Coagulation and thyroiditis are factors associated with adverse pathological features in differentiated thyroid cancer:A retrospective cohort study

Coagulation and thyroiditis are factors associated with adverse pathological features in differentiated thyroid cancer:A retrospective cohort study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Coagulation and thyroiditis are factors associated with adverse pathological features in differentiated thyroid cancer:A retrospective cohort study Xiao Chen, Han-yu Wang, Lu Yu, Jia-qi Liu, Hui Sun This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5272747/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objective Lymph node metastasis (LNM) and capsular invasion (CI) are the main pathological features leading to poor prognosis of differentiated thyroid cancer (DTC), and there is a lack of effective diagnostic methods before surgery. Therefore, this study was designed to analyze a large number of preoperative clinical features of DTC and identify factors closely related to those two pathological features. Methods 4557 patients with DTC, postoperative pathological results showed LNM in 2146 cases and CI in 2783 cases were retrospectively included. The preoperative blood, urine, serum laboratory test and ultrasound of thyroid were performed for data collection. A total of 74 clinical features were analyzed by the methods of principal component analysis (PCA), and key principal components were extracted for regression analysis of LNM and CI as well as subgroup analysis. Results 11 key clinical features were used for principal component analysis, and 6 principal components PC0-PC5 were finally obtained. PC0 is mainly composed of prothrombin time and international normalized ratio, and the score represents better coagulation function and has a protective effect on LNM. PC1 is mainly composed of thyroid peroxidase antibody and thyroid texture, and the score represents the severity of thyroiditis and has a protective effect on LNM and CI. Conclusion Thyroiditis and coagulation function were identified by principal component analysis as protective and risk factors for adverse pathology of DTC, meaning they were closely related to tumor metastasis and invasion. differentiated thyroid cancer coagulation thyroiditis metastasis invasion Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Thyroid neoplasms are common malignancies of the endocrine system, with differentiated thyroid cancer (DTC) accounting for more than 95% of all tumor incidences 1 , 2 . Distinct from other systemic neoplastic diseases, DTC usually have a relatively favorable prognosis and rarely develop distant metastases. At present, the pathological features of DTC with significant correlation with prognosis are mainly reflected in local lymph node metastasis of the tumor, as well as extrathyroidal invasion of the tumor tissue after breaking through the thyroid capsule 3 . When combined with these two types of pathologic features, such tumors tend to be considered to have worse pathologic stage. Because studies have shown that such tumors are more likely to recur after surgery, or have a risk of further distant metastasis, and affect patient survival 4 , 5 . Therefore, prompt surgical treatment of such tumors is required, and the extent of surgical resection is adjusted according to the status of local tumor metastasis or invasion. 6 Because of the indolent characteristics of thyroid tumors, there is still a lack of diagnostic modalities or biomarkers that can make direct predictions about the adverse pathological features of DTC 7 . In some clinical retrospective studies, it was found that metastasis and invasion of DTC may be associated with some baseline clinical characteristics of patients, and artificial diagnostic tools, such as predictive models, were developed based on these characteristics, which somewhat improved the preoperative predictive power for pathological features of DTC 8 . However, such studies included relatively single data types, and most of them predicted the actual pathological status of thyroid tumors or nodules by screening their imaging features under ultrasound. Ultrasound localization and characterization, on the other hand, are very dependent on the subjective experience of the operator and are not fully representative 9 , 10 . Therefore, prognostic factors other than ultrasound in thyroid tumors remain elusive. In particular, biochemical markers of other systems than the thyroid gland, may also have potential pathogenic or protective effects on the pathological and prognostic characteristics of tumors 11 . Therefore, we included postoperative cases of larger patients with thyroid tumors in this study and mainly collected clinical data other than local ultrasound morphology of thyroid tumors. Including a number of laboratory tests of blood and urine samples of patients, combined with general epidemiological characteristics, the main risk factors for poor prognosis of thyroid tumors were explored by means of data dimension reduction using principal component analysis (PCA). Finally, it was found that coagulation function and thyroid inflammatory markers were highly correlated with adverse pathological features of the tumor. It indicates some unique immune and microcirculatory mechanisms that may be possessed in the development of DTC, and provides a new direction for the development of predictors of adverse pathological features of DTC. Materials and methods Study Cohort and Examination Indicators The retrospective cohort selected for this study included all patients who were hospitalized in Wuhan Union Hospital due to DTC and underwent thyroidectomy and lymph node dissection from 2018 to 2021. All of these patients had definitive postoperative pathology reports, which described whether tumor capsule invasion (CI) and lymph node metastasis (LNM) occurred. After exclusion of some cases that may have confounded or significantly biased the results of the analysis (patients who had previously undergone thyroid-related surgery at an outside hospital, or were taking thyroid-related drugs, as well as patients with other major diseases). A total of 4557 patients were included, including 2146 patients with lymph node metastasis and 2783 patients with capsular or extrathyroidal invasion of thyroid tumors. The most recent clinical examination prior to surgery was collected for these patients. Including blood cell count, biochemical and endocrine related tests, urine tests and thyroid ultrasound examination, a total of 74 examination items. Subsequent statistical analyses were performed on these data. This study was approved by the institutional review board of the Wuhan Union Hospital, and the requirement for informed consent was waived. Statistical Analysis Principal component analysis We normalized the data as well as principal component transformation for 74 clinical features and 2 pathological features after performing multiple imputation and obtained 76 principal components. The principal component set (minor principal component set) at the end and the principal component set (major principal component set) at the front end were screened according to the elbow graph and principal component weight graph, respectively. The scores of these principal components were analyzed for clinical correlation with all raw data, and the variables that best represented the structure of the raw data were selected for analysis based on the significance of the correlation (p-value less than 0.05) between each clinical feature and each principal component in the minor principal component set and the major principal component set. For example, when a clinical feature has a significantly larger number of principal components associated with it in the principal component set than in the minor principal component set, the feature is retained and vice versa is removed. We used this approach to screen the original clinical features twice. Finally, the clinical features that best represent the original data features are retained for the final data dimension reduction, and the key principal components after dimension reduction are obtained. Clinicopathological correlation analysis Principal component score after principal component analysis was used as a new clinical feature, and multiple logistic regression analysis was performed with lymph node metastasis and thyroid capsular invasion as two pathological conditions, respectively, after matching the two confounding factors of age and gender to determine the correlation between these principal components and the prognosis of thyroid cancer. Similarly, the weights occupied by the original clinical features in each principal component were determined based on spearman correlation coefficient and pearson correlation coefficient using batch clinical correlation tests. These original clinical features were further examined as independent risk or protective factors for pathological conditions in the original data. Finally, subgroups were differentiated by sex and age, and subgroup analyses of clinical versus pathological characteristics were performed in the subgroup population. The statistical analysis process were completed by python 3.8, R 4.3.1, SPSS 26.0, and p-values less than 0.05 were considered statistically significant. Results Baseline Data The raw data for this study cohort contained a total of 74 clinical features as well as 2 pathological features. Pathologic features as dependent variables were lymph node metastasis and thyroid capsule invasion. The details included in the clinical features are shown in Table S1 . After principal component analysis and factor screening of these clinical characteristics, a total of 11 indicators were finally included in the analysis. Their distribution with age and sex in populations with different pathological characteristics is shown in Table 1 . Table 1 Baseline data of all clinical characteristics screened for principal component analysis LNM(n = 2146) No LNM(n = 2411) CI(n = 2783) No CI(n = 1774) Albumin/globulin ratio 1.70(1.50–1.90) 