| Part Ⅰ Overview and analysis of clinical characteristics of patients with thyroid cancer of Anhui ProvinceBackground and Objective:Thyroid cancer is the most common endocrine-related malignant tumor,accounting for more than 90%of endocrine cancers.The incidence of thyroid cancer has increased significantly over the past four decades.The latest data from the "National Cancer Report 2022" shows that in 2016,the number of cases of thyroid cancer diagnosed nationwide was 50,424,and the number of female patients was more than triple that of male cases;the crude incidence rate of thyroid cancer nationwide was 13.22/100,000,the incidence rate is 9.70/100,000.A series of clinical studies have reported the clinical value of common clinical indicators in differentiating thyroid diseases and predicting the prognosis of thyroid patients.In the first part,according to the pre-set inclusion and exclusion criteria,we collected the clinical data of patients with thyroid cancer and conducted comparative analysis in order to find characteristic indicators that can reflect the clinical conditions of patients with thyroid cancer.Methods:According to the pre-set inclusion and exclusion criteria,after approved by the hospital ethics committee,we included patients who underwent thyroid cancer surgery in our hospital from July 1,2019 to July 31,2021 for follow-up analysis.We first recorded basic demographic information,medical history and living habits,thyroid ultrasonography results,pathological information after surgery,and data such as thyroid hormone levels,blood cell routine,hemostatic indicators,and biochemical indicators.For continuous data,we used independent samples t-test or Wilcoxon rank-sum test to evaluate differences between groups.For categorical variables,the chi-square test was used,and when the cells with the expected frequency less than 5 accounted for more than 25%of the total cells in the chi-square test,in order to reduce the error,we used Fisher’s exact test to calculate the statistical significance.All tests were two-tailed,and a P value<0.05 was considered statistically significant.Result:A total of 253 patients with thyroid cancer who underwent surgery were included in this study,including 232 primary patients and 21 recurrent patients.There was no statistical significance in gender distribution between the two groups(P=0.461).The average age of recurrent patients(51.5± 12.2 years)was higher than that of primary patients(43.0± 11.4 years),and the difference was statistically significant(P=0.00541).There were no significant differences in height,weight,BMI,diabetes,hypertension,alcoholism,smoking and other factors between the two groups.According to ultrasonography and postoperative pathological data,the distribution of primary and recurrent thyroid cancer in pathological type(98.4%of patients was PTC),lymph node metastasis(recurrent 76.2%vs primary 53.0%,P=0.0647),tumor stage(stage Ⅱ:recurrent 52.4%vs primary 6.0%,P<0.001),multifocal thyroid cancer(recurrent 47.6%vs primary 23.7%,P=0.0331)and tumor distant metastasis(recurrent 76.2%vs There are differences in primary 0,M1 stage)and so on.The sensitivity of ultrasonography for abnormal lymph nodes was only 32.456%,but the specificity was 80.567%.Generally,in patients with recurrent thyroid cancer,peripheral blood T4 levels(121 ± 21 vs.109 ± 23.3 nmol/L,P=0.0279)and D dimer(0.484 ± 0.269 vs.0.0819)and phosphorus ion levels(1.11 ± 0.189 vs.1.21 ± 0.201 mmol/L,P=0.0292),while TSH,T3,TgAb,Tg,FT3,FT4,and PTH and other blood indicators related to thyroid function and Other biochemical indicators such as PT,APTT,PLT,RBC,WBC,K,Ca,CL,Na,fasting blood glucose,BUN,and Cr had no significant difference between the two groups.The above results showed that there were significant differences between primary and recurrent thyroid cancer in patients’ age,pathological type,lymph node metastasis,multifocal thyroid cancer,distant tumor metastasis and some blood indicators.The 232 patients with primary thyroid cancer included 109 patients with negative lymph node metastasis and 123 patients with positive lymph node metastasis.The comparison of the two groups showed that there was a statistically significant difference in gender distribution(P=0.0436).The mean ages of the negative lymph node metastasis group and the positive lymph node metastasis group were 45.3 ± 11.8 years old and 40.9 ± 10.8 years old,and younger patients were more likely to Positive lymph node metastasis appeared,and the difference was statistically significant(P=0.00312).However,there was no significant statistical difference between the two groups in terms of