Utility of high-molecular weight cytokeratin 34βE12 in atypical small acinar proliferations in prostatic needle biopsies

Prostatic needle biopsies may contain small acinar proliferations that are atypical but lack sufficient features for an unequivocal diagnosis of malignancy. The presence of a basal cell layer is considered to be an important criterion for benignity. Thus its immunohistochemical detection by the high-molecular weight cytokeratin antibody 34βE12 may be useful in further characterizing atypical small acinar proliferations in prostate biopsies. Of 2,031 biopsies between January 1990 and June 1997 at Albany Medical Center, 27 representative biopsies diagnosed as atypical were retrieved and selected for the study. One hematoxylin and eosin (H and E) and one cytokeratin 34βE12 stained slides were provided for each case, which were reviewed by a panel of five pathologists. The participants were asked to classify the lesions into one of four categories including favor benign/benign, atypical of uncertain significance, suspicious for carcinoma, and diagnostic of carcinoma, based initially on H and E slides and again after evaluation of 34βE12 staining. Each case was assigned a consensus diagnosis when three or more observers agreed. The participants had no knowledge of any clinical information and follow-up data. Consensus diagnosis was achieved in 20 of 27 (74%) cases after review of the 34βE12 staining compared to 18 of 27 (67%) initial consensus diagnoses on the basis of H and E alone. Nine (33%) diagnoses remained unchanged, whereas 15 (56%) were reclassified with the use of 34βE12, including eight cases that were finally judged diagnostic for carcinoma. The lesion was not represented in the immunostained slides in three cases. When the diagnoses of each participant were analyzed separately, a more definitive diagnosis was achieved with the use of 34βE12 in 72% of the cases initially diagnosed as atypical of uncertain significance. Of the cases that were suspicious for carcinoma based on H and E staining 85% were confirmed after review of 34βE12 immunostaining. Also of interest is the correlation between the average number of glands available for diagnosis and the diagnostic categories, ranging from three glands in the atypical cases to 10 glands in the carcinoma category. We conclude that although morphologic evaluation remains the gold standard, 34βE12 can be an extremely useful adjunct in further classifying atypical small acinar proliferations and in substantiating the diagnosis of malignancy in needle biopsies containing small, atypical foci.