Follicular tumour in fine needle aspiration byopsy of thyroid: predictive factors of mallignancy
Alexandra Vieira1, Francisco Carrilho1, Cristina Ribeiro1, Sandra Paiva1, Maria João Martins2, Graça Fernandes2, Jacinta Santos1, Mariana Martinho3, Márcia Alves1, Sofia Gouveia1, Fernanda Xavier da Cunha2 & Manuela Carvalheiro1
Introduction: About 20% of fine needle aspiration biopsy of thyroid (FNA) with result of follicular tumour (FT) is malignant. Several factors have been suggested as indicators of malignancy.
Objectives: To determine predictive factors of malignancy in FNA with result of FT.
Methods: We evaluated retrospectively 140 clinical files of patients with cytology (ultrasound-guided or palpation) of FT. Presence of relationship between malignancy (in histology) and age, sex, family history, cervical symptoms (choking, dysphagia, dysphonia, tightness, discomfort, volume increase), multinodular goiter (MG), palpable nodules/adenopathys, scintigraphic/sonographic features, previous benign cytology, richness of Hurthle cells was investigated.
Results: In 40 clinical cases it wasnt possible histological classification (2 with uncertain malignant potential, 38 non-operated). Of the 100 remaining, 27 were malignant: papillary microcarcinoma (9), minimally invasive follicular carcinoma (4), follicular variant of papillary carcinoma (4), oxyphilic variant of papillary carcinoma (2), multifocal papillary carcinoma (2), multifocal papillary microcarcinoma (2), medullary microcarcinoma (1), papillary carcinoma (1), columnar cell variant of papillary carcinoma (1), poorly differentiated thyroid carcinoma (1). There was no statistically significant relationship between malignancy (excluding microcarcinoma or not) and age, sex, family history, autoimmune thyroiditis, MG, palpable nodules, size/number of nodules, calcifications, nodule solid/heterogeneous, nodule growth (≥3 mm), adenopathys in ultrasonography, scintigraphic result, richness of Hurthle cell, previous cytology benign.
The 2 patients with palpable adenopathys had papillary carcinoma.
There was statistically significant relationship between cervical symptoms and malignancy (P=0.039, OR=2.55; P=0.196 excluding microcarcinoma) as well as between malignancy and ill-defined nodule on ultrasonography (P=0.020, OR=5.93; P=0.05, OR=5.33 excluding microcarcinoma). Many sonographic data are insufficient not allowing for statistical analysis.
Conclusions: Predictive factors of malignancy were: presence of cervical symptoms and ill-defined nodule on ultrasonography. Malignancy prevalence (18% if we exclude the papillary microcarcinoma) was similar to that described in literature. These data reinforce the need for surgery for definitive diagnosis.