Abstract
Objective
The study aimed (1) to investigate the association between aggressive clinicopathological characteristics and the American Thyroid Association (ATA) recurrence risk classification in differentiated thyroid cancer (DTC) patients, and (2) to investigate the prognostic value of preoperative thyroid parameters.
Methods
A total of 3833 patients histologically confirmed DTC were recruited. Preoperative clinical and postoperative pathologic data were retrospectively collected. Participants were stratified into low recurrence risk and intermediate-to-high recurrence risk groups based on the ATA risk stratification system.
Results
The study cohort had a mean age of 48.87 ± 8.08 years, and 1,465 (76.82%) were female. Male (OR = 1.37, p = .024), aged 52 years and older (OR = 2.01, p < .001), larger tumor size (OR = 3.71, p = 0.011), nerve invasion (OR = 6.69, p = .004), margin involvement (OR = 5.46, p < .001), multifocality (OR = 3.71, p < .001), and bilaterality (OR = 3.95, p < .001) were identified as risk factors for a higher ATA recurrence risk classification, in addition to established factors such as lymph node metastasis and angioinvasion, after adjusting for potential confounding variables. Higher preoperative levels of free triiodothyronine (FT3), FT3 to free thyroxine (FT3/FT4), and lower thyroid feedback quantile-based index (TFQI) levels were associated with a higher ATA recurrence risk classification. The comprehensive predictive model incorporating these variables demonstrated excellent discrimination (AUC = 0.836). Furthermore, higher FT3/FT4 levels and lower TFQI levels were associated with higher risk of lymph node metastases and angioinvasion.
Conclusions
Factors such as male sex, older age, multifocality, bilaterality, margin involvement, nerve invasion, larger tumor size, and preoperative thyroid parameters serve as complementary predictors for higher ATA recurrence risk in DTC, in addition to conventional risk factors. These insights contribute to a more nuanced understanding and optimization of current risk stratification methodologies.
Keywords: Differentiated thyroid cancer, risk of recurrence, American Thyroid Association risk classification, pathology, sensitivity to thyroid hormones
KEY MESSAGES
Overall, male sex, older age, multifocality, bilaterality, margin involvement, nerve invasion, larger tumor size and preoperative thyroid parameters were complementary predictors for higher ATA recurrence risk in DTC, beyond traditional risk factors.
Higher FT3/FT4 levels and lower TFQI levels were associated with a higher risk of lymph node metastases and angioinvasion in DTC patients.
Introduction
Thyroid cancer is among the most common malignant tumors. The initial management of differentiated thyroid cancer (DTC) should be guided by pre-treatment risk assessment to avoid unnecessary or overly invasive treatment [1,2]. The American Thyroid Association (ATA) guidelines [3] recommend stratifying the risk of DTC recurrence into low-, intermediate-, and high-risk categories based on clinicopathological features, including invasive pathological variants, extent of lymph node metastasis, vascular invasion, extrathyroidal extension, and distant metastasis. Patients are then managed with appropriate surgical and postoperative strategies [3]. The intermediate- and high-risk DTC patients usually undergo total thyroidectomy, radioiodine, and thyroid-stimulating hormone (TSH) suppression [3]. Postoperative TSH targets and follow-up protocols are also guided by ATA recurrence risk stratification [3]. Subsequent regimen adjustments are guided by the patients’ dynamic risk stratification and their response to therapy [3–5].
Recent research has raised questions about some elements of the ATA risk classification system, including the significance of minor extrathyroidal extension [6] and the exclusion of age [7] as a factor. Emerging evidence suggests that incorporating additional prognostic markers could improve the system’s predictive accuracy [6,8–11]. Furthermore, the results of several studies investigating multifocal disease as a potential risk factor for DTC recurrence are contradictory [12,13]. Therefore, exploring the relationship between aggressive clinicopathologic features and DTC recurrence risk is essential to better understand ATA recurrence risk stratification.
Knowing which patients are more likely to experience recurrence before surgery can aid in optimize treatment decisions. Recently, several studies explored whether some preoperative metrics, such as thyroid cytology and molecular testing results, can predict or optimize ATA recurrence risk stratification [14,15]. Despite the crucial role of the hypothalamic-pituitary-thyroid axis in the pathogenesis and prognosis of thyroid cancer [16], the association between preoperative thyroid parameters and DTC recurrence risk remains poorly characterized. Therefore, the relationship between preoperative clinical parameters, including thyroid parameters, and ATA risk classification warrants further research.
Recently, several studies employed thyroid hormone sensitivity indices to comprehensively evaluate thyroid status. Reports from earlier research showed that impaired thyroid hormones sensitivity is associated with obesity [17], diabetes [18,19], and metabolic dysfunction-associated fatty liver disease [20]. One study found a correlation between decreased levels of log TSH and thyroid feedback quantile-based index (TFQI) and a higher prevalence of thyroid nodules (including thyroid cancer) in participants aged 12–17 [21]. Notably, D. Muhanhali et al. noted that impaired central sensitivity to thyroid hormone is related with an increased risk of papillary thyroid carcinoma (PTC), as well as cervical lymph node metastasis [22]. Moreover, the relationship between thyroid hormone sensitivity and the risk of DTC recurrence warrants further research.
This study was conducted with two primary objectives: (1) to investigate the correlation between aggressive clinic pathologic features and the ATA recurrence risk classification, and (2) to explore the correlation between preoperative thyroid parameters and the risk of DTC recurrence as assessed by the 2015 ATA Risk Stratification System after the initial treatment. This helps to provide additional insights into the stratification of recurrence risk in DTC patients, potentially enhancing clinical decision-making and facilitating individualized management plans.
