The relationship between papillary thyroid carcinoma and preoperative TSH level: A cross-sectional study from Syria

Abstract

Papillary thyroid carcinoma (PTC) is the most common type of thyroid cancer, and thyroid stimulating hormone (TSH) is the major growth factor for thyroid cells. It is also an available, inexpensive test and performed routinely while evaluating thyroid nodules. Yet the relationship between TSH levels and PTC is still controversial. Understanding the relationship between preoperative TSH levels and thyroid cancer helps to break new ground of current prevention, diagnosis, and management approaches of thyroid cancer. A cross-sectional retrospective study of patients underwent total thyroidectomy during 2019 at Al-Assad University Hospital, which included 305 individuals. All patients underwent thyroid ultrasonography and laboratory assessment of serum TSH levels prior to surgery, in addition to recording histological features of resected thyroid samples. The malignancy rate was 21.3%, PTC patients had higher TSH values across the entire study even when gender, age and number of thyroid nodules differed. A statistically significant increase in TSH levels was noticed by moving from the benign thyroid nodular disease (BTND) group to papillary thyroid microcarcinoma (PTMC) group, then to thyroid cancer of larger size (TCLS) group (P = .001). A statistically significant relationship was also found between high TSH levels and lymph node metastases (LNM) (P = .008). TSH concentrations were significantly higher in patients with PTC, and higher TSH values were associated with TCLS and LNM.

1. Introduction

Papillary thyroid carcinoma (PTC) is the most common type of thyroid cancer and has the best prognosis.^[1] The known risk factors for thyroid cancer are mostly non-modifiable including age, sex, race or ethnicity, exposure to ionizing radiation, and positive family history of thyroid cancer.^[1,2] The incidence of thyroid cancer has increased over the past several decades, almost entirely due to an increase in PTC, while thyroid-cancer-related mortality remained largely stable.^[1,3]

Fine-needle aspiration biopsy is the gold standard method for evaluating patients with thyroid nodules. Diagnostic results are obtained in most cases, but they cannot reliably rule out cancer in 20% to 30% of nodules.^[4]

From this standpoint, thyroid stimulating hormone (TSH) has been introduced as a possible predictor of thyroid malignancy in recent years for several reasons. Firstly, serum TSH level is the first laboratory test that is routinely performed when thyroid nodules are detected, and at the same time it is an available and inexpensive test.^[5,6] Secondly, data on other thyroid cancer risk factors and early diagnosis strategies are still insufficient and sometimes expensive and their impact on patient management is still debated.^[4,7] Thirdly, the improved survival rate and decreased recurrence rate in thyroid cancer patients treated with suppressive doses of levothyroxine supports TSH tropic effect on thyroid tissue promoting neoplasia and carcinogenesis along with oncogenes and other growth factors such as IGF1.^[7,8]

Albeit many previous studies confirmed an increased risk of malignancy in the thyroid nodules with higher TSH values, even at normal TSH levels, the relationship between TSH levels and PTC is still controversial.^[4,8,9]

Therefore, we aimed to evaluate the relationship between preoperative serum TSH and PTC. Also, we tried to investigate the relationship between preoperative TSH and clinicopathological features and variants of PTC.

2. Material and methods

We retrospectively collected data from 607 patients who underwent thyroid surgery at Al-Assad University Hospital, Damascus, Syria, between January and December 2019. The study protocol was approved by Research Ethics Committee of Damascus University and in accordance with the ethical standards of the Helsinki declaration.

Three hundred 5 patients who underwent total thyroidectomy with available preoperative serum TSH concentration and had either PTC or benign thyroid nodular disease (BTND) histopathological diagnosis were eligible for the study and their demographic and pathologic data were recorded. Patients with unilateral resection, patients with a final histological diagnosis other than PTC (e.g., Follicular thyroid cancer, medullary thyroid cancer, or anaplastic thyroid cancer), and patients treated with medications that affect thyroid function (e.g., antithyroid therapy or thyroid hormone replacement, corticosteroids and Amiodarone) were excluded as well as patients with a severe systemic disease or pregnancy. Data collected in each patient were: gender, age at diagnosis, thyroid nodules number, preoperative serum TSH, and histological diagnosis.

