Using PSA Intelligently to Manage Prostate Cancer: Part 2 of 2

Quick Search


PCRI Insights August, 2005 vol. 8, no. 3
By Jonathan McDermed, Pharm. D., Diagnostic Products Corporation

Corollary to Part One

In Part One of this article (Insights, August 2003), I reviewed studies that described differences in PSA levels between men who eventually developed, or did not develop, prostate cancer in later life. One of these studies confirmed that 40- to 50-year-old men having a “baseline” PSA level that was higher than average had a greater risk of developing prostate cancer. Conversely, this risk was lower for men having baseline PSA levels below average for their age group. In another study, results showed that men who eventually developed prostate cancer had a more rapid rate of PSA increase during the 10 years preceding the diagnosis than men who did not develop this disease. By combining these two concepts, I showed how prostate cancer screening has evolved during the past decade. As a direct result, the median age of men diagnosed with prostate cancer had declined to 62 in the year 20001, and the prostate cancer death rate today is at a 50-year low.

Once a man is identified as being at risk for prostate cancer (on the basis of an elevated baseline PSA or serial measurements of total PSA), the next step in the diagnostic workup is to measure serum levels of PSA isoforms. PSA is a protease (an enzyme that degrades proteins) and is a single-chain glycoprotein consisting of 237 amino acid residues and approximately 8% carbohydrate.2 Five PSA isoforms exist: two are biologically active forms differing in their carbohydrate side chain, and three are biologically inactive.3 Biologically active forms of PSA that enter the circulation are rapidly inactivated by binding with a number of protease inhibitors, the most common of which is a1-antichymotrypsin (ACT).4,5 The inactive or “nicked” forms of PSA will not bind to protease inhibitors and are referred to as “free” PSA whereas ACT-bound PSA is called “complexed PSA“. Immunoassays for PSA (total PSA) measure the levels of both free and complexed PSA isoforms.

Assays are now commercially available that specifically measure only the free PSA and complexed PSA isoforms.

Free PSA comprises a higher proportion of total PSA (and complexed PSA a lower percentage of the total) in men with enlarged prostates, or benign prostatic hypertrophy (BPH). Conversely, free PSA comprises a lower proportion of total PSA (and complexed PSA a higher proportion of the total) in men with prostate cancer.6 The FDA has approved both a complexed PSA assay and several free PSA assays (used in conjunction with total PSA) as diagnostic tools to help differentiate prostate cancer from benign prostatic hyperplasia (BPH).

Probability of detecting prostate cancer using free-to-total PSA ratiosThe landmark study using Hybritech free and total PSA assays documented the value of measuring the percent free-PSA in men with total PSA values between 4.0 and 10.0 ng/mL (the so-called “gray zone”).7 (See Table 1.) The discriminatory power of percent free PSA was better than total PSA when the percent free PSA was = 10% or > 25% but was not any better in this regard when the percent free PSA was between these limits. When the total PSA value is below 4.0 ng/mL, the discriminatory power using percent free PSA is only marginally better than total PSA. As a result, the FDA-approved free PSA assays on the US market are only indicated for use in men with total PSA measurements between 4.0 and 10.0 ng/mL.

Prostate volume is important to consider when evaluating the percent of free PSA because its power to discriminate between BPH and prostate cancer is greater in men having a smaller gland volume (< 40 cm3).8 A study examining this relationship using DPC’s free* and total PSA assays led directly to the development of an artificial neural network (ANN)9, which is a mathematical model to help urologists decide whether or not a prostate biopsy should be performed. This ANN was designed using patient age, levels of free and total PSA, prostate volume (as determined by transrectal ultrasound), and digital rectal examination findings (positive or negative) as input variables. Together, these factors provide a “risk assessment” for prostate cancer that is better than percent free PSA alone.

In a subsequent multi-center study, this ANN (“ProstataClass”) was validated using more than 1,100 samples from patients with known cancerous or benign biopsy results. Study results demonstrated that ProstataClass was significantly more accurate than either total PSA or percent free PSA for predicting biopsy results (p = 0.01). ProstataClass is available free of charge for urologists’ use from the Charité Hospital in Berlin, Germany and can be accessed via their Web site,

Complexed PSA levels and percent free PSA show similar superiority over total PSA levels in detecting prostate cancer in younger men, where BPH is much less common.11 However, as shown in Table 2, ProstataClass provides even greater specificity for prostate cancer detection in men with low total PSA serum levels (2.0 – 4.0 ng/mL) than complexed PSA or percent free PSA.

