Clinical Significance of Measuring Prostate- Specific Antigen

Transcription

1 Clinical Significance of Measuring Prostate- Specific Antigen Borros M. Arneth, MS, MD (Institute of Clinical Chemistry and Laboratory Medicine, Johannes Gutenberg University Mainz, Mainz, Germany) DOI: /LMEGGGLZ2EDWRXUK Abstract The incidence of prostate-specific antigen (PSA) monitoring has increased in recent years. However, interpretation of the results is often ambiguous and leads to uncertainty for both the patient and the treating physician. Advantages and disadvantages of measuring PSA and possibilities for improved interpretation using the variable PSA quotient are given. After reading this article, readers should be able to better interpret PSA values and assess the risk of prostate carcinoma using PSA quotient. In addition, errors associated with PSA and PSA quotient measurement are compared and discussed. Chemistry exam questions and corresponding answer form are located after this CE Update on page 492. The prostate gland is a chestnut-sized organ that lies directly under the bladder and encloses the upper section of the urethra. It is approximately 20 to 25 g in a young man and 30 g in an older man. During ejaculation, the prostate adds up to 40 g of a milky secretion to the ejaculate, in which prostate-specific antigen (PSA), a protein formed by the prostate gland, is present in high concentrations. 1-3 Prostatic secretions are slightly acidic with a ph around 6.4. The acidity serves to neutralize vaginal alkalinity and prolong the lifespan of spermatozoa. 4 PSA liquefies semen, promoting sperm motility, and serves to dissolve cervical mucus. PSA is present in low concentrations in the blood, but the concentration increases with prostate irritation, prostatic infection, and benign prostatic hyperplasia (BPH). PSA exists in the blood in two forms. Most PSA in blood is bound to serum proteins, some of which are inhibitors of the serine protease activity of PSA. Further, PSA is also present as free PSA. 5 Total PSA is the sum of both bound and free PSA; however, free PSA is measured only if the total PSA is increased. PSA is primarily a tissue-specific marker. From an elevated PSA measurement, it is difficult to differentiate between a benign and malignant transformation of the prostate gland. Distinguishing between the two is where free PSA is useful. Free PSA is more often formed from benign transformations while bound PSA tends to come from malign transformations. Both tests (free and total PSA) have high accuracy and repeatability. With increasing age, enlargement of the prostate gland is common and, in most cases, is benign. However, it often leads to unpleasant symptoms, such as problems with urination. The lifetime incidence of prostate carcinoma is 8% to 14% for Caucasian men 5 but prostate carcinoma will not be symptomatic in all of these men. Prostate carcinoma is rare in men under age 50 and the average age at diagnosis is 75 years. The risk is greater for men who have relatives with prostate carcinomas. 6 Only about 3% of the men with prostate carcinoma in Europe and in the United Glossary of Terms ACT-PSA BPH CV Complex PSA EDTA Free PSA PSA PSA EIA PSA ratio Total PSA Alpha-1-chymotrypsin prostate-specific antigen complex Benign prostatic hyperplasia Coefficient of variation Complexed portion of PSA Ethylenediamine tetraacetic acid Uncomplexed, free portion of PSA Prostate-specific antigen PSA enzyme immunoassay Quotient of free PSA to total PSA Free PSA + complexed PSA States die from this disease. A cure is more often possible if the tumor is recognized at an early stage; however, not all prostate carcinomas are aggressive. Some prostate carcinomas grow rapidly and metastasize. If not recognized at an early stage and treated appropriately, aggressive tumors often lead to death. Others grow slowly, remain asymptomatic, and do not extend to other organs. Cancerous prostate tissue usually releases more PSA and more complexed PSA into the blood than normal, healthy tissue does. Thus, an increased PSA may indicate the presence of prostate carcinoma. The higher the PSA concentration in blood, the more likely one is to find tumors that have extended beyond the prostate gland. 7 Today, asymptomatic and aggressive tumor forms can be differentiated most reliable by means of prostatic biopsy and subsequent histopathological investigation of the tissue. Until recently, PSA was thought to be produced only in the prostate; however, using very sensitive methods, PSA has now been found in low concentrations in many other tissues. The PSA molecule is very similar to kallikrein, which is androgen-dependent and produced in epithelial and glandular tissue. August 2009 Volume 40 Number 8 LABMEDICINE 487

