Background

Thyroid cancer is the most frequent endocrine malignancy, accounting for approximately 1% of all malignant tumors in the United States [1]. However, a recent study [2] by the Office of Cancer Prevention and Treatment indicates that thyroid cancer represents 5.90% of all new malignant diseases in one district of Wenzhou (a coastal city in Southeast China, with a referral population of 7,558,000). The detection rate of thyroid nodules is higher than before because of high-resolution ultrasound [3]. In order to identify patients with thyroid cancer who require early intervention, many predisposing factors for malignancy have been recognized, including young age (<20 years old) or older age (>70 years old), male gender, large (>4 cm) or rapidly growing nodules, and history of radiation exposure, calcification and thyroglobulin (Tg) [4–7]. The role of thyroid-stimulating hormone (TSH) in the growth and development of thyroid cancer has long been known [8]. Recently, Boelaert et al. [9] proposed that the TSH level at presentation is a novel predictor of malignancy in patients with thyroid nodules. Another report indicated that higher TSH levels were associated with greater risks of differentiated thyroid cancer and advanced tumor stage [10]. The findings above need to be verified further.

Hashimoto’s thyroiditis (HT) is an autoimmune disease characterized by widespread lymphocyte infiltration, fibrosis, and parenchymal atrophy [11]. Historically, HT was thought to be associated with malignant lymphoma. However, Dailey et al. in 1955 first reported an increased association between HT and papillary thyroid cancer (PTC) [12]. The coexistence of these two diseases has been variously reported to range from 0.5% to 38.0% [11, 13–18]. A few studies have proposed HT as a risk factor for thyroid cancer [12, 16, 18, 19], whereas others have reported a negative correlation between the two diseases [10, 13, 15]. Nowadays, the causal association between these diseases remains controversial, and few papers have discussed whether HT influences the clinical and pathologic features of thyroid cancer.

Therefore, the current study investigates which clinical features or biochemical criteria predict the likelihood of thyroid malignancy in patients with thyroid nodules, thereby identifying those at greatest risk of harboring thyroid cancer. Other objectives of the current study are to determine the following: 1) the relative incidence of HT and PTC among patients undergoing thyroidectomy, and 2) the possible influence of HT on the clinical behavior and neoplastic aggressiveness of PTC.

Methods

A total of 2,052 patients (1,645 female and 407 male) with thyroid nodules, in whom malignancy was suspected under fine-needle aspiration (FNA) or ultrasonography, or who had rapidly growing nodules or nodules fixed to adjacent structures, and who underwent initial thyroidectomy in the First Affiliated Hospital of Wenzhou Medical College, were retrospectively recruited from June 2006 to August 2008. However, some female patients underwent surgery because of cosmetic problems or anxiety about the effects on their quality of life. The majority of the patients received treatment for thyroid cancer according to the accepted protocol, which involves total or near-total thyroidectomy, followed by thyroid hormone suppression. Cervical lymph node dissection or sentinel lymph node biopsy was performed routinely during thyroid cancer’s operation. Others underwent resection of the whole or partial lobes of the thyroid gland.

The diagnosis was confirmed either on reevaluation of histopathology sections when available, or on reviewing the previous pathology reports. The histologic type of thyroid cancer was assessed by two senior pathologists according to the World Health Organization criteria and classified as PTC in 1004 patients (98.05%), including two with coexisting PTC and malignant lymphoma, follicular thyroid cancer (FTC) in four patients (0.39%), medullary thyroid cancer in twelve patients (1.17%), and anaplastic thyroid cancer in four patients (0.39%). Patients who were characterized by the presence of diffuse lymphocytic and plasma cell infiltration, oxyphilic cells, and lymphoid follicles with reactive germinal centers, were defined as having HT. Whereas patients with lymphocytic infiltration immediately surrounding the malignant tumor (but without other lobe involvement) were considered as having thyroid cancer alone because the surrounding inflammation may represent an immune response to the host tumor [20].