1.70(1.50–1.90) 1.70(1.50–1.90) 1.70(1.50–1.90) Lactate dehydrogenase(U/L) 183(159–209) 186(162–212) 184(160–210) 186(160–212) Prothrombin time(s) 12.60(12.30–13.10) 12.60(12.20–13.00) 12.60(12.20–13.00) 12.60(12.20–13.00) Activated partial thromboplastin time(s) 36.40(34.20–38.70) 36.00(34.00-38.40) 36.30(34.10–38.70) 36.10(34.10–38.30) Parathyroid hormone(pg/ml) 56.00(40.20–74.00) 58.70(43.20-75.58) 56.70(41.60-74.24) 58.40(42.60–75.50) International normalized ratio 0.96(0.92-1.00) 0.96(0.92–0.99) 0.96(0.92-1.00) 0.96(0.92-1.00) Platelets(10^9/L) 235(198–275) 232(193–273) 233(196–274) 234(195–273) Neutrophil ratio(Neu%) 59.40(53.10–65.20) 59.10(53.20–65.00) 59.30(53.39–65.40) 59.15(52.90-64.78) Thyroid peroxidase antibody(U/ml) 1.00(1.00-13.16) 1.00(1.00-20.55) 1.00(1.00-13.90) 1.02(1.00-22.06) Thyroid stimulating hormone(mIU/L) 1.79(1.25–2.61) 1.81(1.25–2.66) 1.82(1.28–2.67) 1.77(1.17–2.60) Age(years) 52(40–52) 52(45–52) 52(43–52) 52(42–52) Texture of thyroid(n%) Homogeneous 528(24.60%) 731(30.32%) 729(26.19%) 799(45.04%) Inhomogeneous 1025(47.76%) 1073(44.50%) 1300(46.71%) 529(29.82%) Missing data 593(27.64%) 607(25.18%) 754(27.00%) 446(25.14%) Gender(n%) Male 644(30.01%) 434(18.00%) 693(24.90%) 385(21.70%) Female 1502(69.99%) 1977(82.00%) 2090(75.10%) 1389(78.30%) Abbreviation: LNM Lymph Node Metastasis, CI Capsular Invasion Factor Screening and Dimensionality Reduction Because the clinical features covered by the original data were too redundant, we screened all 76 features twice by principal component analysis. After the first principal component transformation, 76 features were transformed into 76 principal components, and their weights on the original data are shown in Fig. 1 -A, 1 -B. According to this result, 16 minor principal components at the tail of the elbow graph and principal component weight graph, as well as 16 principal components at the head, were screened as minor principal component sets as well as major principal component sets, respectively. Each principal component in the set was correlated with each clinical feature of the original data, and the number of principal components that were significantly correlated with each clinical feature in the two principal component sets was compared. This result is shown in Figure S1 A-F. Finally, a larger number of clinical features significantly associated with each principal component in the primary principal component set were retained for secondary principal component screening. This process screened 28 clinical features and repeated the first screening process. After the second principal component transformation, the principal component weights corresponding to 28 clinical features are shown in Fig. 1 -C, 1 -D. According to this result, the first 6 principal components and the last 6 principal components were analyzed as the primary and secondary principal component sets for correlation matrix with clinical characteristics, and the results of this analysis are shown in Figure S2 A-F. 11 clinical features were finally selected as the most representative variables for the original data structure for subsequent analysis. Principal component analysis and clinicopathologic correlation analysis Dimensionality reduction of principal components was performed again after screening 11 key clinical features. The results of Fig. 1 -E, 1 -F showed that there were no more principal components at the ends of these 11 principal components that contributed ineffectively to the original data feature degree, indicating that the previous two feature screens yielded good results. Therefore, we chose the first six principal components PC0-PC5 that contributed 80% to the original data structure as the results after data dimension reduction. These 6 principal components were correlated with 11 clinical features, and the clinical features with significant correlation were selected to assess the clinical feature dimension mainly represented by each principal component according to the magnitude of the correlation coefficient. The results of Fig. 2 -A showed that PC0 was mainly associated with Prothrombin time (PT) and International normalized ratio (INR), representing coagulation function. Figure 2 -B suggests that PC1 is mainly associated with Thyroid peroxidase antibody (ATPO) and thyroid texture, representing thyroiditis. We further performed multiple logistic regression between the scores of these principal components and the two pathological features, and found that after including both confounding factors, age and gender, Fig. 2 -C indicated that PC0 and PC1 were associated with lymph node metastasis and showed a significant protective effect on the occurrence of lymph node metastasis. Figure 2 -D suggests that PC1 and PC3 are associated with thyroid capsule invasion, PC1 is a protective factor, and PC3 is a risk factor. Subgroup analysis To further validate the results of principal component analysis. We again performed multiple logistic regression based on pathological features for 11 clinical features and simultaneously matched age and gender factors. Figure 3 A-B suggests that thyroiditis as well as coagulation parameters such as PT, ATPO, and INR still have a significant effect on the development of pathological conditions, consistent with the results of principal component analysis. We then divided the original population into two subgroups according to sex. At the same time, we performed a restrictive cubic spline (RCS) analysis of the age of the patients, and the results of Figs. 4 -A and 4 -B suggest that as continuous variables, age around 52 years is more suitable for stratifying patients because the role of age on disease prognosis changes before and after this value. Patients were divided into older and younger groups by 52 years of age. Multiple logistic regression of PT, ATPO, and INR was repeated in different gender and age subgroups, and Fig. 4 -C and 4 -D showed that PT and ATPO had a more significant effect on LNM in female patients and younger patients compared with male patients and older people. However, the effect of ATPO on CI showed the opposite trend in male and female patients. Figure 4 -E finds that INR no longer significantly contributes to CI after differentiating age and gender subgroups. Discussion Because of the indolent nature of DTC, it may be difficult to identify unique risk factors associated with prognosis in previous single index or small sample clinical studies 12 . With the help of data dimension reduction in machine learning-unsupervised learning, we repeated the comparison of 74 clinical features in the original data according to data heterogeneity and the representativeness of the primary and secondary principal component scores after principal component analysis, and finally selected 11 clinical features that were relatively the most representative of the structural features of the original data for the final analysis. These clinical features mainly cover coagulation function, thyroid and systemic inflammatory indicators, thyroid-related endocrine function and so on. After the last principal component analysis, we found that principal components representing coagulation parameters as well as principal components representing thyroid immune function were strongly associated with two types of poor prognostic features: lymph node metastasis of thyroid tumors and thyroid capsular invasion. After multiple logistic regression matching for age and sex, two confounding factors considered to be directly associated with tumor prognosis in previous studies 13 , 14 , principal component scores remained significantly associated with the occurrence of pathological features. According to our results, there was a significant negative correlation between the score of PC0, the principal component representing coagulation function, and the specific coagulation parameters PT and INR. PC0 was also an independent protective factor for lymph node metastasis status in thyroid tumors. demonstrated that prolonged coagulation time is a risk factor for lymph node local metastasis in thyroid tumors. This conclusion was also demonstrated in our original data. Apart from individual studies that have used coagulation parameters as an indicator to predict thyroid tumors to establish prediction models, there has been little previous literature reporting a direct association between coagulation function and thyroid tumor prognosis 15 . However, the results of multiple logistics regression have excluded the possibility of collinearity and confounding factors. Therefore, we consider that it may be similar to the mechanism of other cancers, and the prolongation of coagulation time is more conducive to the establishment of excess blood supply by the local microenvironment of thyroid tumors, which in turn facilitates tumor growth as well as metastasis 16 , 17 . This may also serve as a novel indicator to assess whether DTC may have adverse pathology, however, deeper mechanisms may need to be further confirmed by prospective studies or basic studies. Abnormal thyroid immune function is mainly characterized by elevated thyroid autoantibody titer levels, that is, the occurrence of Hashimoto 's thyroiditis. In our study, two indicators representing thyroid immunity, namely the level of ATPO and whether thyroid ultrasound texture was uniform, were mainly included after principal component dimension reduction screening. Both measures represent the overall immune status of the thyroid gland and are not thought to be altered by tumor effects. In fact, there have been many similar studies on the relationship between the immune status of the thyroid gland and thyroid nodules as well as thyroid tumors. Most studies have shown that Hashimoto 's thyroiditis has a "bidirectional" effect on thyroid tumors, on the one hand, as a risk factor for thyroid tumorigenesis, and on the other hand, as a protective factor for further tumor metastasis or progression after thyroid tumors have formed 18 , 19 . The mechanism behind this is not clear and may be considered to be associated with the regulation of the local