Impaired thyroid hormone sensitivity increases the risk of papillary thyroid cancer and cervical lymph node metastasis

Background: The association of thyroid hormone sensitivity with papillary thyroid carcinoma (PTC) isunclear. This study investigated the relationship between the thyroid hormone sensitivity indices and the risk of PTC and the in�uence of thyroid hormone sensitivity on the aggressive clinicopathologic features of PTC. Methods: This retrospective studyrecruited 1225 PTC patients and 369 patients with benign nodules undergoing surgery in Zhongshan Hospital in 2020. The thyroid hormone sensitivity indices were thyroid feedback quantile-based index (TFQI), TSH index (TSHI) and thyrotropin thyroxine resistance index (TT4RI). We employed logistic regression models to explore the correlation between the thyroid hormone sensitivity indices and the risk of PTC and its cervical lymph node metastasis (LNM). Results: PTC patients had signi�cantly higher levels of TSH, TFQI, TSHI and TT4RI compared to the patients with benign nodules, but thyroid hormone levels did not differ signi�cantly between the two groups. Logistic regression analysis revealed that the higher levels of TFQI, TSHI, and TT4RI were associated with an increased risk of PTC after adjustment for multiple risk factors (TFQI: OR=1.92, 95%CI: 1.39-2.65, P<0.001; TSHI: OR=2.33, 95%CI:1.67-3.26, P<0.001; TT4RI: OR=2.41, 95%CI:1.73-3.36, P<0.001). In addition, patients with decreased thyroid hormone sensitivity had a higher risk of cervical LNM in multiple logistic regression analysis (TFQI: OR=1.38, 95%CI:1.03-1.86, P=0.03; TSHI: OR=1.37, 95%CI:1.02-1.84, P=0.04; TT4RI: OR= 1.40, 95%CI:1.05-1.89, P=0.02). Conclusion:


Introduction
Thyroid cancer has increased rapidly over the last two decades in China. The age-standardized incidence was 3.21/10 5 in 2005 and increased to 9.61/10 5 in 2015 (1). The most common histological subtype is papillary thyroid carcinoma (PTC) which accounts for 80% of the new cases (1). The pathogenesis of PTC has not yet been fully clari ed, and the most well-recognized risk factors for thyroid cancer include older age, female gender, ionizing radiation exposure and obesity (2)(3)(4). Recently, a lot of studies investigated whether TSH and thyroid hormone had an in uence on the development of thyroid cancer, but the results were inconsistent.
The ndings of epidemiological studies about the correlations between TSH, thyroid hormone and the risk of PTC remain controversial (5)(6)(7)(8)(9)(10)(11)(12). The majority of previous studies reported an increased risk of thyroid cancer with elevated TSH concentrations (5,7,10,11,13), whereas several studies detected no association or inverse association between TSH level and the risk of thyroid cancer (6,8). Although the relationship between TSH and the risk of PTC has not been clearly de ned, it has been demonstrated that TSH suppression therapy improves the outcomes of differentiated thyroid cancer patients who are at high risks of recurrence and death (14). The results about the association of thyroid hormone with the risk of PTC are also inconsistent. Several studies found lower thyroid hormone levels were associated with higher risk of thyroid cancer (5,10,11), while others reported no association (8,12).
Physiologically, the inverse correlation between thyroid hormones and TSH is regulated via the negative feedback loop of the hypothalamic-pituitary-thyroid (HPT) axis (15). Due to the complex interactions within the HPT axis, using a single index may not be su cient to explain the relationship between thyroid function and PTC. Indices of sensitivity to thyroid hormone such as thyrotropin thyroxine resistance index (TT4RI), TSH index (TSHI) and thyroid feedback quantile-based index (TFQI) were proposed to re ect the regulation of thyroid hormone homeostasis (16)(17)(18). Recently, many studies have demonstrated that reduced sensitivity to thyroid hormone in the general population was associated with higher risk of metabolic syndrome, hypertension, obesity, diabetes, osteoarthritis, non-alcoholic fatty liver disease and impaired renal function(18-26).
It is not known whether changes of sensitivity to thyroid hormone impact the development and progression of PTC. In the current study, we aimed to investigate the relationship between sensitivity to thyroid hormone and the risk of PTC and the possible in uence of thyroid hormone sensitivity on the aggressive clinicopathologic features of PTC.