Methods
Participants
A total of 3833 patients histologically confirmed DTC at Wuhan Union Hospital between January 2018 and August 2022 were initially recruited for this study (Figure 1). Participants with severe hepatic and renal abnormalities, missing preoperative data or incomplete histopathologic data were excluded. Ultimately, 1907 participants were included in the study, and their initial recurrence risk was assessed according to the ATA Risk Stratification System [3,15], as shown in Figure 2. The categories were as follows: 1194 patients were in the low-risk group; 713 patients were in the intermediate- and high-risk group. This study was performed in accordance with the Declaration of Helsinki (as revised in 2013) and approved by the Institutional Review Board of Wuhan Union Hospital (approval number: No. 0173). Written informed consent was obtained from all participants.
Figure 1.
Flowchart of participant recruitment.
Figure 2.
Criteria for initial recurrence risk stratification in differentiated thyroid cancer (DTC) as assessed by the 2015 American thyroid Association (ATA) guidelines.
ATA recurrence risk stratification
Figure 2 illustrates the initial recurrence risk stratification for DTC as defined by the 2015 ATA Guidelines. The stratification criteria include key clinicopathological features such as invasive pathological variants, extent of lymph node metastasis, vascular invasion, extrathyroidal extension, and distant metastasis [3]. Patients were categorized into three risk groups based on their estimated recurrence risk: low (≤5% recurrence rate), intermediate (5%–20%), and high (≥20%) [3].
Data collection
Basic information, including gender, age, height, and weight, was collected from the participants. Body mass index (BMI) was calculated as weight (kg)/height (m) [2]. Postoperative pathological data were collected, including tumor size, multifocality, bilaterality, number and size of lymph node metastases, as well as extrathyroidal, vascular, nerve, and thyroid capsule invasion. Multifocality denotes the presence of two or more malignant thyroid nodules, while bilaterality signifies the presence of malignant thyroid nodules in distinct lobes or isthmuses. Hashimoto thyroiditis was defined as a pathology report suggestive of chronic lymphocytic thyroiditis. In instances of multiple malignant nodules, the size of the largest tumor is documented.
Laboratory examinations and thyroid hormone sensitivity indices
Preoperative TSH, free thyroxine (FT4); free triiodothyronine (FT3), thyroglobulin (Tg), Tg antibody (TgAb), and thyroid peroxidase antibody (TPOAb) were quantified using a supersensitive electrochemiluminescence immunoassay (Siemens Centaur XP, Germany). The normal reference ranges were as follows: 0.27–4.2 mIU/L for TSH; 12–22 pmol/L for FT4; 3.1–6.8 pmol/L for FT3; <115.0 IU/mL for TgAb; and <34.0 IU/mL for TPOAb. Thyrotrophic thyroxine resistance index (TT4RI) = FT4 (pmol/L) × TSH (mIU/L) [18]. TSH index (TSHI) = Ln TSH (mIU/L) + 0.1345 × FT4 (pmol/L) [18]. TFQI = empirical cumulative distribution function (cdf) FT4 − (1 − cdf TSH) [17]. The TFQI values span from −1 to 1. Higher positive values of TFQI, TSHI, and TT4RI indicate a greater impairment in central thyroid hormone sensitivity. Peripheral thyroid hormone sensitivity is based on the FT3/FT4 ratio.
Statistical analysis
We assessed the correlation coefficients between thyroid parameters, ATA recurrence risk classification, and demographic characteristics using bivariate Pearson correlation analysis for normally distributed data and Spearman correlation analysis for non-normally distributed data. Subsequently, multivariable regression models were employed to examine the associations between thyroid hormone sensitivity (FT3/FT4, TFQI, TT4RI, and TSHI), clinicopathologic features and the ATA recurrence risk classification. Receiver operating characteristic (ROC) curves were plotted and area under curve (AUC) values were calculated to determine the predictive power of the regression model. Additionally, multivariate logistic regression analysis was performed to explore the relationship between thyroid hormone sensitivity (FT3/FT4 and TFQI) and aggressive clinicopathologic features. The adjusted model was modified for sex, age, BMI, TgAb-positive, TPOAb-positive and Hashimoto thyroiditis. To evaluate the robustness of our findings, we performed sensitivity analyses by repeating the above statistical procedures in the subgroup of patients with normal thyroid function. Statistical analyses were conducted using Empower Stats (http://www.empowerstats.com) and the R package (http://www.Rproject. org). p < .05 (two-sided) was employed to determine statistical significance.
Results
Characteristics of participants
Table 1 illustrates the clinical and pathological characteristics of 1907 DTC patients, categorized based on the ATA risk stratification system for recurrence. The study population had a mean age of 48.87 ± 8.08 years, with a female predominance (76.82%, n = 1465). Based on the ATA risk stratification, 1194 patients (62.61%) were classified as low-risk, while 713 patients (37.39%) were categorized as intermediate- and high-risk. Comparative analysis revealed that patients in the intermediate- and high-risk groups exhibited a significantly higher prevalence of aggressive clinicopathologic features, including larger tumor sizes, invasive pathological variants, lymph node metastasis, vascular invasion, nerve invasion, margin involvement, extrathyroidal invasion, multifocality, and bilobar involvement. Moreover, the intermediate- and high-risk group patients exhibited a higher likelihood of being male and older, compared to those in the low-risk group. They also showed higher levels of FT3 and FT3/FT4, but lower levels of FT4 and TFQI, along with a lower incidence of concurrent Hashimoto thyroiditis (all p < .05).
Table 1.
