The Study Of Recurrent Sites For Thoracic Esophageal Carcinoma After Surgery

PartⅠ: Analysis of the relationship between recurrent site of esophagealcarcinoma after surgery and preoperative lymph node metastasisOBJECTIVE: To analyse local regional recurrent patterns of thoracic esophagealcarcinoma and relationship between lymph node status at operation and the lymphnodes recurrent site.METHODS:We reviewed local-regional recurrence for175cases with esophagealsquamous cell carcinoma after Three-field lymph node dissection(3-FLD) withoutProphylactic radiotherapy from2005to2010and analysed local regional recurrenttype and site of thoracic esophageal carcinoma, recurrent lymph node metastasis ratein each site and relationship between lymph node status at operation and the lymphnodes recurrence using the chi-square test.RESULTS: In175thoracic esophageal carcinomas(TECs) that underwent curativesurgery, lymph node metastasis rate,anastomosis recurrence rate and tumor bedrecurrence rate was90.29%,19.43%and9.71%, respectively. As to the158caseswith lymph node metastasis, mediastinal metastasis rate, cervical metastasis rate andabdominal lymph node metastasis rate was72.78%,49.37%and20.25%(x2=87.769,P=0.000), respectively. In175TECs, superior mediastinal lymph nodemetastasis rate,middle mediastinal nodes rate and inferior mediastinal nodes rate was67.72%,28.48%and1.27%, respectively. There was significant difference among thethree groups(x2=160.802,P=0.000). The presence of cervical lymph nodemetastasis(LNM) was not significantly associated with locoregional recurrence(x2=0.557, P=0.455). The presence of subcarinal LNM was significantly associatedwith locoregional recurrence (x2=4.533,P=0.033).The positive predictive value,negative predictive value, sensitivity, and specificity were32.43%,83.33%,34.29%,and82.14%, respectively. The presence of abdominal LNM was significantlyassociated with locoregional recurrence (x2=34.498,P=0.000). The positive predictivevalue, negative predictive value, sensitivity, and specificity were75%,78.32%,43.63%,and93.33%, respectively. CONCLUSIONS: The main characteristics of local-regional recurrence may belymph node metastasis for esophageal squamous cell carcinoma after radical3-FLD.The presence of abdominal and subcarinal LNM at operation may be associated withcorresponding locoregional recurrence. While the presences of other LNM were not. PartⅡ: The impact of postoperative radiotherapy to the relationshipbetween recurrent site of esophageal carcinoma after surgery andpreoperative lymph node metastasisOBJECTIVE:To analyse local regional recurrent patterns of thoracic esophagealcarcinoma after three-field lymph node dissection(3-FLD)with and withoutpostoperative radiotherapy, and to assess the value of postoperative radiation for localcontrol.METHODS:We reviewed local-regional recurrence for239cases with esophagealsquamous cell carcinoma after3-FLD, including175cases without prophylacticradiotherapy and64cases with radiotherapy from2005to2010and analysed localregional recurrence patterns.RESULTS: In239thoracic esophageal carcinomas(TECs) that underwent curativesurgery, the sequence of lymph nodes recurrence proportion in both two group fromhigh to low is mediastinal, cervical and abdominal. The recurrent proportion of thesuperior mediastinal lymph node in the patients of preoperative superior mediastinalLNM without radiotherapy is higher than that with radiotherapy (x2=8.85,P=0.03).Compared to patients of preoperative cervical LNM (x2=0.026,P=0.873), middlemediastinal LNM (x2=0.069,P=0.793), subcarinal LNM (x2=0.005,P=0.943) butwithout postoperative radiotherapy, the recurrent proportion of corresponding lymphnode was not significantly lower in patients with postoperative radiotherapy.CONCLUSIONS: Postoperative radiotherapy play role in reducing the recurrence inpatients with the superior mediastinal LNM. Postoperative radiotherapy could not significantly reduce the recurrent proportion of cervical, middle(subcarinal) LNM. PartⅢ: The establishment of predictive model for recurrence in each siteof esophageal carcinoma after surgeryOBJECTIVE:To study the influential factors of local recurrence for esophagealcarcinoma after three-field lymph node dissection(3-LND), and to provide evidencefor postoperative therapy.METHODS:We reviewed1004cases with thoracic esophageal carcinoma(TEC)from2005to2009after3-FLD in our hospital. Logistic regression analysis was usedto identify relative factors for cervical, mediastinal, and abdominal lymph noderecurrence, and then the predictive model for recurrence was established.RESULTS:The lymph node recurrence for esophageal carcinoma after3-LNDincluded cervical(33.33%), mediastinal(45%), and abdominal lymph node recurrence(21.4%). Multivariate analysis showed that N stage and preoperative cervical lymphnode metastasis(P=0.002) were risk factors for cervical lymph node recurrence, whilepostoperative chemotherapy was protective factor(P=0.013). The LogisiticRegression Equation for cervical lymph node recurrence was: Logit(Y)=1.602×(X8)+2.178×(X9)+2.857×(X10)+1.313×(X18)-0.607×(X27)-2.498. T stageand N stage(P<0.001)were risk factors for mediastinal lymph node recurrence. TheLogisitic Regression Equation for mediastinal lymph node recurrence was:Logit(Y)=0.994×(X5)+1.718×(X6)+1.855×(X7)+1.498×(X8)+2.210×(X9)+2.529×(X10)-2.218. Primary tumor location and preoperative abdominallymph node metastasis(P<0.001)were risk factors for abdominal lymph noderecurrence. The Logisitic Regression Equation for abdominal lymph node recurrencewas: Logit(Y)=16.803×(X3)+17.961×(X4)+3.128×(X21)-15.159.CONCLUSIONS: For patients with TEC after3-LND, the cervical lymph noderecurrence was associated with N stage and preoperative cervical lymph node metastasis； mediastinal lymph node recurrence was associated with T stage and Nstage；abdominal lymph node recurrence was associated with primary tumor locationand preoperative abdominal lymph node metastasis,. Patients with risk factors abovewere suggested to perform postoperative adjuvant therapy.