Specificity to detect prostate cancer using PSA and PSA isoforms
*DPC’s free PSA assay has not been FDA-approved in the U.S.

Despite early detection, up to 40 percent of men undergoing definitive local treatment will likely experience PSA progression at some time during their lifetime. The ability to detect residual or recurrent disease earlier assumes much more practical importance today as more and more men are being diagnosed and treated for prostate cancer in their 50s and 60s. In the balance of this article, I will discuss how “ultrasensitive” PSA assays can be used to identify men at high risk for recurrent disease. I will also describe how post-treatment PSA doubling time (PSADT) can differentiate men who may or may not be at risk for dying from recurrent prostate cancer, which may be used to guide decision-making regarding the choice of salvage therapy.

Using PSA Intelligently for Monitoring Treated Prostate Cancer Patients

Treatment monitoring and prognosis assessments are the two most common clinical applications for PSA determinations in men treated for prostate cancer. Local treatments include radical prostatectomy (RP), various methods of delivery for radiation treatment (RT) and cryosurgery, whereas systemic treatments include androgen deprivation therapy (ADT), and cytotoxic chemotherapy. Post-treatment PSA levels can provide invaluable information about the effectiveness of the therapy given and the existence of residual cancer in men treated with RP or cryosurgery. In such patients, rising PSA levels can signal cancer activity well before any clinical signs of recurrence appear. This lead-time can be further increased by months and even years when highly sensitive third-generation PSA assays are employed. In the following discussion, I will review a number of studies examining the use of ultrasensitive PSA measurements to identify and clinically monitor men who may experience PSA progression following RP.

Because primary RT in which the gland remains in situ rarely results in an undetectable PSA, even when cure is achieved12, this article will only discuss the use of ultrasensitive PSA assays in conjunction with salvage RT. Disease progression after primary RT is strongly suspected if the PSA rises on consecutive determinations. The use of a highly sensitive PSA assay, such as IMMULITE or IMMULITE 2000 Third Generation PSA, may be used to detect early disease progression following RT, and serial measurements permit accurate calculations of PSADT.

What is an Ultrasensitive PSA Assay?

The analytical sensitivity of an assay (also referred to as the detection limit) is defined as the lowest concentration of the measured analyte that can be distinguished from the zero control. The Yang Pros- Check® and Hybritech Tandem-R® assays, used clinically back in the late 1980s, were the first commercial immunoassays for PSA. These first-generation PSA assays were manually performed radioimmunometric test methods and possessed analytical sensitivities of 0.3 to 0.6 ng/mL.13,14,15

Using the Yang assay as an example, patient values that were below the detection limit for the assay were reported as < 0.3 ng/mL, but could be anywhere between zero and 0.29 ng/mL. Given the high degree of error for PSA measurements approaching the detection limit of such assays, an accurate, reproducible patient result could not be assured unless the PSA level was as high as 0.6 to 0.8 ng/mL.16 This higher value is called an assay’s “functional sensitivity,” which is defined as the lowest concentration measurable with an assay where the coefficient of variation is less than 20%.

Second-generation PSA assays were developed in the mid-1990s and offered roughly a 10-fold improvement in analytical sensitivity, with detection limits of 0.03 to 0.07 ng/mL, depending upon the manufacturer’s claims. Automated immunoassay analyzers, which reduce the inherent errors associated with manual testing, also began to be introduced during this same time frame. In order to differentiate these second-generation PSA assays from the less sensitive Pros-Check® and Tandem-R® methods, the terms hypersensitive and ultrasensitive were often used to describe these assays. Although more sensitive from an analytical standpoint, second-generation PSA assays possess a functional sensitivity of 0.1 to 0.2 ng/mL.

The first third-generation PSA assay was introduced into the U.S. market in 1997 by DPC. This assay offers an additional 10-fold improvement in low-end analytical sensitivity, with a claimed detection limit of 0.003 ng/mL, and a functional sensitivity of 0.01 ng/mL. Recognizing the clinical value of third-generation sensitivity, other manufacturers have introduced more sensitive versions of their own PSA tests.17

Estimating Risk for Post-Treatment Disease Progression

There are many pre- and post-therapy variables shown to have prognostic value with regard to likelihood of disease progression in patients treated for prostate cancer. Of pre-treatment variables, the most widely recognized by investigators include PSA level, Gleason score (pathological grade), and clinical stage. The published literature describes several predictive algorithms, multivariate analyses, and artificial neural networks incorporating these three variables as a means to estimate the risk for poor outcomes following RP, external beam RT, and brachytherapy. Likewise, there are many predictive algorithms that take into consideration a number of pathological post-prostatectomy findings. Rather than discuss them in detail here, I will focus on the use of post treatment PSA values and their kinetics over time as they are used to predict the need for adjunctive or salvage therapies in men treated with RP.

