Female PSA Predicts Breast Cancer Therapy Outcome

By Barbara Biedrzycki, RN, MSN, AOCN, CRNP

Summary—While the screening and diagnostic values of prostate specific antigen (PSA) levels for prostate cancer have been controversial since their conception, scientists have determined that a prostate gland is not needed to produce measurable levels of PSA. Women with breast cancer have been found to have measurable PSA. Also, in half of the women who received megestrol acetate therapy for breast cancer, PSA levels and mortality rates significantly increased.

When investigators found minute amounts of PSA in women’s plasma, they were challenged to identify its significance. Researchers have been aware of this finding since 1996, although the clinical implications have not been rigorously explored until now.1 Although not detectable by the conventional PSA assay methods reporting levels as nanograms per milliliter (ng/mL), up to 50% of female serum contains traces of PSA detected by more highly sensitive assays that report in nanograms per liter (ng/L).

While the source of PSA in female serum is unknown, research suggests it is under the influence of steroids, especially androgens and progestins. PSA has been discovered in female body fluids and organs, including the breast. In lab studies, androgen and progestins up-regulate the production of PSA in breast cells.2

Studies have related increased PSA levels in women with the following:

    • excessive androgen and hirsutism;3
    • pregnancy;4
    • progesterone changes during the menstrual cycle.5

PSA, a glycoprotein previously believed to be produced only in the cytoplasm of benign and malignant prostate cells, has been investigated over the years to improve its diagnostic and prognostic capabilities for the detection and management of prostate cancer.

While not specific for prostate cancer, PSA values in men overlap with benign prostatic hyperplasia and prostate cancer. The test performance of a PSA > 4 ng/mL is 0.67 sensitivity, 0.97 specificity, and 0.43 positive predictive value for screening for prostate cancer.

Measuring PSA velocity, or serial measurements of PSA, looks for changes in PSA > 0.75 ng/mL per year to increase the likelihood of cancer detection. Analysis of free serum and protein-bound levels of PSA are being studied to determine optimal cutoff levels while maximizing sensitivity and specificity.6

Study Methodology

Seventy-six women with locally advanced or metastatic breast cancer participated in this research study to measure PSA levels with highly sensitive assays able to detect PSA at a level of 1 ng/L while they received one of three therapeutic modalities. Fifty-two women received varying doses of megestrol acetate, a synthetic progestin used as hormonal therapy. Fourteen received tamoxifen, an antiestrogen agent, and 10 received cytotoxic anthracycline doxorubicin. Of the women who had disease progression while receiving megestrol acetate, 14 stopped the treatment and were followed to assess the reversibility of increased PSA levels.

Researchers planned to examine the effect of therapy on PSA levels by obtaining those levels pre-treatment and throughout therapy. They intended to determine whether changes would be associated with clinical outcomes. Study investigators examined two main comparisons. First, changes in PSA levels among women receiving the three treatments were explored. Second, variables and clinical outcomes of women in the megestrol acetate treatment group both with and without elevated PSA levels were analyzed.

Study Results

The significant level of PSA change was arbitrarily set by the researchers at 5 ng/L above baseline. This was considered appropriate, as most women had a baseline of less than 3 ng/L, and the assay could detect changes less than 1 ng/L.

The group of 52 women treated with megestrol acetate was divided in half by statistically significant increases in PSA levels. Although researchers had predetermined that changes of 5 ng/L would be considered significant, the data provided dramatic results.

Researchers found the baseline PSA levels were low, with 68% being less than 1 ng/L. Of the 26 women who had PSA level increases, 12% (3) were >1,000 ng/L above baseline, 27% (7) were >100 ng/L, 38% (10) were >10 ng/L, and 23% (6) were in the 5-10 ng/L range. This meant that PSA changes after megestrol acetate treatment reached the highly significant level with p < 0.0001. None of the women treated with tamoxifen or with anthracycline doxorubicin showed an increase in their PSA levels during therapy.

Fourteen women whose disease progressed on megestrol acetate therapy and who had an increase in PSA levels were followed to see if the increase in PSA levels were reversible after termination of therapy. After megestrol acetate was stopped for 27-121 days, none of the women had a higher PSA level than when on therapy. In fact, the trend of PSA level reduction was significant at the P < 0.01 level.

A statistically significant relationship (P = 0.03) was found between increased PSA levels and increased mortality. The death rate was 58% in the group with increased PSA levels, compared with 28% in the stable PSA group. However, in the multivariate analysis adjusted for all variables, the effect of the PSA level on the death rate was not statistically significant. The analysis suggests that an association of increased PSA levels and the risk of death may be influenced by the type of metastases, local or distant.

The researchers concluded that "megestrol acetate treatment induces PSA production in approximately 50% of breast cancer patients and that this induction, identified by measuring plasma PSA, is associated with increased patient mortality."1 The researchers acknowledge that they have no explanation for the mechanism of this PSA stimulation and increased risk of death. They recommend additional research to determine if megestrol acetate therapy is the most appropriate therapy for these patients.

Implications for Practice

While this research certainly stimulates more questions than the scientific conclusions produced, it challenges us to "see outside the box," to be open to new data defying previously held scientifically based medical beliefs. The importance of keeping informed about relevant research can not be overemphasized.

This landmark study shows that PSA potentially may be used as a prognostic biochemical marker for women with metastatic breast cancer. Astute advanced practice nurses, recognizing that "prostate specific antigen" is not only a misnomer, but, perhaps a very important element leading to the cure for breast cancer, will be alert for future research reports on PSA.


1. Daimandis EP, Helle SI, Yu H, et al. Prognostic value of plasma prostate specific antigen after megestrol acetate treatment in patients with metastatic breast carcinoma. Cancer 1999;85:891-98.

2. Yu H, Diamandis EP, Zarghami N, et al. Induction of prostate specific antigen production by steroids and tamoxifen in breast cancer cell lines. Breast Cancer Res Treat 1994;32:301-10.

3. Melegos D, Yu H, Ashok M, et al. Prostate specific antigen in female serum — a potential new marker of androgen excess. J Clin Endocrinol Metab 1997; 82:777-80.

4. Melegos DN, Yu H, Allen LC, et al. Prostate specific antigen in amniotic fluid of normal and abnormal pregnancies. Clin Biochem 1996;29:555-62.

5. Zarghami N, Grass L, Sauter ER, et al. Prostate specific antigen levels in serum during the menstrual cycle. Clin Chem 1997;43:1862-67.

6. Presti JC, Stoller ML, Carroll PR. Urology In: Current Medical Diagnosis and Treatment 1999. Stamford, CT: Appleton & Lange; 1999:894-931.