Study population
In this retrospective study, the medical records of patients received surgery for thyroid nodules in Zhongshan Hospital a liated to Fudan University in 2020 were reviewed. 1725 patients with malignant nodules and 510 patients with benign thyroid nodules were screened in the current study. The exclusion criteria were as follows: (1) patients with non-PTC malignant nodules; (2) previous history of thyroid surgery; (3) missing data; (4) previous history of malignancy; (5) thyroid dysfunction taking anti-thyroid or thyroid replacement drugs; (6) abnormal baseline thyroid function (clinical/subclinical hypothyroidism or hyperthyroidism). Finally, we recruited 1225 PTC patients and 369 patients with benign nodules and analyzed the association of thyroid hormone sensitivity with PTC risk in these patients. Among 1225 PTC patients, 920 PTC patients underwent cervical lymph node dissection, and 305 PTC patients did not undergo cervical lymph node dissection. The relationship between sensitivity to thyroid hormone and the risk of cervical lymph node metastasis was further analyzed in 920 PTC patients who underwent cervical lymph node dissection. The owchart of patient recruitment is shown in Fig. 1. This study was approved by the Ethics Committees of Zhongshan Hospital and all patients gave written informed consent.
TFQI ranges from − 1 to + 1 with positive values indicating poor thyroid hormone sensitivity and negative values good thyroid hormone sensitivity, and 0 means that central sensitivity to FT 4 is normal.

Pathological variables
Total or near total thyroidectomy or thyroid lobectomy with or without cervical lymph node dissection were applied to all enrolled patients. PTC patients with cervical lymph node dissection underwent central lymph node dissection or central plus lateral lymph node dissection. Lateral lymph node dissection was only performed in patients with clinically suspicious lateral lymph node metastasis based on preoperative imaging results or con rmed by ne-needle aspiration. The pathological variables included tumor size, cervical lymph node metastasis (LNM), central LNM, number of metastatic cervical lymph node, number of metastatic central lymph node, the ratio of cervical LNM, the ratio of central LNM, multifocality, bilaterality, thyroid capsule invasion, extrathyroidal extension, distant metastasis, BRAF V600E mutation, tumor, node, and metastasis (TNM) stage. The ratio of cervical or central LNM were calculated as the number of metastatic cervical or central lymph node divided by the total number of dissected cervical or central lymph node. Multifocality represented the existence of two or more malignant thyroid nodules.
Bilaterality indicated that the malignant thyroid nodules presenting in different lobes or isthmus. If multiple malignant nodules were present, tumor size was recorded based on the largest one. Extrathyroidal extension was de ned as invasion into sternothyroid muscle, subcutaneous soft tissue, larynx, trachea, esophagus, recurrent laryngeal nerve, or vasculature. PTC was staged according to the Eighth Edition American Joint Committee on Cancer TNM staging system (27).

Statistical analysis
Statistical analysis was performed using SPSS version 23.0 Software and statistically signi cant differences were considered by those with p-values < 0.05 in all analysis. Results were expressed as means (standard deviation, SD) or median (interquartile range, IQR) for continuous variables and as proportions for categorical variables. Comparison of normally distributed data was performed using the t test, and Mann-Whitney U test was performed to compare differences in non-normally distributed continuous variables. The chi-square test was used to compare categorical variables. We constructed multiple logistic regression models to explore the associations of thyroid hormone sensitivity (TFQI, TSHI and TT4RI) with PTC and cervical LNM, and estimated odds ratios (OR) and corresponding 95% con dence intervals (CI). Graphs were generated using GraphPad 9.2.