Expected PSA Levels After RP

PSA levels that are measured three or more weeks following a successful RP should be zero, or at least very close to zero, and stable. The presence of PSA in the blood after RP indicates a failure to remove the tumor completely, and the reappearance of PSA at a later date indicates tumor recurrence. Exceptions to this include cases where unilateral or bilateral nerve-sparing surgery or laparoscopic procedures leave benign tissue behind. In such patients, PSA levels will often be detectable using a third-generation PSA test, albeit at a very low concentration.

The functional sensitivity of the first and second-generation PSA assays significantly limits their use for early post-operative detection of surgical failure in most cases. However, a number of clinical studies have been published using PSA assays with third-generation sensitivity post-RP. These studies have clearly established the value of these highly sensitive assays for detecting early prostate cancer progression following RP.18,19,20,21 In a landmark study by Witherspoon et al, DPC’s IMMULITE Third Generation PSA assay appeared to (1) identify men with apparently organ-confined prostate cancer destined to fail surgery and (2) provide an average 18-month lead time in detecting disease progression compared to a conventional PSA assay (Figure 1).18

Post-operative PSA levels over time
Figure 1. Post-operative PSA levels over time in a 73-year-old man who underwent RP. No tumor was present at the surgical margins, seminal vesicles or regional lymph nodes and the postoperative baseline PSA was 0.004 µg/L. PSA at 4.2 years after prostatectomy became detectable at 0.10 µg/L using a conventional PSA assay. Thus, the PSA was noted to be rising more than two years earlier using the IMMULITE Third Generation PSA assay.1

Vassilikos et al reported results in 197 men undergoing RP over a four-year follow-up period using an in-house ultrasensitive PSA assay and they were able to define clinical outcomes for three groups of patients.19 Sixty-two percent of the men did not show any significant changes in serum PSA values during follow-up and had no evidence of clinically recurrent cancer. Fifteen percent of the men showed very slow PSA increases over time, but none of the measurements exceeded 0.1 ng/mL within four years and no clinically recurrent cancer developed. Twenty-three percent of the men demonstrated relatively significant increases of serum PSA, which was first detected an average of 18 months earlier using the ultrasensitive PSA test compared to a less sensitive PSA assay (an analytical sensitivity of 0.1 ng/mL).

Of the 167 men having information on PSA recurrence using both testing methods, 80% had agreement on their recurrence status, including 105 patients in remission and 31 with biochemical recurrence determined by both methods. Of the additional 31 patients in remission (according to the regular PSA test), 26 (84%) were in slow recurrence and five (16%) were in fast recurrence as determined by the ultrasensitive PSA assay. Overall, using the ultrasensitive PSA test, 31 patients (30%) who were considered in remission by the regular PSA test would be reclassified as having biochemical recurrence.

Vassilikos et al point out that pathological findings differed among the three patient groups. Although fewer patients with slow recurrence had unfavorable clinical and pathological features compared to patients with fast PSA recurrence, the group with slow recurrence still had a higher percentage of unfavorable clinical and pathological features than those in remission (Table 3).

Relation of Clinical and Pathological Features With Recurrence Status

In a subsequent study, Doherty et al evaluated the usefulness of DPC’s Third Generation PSA assay for early detection of biochemical recurrence in 200 post-prostatectomy patients.20 The authors measured a single PSA level four to six weeks postop (nadir) and defined an undetectable PSA value as = 0.01 ng/mL (the functional sensitivity of this assay). Results showed the two year biochemical disease-free survival (BDFS) for the 134 patients with evaluable PSA data to be 61.1% (95% confidence interval: 51.6-70.6%). Only two of 73 (2.7%) patients with an undetectable PSA nadir biochemically relapsed compared to 47 of 61 (77.0%) who did not reach this PSA level. Using Cox multivariate analysis, the authors confirmed that an undetectable PSA nadir was the strongest independent variable predicting a favorable BDFS (p < 0.001): it exceeded other known unfavorable pathological features such as baseline PSA, Gleason score, positive surgical margin status and seminal vesicle involvement (Tables 4a and 4b).