Clinical and pathologic characteristics of 1907 patients with DTC.
| Overall | ATA low risk | ATA Intermediate and high risk | p value | |
|---|---|---|---|---|
| N | 1907 | 1194 | 713 | |
| Age (year) | 48.87 ± 8.08 | 48.47 ± 8.47 | 49.55 ± 7.34 | .005 |
| Female, n (%) | 1465 (76.82) | 959 (80.32) | 506 (70.97) | <.001 |
| BMI (kg/m2) | 23.23 ± 3.20 | 23.16 ± 3.14 | 23.36 ± 3.29 | .204 |
| FT3 (pmol/L) | 4.40 ± 0.71 | 4.36 ± 0.67 | 4.47 ± 0.77 | .002 |
| FT4 (pmol/L) | 13.36 ± 2.24 | 13.48 ± 2.22 | 13.17 ± 2.27 | .004 |
| TSH (mIU/L) | 1.74 (1.23, 2.57) | 1.75 (1.23, 2.60) | 1.71 (1.22, 2.50) | .829 |
| FT3/FT4 | 0.33 ± 0.06 | 0.33 ± 0.05 | 0.34 ± 0.06 | <.001 |
| TFQI | 0.00 ± 0.27 | 0.01 ± 0.28 | −0.03 ± 0.26 | <0.001 |
| TT4RI | 22.89 (15.88, 33.01) | 23.20 (16.18, 34.10) | 22.30 (15.44, 31.96) | .203 |
| TSHI | 2.34 (1.97, 2.72) | 2.36 (1.98, 2.75) | 2.31 (1.92, 2.67) | .162 |
| Tg (ng/mL) | 13.89 (5.39, 30.82) | 12.79 (4.76, 26.36) | 17.08 (6.29, 35.58) | .533 |
| TgAb-positive, n (%) | 221 (11.59%) | 143 (11.98) | 78 (10.94) | .878 |
| TPOAb-positive, n (%) | 304 (15.94%) | 205 (17.17) | 99 (13.88) | .165 |
| Hashimoto thyroiditis, n (%) | 702 (36.81) | 461 (38.61) | 241 (33.80) | .035 |
| Tumor size (mm) | 8.39 ± 7.49 | 5.95 ± 4.69 | 12.48 ± 9.32 | <.001 |
| Angioinvasion, n (%) | 42 (2.20) | 3 (0.25) | 39 (5.47) | <.001 |
| Nerve invasion, n (%) | 19 (1.00) | 3 (0.25) | 16 (2.24) | <.001 |
| Margin involvement, n (%) | 1055 (55.32) | 491 (41.12) | 564 (79.10) | <.001 |
| Extrathyroidal invasion, n (%) | 243 (12.74) | 0 (0.00) | 243 (34.08) | <.001 |
| LNM, n (%) | 834 (43.73) | 283 (23.70) | 551 (77.28) | <.001 |
| Multifocal, n (%) | 678 (35.55) | 293 (24.54) | 385 (54.00) | <.001 |
| Bilobar, n (%) | 480 (25.17) | 184 (15.41) | 296 (41.51) | <.001 |
| Invasive pathological variants | 11 (0.58) | 0 (0.00) | 11 (1.54) | <.001 |
Data are presented as mean ± standard deviation (SD), median (interquartile range), or n (%).
DTC: differentiated thyroid cancer; ATA: American Thyroid Association; BMI: body mass index; FT3: free triiodothyronine; FT4: free thyroxine; TSH: thyroid-stimulating hormone; TFQI: thyroid feedback quantile-based index; TT4RI: thyrotrophic thyroxine resistance index; TSHI: TSH index; Tg: thyroglobulin; TgAb: Tg antibody; TPOAb: thyroid peroxidase antibody; LNM: lymph node metastasis.
The relationship between clinicopathologic features and ATA recurrence risk classification in DTC patients
The correlations between thyroid parameters and postoperative pathologic features were showed in Figure 3. Then, multivariable regression models were used to investigate the relationship between thyroid hormone sensitivity, clinicopathologic features and the ATA recurrence risk classification as shown in Table 2. After adjusting for age, sex, BMI, TgAb-positive, TPOAb-positive and Hashimoto thyroiditis, each standard deviation (SD) increase in tumor size was associated with a 3.71-fold [OR (95% CI) 3.71 (3.08, 4.07), p = .011] higher prevalence of intermediate and high risk of recurrence. Male (OR = 1.37, p = .024), aged 52 years and older (OR = 2.01, p < .001), nerve invasion (OR = 6.69, p = .004), margin involvement (OR = 5.46, p < .001), multifocality (OR = 3.71, p < .001), and bilaterality (OR = 3.95, p < .001) were risk factors for higher ATA recurrence risk classification, in addition to established factors such as lymph node metastasis and angioinvasion.
Figure 3.
Correlation plots showing correlation values of thyroid parameters and postoperative pathologic parameters of differentiated thyroid carcinoma.
Table 2.
