Complexed PSA Test More Accurate Than Traditional PSA Test in Detecting Prostate Cancer

Study Indicates cPSA Test More Predictive of Cancer, Results in Fewer False Diagnoses and Unnecessary Biopsies

PSA Rising - October 11, 2003: A new study on detection of prostate cancer finds that measurement of complexed PSA (cPSA, or serum concentration of complexed prostate specific antigen) is more accurate than the total PSA (tPSA) test currently used by most physicians.

The study, published in the November issue of The Journal of Urology, showed that the cPSA test is more predictive of prostate cancer and results in fewer false diagnoses and unnecessary biopsies compared to the traditional tPSA test.

"These findings are important because they suggest that use of the cPSA test can decrease the number of unnecessary prostate cancer biopsies particularly if a lower PSA threshold is utilized. This approach will not only save healthcare expenditures but will also spare many men the discomfort and anxiety associated with biopsy procedures," said investigator Richard Babaian, M.D., from MD Anderson Cancer Center. "In view of these findings, physicians and patients should consider using the cPSA test when screening for prostate cancer."

PSA, a protein produced in the prostate, has proven to be an extremely useful marker for early detection of prostate cancer and in checking for disease progression and the effects of treatment. PSA serum levels of 4.0 ng/ml or less are usually considered normal; higher levels (4 to 10 ng/ml or higher) are often found in men with prostate cancer. However, current PSA testing generates up to 60 percent "false positive" diagnoses because PSA levels can also increase due to enlargement of the prostate, a non-cancerous condition increasingly common as men get older, as well as acute infections of the prostate (prostatitis), and other factors. On the other hand, testing can also generate "false negatives" because a significant number of cases of prostate cancer have been found in men whose PSA was "normal," between 2.5 to 4 ng/ml.(ii)

The ability of a test to detect cancer is defined as sensitivity; the ability to detect those without cancer is defined as specificity. Sensitivity can be increased by lowering the tPSA "cutoff" value at which a physician suspects cancer, but doing so will decrease specificity, since it will also increase the number of men suspected of cancer but who don't actually have it.

Therefore, much research has focused on ways to improve the accuracy of PSA testing, i.e., to increase both sensitivity and specificity.

Some studies have shown that the cPSA test is equivalent to the tPSA test while others have shown it is better than tPSA. Other PSA tests include those that measure free (vs. complexed) PSA and the percentage of tPSA comprised of free PSA and of cPSA.

Study and Findings

In this study, conducted at seven prostate cancer treatment centers, blood samples were collected from 831 men who were about to undergo initial prostate biopsy. The biopsies showed that 313 (37.5 percent) had prostate cancer.

Knowing which patients turned out actually to have prostate cancer, the investigators compared the accuracy of the predictions of the cPSA and tPSA tests. They found that cPSA was significantly more predictive of cancer than tPSA, and provided an improvement in specificity (reduced false positives) compared to tPSA. For example, for men whose tPSA levels were in range of 2 to 4 ng/ml (cPSA range 1.5 to 3.2 ng/ml), using a cutoff of 2.2 ng/ml for cPSA and 2.5 ng/ml for tPSA resulted in both tests correctly identifying cancer 85 percent of the time, which is considered highly sensitive.

At this 85 percent sensitivity level, the cPSA test correctly identified 35 percent of patients who did not have cancer, compared to 21.2 percent for the tPSA test. So the cPSA test would have spared significantly more men -- 14 percent -- who did not have cancer from undering an unnecessary biopsy and the inconvenience and anxiety associated with that procedure.

"Urologists are increasingly looking at the lower range of PSA values to maximize the detection of prostate cancer," said Dr. Babaian. "Since this lower range also includes many men without cancer, the enhanced specificity of the cPSA test to detect those who are cancer free and avoid unnecessary biopsies is an important advantage over the standard tPSA test."

