recurrence is rare. Dangerous regions include supraclavicular nodes, recurrent laryngeal nerve nodes, azygos nodes as well as subcarinal nodes. Key words: Esophageal neoplasms/surgery; Neoplasms recurrence,postoperation; Lymph node metastasis, postoperation

To investigate the characteristics of lymphatic metastasis in different types of adenocarcinoma of the esophagogastric junction (AEG) and provide guidance for surgical approach adoption. Clinical data of 228 patients with AEG undergoing surgery were analyzed retrospectively. According to Siewert classification, there were 9 cases of type I (3.9%) who all underwent left thoracoabdominal approach procedures. A total of 121 patients belonged to type II (53.1%), of whom 12 underwent left transthoracic approach, 48 left thoracoabdominal approach, and 61 transabdominal approach. Ninety-eight patients belonged to type III (43%), of whom 22 underwent left thoracoabdominal approach procedures and 76 transabdominal. The pattern of lymph node metastasis was analyzed and the association between surgical approach and oncological clearance was examined. The resection margin was positive in 20(8.8%) patients, including 10 with type II (8.3%) and 10 with type III (10.2%), and the difference was not statistically significant (P>0.05). The rate of positive resection margin was 12.4%(17/137) in the transabdominal group and 16.7%(2/12) in the left transthoracic group, both significantly higher than the left thoracoabdominal group (1.1%, 1/88) (both P<0.05). Lymph node metastasis was found in 159(69.7%) patients. The metastasis was found in 4 of 9 patients with type I cancer and two were thoracic metastasis, no metastasis was found in the upper mediastinum. For type II cancer, the rate of lymph node metastasis was 66.9%(81/121), including thoracic metastasis ( n=32, 26.4%) and abdominal metastasis (n=81, 66.9%). For type III cancer, the rate of lymph node metastasis was 66.9%(81/121), including thoracic metastasis (n=15, 15.3%) and abdominal metastasis (n=69, 70.4%). For type I AEG, left thoracoabdominal approach should be used because the pattern of lymph node metastasis is similar to that of the distal esophageal carcinoma. For type II , left thoracoabdominal approach should be used to ensure adequate resection of the tumor and clearance of lymph node in the lower esophagus and upper mediastinum because of high rate of intrathoracic lymph node metastasis. For type III cancer, transabdominal incision offers better benefit with less impact on respiratory function. However, thoracic incision should be used to ensure adequate clearance for tumors of larger size and significant external invasion.