Development and internal validation of a nomogram for predicting the initial response to radioactive iodine therapy in differentiated thyroid cancer with concurrent Hashimoto’s thyroiditis

Introduction

Background

Thyroid cancer (TC) is the most prevalent malignant tumor of the endocrine system, ranking seventh globally in terms of incidence as of 2022, with a continuing upward trend (1). Differentiated thyroid cancer (DTC) represents the most common subtype of TC, primarily comprising papillary thyroid carcinoma (PTC) and follicular thyroid carcinoma (FTC), which together account for over 90% of all TC cases (2). Radioactive iodine (RAI) therapy serves as a critical postoperative treatment for DTC, playing a vital role in the elimination of residual thyroid tissue and in reducing the risks of recurrence and metastasis (3). Hashimoto’s thyroiditis (HT) is an autoimmune disease characterized by chronic lymphocytic infiltration (4), with its prevalence showing an upward trend in recent years (5). Multiple studies have confirmed an association between HT and DTC (6,7), with data indicating that HT and DTC coexist in approximately 25% of cases (5). This coexistence highlights the necessity for vigilance in diagnosing and managing HT alongside DTC. Previous research has demonstrated that HT can influence the clinical and pathological characteristics of DTC patients (8,9) and may also affect the prognosis of RAI therapy to some extent (10).

Rationale and knowledge gap

Current studies analyzing factors influencing the efficacy of RAI therapy following DTC surgery (11,12) or constructing predictive models (13) predominantly focus on the overall DTC population, leaving a gap regarding research specifically addressing DTC patients with HT. Notably, serum antithyroglobulin antibodies (TgAb) are frequently elevated in DTC patients with HT, and elevated TgAb levels can interfere with thyroglobulin (Tg) measurement, leading to falsely low or negative Tg results. This situation creates significant uncertainty in Tg-based efficacy assessments and prognostic evaluations, posing substantial challenges for clinical decision-making.

Objective

In response to these challenges, this study specifically focused on DTC patients with concomitant HT, developing a predictive model for short-term response to RAI therapy based on clinicopathological features. This model incorporated TgAb as a prognostic marker within a multivariable framework. By predicting the short-term response (excellent response vs. non-excellent response), the clinical value of this model lies in its potential to forecast early treatment responses and to inform individualized follow-up strategies. We present this article in accordance with the TRIPOD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0252/rc).

Methods

Study population

This research conducts a retrospective analysis of clinical records for 461 individuals diagnosed with DTC (Figure 1), all of whom initially underwent RAI therapy at our hospital between August 2020 and April 2025. The inclusion criteria were as follows: (I) a pathological diagnosis of DTC with first-time RAI therapy; (II) postoperative pathology indicating concomitant HT or chronic lymphocytic thyroiditis; (III) complete clinical and follow-up data before and after the initial RAI therapy. The exclusion criteria comprised: (I) incomplete data or lack of regular follow-up; (II) presence of other malignant tumors; (III) pathological types of undifferentiated TC or medullary carcinoma. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The First Affiliated Hospital, Jiangxi Medical College, Nanchang University (approval No. IIT [2026] Clinical Ethics Review No. 065) and was simultaneously registered on the Chinese Clinical Trial Registry (ChiCTR2600118586). Given the retrospective nature of the study, the requirement for informed consent was waived.

Figure 1 Research flowchart. LASSO, least absolute shrinkage and selection operator; RAI, radioactive iodine.

Treatment, follow-up and evaluation

Pre-treatment preparation

All patients discontinued levothyroxine sodium tablets for a duration of 3 to 4 weeks prior to treatment to achieve thyroid-stimulating hormone (TSH) levels exceeding 30 mU/L, while adhering to a strict low-iodine diet for 1 to 2 weeks.