immune microenvironment of the thyroid 20 , 21 . In our study, both PC1 score, which represents the principal component of thyroid immunity, and thyroid immunity index in the original data suggest a protective effect against tumor metastasis and thyroid capsule invasion, which is consistent with most previous studies. This further illustrates the need to pay attention to the specific population of DTC with Hashimoto 's thyroiditis and develop individualized diagnosis and treatment strategies in the future. Our findings also provide some other interesting findings, such as the association of immune-related and coagulation-related principal components with pathological features seems to be more significant in female patient populations and younger patient populations younger than 52 years after further differentiation of subgroups according to age and sex in the original population. Of course, this may be associated with uneven sample sizes. However, consistent with previous studies of thyroid cancer risk prediction in this specific population, our results also suggest that more attention should be paid to screening for indicators related to specific populations to better serve a guiding role in clinical practice 22 , 23 . In addition, in addition to PC0 and PC1, there are some other principal components and the main clinical features they represent that somewhat suggest an association with thyroid tumors. For example, the neutrophil ratio represented by PC5 and the globulin ratio represented by PC3 also seem to predict a specific relationship between systemic immune status and adverse pathology of DTC. Platelet count and PTH levels, on the other hand, also suggest an association with thyroid tumor invasion and metastasis, respectively. These features have also been reported in some studies because the systemic immune and hematologic conditions are very complex, so they may similarly be reflected in the local microenvironment of thyroid, especially thyroid tumors, through specific pathways or mechanisms 24 , 25 .However, in our study, the contribution of these indicators to principal components and the association with pathological features was not as significant as that of coagulation and thyroid immunity, and more meticulous matching and correction may be needed to clarify this conclusion. Our study has limitations because as a retrospective study, bias during data collection cannot be avoided. However, since the purpose of hospitalization for patients with DTC is very clear, no other treatment except surgery will be performed, and there are few complications, we believe that the study results are credible. Conclusion After preoperative screening of patients for multidimensional clinical features, thyroiditis and coagulation abnormalities were identified by principal component analysis as independent protective and risk factors for adverse pathology of DTC, meaning they were closely related to tumor metastasis and invasion. It is necessary to validate the relevant indicators at the mechanistic level to help us provide a deeper understanding of the immune-related mechanisms of DTC and the role of the tumor microenvironment and develop accurate diagnosis and treatment strategies for DTC. Declarations Ethics approval and consent to participate This study was carried out in accordance with the Declaration of Helsinki. This study was approved by the institutional review board of the Wuhan Union Hospital, and the requirement for informed consent was waived. Consent for publication Not applicable . Availability of data and materials De-identified datasets analyzed in this study are available from the corresponding author upon reasonable request. Competing interests The authors declare no conflict of interest. Funding This work was funded by the Hubei Province Key Laboratory of Molecular Imagine, Grant(No.2021fzyx016), the Principal Investigator is Han-yu Wang, the Wuhan Knowledge Innovation Project,Grant(No.2023020201010162), the Principal Investigator is Hui Sun, and the Technology Innovation Project of Hubei Province, Grant(No.2023BCB131), the Principal Investigator is Hui Sun. Author Contributions All authors made substantial contributions to the conception and design of this study. XC performed the data analyses and wrote the manuscript; HYW performed the data collection and prepared the manuscript. HS contributed to the conception of the study and provided professional comments on the content. LY and JQL helped with data collection. Acknowledgements The authors are grateful to all the organizations and people who participated in the study, Huazhong University of Science and Technology. References Singh Ospina N, Iñiguez-Ariza NM, Castro MR. Thyroid nodules: diagnostic evaluation based on thyroid cancer risk assessment. BMJ (Clinical research ed). 2020;368:l6670.10.1136/bmj.l6670. Lim H, Devesa SS, Sosa JA, Check D, Kitahara CM. Trends in Thyroid Cancer Incidence and Mortality in the United States, 1974–2013. Jama. 2017;317(13):1338 – 48.10.1001/jama.2017.2719. Filetti S, Durante C, Hartl D, Leboulleux S, Locati LD, Newbold K et al. Thyroid cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up†. Annals of oncology: official journal of the European Society for Medical Oncology. 2019;30(12):1856 – 83.10.1093/annonc/mdz400. Kim HJ. Updated guidelines on the preoperative staging of thyroid cancer. Ultrasonography (Seoul, Korea). 2017;36(4):292 – 9.10.14366/usg.17023. Kim M, Jeon MJ, Oh HS, Park S, Song DE, Sung TY et al. Prognostic Implication of N1b Classification in the Eighth Edition of the Tumor-Node-Metastasis Staging System of Differentiated Thyroid Cancer. Thyroid: official journal of the American Thyroid Association. 2018;28(4):496-503.10.1089/thy.2017.0473. Montgomery KB, Fazendin JM, Chen H, Broman KK. Contemporary trends in extent of surgery for differentiated thyroid cancer with extrathyroidal extension. American journal of surgery. 2023;10.1016/j.amjsurg.2023.09.021.10.1016/j.amjsurg.2023.09.021. Singh Ospina N, Brito JP. Novel assessment of epigenetic imprinting biomarkers for the diagnosis of thyroid nodules. Nature reviews Clinical oncology. 2023;20(3):139 – 40.10.1038/s41571-022-00720-3. Zheng H, Lai V, Lu J, Kang JK, Chou J, Burman KD et al. Clinical Factors Predictive of Lymph Node Metastasis in Thyroid Cancer Patients: A Multivariate Analysis. Journal of the American College of Surgeons. 2022;234(4):691-700.10.1097/xcs.0000000000000107. Wang J, Dong C, Zhang YZ, Wang L, Yuan X, He M et al. A novel approach to quantify calcifications of thyroid nodules in US images based on deep learning: predicting the risk of cervical lymph node metastasis in papillary thyroid cancer patients. European radiology. 2023;10.1007/s00330-023-09909-1.10.1007/s00330-023-09909-1. Jiang L, Guo S, Zhao Y, Cheng Z, Zhong X, Zhou P. Predicting Extrathyroidal Extension in Papillary Thyroid Carcinoma Using a Clinical-Radiomics Nomogram Based on B-Mode and Contrast-Enhanced Ultrasound. Diagnostics (Basel, Switzerland). 2023;13(10).10.3390/diagnostics13101734. Guo K, Qian K, Shi Y, Sun T, Chen L, Mei D et al. Clinical and Molecular Characterizations of Papillary Thyroid Cancer in Children and Young Adults: A Multicenter Retrospective Study. Thyroid: official journal of the American Thyroid Association. 2021;31(11):1693 – 706.10.1089/thy.2021.0003. Nixon AM, Provatopoulou X, Kalogera E, Zografos GN, Gounaris A. Circulating thyroid cancer biomarkers: Current limitations and future prospects. Clinical endocrinology. 2017;87(2):117 – 26.10.1111/cen.13369. Shobab L, Burman KD, Wartofsky L. Sex Differences in Differentiated Thyroid Cancer. Thyroid: official journal of the American Thyroid Association. 2022;32(3):224 – 35.10.1089/thy.2021.0361. LeClair K, Bell KJL, Furuya-Kanamori L, Doi SA, Francis DO, Davies L. Evaluation of Gender Inequity in Thyroid Cancer Diagnosis: Differences by Sex in US Thyroid Cancer Incidence Compared With a Meta-analysis of Subclinical Thyroid Cancer Rates at Autopsy. JAMA internal medicine. 2021;181(10):1351 – 8.10.1001/jamainternmed.2021.4804. Gu J, Xie R, Zhao Y, Zhao Z, Xu D, Ding M et al. A machine learning-based approach to predicting the malignant and metastasis of thyroid cancer. Frontiers in oncology. 2022;12:938292.10.3389/fonc.2022.938292. Bauer AT, Gorzelanny C, Gebhardt C, Pantel K, Schneider SW. Interplay between coagulation and inflammation in cancer: Limitations and therapeutic opportunities. Cancer treatment reviews. 2022;102:102322.10.1016/j.ctrv.2021.102322. Falanga A, Marchetti M, Vignoli A. Coagulation and cancer: biological and clinical aspects. Journal of thrombosis and haemostasis: JTH. 2013;11(2):223 – 33.10.1111/jth.12075. Ferrari SM, Fallahi P, Elia G, Ragusa F, Ruffilli I, Paparo SR et al. Thyroid autoimmune disorders and cancer. Seminars in cancer biology. 2020;64:135 – 46.10.1016/j.semcancer.2019.05.019. Xu J, Ding K, Mu L, Huang J, Ye F, Peng Y et al. Hashimoto's Thyroiditis: A Double-Edged Sword in Thyroid Carcinoma. Frontiers in endocrinology. 2022;13:801925.10.3389/fendo.2022.801925. Dias Lopes NM, Mendonça Lens HH, Armani A, Marinello PC, Cecchini AL. Thyroid cancer and thyroid autoimmune disease: A review of molecular aspects and clinical outcomes. Pathology, research and practice. 2020;216(9):153098.10.1016/j.prp.2020.153098. Ehlers M, Schott M. Hashimoto's thyroiditis and papillary thyroid cancer: are they immunologically linked? Trends in endocrinology and metabolism: TEM. 2014;25(12):656 – 64.10.1016/j.tem.2014.09.001. Deng Y, Li H, Wang M, Li N, Tian T, Wu Y et al. Global Burden of Thyroid Cancer From 1990 to 2017. JAMA network open. 2020;3(6):e208759.10.1001/jamanetworkopen.2020.8759. Sawka AM, Ghai S, Rotstein L, Irish JC, Pasternak JD, Gullane PJ et al. Gender Differences in Fears Related to Low-Risk Papillary Thyroid Cancer and Its Treatment. JAMA otolaryngology– head & neck surgery. 2023;149(9):803 – 10.10.1001/jamaoto.2023.1642. Russo E, Guizzardi M, Canali L, Gaino F, Costantino A, Mazziotti G et al. Preoperative systemic inflammatory markers as prognostic factors in differentiated thyroid cancer: a systematic review and meta-analysis. Reviews in endocrine & metabolic disorders. 