Utility of cell blocks in the diagnosis of thyroid aspirates.

Cell blocks (CBs) are often prepared with fine-needle aspirates (FNAs) from multiple organs as an adjunct to smears in the diagnosis of aspirated lesions. However, the literature contains few reports on their utility with regard to specific organ sites. At our institution, CBs are made routinely on FNAs when there is sufficient material remaining after smear preparation, with thyroid representing the largest volume. The aim of this study was to determine the utility of CBs in the diagnosis of thyroid lesions. From January 2002 to April 2004, 546 thyroid FNAs were performed. Eighty-two (15%) cases, from 60 females and 20 males (age range, 17-88 yr; mean, 50 yr), had CBs and formed the basis of this study. Seventy-four (90%) of the cases were performed by the radiologist or the clinician and 8 (10%) by the pathologist, all of which had an immediate assessment for adequacy. One to 7 passes were performed with an average of 3/case. The needles were immediately rinsed in Hanks' Balanced Salt Solution after smear preparation. CBs were made on bloody specimens/those with tissue fragments. Cell-block slides were reviewed for the presence of cellular elements and classified into three categories: (1) contributory, (2) noncontributory, or (3) provides additional information. Of the 82 cases, 23 (28%) were neoplastic, 51 (62%) were nonneoplastic, and 8 (10%) were nondiagnostic. Fifteen of the neoplastic cases had confirmatory biopsies, 9 of which were papillary carcinoma. The overall cellularity of the CBs was low, varying from 0 to 2 follicular groups in the noncontributory CBs and 3 to 6 follicular groups or papillary formations in the contributory CBs. CBs were contributory in 25 (31%) cases: 5 neoplastic (1 follicular neoplasm, 3 papillary carcinoma, and 1 suspicious for papillary carcinoma), 18 nonneoplastic, and 2 nondiagnostic. CBs were noncontributory in 56 (68%) cases: 18 neoplastic (4 papillary carcinomas, 1 suspicious for papillary carcinoma, 4 Hürthle cell neoplasms, and 9 follicular neoplasms), 33 nonneoplastic, and 5 nondiagnostic. One case was categorized as provided additional information because the CB showed material that was not present on the slides; however, it was still nondiagnostic. In summary, CBs did not help in the majority of cases. They were contributory in only 25 (31%) of the 82 cases, and of the 23 neoplastic cases, only 5 (22%) CBs were contributory. The contribution of the CBs in the diagnosis of thyroid lesions was minimal because of the low cellularity. On-site assessment of specimen adequacy often results in fewer passes, thus contributing to the low cellularity present in cell-block preparations. Ancillary studies may require additional passes.