Multivariate analysis for factors predicting ATA intermediate and high initial risk of recurrence in DTC patients.
| ATA intermediate and high initial risk of recurrence |
||||
|---|---|---|---|---|
| Unadjusted model | Adjusted model | |||
| OR (95% CI) | p value | OR (95% CI) | p value | |
| FT3 per SD | 1.16 (1.05, 1.27) | .003 | 1.15 (1.03, 1.29) | .011 |
| FT4 per SD | 0.86 (0.78, 0.95) | .004 | 0.92 (0.82, 1.03) | .148 |
| TSH per SD | 0.99 (0.90, 1.09) | .829 | 0.99 (0.89, 1.10) | .813 |
| FT3/FT4 per SD | 1.30 (1.18, 1.43) | <.001 | 1.23 (1.09, 1.37) | <.001 |
| TFQI per SD | 0.84 (0.77, 0.93) | <.001 | 0.86 (0.76, 0.96) | .007 |
| TT4RI per SD | 0.94 (0.85, 1.04) | .205 | 0.95 (0.85, 1.06) | .364 |
| TSHI per SD | 0.94 (0.86, 1.03) | .164 | 0.92 (0.84, 1.02) | .128 |
| Hashimoto thyroiditis (yes vs no) | 0.81 (0.67, 0.99) | .035 | 0.86 (0.70, 1.05) | .145 |
| Male vs female | 1.67 (1.35, 2.07) | <.001 | 1.37 (1.04, 1.79) | .024 |
| Age ≥ 52 years vs < 52 years | 2.15 (1.73, 2.67) | <.001 | 2.01 (1.58, 2.58) | <.001 |
| BMI ≥ 25 kg/m2 vs < 25 kg/m2 | 1.30 (1.05, 1.61) | .015 | 1.10 (0.86, 1.42) | .440 |
| Tumor size per SD | 4.16 (3.50, 4.94) | <.001 | 3.71 (3.08, 4.07) | <.001 |
| Angioinvasion (yes vs no) | 22.97 (7.08, 74.51) | <.001 | 26.47 (6.28, 111.62) | <.001 |
| Nerve invasion (yes vs no) | 9.11 (2.65, 31.39) | <.001 | 6.69 (1.85, 24.22) | .004 |
| Margin involvement (yes vs no) | 5.42 (4.37, 6.71) | <.001 | 5.46 (4.23, 7.05) | <.001 |
| LNM (yes vs no) | 10.95 (8.79, 13.65) | <.001 | 10.26 (7.92,13.28) | <.001 |
| Multifocal (yes vs no) | 3.61 (2.96, 4.40) | <.001 | 3.71 (2.94, 4.69) | <.001 |
| Bilobar (yes vs no) | 3.90 (3.14, 4.84) | <.001 | 3.95 (3.08, 5.08) | <.001 |
The evaluation of the odds ratio (OR) and 95% confidence interval (CI) was conducted using multivariable logistic regression models.
Adjusted model: adjusted for sex, age, BMI, TgAb-positive, TPOAb-positive and hashimoto thyroiditis.
ATA: American Thyroid Association; DTC: differentiated thyroid cancer; FT3: free triiodothyronine; SD: standard deviation; FT4: free thyroxine; TSH: thyroid-stimulating hormone; TFQI: thyroid feedback quantile-based index; TT4RI: thyrotrophic thyroxine resistance index; TSHI: TSH index; BMI: body mass index; LNM: lymph node metastasis; TgAb: thyroglobulin antibody; TPOAb: thyroid peroxidase antibody.
The relationship between thyroid hormone sensitivity and ATA recurrence risk classification in DTC patients
After adjusting for age, sex, BMI, TgAb-positive, TPOAb-positive and Hashimoto thyroiditis, each SD increase in FT3 and FT3/FT4 was associated with a 1.15-fold [OR (95% CI) 1.15 (1.03, 1.29), p = .011] and 1.23-fold [OR (95% CI) 1.23 (1.09, 1.37), p < .001] higher prevalence of intermediate and high risk of recurrence, respectively (Table 2). A higher ATA recurrence risk classification [OR (95% CI) 0.86 (0.76, 0.96), p = .007] was observed in patients with decreased TFQI levels. Notably, the relationship between FT3/FT4 and ATA recurrence risk classification remained statistically significant even after adjusting for aggressive histological characteristics (Supplementary Table 1). A linear relationship between thyroid sensitivity and ATA recurrence risk stratification in DTC patients with normal thyroid function were observed (Supplementary Figure 1).
ROC analysis of clinicopathologic features for predicting higher ATA recurrence risk
Figure 4 presents the ROC curve with the AUC for the multivariate logistic regression analysis assessing factors predicting ATA intermediate and high initial recurrence risk. The model included demographic variables (male sex and age ≥52 years), preoperative thyroid parameters (FT3, FT3/FT4, and TFQI), and aggressive clinicopathological features (tumor size, nerve invasion, margin involvement, multifocality, and bilaterality). Notably, established risk factors such as lymph node metastasis and angioinvasion were not included in this model. The AUC value of the model was 0.836, indicating a robust predictive performance for identifying patients at higher risk of recurrence.
Figure 4.
Receiver operating characteristic (ROC) curve with area under curve (AUC) of multivariate logistic analysis for factors predicting ATA intermediate and high initial risk of recurrence in DTC patients.
The relationship between thyroid hormone sensitivity and aggressive clinicopathologic features in DTC patients
Multivariate regression analysis also confirmed that higher levels of FT3/FT4 were independent risk factor for a higher risk of lymph node metastases [OR (95% CI) 1.16 (1.04, 1.29), p = .008], vascular invasion [OR (95% CI) 1.49 (1.14, 1.94), p = .003] and invasive pathological variants [OR (95% CI) 1.38 (1.01, 1.89), p = .042], after controlling for potential confounding variables (Table 3). Moreover, lower levels of TFQI were associated with a higher risk of lymph node metastases and vascular invasion. No significant correlations were observed between FT3/FT4 or TFQI with other aggressive clinicopathologic features, including nerve invasion, margin involvement, extrathyroidal invasion, multifocality, bilobar involvement, or tumor size.
Table 3.