2023;10.1007/s11154-023-09845-x.10.1007/s11154-023-09845-x. Gambardella C, Mongardini FM, Paolicelli M, Bentivoglio D, Cozzolino G, Ruggiero R et al. Role of Inflammatory Biomarkers (NLR, LMR, PLR) in the Prognostication of Malignancy in Indeterminate Thyroid Nodules. International journal of molecular sciences. 2023;24(7).10.3390/ijms24076466. Additional Declarations No competing interests reported. Supplementary Files FigureS1.jpg Figure S2: Heatmap of the clinical correlation of 28 clinical characteristics with principal component scores after the second principal component analysis. (S2-A, B, C) Association of the first 6 major principal component collections (PC0-PC5) with each clinical feature. (S2-D, E, F) Association of the post-6 minor principal component collections (PC22-PC27) with each clinical feature. Numbers are p-values indicating whether there is a significant association between clinical characteristics and principal components. FigureS2.jpg Figure S1: Heatmap of the clinical correlation of 74 clinical characteristics with principal component scores after the first principal component analysis. (S1-A, B, C, D) Association of the first 16 major principal component collections (PC0-PC15) with each clinical feature. (S1-E, F, G, H) Association of the post-16 minor principal component collections (PC60-PC75) with each clinical feature. Numbers are p-values indicating whether there is a significant association between clinical characteristics and principal components. TableS1.xlsx Table S1: All clinical characteristics collected and used for principal component analysis and factor screening were included in this study. The number included in each category of clinical features or the abbreviated form of a specific measure is described in parenthesis. These abbreviated forms are also used in the results of the clinical relevance matrix. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5272747","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":378217387,"identity":"df95dccd-0ad2-453c-9d8c-81bc69df4dba","order_by":0,"name":"Xiao Chen","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiao","middleName":"","lastName":"Chen","suffix":""},{"id":378217388,"identity":"088e4d0f-a830-4751-9053-37b6cf655888","order_by":1,"name":"Han-yu Wang","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Han-yu","middleName":"","lastName":"Wang","suffix":""},{"id":378217389,"identity":"653d1ebb-d965-4587-a424-0197f0d8694e","order_by":2,"name":"Lu Yu","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Lu","middleName":"","lastName":"Yu","suffix":""},{"id":378217390,"identity":"dad6bfdd-c850-4fdd-823c-d24dc4b9ec59","order_by":3,"name":"Jia-qi Liu","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jia-qi","middleName":"","lastName":"Liu","suffix":""},{"id":378217394,"identity":"53a9cd34-b343-4eed-8cde-b7a60e9dd3c4","order_by":4,"name":"Hui Sun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIie3OIQ+CQBTA8TvdtEDHOeUrHGOjwIc5ClcOMsHwgsNoJfghaGpzY8Ny9os4g4VgNDlxzAoX3bzf9rYX3n97CGnaz0qDJem2sWoiIrdNMKgneF2GhXJi53F1MyFie4vda5T6IUwvp96EyIS55iGIjzl3AAkWgpHQ/sTi3twUUVxIjgFnZQiWQQYe+yRZyYhkV8AvhQTJLqFEUgcwKCRENN5sJyKnEI2T04q5mcEHHttwz2rSwCZnVj8eK3+xnYqBx1oj47vRdiaD9y38VLnSNE37X29F00cByxK2MgAAAABJRU5ErkJggg==","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Hui","middleName":"","lastName":"Sun","suffix":""}],"badges":[],"createdAt":"2024-10-16 05:23:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5272747/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5272747/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":70381232,"identity":"5d81c560-e5b5-45b6-80a5-711300db9bf2","added_by":"auto","created_at":"2024-12-02 16:06:26","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3227084,"visible":true,"origin":"","legend":"\u003cp\u003eElbow diagram and principal component weight diagram for dimension reduction of data by principal component analysis. (1-A, B) 76 principal components were entered and principal components from PC60 onwards were classified as minor principal components. (1-C, D) Twenty-eight principal components were entered and principal components from PC22 onwards were classified as minor principal components. (1-E, F) 11 principal components are input, and finally the first 6 principal components (PC0-PC5) whose cumulative explanation reaches 0.8 are selected as the final dimension reduction results\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/b1992ebd054107c54a917579.jpg"},{"id":70381233,"identity":"3a126b3c-64f5-43d3-8555-28befc703aed","added_by":"auto","created_at":"2024-12-02 16:06:26","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":553225,"visible":true,"origin":"","legend":"\u003cp\u003eAbbreviation：LNM Lymph nodes metastases, CI Capsular invasion. Weight plot of correlation coefficients between two principal components PC1, PC0 and 11 clinical features (2-A, B). r value is pearson 's correlation coefficient for continuous variables and spearman' s correlation coefficient for categorical variables. Absolute values of r value greater than 0.7 were considered strongly correlated with principal components. Forest plot of multiple logistics regression between six principal components (PC0-PC5) combined with age, gender and two pathological features LNM (2-C), CI (2-D). p \u0026lt; 0.05 was considered an independent risk factor.\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/dddd0ffbb2fed2bc2b8c5bbf.jpg"},{"id":70381231,"identity":"11845f60-c2a2-4a51-9b4f-8de101da205d","added_by":"auto","created_at":"2024-12-02 16:06:25","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":515457,"visible":true,"origin":"","legend":"\u003cp\u003eAbbreviation：LNM Lymph nodes metastases, CI Capsular invasion. Forest plot of multiple logistics regression of 11 clinical characteristics versus age and gender on two pathological characteristics LNM (3-A) and CI (3-B). p \u0026lt; 0.05 was considered an independent risk factor.\u003c/p\u003e","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/cad6f87e96cd67adc4657351.jpg"},{"id":70381044,"identity":"041981e3-46c3-49fb-bdc4-0b25962269aa","added_by":"auto","created_at":"2024-12-02 15:58:25","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1122499,"visible":true,"origin":"","legend":"\u003cp\u003eAbbreviation：LNM Lymph nodes metastases, CI Capsular invasion. Restrictive cubic spline (RCS) plots of age versus LNM (4-A), CI (4-B), respectively. Multiple logistics regression of ATPO (4-C), PT (4-D), and INR (4-E) on the two pathological conditions in different gender and age subgroups, respectively. Covariates included age, gender, PT (dependent variable LNM), INR (dependent variable CI) for ATPO, age, gender, ATPO for PT (dependent variable LNM), and age, gender, ATPO for INR (dependent variable CI).\u003c/p\u003e","description":"","filename":"Figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/1b1f4520fb09748f2a89bbb5.jpg"},{"id":72386888,"identity":"1eb689d4-f14d-45d6-ac6c-436ba7f4ad5c","added_by":"auto","created_at":"2024-12-26 10:17:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5851569,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/ce35774d-3ab0-4cb7-903a-2549b9dcb4f1.pdf"},{"id":70381235,"identity":"36eb2fbf-fbd0-4217-b099-58a3f26a43ca","added_by":"auto","created_at":"2024-12-02 16:06:29","extension":"jpg","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1798185,"visible":true,"origin":"","legend":"\u003cp\u003eFigure S2: Heatmap of the clinical correlation of 28 clinical characteristics with principal component scores after the second principal component analysis. (S2-A, B, C) Association of the first 6 major principal component collections (PC0-PC5) with each clinical feature. (S2-D, E, F) Association of the post-6 minor principal component collections (PC22-PC27) with each clinical feature. Numbers are p-values indicating whether there is a significant association between clinical characteristics and principal components.\u003c/p\u003e","description":"","filename":"FigureS1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/ff15d989b54e15907afda4ad.jpg"},{"id":70381038,"identity":"949ec8b3-3231-4d73-8b1a-ed20adf7928f","added_by":"auto","created_at":"2024-12-02 15:58:25","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":533263,"visible":true,"origin":"","legend":"\u003cp\u003eFigure S1: Heatmap of the clinical correlation of 74 clinical characteristics with principal component scores after the first principal component analysis. (S1-A, B, C, D) Association of the first 16 major principal component collections (PC0-PC15) with each clinical feature. (S1-E, F, G, H) Association of the post-16 minor principal component collections (PC60-PC75) with each clinical feature. Numbers are p-values indicating whether there is a significant association between clinical characteristics and principal components.\u003c/p\u003e","description":"","filename":"FigureS2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/647a841ccc81d4c51fc48d75.jpg"},{"id":70381041,"identity":"c39d0da1-6dae-41c1-a013-60638b06e058","added_by":"auto","created_at":"2024-12-02 15:58:25","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":11965,"visible":true,"origin":"","legend":"\u003cp\u003eTable S1: All clinical characteristics collected and used for principal component analysis and factor screening were included in this study. The number included in each category of clinical features or the abbreviated form of a specific measure is described in parenthesis. These abbreviated forms are also used in the results of the clinical relevance matrix.\u003c/p\u003e","description":"","filename":"TableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5272747/v1/55a92683b6c64eb17eb8fd86.