The investigators compared the predictive value of the tests in other ranges as well. For the range of 2 to 10 ng/ml tPSA (1.5 to 8.3 ng/ml for cPSA), the cutoffs of 3.37 ng/ml tPSA and 2.87ng/ml cPSA provided 85 percent sensitivity, but the cPSA test correctly identified 35 percent of patients who did not have cancer, compared to 29 percent with the tPSA test -- saving unnecessary biopsies in six percent more of the men without cancer, also statistically significant.

Other recent studies in Germany and Japan (see below) arrived at somewhat different conclusions by comparing the array of tests available today. One thing clear is, patients do have more chance today of receiving more accurate tests for prostate cancer. Chances of avoiding unnecessary biopsies while finding significant tumors are improving.

Prostate cancer is the second leading cause of male cancer death in the U.S. The American Cancer Society estimates that 220,900 men will be diagnosed with prostate cancer this year in the U.S., with 28,900 deaths attributable to prostate cancer.


J Urol 2003; (Volume 170, November 2003) Complexed PSA improves specificity for prostate cancer detection: results of a prospective multicenter clinical trial. Partin AW, Brawer MK, Bartsch G, Hominger W, Taneja S, Lepor H, Babaian RJ, Childs S, Stamey T, Fritsche HA, Sokoll LK, Chan DW, Thiel RP, Cheli CD.

J Urol. 2003 Oct;170(4 Pt 1):1175-9. A multicenter clinical trial on the use of complexed prostate specific antigen in low prostate specific antigen concentrations. Lein M, Kwiatkowski M, Semjonow A, Luboldt HJ, Hammerer P, Stephan C, Klevecka V, Taymoorian K, Schnorr D, Recker F, Loening SA, Jung K. Department of Urology, University Hospital Charite, Humboldt University Berlin, Germany.

"The areas under the curves (AUC) for cPSA, tPSA and free-to-total PSA (f/tPSA) showed no significant differences in the tPSA ranges of 0 to 6 (700 cases), 0 to 4 (510) and 0.5 to 2.5 ng/ml (253). Within the tPSA range of 2.5 to 4 ng/ml (230 cases) the AUC for cPSA (0.61) was significantly larger than that for tPSA (Roche 0.51, Bayer 0.54) but did not differ from the AUC of f/tPSA (Roche). On the basis of the cutoffs for 95% specificity or sensitivity, no significant increase in corresponding sensitivity or specificity was found between tPSA with cPSA. CONCLUSIONS: In the tPSA range of less than 4 ng/ml no improvement in diagnostic accuracy was shown between cPSA with tPSA or the ratio of f/tPSA. The search for a useful marker in the low PSA range must continue."

Int J Urol. 2003 Apr;10(4):201-6. Predictors of prostate cancer on repeat prostatic biopsy in men with serum total prostate-specific antigen between 4.1 and 10 ng/mL. Okegawa T, Kinjo M, Ohta M, Miura I, Horie S, Nutahara K, Higashihara E.Department of Urology, Kyorin University School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo, Japan

"fPSA/tPSA ratio, fPSA/cPSA ratio, tPSATZ and cPSATZ enhance the specificity of PSA testing compared to tPSA or cPSA when determining which patients should undergo repeat biopsy."

Urology. 2002 Oct;60(4 Suppl 1):31-5. Complexed prostate-specific antigen for early detection of prostate cancer in men with serum prostate-specific antigen levels of 2 to 4 nanograms per milliliter. Horninger W, Cheli CD, Babaian RJ, Fritsche HA, Lepor H, Taneja SS, Childs S, Stamey TA, Sokoll LJ, Chan DW, Brawer MK, Partin AW, Bartsch G. Department of Urology, University of Innsbruck, Austria

JAMA. 1997 May 14; 277(18):1452-5. Enhancement of specificity with free PSA measurements. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination. Catalona WJ, Smith DS, Ornstein DK

Edited by J. Strax. Updated 2003-10-11