Introduction
Thyroid nodules are frequently found in cases of Graves' disease.Reference Belfiore, Russo, Vigneri and Filetti1–Reference Mishra and Mishra4 The prevalence of these nodules, detected on palpation or via imaging modalities, is higher than in the normal population.Reference Carnell and Valente2 The reported incidence of nodules in Graves' disease varies between 22 and 45 per cent.Reference Carnell and Valente2–Reference Mishra and Mishra4 The incidence of malignancy in these nodules ranges from 0.4 to 9.8 per cent.Reference Silver and Parangi5 Although the risk of carcinoma developing within a nodule is high, the treatment of Graves' disease with nodules is still controversial.Reference Cantalamessa, Baldini, Orsatti, Meroni, Amodei and Castagnone6–Reference Terzioğlu, Tezelman, Onaran and Tanakol11 A higher detection rate of nodules in Graves' disease, using ultrasonography (US), forms the basis of this debate. This makes nodules in Graves' disease a common entity.Reference Mazzeferri9 The diagnosis and treatment of nodules and the risk of malignancy in Graves' disease is currently a much discussed topic. There is as yet no consensus on the best pre-operative investigation or on the need for surgery in Graves' disease with nodules.
The aim of this study was to evaluate pre-operatively the risk of thyroid carcinoma in Graves' disease patients, both with and without nodules, and then, following surgical treatment, to determine the association between thyroid nodules and carcinoma in these patients.
Methods
From September 1996 to December 2005, 150 consecutive patients with Graves' disease who underwent surgical treatment were studied retrospectively. The indications for surgical treatment were as follows: recurrent disease after anti-thyroid treatment (36 patients), recurrent disease in association with a large goitre (67), severe ophthalmopathy (27), serious adverse effects of anti-thyroid drugs (eight), patient preference (seven) or fine needle aspiration biopsy (FNAB) findings (five). The patients were divided into two groups according to pre-operative US findings. Patients with diffuse parenchyma formed group one. In this group, any nodule or suspicious area could not be identified by US examination before the operation. Patients with nodules formed group two.
None of the patients had previously been treated with radioiodine therapy or had received external irradiation to the neck. Patients were treated with methimazole or propylthiouracil pre-operatively. They also received Lugol solution for 10 days before surgical intervention. The diagnosis of Graves' disease was based on history and signs of hyperthyroidism, elevated free and total thyroid hormone concentrations, undetectable or reduced levels of thyroid-stimulating hormone (TSH), the presence of thyroid antibodies in the serum, and increased iodine uptake or diffuse uptake on thyroid scan on admission. (Normal ranges were taken as: tri-iodothyronine (T3), 0.8–2 ng/ml; thyroxine (T4), 5–12 µg/dl, free thyroxine, 10–25 pmol/l, TSH, 0–4.2 mIU/l; anti-TSH antibodies, 0–35 IU/ml.)
During pre-operative evaluation, clinical examinations and thyroid US (5–7.5 MHz probe) were performed on all patients. Fine needle aspiration biopsy of US-detected nodules was performed in 30 (20 per cent) of the 150 patients. Ultrasonography-guided biopsy was routinely employed for nodules >1 cm in diameter, whether palpable or not. For nodules <1 cm in diameter, FNAB was performed when US indicated suspicious findings (i.e. hypoechogenic nodule in association with punctuate calcification and/or intranodular vascularity and/or irregular borders). At the time of FNAB, all patients were euthyroid. After surgical intervention, histopathological examinations were carried out on all specimens.
Microcarcinoma was defined as carcinoma of diameter <10 mm. Histological criteria for radioiodine ablation therapy for thyroid carcinoma after surgery included: tumour size >10 mm; any size of tumour with extension through the thyroid capsule; and multifocality (i.e. three or more tumour foci). Chi-square and Wilcoxon rank test were used for statistical analysis, and p < 0.05 was accepted as significant. Statistical calculations were performed using the Statistical Package for the Social Sciences version 11.0 for Windows software (SPSS Inc, Chicago, Illinois, USA).
The study plan was reviewed and approved by our institutional ethical committee.