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Coagulation and thyroiditis are factors associated with adverse pathological features in differentiated thyroid cancer:A retrospective cohort study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThyroid neoplasms are common malignancies of the endocrine system, with differentiated thyroid cancer (DTC) accounting for more than 95% of all tumor incidences\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Distinct from other systemic neoplastic diseases, DTC usually have a relatively favorable prognosis and rarely develop distant metastases. At present, the pathological features of DTC with significant correlation with prognosis are mainly reflected in local lymph node metastasis of the tumor, as well as extrathyroidal invasion of the tumor tissue after breaking through the thyroid capsule \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. When combined with these two types of pathologic features, such tumors tend to be considered to have worse pathologic stage. Because studies have shown that such tumors are more likely to recur after surgery, or have a risk of further distant metastasis, and affect patient survival \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Therefore, prompt surgical treatment of such tumors is required, and the extent of surgical resection is adjusted according to the status of local tumor metastasis or invasion.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eBecause of the indolent characteristics of thyroid tumors, there is still a lack of diagnostic modalities or biomarkers that can make direct predictions about the adverse pathological features of DTC\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. In some clinical retrospective studies, it was found that metastasis and invasion of DTC may be associated with some baseline clinical characteristics of patients, and artificial diagnostic tools, such as predictive models, were developed based on these characteristics, which somewhat improved the preoperative predictive power for pathological features of DTC\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. However, such studies included relatively single data types, and most of them predicted the actual pathological status of thyroid tumors or nodules by screening their imaging features under ultrasound. Ultrasound localization and characterization, on the other hand, are very dependent on the subjective experience of the operator and are not fully representative \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Therefore, prognostic factors other than ultrasound in thyroid tumors remain elusive. In particular, biochemical markers of other systems than the thyroid gland, may also have potential pathogenic or protective effects on the pathological and prognostic characteristics of tumors\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTherefore, we included postoperative cases of larger patients with thyroid tumors in this study and mainly collected clinical data other than local ultrasound morphology of thyroid tumors. Including a number of laboratory tests of blood and urine samples of patients, combined with general epidemiological characteristics, the main risk factors for poor prognosis of thyroid tumors were explored by means of data dimension reduction using principal component analysis (PCA). Finally, it was found that coagulation function and thyroid inflammatory markers were highly correlated with adverse pathological features of the tumor. It indicates some unique immune and microcirculatory mechanisms that may be possessed in the development of DTC, and provides a new direction for the development of predictors of adverse pathological features of DTC.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Cohort and Examination Indicators\u003c/h2\u003e \u003cp\u003eThe retrospective cohort selected for this study included all patients who were hospitalized in Wuhan Union Hospital due to DTC and underwent thyroidectomy and lymph node dissection from 2018 to 2021. All of these patients had definitive postoperative pathology reports, which described whether tumor capsule invasion (CI) and lymph node metastasis (LNM) occurred. After exclusion of some cases that may have confounded or significantly biased the results of the analysis (patients who had previously undergone thyroid-related surgery at an outside hospital, or were taking thyroid-related drugs, as well as patients with other major diseases). A total of 4557 patients were included, including 2146 patients with lymph node metastasis and 2783 patients with capsular or extrathyroidal invasion of thyroid tumors. The most recent clinical examination prior to surgery was collected for these patients. Including blood cell count, biochemical and endocrine related tests, urine tests and thyroid ultrasound examination, a total of 74 examination items. Subsequent statistical analyses were performed on these data. This study was approved by the institutional review board of the Wuhan Union Hospital, and the requirement for informed consent was waived.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003e \u003cstrong\u003ePrincipal component analysis\u003c/strong\u003e \u003cp\u003e We normalized the data as well as principal component transformation for 74 clinical features and 2 pathological features after performing multiple imputation and obtained 76 principal components. The principal component set (minor principal component set) at the end and the principal component set (major principal component set) at the front end were screened according to the elbow graph and principal component weight graph, respectively. The scores of these principal components were analyzed for clinical correlation with all raw data, and the variables that best represented the structure of the raw data were selected for analysis based on the significance of the correlation (p-value less than 0.05) between each clinical feature and each principal component in the minor principal component set and the major principal component set. For example, when a clinical feature has a significantly larger number of principal components associated with it in the principal component set than in the minor principal component set, the feature is retained and vice versa is removed. We used this approach to screen the original clinical features twice. Finally, the clinical features that best represent the original data features are retained for the final data dimension reduction, and the key principal components after dimension reduction are obtained.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eClinicopathological correlation analysis\u003c/strong\u003e \u003cp\u003ePrincipal component score after principal component analysis was used as a new clinical feature, and multiple logistic regression analysis was performed with lymph node metastasis and thyroid capsular invasion as two pathological conditions, respectively, after matching the two confounding factors of age and gender to determine the correlation between these principal components and the prognosis of thyroid cancer. Similarly, the weights occupied by the original clinical features in each principal component were determined based on spearman correlation coefficient and pearson correlation coefficient using batch clinical correlation tests. These original clinical features were further examined as independent risk or protective factors for pathological conditions in the original data. Finally, subgroups were differentiated by sex and age, and subgroup analyses of clinical versus pathological characteristics were performed in the subgroup population.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eThe statistical analysis process were completed by python 3.8, R 4.3.1, SPSS 26.0, and p-values less than 0.05 were considered statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eBaseline Data\u003c/h2\u003e \u003cp\u003eThe raw data for this study cohort contained a total of 74 clinical features as well as 2 pathological features. Pathologic features as dependent variables were lymph node metastasis and thyroid capsule invasion. The details included in the clinical features are shown in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. After principal component analysis and factor screening of these clinical characteristics, a total of 11 indicators were finally included in the analysis. Their distribution with age and sex in populations with different pathological characteristics is shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline data of all clinical characteristics screened for principal component analysis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLNM(n\u0026thinsp;=\u0026thinsp;2146)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNo LNM(n\u0026thinsp;=\u0026thinsp;2411)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCI(n\u0026thinsp;=\u0026thinsp;2783)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eNo CI(n\u0026thinsp;=\u0026thinsp;1774)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlbumin/globulin ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.70(1.50\u0026ndash;1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.70(1.50\u0026ndash;1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.70(1.50\u0026ndash;1.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.70(1.50\u0026ndash;1.90)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLactate dehydrogenase(U/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e183(159\u0026ndash;209)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e186(162\u0026ndash;212)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e184(160\u0026ndash;210)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e186(160\u0026ndash;212)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProthrombin time(s)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.60(12.30\u0026ndash;13.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12.60(12.20\u0026ndash;13.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12.60(12.20\u0026ndash;13.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e12.60(12.20\u0026ndash;13.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eActivated partial thromboplastin time(s)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36.40(34.20\u0026ndash;38.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e36.00(34.00-38.