BMI,height,weight,history of diabetes,history of hypertension,history of alcoholism,history of smoking,long-term consumption of iodized salt,and life stress.The preoperative ultrasound TI-RADS score showed a significant difference between patients with negative and positive lymph node metastasis,and the preoperative TI-RADS score of positive patients was higher.In terms of abnormal lymph nodes and calcifications indicated by ultrasound,the positive lymph node metastasis group had more abnormal lymph node risk indicators(78.9%vs.67.9%,P=0.0728),but there was no significant difference between the two groups in the indications of calcifications(P=0.19).In terms of the largest tumor diameter,the lymph node metastasis negative group was 0.808 ± 0.616,and the positive group was 1.11 ± 0.832,and there was a significant statistical difference between the two groups(P=0.00328).At the same time,the proportion of multiple lesions in the positive group(26.8%)was also higher than that in the negative group(20.2%).After observing the distribution of peripheral blood biochemical indicators,we found that the positive group was in PT%(110 ± 13.0%vs.107 ± 11.5%,P=0.034)and WBC(6.11 ±1.76×10^9/L vs.6.59±1.85×10^9/L,P=0.0495)was higher than that in the negative group,but there was no difference in TSH,T3,T4,TgAb,Tg,FT3,FT4 and PTH between the two groups.In addition,we also found that BMI(R=0.19),BUN(R=0.17)were positively correlated with PT%,while PT%was negatively correlated with peripheral blood T3(R=-0.17),T4(R=-0.13)levels.This finding suggests that age,sex,tumor diameter,some ultrasonographic scores,and certain blood test parameters may be critical in primary thyroid cancer and lymph node metastases.Conclusion:1.The average age of patients with recurrent thyroid cancer is significantly higher than that of primary patients,and the frequency of abnormal lymph nodes indicated by ultrasound in patients with recurrent thyroid cancer is higher,but the incidence of calcifications is lower.2.Compared with patients with primary thyroid cancer,recurrent patients are more likely to have high tumor stage and lymph node metastasis,and 76.2%of recurrent patients have distant metastasis of tumor.3.The levels of T4、D-Dimer and phosphorus ions in patients with recurrent thyroid cancer were significantly higher than those in primary patients.4.Among patients with primary thyroid cancer,male and younger patients are more likely to have positive lymph node metastasis,and patients with positive lymph node metastasis tend to have higher TI-RADS scores.5.The tumor diameter of positive lymph node metastasis group was generally larger,and the proportion of multiple lesions was also higher.6.There were significant differences in peripheral blood prothrombin activity and white blood cell counts between patients with positive lymph node metastasis and patients with negative lymph node metastasis.Part Ⅱ Construction and validation of a predictive model for the risk of pathological lymph node metastasis based on genetic alterations Background and Objective:The treatment and prognosis of thyroid cancer are related to many factors,among which lymph node metastasis(LNM)is an important and common risk factor.Early detection and prediction of lymph node metastasis has become a challenge in clinical practice in recent years.The 5-year survival rate of global PTC patients is about 98.2%,but the 5-year survival rate of Chinese thyroid cancer patients is only 84.3%.Lymph node metastasis often leads to poor prognosis of thyroid cancer patients.With the application of more and more molecular detection in clinical practice,a series of molecular markers of thyroid cancer have been proved to be related to the prognosis of the disease.In the second part,we detected the molecular characteristics of the thyroid cancer patients included in the analysis in the first part,and determined the predictive factors that could reflect the presence of LNM,in order to construct a new prediction model that can predict LNM in thyroid cancer patients.Methods:We first obtained the corresponding pathologically confirmed tissue wax block samples from the 253 thyroid cancer patients included in the analysis in the first part,DNA was extracted using HiPure FFPE DNA Kit and RNA was extracted using RNA Pure FFPE Kit fixed tissue RNA extraction kit,and quantified.Then construct DNA and RNA libraries for library quality inspection.Afterwards,we constructed DNA and RNA libraries and performed library quality checks.Next Generation Sequencing(NGS)was used to sequence the library,and the sequencing results were analyzed,and specific SNV&Indel filter conditions were