The role of Graves’ disease in the development of thyroid nodules and thyroid cancer

Abstract

Aim

The prevalence of thyroid nodules and the risk of thyroid cancer in patients with Graves’ disease is uncertain. We aimed to evaluate the prevalence of thyroid nodules and cancer in patients with Graves’ disease.

Methods

Retrospective observational study of adult subjects with Graves' disease (positive autoantibodies thyrotropin receptor antibodies (TRAbs)) between 2017 and 2021 at our center was done. We evaluated the prevalence of thyroid nodules and cancer in this population and characterized the predictive factors for thyroid malignancy using linear and logistic regression models.

Results

We evaluated a total of 539 patients with Graves' disease during a median follow-up of 3.3 years (25th–75th percentiles 1.5–5.2 years). Fifty-three percent had thyroid nodules and 18 (3.3%) were diagnosed with thyroid cancer (12 papillary microcarcinomas). All tumors were classified using TNM classification as T1, and only one had lymph node metastasis; there were no recordings of distant metastasis. Sex, age, body mass index, smoking, TSH, and TRAbs levels were not significantly different between patients with and without thyroid cancer. Patients with multiple nodules on ultrasound (OR 1.61, 95%CI 1.04–2.49) and with larger nodules (OR 2.96, 95%CI 1.08–8.14, for 10 mm increase in size) had a greater risk of thyroid cancer diagnosis.

Conclusion

Patients with Graves’ disease had a high prevalence of thyroid nodules and their nodules had a significant risk of thyroid cancer. The risk was higher in those with multiple and larger nodules. Most had low-grade papillary thyroid cancer. More studies are needed to clarify the clinical relevance of these findings.

Introduction

Graves' disease is an autoimmune thyroid disease, being the most frequent cause of hyperthyroidism (1). It is caused by the production of autoantibodies (thyrotropin receptor antibodies (TRAbs)) against the thyrotropin receptor (TSH-R) (2). These act as TSH-R agonists and induce hypersecretion of thyroid hormones (3). In addition, TSH-R is also involved in the normal physiology and growth of the thyroid gland, participating in its organogenesis, cell differentiation, and iodine uptake, which may lead to the formation of goiter when faced with a strong stimulus (4).

Differentiated thyroid carcinoma, which originates in the epithelial follicular thyroid cells, represents about 95% of thyroid cancer cases. It includes papillary thyroid carcinoma (PTC), follicular thyroid carcinoma, and oncocitic cell carcinoma of the thyroid (5).

PTC is the most common histological subtype of differentiated thyroid malignancy, constituting about 80% of cases (6). The age of presentation, gender, histology, and stage of the tumor are factors that influence its prognosis (7).

Graves' disease has been associated with an increased risk of thyroid cancer (8). Several case reports suggest an aggressive behavior of thyroid cancer in patients with Graves’ disease (9). Infiltrative immune cells in Graves' disease may also contribute to an increased risk of thyroid cancer due to a link between tissue inflammation and carcinogenesis (10). The mechanism for this increased risk is still not fully understood, but the binding of TRAbs to TSH-R is probably one of the main mechanisms (11). By positively stimulating thyroid growth, production of growth factors and angiogenesis, TRAbs may promote tumor development and invasion (12). However, predictive factors for thyroid cancer in patients with Graves' disease are still unknown (13).

Thus, our aim was to evaluate the prevalence of thyroid nodules and thyroid cancer in patients with Graves’ disease and to characterize the predictive factors of thyroid malignancy.

Material and methods

Study design and participants

We performed a retrospective longitudinal observational study at Centro Hospitalar Universitário de São João (CHUSJ) in Porto, Portugal. The studied population consisted entirely of adult subjects with positive TRAbs, enrolled between 2017 and 2021 (all participants under 18 years of age were excluded from this selection). Data was collected until May 2022.

Our study was approved by the Ethics Committee of CHUSJ/ Faculty of Medicine of the University of Porto (process number 204-22).

The data analyzed in this study were collected retrospectively from electronic clinical records in May 2022.