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36.30(34.10\u0026ndash;38.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e36.10(34.10\u0026ndash;38.30)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParathyroid hormone(pg/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e56.00(40.20\u0026ndash;74.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58.70(43.20-75.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e56.70(41.60-74.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e58.40(42.60\u0026ndash;75.50)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInternational normalized ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.96(0.92-1.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.96(0.92\u0026ndash;0.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.96(0.92-1.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.96(0.92-1.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlatelets(10^9/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e235(198\u0026ndash;275)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e232(193\u0026ndash;273)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e233(196\u0026ndash;274)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e234(195\u0026ndash;273)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeutrophil ratio(Neu%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e59.40(53.10\u0026ndash;65.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e59.10(53.20\u0026ndash;65.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e59.30(53.39\u0026ndash;65.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e59.15(52.90-64.78)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThyroid peroxidase antibody(U/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.00(1.00-13.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.00(1.00-20.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.00(1.00-13.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.02(1.00-22.06)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThyroid stimulating hormone(mIU/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.79(1.25\u0026ndash;2.61)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.81(1.25\u0026ndash;2.66)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.82(1.28\u0026ndash;2.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.77(1.17\u0026ndash;2.60)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge(years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e52(40\u0026ndash;52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e52(45\u0026ndash;52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e52(43\u0026ndash;52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e52(42\u0026ndash;52)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTexture of thyroid(n%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHomogeneous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e528(24.60%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e731(30.32%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e729(26.19%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e799(45.04%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInhomogeneous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1025(47.76%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1073(44.50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1300(46.71%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e529(29.82%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMissing data\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e593(27.64%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e607(25.18%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e754(27.00%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e446(25.14%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender(n%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e644(30.01%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e434(18.00%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e693(24.90%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e385(21.70%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1502(69.99%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1977(82.00%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2090(75.10%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1389(78.30%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003eAbbreviation: LNM Lymph Node Metastasis, CI Capsular Invasion\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eFactor Screening and Dimensionality Reduction\u003c/h3\u003e\n\u003cp\u003eBecause the clinical features covered by the original data were too redundant, we screened all 76 features twice by principal component analysis. After the first principal component transformation, 76 features were transformed into 76 principal components, and their weights on the original data are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-A, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-B. According to this result, 16 minor principal components at the tail of the elbow graph and principal component weight graph, as well as 16 principal components at the head, were screened as minor principal component sets as well as major principal component sets, respectively. Each principal component in the set was correlated with each clinical feature of the original data, and the number of principal components that were significantly correlated with each clinical feature in the two principal component sets was compared. This result is shown in Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eA-F. Finally, a larger number of clinical features significantly associated with each principal component in the primary principal component set were retained for secondary principal component screening. This process screened 28 clinical features and repeated the first screening process. After the second principal component transformation, the principal component weights corresponding to 28 clinical features are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-C, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-D. According to this result, the first 6 principal components and the last 6 principal components were analyzed as the primary and secondary principal component sets for correlation matrix with clinical characteristics, and the results of this analysis are shown in Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003eA-F. 11 clinical features were finally selected as the most representative variables for the original data structure for subsequent analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003ePrincipal component analysis and clinicopathologic correlation analysis\u003c/h2\u003e \u003cp\u003eDimensionality reduction of principal components was performed again after screening 11 key clinical features. The results of Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-E, \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-F showed that there were no more principal components at the ends of these 11 principal components that contributed ineffectively to the original data feature degree, indicating that the previous two feature screens yielded good results. Therefore, we chose the first six principal components PC0-PC5 that contributed 80% to the original data structure as the results after data dimension reduction. These 6 principal components were correlated with 11 clinical features, and the clinical features with significant correlation were selected to assess the clinical feature dimension mainly represented by each principal component according to the magnitude of the correlation coefficient. The results of Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e-A showed that PC0 was mainly associated with Prothrombin time (PT) and International normalized ratio (INR), representing coagulation function. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e-B suggests that PC1 is mainly associated with Thyroid peroxidase antibody (ATPO) and thyroid texture, representing thyroiditis. We further performed multiple logistic regression between the scores of these principal components and the two pathological features, and found that after including both confounding factors, age and gender, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e-C indicated that PC0 and PC1 were associated with lymph node metastasis and showed a significant protective effect on the occurrence of lymph node metastasis. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e-D suggests that PC1 and PC3 are associated with thyroid capsule invasion, PC1 is a protective factor, and PC3 is a risk factor.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSubgroup analysis\u003c/h3\u003e\n\u003cp\u003eTo further validate the results of principal component analysis. We again performed multiple logistic regression based on pathological features for 11 clinical features and simultaneously matched age and gender factors. Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-B suggests that thyroiditis as well as coagulation parameters such as PT, ATPO, and INR still have a significant effect on the development of pathological conditions, consistent with the results of principal component analysis. We then divided the original population into two subgroups according to sex. At the same time, we performed a restrictive cubic spline (RCS) analysis of the age of the patients, and the results of Figs.\u0026nbsp;\u0026lt;link rid=\"fig6\"\u0026gt;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u0026lt;/link\u0026gt;\u003c/span\u003e-A and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u003c/span\u003e-B suggest that as continuous variables, age around 52 years is more suitable for stratifying patients because the role of age on disease prognosis changes before and after this value. Patients were divided into older and younger groups by 52 years of age. Multiple logistic regression of PT, ATPO, and INR was repeated in different gender and age subgroups, and Fig.