Treatment protocol

Upon admission, each case was meticulously assessed by a medical team comprising two senior physicians. Evaluation of recurrence risk stratification was performed based on pathology, serology, and imaging results. Although current clinical guidelines do not specify exact dosages, the administered dose of iodine-131 was tailored to each patient’s individualized risk profile, referencing the recommended dosage ranges. A post-therapy RAI whole-body scan (RxWBS) was performed 72 hours after treatment.

Follow-up schedule

Patients completed the first three follow-ups at 6–8 weeks, 3 months, and 6 months post-iodine therapy. Each visit included measurements of serum TSH, Tg, and TgAb levels, as well as scheduled neck ultrasound, chest computed tomography (CT), and diagnostic whole-body iodine-131 scan (DxWBS). After the initial iodine therapy, patients were instructed to discontinue levothyroxine sodium tablets for 3 to 4 weeks between 6 and 12 months post-treatment. A comprehensive evaluation was conducted during this period, with the results serving as the final outcome for this study.

Serum Tg and TgAb concentrations were measured using electrochemiluminescence immunoassay (ECLIA) on the Roche cobas e801 analyzer (Roche Diagnostics, Switzerland), with measurement ranges of 0.04–500.00 ng/mL for Tg and 10–4,000 IU/mL for TgAb (reference ranges: 3.5–77.00 ng/mL and 0–115 IU/mL, respectively). TgAb levels above 115 IU/mL were defined as positive. Serum TSH levels were assessed using chemiluminescent immunoassay on the Bayer ADVIA Centaur system (Bayer, Germany), with a measurement range of 0.04–100.00 mIU/L (reference range, 0.27–4.20 mIU/L). Blood samples were uniformly collected from all patients after thyroid hormone withdrawal prior to initial RAI therapy for the measurement of serum Tg, TgAb, and TSH. Consistent analytical platforms and standardized testing protocols were utilized throughout the study period.

As per the 2015 American Thyroid Association (ATA) guidelines, efficacy assessments were conducted 6 to 12 months following the initial RAI therapy. Clinical outcomes were explicitly categorized into four distinct groups:

• Excellent response (ER): no clinical or radiographic evidence of disease, with suppressed Tg <0.2 ng/mL or stimulated thyroglobulin (sTg) <1 ng/mL in the absence of antibody interference.
• Biochemical incomplete response (BIR): abnormal Tg or rising TgAb levels in the absence of localizable structural disease (suppressed Tg > 1 ng/mL, sTg > 10 ng/mL, or rising TgAb trends).
• Structural incomplete response (SIR): structural or functional evidence of disease identified on imaging, regardless of Tg or TgAb levels.
• Indeterminate response (IDR): non-specific biochemical or structural findings that cannot be confidently classified as benign or malignant (e.g., detectable but low Tg, stable or declining TgAb).

For binary modeling, patients with BIR, IDR, and SIR were combined into a non-ER (NER) group, resulting in an outcome defined as ER versus NER.

Data collection

Clinical, pathological, laboratory, and imaging-related data were systematically collected from patient cohorts. Clinical information encompassed age, gender, iodine dose, and the interval between surgery and RAI therapy. Pathological features included histological type, primary tumor location, maximum tumor diameter, number of lesions, lymph node metastasis rate, extrathyroidal extension (ETE), and the presence of nodular goiter (NG). Laboratory indicators comprised sTg, TgAb, TSH, platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), neutrophil-to-lymphocyte ratio (NLR), systemic immune-inflammation index (SII), and systemic inflammatory response index (SIRI). Patients were further stratified according to the 2015 ATA risk stratification system for recurrence. Tumor staging was determined based on the 8th edition of the American Joint Committee on Cancer (AJCC) TNM staging system, including T, N, and M classifications.