Age was evaluated as a continuous and categorical (<55 years, ≥55 years) variable. Preoperative thyroid ultrasonography report confirmed the presence of either solitary nodule (defined as 1 nodule within the thyroid gland) or multi-thyroid nodule (defined as 2 or more nodules within the thyroid gland) in all patients.

A biochemical evaluation of all patients through measurement of TSH concentrations, in keeping with the American Thyroid Association guidance, was performed. TSH levels were checked within a week preceding surgery for most patients and did not exceed 3 months period for the rest. Most studies report that the lower TSH limit (2.5th percentile) lies between 0.2 and 0.4 mIU/L, but upper limits (97.5th percentile) vary between 2.4 and 4.2 mIU/L as related to ethnicity or geographic location.^[10] The mean of normal TSH values is only between 1.18 and 1.40 mU/L and more than 95% of the normal population will have a TSH level <2.5 mU/L.^[11]

An automated electrochemiluminescence immunoassay (Cobas 6000 analyzer, series e601, Roche Diagnostic) was used for TSH measurement with 0.27 to 4.2 mIU/L as the normal reference range. A TSH of <0.1 mIU/L was considered maximally suppressed and a TSH of 10 mIU/L or greater was considered overt hypothyroidism. A morning blood sample from a vein had been used for TSH test and evaluated as a continuous variable and categorically within the following 3 groups: subclinical hyperthyroidism (TSH < 0.27 mIU/L [group 1]), euthyroid group (TSH 0.27–4.2 mIU/L [group 2]), and subclinical hypothyroidism (TSH > 4.2 mIU/L [group 3]).^[10,11]

The cytological results were classified according to the criteria of the Bethesda system for cytological classification of thyroid nodules into 6 categories.^[12] Histopathologically, Patients were classified as either benign or malignant. The final histological diagnosis was mentioned as follows: non-diagnostic or unsatisfactory, Benign, AUS/FLUS; Atypia of Undetermined Significance/Follicular Lesion of Undetermined Significance, FN/SFN; follicular neoplasm/suspected follicular neoplasm, Suspicious for Malignancy, and Malignant.

In patients with PTC; primary tumor size, the number of tumor foci, and histological variants as the presence of cervical lymph node metastases (LNM), extrathyroidal extension, lymphovascular invasion, capsular invasion, and chronic lymphocytic thyroiditis were recorded. Tumor multifocality was defined as 2 or more tumor foci within the thyroid gland. A tumor that is measuring 1 cm or less in diameter was classified as PTMC meanwhile tumor larger than 1 cm was considered thyroid cancer of larger size (TCLS). The histological variant of PTC was available for 52 patients, classified as nonaggressive (classical/conventional and follicular variants), and aggressive (tall cell, diffuse sclerosing, and columnar variants).

2.1. Statistical analysis

All analyses were performed using SPSS statistical software (version 25). Microsoft Office Excel 2016 program was used to prepare graphics. Distribution of age, tumor size and TSH were determined by Shapiro–Wilk test and normality graphics. TSH and tumor size were defined as median (range) since they were not normally distributed while age was normally distributed and defined as mean ± SD. Categorical variables were presented as percentage.

Mann–Whitney U test was used to compare TSH, in the pairwise comparisons of groups. Kruskall–Wallis test was used to determinate whether there were significant differences between patients with benign, PTMCs, and PTC. P value < .05 was considered statistically significant.

3. Results

Patient characteristics are presented in Table 1. Of 305 patients, 242 (79.3%) were females, 41 (13.4%) had diabetes mellitus, and 64 (21%) hypertension. The mean of age of the participants was 47.79 ± 12.46 years, and 213 (69.8%) of patients aged younger than 55 years. Mean serum TSH was 1.2 mIU/L, and the majority of patients (92.5 %) showed serum TSH levels within the normal range. Multiple nodules were noted in 254 patients (83.3%). Final pathology report showed no evidence of malignancy in 240 patients (78.7%), whereas malignant lesions were present in 65 cases (21.3%).

Table 1.

Demographic characteristics of the total sample.