Baseline characteristics
Comparison of baseline characteristics between the PTC patients and patients with benign thyroid nodules was presented in Table 1. A total of 1225 PTC patients were enrolled in the study, which consisted of 857 (69.96%) females and 368 (30.04%) males with an average age of 44 years. Among 1225 PTC patients, 920 PTC patients underwent cervical lymph node dissection, and 305 PTC patients did not undergo cervical lymph node dissection. There were 369 patients with benign nodules included in the analysis, of whom 262 (71.00%) were females and 107 (29.00%) were males, and the mean age of them was 50 years. The PTC patients were younger compared to the patients with benign nodules (P < 0.001) ( Table 1). The thyroid hormone levels did not differ signi cantly between the PTC patients and patients with benign nodules, but the TSH level in PTC patients was higher than that of patients with benign nodules (P < 0.001) ( Table 1). The PTC patients had signi cantly higher levels of TFQI, TSHI and TT4RI compared to the patients with benign nodules (all P values < 0.001) ( Table 1, Fig. 2A-C). In addition, the mean nodule size of benign nodules was larger than that of PTC (P < 0.001). The proportions of diabetes and hypertension in PTC patients were lower than those of patients with benign nodules (P = 0.03 and 0.01). Comparisons of the characteristics of PTC patients who underwent cervical lymph node dissection by medians of thyroid hormone sensitivity indices The characteristics of PTC patients according to the median of TFQI were shown in Table 2. Those with higher TFQI levels were more likely to be younger (P < 0.001) and had higher proportion of male (P = 0.02) than the patients with lower TFQI levels. The proportions of LNM and central LNM were higher (P = 0.001 and P = 0.002) and the tumor size was larger (P = 0.01) in the group with TFQI > 0.001 comparing to the group with TFQI ≤ 0.001. The number of metastatic cervical and central lymph node and ratio of central LNM were also higher in patients with higher TFQI level than those with lower TFQI levels (p < 0.001, p = 0.003 and P = 0.001). In addition, FT 3 , FT 4 , TSH, TSHI and TT4RI levels were higher in the group with higher TFQI (P < 0.001).  Table 3 showed the characteristics of patients with PTC by median of TSHI. Compared with patients with TSHI ≤ 2.745, the age was younger (P = 0.01), the tumor size was larger(P = 0.03), and the proportions of LNM (P = 0.01) and central LNM (P = 0.02) were higher in those with TSHI > 2.745. The number of metastatic cervical and central lymph node and ratio of central LNM were also higher in the group with higher TSHI level than the group with lower TSHI level (P = 0.002, 0.02 and 0.02). In addition, there were signi cant differences in FT 3 , FT 4 , TSH, TFQI and TT4RI levels between the two groups (P < 0.001). The characteristics of PTC patients according to the median of TT4RI were shown in Table 4. Compared with patients with TT4RI ≤ 28.39, the age, gender and tumor size did not differ signi cantly from those with TT4RI > 28. 39. The patients with higher TT4RI had higher proportions of cervical LNM and central LNM (P = 0.01 and P = 0.02). The number of metastatic cervical lymph node and ratio of central LNM were also higher in the group with higher TT4RI level than the group with lower TT4RI level (P = 0.03and 0.03). In addition, there were signi cant differences in FT 4 , TSH, TFQI and TT4RI levels between the two groups (P < 0.001).

Association of thyroid hormone sensitivity with PTC
The results of multiple logistic regression analysis for the association of thyroid hormone sensitivity indices with PTC were shown in Fig. 3 Association of thyroid hormone sensitivity with cervical LNM PTC patients with cervical LNM had signi cantly higher levels of TT4RI, TSHI and TFQI, which indicated worse thyroid hormone sensitivity (P = 0.01, P = 0.002 and P = 0.002, Fig. 4A-C). The results of the multiple regression analyses for the association of thyroid hormone sensitivity indices with cervical LNM in PTC patients were shown in Fig. 5 for each model in TSHI and TT4RI analysis were consistent with the above analysis, and the results were similar. Higher levels of TSHI and TT4RI were signi cantly associated with cervical LNM in the crude and adjusted logistic models.

