Hashimoto's thyroiditis in papillary thyroid carcinoma: a 22-year study.

OBJECTIVE: We retrospectively investigated whether there is a relationship between Hashimoto's thyroiditis (HT) and papillary thyroid carcinoma and studied the effect of HT on prognostic factors.
METHODS: 1080 patients, who underwent thyroidectomy in our hospital and received a diagnosis of papillary thyroid carcinoma, were included in the study. In histopathological specimens, the diagnosis of papillary thyroid carcinoma was reconfirmed and non-neoplastic areas in the same specimen were evaluated in terms of HT.
RESULTS: HT was detected in non-neoplastic areas of specimens in 36.1% (n = 390) of 1080 patients with a diagnosis of papillary thyroid carcinoma whereas HT was not observed in 63.9% (n = 690). There was a significant positive correlation between presence of HT and multifocal location (p < 0.05, χ2 = 38.5). There was no significant relationship between extrathyroidal tissue invasion and HT (p > 0.05).
CONCLUSION: We assume that patients with HT developing papillary thyroid carcinoma have an increased risk of having multifocal tumour, and thus surgical intervention should be tailored according to this risk.

Introduction

Hashimoto’s thyroiditis (HT) is a common autoimmune disorder of the thyroid gland. It can be encountered in all age groups. It is 10-20 times more common in females than in males. Although its aetiology is not fully elucidated, it is known that it is related to the release of cytokines, which damage thyroid follicle cells, following the activation of the helper T lymphocytes. Hypothyroidism is found in the majority of patients with HT [1].

HT has two major components from a histopathological point of view. The first is lymphoplasmacytic cell infiltration into stroma. In addition, histiocytes and scattered multinucleated giant cells may be also observed. The second main feature is oxyphilic changes in the epithelium paving the thyroid follicles. These cells, also called as Askanazy or Hurthle cells, have nuclei that are larger than normal, as well as prominent nucleoli and wide acidophilic and granular cytoplasm. In addition, similar to that in papillary thyroid carcinoma (PTC), cells may contain optically clear nuclei, which tend to overlap follicle epithelium. In HT, the number of the thyroid follicles decrease in parallel with increased inflammatory infiltration. Follicles with an intact appearance are usually smaller than normal and contain less colloid than normal. Limited fibrosis in the thyroid glands is observed in pathological examinations of patients with HT. However, in the fibrous variant of the disease, connective tissue constitutes more than 30% of the thyroid and this variant is encountered in 12% of all patients [2,3].

Following the first description of the relationship between inflammation and cancer by Virchow in 1863, the relationship of cancer with inflammation of organs such as the lung, stomach, bowel and skin has been investigated [4]. The relationship between HT and PTC was first reported by Dailey and colleagues in 1955 [5]. Although this relationship was investigated in several studies, consensus has not yet been achieved [6-9].

In this retrospective study, in which archived data of the last 22 years were screened, our objective was to investigate the presence of HT in PTC and the effect of HT on prognostic factors.

Materials and methods

The study protocol was approved by the local Ethics Committee (2017-14/24). Patient approval was not necessary because this is a retrospective study. We included 1080 patients whose pathology specimens were available for review and diagnosed with PTC who underwent thyroidectomy between 1995 and 2017 in our hospital. The examination of haematoxylin-eosin-stained histopathological materials of these patients was repeated by a single pathologist. With this examination, first the diagnosis of PTC was confirmed and secondly the non-neoplastic fields in the same specimen were evaluated for the presence of HT. Regarding HT, the presence of Hurthle cells and lymphocytic inflammation with a germinal centre in the thyroid parenchyma and stroma were the main findings looked for (Fig. 1). Moreover, tumour size, extrathyroidal tissue infiltration and multifocality in PTC fields were evaluated.

Figure 1.

Lymphoid follicular and Hurthle cells surrounding the papillary carcinoma.