\u0026nbsp;\u0026lt;link rid=\"fig6\"\u0026gt;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u0026lt;/link\u0026gt;\u003c/span\u003e-C and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u003c/span\u003e-D showed that PT and ATPO had a more significant effect on LNM in female patients and younger patients compared with male patients and older people. However, the effect of ATPO on CI showed the opposite trend in male and female patients. Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u003c/span\u003e-E finds that INR no longer significantly contributes to CI after differentiating age and gender subgroups.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eBecause of the indolent nature of DTC, it may be difficult to identify unique risk factors associated with prognosis in previous single index or small sample clinical studies\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. With the help of data dimension reduction in machine learning-unsupervised learning, we repeated the comparison of 74 clinical features in the original data according to data heterogeneity and the representativeness of the primary and secondary principal component scores after principal component analysis, and finally selected 11 clinical features that were relatively the most representative of the structural features of the original data for the final analysis. These clinical features mainly cover coagulation function, thyroid and systemic inflammatory indicators, thyroid-related endocrine function and so on. After the last principal component analysis, we found that principal components representing coagulation parameters as well as principal components representing thyroid immune function were strongly associated with two types of poor prognostic features: lymph node metastasis of thyroid tumors and thyroid capsular invasion. After multiple logistic regression matching for age and sex, two confounding factors considered to be directly associated with tumor prognosis in previous studies\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, principal component scores remained significantly associated with the occurrence of pathological features.\u003c/p\u003e \u003cp\u003eAccording to our results, there was a significant negative correlation between the score of PC0, the principal component representing coagulation function, and the specific coagulation parameters PT and INR. PC0 was also an independent protective factor for lymph node metastasis status in thyroid tumors. demonstrated that prolonged coagulation time is a risk factor for lymph node local metastasis in thyroid tumors. This conclusion was also demonstrated in our original data. Apart from individual studies that have used coagulation parameters as an indicator to predict thyroid tumors to establish prediction models, there has been little previous literature reporting a direct association between coagulation function and thyroid tumor prognosis\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. However, the results of multiple logistics regression have excluded the possibility of collinearity and confounding factors. Therefore, we consider that it may be similar to the mechanism of other cancers, and the prolongation of coagulation time is more conducive to the establishment of excess blood supply by the local microenvironment of thyroid tumors, which in turn facilitates tumor growth as well as metastasis \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. This may also serve as a novel indicator to assess whether DTC may have adverse pathology, however, deeper mechanisms may need to be further confirmed by prospective studies or basic studies.\u003c/p\u003e \u003cp\u003eAbnormal thyroid immune function is mainly characterized by elevated thyroid autoantibody titer levels, that is, the occurrence of Hashimoto 's thyroiditis. In our study, two indicators representing thyroid immunity, namely the level of ATPO and whether thyroid ultrasound texture was uniform, were mainly included after principal component dimension reduction screening. Both measures represent the overall immune status of the thyroid gland and are not thought to be altered by tumor effects. In fact, there have been many similar studies on the relationship between the immune status of the thyroid gland and thyroid nodules as well as thyroid tumors. Most studies have shown that Hashimoto 's thyroiditis has a \"bidirectional\" effect on thyroid tumors, on the one hand, as a risk factor for thyroid tumorigenesis, and on the other hand, as a protective factor for further tumor metastasis or progression after thyroid tumors have formed \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. The mechanism behind this is not clear and may be considered to be associated with the regulation of the local immune microenvironment of the thyroid \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. In our study, both PC1 score, which represents the principal component of thyroid immunity, and thyroid immunity index in the original data suggest a protective effect against tumor metastasis and thyroid capsule invasion, which is consistent with most previous studies. This further illustrates the need to pay attention to the specific population of DTC with Hashimoto 's thyroiditis and develop individualized diagnosis and treatment strategies in the future.\u003c/p\u003e \u003cp\u003eOur findings also provide some other interesting findings, such as the association of immune-related and coagulation-related principal components with pathological features seems to be more significant in female patient populations and younger patient populations younger than 52 years after further differentiation of subgroups according to age and sex in the original population. Of course, this may be associated with uneven sample sizes. However, consistent with previous studies of thyroid cancer risk prediction in this specific population, our results also suggest that more attention should be paid to screening for indicators related to specific populations to better serve a guiding role in clinical practice \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. In addition, in addition to PC0 and PC1, there are some other principal components and the main clinical features they represent that somewhat suggest an association with thyroid tumors. For example, the neutrophil ratio represented by PC5 and the globulin ratio represented by PC3 also seem to predict a specific relationship between systemic immune status and adverse pathology of DTC. Platelet count and PTH levels, on the other hand, also suggest an association with thyroid tumor invasion and metastasis, respectively. These features have also been reported in some studies because the systemic immune and hematologic conditions are very complex, so they may similarly be reflected in the local microenvironment of thyroid, especially thyroid tumors, through specific pathways or mechanisms \u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.However, in our study, the contribution of these indicators to principal components and the association with pathological features was not as significant as that of coagulation and thyroid immunity, and more meticulous matching and correction may be needed to clarify this conclusion.\u003c/p\u003e \u003cp\u003eOur study has limitations because as a retrospective study, bias during data collection cannot be avoided. However, since the purpose of hospitalization for patients with DTC is very clear, no other treatment except surgery will be performed, and there are few complications, we believe that the study results are credible.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eAfter preoperative screening of patients for multidimensional clinical features, thyroiditis and coagulation abnormalities were identified by principal component analysis as independent protective and risk factors for adverse pathology of DTC, meaning they were closely related to tumor metastasis and invasion. It is necessary to validate the relevant indicators at the mechanistic level to help us provide a deeper understanding of the immune-related mechanisms of DTC and the role of the tumor microenvironment and develop accurate diagnosis and treatment strategies for DTC.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate\u003c/p\u003e\n\u003cp\u003eThis study was carried out in accordance with the Declaration of Helsinki. This study was approved by the institutional review board of the Wuhan Union Hospital, and the requirement for informed consent was waived.\u003c/p\u003e\n\u003cp\u003eConsent for publication\u003c/p\u003e\n\u003cp\u003eNot applicable\u003cstrong\u003e.\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDe-identified datasets analyzed in this study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e\n\u003cp\u003eFunding\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis work was funded by the Hubei Province Key Laboratory of Molecular Imagine, Grant(No.2021fzyx016), the Principal Investigator is Han-yu Wang, the Wuhan Knowledge Innovation Project,Grant(No.2023020201010162), the Principal Investigator is Hui Sun, and the Technology Innovation Project of Hubei Province, Grant(No.2023BCB131), the Principal Investigator is Hui Sun.\u003c/p\u003e\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\u003cp\u003eAll authors made substantial contributions to the conception and design of this study. XC performed the data analyses and wrote the manuscript; HYW performed the data collection and prepared the manuscript. HS contributed to the conception of the study and provided professional comments on the content. LY and JQL helped with data collection.\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe authors are grateful to all the organizations and people who participated in the study, Huazhong University of Science and Technology.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSingh Ospina N, I\u0026ntilde;iguez-Ariza NM, Castro MR. Thyroid nodules: diagnostic evaluation based on thyroid cancer risk assessment. BMJ (Clinical research ed). 2020;368:l6670.10.1136/bmj.l6670.