Clinical characteristics and study outcomes

The following parameters were collected from the patients’ clinical records: sex, age, age at diagnosis of Graves' disease, initial TRAbs, free tri-iodothyronine (FT3), free thyroxine (FT4), thyroid stimulating hormone (TSH), height, weight, smoking habits, treatment for Graves’ disease (type of antithyroid drug, maximum dose, duration and number of iodine ablative doses, and thyroidectomy), diagnosis of ophthalmopathy, diagnosis of dermopathy, thyroid ultrasound results, results of biopsies, and thyroid carcinoma (histology, classification, tumor size, and clinical evolution in the follow-up). TSH, FT3, and FT4 were measured in serum samples by an electrochemiluminescence immunoassay using an Abbott Architect i2000 analyzer (Abbott Diagnostics). TRABs were measured by radioimmunoassay. The normal values for these assays are TSH 0.35–4.94 mU/mL, FT4 0.70–1.48 ng/dL, FT3 1.71–3.71 pg/mL, and TRAbs <1.8 U/L.

The ultrasound evaluated was the first performed after the diagnosis of Graves' disease. The nodules were identified and recorded by the radiologist performing the evaluation according to usual clinical practice. All data were taken from clinical records. Biopsies were performed according to the usual clinic practice. We defined a ‘thyroid nodule’ as all nodules with a solid component reported in the clinical evaluation by the radiologist. Simple cysts were not classified as nodules. The decision to perform FNA (fine needle aspiration) biopsy was made by the assistant physician based on the suspicious ultrasound characteristics: solid hypoechoic nodule, existence of microcalcifications, hypervascularization, irregular margins, anteroposterior diameter greater than the transversal one, and invasion of extrathyroid tissues. Also, we considered suspicious characteristics as: personal history with increased risk of thyroid cancer (history of head and neck radiation therapy or exposure to ionizing radiation in childhood or adolescence), a hard and adherent nodule on palpation, or compressive symptoms.

Thyroid nodule cytopathology was classified according to the Bethesda System for Reporting Thyroid Cytopathology (14).

Statistical analysis

Baseline characteristics of patients with and without diagnosis of thyroid cancer were compared with Student’s t-test, Mann–Whitney U-test for normal and nonnormal continuous variables, respectively, and the chi-square test for categorical variables.

The association of predictors with the presence of thyroid nodules and the diagnosis of thyroid cancer was evaluated using linear regression (for continuous outcomes) and logistic regression (for dichotomic outcomes).

Continuous variables are described as mean ± standard deviation or median (25–75th percentiles) for normal and nonnormal continuous variables, respectively, and categorical variables as proportions (percentages).

A two-sided P-value of <0.05 was considered statistically significant. Analyses were performed with Stata (version 17.0).

Results

The baseline characteristics of the 539 patients with Graves' disease are shown in Table 1. Eighteen percent were men and their mean age at diagnosis was 45.7 ± 15.1 years. Regarding treatment, 84.1% received methimazole and 9.5% received propylthiouracil. The median duration of antithyroid treatment was 20.0 (11.0–36.0) months. The follow-up time was 3.3 (1.5–5.2) years. The median duration of Graves' disease was longer in patients with thyroid nodules than in those without (275.5 days in those with thyroid nodules vs 126 days in patients without nodules, P = 0.13).

Table 1.

Baseline characteristics. Continuous variables are presented as mean ± s.d. or median (25–75th percentiles) for normal and nonnormal continuous variables, respectively, and categorical variables as n (%).

Characteristics	Values
Total (n)	539
Male	98 (18.2%)
Age, years	45.7 ± 15.1
BMI, kg/m2	25.6 ± 4.9
Ophthalmopathy	128 (23.9%)
Dermopathy	7 (1.3%)
Smoking habits
Non-smoker	321 (61.3%)
Current smoker	163 (31.1%)
Ex-smoker	40 (7.6%)
Antithyroid drug
Methimazole, mg	408 (84.1%)
Propylthiouracil, mg	46 (9.5%)
Initial TRAbs, U/L	7.5 (3.7– 21.8)
TSH, mU/L	0.001 (0.0–0.008)
FT4, ng/dL	1.7 (1.1–2.3)
FT3, pg/mL	5.7 (3.7–10.1)

BMI, body mass index; FT3, free triiodothyronine; FT4, free thyroxine; TRAbs, anti-TSH receptor antibodies; TSH, thyroid-stimulating hormone.

Eighty percent of the patients did not receive any ablative dose with radioactive iodine, 18.3% received one, 1.7% received two, and 0.2% received three doses. From the 539 patients evaluated, 361 (67.0%) patients had thyroid function results available before starting antithyroid drugs. At evaluation, the median values of thyroid function were the following: TSH 0.001 (0.0–0.008) mU/L, FT4 1.7 (1.1–2.3) ng/dL, and FT3 5.7 (3.7–10.1) pg/mL.