Odds ratios (or) of thyroid hormone sensitivity indices to histological characteristics in DTC patients.
| Thyroid hormone sensitivity indices |
||||
|---|---|---|---|---|
| FT3/FT4 | TFQI | |||
| Outcome variables | OR (95% CI) | p value | OR (95% CI) | p value |
| Angioinvasion | 1.49 (1.14, 1.94) | .003 | 0.67 (0.47, 0.96) | .028 |
| Nerve invasion | 0.81 (0.47, 1.40) | .456 | 1.04 (0.60, 1.79) | .901 |
| Margin involvement | 1.05 (0.94, 1.16) | .401 | 1.09 (0.98, 1.22) | .106 |
| Lymph node metastases | 1.16 (1.04, 1.29) | .008 | 0.84 (0.75, 0.94) | .002 |
| Multifocal | 1.02 (0.92, 1.14) | .681 | 0.95 (0.85, 1.06) | .385 |
| Bilobar | 0.95 (0.85, 1.08) | .446 | 0.98 (0.87, 1.11) | .727 |
| Extrathyroidal invasion | 1.01 (0.87, 1.17) | .925 | 1.03 (0.88, 1.20) | .757 |
| Invasive pathological variants | 1.38 (1.01, 1.89) | .042 | 0.75 (0.39, 1.46) | .401 |
| Tumor size | 0.35 (−0.06, 0.76) | .094 | 0.26 (−0.13, 0.65) | .195 |
The evaluation of the OR and 95% confidence interval (CI) was conducted using multivariable logistic regression models.
Adjusted model: adjusted for sex, age, BMI, TgAb-positive, TPOAb-positive, hashimoto thyroiditis.
DTC: differentiated thyroid cancer; FT3: free triiodothyronine; FT4: free thyroxine; TFQI: thyroid feedback quantile-based index; BMI: body mass index; TgAb: thyroglobulin antibody; TPOAb: thyroid peroxidase antibody.
Sensitivity analysis
We also investigated the correlation between thyroid parameters, aggressive clinicopathologic features and ATA recurrence risk classification for DTC in the population with normal thyroid function. Most of the findings were consistent with previous results, as detailed in Supplementary Table 2 and Supplementary Table 3.
Discussion
This study provides comprehensive evidence that tumor size, multifocality, bilaterality, margin involvement, and nerve invasion serve as independent predictors of higher ATA recurrence risk classification in DTC patients, complementing established risk factors such as invasive pathological variants, lymph node metastasis, vascular invasion, and extrathyroidal invasion. Furthermore, the study demonstrated that higher preoperative FT3 levels and increased thyroid hormones sensitivity were associated with an increased risk of recurrence, as well as an increased risk of lymph node metastases and angioinvasion. Notably, our multivariate model exhibited excellent predictive performance (AUC = 0.836), highlighting the clinical relevance of these parameters in risk stratification.
Although thyroid cancer is associated with a low mortality rate, persistent or recurrent disease remains a significant clinical concern [2]. The ATA guidelines [3] recommend stratifying the risk of recurrence of DTC into low-, intermediate-, and high-risk categories to guide surgical and postoperative management strategies [23]. However, recent studies have questioned certain aspects of the ATA risk classification system, such as the clinical significance of minor extrathyroidal extension [6], and the exclusion of age [7] as a predictive factor. Additionally, some studies have proposed incorporating additional predictive indicators to enhance the accuracy of the system, such as stimulated Tg before radioiodine ablation [10,11], postoperative Tg [9], extranodal extension status, and metastatic lymph node ratios [10]. Our findings align with previous studies that has identified tumor size, advanced age, and male sex as predictors of DTC recurrence, in addition to well-established pathological features including invasive variants, lymph node metastasis, vascular invasion, and extrathyroidal extension [8,23–25]. This study further validates and extends the generalizability of these clinical predictors.
Nerve invasion is an infrequent pathological characteristic in PTC, yet it has been documented as being correlated with extrathyroidal invasion and an increased risk of recurrence [26]. Margin involvement, another pathological characteristic, has been linked to decreased survival rates in PTC [27,28]. In alignment with these observations, our study identified both margin involvement (OR = 5.46, p < .001) and nerve invasion (OR = 6.69, p = .004) as independent predictors of higher ATA recurrence risk classification.
Multifocality and bilaterality represent common pathological characteristics in DTC. In 9.2% of multifocal PTC cases, molecular drivers were found to be inconsistent between tumor foci [29], and single-cell transcriptome analyses have demonstrated heterogeneity between bilateral PTC tumors [30], This suggests the potential need for individualized therapeutic approaches for such patients. Our study identified prevalence rates of 35.6% for multifocality and 25.2% for bilaterality, aligning with previous findings [12]. Notably, our comprehensive analysis revealed significant correlations between postoperative pathological features, indicating that multifocality and bilaterality are significantly associated with lymph node metastasis, tumor size, margin involvement, and extrathyroidal invasion. These findings are consistent with several studies suggesting a more aggressive disease profile in multifocal and bilateral DTC [13,31–34]. However, the role of multifocality and bilaterality as risk factors for the progression and recurrence of DTC remains controversial [35]. While some studies have identified multifocality as an independent risk factor for recurrence [31,36,37], others have failed to demonstrate its prognostic value [38,39]. Notably, a recent study suggests that bilaterality, rather than multifocality, may serve as a more clinically significant predictor of disease recurrence [12]. This study suggests that patients with unilateral multifocal lesions may not require total thyroidectomy, whereas those with bifocal lesions must be managed with caution and close follow-up [12]. The observed inconsistencies may arise from variations in study populations, sample sizes, and outcome definitions. Importantly, our study, which is based on a large sample population, provides robust evidence that factors such as tumor size, multifocality, margin involvement, nerve invasion, and bilaterality are significantly associated with a higher ATA recurrence risk classification. The strong predictive performance of our multivariate model (AUC = 0.836) further underscores the clinical relevance of these factors in recurrence risk stratification. These findings may offer complementary prognostic information to existing risk stratification approaches.