Results
Pre-operative findings
The mean ± standard deviation (SD) for patients' ages was found to be 37 ± 11 years (range 17–64 years). The female/male ratio was 105/45. During clinical examination, a palpable nodule was found in 24 (16 per cent) Graves' disease patients. The duration of hyperthyroid clinical symptoms before surgical intervention ranged from five to 96 months (mean ± SD, 29 ± 20 months). Of the 150 patients, 28 (18.6 per cent) were found to have ophthalmopathy. Papillary carcinoma was diagnosed in five of 30 patients undergoing FNAB. The mean ± SD of T3, T4, free T4, TSH and anti-TSH antibodies were as follows: 1.1 ± 0.5 ng/ml, 8.9 ± 3.9 µg/dl, 17 ± 4 pmol/l, 2.5 ± 3.1 mIU/l and 319 ± 27 IU/l, respectively.
Peri-operative findings
Total or near-total thyroidectomy was performed in all patients. Of the 150 Graves disease patients, 95 (63 per cent) underwent total thyroidectomy and 65 (37 per cent) underwent near-total thyroidectomy. Subtotal thyroidectomy was not performed in any patient. Total thyroidectomy was performed via extracapsular dissection. Near-total thyroidectomy was performed using the capsular dissection method, leaving less than 1 g of remnant tissue. The range of the resected tissue weight was found to be 25 to 155 g (mean ± SD, 60 ± 28 g).
Pre-operative ultrasound findings
Ultrasound findings were as follows: diffuse parenchyma (i.e. group one), n = 70; nodule(s) (group two), n = 80. The characteristics of the ultrasonographically identified nodules are shown in Table I.
Histopathological characteristics of nodules
*n=32; †n = 48. SD = standard deviation
No statistically significant differences were observed between the two groups, comparing ophthalmopathy ratios and tissue weights (p > 0.05). The female patients in groups 1 and 2 (64% and 75%) had significantly higher than the male patients (36% and 25%) (p < 0.002). Group one patients had a significantly reduced mean age and Graves' disease duration, compared with group two (p < 0.005) (Table II).
Clinical characteristics of groups 1 & 2
*n=70; †n = 80; **n = 150. ‡p < 0.05, compared with group 1. SD = standard deviation; yrs = years; F = female; M = male; Oph = ophthalmopathy; GD = Graves' disease; mths = months; wt = weight
Of the 150 patients, 18 (12 per cent) had well differentiated thyroid carcinoma. Sixteen of these patients were female, with a mean ± SD duration of Graves' disease of seven ± 10 months. Two of the 18 carcinoma patients had ophthalmopathy. When comparing the 18 patients diagnosed with carcinoma with those Graves' disease patients without carcinoma, no significant difference was seen regarding age, sex, ophthalmopathy, resected tissue, thyroid hormone levels and thyroid antibodies (Table III). A nodule had been palpable in only one thyroid cancer patient pre-operatively. Mean ± SD of tumour size was 6.7 ± 4.3 mm (range 2–20 mm). All tumours were papillary carcinomas, with four cases of the follicular variant. Six tumours showed microscopic extrathyroid invasion, nine showed thyroid capsule invasion and another nine exhibited vascular invasion. The carcinomas were multifocal in two patients (bilateral in one). Fifteen tumours were microcarcinoma. The patients were classified as low risk group according to their MACIS score (metastasis, age, completeness of resection, invasion, and tumorsize) and tumour–node–metastasis scores. The clinical and biological characteristics of the carcinomas are shown in Table IV. Fifty per cent of carcinoma patients received radioiodine ablation therapy, according to the criteria mentioned above (see Methods section). Loco-regional recurrence or distant metastasis did not occur during the follow-up period.