Variables		Total sample (N = 305)
		Number	Percent
Gender	Female	242	79.3
	Male	63	20.7
	Total	305	100
Age	<55 yr	213	69.8
	≥55 yr	92	30.2
	Total	305	100
DM	No	264	86.6
	Yes	41	13.4
	Total	305	100
HTN	No	241	79
	Yes	64	21
	Total	305	100
TSH categories (mIU/L)	0.1–0.26	20	6.6
	0.27–4.19	282	92.5
	4.2–10	3	1
	Total	305	100
Number of nodules in US	Solitary	51	16.7
	Multiple	254	83.3
	Total	305	100
Histopathology	Benign	240	78.7
	Malignant	65	21.3
	Total	305	100
FNA	Non-diagnostic or unsatisfactory	13	11.7
	Benign	40	36
	AUS/FLUS	32	28.8
	FN/SFN	11	9.9
	Suspicious for malignancy	14	12.6
	Malignant	1	0.9
	Total	111	100

AUS/FLUS = atypia of undetermined significance/follicular lesion of undetermined significance, DM = diabetes mellitus, FN/SFN = follicular neoplasm/suspected follicular neoplasm, FNA = fine needle aspiration, HTN = hypertension, TSH = thyroid stimulating hormone, US = ultrasound.

Cytological diagnosis was available for 111 patients, and the results were as follows: 11.7% (n = 13) unsatisfactory, 36% (n = 40) benign, 28.8% (n = 32) Atypia of Undetermined Significance/Follicular Lesion of Undetermined Significance (AUS/FLUS), 9.9% (n = 11) follicular neoplasm/suspected follicular neoplasm (FN/SFN), 12.7% (n = 14) suspicious for malignancy, and 0.9% (n = 1) malignant (Table 1).

Table 2 shows the following; preoperative serum TSH level was higher in males, individuals aged <55 years, and individuals with solitary thyroid nodule. However, only age differed noteworthy among the evaluated variables (P = .000). Median TSH was also significantly higher in PTC patients compared to benign ones (1.41 mIU/L [0.13–8.01] vs 1.13 mIU/L [0.11–3.6], P < .001). When PTC group was subdivided according to tumor size; patients with TCLS had higher TSH than PTMC patients (1.52 mIU/L [0.13–8.01] vs 1.3 mIU/L [0.45–3.84]), and patients with benign lesions had the lowest TSH levels (1.13 mIU/L [0.11–3.6]). A remarkable statistical significance was noticed when the 3 groups were compared to each other (P < .001). Pairwise comparisons between groups showed a statistically significant difference only between benign and TCLS groups (1.13 mIU/L [0.11–3.6] vs 1.52 mIU/L [0.13–8.01], P = .000).

Table 2.

Comparing serum TSH levels according to study variants (Total sample: N = 305).

			N	Median	Rang	P value
Histopathological diagnosis	Benign		240	1.13	0.11–3.6	.001
	Malignant		65	1.48	0.13–8.01
	Benign		240	1.13	0.11–3.6	.001
	Malignant	PTMC	27	1.3	0.45–3.84
		TCLS	38	1.52	0.13–8.01
	Pairwise comparisons	Benign	240	1.13	0.11–3.6	.212
		PTMC	27	1.3	0.45–3.84
		Benign	240	1.13	0.11–3.6	.000
		TCLS	38	1.52	0.13–8.01
Gender	Male		63	1.32	0.11–5.52	.254
	Female		242	1.18	0.12–8.01
	Male	Benign	49	1.2	0.11–2.54	.038
		Malignant	14	1.61	0.51–5.52
	Female	Benign	191	1.13	0.12–3.6	.006
		Malignant	51	1.46	0.13–8.01
Age	<55 yr		213	1.35	0.11–8.01	.000
	≥55 yr		92	0.83	0.11–3.5
	<55 yr	Benign	160	1.26	0.11–3.6	.003
		Malignant	53	1.5	0.13–8.01
	≥55 yr	Benign	80	0.83	0.11–2.97	.509
		Malignant	12	0.86	0.27–3.5
No. of nodules	Solitary		51	1.24	0.21–8.01	.162
	Multiple		254	1.19	0.11–5.86
	Solitary	Benign	42	1.18	0.21–3.04	.529
		Malignant	9	1.7	0.47–8.01
	Multiple	Benign	198	1.13	0.11–3.6	.001
		Malignant	56	1.47	0.130–5.86

PTMC = papillary thyroid microcarcinoma (≤1cm), TCLS = thyroid cancer of larger size (>1cm), TSH = thyroid stimulating hormone.

