Study On Proteomics Diagnostic Model With Serum And Saliva Of Thyroid Carcinoma In Combination Of Disease And Syndrome

Thyroid carcinoma is a common malignant in head and neck and the most common endocrine malignancy.Incidence of thyroid carcinoma increased significantly in recent years in our country,and there is getting younger and younger,particularly in young women,has become one of the common malignant tumors.Differentiated cancer which accounted for 80%-90% of all thyroid cancer and operation therapy is the basic treatment. Therefore,early diagnosis of thyroid cancer has important significance.Current clinical preoperative diagnosis of thyroid cancer mainly depends on B-,CT,MRI,radioisotope scanning,fine needle aspiration cytology,etc,but still can not meet the current diagnostic techniques in clinical non-invasive,early diagnosis of thyroid cancer-specific requirements.Therefore, the search for the ideal of the early diagnosis of thyroid cancer biomarkers,lower preoperative misdiagnosis rate,has become an important problem demanding prompt solution.Surface enhanced laser desorption/ionization time of flight mass spectrometry(SELDI-TOF-MS) is a new technology for proteomics research. Which is a principle of gene chips,to chromatography,mass spectrometry and protein chip technology combined research methods, with a rapid, sensitive,automated,high throughput characteristics,in the early detection of disease especially cancer is of great significance.ObjectiveTO explore the serum and saliva protein fingerprints of thyroid cancer,benign nodular goiter patients and healthy human by SELDI-TOF-MS and its supporting protein chip,screening the specific protein markers and construct diagnostic model of thyroid cancer,and provide a simple method for thyroid cancer diagnosed before surgery.At the same time discuss the differences in serum and saliva protein mass spectrometry of thyroid cancer with different syndromes and build diagnostic model,to provide an objective basis for thyroid cancer with the dialectical type of Chinese medicine,also to establish a non-invasive diagnosis of salivary proteome technology to explore new ways and methods.MethodsSerum protein fingerprints were detected in 40 cases of thyroid cancer and 30 healthy human by PBSⅡtype SELDI-TOF-MS technology platform and using CM10 chips,the raw data was corrected with Proteinchip Software3.2.1 and brought the total ionic strength and molecular weight to achieve uniform-mity,then analyzed the data obtained with Biomarker Wizard and Biomarker Pattern Software5.0.2,selected the group-specific differences in protein expressi-on,combined with bioinformatics method to establish differences between groups in serum protein diagnostic decision tree classification model;at the same time,establish differences between liver stagnation and Qi stagnation group and syndrome of intermingled phlegm and blood stasis group in serum protein diagnostic decision tree classification model.Saliva protein fingerprints were detected in 49 cases of thyroid cancer,34 cases of benign nodular goiter patients and 43 cases healthy human by NP20 chips,selected the group-specific differences in protein expression,combined with bioinformatics method to establish differences between groups in saliva protein diagnostic decision tree classification models.Results1.Comparison of serum protein differences between thyroid cancer and Healthy control groupEighteen Serum protein peaks had a significant difference in expression between the two groups(P<0.01),M/Z of 2050.69,2940.50,3942.92,5345.54,6918.33 peaks were selected and built the diagnostic decision tree classification model.The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of cross verification of this model were 81.4%,80.0%,83.3%,86.5% and 75.8%,respectively.The Youden'index and the area under the ROC curve were 0.633 and 0.997.2.Comparison of serum protein differences between liver stagnation and Qi stagnation group and syndrome of intermingled phlegm and blood stasis group Twenty two Serum protein peaks had a significant difference in expression between the two groups(P<0.01),M/Z of 4099.85,9418.17 were peaks selected and built the diagnostic decision tree classification model.The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of cross verification of this model were 82.5%,73.7%,90.5%,87.5% and 79.2%,respectively.The Youden'index and the area under the ROC curve were 0.642 and 0.982.3.Comparison of saliva protein differences between thyroid cancer and healthy control groupTwenty eight saliva protein peaks had a significant difference in expression between the two groups(P<0.01),M/Z of 3491.10,3642.28,4315.10,7424.63 peaks were selected and built the diagnostic decision tree classification model.The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of cross verification of this model were 81.8%,88.9%,74.4%,78.4%and 86.5%,respectively.The Youden'index and the area under the ROC curve were 0.633 and 0.953.4.Comparison of saliva protein differences between thyroid cancer and benign nodular goiter group.Forty nine saliva protein peaks had a significant difference in expression between the two groups(P<0.01),M/Z of 2587.38.3346.06.3584.20.8054.57 peaks were selected and built the diagnostic decision tree classification model.The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of cross verification of this model were89.1%,87.7%,91.1%,93.4% and 83.7%,respectively.The Youden'index and the area under the ROC curve were 0.788 and 0.954.5.Comparison of saliva protein differences between liver stagnation and Qi stagnation group and syndrome of intermingled phlegm and blood stasis group Twenty nine saliva protein peaks had a significant difference in expression between the two groups(P<0.01),M/Z of 3809.85.3110.04.4970.80 peaks were selected and built the diagnostic decision tree classification model.The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of cross verification of this model were85.7%,84.6%,86.9%,88.0%and 83.3%,respectively.The Youden'index and the area under the ROC curve were 0.715 and 0.967.Conclusion1.In this study, serum samples and saliva samples were measured by CM 10 and NP20 chips respectively with good stability and repeatability.2.BPS analysis selected M/Z of 2050.69.2940.50.3942.92.5345.54.6918.33 peaks to build differences in serum protein diagnostic decision tree classification model between thyroid cancer group and healthy control group, the total accuracy was 81.4%. Screening for thyroid cancer from healthy people with good value.3.BPS analysis selected M/Z of 4099.85,9418.17peaks to build differences in serum protein diagnostic decision tree classification model between liver stagnation and Qi stagnation group and syndrome of intermingled phlegm and blood stasis group, the total accuracy was 82.5%.So it provided an objective basis to thyroid cancer syndrome differentiation type of serum protein level.4.BPS analysis selected M/Z of 3491.10,3642.28,4315.10,7424.63 peaks to build differences in saliva protein diagnostic decision tree classification model between thyroid cancer group and healthy control group, the total accuracy was 81.8%. Screening for thyroid cancer from healthy people with good value and for the non-invasive diagnosis of thyroid cancer provides a new way.5.BPS analysis selected M/Z of 2587.38,3346.06,3584.20,8054.57 peaks to build differences in saliva protein diagnostic decision tree classification model between thyroid cancer and benign nodular goiter group, the total accuracy was 89.1%.To differentiate benign nodules and thyroid cancer, judge of preoperative value has a very good value.6.BPS analysis selected M/Z of 3809.85,3110.04,4970.80 peaks to build differences in saliva protein diagnostic decision tree classification model between liver stagnation and Qi stagnation group and syndrome of intermingled phlegm and blood stasis group, the total accuracy was 85.7%.So it provided an objective basis to thyroid cancer syndrome differentiation type of serum protein level.Initially established the dialectical mode of traditional Chinese medicine-saliva protein pattern of thyroid cancer,and also for actual research TCM provided new ways and means.