Pitfalls in fine needle aspiration cytology of extraadrenal paraganglioma. A report of 2 cases.

BACKGROUND: Fine needle aspiration (FNA) is commonly used as an initial diagnostic tool in the evaluation of mass lesions. However, extraadrenal paragangliomas (EAPs) are not commonly seen on FNA and, when encountered, can pose a diagnostic challenge since they have a wide variety of morphologic patterns. CASES: Two intraabdominal EAPs were misdiagnosed as an anaplastic adenocarcinoma of the pancreas with giant cell features and papillary adenocarcinoma of the pancreas. The sources of the pitfalls were inaccurate radiologic localization, misinterpretation of vascular-rich tissue fragments as papillary structures, presence of acinarmicroglandular structures, marked cellular pleomorphism (including multinucleation), prominent nucleoli and some cells with "squamoid" cytoplasm. CONCLUSION: The cytologic features of EAP, although suggestive, are not specific. A high index of suspicion and knowledge of clinical information, exact anatomic location and cytologic morphology combined with appropriate ancillary studies are the key to an accurate diagnosis.

Cytohistologic correlation of cirrhosis and hepatocellular carcinoma. Pitfall in diagnosis?

OBJECTIVE: To compare the diagnostic criteria for cirrhosis and hepatocellular carcinoma (HCC) noted on liver fine needle aspirates (FNAs) and their corresponding liver needle core biopsies (NCBs). STUDY DESIGN: We reviewed FNA slides from 15 cases of cirrhosis and 6 cases of HCC and their corresponding NCBs. We compared a variety of specific nonarchitectural criteria, including small cell dysplasia (SCD) and large cell dysplasia (LCD), for distinguishing cirrhosis from HCC. RESULTS: FNA smears diagnostically correlated with NCBs. The cytologic criterion with the greatest correlation in predicting HCC on FNA was SCD. This was not noted in all the core biopsies, probably due to sampling error. LCD was seen more frequently in cirrhosis than HCC on both cytology and histology and therefore was not a criterion useful in establishing a diagnosis of malignancy. The remaining cytologic criteria had good correlations but did not aid in diagnosing HCC. CONCLUSION: FNA has good cytohistologic correlation with NCB for both cirrhosis and HCC. There is an association of SCD with HCC; however, LCD is not a reliable "precancerous" change as it is commonly seen in cirrhosis and HCC. Therefore, the presence of SCD on FNA should be reported and is an indication for close clinical follow-up to exclude HCC.

Significance of extensive hyperkeratosis on cervical/vaginal smears.

OBJECTIVE: To assess the significance of reporting hyperkeratosis on cervical/vaginal (CV) smears. STUDY DESIGN: Cases diagnosed with extensive hyperkeratosis (E-HK) and without prior or concurrent history of neoplasia, squamous intraepithelial lesion or atypical squamous cells of undetermined significance (ASCUS) were retrieved from our files for the period January 1994-August 2001. E-HK is defined in our practice as patches of anucleated squames with irregular, angulated edges present in at least 5 low-power (10 x eyepiece and 10 x objective) fields on a conventional CV smear. On liquid-based preparations, we use 3 low-power fields. Only cases with a follow-up CV smear and/or cervical biopsy (CB) were selected. RESULTS: Among 328 cases of E-HK, 138 patients met the study selection criteria. Eighty-one cases had negative CV smears and/or CB, 17 (12.3%) patients had persistent E-HK, and a subsequent diagnosis of ASCUS or higher was made in 40 patients (28.9%). Among the 40 cases with subsequent abnormalities, 13 (9.4%) were diagnosed with ASCUS, 24 (17.4%) with HPV or dysplasia, and 3 (2.1%) with malignancy. CONCLUSION: While isolated, anucleate squames may have no clinical importance in patient management, E-HK can be a significant marker of underlying neoplastic disease. This should be kept in mind as one decides how to report CV cytology based on 2001 Bethesda System recommendations.

Cytologic features of granulosa cell tumors in fluids and fine needle aspiration specimens.