The whole study group was subdivided by gender, age and number of thyroid nodules; then TSH levels were analyzed according to final pathology diagnosis (PTC vs benign). PTC patients exhibited higher median TSH levels compared to those with benign lesions in both males and females’ groups (P = .038 and .006, respectively). Interestingly, TSH level was again higher again in PTC patients irrespectively of age group and number of thyroid nodules. Of note that TSH difference between PTC versus benign subgroup was statistically significant only in those aged <55 years and who had multiple thyroid nodules (P = .003 and .001, respectively).

In PTC group (Table 3), studying preoperative serum TSH levels according to clinical features (gender, age and number of thyroid nodules), revealed that it was significantly higher in patients aged < 55 years old compared to those aged ≥ 55 (1.5 mIU/L [0.13–8.01] vs 0.86 mIU/L [0.27–3.5], P = .032). When pathological features were evaluated, TSH level was higher in TCLS comparing to PTMC, but the difference between the groups was not statistically significant. Likewise, preoperative serum TSH level was higher in patients presented with extrathyroidal invasion, lymphovascular invasion, capsular invasion, LNM, and chronic thyroiditis compared to those without, but the difference was statistically significant only in patients with LNM compared to those without LNM (2.3 mIU/L [1.07–5.52] vs 1.44 mIU/L [0.13–8.01], P = .008). There was not any difference in serum TSH level in patients with unifocal versus multifocal tumor as well as patients with aggressive variant PTC verses nonaggressive variants.

Table 3.

Comparing serum TSH levels according to the clinicopathological features in PTC patients (N = 65).

feature	N	Median	Rang	P value
Age
<55 yr	53	1.5	0.13–8.01	.032*
≥55 yr	12	0.86	0.27–3.5
Gender
Male	14	1.61	0.13–8.01	.271
Female	51	1.46	0.5–5.52
No. of nodules on ultrasound
Solitary	9	1.7	0.47–8.01	.805
Multiple	56	1.47	0.13–5.86
Tumor size
<1 cm	27	1.3	0.45–3.84	.110
>1 cm	38	1.52	0.13–8.01
Histological variant (N = 52)
Aggressive	3	1.53	1.48–5.52	.475
Non-aggressive	49	1.55	0.13–8.01
Tumor multifocality
Present	38	1.48	0.13–8.01	.989
Absent	27	1.43	0.27–5.86
Lymphovascular invasion
Present	28	1.64	0.13–8.01	.128
Absent	37	1.35	0.45–5.86
Extrathyroidal invasion
Present	12	1.56	0.7–5.52	.310
Absent	53	1.48	0.13–8.01
Capsular invasion
Present	18	1.46	0.5–5.86	.681
Absent	47	1.49	0.13–8.01
LNM
Present	12	2.3	1.07–5.52	.008
Absent	53	1.44	0.13–8.01
Chronic thyroiditis
Present	12	1.7	0.13–5.52	.543
Absent	53	1.48	0.27–8.01

LNM = lymph node metastases, PTC = papillary thyroid carcinoma, TSH = thyroid stimulating hormone.

4. Discussion

It is well established that thyroid proliferation and differentiation as well as thyroid carcinogenesis depend on a precise regulation process, in particular hormonal influence (TSH, estrogen), environmental influence (iodine status), and other molecules secreted in situ (such as vascular Endothelial Growth Factor and Insulin-like growth factor-1). In addition, some genetic mutations appear to be closely associated with the risk of differentiated thyroid cancer, such as BRAF mutations.^[8,13,14]