In addition to the demographic data (age and gender), data on the thyroid-associated parameters [serum free T4 (fT4), free T3 (fT3), TSH, Tg, and thyroglobulin antibody (TgAb), antimicrosomal antibody (TMAb)], tumor-associated status (tumor size, number and region of positive lymph nodes, and metastasis (TNM) stage), and thyroid disorders (presence or absence of HT, Graves disease, follicular adenoma, or malignant lymphoma) were obtained from a review of the medical records from the Patient Information Inquiry System. The normal ranges of serum fT3, fT4, TSH, Tg, TgAb, and TMAb were 3.67 to 10.43 pmol/L, 7.5 to 21.1 pmol/L, 0.35 to 4.94 mIU/L, 4.14 to 14.46 ng/ml, 0.0% to 30.0%, and 0.0% to 20.0%, respectively. We had free thyroid hormones, TSH, Tg, TgAb, and TMAb measured simultaneously with surgery. However, the measurement results of patients known to have taken thyroid medication were excluded. This research was approved by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical College.

The final diagnostic outcome was defined as the presence or absence of thyroid cancer. Data were analyzed using Statistical Package for Social Science (SPSS) version 13.0. The odds ratio (OR) in multivariate logistic regression analysis, with the relative 95% CI was calculated to assess the relevance of thyroid-associated clinical or biochemical conditions to predict the final diagnostic outcome. Confounding factors, including age, gender, and thyroid disorders, were investigated statistically. Comparisons of frequency distributions and numerical variable data were performed using the chi-square (χ2) test. Observed differences were considered statistically significant if the probability of chance occurrence was less than 0.05.

Results

The clinical and biochemical features of patients with and without thyroid cancer were compared. Of the 1,024 patients with thyroid cancer, female patients (n = 843) accounted for the majority (female/male ratio = 4.66/1.00). The age distribution (range 10 to 84 years, mean 45.0 ± SD 12.1 years) revealed a predominance in the 30 to 44-year-old age group (n = 449 patients, 43.8%) with thyroid cancer. However, the high tumor incidence in the 45 to 60-year-old age group (n = 370 patients, 36.1%) should not be neglected. Comparison between the two groups highlighted a statistically significant difference in gender (χ2 = 5.990, P = 0.014), age (χ2 = 13.588, P = 0.018) (Table 1). To avoid the interference of several extreme values, fT3, fT4, TSH, Tg, TgAb, and TMAb were treated as categorical variables in the analysis. Significant differences were detected when the following values were compared: TSH (χ2 = 20.160, P <0.001), TgAb (χ2 = 14.142, P <0.001) and TMAb (χ2 = 15.026, P <0.001) but not fT3 (χ2 = 4.784, P = 0.306), fT4 (χ2 = 4.548, P = 0.337), and Tg (χ2 = 3.431, P = 0.180) among patients with and without thyroid cancer (Table 2). Moreover, the presence of HT was associated with thyroid cancer (χ2 = 67.421, P <0.001). Regarding calcification of the thyroid nodules, the incidence of microcalcification was higher in patients with thyroid cancer than in the controls (71.4% vs 16.2%, χ2 = 633.792, P <0.001). However, the prevalence of thyroid cancer was similar in patients with solitary nodules and patients with multiple nodules (χ2 = 1.456, P = 0.228) (Table 1).

Table 1 Comparison of characteristics of patients with and without thyroid cancer
Full size table
Table 2 Comparison of the biochemical features of patients with and without thyroid cancer
Full size table

Subsequently, a binary logistic regression analysis was performed to define the risk predictors for thyroid cancer, which simultaneously analyzed gender, age, serum fT3, fT4 and TSH concentration, HT, and calcification. In addition, considering the feedback inhibition regulation of TSH by serum fT3 and fT4 concentrations, the interaction of variables between TSH and/or fT3, and/or fT4 were also taken into account in the analysis. Finally, four variables, namely, calcification, HT, TSH, and age, were entered into the multivariate model via forward stepwise regression (likelihood ratio). An increased risk of thyroid cancer was detected among the patients with HT (OR = 3.732, 95%CI = 2.563–5.435), microcalcification (OR = 14.486, 95%CI = 11.374–18.449). There was significantly increased risk of malignancy in patients with serum TSH of 1.97 to 4.94 mIU/L, compared to patients with TSH below 0.35 mIU/L (OR = 1.951, 95% CI = 1.201–3.171, P = 0.007) (Table 3).

