Spindle cell foci of the thyroid-mimicking malignancy: a diagnostic pitfall.

Abstract

To the Editor: We recently published a case series of spindle cell foci of the thyroid associated with multinodular goiter, follicular adenomas, and minimally invasive follicular carcinoma.1 Previously, Vergilio et al2 described spindle cell foci of the thyroid gland in association with papillary thyroid carcinoma and microcarcinoma. This uncommon entity is usually an incidental finding that consists of a spindle cell proliferation with nodular or diffuse distribution, bland cytologic features, low Ki-67 index, and an immunophenotype that supports a metaplastic follicular origin of the spindle cells. We previously emphasized the potential for confusion of spindle cell foci of the thyroid with medullary thyroid carcinoma, which is well demonstrated in this recent case that we reviewed on our consultation service. We report a 61-year-old woman who underwent a total thyroidectomy followed by cervical lymph node dissection after needle biopsy raised the concern for medullary thyroid carcinoma. Gross examination demonstrated the presence of two solid nodules measuring 1.6 and 0.7 cm in maximum diameter. Microscopic examination of sections of the nodule on the left lobe demonstrated an adenomatous nodule with macrofollicular architecture, extensive degenerative changes that included fibrosis and sclerosis along with florid spindle cell foci with infiltrative borders, and areas of eosinophilic interstitial deposits (Fig. 1). The spindle cells were arranged in sheets as well as in short fascicles, had elongated plump nuclei, fine chromatin, inconspicuous nucleoli, and variable eosinophilic cytoplasm (Figs. 1B, C). Intermixed cells with clear or foamy appearance were present. There was no evidence of mitotic activity or necrosis. Areas resembling desmoplastic stromal reaction associated with spindle cells were noted (Fig. 1E). The areas of sclerosis and fibrosis had an amorphous appearance simulating amyloid (Fig. 1F). A focus showing nuclear pleomorphism that merged with the spindle cell areas was noted in one section (Fig. 1F). Rare nuclear pseudoinclusions were also noted. Both the bland spindle cell component and the atypical cells had a low proliferation index on Ki-67 stain (<1%). All lymph nodes were negative for malignancy. Using immunohistochemistry, we found that the spindle cells were strongly positive for thyroid transcription factor-1 (TTF-1) and thyroglobulin but negative for cytokeratin AE1/3, synaptophysin, carcinoembryonic antigen (CEA), calcitonin, and CD56. Congo red stain was negative. Low Ki-67 along with the positive TTF-1 staining helps in excluding the possibility of anaplastic carcinoma. On the basis of these features, a diagnosis of adenomatous nodule with spindle cell foci consistent with spindle cell metaplasia of follicular cells was rendered. The atypical nuclei were interpreted as being consistent with “endocrine atypia”. As in the case presented here, all previous cases showed absence of necrosis, inflammation, or vascular or perineural invasion.1 The majority of cases of spindle cell foci of the thyroid also demonstrate positive staining with broad-spectrum cytokeratins and are usually negative with high-molecular weight cytokeratins, cytokeratin-19, smooth muscle actin, desmin, calcitonin, chromogranin, and synaptophysin.1 It is important to highlight that spindle cell foci of the thyroid can be negative with cytokeratin as in the present case. This should not lead to the erroneous conclusion that the cells in question are not follicular in origin, as they retain expression of TTF-1. Importantly, we have shown that all the cases of spindle cell foci of the thyroid have a low Ki-67 index (<1%).1 Both spindle cell foci of the thyroid and medullary thyroid carcinoma are positive for TTF-1, with the exception of some medullary thyroid carcinomas with spindle cell morphology that can lack TTF-1 expression.3 However, medullary thyroid carcinomas are reactive for calcitonin, CEA, and neuroendocrine markers in the great majority of the cases, whereas spindle cell foci of follicular origin are negative.4–6 Although it is well recognized that medullary thyroid carcinoma may lack immunohistochemical staining for CEA and calcitonin in a small subset of cases,7 it would be highly unusual for concurrent absence of staining for synaptophysin, as was seen in this case. The current case differed from previously reported cases of spindle cell foci of the thyroid in that there were prominent areas of fibrosis and sclerosis, mimicking amyloidosis, intermixed with areas showing “endocrine atypia.” Such “endocrine atypia” can include occasional large, irregular and angulated nuclei with coarse chromatin. Poor correlation exists between these histopathologic findings and aggressiveness of the lesion. Therefore, this feature should never be used as the sole criterion for defining malignancy. In summary, this case of spindle cell foci of the thyroid represents a variant of spindle cell metaplasia with marked nuclear pleomorphism and abundant intercellular fibrosis and sclerosis, mimicking spindle cell variant of medullary carcinoma. We emphasize the lack of morphologic features commonly seen in malignant lesions including lack of mitoses or apoptosis, absence of inflammation, necrosis, and vascular or perineural invasion. Because morphologic appearance by itself could be deceiving in the presence of uncommon lesions like the one presented here, we encourage the use of a panel of immunohistochemical markers including Ki-67 index as complementary diagnostic tools to avoid the misinterpretation as a malignant process. The authors declare no conflict of interest. LETTER TO THE EDITOR