Low Value of Color Flow-Doppler in Predicting Malignancy of Thyroid Follicular Neoplasms Dear Dr. Bedrossian: Follicular neoplasms (FN) represent the grey area of thyroid cytology in which only histologic evaluation of capsular and vascular invasion can discriminate malignant from benign tumors. As a consequence, in the presence of nodules cytologically classified as FN, thyroidectomy is warranted. As only 20% of FN are found to be malignant, the identification of parameters associated with malig- nancy is of particular importance 1 ; in the literature, how- ever, only a few papers have focused on the analysis of ultrasound characteristics. 2–4 The aim of this study was to evaluate the efficacy of color flow-doppler (CFD) charac- teristics in predicting malignancy in a series of 93 nodules cytologically classified as FN, of which 18 with Hu ¨rthle cells. All the patients (73 females and 20 males, mean age 50 6 15 yr) underwent total thyroidectomy. Patients with toxic or pretoxic nodules, as assessed by TSH, FT 3 , FT 4 serum levels and scintiscan, and with serum Calcitonin > 10 ng/L were not included in the series. Thyroid ultra- sonography was performed by the same endocrinologist (PT) using an ESAOTE MyLab60 system equipped with an 8-14 MHz linear probe, and all clips were stored and blindly reviewed by the second author (SS). The pattern of nodular CFD was defined as follows: CFD-1, absent signal; CFD-2, spot or complete perinodular signal; CFD- 3, intranodular signal (Fig. 1). Cytologic and histologic specimens were evaluated by two different cytopatholog- ists and only concordant diagnoses were included in the study. Cytologic samples were classified according to British Thyroid Association Thyroid Cancer Guidelines. Follicular neoplasms were defined using the following cytologic criteria: monotonous cellular population, scant colloid, and lack of features of papillary carcinoma. 5 All cytologic specimens that met these criteria were included in the indeterminate category (THY-3) of the above guidelines. Color flow-doppler characteristics were com- pared by v 2 test or Fisher exact test, and the statistical significance was set at P < 0.05. Sensitivity, specificity, and accuracy were obtained using the Galen and Gambino predictivity test. Histologic diagnosis showed that 15 (16.1%) nodules harbored cancer (9 follicular variant of papillary, 2 follic- ular, 3 Hu ¨rthle, 1 medullary), while 78 (83.9%) were benign (65 nodular hyperplasias, 6 follicular adenomas, 5 Hashimoto hyperplasias, 2 oncocitic adenomas). Regard- ing CFD, 1/5 (20%) of nodules with CFD-1, 7/44 (15.9%) of nodules with CFD-2, and 7/44 (15.9%) of nodules with CFD-3 were malignant. Comparison between malignant and benign tumors showed no difference in CFD classes. Analysis of sensitivity, specificity, and accuracy showed a low sensitivity for CFD-1, CFD-2, and CFD-3, whereas CFD-1 had higher specificity and accuracy (Table I). Dif- ference was recorded neither between FN with Hu ¨rthle cells and without Hu ¨rthle cells, nor between different can- cer histologic types. These data showed that CFD was not associated with cancer and that absent vascular signal had the higher specificity and accuracy in predicting malignancy. The latter is not in agreement with what has previously been Fig. 1. Schematic illustration of the color flow-doppler (CFD) character- istics. The description of the classes is detailed in the manuscript. *Correspondence to: Salvatore Sorrenti, Ph.D., Department of Surgical Science, ‘‘Sapienza’’ University of Rome, Viale del Policlinico, 155, Rome 00161, Italy. E-mail: salvatore.sorrenti@uniroma1.it Received 15 November 2008; Accepted 18 December 2008 DOI 10.1002/dc.21050 Published online 13 February 2009 in Wiley InterScience (www. interscience.wiley.com). ' 2009 WILEY-LISS, INC. Diagnostic Cytopathology, Vol 37, No 5 391