set.Independent predictors of lymph node metastasis were screened by multivariate logistic regression analysis.Based on the predictors,a nomogram for assessing the risk of lymph node metastasis was constructed,and its accuracy was tested using calibration curves,decision curves,and clinical impact curves.To verify the predictive ability of the nomogram,we obtained the clinical and pathological information,gene mutation matrix,and gene expression profile of the TCGA-THCA cohort,and further evaluated the predictive ability of the nomogram for lymph node metastasis in an external cohort.Find the downstream genes that are affected when the RET gene is changed,and identify the signaling pathways that may be affected by enrichment analysis.T test or Mann-Whitney U test was used to evaluate the significant difference of continuous variables.Multivariate logistic regression analysis was used to explore the influence of each feature on the results.All statistical analyzes were completed in the R software environment.Results:Among the 253 thyroid cancer patients we analyzed,the BRAF gene was the most common,with mutations in this gene occurring in 64.4%of patients.We noticed that in the primary and recurrent groups,the BRAF gene(90.5%vs.62.1%,P=0.015)and TERT gene(33.3%vs.2.6%,P<0.001)mutation rates in the recurrent group Significantly higher than the primary group.Moreover,90.5%of patients in the recurrent group had at least one mutated gene,which was significantly higher than 88.0%in the primary group(P<0.001).In the lymph node metastasis-negative and positive groups,although there was no significant difference in the distribution of mutations in most genes,we found that the proportion of gene fusions of the RET gene was as high as 13.0%in the lymph node metastasis-positive group,which was significantly higher than 1.8%in the lymph node metastasis-negative group%(P=0.001).Using multivariate logistic regression analysis,we identified age,tumor diameter,and RET fusion type as significant predictors of lymph node metastasis.We constructed a nomogram for the prediction of lymph node metastasis,including the four factors of age,gender,tumor diameter,and RET change.This prediction model demonstrated high clinical net benefit under the evaluation of DCA curve and clinical impact curve,and Under ROC curve analysis,its AUC value reached 0.724,and its sensitivity and specificity were 82.7%and 52.9%respectively.It is an effective tool for predicting the probability of lymph node metastasis.In the validation of the TCGA-THCA cohort(390 thyroid cancer patients),we found that men were more likely to have positive lymph node metastasis(32.0%vs.19.0%,P=0.0058),and patients with positive lymph node metastasis had larger tumor diameters than those with negative lymph node metastasis.Larger(≥ 1 cm:95.9%vs.90.1%,P=0.0396),and a greater proportion of RET gene alterations(7.3%vs.1.7%,P=0.0208).The DCA curve and the clinical impact curve further validated the high clinical net benefit of the nomogram prediction model.Further studies revealed a potential correlation between RET fusions and lymph node metastasis in thyroid cancer patients.There are 133 upregulated genes in RET fusion samples.The biological processes and molecular functions associated with these genes have undergone significant changes,including extracellular matrix organization,granulocyte migration and cell adhesion.In addition,network analysis showed that LOX,DPP4,CCL17,CCL13,CCL18,CYP1B1,COL8A2,etc.were key genes.RET gene mutations also trigger the activation or inhibition of multiple key signaling pathways,such as protein digestion and absorption,ECM-receptor interaction,interaction between cytokines and cytokine receptors,chemokine signaling pathways,etc.Conclusion:1.The mutation rates of BRAF gene and TERT gene in recurrent patients were significantly higher than those in primary patients.2.In patients with primary thyroid cancer with positive lymph node metastasis,the fusion mutation rate of RET gene was significantly higher than in patients with negative lymph node metastasis.3.Multivariate logistic regression analysis,patient age,tumor diameter and RET fusion were confirmed as independent predictors of lymph node metastasis.4.We constructed a positive prediction nomogram for lymph node metastasis based on patient gender,age,tumor diameter,and RET gene changes,and verified the good predictive performance of the nomogram in an external cohort.5.RET gene fusion may affect the development of thyroid cancer by affecting signaling pathways related to extracellular matrix,oxidoreductase and chemokine activity. |
PDF Full Text Request