Univariate and Multivariate predictors of BDFS20

Shen and associates from the New York University Medical Center published the most recent study evaluating DPC’s Third Generation PSA assay in March 2005.21 The 545 evaluable patients included in this report comprise the largest published series thus far to evaluate the value of ultrasensitive PSA measurements in the post-RP setting. Following RP, patients returned to the clinic for serum sampling at three months, six months, 12 months, then once a year thereafter, and the group was followed for an average of 3.1 years. Results confirmed earlier reports, demonstrating that men with an undetectable PSA nadir (< 0.01 ng/mL) post-RP had a significantly lower biochemical relapse rate than men with a PSA nadir of 0.01 ng/mL (p < 0.01), 0.02 ng/mL (p< 0.025) or > 0.04 ng/mL (p < 0.01). Using multivariate logistic regression analysis, these authors showed that a PSA nadir of 0.01 ng/mL (p < 0.05), 0.02 ng/mL (p < 0.05) and > 0.04 ng/mL (p < 0.01) independently predicted an increased risk of biochemical relapse compared to a nadir of less than 0.01 ng/mL. (See Table 5.)

Relation of PSA Nadir Post-RP, Relapse Rate and Tme to Relapse

Using Third-Generation PSA to Manage Post-RP Biochemical Relapse

For patients suffering PSA progression following primary surgery, viable treatment options include watchful waiting (no treatment), salvage radiation to the prostatic fossa (with or without radiation to the pelvic lymph nodes), or androgen deprivation therapy (ADT). Salvage radiation treatment will benefit only those patients with proven residual cancer in the prostatic fossa, whereas ADT can potentially benefit those with residual cancer and/or metastatic disease. Serial PSA measurements in the ensuing months following surgery can be used to help determine the likelihood that a patient is more or less likely to benefit from salvage radiation therapy.11,22,23,24

Undetectable baseline PSA levels defined as < 0.2 ng/L following RP, which later become detectable and progressively rise, suggest locally recurring cancer in the prostatic fossa11 (as indicated by the patient example in Figure 1). Detectable PSA levels at baseline that show progressive increases over time likely represent microscopic metastatic disease that was present prior to RP.22,23 In the former scenario, salvage external beam RT (EBRT) may be indicated, whereas systemic salvage treatment using ADT would appear to be indicated in the latter scenario.24

A recent paper examined the four-year benefits of salvage EBRT for post-RP PSA relapse.25 This retrospective study involved 501 men from five academic institutions. Overall, the results showed an overall four year progression-free probability of only 45%. Analysis of the data revealed several factors that significantly predicted disease progression following EBRT:

Adverse Predictive
p Value
Pathology Gleason
score of 8, 9 or 10
< 0.001
level > 2.0 ng/mL
< 0.001
Negative surgical
< 0.001
PSADT < 10 mos.
< 0.001

Interestingly, the four-year progression-free probability for men receiving salvage EBRT and having none of these adverse predictive factors was 77%, indicating that such therapy can provide durable responses in selected patients. Since these criteria include both a pre- EBRT PSA value < 2.0 ng/mL and a PSADT > 10 months, the lead-time afforded by a third-generation PSA assay with its superior low-end precision can prove useful for identifying patients likely to have an early disease relapse, as well as for guiding medical decision-making regarding salvage therapy options.


Third-generation PSA assays effectively address both the requirements of modern prostate cancer screening strategies and the early detection of recurrence following definitive treatment. It has become essential for an all-purpose PSA assay to have third-generation performance capabilities to meet the needs of prostate cancer patients and their clinicians today. 

Part 1 of 2


1. Moul JW, Wu H, Sun L, et al. Epidemiology of radical
prostatectomy for localized prostate cancer in the era of prostate-specific antigen: an overview of the Department of Defense Center for Prostate Disease Research national database. Surgery 132: 213-9, 2002.

2. Armbruster DA. Prostate-specific antigen: Biochemistry, analytical methods, and clinical application. Clin Chem 39: 181-95, 1993.

3. Nurmikko P, Pettersson K, Piironen T, Hugosson J, and Lilja H. Discrimination of Prostate Cancer from Benign Disease by Plasma Measurement of Intact, Free Prostate-specific Antigen Lacking an Internal Cleavage Site at Lys145-Lys146. Clin Chem 47: 1415-23, 2001.

