False positive labeling of prostate cancer with high molecular weight cytokeratin: p63 a more specific immunomarker for basal cells.

Occasional nonspecific staining of prostate cancer cells with high molecular weight cytokeratin (HMWCK) can lead to false-negative diagnoses. We compared p63 and HMWCK immunostaining to check their specificity for basal cell identification. Out of 6887 prostate cancer cases sent in consultation to one of the authors over 1.5 years, we identified 22 (0.3%) cases with HMWCK labeling of cancer cells, including 20 needle biopsies and 2 transurethral resections of prostate (TURP). Cases were sent in consultation because of the confusing immunostaining pattern, where prostate cancer cells labeled with HMWCK at the outside institutions. In 6 cases, p63 immunostains were also received from the outside institution, whereas in the remaining 16 cases p63 immunohistochemistry was performed at our institution. In 14 cases, we used either an extra destained hematoxylin and eosin slide or a negative control slide for immunohistochemistry with antibodies to p63, and in the 2 remaining cases submitted unstained slides were used. The Gleason scores were 3+3=6 in 20 cases and 4+4=8 in 2 cases. The size of the tumor on needle biopsy ranged from 0.5 to 6.0 mm (mean 1 mm) and on the 2 TURP cases consisted of 44 and 68 cancer glands, respectively. The number of tumor cells positive for HMWCK in each of the needle biopsy cases ranged from 3 to 48 (mean 13 cells), whereas on the 2 TURP cases 26 and 10 cells were labeled with HMWCK. Corresponding stains for p63 on the same cases were negative in 18 cases. In 3 of 4 cases, p63 labeled 1, 1, and 2 tumor cells, respectively. The fourth case had 5 positive cells on p63 staining with 4 positive for HMWCK. To assess whether overstaining was a factor, we evaluated the intensity of HMWCK staining in the basal cells of the benign glands, which was moderate in 6 and strong in 16 cases. The cytoplasm of benign secretory cells showed focal weak (n=3), diffuse weak (n=1), and focal moderate (n=2) staining for HMWCK. HMWCK labeling of prostate cancer cells is uncommon and does not seem to be solely attributable to overstaining. p63 is a more specific marker for basal cells than HMWCK, with less labeling of tumor cells. Recognition of this phenomenon and performing stains for p63 when it occurs can help prevent underdiagnosing prostatic carcinoma.