In recent years, several studies have discussed the relationship between TSH and thyroid cancers, however, the pathogenesis is still in debate. It is not well-defined whether TSH is a carcinoma initiator; which has a fundamental role in the generation of thyroid cancer, or it might also be involved in the promotion and progression of thyroid tumor growth and aggressiveness. On the other hand, data on autoimmune thyroid disease and thyroid cancer support the role of TSH receptor activation in tumorigenesis.^[8,14,15]

In this retrospective cross-sectional study of Syrian patients, the mean age was 47.7 ± 12.46 years with a female predominance (79.3%), which consistent with the fact that thyroid nodules and tumors are more common in middle aged patients and females.^[1,2] Additionally, 69.8% of patients aged <55 years, this observation perhaps because this age group is more likely to undergo surgery compared to older patients whose comorbidities hinder anesthesia and surgery.

Similar to other studies,^[4,16] our study showed that 83.3% (254 patients) of participants had multiple thyroid nodules compared to patients with single thyroid nodule, whom comprised 16.7% (51 patients) of the all sample. This observation explained by variety of reasons for thyroidectomy in patients with multiple thyroid nodules; including not only related to suspicion for malignancy but also by benign causes such as a large thyroid gland and compressive symptoms. On the other hand, suspected malignancy is usually the most common reason for patients with single thyroid nodule to undergo surgery.

Previous studies reported a malignancy rate between 20% and 44%.^{[4,8,9,16,17]} However, the prevalence of malignancy in our study was (21.3%) and corresponded with previous reports, these results do not indicate the true prevalence of malignancy in thyroid nodules which ranges between 5% and 10%.^[1,18] This high rate of malignancy in the current and previous studies could explained by the fact that study samples were a carefully chosen group from patients with BTND who underwent thyroidectomy for benign or most importantly for malignant suspicion.

Our study confirmed the relationship between TSH levels and final pathological outcome where the median TSH was significantly higher in PTC patients compared to benign ones (P = .001) in agreement with previous studies.^{[4,8,9,16,19]} We have also confirmed a higher preoperative serum TSH in patients with malignancy, where the lowest level was in BTND (1.13 mIU/L), higher in PTMC (1.3 mIU/L), and highest in TCLS (1.52 mIU/L).

Similar to studies by Carles Zafon et al,^[16] He et al,^[17] and Tam et al,^[9] our study confirmed that these differences have persisted when comparing all 3 groups (benign, PTMC, and TCLS) (P = .001). However, significant differences showed when comparing between benign lesions and TCLS (P = .000), there was no difference when comparing between PTMC and TCLS (P = .212). This supports the hypothesis that TSH is likely a carcinoma propagator and promotes thyroid cancer growth and aggressiveness. Furthermore, Haymart et al pointed out that TSH stimulation may cause occult microcarcinomas to grow to a detectable size.^[20] Another series of 11,919 patients recorded that the probability of PTC increases by about 11% for each mIU/L increase of TSH.^[21]

Although age, gender, and thyroid nodules are established known risk factors for thyroid cancer but there is still an ongoing debate on the relationship between these risk factors, cancer, and TSH.^[1,2] Data from 2 studies by Haymart et al^[8,20] demonstrated that gender significantly affects TSH levels between benign and malignant lesions. Although, other studies by Demircioglu et al^[22] and Golbert et al^[4] did not show a gender effect on TSH levels.

Another contradictory issue is whether age significantly affects TSH levels between benign and malignant lesions. Multiple population-based studies, including the National Cancer Database Report and the Surveillance, Epidemiology, and End Results, have shown an important prognostic indicator of age for thyroid cancer.^[23,24] Two previous studies attempted to study TSH values in BTND and PTC patients. In the first study by Haymart et al there was a trend of rising mean TSH with age. It also showed that mean TSH was significantly higher in cancer patients regardless of age (<45 years or ≥45 years).^[20] Meanwhile, serum TSH levels decreased progressively with age in the second study, but again it was significantly higher in PTC than in BTND in all age groups.^[19]

Here, we found a significant association between age and TSH levels in the whole sample as well as the malignant group (P = .000 and .032, respectively), where higher TSH levels were observed in patients younger than 55 years. Furthermore, by comparing benign and malignant lesions, only subgroup analysis of patients younger than 55 years showed a significant difference in serum TSH (P = .000). Although TSH values were higher in males and individuals with solitary thyroid nodules in both total and malignant samples in our study, these differences were not associated with statistical significance. Moreover, a subgroup analysis comparing TSH levels between benign and malignant lesions according to gender (Table 2) revealed that PTC patients have remarkably higher TSH values in both male and females. Also, the same comparison (benign vs malignant) showed significant difference (P = .00) only in multiple nodules group (Table 2) and revealed remarkably higher TSH values in PTC patients compared to benign patients (1.47 vs 1.13).