Table 2 shows the data related to the first ultrasounds performed after the diagnosis of Graves' disease. Of the 539 patients, 97.6% were evaluated with thyroid ultrasound, 55.0% had at least one nodule, and 39.1% had multiple nodules. The median size of the largest nodule was 10.8 (7.0–17.0) mm. One hundred and forty-four had nodules with maximum diameter >1 cm.

Table 2.

Results from thyroid ultrasound. Nonnormal continuous variables are presented as median (25–75th percentiles) and categorical variables as proportions (percentages).

Thyroid ultrasound	Results
Ultrasound performed	521 (97.9%)
Thyroid nodules	287 (55.0%)
Multiple nodes	204 (39.1%)
Maximal nodule diameter, mm	10.8 (7.0, 17.0)

Table 3 shows the data related to the biopsies performed. Of the 539 patients, 13.0% performed one, 3.4% performed two, and 2.1% performed three FNA. With regard to indeterminate nodules, most patients with FLUS/AUS repeated FNA and patients with ‘Follicular neoplasm or suspicious for follicular neoplasm’ were proposed to thyroidectomy, as usual in our clinical practice.

Table 3.

Results from thyroid biopsies (n = 139). Results shown according to Bethesda classification.

Bethesda classification	Results
I–Non-diagnostic or inadequate	23 (16.5%)
II–Benign	99 (71.2%)
III–Atypia/follicular lesion of undetermined significance	4 (2.9%)
IV–Follicular neoplasm or suspicious for follicular neoplasm	7 (5.1%)
V–Suspicious for malignancy	1 (0.7%)
VI–Malignant	3 (2.2%)
Suspicious for lymphoma	2 (1.4%)

From the patients who underwent total thyroidectomy (n = 78, 14.5%), 17 (23.0%) had thyroid cancer. Eight of these patients did not undergo a FNA prior to surgery. Of the 9 patients submitted to FNA prior to surgery, 2 were classified as malignant in the FNA, 2 had follicular neoplasm or suspicious follicular neoplasm, and 5 had benign results. All benign results were from a nodule that did not correspond to the thyroid cancer diagnosed after surgery. In one patient, the FNA result was ‘VI–Malignant’; however, considering the patient's severe cardiovascular disease, a conservative treatment approach was used instead of surgery.

Table 4 shows data for patients submitted to thyroid surgery and the characteristics of the thyroid cancers identified. From the patients submitted to thyroid surgery (n = 78), 94.9% underwent total thyroidectomy and 5.1% underwent hemithyroidectomy. Eighteen patients had the diagnosis of thyroid cancer (3.3%). Of the 17 operated cancers, 5 were discovered on preoperatory FNA of thyroid nodule, 1 was discovered by FNA of cervical lymph node, and 11 were an incidental finding on thyroidectomy. Eleven had suspicious nodules on ultrasound, which were requested for FNA and, after surgery, it was confirmed that they were malignant. Regarding the histological classification of the tumors, 10 (55.6%) were papillary microcarcinomas, 3 (16.7%) were classic papillary carcinomas, 4 (22.2%) were follicular variant papillary carcinomas, and 1 (5.6%) was NIFTP (non-invasive follicular thyroid neoplasm with papillary-like nuclear features). Of the 10 microcarcinomas, 3 were discovered on preoperatory FNA, 1 discovered by FNA of a cervical lymph node, and 6 were an incidental finding on thyroidectomy. The mean size of the tumors was 7.3 ± 4.1 mm. All tumors were T1 in T category of TMN classification – 12.5% pT1a, 56.3% pT1aN0R0, 12.5% pT1(m)Nx, 12.5% pT1bNxR0, and 6.25% pT1N1R0. All patients had excellent response (negative imaging and either suppressed Tg <0.2 ng/mL or TSH-stimulated Tg <1 ng/mL) at the timing of last evaluation (15).

Table 4.

Thyroid surgery and histology of thyroid cancers.

Thyroid surgery	Results
Total thyroidectomy	74 (94.9%)
Hemithyroidectomy	4 (5.1%)
Thyroid cancer	17 (3.1%)
Tumor size, mm	7.3 ± 4.1
Classificationa
Papillary microcarcinoma	10 (55.6%)
Classic papillary carcinoma	3 (16.7%)
Follicular variant papillary carcinoma	4 (22.2%)
NIFTP- non-invasive follicular thyroid neoplasm with papillary-like nuclear features	1 (5.6%)
Tumor stageb
pT1a	2 (12.5%)
pT1aN0R0	9 (56.3%)
pT1(m)Nx	2 (12.5%)
pT1bNxR0	2 (12.5%)
pT1N1R0	1 (6.25%)

^aLloyd RV, Osamura RY, Klöppel G. editors. WHO Classification of Tumours of Endocrine Organs (4th edition). Lyon: IARC Press, 2017 .