Clinical and laboratory features in graves’ disease with and without carcinoma
*n=18; †n = 132. ‡p > 0.05, comparing patients with and without carcinoma, for all parameters. GD = Graves' disease; Ca = carcinoma; yrs = years; F = female; M = male; Oph = ophthalmopathy; SD = standard deviation; wt = weight; T3 = tri-iodothyronine; T4 = thyroid hormone; f = free; TSH = thyroid-stimulating hormone; anti-TSH = anti-TSH antibodies
Clinical characteristics and histological features of graves’ disease patients with thyroid carcinoma
Pt = patient; yrs = years; grp = group; ca = carcinoma; F = female; M = male; −= negative findings; += positive findings; FV = follicular variant
Papillary carcinoma was found in seven patients in group one and in 11 patients in group two. The incidences of carcinoma were 10 per cent in group one and 13.7 per cent in group two (p > 0.05). The tumour size in group two patients was significantly greater than that in group one patients. Five group two carcinoma cases were detected within the parenchyma outside a nodule. This means that 12 of 18 carcinomas (67 per cent) were found in the parenchyma outside nodules and six carcinomas (33 per cent) were found inside nodules. In the 12 patients in whom papillary cancer was detected in the parenchyma outside nodules, the mean ± SD tumour diameter was 5.7 ± 3 mm. The mean ± SD tumour diameter for the six patients in whom papillary cancer was detected inside a nodule was 8.8 ± 6 mm. The size of tumours found within a nodule was significantly greater than that of those found in the parenchyma outside nodules (p < 0.05). The ratios for thyroid capsule, vascular and extrathyroid invasion were similar amongst the groups. The Macis scores of the groups did not show any statistically significant difference.
Discussion
Thyroid nodules are a common finding in the general population. Palpation findings in iodine-sufficient areas show a nodule prevalence of 3.2–4.7 per cent.Reference Silver and Parangi5, Reference Brander, Viikinkoski, Nickels and Kivisaari12 In endemic areas where iodine intake is deficient, nodules are encountered more frequently (up to 15 per cent).Reference Silver and Parangi5 When ultrasonography is used to examine thyroid morphology, nodule prevalence is found in up to approximately 50 per cent of patients in endemic areas.Reference Brander, Viikinkoski, Nickels and Kivisaari12 Several autopsy studies have provided estimations of thyroid nodule prevalence ranging from 40 to 50 per cent.Reference Komorowski and Hanson13, Reference Sampson, Woolner, Bahn and Kurland14 In a Graves' disease patient, the prevalence of palpable thyroid nodules is approximately three times greater than that in the general population.Reference Belfiore, Russo, Vigneri and Filetti1–Reference Mishra and Mishra4 Palpable nodule prevalence is 10–15 per cent in Graves' disease patients, but only approximately 5 per cent in the general population.Reference Belfiore, Russo, Vigneri and Filetti1, Reference Kraimps, Bouin-Pineau, Mathonnet, De Calan, Ronceray and Visset3, Reference Silver and Parangi5 In this series of 150 patients, palpable nodules were found in 24 patients (16 per cent). In the present study, patients with ultrasonographically identified nodules had a significantly longer duration of Graves' disease, compared with patients without nodules. The mean age of group one (diffuse parenchyma) was younger than that of group two (nodules).
