Affiliations 

  • 1 Department of Pathology, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor
  • 2 Departments of Community Health, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor
  • 3 Departments of Pathology, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor
  • 4 Department of Pathology, Faculty of Medicine, Hospital Universiti Kebangsaan Malaysia
Med J Malaysia, 2007 Mar;62(1):36-9.
PMID: 17682568 MyJurnal

Abstract

The diagnosis of prostatic carcinoma (Pca) on routine biopsies may be challenging, and to date the commonly used marker to distinguish prostate carcinoma from benign prostatic lesions has been High Molecular Weight-Cytokeratin (HMW-CK). However, the antigen of HMW-CK is susceptible to the effect of formalin fixation and causes frequent loss or patchy staining in the obviously benign glands. More recently, antibodies to p63 have been reported to be more sensitive than HMW-CK for the detection of prostatic basal cells. p63, a homologue of tumour suppressor gene p53, is essential for prostate development and is selectively expressed in the nuclei of basal cells of normal prostate glands. The objective of this study is to compare the sensitivity and specificity of HMW-CK and p63 in distinguishing prostatic carcinomas from benign prostatic lesions, as well as determining their positive predictive values. Seventy-two cases from HUKM (comprising 29 prostatic carcinomas and 43 benign prostatic hyperplasias) were stained for both HMW-CK and p63. The sensitivity of p63 and HMW-CK in identifying basal cells in benign glands was 88.37% and 90.70% respectively. The specificity of both reagents was 100%, and the positive predictive value for both reagents was also 100%. Thus, p63 is a useful complementary basal cell specific stain to HMW-CK, and would be very helpful to practicing pathologists in dealing with difficult cases.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.