Statistical analysis

Statistical analyses were conducted using R software (version 4.5.1). Continuous variables were expressed as mean ± standard deviation when normally distributed and were compared using the independent samples t-test. Non-normally distributed continuous variables were expressed as median (interquartile range, IQR) and compared using the Mann-Whitney U test. Categorical variables were summarized as frequencies and percentages, and comparisons were made using the chi-square test or Fisher’s exact test, as appropriate. A logarithmic transformation was applied to sTg, TgAb, NLR, PLR, LMR, SII, and SIRI to mitigate right-skewed distributions. All transformed variables were calculated using a log₁₀(x+1) transformation to manage zero or near-zero values prior to analysis. The dataset was randomly split into training and validation cohorts at a 7:3 ratio using stratified random sampling. Feature selection was performed utilizing least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation (package: glmnet). Selected variables were subsequently incorporated into multivariable logistic regression models (package: stats) to construct the predictive model. A nomogram was developed for individualized risk estimation (package: rms). Model discrimination was evaluated through receiver operating characteristic (ROC) analysis and area under the curve (AUC) calculation (package: pROC). Calibration performance was assessed using bootstrap resampling with 1,000 iterations (package: rms). Decision curve analysis (DCA) was conducted to evaluate clinical net benefit (package: dcurves). A two-sided P value <0.05 was considered statistically significant.

Results

Baseline characteristics

This study involved a total of 461 patients diagnosed with DTC concurrently with HT who underwent RAI therapy. Based on treatment responses (refer to Table 1), patients were categorized into the ER group (n=235) and the NER group (n=226). The overall mean age of the cohort was 39.8 years, with female patients constituting 81% of the sample.

Comparisons of baseline characteristics revealed statistically significant differences between the ER and NER groups regarding N stage, M stage, ETE, ATA risk stratification, presence of NG, TgAb levels, sTg, number of lesions, metastatic lymph node rate, and the administered dose of iodine-131 (P<0.05) (Table 1).

The administered dose of iodine-131 ranged from 50 to 250 mCi, with a median of 150 mCi (IQR, 150–160 mCi), primarily clustered at distinct levels: 150 mCi (n=242), 160 mCi (n=64), 180 mCi (n=59), 100 mCi (n=43), 120 mCi (n=21), and 200 mCi (n=19). For clarity and to eliminate categorical overlap, these doses were reclassified into three mutually exclusive groups: ≤100, >100 to ≤150, and >150 mCi.

Baseline comparison between training set and validation set

The training and validation sets showed no statistically significant variations in the distribution of most clinical traits [P>0.05 and standardized mean difference (SMD) <0.20], indicating that the baseline characteristics of both groups are largely comparable (Table 2).

Variable selection

To address multicollinearity and mitigate model overfitting, LASSO regression was applied for initial variable selection. By incorporating a penalty parameter (λ), regression coefficients were progressively reduced, controlling the model’s complexity. As λ increased, the penalty on coefficients became more stringent, leading to the reduction of less informative variables to zero. LASSO regression was conducted within the training set to avoid data leakage, with the optimal penalty parameter selected through cross-validation within the glmnet framework. The optimal value of λ.1se was determined to be 0.03466. Based on this criterion, candidate variables were retained for further analysis and subsequently entered into a multivariable logistic regression model to construct the final predictive model. Ultimately, five predictors were identified: number of lesions, N stage, lymph node metastasis rate, TgAb, and sTg (Figure 2A,2B).

Figure 2 LASSO regression. (A) Cross-validation plot. (B) Coefficient path plot. LASSO, least absolute shrinkage and selection operator; LN, lymph node; N, node; sTg, stimulated thyroglobulin; TgAb, thyroglobulin antibody.

Construct a logistic model

The variables selected via LASSO regression were utilized in a multivariable logistic regression model, with treatment efficacy serving as the dependent variable. The results of this analysis are detailed in Table 3.

