Analysis Of Lymph Node Metastasis In Early Stage Non-small Cell Lung Cancer&Exploration Of New Tumor Marker For Lung Cancer

Objective:With the prevalence of LDCT(low dose computed tomography)and the realization of importance on the health control,there has been more and more early stage non-small cell lung cancers(NSCLC)founded than before.It is crucial to assess the lymph node status aiming to choosing the suitable surgical procedure for pathology NO stage NSCLC patients,for the segmentectomy could work the same as the standardized lobectomy.In order to study the correlated factors of lymph node metastasis and assist surgeons make the medical treatment,a predictive model was developed by analyzing the clinicopathological characteristics of clinical Tla7th stage NSCLC patients.Method:Information of patients from 2 hospitals,who had undergone the standardized lobectomy and the systematic lymph node dissection from 2010-2015,was collected as the model group retrospectively.For the validation group,patients with the same clinical stage NSCLC like the model group were enrolled from January to April in 2016.In the study of clinical T1aN0-2MO7th stage NSCLC,all the patients were from one hospital.And patients from the two hospitals were mixed for the study of clinical T1aNOMO7th stage NSCLC.Univariate logistic regression analysis was used to determine the relationship between positive lymph nodes and variables,and factors with a p value less than 0.05 were included in binary logistic regression.Categorical data was compared using the Pearson’s chi-square test or the Fisher exact test,and independent sample t test for continuous variables.The backward stepwise selection procedure was employed to screen significant factors(p<0.05),which were used in the predictive model.The concordance between predicted and observed probabilities was determined with application of the Hosmer-Lemeshow goodness-of-fit test(p>0.05).AUC(area under the receiver)of the ROC(receiver operating characteristic)was used to assess the discriminative ability.The value of AUC ranges from 0.5 to 1.0,with a value of 0.5 indicating the model unable to distinguish between patients with or without the clinical event,and a value of 1.0 indicating perfect discrimination.The statistical analysis was performed using the statistical software SPSS for IBM 22.0 and R(version 3.3.1).The institutional review board of Cancer Hospital approved the study.As it was a retrospective study,the necessity to obtain written informed consent from each patient was waived.Result:①474 patients were enrolled in the group of clinical T1aNO-2MO7th stage,and the ratio of male and female patients was 243:231.The mean age was 58.4±8.7 years,and the percentage of positive lymph nodes was 17.3%.No patients in the pure GGO group was confirmed with positive lymph nodes,and only 1 was in the mixed group.The left 81 patients were all in the pure solid tumor group.In the clinical NO stage group(408 patients),there were 45 patients with positive nodes;in the clinical N1-2 stage group(66 patients),the percentage was 43.9%(37/66).With the statistical analysis,4 factors were included in the predictive model:the tumor location,the consolidation size,the status of tumor marker and the clinical N stage.The result of Hosmer-Lemeshow goodness-of-fit test was not statistically significant(p=0.766),indicating a high concordance between the predicted and observed probabilities.The AUC was good at 0.842(95%CI:0.797-0.886).For validation group,the accuracy of the prediction model was reasonable,and the AUC was 0.810(95%CI:0.731-0.889).②498 patients from the 2 hospitals were in the group of clinical T1aNOMO7th stage,lymph node disease(pN1 or pN2)was found in 50 patients(the incidence was 10%).The central tumor location,the size of consolidation,and the abnormal status of tumor-marker were identified as independent risk factors that be used to develop the model.The calibration of the model was good(Hosmer-Lemeshow goodness-of-fit test:p=0.265),and the area under the receiver operating characteristic cure(ROC)was 0.755(95%confidential interval,0.696 to 0.813).For the validation group,the area under the ROC was 0.794(95%confidential interval,0.713 to 0.876).Conclusion:There were 4 factors were considered as indicators for the lymph node metastasis:the tumor location,the consolidation size,the tumor marker status and the clinical lymph node stage(except the clinical NO stage patients).The 2 predictive models were used to assess the likelihood of lymph node metastasis efficiently.For patients with high predictive level,the more advanced or the further test was suggested for more accurate lymph node stage and making the suitable treatment decision.Furthermore,the standardized lobectomy combined with systematic lymph node dissection was performed of priority.For patients with low predictive level,the surgical treatment should be suggested if there is no contraindication.And for patients with no positive lymph nodes confirmed in surgery,the segmentectomy may be chosen as well.Aim Lung cancer has already been the most frequent malignant tumor in China,and some patients with more advanced stages disease were not candidates for the lack of early detection.Although,the LDCT screening programe has done a lot in founding small lung nodules,it did a little in differentiating the benign and malignant.It’s a crutial key to explore new tumor marker combined the LDCT screening for the controlling of lung cancer.Method A triplex real-time PCR assay for methylated DNA of SHOX2 and PTGER4 and ACTB as control assay was developed.Total DNA was extracted from 3.5ml plasma samples and bisulfite converted utilizing a commercially available kit.After purification DNA was assayed in PCR triplicates.Cycle threshold values were aggregated utilizing a predefined algorithm.Receiver operating characteristic and the area under the curve(AUC)were analyzed.Result 102 patients were enrolled in the study,and 13 patiets were with benign nodules,the left 89 patients were with lung cancers.By the kit,34 patients were diagnosed with positive results(3 were with benign tumors and 31 were with lung cancers),and 68 patients were with negative results.The sensitivity is 34.8%,and the specificity is 76.9%.17 patients were diagnosed with squamous cell lung cancer,70 patients were with adenocarcinoma,and 2 were with small cell lung cancer.The sensitivity for pT1,pT2,pT3,pT4 stage patients were 18.9%,51.7%,83.3%,100%(p<0.001).In patients with well differentiation group,the sensitivity was 20.8%;in the media differentiation group,the sensitivity was 31.1%;in the poor differentiation group,the sensitivity was 60.0%(p=0.009).Conclusion The methylation of SHOX2 and EP4 in patients with different stages and differentiation level of lung cancers are different,statistically.The clinical utility of the combination of DNA methylation markers SHOX2 and EP4 in distinguishing lung cancers from non-malignant diseases still needs to be studied with more enrolled patients.