Table 3 Independent risk predictors for diagnosis of thyroid malignancy under multivariate logistic regression analysis (n = 1,789 patients)
Full size table

The patients suffering from PTC (n = 1004, 98.05%) were selected and divided into patients with HT (group I) and patients without HT (group II) based on the final histologic examination. Tables 4 and 5 show the clinical and pathologic features of the two groups; 18.63% of the patients with PTC had concurrent HT, which was more frequent than in the benign group (P <0.001). Moreover, the female patients constituted an overwhelming 97.3% (n = 182) of the patients with coexisting PTC and HT. The mean age at initial thyroidectomy was similar between the groups (χ2 = 7.298, P = 0.063), as well as thyroid-associated disorders (χ2 = 3.322, P = 0.325) (Table 4). Furthermore, the tumor size (χ2 = 2.975, P = 0.209), frequency of occult PTC (χ2 = 2.872, P = 0.090), extrathyroidal invasion (χ2 = 0.885, P = 0.347), capsular infiltration (χ2 = 0.891, P = 0.345), angioinvasion (χ2 = 0.429, P = 0.512), and lymph node metastases (χ2 = 0.227, P = 0.634) did not differ between patients with and without HT. No distant metastasis was observed in the two groups. Although the majority of positive lymph nodes (58.3%) in patients with HT were distributed in level VI (classification references [21]) and a higher proportion was observed in group I with stage I (85.6%), the difference was not significant compared with group II (χ2 = 4.703, P = 0.427; χ2 = 0.540, P = 0.910, respectively), in other words, the pathologic TNM (pTNM) classification in the two groups were similar (shown in Table 5).

Table 4 Clinical features of patients with papillary thyroid cancer by group in patients with and without Hashimoto’s thyroiditis (HT)
Full size table
Table 5 Pathologic features of patients with papillary thyroid cancer (PTC) by group in patients with and without Hashimoto’s thyroiditis (HT)
Full size table

Discussion

As shown in the current study, the peak incidence of thyroid cancer in Wenzhou occurs in the 30 to 44-year-old age group. Moreover, when the association between gender and thyroid cancer was assessed, significantly increased rates of malignancy were detected among the female patients (82.3%) (χ2 = 5.990, P = 0.014). However, further logistic regression analysis did not identify gender as an independent risk predictor. These results may challenge a previously recognized view of an increased risk of underlying malignancies in male patients [5, 9, 22, 23]. Furthermore, the patient age was demonstrated as a risk predictor for malignancy. Increased adjusted ORs for thyroid cancer were observed in the older age group (75 years) in accordance with previous reports [22, 24], although no significant difference was observed compared with the young age group (<15 years, P = 0.362).

The risk of thyroid cancer has been shown to increase with serum TSH concentrations, and even within normal ranges, higher TSH levels are associated with a higher incidence and more advanced stage of thyroid malignant tumor [9, 10, 24, 25]. Serum TSH has a tropic effect on thyroid nodules and suppression of TSH concentration by administration of exogenous thyroxine may interfere with the growth of established nodules, as well as the formation of new thyroid nodules [26–29]. TSH suppression is also associated with a decreased frequency of PTC [30].

We demonstrated that the risk of thyroid cancer increased in parallel with the serum TSH concentration within the normal range at presentation. This observation is supported by a previous study [24]. In our study, serum TSH concentration (>4.94 mIU/L) was not associated with the greatest risk of thyroid cancer, which is inconsistent with other studies [8, 9]. We have no vigorous explanation for this inconsistency. One reason may be that the sample size of the group with serum TSH concentrations >4.94 mIU/L was smaller than the other groups. In addition, the mechanism of TSH on thyroid disease may differ between populations. Furthermore, several arguments weaken the role of TSH in the development or progression of malignant thyroid tumors: 1) TSH receptor mutations in regions functionally associated with increased signal transduction do not often occur in malignant thyroid tumors [31]; 2) other growth factors such as insulin-like growth factor-I (IGF-I) have been demonstrated to have a more important role in the growth of thyroid cancer in in vitro studies [32, 33], and TSH needs cooperation with insulin/IGF-1 to educe its proliferative effects [34]; 3) Shi et al. [35] reported that the relationship between TSH receptor mRNA levels and the aggressiveness of cancer was inversed; 4) thyroid cancer occurs in patients with serum TSH concentrations suppressed by hyperfunctioning nodules in the contralateral lobe [36]; 5) a recent study showed that patients carrying one of two alleles associated with an increased risk of differentiated thyroid cancer have lower serum TSH concentrations [37].