Validation of the nomogram and ROC curve

Based on the aforementioned predictive variables, a nomogram was developed to visualize the individualized probability of developing a NER in patients with DTC accompanied by HT undergoing initial RAI therapy (Figure 3). To convert the nomogram into a predicted probability for specific patients, a straightforward scoring system was applied: first, the patient’s specific value for each risk factor (multifocality, lymph node metastasis rate, sTg, TgAb, and N stage) was identified on its respective axis, and a vertical line was drawn upward to the top “Points” scale (ranging from 0 to 100) to determine the points assigned to that variable. Following this, the points from all five variables were summed to obtain the “Total Points” (ranging from 0 to 160). Finally, a vertical line was drawn straight down from the “Total Points” axis to the bottom “Risk of NER” scale to derive the patient’s individualized predicted probability. Consequently, a higher total score corresponds directly to an increased probability of experiencing an NER.

Figure 3 Nomogram. LN, lymph node; N, node; NER, non-excellent response; sTg, stimulated thyroglobulin; TgAb, thyroglobulin antibody.

Subsequently, the discriminative performance of the prediction model was evaluated using ROC curves (Figure 4). The 95% confidence intervals (CIs) for the AUCs were calculated using DeLong’s non-parametric method. In the training cohort, the AUC for predicting NER was 0.864 (95% CI: 0.825–0.904). Utilizing an optimal cutoff value of 0.562, determined by maximizing Youden’s index, the model achieved a sensitivity of 0.734 and a specificity of 0.878. In the validation cohort, the AUC was 0.882 (95% CI: 0.824–0.939), with a sensitivity of 0.765 and a specificity of 0.873 at an optimal cutoff of 0.568. Overall, the model demonstrated robust discriminative performance in both cohorts.

Figure 4 ROC curves. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

Calibration curve and DCA curve

Calibration curves (Figure 5A,5B) illustrated acceptable agreement between the nomogram-predicted probabilities and the observed frequencies of NER in both the training and internal validation cohorts. Visually, the calibration plots were generally positioned close to the ideal 45-degree reference line. Quantitatively, the model exhibited satisfactory calibration performance, with a Brier score of 0.148 in the training cohort (intercept =0.005, slope =0.941) and 0.137 in the internal validation cohort (intercept =−0.159, slope =1.028). Additionally, DCA was conducted to assess the clinical utility of the nomogram (Figure 5C,5D). The threshold probability range was evaluated across the full theoretical spectrum (0–1.0), while the interpretation centered on the clinically relevant range (approximately 0% to 80%). Within this range, the model consistently demonstrated a higher net benefit compared to the default “treat-all” and “treat-none” strategies. A similar trend was observed in the internal validation cohort, indicating the model’s potential utility for individualized risk stratification and decision support concerning initial RAI therapy. However, these findings require confirmation in external validation cohorts prior to clinical application.

Figure 5 Model performance. (A,B) Calibration curve. (C,D) DCA curve. DCA, decision curve analysis.

Discussion

This study presents a retrospective analysis of 461 patients diagnosed with DTC and HT who underwent initial RAI therapy. Comprehensive evaluations of clinical, pathological, laboratory, and imaging data were conducted. Utilizing stratified random sampling, the data were divided into training and validation sets. LASSO and multivariable binary logistic regression analyses were employed to identify five independent predictors: the number of lesions, lymph node metastasis rate, lymph node metastasis stage, TgAb, and sTg. Based on these predictors, a binary predictive model was constructed, demonstrating acceptable discrimination and calibration in the internal validation cohort. This study provides a preliminary contribution to the development of predictive models for RAI treatment efficacy in patients with DTC and HT, offering a visual tool for identifying populations at risk of NER.

sTg was identified as one of the most statistically significant predictors within the model, aligning with its established clinical value in postoperative surveillance and assessment of RAI efficacy for DTC. In patients exhibiting negative or low-titer TgAb, sTg has been validated as a critical indicator for monitoring biochemical recurrence and evaluating therapeutic response; additionally, a propensity score-matched study illustrated that pre-ablation sTg levels were independently associated with RAI-refractory disease (14). However, in TgAb-positive patients, sTg measurements may face assay interference, limiting the interpretability of sTg as a standalone biomarker. The model’s added value stems from the integration of sTg with TgAb and other clinicopathological variables within a multivariable framework, thus addressing the limitations of any single indicator in this specific patient population.

