Thyroid Nodules as a Risk Factor for Thyroid Cancer in Patients With Graves' Disease

A Systematic Review and Meta-Analysis of Observational Studies in Surgically Treated Patients

Anastasios Papanastasiou; Konstantinos Sapalidis; Dimitrios G. Goulis; Nikolaos Michalopoulos; Evangelia Mareti; Stylianos Mantalovas; Isaak Kesisoglou


Clin Endocrinol. 2019;91(4):571-577. 

In This Article


The current meta-analysis aimed to investigate if the presence of thyroid nodules in patients with GD is associated with a higher risk of thyroid cancer. The study concluded that the presence of at least one nodule is associated with a fivefold increase in thyroid cancer risk, without difference according to the number of nodules (solitary versus multiple). When patients with malignant or suspicious cytology were removed, a fourfold higher risk was noted.

The prevalence of thyroid cancer in this study was 11.5% (range 3.8%-29.2%). It should be noted, however, that this range concerns only surgical series of patients with GD. In recent decades, an increase in the frequency of thyroid cancer was observed in patients with GD. This is probably due to the fact that patients with GD are treated mostly with total thyroidectomy, and a more intense and extensive histological examination is applied.[11,12]

This meta-analysis has some limitations. First, all the included studies were retrospective. Second, high heterogeneity was noticed among studies for the primary outcome, mainly because of confounding bias. More specifically, the patient characteristics and the indications for surgery varied among studies. For this reason, a sensitivity analysis was performed without the patients with malignant or suspicious cytology. Additionally, in the Kraimps et al[9] study, the diagnosis of GD was not clarified, although all patients underwent clinical, biochemical, ultrasonographic and scintigraphic evaluation. In the Ergin et al[10] study, the patients with suspected papillary thyroid cancer following fine-needle aspiration biopsy (FNAB) were excluded, resulting in a profound selection bias. Finally, in the same study, an attrition bias was identified, as no information regarding the preoperatively detected nodules was available for 19 patients.

Some meta-analyses have focused on thyroid cancer risk in patients with GD. None of them had assessed thyroid nodules as a risk factor for thyroid cancer in patients with GD. Staniforth et al[13] reported no difference in the odds of developing thyroid cancer between patients with GD and toxic multinodular goitre or toxic adenoma. The overall prevalence of thyroid cancer in patients with GD was 7% (95% CI 0.04-0.12), being 23% and 5% in the presence or absence of thyroid nodules, respectively. Although these results are quite similar to those of the current meta-analysis, some differences should be mentioned. In Staniforth's et al study, the association of thyroid nodules with thyroid cancer in GD was a secondary outcome and no meta-analysis was performed. Furthermore, the present study had more strict inclusion criteria to diminish the selection bias, considering that only nonrandomized studies were available in the literature. Jia et al[14] concluded that no difference existed in the prevalence of incidental thyroid cancer between patients with GD and toxic multinodular goitre or toxic adenoma. However, the data were inadequate to confirm any association between thyroid nodules and thyroid cancer in patients with hyperthyroidism.

Ultrasonography is considered the gold standard for thyroid imaging, because of its high sensitivity in the identification of thyroid nodules.[2,15] Every thyroid nodule should be characterized by specific ultrasonographic features, such as microcalcifications, a 'taller-than-wide' shape, irregular margins and central vascularization. Although a combination of these features increases the risk of thyroid cancer, in isolation they are neither sensitive nor specific enough to decide if further investigation would be necessary.[16] Based on the present study, the probability of thyroid cancer is enhanced with the ultrasonographic presence of one or more thyroid nodules in patients with GD. In such cases, a FNAB could be performed for every suspicious nodule to determine the existence of thyroid cancer. If this is not feasible, surgical management with total or near-total thyroidectomy may be considered to avoid thyroid cancer underdiagnosis in this group of patients.

In conclusion, the preoperative detection of thyroid nodules was associated with a higher prevalence of thyroid cancer in patients with GD. However, the retrospective data of the selected studies had led to confounding bias and high heterogeneity. Consequently, further prospective research on cancer risk in patients with GD is required to confirm this finding.