Discussion
To the best of our knowledge, it is the rst study that investigated the association of thyroid hormone sensitivity with PTC. We found that higher TFQI, TSHI and TT4RI levels, which represented reduced sensitivity to thyroid hormone, were associated with an increased risk of PTC. In addition, impaired thyroid hormone sensitivity was associated with a higher risk of cervical LNM in PTC patients.
All previous studies analyzed the association between thyroid function and PTC using a single index, and the results were inconsistent. The majority of these studies reported an increased risk of thyroid cancer with higher TSH concentrations within the normal ranges (10,11,13). An meta-analysis of 28 studies suggested that an increased thyroid cancer risk was related to high TSH concentrations at normal and subnormal levels (7). However, two other studies observed an inverse association between the risk of PTC  (30). Furthermore, some genetic study observed an inverse association between TSH levels and thyroid cancer. A GWAS meta-analysis of 22.4 million genetic markers in 119,715 individuals showed that genetically predicted increased TSH levels are associated with a reduced risk of thyroid cancer (31). A two-sample Mendelian randomization study using TSH index variants as instrumental variables also showed that lower TSH causes an increased risk of thyroid cancer (32). Inconsistency in previous studies reveals that TSH or thyroid hormone levels alone may be insu cient to explain the relationship between thyroid homeostasis and PTC progression.
Resistance to thyroid hormone (RTH) is an inherited syndrome de ned by impaired responsiveness to thyroid hormones and mostly caused by mutations of the thyroid hormone receptor beta (TRβ) gene (33).
Recently, some researchers hypothesized that acquired resistance to thyroid hormone might exist in general populations as a more common disease and not just as an inherited rare disorder(18, 34).
Therefore, new indices derived from thyroid hormones and TSH were calculated for estimating the sensitivity of pituitary to thyroid hormones, which was named as TT4RI and TSHI (16,17). In 2019, Laclaustra et al. reported that TFQI was better in assessing sensitivity to thyroid hormones than TSHI and TT4RI(18). They also proposed that all these resistance to thyroid hormone indices measure central sensitivity, representing the degree of pituitary gland inhibition by FT 4 (18). Generally, central resistance may be related to a general decrease in thyroid hormone sensitivity, not only centrally but also in the periphery, which means peripheral resistance could also be present among participants with higher values of TFQI, TT4RI and TSHI(18).
There exists complicated and feedback relationship between TSH and thyroid hormone, which was regulated via the negative feedback loop of the HPT axis. In addition, the tissue enrichment analyses of genes in a GWAS meta-analysis revealed that TSH-lowering alleles were generally associated with higher free thyroxine levels (35). Due to the complex interactions between TSH and thyroid hormone, using a single index may cause the inconsistent results of the relationship between thyroid hormone or TSH levels and thyroid cancer risk in previous studies. Indices of sensitivity to thyroid hormone are composite indices that re ect the regulation of thyroid hormone homeostasis, and they are more su cient to explain the relationship between thyroid function and diseases than a single index.
In this study, we investigated the correlation between sensitivity to thyroid hormones and PTC for the rst time. Patients with PTC had a signi cantly higher levels of TFQI, TSHI and TT4RI compared to the   patients with benign nodules. Those with higher TFQI, TSHI and TT4RI levels were more likely to have  larger tumor size and lymph node metastasis than the patients with lower TFQI, TSHI and TT4RI levels. Moreover, the results of multiple logistic regression showed that reduced thyroid hormone sensitivity was related to an increased risk of PTC after adjusting for multiple risk factors. Furthermore, PTC patients with decreased sensitivity to thyroid hormones had higher risk of cervical LNM. Our results suggested that impaired sensitivity to thyroid hormone is associated with the development and progression of PTC.
Our study provided evidence that using the merged indices of thyroid homeostasis including TFQI, TSHI and TT4RI may be superior to the single indices of thyroid function, which should be validated in the future research.
The mechanisms about how impaired thyroid hormone sensitivity affect the development and progression of PTC were unclear. We hypothesized that there may be the following mechanisms. First, patients with reduced sensitivity to thyroid hormone had relatively higher TSH and thyroid hormone levels. The elevated TSH and thyroid hormones may promote tumor progression through pathways regulating both tumor growth, tumor evasion and tumor-related angiogenesis(38-40). Secondly, as thyroid hormone resistance is associated with Metabolic Syndrome and its components(18-21, 26), and insulin resistance may play a role in the development of thyroid cancer (41)(42)(43), insulin resistance may partially explain the relationship between impaired sensitivity to thyroid hormones and the development and progression of PTC.
There were some limitations in our study. First, our study was a retrospective study included a relatively small number of patients, which was insu cient to assess the causal relationship between thyroid hormone sensitivity and PTC. Large prospective studies from different areas and different populations are required to con rm our results. Secondly, we demonstrated that impaired sensitivity to thyroid hormone is associated with the risk of PTC and aggressiveness of PTC, but the exact mechanism remains unclear. Further investigations are necessary to understand its underlying mechanism.
In conclusion, the present study showed that impaired sensitivity to thyroid hormones was associated with an increased risk of PTC, and it is also associated with a higher risk of cervical LNM in PTC patients.
Thyroid hormone sensitivity indices may be new valuable indicators for predicting the development and progression of PTC, which need to be veri ed in large prospective cohort studies in the future.

Declarations
Funding This work was supported by the National Natural Science Foundation of China (grant no.: 81972497).
Competing Interests Figure 1 Flowchart of participant recruitment.

Figure 2
Comparison of thyroid hormone sensitivity indices between PTC patients and patients with benign nodules.
Scatter diagrams A-Cshow the comparisons of TT4RI, TSHI, and TFQI between PTC patients and patients with benign nodules respectively. Thick black lines represent the mean value and thinner black lines represent the standard deviation for TSHI and TFQI. Thick black lines represent the median and thinner black lines represent the lower and upper quartiles for TT4RI.

Figure 3
Page 24/25 Comparison of thyroid hormone sensitivity indices between PTC patients with and without cervical LNM.
Scatter diagrams A-Cshow the comparisons of TT4RI, TSHI, and TFQI in patients with and without cervical LNM respectively. Thick black lines represent the mean value and thinner black lines represent the standard deviation forTSHI and TFQI. Thick black lines represent the median and thinner black lines represent the lower and upper quartiles for TT4RI.