TgAb emerged as another significant predictor in the model. Clinically, DTC patients with HT typically exhibit a higher positive rate and baseline level of TgAb compared to those with isolated DTC. Previous studies have confirmed TgAb as an independent risk factor for lymph node metastasis in DTC patients (15), and higher baseline levels of TgAb prior to RAI therapy often correlate with poorer efficacy and prognosis (16). Furthermore, the dynamic changes in TgAb levels provide substantial value in the comprehensive management of the disease, serving as an important indicator of tumor status and prognosis, whether in preoperative and postoperative stages (17,18) or before and after RAI therapy (19,20). Notably, some studies suggest that TgAb may surpass Tg in monitoring recurrence following curative resection of DTC (21). Although the sample size of the recurrence group in this cohort was limited and necessitates further validation, this finding offers new insights into serological monitoring.

The number of tumor foci serves as a critical indicator of disease burden and invasion extent. A cohort study conducted by Woo Ri Choi, encompassing 2,390 samples, demonstrated that tumor multifocality constitutes an independent risk factor for postoperative recurrence of PTC (22). Furthermore, propensity score matching analysis revealed that an increase in the number of tumor foci significantly heightened the risk of lymph node metastasis in the lateral neck region among PTC patients (23). This study reinforces the notion that tumor multifocality is a predictor of NER to RAI therapy. This finding indicates that closer surveillance may be warranted for DTC patients with multifocal lesions, considering their potential for suboptimal treatment responses. The metastasis rate in cervical lymph nodes among DTC patients is notably significant, ranging from 20% to 50%. When micro-metastases are included, this percentage can escalate to as much as 90%. The lymph node metastasis rate is defined as the ratio of the number of positive lymph nodes in the surgical clearance specimen to the total number of cleared lymph nodes. Prior studies have established a correlation between the lymph node metastasis rate and patient prognosis (24), consistent with our findings, which indicate that this rate serves as a predictor of NER to RAI therapy.

Inflammatory and immune-related markers have undergone extensive investigation in TC, though their predictive value remains contentious (25-29). In the current cohort, none of the conventional peripheral inflammatory indicators (NLR, PLR, LMR, SII, and SIRI) emerged as predictors of RAI therapy response. This negative finding suggests that systemic inflammatory indices may possess limited additive value in DTC patients with concurrent HT, potentially due to the inadequacy of peripheral blood measurements in reflecting the localized immune microenvironment of the thyroid. Other factors, such as tumor heterogeneity and variations in sampling timing, may also account for the discrepancies observed in the literature.

Previous studies have typically regarded TC patients as a homogeneous group; however, patients with concurrent HT exhibit a distinct immune microenvironment (30) and unique biological behavioral characteristics (31). Guided by principles of precision medicine, the present study specifically targeted this subgroup to develop a predictive model for treatment efficacy. The clinical value of this model lies in its ability to preliminarily assess early treatment responses (differentiating ER from NER) in DTC patients with concurrent HT, thus providing a reference for the initial stratification of follow-up intensity. Among the five identified predictors, the number of lesions and the lymph node metastasis rate primarily capture tumor burden and invasiveness, and their predictive value is well recognized. The lymph node metastasis stage (N1a/N1b) further refines the anatomical extent of regional metastasis, complementing the lymph node metastasis rate by assessing lymph node involvement from multiple perspectives. Serological markers, such as sTg and TgAb, offer enhanced dynamic monitorability. Based on the model’s findings, closer follow-up intervals and increased imaging surveillance may be warranted for patients with elevated sTg levels. Notably, the TgAb levels used in this study were measured at baseline prior to radioiodine therapy. While baseline TgAb levels were confirmed as an independent predictor of ER/NER, the postoperative dynamic trend regarding prognosis has not been validated in this study. Although previously reported in the literature (19,32), whether this association can inform follow-up strategies specifically for this patient population requires further investigation.