OBJECTIVE: To describe the cytologic features of granulosa cell tumors in fluids and fine needle aspiration specimens, with histologic confirmation. STUDY DESIGN: Histologically confirmed granulosa cell tumors, 6 adult type and 1 juvenile type, were identified. All patients had local recurrences or metastases. Eleven specimens from 7 patients, including cytologic samples, cell blocks and histology, were reviewed. Inhibin immunostaining was performed on cell blocks to aid identification of this group of tumors in the cytologic and histologic samples. RESULTS: The patients were 22-72 years old. Sites included ovary and peritoneum; there were pelvic recurrences and metastatic lesions in the spleen, liver, perirectum and cervical lymph node. Cytologic features of adult granulosa cell tumors included 3-dimensional clusters, resettes loose monolayers and individual cells. Other features were Call-Exner bodies, vacuolated cytoplasm, exuberant capillaries associated with papillarylike fronds, a second population of elongated theca cells, and prominent or rare nuclear grooves. In juvenile granulosa cell tumor the features observed were monolayers, loosely cohesive sheets, single cells, occasional larger pleomorphic cells with nuclear clefting and nuclear protrusions, vacuolated cytoplasm, finely granular chromatin and frequent mitoses. The overall cytologic and histologic correlation was good. Inhibin was focally positive in one peritoneal fluid, correlating with the focal pattern of staining seen on histology. CONCLUSION: A definitive cytologic diagnosis of granulosa cell tumor can be made based on the above criteria. Aggressive tumors are discohesive and show pleomorphism and nuclear protrusions. Inhibin stain may be helpful in identifying granulosa cell tumors in cell block specimens.

Fine-needle aspiration cytomorphology of sebaceous lymphadenoma of the salivary gland.

Sebaceous lymphadenoma (SLA) is a histologic variant of lymphadenoma or sebaceous adenoma of the salivary gland. The fine-needle aspiration (FNA) cytology of SLA has not been well-described. In this report, we reported a case that involved an 81-year-old male with a long-standing left parotid mass and was diagnosed on FNA cytology. The FNA smears showed clusters of three types of epithelial cells in a background of abundant lymphoid cells, macrophages and abundant proteinaceous materials. The predominant epithelial cells were large polygonal cells with abundant cytoplasm filled with multiple, uniform, small, and clear vacuoles, ill-defined cytoplasmic borders, and small centrally located round nuclei with finely granular chromatin, conspicuous nucleoli and indented nuclear membranes apparently imprinted by cytoplasmic vacuoles. These cells were surrounded by polygonal or flat cells with less or more dense cytoplasm, indistinct cell borders and round or oval small nuclei with smooth nuclear membranes, which correspond to germinative or basaloid cells. Some cells had granular cytoplasm. Large three dimensional clusters of nonkeratinized squamous cells have oval nuclei containing evenly distributed chromatin, and scant to moderate dense cytoplasm that were arranged in a "stream of fish" pattern. Rare granulomas and cystic contents (degenerated cells, inflammatory cells, macrophages, and abundant granular debris/proteinaceous material) were also seen. The diagnosis of SLA was confirmed by the surgical resection.

Cytomorphology of intraductal oncocytic papillary neoplasm of the liver.

We describe the first cytology case report of an intraductal oncocytic papillary neoplasm (IOPN) of the liver. A 51-year-old male presented with recurrent cholangitis. Magnetic resonance imaging and endoscopic retrograde cholangiopancreatogram revealed a 1.1 × 0.9 cm polypoid lesion within the left intrahepatic bile duct. Fine-needle aspiration and needle core biopsy (NCB) revealed nests, 3-dimensional or papillary clusters of columnar or cuboidal cells with loss of polarity. The nuclei were uniform with even chromatin, and cytoplasm was granular or vacuolated. No mitosis or necrosis was seen. The cytologic and histologic diagnosis was "consistent with Intraductal Oncocytic Papillary Neoplasm (IOPN), intermediate grade (borderline)." The patient then underwent a left lateral liver segmentectomy. Microscopic examination showed histology similar to the NCB with no stromal invasion identified. Hepatic IOPN poses a diagnostic challenge due to its broad differential diagnoses. Both malignant and non-malignant IOPNs may present with similar clinical symptoms, pathology, histology, cytomorphology, and immunohistochemistry. Hepatic IOPN should be excised as it is a precursor lesion of adenocarcinoma.

Thin core needle biopsy crush preparations in conjunction with fine-needle aspiration for the evaluation of thyroid nodules: a complementary approach.