Worth to be mentioned here, it is well documented that in iodine deficient areas, longstanding iodine deficiency causes higher frequency of thyroid nodularity and autonomy in older people. Based on these data Fiore et al suggested that higher levels of TSH in PTC are not due to an increase of thyrotropin in patients with thyroid cancer, but are mainly related to the reduction of serum TSH in patients with nodular goiter.^[19] Unfortunately, a radionuclide thyroid scan was not performed to confirm thyroid autonomy in these patients.

Many studies have confirmed the important relationship between TSH and aggressive features of PTC like tumor stage, size, and LNM. Haymart et al^[8] showed for the first time that higher TSH is associated with not only the incidence of differentiated thyroid cancer but also with advanced stage of tumors. These results were confirmed in subsequent studies.^[19,25,26] Conversely, other studies have failed to show a significant effect of serum TSH concentration on tumor stages.^[9,17,27] Another 2 meta-analyses including 28 studies and 56 studies showed that high TSH was significantly related to tumor size and invasion even at normal and subnormal levels.^[26,28] Tam et al^[9] revealed for the first time that patients with aggressive variant PTC had higher serum TSH than nonaggressive ones. Furthermore, in this study serum TSH was higher in patients with bilateral tumor, capsular invasion, and LNM than in those without.^[9] Demircioglu et al and Besler et al had similar results.^[22,29] Recently, Kim et al^[30] have shown that preoperative TSH levels are a considerable predictive factor for the presence of extrathyroidal spread and lateral LNM in patients with thyroid cancer.

Similar to the current literature, our findings demonstrated a significant increment in TSH levels for the presence of cervical LNM (P = .008) but we did not find an association between serum TSH concentration and tumor size, the present of Lymphovascular invasion, extrathyroidal invasion, and capsular invasion (Table 3).

Some limitations should be pointed out in this study such as being a retrospective study performed in a single center. Since we included patients with nodular thyroid disease who underwent surgical treatment, the fact that our patient sample is highly selected and therefore potential selection bias is possible.

On the other hand, since patients with L-thyroxine (LT4) treatment were excluded, TSH levels in this group had not been evaluated. In this regard, the treatment with LT4 to reduce thyroid nodule size and/or growth is controversial.^[15] A large series of 27,914 patients by Fiore et al^[31] found a significantly lower frequency of PTC in LT4-treated patients (3.2%) than in untreated patients (5.1%). However, considering the doubts about its effectiveness in patients with nodular goiter, the American Thyroid Association strongly recommends against the use of LT4.^[32] Despite our results and several data that confirmed the relationship between TSH and thyroid cancer, future studies should focus on the treatment with LT4 to reduce thyroid nodule size and/or growth.

In conclusion, patients with PTC had significantly higher TSH levels than BTND patients. Furthermore, they had higher TSH levels across the entire study irrespective of sex, age group, and the number of thyroid nodules. We also observed a statistically significant relationship between high TSH levels and LNM. According to these results, we suggest that TSH is involved in PTC progression, therefore it should be considered alongside with other clinical, radiographic and cytological features while evaluating the risk of PTC in patients with nodular thyroid diseases.

Acknowledgments

We would thank Prof Zaynab Alourfi for reviewing the last copy of the article and for important advices for literature writing.

Author contributions

Conceptualization: Marjan Shahrokh.

Data curation: Marjan Shahrokh.

Methodology: Marjan Shahrokh.

Supervision: Younes Kabalan.

Writing – original draft: Marjan Shahrokh, Mohammad Alsultan.

Writing – review & editing: Marjan Shahrokh, Mohammad Alsultan, Younes Kabalan.