Recent research has significantly advanced the development of predictive models for TC. In addition to the previously mentioned models predicting lymph node metastasis and responses to RAI treatment, several innovative clinical application studies have emerged. For instance, a study utilizing the Surveillance, Epidemiology, and End Results (SEER) database developed a nomogram to estimate the risk of distant metastasis in TC (33). Other studies have created nomograms to forecast specific survival rates in patients with poorly differentiated thyroid carcinoma (34) and the survival rates of patients with DTC accompanied by distant metastasis (35). Furthermore, specific predictive models have been developed for assessing RAI therapy efficacy in patients with RAI-resistant TC (36). These models provide highly individualized prognostic information for TC patients. However, prognostic prediction tools tailored specifically for patients with DTC accompanied by HT remain relatively scarce. To deliver precise individualized predictions for this specific demographic, data from our center were integrated to create a web-based calculator utilizing the Shiny framework, following the construction concepts of the aforementioned predictive models. Users can input a patient’s predictive indicators into this online calculator to obtain the probability of NER (https://ev827625936en.shinyapps.io/DTC-HT-RAIT-shiny/).

Conclusions

In summary, the predictive model for RAI therapy efficacy in DTC patients with concurrent HT constructed within this study represents a preliminary exploration of precision medicine in this area. The model may provide valuable references for individualized treatment planning, although its clinical utility necessitates further validation. The nomogram demonstrated acceptable predictive performance in the internal validation cohort, providing clinicians with a visual tool for early assessment of RAI therapy response. Nevertheless, several limitations should be acknowledged. As a single-center retrospective study with a relatively limited sample size, the model underwent only internal validation without external verification. The stringent exclusion criteria (employing complete-case analysis) may have introduced selection bias, collectively limiting the generalizability of our findings. Additionally, the number of cases with positive surgical margins or vascular invasion was insufficient for inclusion in the analysis, constraining the development and evaluation of the predictive model and failing to adequately capture the prognostic impact of these high-risk pathological features. The study also did not encompass key molecular characteristics (such as BRAF and TERT promoter mutations), which may not only restrict the predictive performance of the model but also omit exploration of underlying molecular mechanisms. Therefore, subsequent multicenter, large-sample, prospective studies that incorporate molecular markers are essential for further validation and optimization of the results of this study.

Acknowledgments

The authors are grateful to all the organizations and people who participated in the study.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0252/rc

Data Sharing Statement: Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0252/dss