4. Lilja H.A kallikrein-like serine protease in prostatic fluid cleaves the predominant seminal vesicle protein. J Clin Invest 76: 1899-903, 1985.

5. Denis LJ, Murphy GP, and Schroder FH. Report of the consensus workshop on screening and global strategy for prostate cancer. Cancer 75: 1187-207, 1995.

6. Graves HC, Wehner N and Stamey TA. Comparison of a polyclonal and monoclonal immunoassay for PSA: need for an international antigen standard. J Urol 144: 1516- 22, 1990.

7. Catalona WJ, Partin AW, Slawin KM, et al. Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multi-center clinical trial. JAMA 279: 1542-7, 1998.

8. Stephan C, Lein M, Jung K, Schnorr D, Loening SA. The influence of prostate volume on the ratio of free to total prostate specific antigen in serum of patients with prostate carcinoma and benign prostate hyperplasia. Cancer 79: 104-109, 1997.

9. Stephan C, Jung K, Cammann H, Vogel B, Brux B, Kristiansen G, et al. An artificial neural network considerably improves the diagnostic power of percent free prostate specific antigen in prostate cancer diagnosis— results of a five year prospective investigation. Int J Cancer 99:466–73, 2002.

10. Stephan C, Cammann H, Semjonow A, et al. multi-center evaluation of an artificial neural network to increase the prostate cancer detection rate and reduce unnecessary biopsies. Clin Chem 48:8 1279-1287, 2002.

11. Partin AW, Brawer MK, Bartsch G, et al. Complexed prostate specific antigen improves specificity for prostate cancer detection: results of a prospective multi-center clinical trial. J Urol 170: 1787-91, 2003.

12. Forman JD, Meetze K, Pontes W, et al. Therapeutic irradiation for patients with an elevated post-prostatectomy prostate specific antigen level. J Urol 158: 1436-9, 1997

13. Stamey TA, Kabalin JN, McNeal JE, et al. Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. II. Radical prostatectomy treated patients. J Urol 141: 1076-83, 1989.

14. Lange PH, Ercole CJ, Lightner DJ, Fraley EE, Vessella R. The value of serum prostate specific antigen determinations before and after radical prostatectomy. J Urol 141: 873-9, 1989.

15. Hudson MA, Bahnson RR, Catalona WJ. Clinical use of prostate specific antigen in patients with prostate cancer. J Urol 142: 1011-7, 1989.

16. Graves HCB, Wehner N, Stamey T. Ultrasensitive radioimmunoassay of prostate-specific antigen. Clin Chem 38: 735-42, 1992.

17. Ellis WJ, Vessella RL, Noteboom JL, et al. Early detection of recurrent prostate cancer with an ultrasensitive chemiluminescent prostate-specific antigen assay. Urology 50: 573-9, 1997.

18. Witherspoon LR, Lapeyrolerie T. Sensitive prostate specific antigen measurements identify men with long disease- free intervals and differentiate aggressive from indolent cancer recurrences within 2 years after radical prostatectomy. J Urol 157: 1322-8, 1997.

19. Vassilikos EJK, Yu H, Trachtenberg J, et al. Relapse and cure rates of prostate cancer patients after radical prostatectomy and 5 years of follow-up. Clin Biochem 33: 115- 23, 2000.

20. Doherty AP, Bower M, Smith GL, et al. Undetectable ultrasensitive PSA after radical prostatectomy fpr prostate cancer predicts relapse-free survival. Br J Cancer 83: 1432-6, 2000.

21. Shen S, Lepor H, Yaffee R, Taneja SS: Ultrasensitive serum prostate specific antigen nadir accurately predicts the risk of early relapse after radical prostatectomy. J Urol 173: 777-80, 2005.

22. Cadeddu JA, Partin AW, DeWeese TL, Walsh PC. Long term results of radiation therapy for prostate cancer recurrence following radical prostatectomy. J Urol 159: 173-7, 1998.

23. Egawa S, Matsumoto K, Suyama K, et al. Limited suppression of prostate-specific antigen after salvage radiotherapy for its isolated elevation after radical prostatectomy. Urology 53: 148-54, 1999.

24. Fowler JE, Pandey P, Braswell NT, et. al: Prostate specific antigen progression rates after radical prosatectomy or radiation therapy for localized prostate cancer. Surgery 116:302-306, 1994.

25. Stephenson AJ, Shariat SF, Zelefsky MY, et al. Salvage radiotherapy for recurrent prostate cancer after radical prostatectomy. JAMA 291: 1325-32, 2004.