Taken together, TSH is unlikely to act versatilely in cancer development. These findings indicate that serum TSH concentration can be used as an adjunct biochemical predictor of thyroid cancer. However, this requires further investigation.

In contrast to previous studies [7, 38] that reported that Tg levels were significantly higher in patients with thyroid cancer than in those with benign disorders, no significant difference in Tg concentration was observed in patients with malignancy. Limited to the current measurement, whether preoperative serum Tg measurement helps differentiate benign from malignant thyroid disorders needs further verification. Ultrasonography cannot reliably distinguish benign from malignant lesions [39]. The current study demonstrates similar incidences of malignancy between patients with solitary nodules and those with diffuse or multinodular lesions, as reported by ultrasonography, although others found that the presence of solitary nodules found by palpation is associated with increased malignancy rates [5, 9, 40, 41]. In addition, a higher risk of malignancy was detected in patients with microcalcification (OR = 14.486, P <0.001, Table 3), which is consistent with recent investigations in patients evaluated through ultrasonography [42, 43]. Nevertheless, approximately 13.5% of patients with thyroid cancer (n = 138) were confirmed to have coexisting coarse calcifications. Accordingly, ultrasonographic findings, such as solitary nodules or multinodular types, as well as microcalcifications or coarse calcifications, should not prevent further evaluation of benign or malignant lesions.

In the present study, a vast majority of thyroid malignancies were PTC (98.05%) and HT was found to coincide highly with PTC (18.63%). Moreover, further regression analysis identified the presence of HT as an independent risk predictor for malignancy (OR = 3.732, P <0.001, Table 3). These data confirm the large American dataset from Dr. Skahen’s report of 1955 [12]. Subsequent studies showed a similar association between PTC and HT. Although according to previous reports, hypothyroidism induced by HT could induce high serum TSH concentration, which may enhance the proliferative activity of follicular epithelia and result in the development of PTC [44], the oncogenic effect that causes PTC in patients with HT are still unknown. According to Harach et al. [45], iodine intake may play a role in the tumorigenesis of thyroiditis. Although Wenzhou is an iodine-sufficient area, determining the correlation between HT and PTC in terms of iodine intake is difficult because no information on iodine intake was available in the majority of patients studied. Some molecular mechanisms, such as the PI3K/Akt or RET/RAS/ERK pathways may also contribute to tumorigenesis [20, 46]. At present, no evidence suggests that HT is a premalignant lesion for PTC. Furthermore, in contrast to previous studies that reported that the presence of HT in PTC is associated with better prognosis, lower recurrence, and a less aggressive disorder than cases without HT [47–49], the current results did not find that HT has a protective effect on tumor aggressiveness in patients with PTC, which is consistent with the recent reports by Del Rio et al. [50]. Recently, Kim et al. reported that HT was positively associated with multifocality and smaller size but not with extrathyroidal invasion, nodal metastasis, or TNM stage [51]. These divergences may be due, at least in part, to differences in patient selection (this retrospective study only included patients undergoing thyroidectomy, thus, a percentage of patients with HT that are treated conservatively may have been omitted), or to indications for thyroidectomy, inclusion of different histologic types (others included PTC as well as FTC), or varying definitions of thyroiditis. Nevertheless, the current results indicate that the presence of HT represents an increased risk of developing thyroid cancer, particularly PTC, but may have a minimum effect on the aggressiveness of the malignancy.

Conclusions

In conclusion, the risk of thyroid malignancy increases with the presence of HT and microcalcification as evaluated by ultrasonography. Raised TSH levels within the normal range are also independently associated with the likelihood of thyroid malignancy. Coexisting HT in PTC does not have a significant effect on the biologic behavior of PTC. Based on the current findings, because the study was retrospective, we do not advocate a more aggressive role for surgical intervention for patients with HT, microcalcification, or even high TSH levels. Prospective large-scale studies are required to verify or establish the potentially important factors for predicting thyroid malignancy, and the follow-up of the patients with HT in PTC should be done to verify long-term prognosis in clinical practice.

Consent

Written informed consent was obtained from the patient for publication of this report and any accompanying images.