Polyclonal Anti-PSA Is More Sensitive but Less Specific Than Mon
Prostate-specific antigen (PSA) production by nonprostatic tissues has been reported, casting doubts on its specificity. The immunohistochemical relative specificity and sensitivity of PSA expression using monoclonal and polyclonal anti-PSA was analyzed on 60 prostate carcinomas, 40 normal seminal vesicles, and 310 nonprostatic tumors. All nonprostatic tumors proved negative with both antibodies. However, 13 (32%) seminal vesicles showed immunoreactivity with polyclonal anti-PSA, but none showed immunoreactivity with the monoclonal antibody. The sensitivity of the 2 antibodies for prostate cancer varied with tumor grade. In Gleason pattern 3, both antibodies showed diffuse immunostaining in all cases. In Gleason pattern 5, polyclonal anti-PSA showed diffuse (>95%) tumor cell positivity in 18 cases (90%), while with the monoclonal antibody, 7 cases (35%) showed only focal (<10%) tumor cell immunoreactivity. Thus, monoclonal anti-PSA seems to be useful in small gland proliferations in which the differential diagnosis includes seminal vesicle, while for poorly differentiated neoplasms, polyclonal anti-PSA is considered superior. Sections of high-grade prostate cancer should be included as positive controls for PSA immunostaining.

Immunohistochemical analysis using antibodies directed against prostate-specific antigen (PSA), a secretory product of prostatic epithelium, is widely used to establish the prostatic origin of metastatic adenocarcinoma in diagnostic histopathology practice. This has important clinical relevance, as the confirmation of prostatic origin would result in the patient being administered specific treatment, which is hormonal manipulation in most instances. PSA also is useful to distinguish seminal vesicle epithelium from prostatic cancer in needle biopsy specimens, a differential diagnosis of small acinar proliferation that is complicated further by the fact that pigment resembling that of the seminal vesicle may be seen in benign and malignant prostatic epithelium. PSA also is used to distinguish prostatic adenocarcinomas from other nonprostatic small acinar proliferations or normal histoanatomic variants such as Cowper glands, nephrogenic adenoma, and hyperplastic mesonephric remnants, as well as adenocarcinomas of rectum, urethra, and urinary bladder, which may have very similar histologic features and manifest in a pelvic location.

The use of PSA immunohistochemical analysis in the aforementioned clinical settings is based on the premise that PSA is specific for prostatic origin. However, there have been several individual case reports and small series documenting PSA immunoreactivity in a variety of nonprostatic tumors, including salivary gland neoplasms, carcinoma of the breast, urothelial (transitional cell) carcinoma, adenocarcinoma of the urinary bladder, colonic adenocarcinoma, and lung carcinoma. PSA immunoreactivity also has been reported in benign seminal vesicle epithelium.

PSA immunoreactivity can be demonstrated using polyclonal and monoclonal antibodies that may differ in their sensitivity and specificity for PSA. Traditionally, polyclonal anti-PSA is used widely in the United States, while several major centers in the United Kingdom use the monoclonal antibody. However, this choice of antibody type does not seem to be based on any detailed comparison of the 2 antibodies. Previous studies evaluating PSA immunohistochemical analysis generally have focused on either the sensitivity or the specificity of immunostaining, but not both together. However, a very specific technique may lack sensitivity, while improvements in sensitivity may be at the expense of specificity.

The aim of the present study was to systematically analyze the immunohistochemical specificity and sensitivity of PSA using both monoclonal and polyclonal anti-PSA applied with a contemporary antigen-retrieval method on a large series of cases spanning the range of prostatic carcinoma differentiation and in a diverse group of nonprostatic tissues, including those with reported cross-reactivity for PSA and tumors likely to be in the differential diagnosis of prostate cancer such as carcinoid tumors and urothelial carcinoma.