The ability to determine post-surgical disease status before surgery helps to optimize the treatment plan. However, existing risk stratification models for recurrence predominantly rely on postoperative histopathologic characteristics and do not incorporate a comprehensive dataset that integrates clinical and histologic factors [1,2]. While several studies have investigated the predictive value of preoperative metrics such as thyroid cytology and molecular testing for optimizing ATA recurrence risk stratification [14,15], the association between preoperative thyroid parameters and DTC recurrence risk has not been extensively investigated. Our study indicated that increased levels of preoperative FT3 and increased sensitivity to thyroid hormones, both centrally and peripherally, were linked to a higher ATA recurrence risk classification.
The role of thyroid hormone signaling in thyroid cancer progression remains incompletely understood [16]. Our study provides novel evidence that higher FT3/FT4 levels and lower TFQI levels were significantly associated with a higher risk of lymph node metastasis and angioinvasion in DTC. These findings align with some preclinical research demonstrating that thyroid hormone, particularly thyroxine and to a lesser extent triiodothyronine, can promote tumor proliferation and angiogenesis via classical thyroid hormone receptors and the integrin αvβ3 pathway [16,40]. Miyauchi et al. [41] reported cases of patients with large or massive metastatic follicular thyroid carcinoma who exhibiting decreased serum FT4 levels, as well as increased serum FT3 levels and higher FT3/FT4 ratios. While our results appear to contrast with those of Muhanhali et al. who reported that impaired central thyroid hormone sensitivity increases the risk of cervical lymph node metastasis [22]. However, numerous studies have established that autoimmune mechanisms significantly influence the clinical behavior and prognosis of DTC [42,43], a factor not accounted for in Muhanhali et al.’s analysis. Moreover, the thyroid autoimmune process may also affect thyroid function [44]. In contrast, our study rigorously adjusted for potential confounders, including TgAb/TPOAb positivity and Hashimoto thyroiditis, with results further validated through sensitivity analyses.
This study has some limitations. Firstly, while we utilized the ATA risk stratification system rather than actual 1-year recurrence outcomes due to the generally favorable prognosis and low recurrence rates observed in DTC patients. Our substantial sample size provided robust statistical power to investigate associations between risk stratification and clinicopathologic characteristics. Secondly, we did not incorporate molecular profiling data, such as BRAF mutation status, which could potentially enhance the precision of risk assessment. Lastly, given the retrospective nature of this study, our findings would benefit from external validation across diverse populations to confirm their generalizability.
Conclusions
This comprehensive study elucidates that multifocality, bilaterality, margin involvement, nerve invasion, and tumor size independently predict higher ATA recurrence risk in DTC, beyond traditional risk factors. Furthermore, preoperative thyroid parameters, specifically higher FT3/FT4 levels and decreased TFQI levels, are associated with recurrence risk and aggressive clinicopathological features, suggesting their potential utility in risk stratification. These findings collectively advance our understanding of DTC recurrence risk determinants and support the potential for optimizing current risk stratification system.
Supplementary Material
Acknowledgements
Our gratitude extends to all participants of this study for their invaluable contributions. Also, the authors would like to express their gratitude to EditSprings (https://www.editsprings.cn) for the expert linguistic services provided.
Funding Statement
This work was supported by the Hubei Provincial Office of Science and Technology under Grant number [2023BCB131], and Wuhan Municipal Science and Technology Bureau under Grant number [2023020201010161].
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
Data underlying the findings of this study are available upon request from the corresponding author.
References
- 1.Chen DW, Lang BHH, McLeod DSA, et al. Thyroid cancer. Lancet. 2023;401(10387):1531–1544. doi: 10.1016/s0140-6736(23)00020-x. [DOI] [PubMed] [Google Scholar]
- 2.Boucai L, Zafereo M, Cabanillas ME.. Thyroid cancer: a review. JAMA. 2024;331(5):425–435. doi: 10.1001/jama.2023.26348. [DOI] [PubMed] [Google Scholar]
- 3.Haugen BR, Alexander EK, Bible KC, et al. 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid. 2016;26(1):1–133. doi: 10.1089/thy.2015.0020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tuttle RM, Alzahrani AS.. Risk stratification in differentiated thyroid cancer: from detection to final follow-up. J Clin Endocrinol Metab. 2019;104(9):4087–4100. doi: 10.1210/jc.2019-00177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Abelleira E, Jerkovich F.. Dynamic risk assessment in patients with differentiated thyroid cancer. Rev Endocr Metab Disord. 2024;25(1):79–93. doi: 10.1007/s11154-023-09857-7. [DOI] [PubMed] [Google Scholar]
