Utilization of microdissection and the polymerase chain reaction for the diagnosis of adrenal cortical carcinoma in fine-needle aspiration cytology

BACKGROUND, Loss of heterozygosity (LOH) for several tumor suppressor genes (including loci on 3p, 1, and 17p,) has been documented in surgical specim ens of adrenal cortical carcinomas (ACCA) without accompanying losses in be nign hyperplastic and adenomatous adrenal cortical lesions (ACL). This disp arate pattern of LOH raises the possibility of exploitation of these differ ences for diagnostic utilization. Cytologic differentiation of benign versu s malignant ACL may be impossible based solely on fine-needle aspiration (F NA) material. The authors attempted to extrapolate the genetic findings on surgical specimens to FNA specimens of ACL to determine whether LOH studies could be utilized as a definitive diagnostic tool. METHODS. Microdissection of archival material was performed on FNAs of ten ACCAs (stained with the Papanicolaou and Diff-Quik stains) with correspondi ng histologic material (stained with hematoxylin and eosin), one FNA of a b enign ACL, and three touch preparations of benign adrenal cortex. LOH analy sis was performed by polymerase chain reaction (PCR) with flanking markers for the following putative tumor suppressor genes: p53 (17p13; TP53), 1p (1 p36; D1S165), and the von Hippel-Lindau gene at 3p25 (D3S1038 and D3S1110). RESULTS. Similar results were obtained with cytologic and histologic materi al. As expected, benign ACL showed no LOH for the markers examined. Of the informative ACCA cases, 70% showed LOH for at least 1 of the 3 markers test ed on both FNA and histologic samples. For all cases with amplifiable DNA, there was a 100% concordance rate for LOH between cytologic and histologic material, with at least 7 of the 10 cytologic samples originating from meta static lesions and all of the surgical material originating from the primar y adrenal neoplasm. CONCLUSIONS. The results of this study suggest that the combination of micr odissection and PCR for LOH of p5, 1p, and 3p25 from FNA material has the p otential to be utilized to distinguish ACCA from benign ACL in informative cases. It also shows a 100% concordance rate between metastatic and primary ACCAs for the losses observed, a finding that can be extremely useful for the definitive identification of metastatic lesions. Archival cytologic pre parations of ACCA are a reliable source of DNA for LOH studies. [See editor ial counterpoint on pages 173-5 and reply to counterpoint on pages 176-7, t his issue.] (C) 1999 American Cancer Society.