^bTumor stage shown according to the TNM staging AJCC UICC 8th edition.

Patients with thyroid cancer were more likely to have multiple nodules on ultrasound (64.7% in patients with multiple nodules vs 38.2% in patients without nodules or with a single nodule, P = 0.028) and to have larger nodules (17.8 (8–28) vs 10.4 (7–17), P = 0.048) that patients without thyroid cancer diagnosis (Table 5). The odds ratio of having thyroid cancer was 1.6 (95%CI, 1.04–2.49) for patients with multiple thyroid nodules (vs no nodules or single nodule) and 3.0 (95%CI, 1.1–8.1) for each 10 mm increase in the size of the largest nodules. The differences between sex, age, BMI, ophthalmopathy, and dermopathy were not statistically significant. Regarding smoking habits, treatment with antithyroid drugs, and the number of ablative doses, the differences were also not significant between groups. Thyroid function biochemical values at diagnosis of Graves' disease (TSH, FT4, FT3, and TRAbs) did not show a significant association.

Table 5.

Predictors of thyroid cancer. Continuous variables are presented as mean ± s.d.or median (25–75th percentiles) for normal and nonnormal continuous variables, respectively, and categorical variables as n (%).

	Without thyroid cancer	With thyroid cancer	P value
n	521	18
Male	96 (18.4%)	2 (11.1%)	0.43
Age, years	45.7 ± 15.0	46.4 ± 16.2	0.86
BMI, kg/m2	25.6 ± 4.9	26.0 ± 5.4	0.76
Ophthalmopathy	124 (23.9%)	4 (22.2%)	0.87
Dermopathy	7 (1.4%)	0 (0.0%)	0.63
Smoking habits			0.91
Nonsmoking	311 (61.3%)	10 (58.8%)
Smoker	157 (31.0%)	6 (35.3%)
Ex-smoker	39 (7.7%)	1 (5.9%)
Antithyroid drug			0.31
Methimazole	394 (84.2%)	14 (82.4%)
Propylthiouracil	43 (9.2%)	3 (17.6%)
Ablative doses			0.95
0	408 (79.7%)	14 (82.4%)
1	94 (18.4%)	3 (17.6%)
2	9 (1.7%)	0 (0.0%)
3	1 (0.2%)	0 (0.0%)
One or more thyroid nodules	274 (54.3%)	13 (76.5%)	0.07
Multiple nodes	193 (38.2%)	11 (64.7%)	0.028
Largest nodule dimension, mm	10.4 (7.0–17.0)	17.8 (8.0–28.0)	0.048
TSH, mU/L	0.0 (0.0–0.0)	0.0 (0.0–0.0)	0.25
FT4, ng/dL	1.7 (1.1–2.3)	1.4 (1.1–2.2)	0.51
FT3, pg/mL	5.8 (3.6–10.1)	5.5 (4.2–9.5)	0.96
TRAbs	7.5 (3.8–21.8)	5.2 (3.4–10.7)	0.55
Antithyroid drug duration, months	20.0 (10.0–35.0)	16.0 (10.0–37.0)	0.63

BMI, body mass index; FT3, free triiodothyronine; FT4, free thyroxine; TRAbs, anti-TSH receptor antibodies; TSH, thyroid-stimulating hormone.

Discussion

In this retrospective cohort study, patients with Graves' disease have a high prevalence of thyroid nodules and a relevant risk of thyroid malignancy (3.3% of patients during a median follow-up of 3.3 years) compared to the general population (9, 16, 17). Hence incidence of thyroid cancer is higher from what is reported for the Portuguese population (0.40/100,000 in women and 0.34/100,000 in men) (18). Unfortunately, no robust data for the prevalence of thyroid nodules is available in Portugal. However, studies in Europe suggest a prevalence of 50% in adult females and 30% in males (18).

All of the thyroid cancers diagnosed were low-grade papillary carcinomas. Patients with multiple nodules and larger nodules had a higher risk of thyroid malignancy.

Our study is in agreement with previous literature that suggests that thyroid nodules are frequent in patients with Graves’ disease (9, 19).