- 6.Alzahrani AS. The risk of expanding risk stratification in thyroid cancer. J Clin Endocrinol Metab. 2023;108(10):e1147–e1148. doi: 10.1210/clinem/dgad182. [DOI] [PubMed] [Google Scholar]
- 7.Maino F, Botte M, Dalmiglio C, et al. Prognostic factors improving ATA risk system and dynamic risk stratification in low- and intermediate-risk DTC patients. J Clin Endocrinol Metab. 2024;109(3):722–729. doi: 10.1210/clinem/dgad591. [DOI] [PubMed] [Google Scholar]
- 8.Grani G, Gentili M, Siciliano F, et al. A Data-driven approach to refine predictions of differentiated thyroid cancer outcomes: a prospective multicenter study. J Clin Endocrinol Metab. 2023;108(8):1921–1928. doi: 10.1210/clinem/dgad075. [DOI] [PubMed] [Google Scholar]
- 9.Giovanella L, Milan L, Roll W, et al. Postoperative thyroglobulin as a yard-stick for radioiodine therapy: decision tree analysis in a European multicenter series of 1317 patients with differentiated thyroid cancer. Eur J Nucl Med Mol Imaging. 2023;50(9):2767–2774. doi: 10.1007/s00259-023-06239-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang H, Li Q, Tian T, et al. Improving the risk prediction of the 2015 ATA recurrence risk stratification in papillary thyroid cancer. J Clin Endocrinol Metab. 2025;110(2):534–541. doi: 10.1210/clinem/dgae465. [DOI] [PubMed] [Google Scholar]
- 11.Wijewardene A, Gill AJ, Gild M, et al. A retrospective cohort study with validation of predictors of differentiated thyroid cancer outcomes. Thyroid. 2022;32(10):1201–1210. doi: 10.1089/thy.2021.0563. [DOI] [PubMed] [Google Scholar]
- 12.Rodriguez Schaap PM, Lin JF, Metman MJH, et al. Bilaterality, not multifocality, is an independent risk factor for recurrence in low-risk papillary thyroid cancer. J Natl Cancer Inst. 2023;115(9):1071–1076. doi: 10.1093/jnci/djad105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Feng J-W, Qu Z, Qin A-C, et al. Significance of multifocality in papillary thyroid carcinoma. Eur J Surg Oncol. 2020;46(10 Pt A):1820–1828. doi: 10.1016/j.ejso.2020.06.015. [DOI] [PubMed] [Google Scholar]
- 14.Liu JB, Baugh KA, Ramonell KM, et al. Molecular testing predicts incomplete response to initial therapy in differentiated thyroid carcinoma without lateral neck or distant metastasis at presentation: retrospective cohort study. Thyroid. 2023;33(6):705–714. doi: 10.1089/thy.2023.0060. [DOI] [PubMed] [Google Scholar]
- 15.Croce L, Teliti M, Chytiris S, et al. The American Thyroid Association risk classification of papillary thyroid cancer according to presurgery cytology. Eur J Endocrinol. 2024;190(2):165–172. doi: 10.1093/ejendo/lvae012. [DOI] [PubMed] [Google Scholar]
- 16.Abaandou L, Ghosh R, Klubo-Gwiezdzinska J.. The role of the hypothalamic-pituitary-thyroid axis in thyroid cancer. Lancet Diabetes Endocrinol. 2025;13(4):333–346. doi: 10.1016/s2213-8587(24)00364-4. [DOI] [PubMed] [Google Scholar]
- 17.Laclaustra M, Moreno-Franco B, Lou-Bonafonte JM, et al. Impaired sensitivity to thyroid hormones is associated with diabetes and metabolic syndrome. Diabetes Care. 2019;42(2):303–310. doi: 10.2337/dc18-1410. [DOI] [PubMed] [Google Scholar]
- 18.Mehran L, Delbari N, Amouzegar A, et al. Reduced sensitivity to thyroid hormone is associated with diabetes and hypertension. J Clin Endocrinol Metab. 2022;107(1):167–176. doi: 10.1210/clinem/dgab646. [DOI] [PubMed] [Google Scholar]
- 19.Wan H, Yu G, He Y, et al. Associations of thyroid feedback quantile-based index with diabetes in euthyroid adults in the United States and China. Ann Med. 2024;56(1):2318418. doi: 10.1080/07853890.2024.2318418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wan H, Yu G, Xu S, et al. Central sensitivity to free triiodothyronine with MAFLD and its progression to liver fibrosis in euthyroid adults. J Clin Endocrinol Metab. 2023;108(9):e687–e697. doi: 10.1210/clinem/dgad186. [DOI] [PubMed] [Google Scholar]
- 21.Suzuki S, Suzuki S, Iwadate M, et al. Possible association between thyroid nodule formation and developmental alterations in the pituitary-thyroid hormone axis in children and adolescents: the Fukushima health management survey. Thyroid. 2022;32(11):1316–1327. doi: 10.1089/thy.2022.0327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Muhanhali D, Deng L, Ai Z, et al. Impaired thyroid hormone sensitivity increases the risk of papillary thyroid cancer and cervical lymph node metastasis. Endocrine. 2024;83(3):659–670. doi: 10.1007/s12020-023-03508-2. [DOI] [PubMed] [Google Scholar]
- 23.Attia A, Touma E, Lussey-Lepoutre C, et al. Consideration of early dynamic risk stratification to guide discharge from oncologic follow-up in patients with differentiated thyroid cancer. Thyroid. 2024;34(12):1465–1475. doi: 10.1089/thy.2024.0119. [DOI] [PubMed] [Google Scholar]
- 24.Piccardo A, Siri G, Raffa S, et al. How to better stratify the risk of differentiated thyroid carcinomas: the key role of radioactive iodine therapy, age, and gender. Eur J Nucl Med Mol Imaging. 2021;48(3):822–830. doi: 10.1007/s00259-020-05020-5. [DOI] [PubMed] [Google Scholar]
