68-4

51/136

the cartridge. The data of the fluorescent colors that appear on the upper surface of the cartridge are processed into fluorescence images for analysis and assessment. We defined CTC positive when CTC count is at least one in 7.5 ml of blood sample.We measured serum tumor makers (cytokeratin fragment [CYFRA], squamous cell carcinoma antigen [SCC], carbohydrate antigen 19-9 [CA19-9], and carcinoembryonic antigen [CEA]) of same samples. Blood samples for CTCs and tumor makers were collected from first visit of our depart-ment to beginning of any treatment. Statistical AnalysesAll statistical analyses were performed using IBM SPSS advanced statistics ver.25. The chi- Number of casesAgeSexTreatmentPrimary tumor locationPredominant Histological typecT classification*Clinical stage ** UICC TNM Classification of Malignant Tumors, 7th edCe: Cervical esophagus, Ut: Upper thoracic esophagus, Mt: Middle thoracic esophagus, Lt: Lower thoracic esophagusSqCC: Squamous cell carcinomaTable 1　Patients’ characteristics38 cases66 (47-84) years old MaleFemale Surgery Esophagectomy with three(two) -fields lymph node dissection Laryngectomy with cervical esophagecto-my Endoscopic treatmentChemotherapy alone/ Chemoradiotherapy CeUtMtLtWell-differentiated SqCCModerately differentiated SqCCPoorly differentiated SqCCOthers (adenocarcinoma, Basaloid carcinoma)cT1　cT2cT3cT4ⅠIIIIIIVsquare test was applied for the differences of patients’ characteristics and CTCs status between the CTCs-positive and the CTCs-negative groups. Serum CYFRA, CA19-9, SCC and CEA levels are divided into two groups of over and within normal range. Differences were considered significant at p-value <0.05. Values are expressed as median (range).OS and disease-free survival (DFS) rates were calculated using the Kaplan–Meier method, and univariate analyses were performed by the Log-rank test. The DFS was defined as duration (days) from the date of surgery (or the first treat-ment day of chemotherapy or chemoradiotherapy patients) to the first relapse of cancer or death from any cause. Cox hazards regression analysis 33 (86.8%)5 (13.2%)35 28 (73.7%)1 (2.6%)6 (15.8%)3 (7.8%)4 (10.5%)7 (18.4%)14 (36.8%)13 (34.2%)10 (26.3%)21 (55.3%)2 (5.3%)5 (13.2%)15 (39.5%)4 (0.5%)15 (39.5%)4 (10.5%)15 (39.5%)5 (13.2%)14 (36.8%)4 (10.5%)371