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0252/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0252/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The First Affiliated Hospital, Jiangxi Medical College, Nanchang University (approval No. IIT [2026] Clinical Ethics Review No. 065). Given the retrospective nature of the study, the requirement for informed consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63.
2. Oh JM, Ahn BC. Molecular mechanisms of radioactive iodine refractoriness in differentiated thyroid cancer: Impaired sodium iodide symporter (NIS) expression owing to altered signaling pathway activity and intracellular localization of NIS. Theranostics 2021;11:6251-77. [Crossref] [PubMed]
3. Ringel MD, Sosa JA, Baloch Z, et al. 2025 American Thyroid Association Management Guidelines for Adult Patients with Differentiated Thyroid Cancer. Thyroid 2025;35:841-985. [Crossref] [PubMed]
4. Barros Silva AC, Damas II, Moma CA, et al. Interplay between expression of PD-L1 on thyrocytes and intrathyroidal lymphocytes and FOXP3 as a marker of regulatory T lymphocytes in Hashimoto thyroiditis. Am J Physiol Endocrinol Metab 2026;330:E38-46. [Crossref] [PubMed]
5. Wang L, Zhu X, Xu S, et al. Hashimoto's thyroiditis: from pathogenesis to clinical management. Front Endocrinol (Lausanne) 2026;17:1729316. [Crossref] [PubMed]
6. Hanege FM, Tuysuz O, Celik S, et al. Hashimoto's thyroiditis in papillary thyroid carcinoma: a 22-year study. Acta Otorhinolaryngol Ital 2021;41:142-5. [Crossref] [PubMed]
7. Paparodis RD, Karvounis E, Bantouna D, et al. Incidentally Discovered Papillary Thyroid Microcarcinomas Are More Frequently Found in Patients with Chronic Lymphocytic Thyroiditis Than with Multinodular Goiter or Graves' Disease. Thyroid 2020;30:531-5. [Crossref] [PubMed]
8. Wang L, Chen J, Yuan X, et al. Lymph node metastasis of papillary thyroid carcinoma in the context of Hashimoto's thyroiditis. BMC Endocr Disord 2022;22:12. [Crossref] [PubMed]
9. Parvathareddy SK, Siraj AK, Siraj N, et al. Evaluating the Influence of Hashimoto's Thyroiditis on Clinico-Pathological Characteristics and Prognostic Outcomes of Middle Eastern Differentiated Thyroid Carcinoma. Int J Endocrinol 2024;2024:9929782. [Crossref] [PubMed]
10. Tang Q, Pan W, Peng L. Association between Hashimoto thyroiditis and clinical outcomes of papillary thyroid carcinoma: A meta-analysis. PLoS One 2022;17:e0269995. [Crossref] [PubMed]
11. Wang C, Qin G, Wang S, et al. Clinical outcome and influencing factors of differentiated thyroid cancer with biochemical incomplete response. Future Oncol 2025;21:2911-9. [Crossref] [PubMed]
12. Li J, Geng H, Nan J, et al. Prognostic Factors Affecting Iodine-131 Treatment Efficacy in Patients With Differentiated Thyroid Carcinoma: A Risk-Stratified Analysis. Nutr Cancer 2026;78:181-8. [Crossref] [PubMed]
13. Bülbül O, Nak D. Predicting excellent response to radioiodine in differentiated thyroid cancer using machine learning. Acta Otorhinolaryngol Ital 2024;44:261-8. [Crossref] [PubMed]
14. Cheng X, Xu S, Zhu Y, et al. Markedly elevated serum preoperative thyroglobulin predicts radioiodine-refractory thyroid cancer. Eur J Clin Invest 2022;52:e13721. [Crossref] [PubMed]
15. Lai Y, Gu Y, Yu M, et al. Thyroglobulin Antibody (TgAb) Positive is an Independent Risk Factor for Lymph Node Metastasis in Patients with Differentiated Thyroid Carcinoma. Int J Gen Med 2023;16:5979-88. [Crossref] [PubMed]
16. Han N, Lu C, Li J, et al. Stimulated thyroglobulin and pre-ablation antithyroglobulin antibody products can predict the response to radioiodine therapy of TgAb-positive differentiated thyroid cancer patients: a retrospective study. Front Endocrinol (Lausanne) 2023;14:1222470. [Crossref] [PubMed]
17. Jeong HJ, Lee JS, Lee JS, et al. Anti-thyroglobulin antibody levels post-thyroidectomy and papillary thyroid carcinoma recurrence. BMC Cancer 2025;25:1371. [Crossref] [PubMed]
