The thyroid fine-needle aspiration (FNA) diagnosis of "follicular neoplasm" does not differentiate between a benign and malignant tumor. Often cases diagnosed as "follicular or Hürthle-cell neoplasm" undergo surgical excision for further characterization. The aim of this study was to identify clinical features that may help in predicting malignancy in patients with an FNA diagnosis of follicular neoplasm. One hundred eighty-four cases in 167 patients were diagnosed as "follicular neoplasm" among 1,024 thyroid FNA evaluated with on-site interpretation from 1998-2000. The cases were evaluated for the following variables: histologic diagnosis, age, sex, and size of the nodule. One hundred thirty-nine patients were female, and 28 were male (age range, 23-80 yr). Among 122 patients (67%) undergoing surgical excision (lobectomy, 96; total thyroidectomy, 26), malignancy was identified in 37 cases (31%). Nonpapillary (follicular/Hürthle) carcinoma was diagnosed in 11 (9%), follicular variant of papillary carcinoma in 25, and medullary carcinoma in 1 case. The risk of malignancy was greater in males (47% vs. 29%, P < 0.0004) than females, in nodules measuring 3 cm or more (55% vs. 23%, P < 0.0001), than in nodules measuring less than 3 cm, and in patients 40 or more yr old (20% vs. 10%, P = 0.0001) than in patients younger than 40 years. The diagnosis "follicular neoplasm" is indeterminate, and the majority of cases (70% in the current study) are benign. However, clinical features, including gender, nodule size, and age, can be a part of the decision analysis in selecting patients for surgery.

In this report we describe our institutional experience with fine-needle aspiration (FNA) of the thyroid. Six hundred sixty-two FNAs were performed in 616 patients in a 3 1/2 year period. The cytological diagnoses were categorized as: negative for malignancy, 455 (69%); indeterminate, 30 (4%); neoplasm/malignant, 105 (16%), and nondiagnostic, 72 (11%) including 29 cases from outside institutions. Surgical follow-up was available in 140 (21%) cases, 95 of which had preceding cytological diagnoses of positive or indeterminate for neoplasm/malignancy. For those cases with definite benign or neoplastic/malignant cytodiagnosis, a sensitivity of 92% and specificity of 84% was achieved. The cumulative false-positive and false-negative rates for these categories were 16% and 7.5%, respectively. On comparing discrepant diagnosis between cytological and histological specimens, two major factors were identified as causes for diagnostic misinterpretations: overlapping cytological features among follicular-derived lesions and inadequate/suboptimal specimens. The cytological features of follicular variant of papillary carcinoma were found to overlap those of hyperplastic/adenomatous nodules and follicular neoplasms due to the presence of abundant thin colloid, monolayer sheets of follicular cells and subtle nuclear features of papillary carcinoma. We suggest that awareness of variable cytological features in follicular lesions (especially in follicular variant of papillary carcinoma), following strict criteria of specimen adequacy in thyroid FNA, and clinicopathological correlation can markedly reduce false-negative results.

A b s t r a c tThe cytologic diagnosis of follicular variant of papillary thyroid carcinoma (FVPTC)

The current FDA-approved standard of care for nonsmall cell lung cancer is Carboplastin/Taxol/Avastin based upon an impressive survival benefit; however, patients with squamous carcinoma (SQCC) cannot receive Avastin because of a 30% mortality rate due to fatal hemoptysis. In this study we evaluated the role of cytomorphology and immunohistochemistry in differentiating SQCC from adenocarcinoma (ADC) in lung FNA specimens. The case cohort included 53 FNA cases of nonsmall cell lung carcinoma with surgical pathology follow-up. All FNA specimens were reviewed independently by a panel of cytopathologists to differentiate between SQCC and ADC. The cell block material was available in 23 cases (11 ADC and 12 SQCC) to perform immunohistochemical stains for TTF-1, CK7, CK20, P63, and CK5/6. On surgical resection, 35/53 (66%) cases were diagnosed as ADC and 18/53 (34%) as SQCC. The number of cases classified correctly on the basis of cytomorphology was 66% for ADC and 53% for SQCC (combined accuracy 60%). By immunohistochemical staining, 14/23 (61%) cases expressed TTF-1. Nine cases were TTF-1 negative; eight of the TTF-1 negative cases (89%) were SQCC. Twenty-three cases expressed CK7 (87%); one ADC case (4%) showed focal CK20 positivity. Both P63 and CK5/6 expression was seen in 9/12 (75%) SQCC cases; none of the ADC cases showed this dual expression. Cytomorphology alone may not be able to stratify all cases of nonsmall cell lung carcinoma into ADC and SQCC in FNA specimens. The immune-panel of TTF-1, CK7, CK20, P63, and CK5/6 is useful in differentiating SQCC from ADC.

The purpose of the present study was to determine the role of repeat fine-needle aspiration biopsy (FNAB) in the evaluation of thyroid nodules initially classified as "nondiagnostic" due to limited cellularity or as "indeterminate for neoplasm." We reviewed a cohort of 431 patients (352 females, 79 males; average age 50 yr); 237 patients were classified as "nondiagnostic" due to limited cellularity and 194 as "indeterminate for neoplasm" over a 3-yr period (1999-2002). Repeat FNAB under ultrasound guidance was performed in 226 patients (226/431, 52%); surgical pathology results were available in 101 patients. Repeat FNAB diagnoses were: benign 70 (31%), follicular/Hürthle cell neoplasm 62 (27%), suspicious for papillary carcinoma 25 (12%), malignant 17 (7%), and nondiagnostic 52 (23%) cases. Surgical follow-up was available in 101 (45%) patients; malignancy was identified in 50 (49%) patients. The malignancy rate was 51% and 48% in cases in which initial FNAB was nondiagnostic and indeterminate for neoplasm, respectively. There were no false-positives and all malignant cases undergoing surgery were found to be malignant. This study demonstrates that repeat FNAB is warranted in patients with thyroid nodules diagnosed on initial FNAB as nondiagnostic and indeterminate for neoplasm since it can yield a definitive diagnosis in the majority of cases with an overall malignancy rate of 49%.