Detection Of Human Salivary Molecular Biomarkers For Noninvasive Diagnosis Of Thyroid Cancer

Thyroid cancer（TC）is the most common head and neck endocrine cancer in people aged 15-29,ranking the 9th in the world for malignant tumors and the 5th cancer affecting women’s health in the world.According to statistics,the early 10-year survival rate of TC is as high as about 90%,while lymph node metastasis and lung metastasis are prone to occur in the middle and late stages,resulting in a significant decrease in the 10-year survival rate.Therefore,early diagnosis is the most effective strategy to improve patient prognosis and reduce mortality.Traditional clinical examinations,including imaging examinations and histopathological examinations,are mostly invasive and are not suitable for routine large-scale examinations.As a class of molecular substances that can reflect the pathological information of the body and provide auxiliary basis for diagnosis,biomarkers are expected to improve the accuracy of clinical diagnosis and have received attention in recent years,and have an important significance to the diagnosis of TC.At the same time,a bodily fluid with great potential for clinical research,saliva,is emerging as a rich source of disease biomarkers.In view of this,it is hoped to establish a non-invasive method for the preoperative early diagnosis of TC based on salivary markers.For this reason,this paper mainly studies from the following two aspects.1.In this paper,a simple,rapid and high-throughput analytical method for the simultaneous determination of 10 amino acids in saliva by ultra-high performance liquid chromatography-high resolution mass spectrometry（UPLC-HRMS）was established.The method validation results showed that this method had good linearity（R²>0.99）,recovery（92.2-110.3%）,and precision（intra-day and inter-day precision RSD<7%and RSD<9%,respectively）.The amino acid metabolism levels in saliva from 61 papillary thyroid carcinoma（PTC）patients and61 healthy controls（HC）were analyzed by the nonparametric Mann-Whitney U test.showed that the concentrations of 10 amino acids were significantly different between PTC and HC（P<0.05）,suggesting that they could serve as potential PTC disease markers in saliva.Receiver characteristic operating curve（ROC）analysis was then used to determine the diagnostic value of potential markers.As a single marker,their area under the curve（AUC）ranged from 0.678to 0.833,while those established by binary logistic regression method.The AUC obtained by the combined diagnostic index was 0.936（sensitivity:91.2%;specificity:85.2%）.The results indicated that the salivary amino acid biomarkers have potential application value for the early diagnosis of PTC,and may provide a simple auxiliary way for its non-invasive diagnosis.2.In order to find more comprehensive and accurate molecular markers for the early diagnosis of PTC,we systematically studied and established a multi-platform method built on the combination of UPLC-HRMS and GC-HRMS based on non-targeted metabolomics to analyze volunteer’s saliva samples（PTC:51,HC:47）,so as to improve the accuracy and precision of metabolite identification and determine the complete PTC saliva metabolic network in clinical samples.PCA results showed that PTC and HC were clearly classified.Furthermore,OPLS-DA,nonparametric test,and volcano plot analysis were used to obtain significant markers which satisfying VIP>1.0,P<0.05,and|Log₂FC|>1.0 simultaneously,including amino acids,lipids,carbohydrates,etc.,indicated that the multi-platform combination achieved greater coverage of the PTC metabolome.Besides,the AUC of the above markers was 0.622-0.984 by ROC analysis,reflecting that all the markers identified have independent PTC predictive potential.In addition,a number of combined diagnostic index（Com.Index）were established by binary logistic regression.Among them,the AUC of Com.Index 3 based on GC-MS analysis reached 1.000,and the sensitivity and specificity were both 100.0%,and the AUC of Com.Index 4 under LC-MS was 0.996,and the sensitivity and specificity reached 95.7%and 97.8%,respectively.Data showed that compared with a single marker,Com.Index has higher accuracy in diagnosing and predicting PTC.The results of the study show that the multi-platform combined salivary metabolic fingerprint study of UPLC-HRMS and GC-HRMS based on non-targeted metabolomics can be used to distinguish and predict patients and controls,which provides a useful idea for non-invasive clinical screening of PTC.