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim H, Devesa SS, Sosa JA, Check D, Kitahara CM. Trends in Thyroid Cancer Incidence and Mortality in the United States, 1974\u0026ndash;2013. Jama. 2017;317(13):1338\u0026thinsp;\u0026ndash;\u0026thinsp;48.10.1001/jama.2017.2719.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFiletti S, Durante C, Hartl D, Leboulleux S, Locati LD, Newbold K et al. Thyroid cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up\u0026dagger;. Annals of oncology: official journal of the European Society for Medical Oncology. 2019;30(12):1856\u0026thinsp;\u0026ndash;\u0026thinsp;83.10.1093/annonc/mdz400.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim HJ. Updated guidelines on the preoperative staging of thyroid cancer. Ultrasonography (Seoul, Korea). 2017;36(4):292\u0026thinsp;\u0026ndash;\u0026thinsp;9.10.14366/usg.17023.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim M, Jeon MJ, Oh HS, Park S, Song DE, Sung TY et al. Prognostic Implication of N1b Classification in the Eighth Edition of the Tumor-Node-Metastasis Staging System of Differentiated Thyroid Cancer. Thyroid: official journal of the American Thyroid Association. 2018;28(4):496-503.10.1089/thy.2017.0473.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontgomery KB, Fazendin JM, Chen H, Broman KK. Contemporary trends in extent of surgery for differentiated thyroid cancer with extrathyroidal extension. American journal of surgery. 2023;10.1016/j.amjsurg.2023.09.021.10.1016/j.amjsurg.2023.09.021.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh Ospina N, Brito JP. Novel assessment of epigenetic imprinting biomarkers for the diagnosis of thyroid nodules. Nature reviews Clinical oncology. 2023;20(3):139\u0026thinsp;\u0026ndash;\u0026thinsp;40.10.1038/s41571-022-00720-3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng H, Lai V, Lu J, Kang JK, Chou J, Burman KD et al. Clinical Factors Predictive of Lymph Node Metastasis in Thyroid Cancer Patients: A Multivariate Analysis. Journal of the American College of Surgeons. 2022;234(4):691-700.10.1097/xcs.0000000000000107.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J, Dong C, Zhang YZ, Wang L, Yuan X, He M et al. A novel approach to quantify calcifications of thyroid nodules in US images based on deep learning: predicting the risk of cervical lymph node metastasis in papillary thyroid cancer patients. European radiology. 2023;10.1007/s00330-023-09909-1.10.1007/s00330-023-09909-1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang L, Guo S, Zhao Y, Cheng Z, Zhong X, Zhou P. Predicting Extrathyroidal Extension in Papillary Thyroid Carcinoma Using a Clinical-Radiomics Nomogram Based on B-Mode and Contrast-Enhanced Ultrasound. Diagnostics (Basel, Switzerland). 2023;13(10).10.3390/diagnostics13101734.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo K, Qian K, Shi Y, Sun T, Chen L, Mei D et al. Clinical and Molecular Characterizations of Papillary Thyroid Cancer in Children and Young Adults: A Multicenter Retrospective Study. Thyroid: official journal of the American Thyroid Association. 2021;31(11):1693\u0026thinsp;\u0026ndash;\u0026thinsp;706.10.1089/thy.2021.0003.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNixon AM, Provatopoulou X, Kalogera E, Zografos GN, Gounaris A. Circulating thyroid cancer biomarkers: Current limitations and future prospects. Clinical endocrinology. 2017;87(2):117\u0026thinsp;\u0026ndash;\u0026thinsp;26.10.1111/cen.13369.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShobab L, Burman KD, Wartofsky L. Sex Differences in Differentiated Thyroid Cancer. Thyroid: official journal of the American Thyroid Association. 2022;32(3):224\u0026thinsp;\u0026ndash;\u0026thinsp;35.10.1089/thy.2021.0361.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeClair K, Bell KJL, Furuya-Kanamori L, Doi SA, Francis DO, Davies L. Evaluation of Gender Inequity in Thyroid Cancer Diagnosis: Differences by Sex in US Thyroid Cancer Incidence Compared With a Meta-analysis of Subclinical Thyroid Cancer Rates at Autopsy. JAMA internal medicine. 2021;181(10):1351\u0026thinsp;\u0026ndash;\u0026thinsp;8.10.1001/jamainternmed.2021.4804.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu J, Xie R, Zhao Y, Zhao Z, Xu D, Ding M et al. A machine learning-based approach to predicting the malignant and metastasis of thyroid cancer. Frontiers in oncology. 2022;12:938292.10.3389/fonc.2022.938292.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBauer AT, Gorzelanny C, Gebhardt C, Pantel K, Schneider SW. Interplay between coagulation and inflammation in cancer: Limitations and therapeutic opportunities. Cancer treatment reviews. 2022;102:102322.10.1016/j.ctrv.2021.102322.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFalanga A, Marchetti M, Vignoli A. Coagulation and cancer: biological and clinical aspects. Journal of thrombosis and haemostasis: JTH. 2013;11(2):223\u0026thinsp;\u0026ndash;\u0026thinsp;33.10.1111/jth.12075.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFerrari SM, Fallahi P, Elia G, Ragusa F, Ruffilli I, Paparo SR et al. Thyroid autoimmune disorders and cancer. Seminars in cancer biology. 2020;64:135\u0026thinsp;\u0026ndash;\u0026thinsp;46.10.1016/j.semcancer.2019.05.019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Ding K, Mu L, Huang J, Ye F, Peng Y et al. Hashimoto's Thyroiditis: A Double-Edged Sword in Thyroid Carcinoma. Frontiers in endocrinology. 2022;13:801925.10.3389/fendo.2022.801925.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDias Lopes NM, Mendon\u0026ccedil;a Lens HH, Armani A, Marinello PC, Cecchini AL. Thyroid cancer and thyroid autoimmune disease: A review of molecular aspects and clinical outcomes. Pathology, research and practice. 2020;216(9):153098.10.1016/j.prp.2020.153098.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEhlers M, Schott M. Hashimoto's thyroiditis and papillary thyroid cancer: are they immunologically linked? Trends in endocrinology and metabolism: TEM. 2014;25(12):656\u0026thinsp;\u0026ndash;\u0026thinsp;64.10.1016/j.tem.2014.09.001.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng Y, Li H, Wang M, Li N, Tian T, Wu Y et al. Global Burden of Thyroid Cancer From 1990 to 2017. JAMA network open. 2020;3(6):e208759.10.1001/jamanetworkopen.2020.8759.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSawka AM, Ghai S, Rotstein L, Irish JC, Pasternak JD, Gullane PJ et al. Gender Differences in Fears Related to Low-Risk Papillary Thyroid Cancer and Its Treatment. JAMA otolaryngology\u0026ndash; head \u0026amp; neck surgery. 2023;149(9):803\u0026thinsp;\u0026ndash;\u0026thinsp;10.10.1001/jamaoto.2023.1642.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRusso E, Guizzardi M, Canali L, Gaino F, Costantino A, Mazziotti G et al. Preoperative systemic inflammatory markers as prognostic factors in differentiated thyroid cancer: a systematic review and meta-analysis. Reviews in endocrine \u0026amp; metabolic disorders. 2023;10.1007/s11154-023-09845-x.10.1007/s11154-023-09845-x.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGambardella C, Mongardini FM, Paolicelli M, Bentivoglio D, Cozzolino G, Ruggiero R et al. Role of Inflammatory Biomarkers (NLR, LMR, PLR) in the Prognostication of Malignancy in Indeterminate Thyroid Nodules. International journal of molecular sciences. 2023;24(7).10.3390/ijms24076466.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"differentiated thyroid cancer, coagulation, thyroiditis, metastasis, invasion","lastPublishedDoi":"10.21203/rs.3.rs-5272747/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5272747/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eLymph node metastasis (LNM) and capsular invasion (CI) are the main pathological features leading to poor prognosis of differentiated thyroid cancer (DTC), and there is a lack of effective diagnostic methods before surgery. Therefore, this study was designed to analyze a large number of preoperative clinical features of DTC and identify factors closely related to those two pathological features.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003e4557 patients with DTC, postoperative pathological results showed LNM in 2146 cases and CI in 2783 cases were retrospectively included. The preoperative blood, urine, serum laboratory test and ultrasound of thyroid were performed for data collection. A total of 74 clinical features were analyzed by the methods of principal component analysis (PCA), and key principal components were extracted for regression analysis of LNM and CI as well as subgroup analysis.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003e11 key clinical features were used for principal component analysis, and 6 principal components PC0-PC5 were finally obtained. PC0 is mainly composed of prothrombin time and international normalized ratio, and the score represents better coagulation function and has a protective effect on LNM. PC1 is mainly composed of thyroid peroxidase antibody and thyroid texture, and the score represents the severity of thyroiditis and has a protective effect on LNM and CI.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThyroiditis and coagulation function were identified by principal component analysis as protective and risk factors for adverse pathology of DTC, meaning they were closely related to tumor metastasis and invasion.\u003c/p\u003e","manuscriptTitle":"Coagulation and thyroiditis are factors associated with adverse pathological features in differentiated thyroid cancer:A retrospective cohort study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-02 15:58:20","doi":"10.21203/rs.3.rs-5272747/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"58477dcd-a743-47bb-a182-73e8b3661628","owner":[],"postedDate":"December 2nd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-01-20T05:23:25+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-02 15:58:20","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5272747","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5272747","identity":"rs-5272747","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}