Some studies have suggested that Graves' disease patients might have an increased prevalence of differentiated thyroid carcinoma and benign thyroid nodules. The incidence of thyroid carcinoma in our study was 12 per cent (18/150), a finding which contrasts with previous reports showing an incidence of thyroid malignancy in Graves' disease of 0.4 to 9.8 per cent.Reference Cantalamessa, Baldini, Orsatti, Meroni, Amodei and Castagnone6–Reference Mazzeferri9, Reference Terzioğlu, Tezelman, Onaran and Tanakol11
The incidental carcinoma ratio in patients who had undergone surgery for benign thyroid disease was 10–15 per cent.Reference Giles, Boztepe, Terzioğlu and Tezelman15 It should be mentioned that autopsy studies have reported thyroid microcarcinoma prevalence rates of between 3 and 36 per cent.Reference Komorowski and Hanson13, Reference Sampson, Woolner, Bahn and Kurland14 The findings in Graves' disease patients resemble autopsy findings when patients with thyroid carcinoma within dominant nodules are excluded. The incidence varies in different series, and a higher rate of incidence in recent series is possibly due to pre-operative US scanning, a preference for total thyroidectomy, and an increased number of histological sections per specimen.Reference Brander, Viikinkoski, Nickels and Kivisaari12, Reference Giles, Boztepe, Terzioğlu and Tezelman15
• Thyroid nodules are frequently found in Graves' disease
• The aim of this study was to evaluate the risk of thyroid carcinoma in Graves' disease patients with and without ultrasonographically identified nodules
• Of the 150 patients, 18 (12 per cent) were found to have papillary thyroid carcinoma
• Carcinoma can occur in Graves' disease without nodules. The absence of nodules on ultrasonographic examination does not reduce the risk of malignancy
In Graves' disease patients, the malignancy rate for palpable thyroid nodules is reported to range from 2.3 to 48 per cent (mean 16.9 per cent), whereas, in the general population, the malignancy rate for palpable nodules is approximately 5 per cent.Reference Cantalamessa, Baldini, Orsatti, Meroni, Amodei and Castagnone6, Reference Chao, Lin, Jeng and Chen7, Reference Mazzeferri9, Reference Terzioğlu, Tezelman, Onaran and Tanakol11 In this series, only one of 18 carcinoma patients (5.5 per cent) had a palpable nodule. The prevalence of clinical thyroid carcinoma increases in Graves' disease patients, and a thyroid nodule diagnosed in a Graves' disease patient has a higher risk of malignancy, compared with one in a euthyroid patient.Reference Chao, Lin, Jeng and Chen7, Reference Mazzeferri9
Series including Graves' disease patients with coexisting carcinomas have reported that almost all carcinomas are found within nodules.Reference Cantalamessa, Baldini, Orsatti, Meroni, Amodei and Castagnone6–Reference Terzioğlu, Tezelman, Onaran and Tanakol11 In a previous study, patients with nodules and diffuse parenchyma, diagnosed by palpation and imaging modalities, were divided into groups and their carcinoma ratios compared. More papillary thyroid carcinoma was found in the groups with nodules (detected by palpation and/or imaging modalities), compared with the diffuse parenchyma group.Reference Gerenova, Buysschaert, Burbure and Daumerie8
Many studies have reported that the incidence of nodules in Graves' disease patients is higher than that in the general population. However, the clinical significance of most of these nodules remains controversial, as thyroid nodules are frequently detected in ultrasonographic studies of the thyroid.Reference Belfiore, Russo, Vigneri and Filetti1, Reference Kraimps, Bouin-Pineau, Mathonnet, De Calan, Ronceray and Visset3 Although the treatment of nodules in Graves' disease patients is still controversial, it is currently accepted that there is a high incidence of carcinoma in nodules in Graves' disease patients, and that therefore early thyroidectomy is recommended in these cases, particularly in endemic regions.Reference Belfiore, Russo, Vigneri and Filetti1, Reference Kraimps, Bouin-Pineau, Mathonnet, De Calan, Ronceray and Visset3, Reference Chao, Lin, Jeng and Chen7, Reference Mazzeferri9 In our study, the incidence of carcinoma in the parenchyma was higher than of carcinoma found within nodules (p = 0.021).
The treatment of microcarcinomas is still a matter of discussion. Many remain occult, and may be diagnosed as an incidental finding during surgery for benign thyroid disorders.Reference Giles, Boztepe, Terzioğlu and Tezelman15, Reference Ringel and Ladenson16 However, some microcarcinomas may have a negative outcome, such as distant metastasis and patient death.Reference Okamoto, Iihara and Obara10, Reference Ringel and Ladenson16–Reference Pellegriti, Scollo, Lumera, Regalbuto, Vigneri and Belfiore18 The pathogenic relationship between Graves' disease and papillary thyroid carcinoma is still obscure. Although some authors have reported that carcinoma in Graves' disease patients behaves more aggressively, others have refuted this finding.Reference Carnell and Valente2, Reference Komorowski and Hanson13, Reference Hales, McElduff, Crummer, Clifton-Bligh, Delbridge and Hoschl19
Conclusion
Differentiated papillary carcinomas can occur in Graves' disease patients without ultrasonographically identified nodules, and the absence of nodules does not diminish such patients' chances of developing malignant disease.