BACKGROUND: Fine-needle aspiration (FNA) is widely accepted as the initial test to evaluate thyroid nodules; however, inadequate and suboptimal specimens have been 1 of its limitations. Unsatisfactory rates of 4.1% to 43% have been reported, but suboptimal specimens with adequate epithelial cells and other limiting factors, such as clotting, often are not addressed. The authors' institution has a low unsatisfactory rate, especially for thyroid biopsies performed under ultrasound in the Interventional Radiology (IR) Department. In addition to on-site evaluation for all cases, they concomitantly use thin, 22/20-gauge core needle biopsy (CB) crush preparations (CP) for unsatisfactory/suboptimal FNAs. The CB usually is done after 2 FNA passes and, in most cases, is exhausted by making an air-dried CP, which is evaluated on site for adequacy; any residual tissue is processed for tissue sections. Experience is required to interpret CP on air-dried smears. In this report, the authors describe a complementary approach to thyroid biopsy that has worked well. METHODS: All thyroid FNA and CB/CP that were performed in the IR Department during the year 2005 were reviewed. Follow-up histology and records of all procedural complications were retrieved. RESULTS: Seven hundred thirteen thyroid biopsies qualified, 225 biopsies (31%) had FNA with CB/CP (85% had only CP for evaluation), and 488 biopsies (69%) had only FNA. The final unsatisfactory rate in IR for FNA was 8.7%; this was reduced to 3.4% with the use of CB/CP. The addition of CB also helped to obtain a more definitive diagnosis in suboptimal FNA specimens. Cytologic-histologic correlation was comparable for FNA only cases and FNA/CB cases. There were no significant procedural complications in CB cases. CONCLUSIONS: FNA in conjunction with a thin CB/CP performed during the same procedure is a safe technique that can reduce the rate of unsatisfactory and suboptimal thyroid biopsy.

Microcystic adenoma of the pancreas: clinical, radiologic, and cytologic features.

BACKGROUND: Cytology literature on pancreatic microcystic adenoma is sparse. It is important to separate microcystic adenoma from adenocarcinoma and mucinous cystic neoplasms on aspiration cytology, because patients with microcystic adenoma can be treated conservatively unless they are symptomatic. Potential pitfalls with endoscopic ultrasound (EUS) sampling of these lesions is discussed. METHODS: From January 1991 through June 2003, 10 patients with microcystic adenoma of the pancreas were diagnosed on fine-needle aspiration cytology. An additional patient, who was diagnosed with a mucinous cystic neoplasm by EUS sampling, was rediagnosed with microcystic adenoma on the excised specimen. Aspirate smears, cell blocks, core biopsies, subsequent excision (if any), and special stains were reviewed. Imaging studies and clinical data were available from the majority of patients. CLINICAL RESULTS: The patients included 5 females and 6 males who ranged in age from 45 years to 84 years. Radiology studies showed tumors, which were heterogeneous with areas of fluid density and septations, located in the head, body, or tail of the pancreas. The masses ranged in size from 1 cm to 17 cm. The radiographic impression was highly suggestive of microcystic adenoma in six patients; detailed radiologic information was not available from three patients. On follow-up, six patients were alive and well at the last follow-up available, two patients died of unrelated sepsis, and three patients were lost to follow-up. Three of 11 patients underwent a Whipple resection. Cytology results: The cytologic features identified included the following: 1) Bland tumor cells were seen in sheets or small groups with a lack of nuclear abnormalities and moderate-to-scant cytoplasm with occasional clearing or vacuolation. Naked nuclei were present occasionally. Tumor cells were distinguishable from acinar cells based on larger cell size and granular cytoplasm in which prominent nucleoli were seen. 2) Relatively acellular, fibrovascular stroma was seen, usually located between tumor cells. 3) Calcifications were seen in four of eleven tumors. One tumor sampled by EUS revealed fragments of glandular-type epithelium with minimal atypia and was diagnosed erroneously as a mucinous cystic neoplasm. Cell blocks or core biopsies from most tumors showed fragments of dense stroma and cystic spaces lined by flattened epithelial cells. Subsequent Whipple resection in three patients showed histologic features of microcystic adenoma. Special stains performed in select tumors were positive for cytokeratin, carbohydrate antigen 19.9, and periodic acid-Schiff stain. Calretinin staining was negative in the tumor cells. CONCLUSIONS: A cytologic diagnosis of microcystic adenoma is possible based on the criteria described above. Cell block and/or core biopsy, special stains, and radiologic information are key in making a definitive diagnosis. Patients with microcystic adenoma are spared a major surgical procedure unless they are symptomatic. With the EUS-guided modality of pancreatic sampling, caution should be exercised in misinterpreting benign glandular epithelium derived from the stomach or small bowel as a mucinous cystic neoplasm.