Regarding the risk of thyroid carcinoma in Graves’ disease, some studies have suggested an increased risk in patients with Graves’ disease (13, 16, 20, 21), while others suggest a similar risk (22, 23). The proposed mechanism for an increased risk in those with Graves’ disease is a stimulatory effect of TRAbs on the TSH receptor (2). This is aligned with the studies performed in patients without Graves’ disease in which higher levels of TSH were associated with an increased risk of thyroid cancer (24, 25). Some studies have also reported an increased risk of disease progression and distant metastasis in patients with Graves' disease (19, 20, 26).

Most previous studies compared the risk of thyroid cancer in patients with Graves' disease and in euthyroid individuals. Such approaches may be of interest to understand the increased risk of thyroid cancer in patients with Graves’ disease. However, from a clinical perspective, it is more relevant to identify, within those with Graves’ disease, the individuals at an increased risk of thyroid cancer. Thus, we can act earlier and more effectively in these patients.

It is known that there are two well-established risk factors for the development of PTC: previous radiation exposure (27) and family history (28). A number of other possible risk factors have been reported, such as elevated TSH levels (24). Although Graves’ disease has been proposed as a risk factor, the mechanism is still not fully characterized (29). TRAbs, by promoting the mechanisms of angiogenesis and production of growth factors (8), can stimulate the formation of nodules, which may have malignant potential (12).

Therefore, it is very important to do a follow-up on patients with Graves' disease in the long term, using complementary diagnostic tools, such as ultrasound and FNA, when necessary (30). Our study suggests that patients who had more and larger nodules have an increased risk of developing thyroid cancer (9, 19). The prevalence and clinical significance of incidental thyroid carcinoma in patients with Graves' disease are relevant data when deciding on therapy (medical vs surgical) (31).

Our study has limitations that we must acknowledge. The first limitation concerns the retrospective design of our study, which limits the collection of more detailed data. The second limitation is that the study was carried out in a single center, which may decrease the generalizability of our findings. It is important to mention that the prevalence of thyroid cancer may have been higher due to the fact that these patients, with Graves' disease, had a more regular follow-up. The prevalence of thyroid cancer in this population cannot be directly compared to the prevalence in the general population as some patients were submitted to thyroidectomy due to hyperthyroidism (this may increase the finding of incidental thyroid cancers). Moreover, microcarcinomas are known to exist in many undiagnosed people and have no future repercussions (32). The selection bias inherent in the clinical care of these patients is also a limitation with respect to the detection of malignancy. It would have been relevant to have also a control group of euthyroid patients. In all patients, only the first ultrasound after the diagnosis of Graves’ disease was evaluated. As such, unfortunately, we cannot evaluate the presence of nodularity prior to the development of Graves' hyperthyroidism. Additionally, the significant association between thyroid nodule size and cancer diagnosis may possibly be in part attributed to the indication for FNA which lead to an identification of larger thyroid cancers but not smaller thyroid cancers. Finally, our study may have been underpowered to detect other predictor of malignancy in Graves’ disease given the low number of thyroid cancers detected (18 patients) despite the inclusion of more than 500 participants with Graves’ disease. Regarding the strengths of our study, we analyzed the data of all patients in a hospital who had a diagnosis of Graves’ disease (identified by positive TRAbs over 5 years). Instead of defining a control group made by euthyroid patients, we chose one with Graves' disease because we found it more clinically relevant. Within patients with Graves' disease, we identified those who had thyroid cancer and compared them with those who did not. We also believe that the large set of potential predictors analyzed is a strength of our study.

In conclusion, Graves’ disease patients with more and larger nodules had an increased risk of thyroid cancer. Further studies are needed to better understand the optimal approach to the follow-up and evaluation of the risk of thyroid cancer in patients with Graves’ disease.

Declaration of interest

The authors have no relevant financial or non-financial interests to disclose.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Ethics approval

The authors state that the procedures followed were in accordance with the regulations of Ethics Committee of the Centro Hospitalar Universitário de São João/ Faculty of Medicine of University of Porto and with those of the Code of Ethics of the World Medical Association (Declaration of Helsinki).

Availability of data and material

The datasets analyzed during the current study are not publicly available but are available from the corresponding author on rational request.

Author contribution statement

JSN and MNS conceived and designed the study. JSN provided methodological and statistical advice. MNS drafted the first version of the manuscript. All authors provided clinical feedback in interpreting the results, contributed critically to subsequent revisions, and approved the final version of the manuscript.