18. Xu S, Huang H, Qian J, et al. Prognostic value of the preoperative and early trends in postoperative serum thyroglobulin antibody levels among patients with papillary thyroid carcinoma and concomitant Hashimoto's thyroiditis. Endocrine 2023;80:392-8. [Crossref] [PubMed]
19. Ge H, Chen W, Lin Z, et al. Analysis of the prognostic value of thyroglobulin antibody change trends during follow-up after (131)I treatment in patients with differentiated thyroid carcinoma. Front Oncol 2025;15:1496594. [Crossref] [PubMed]
20. Zhao Y, Mu Z, Liang D, et al. Prognostic value of postoperative anti-thyroglobulin antibody in patients with differentiated thyroid cancer. Front Endocrinol (Lausanne) 2024;15:1354426. [Crossref] [PubMed]
21. Zahra HO, Omran GA, Gewely AG, et al. Prognostic Value of Serum Thyroglobulin and Anti-Thyroglobulin Antibody in Thyroid Carcinoma Patients following Thyroidectomy. Diagnostics (Basel) 2021;11:2080. [Crossref] [PubMed]
22. Choi WR, Roh JL, Gong G, et al. Multifocality of papillary thyroid carcinoma as a risk factor for disease recurrence. Oral Oncol 2019;94:106-10. [Crossref] [PubMed]
23. Li P, Jiang W, Ding Y, et al. The impact of multifocality on lateral lymph node metastasis in papillary thyroid carcinoma. Eur J Surg Oncol 2023;49:106944. [Crossref] [PubMed]
24. Mansour J, Sagiv D, Alon E, et al. Prognostic value of lymph node ratio in metastatic papillary thyroid carcinoma. J Laryngol Otol 2018;132:8-13. [Crossref] [PubMed]
25. Martin S, Mustata T, Enache O, et al. Platelet Activation and Inflammation in Patients with Papillary Thyroid Cancer. Diagnostics (Basel) 2021;11:1959. [Crossref] [PubMed]
26. Kars A, Sahin A, Kılıc K, et al. Systemic immune inflammation index in differentiated thyroid cancers. Acta Otorhinolaryngol Ital 2022;42:150-4. [Crossref] [PubMed]
27. Chen W, Yu J, Lei K, et al. Analysis of risk factors for lymph node metastasis in 241 patients with thyroid carcinoma and establishment of a prediction model. Am J Cancer Res 2024;14:3104-16. [Crossref] [PubMed]
28. Wang Y, Chang J, Hu B, et al. Systemic Immune-Inflammation Index and Systemic Inflammation Response Index Predict the Response to Radioiodine Therapy for Differentiated Thyroid Cancer. J Inflamm Res 2024;17:8531-41. [Crossref] [PubMed]
29. Chen H, Zhu L, Zhuang Y, et al. Prediction Model of Cervical Lymph Node Metastasis in Papillary Thyroid Carcinoma. Cancer Control 2024;31:10732748241295347. [Crossref] [PubMed]
30. Sulaieva O, Selezniov O, Shapochka D, et al. Hashimoto's thyroiditis attenuates progression of papillary thyroid carcinoma: deciphering immunological links. Heliyon 2020;6:e03077. [Crossref] [PubMed]
31. Aydoğan Bİ, Mutlu ABB, Yüksel S, et al. The Association of Histologically Proven Chronic Lymphocytic Thyroiditis with Clinicopathological Features, Lymph Node Metastasis, and Recurrence Rates of Differentiated Thyroid Cancer. Endocr Pathol 2021;32:280-7. [Crossref] [PubMed]
32. Eraslan Aydemir E, Özkara Akkuş F, Erdoğan MF. Anti-Thyroglobulin Antibodies as a Surrogate Tumor Marker for the Follow-Up of the Differentiated Thyroid Carcinoma: Clinical Implications and Pitfalls. Balkan Med J 2026;43:230-7. [Crossref] [PubMed]
33. Wang W, Shen C, Yang Z. Nomogram individually predicts the risk for distant metastasis and prognosis value in female differentiated thyroid cancer patients: A SEER-based study. Front Oncol 2022;12:800639. [Crossref] [PubMed]
34. Jin S, Liu H, Yang J, et al. Development and validation of a nomogram model for cancer-specific survival of patients with poorly differentiated thyroid carcinoma: A SEER database analysis. Front Endocrinol (Lausanne) 2022;13:882279. [Crossref] [PubMed]
35. Ma Q, Chen Z, Fang Y, et al. Development and validation of survival nomograms for patients with differentiated thyroid cancer with distant metastases: a SEER Program-based study. J Endocrinol Invest 2024;47:115-29. [Crossref] [PubMed]
36. Meng C, Song J, Long W, et al. A user-friendly nomogram for predicting radioiodine refractory differentiated thyroid cancer. Front Endocrinol (Lausanne) 2023;14:1109439. [Crossref] [PubMed]