Abstract
We read with interest the article by Brito et al. (1). Although some studies have reported data on the comparative prevalence of thyroid carcinoma in single nodule (SN) and multinodular goiter (MNG), no certain evidence of a difference in the risk profile of thyroid cancer in the two phenotypes has been provided until now (2,3). Since differences in thyroid cancer risk might impact recommendations and practice, Brito et al. (1) carried out a systematic analysis of 14 trials in order to summarize the available data and to verify the reasons for differing prevalence estimates across studies. Applying a meta-analysis method, the authors found that thyroid cancer is less frequent in MNG than SN, particularly outside the United States and perhaps in iodine-deficient areas. However, as correctly stated by the authors, the meta-analysis presents several limitations, resulting in low-quality evidence. Particularly, the impact of iodine deficiency on cancer prevalence mainly derives from the cytological findings of an Italian study (4), carried out in a very large series of patients and accounting for almost one fourth of the weight of the meta-analysis. In this setting, the reported similar cancer prevalence between studies carried out in the United States versus Italy appears somewhat surprising, as stated in the meta-analysis subgroup analysis. Thus, the Italian studies showed either higher cancer prevalence in SNs or lower overall prevalence compared to those conducted in the United States (5). Moreover, the overall summary of the meta-analysis of the iodine status subgroup depicted in table 4 appears inconsistent with the conclusion of lower cancer prevalence in MNGs of patients living in iodine-deficient areas.
Although Brito et al. (1) describe the prevalence of papillary thyroid cancer (PTC) where available, they did not perform a subanalysis on the basis of cancer phenotypes (follicular thyroid cancer (FTC) vs. PTC, with histological variants). Giving that the relative prevalence of the two phenotypes may be influenced by iodine intake (6,7) and that MNG is more frequent in iodine-deficient areas, the comparison among countries and the possible role of iodine intake might have been affected by including a histological subanalysis. Moreover, the studies reporting on the results of fine-needle aspiration cytology (FNAC) lack the fraction of cancers harboring in the subgroup of nodules with follicular pattern at FNAC examination (indeterminate cytology). Since the prevalence of cancer among thyroid nodules with indeterminate cytology is not negligible, their inclusion in the comparative analysis could modify the risk profile of thyroid cancer in SN and MNG. In this setting, we would like to report our experience on the prevalence of malignancy in a series of consecutive thyroid nodules (169 SN, 140 MNG) with indeterminate cytology, operated on at the same institution over a five-year period (8). The study population comprised patients living in a mildly iodine-deficient area (mid/southern Italy) (9). Overall, we observed 20.1% prevalence of cancer, higher in SNs (23.1%) than MNGs (16.4%); the analysis of the prevalence of cancer phenotypes revealed follicular variants of PTC in 46/62 (74.2%), minimally invasive FTC in 12/62 (19.4%), and invasive FTC in 4/62 (6.4%). The subanalysis showed that SNs harbored the FTC phenotype more than twice that diagnosed in MNGs (33% vs. 13%).
According to the meta-analysis by Brito et al. (1), our data confirm a lower prevalence of differentiated thyroid cancer in MNG compared to SN, not affected by cancer phenotype. Nonetheless, the overall higher cancer prevalence of our study compared to that reported in the meta-analysis clearly depends on the selection of nodules with indeterminate cytology. Thus, although differentiating the two phenotypes may be useful for planning the clinical workup of patients with thyroid nodules, especially in iodine-deficient areas, FNAC still remains the cornerstone for stratifying the risk of malignancy.