IBM SPSS Statistics 22 (IBM SPSS, Turkey) program was used to evaluate data. Fisher’s Exact and Chi-square tests were used to compare data. Significance was set at 0.05.

Results

57.5% of patients (n = 621) were female and 42.5% (n = 459) were male. The mean age of patients was 47 ± 14 years (23-78 years). HT was observed in non-neoplastic fields in 36.1% of 1080 patients diagnosed with PTC (n = 390). The remaining 690 patients (63.9%) had no HT findings. 56.5% of female patients with PTC (n = 351) were HT-negative and 43.5% (n = 270) were HT-positive, whereas 73.8% of males (n = 339) were evaluated as HT-negative and 26.2% (n = 120) as positive.

Of 690 HT-negative patients, 83.5% had PTC with unifocal involvement and 16.5% had multifocal PTC. On the other hand, of 390 HT-positive patients, 53% had unifocal PTC and 47% had multifocal PTC. The correlation analysis showed a significantly positive correlation between HT presence and multifocality (p < 0.05, χ2 = 38.5).

Evaluation of extrathyroidal tissue invasion showed that there was extrathyroidal infiltration in 13.9% of HT-positive and 16.9% of HT-negative specimens. No significant relationship was found between extrathyroidal tissue invasion and HT (p > 0.05).

Moreover, there was no significant relationship of HT presence with age, tumour size and lymphovascular infiltration in patients diagnosed with PTC (p > 0.05).

Discussion

Despite developments in surgical techniques, the underlying causes of PTC have not yet been fully elucidated. It has not been clearly demonstrated whether papillary carcinoma is a reactive response to HT or if HT is a precursor to PTC development [8]. A number of studies have been conducted to answer this question [10-16]. In the first study conducted by Dailey, the incidence of PTC in patients with HT was 35% [5]. In subsequent studies, Schäffler reported the lowest rate (8%) and Repplinger reported the highest rate (63%) [10,17]. Matesa-Anić et al. [11], in their study on 10,580 subjects, which is the largest study on this topic, found an incidence of 42%. In this study, different from other studies, the rate of HT in patients with PTC was investigated and 36.1% of 360 patients with PTC had findings of HT in non-neoplastic fields. Konturek and his colleagues showed that the incidence of PTC was 3 times higher in patients with HT [1].

The relationship between HT and differentiated thyroid carcinoma was investigated in several studies and some investigators have suggested that there was an association between PTC and HT. Regarding its pathogenesis, many authors have tried to demonstrate the relationship between HT and PTC intracellular signaling proteins. In particular, phosphatidylinositol 3-kinase (PI3K), plays a critical role in the balance between cell life and apoptosis [18]. Cunha and colleagues found that there were different cell combinations and that CD4, CD33 and CD11b-positive lymphocytes were involved in the development of thyroid carcinoma [19].

It has been reported that extrathyroidal invasion was significantly decreased in PTC patients with HT [9,20]. In our study, evaluation of extrathyroidal tissue invasion showed that 13.9% of HT-negative patients and 16.9% of HT-positive patients had extrathyroidal infiltration. However, no significant relationship was seen between extrathyroidal tissue invasion and HT (p > 0.05).

Kim et al. [21] determined that multifocal and bilateral thyroid carcinoma was more common in HT patients. Sakorafas and Falvo found similar results [22,23]. In our study, similar to the literature, multifocal PTC was more common in patients with HT and the relationship was statistically significant.

As our study had a retrospective design, we had limited access to clinical, radiological and biochemical data. However, regarding the data we were able to obtain, the risk of developing PTC in HT patients should be taken into consideration and follow-up should be planned accordingly.

Conclusions

Biopsy of all suspected nodes will increase the probability of making the correct diagnosis in patients scheduled for fine needle aspiration biopsy. Lobectomy can be performed as an option in patients who are diagnosed with PTC and have small and single nodules. Since the risk of having multifocal disease is high in HT patients developing PTC, total thyroidectomy should be planned and a selective neck dissection should be carried out in lymph node-positive patients.