- 25.Zuhur SS, Aggul H, Çelik M, et al. Can age at diagnosis and sex improve the performance of the american thyroid association risk stratification system for prediction of structural persistent and recurrent disease in patients with differentiated thyroid carcinoma? A multicenter study. Endocr Pract. 2022;28(1):30–35. doi: 10.1016/j.eprac.2021.09.001. [DOI] [PubMed] [Google Scholar]
- 26.Limberg J, Lee-Saxton YJ, Egan CE, et al. Perineural invasion in papillary thyroid cancer: a rare indicator of aggressive disease. Ann Surg Oncol. 2023;30(6):3570–3577. doi: 10.1245/s10434-023-13307-8. [DOI] [PubMed] [Google Scholar]
- 27.Lee CI, Kutlu O, Khan ZF, et al. Margin positivity and survival in papillary thyroid microcarcinoma: a national cancer database analysis. J Am Coll Surg. 2021;233(4):537–544. doi: 10.1016/j.jamcollsurg.2021.06.011. [DOI] [PubMed] [Google Scholar]
- 28.Khan ZF, Kutlu O, Picado O, et al. Margin positivity and survival outcomes: a review of 14,471 patients with 1-cm to 4-cm papillary thyroid carcinoma. J Am Coll Surg. 2021;232(4):545–550. doi: 10.1016/j.jamcollsurg.2020.12.018. [DOI] [PubMed] [Google Scholar]
- 29.Rivera JP, Yeh YC, Chen PC, et al. Multifocal papillary thyroid carcinomas with discordant molecular drivers: emphasizing the morphology and collision tumors. Am J Surg Pathol. 2024;48(11):1359–1371. doi: 10.1097/pas.0000000000002256. [DOI] [PubMed] [Google Scholar]
- 30.Wang T, Shi J, Li L, et al. Single-cell transcriptome analysis reveals inter-tumor heterogeneity in bilateral papillary thyroid carcinoma. Front Immunol. 2022;13:840811. doi: 10.3389/fimmu.2022.840811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Joseph KR, Edirimanne S, Eslick GD.. Multifocality as a prognostic factor in thyroid cancer: A meta-analysis. Int J Surg. 2018;50:121–125. doi: 10.1016/j.ijsu.2017.12.035. [DOI] [PubMed] [Google Scholar]
- 32.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(9):106944. doi: 10.1016/j.ejso.2023.05.019. [DOI] [PubMed] [Google Scholar]
- 33.Parvathareddy SK, Siraj AK, Annaiyappanaidu P, et al. Bilateral multifocality is an independent predictor of patients’ outcome in Middle Eastern papillary thyroid carcinoma. Front Endocrinol (Lausanne). 2022;13:1060301. doi: 10.3389/fendo.2022.1060301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Rodriguez Schaap PM, Papachristos A, Serrao-Brown H, et al. Predictors of bilateral disease in low-risk papillary thyroid cancer: histopathologic insights and preoperative ultrasonography. Ann Surg Oncol. 2025;32(4):2335–2343. doi: 10.1245/s10434-024-16352-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wang Z, Ji X, Zhang H, et al. Clinical and molecular features of progressive papillary thyroid microcarcinoma. Int J Surg. 2024;110(4):2313–2322. doi: 10.1097/js9.0000000000001117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Evans Harding N, Simo R, Li L, et al. A quantitative assessment of the number of disease foci in papillary thyroid cancer. Eur J Surg Oncol. 2023;49(7):1141–1146. doi: 10.1016/j.ejso.2022.11.592. [DOI] [PubMed] [Google Scholar]
- 37.Sun Y, Liu Y, Li H, et al. The significance and prognostic value of multifocal papillary thyroid carcinoma in children and adolescents. BMC Cancer. 2024;24(1):690. doi: 10.1186/s12885-024-12403-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Geron Y, Benbassat C, Shteinshneider M, et al. Multifocality is not an independent prognostic factor in papillary thyroid cancer: a propensity score-matching analysis. Thyroid. 2019;29(4):513–522. doi: 10.1089/thy.2018.0547. [DOI] [PubMed] [Google Scholar]
- 39.Scholfield DW, Lopez J, Eagan A, et al. Is multifocality a predictor of poor outcome in childhood and adolescent papillary thyroid carcinoma? J Clin Endocrinol Metab. 2023;108(12):3135–3144. doi: 10.1210/clinem/dgad369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schmohl KA, Han Y, Tutter M, et al. Integrin αvβ3-dependent thyroid hormone effects on tumour proliferation and vascularisation. Endocr Relat Cancer. 2020;27(12):685–697. doi: 10.1530/erc-20-0353. [DOI] [PubMed] [Google Scholar]
- 41.Miyauchi A, Takamura Y, Ito Y, et al. 3,5,3’-Triiodothyronine thyrotoxicosis due to increased conversion of administered levothyroxine in patients with massive metastatic follicular thyroid carcinoma. J Clin Endocrinol Metab. 2008;93(6):2239–2242. doi: 10.1210/jc.2007-2282. [DOI] [PubMed] [Google Scholar]
- 42.Paparodis R, Livadas S, Karvounis E, et al. Elevated preoperative TPO Ab titers decrease risk for DTC in a linear fashion: a retrospective analysis of 1635 cases. J Clin Endocrinol Metab. 2023;109(1):e347–e355. doi: 10.1210/clinem/dgad408. [DOI] [PubMed] [Google Scholar]
- 43.De Leo S, D’Elia S, Grani G, et al. A prospective multicenter study examining the relationship between thyroid cancer treatment outcomes and the presence of autoimmune thyroiditis. Thyroid. 2023;33(11):1318–1326. doi: 10.1089/thy.2023.0052. [DOI] [PubMed] [Google Scholar]
- 44.Razvi S. Thyroid-function reference ranges in the diagnosis of thyroid dysfunction in adults. Nat Rev Endocrinol. 2024;20(5):253–254. doi: 10.1038/s41574-024-00972-2. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data underlying the findings of this study are available upon request from the corresponding author.