2 Salient Points PSA is a glycoprotein produced predominantly, although not exclusively, by the prostate gland. PSA in blood may increase as a result of a variety of pathologic conditions of the prostate and may be a suitable marker for prostate carcinoma. For men over age 50, annual PSA screening is often recommended, although the value of this screening is disputed. Particularly at values below 10 µg/l (=10 ng/ml), the positive predictive value is rather poor for a carcinoma. Ultrasound, the rate of increase of PSA (PSA velocity), age-dependent reference values, and free PSA can significantly increase the predictive value of PSA measurement in cancer screening. 8 Blood from which PSA is attained should be drawn before digital rectal examination of the prostate gland The reference range for serum PSA is 0 4 µg/l and is dependent on the method used. PSA is usually elevated in the presence of prostate carcinoma; however, many benign conditions also result in PSA elevation. 6 The most important reason to measure PSA in blood is to screen for prostate carcinoma in men over age 50. In addition to screening, PSA is measured to evaluate the success of treatment and progression of disease when a known prostate carcinoma is present. However, an increased value does not always indicate carcinoma since studies have shown that about 70% of men with PSA values from 4 10 µg/l are free of prostate cancer. 13,14 Causes other than prostate carcinoma that can lead to elevated PSA values include: 1. Prostatic hyperplasia, one benign tumor, or several benign tumors of the prostate gland. Adenomatous hyperplasia is a common condition, occurring in more than half of all men over the age of ,16 2. Inflammation of the prostate gland (prostatitis), which is less common Prostate infarction and other causes, which are rare. 4. Manipulations to the prostate gland or ejaculation. 18 On the other hand, a normal value does not absolutely indicate that carcinoma is not present, since a considerable percentage of men with carcinoma of the prostate have serum PSA levels under 4 µg/l. 13 Also, overweight men have been shown to have lower PSA values. 19,20 In order to increase the predictive value of PSA measurements, the following 5 points have been found useful: 1. Digital rectal examination and ultrasonic investigation of the prostate gland 21 should be performed routinely. 2. Age-specific reference ranges (Table 1) should be considered. In younger men, elevated PSA and decreased free PSA fractions raise suspicion of prostate cancer even at lower values than are suspicious in older men. Table 1_Age-Specific Reference Ranges 3. PSA velocity should be determined. Prostate carcinoma usually produces more PSA and grows faster than does BPH. 22 Seventy percent of all men with prostate carcinoma show an increase in PSA of greater than 0.75 µg/l per year, whereas only 5% of all men without prostate carcinoma show a rise of this magnitude. In order to evaluate PSA velocity, several accurate measurements with comparable test methods and same sample type (ie, serum versus plasma) are necessary. 4. PSA density, free PSA, 23 and the free-to-total PSA quotient should be determined. Prostate carcinomas release much more PSA into the blood than benign prostate tumors. A small carcinoma can dramatically increase PSA levels, whereas the same increase can only be caused by a relatively large benign prostate tumor. PSA density is calculated by dividing the PSA concentration by the volume of the prostate gland as measured by ultrasound. High PSA values with a small prostate size (resulting in a high PSA density) raise suspicion for prostate carcinoma, whereas a low PSA density predicts benign tumors. One disadvantage of PSA density is that when a small carcinoma develops in a large benign tumor, the PSA density will be relatively low despite the presence of a carcinoma. PSA density is thus more useful when the prostate gland is small than when it is large. 7 Most PSA in the blood is bound to other proteins. 24 Normally, more than 23% of serum PSA is in its free form (free PSA). 25,26 Benign prostate tumors (prostatic hyperplasia, BPH) may be associated with elevated PSA, especially free PSA. With prostate carcinoma, however, the proportion of free PSA decreases. Why this occurs is unclear. One theory hypothesizes that prostate carcinoma produces not only PSA but also proteins to which the PSA is bound. As a result, the portion of free PSA decreases. The PSA quotient is computed by dividing the free PSA by the total PSA. The reference range for the PSA quotient is greater than 0.23 for PSA values of 4 to 10 µg/l and above 0.15 for PSA values below 4 µg/l. A lower quotient strongly suggests prostate carcinoma. A higher quotient suggests BPH. 5. Alpha-1-chymotrypsin-bound PSA (ACT-PSA), complexed PSA (cpsa), or both 24 should be found. As mentioned in the previous section, a majority of PSA in the blood is present in the bound form. Bound PSA, detectable in the blood, is often complexed with alpha-1-chymotrypsin. ACT-PSA may be measured instead of free and total PSA. The tests for cpsa measure different forms of bound PSA, among them ACT-PSA. These tests may replace traditional testing for free and total PSA. Some studies have shown improved discrimination between prostatic hyperplasia and prostate carcinoma through measuring cpsa. 24 Other studies, however, found no advantage in measuring cpsa or ACT-PSA over conventional measurement of free and total PSA. This issue is not resolved; however, new tests (cpsa, ACT-PSA) are likely to be at least as useful as traditional tests. % Free PSA Age Age Age >70 <10 49% 58% 65% % 34% 41% % 24% 30% >25 9% 12% 16% Probability in percent of a carcinoma depending on % free PSA and age (Roche Diagnostics Elecsys test 14 ). PSA, prostate-specific antigen. Tissue Biopsy For men with a PSA over 4.0 ng/ml, guidelines for the diagnosis of prostate carcinoma recommend that tissue samples should be obtained by needle biopsy and be examined by a pathologist. 27 At the present time, definitive diagnosis of prostate carcinomas can be made only through pathologic Downloaded 488 from LABMEDICINE Volume Number 8 August 2009 labmedicine.com

3 examination but prostate carcinoma cannot be definitively excluded by this method. For small tumors, the risk remains that the tumor was not contained within the biopsy specimen, and an unrecognized cancer remains. In addition, prostate biopsy is an invasive procedure, which introduces the risk of contamination of the biopsy site by tumor cells. Table 2_Interpretation of Prostate-Specific Antigen Values 32 PSA (ng/ml) % % >10 >49% Probability in Percent that a Carcinoma is Present Prostate-Specific Antigen Measurement Anti-PSA monoclonal antibodies are added to serum (or plasma) and incubated using standard methods and standard conditions. 28 Anti-PSA monoclonal antibodies bound to PSA are detected, and the probability that a carcinoma is present increases with increasing PSA. The following values are to be understood only as reference points as different studies have yielded different results (Table 2). The probability that a carcinoma is present rises as the proportion of free PSA decreases. The following values are reference points; different studies yield different results (Table 3). Specimens In principle, serum or plasma is suitable. If free PSA is desired, plasma (with heparin or EDTA anticoagulant) is recommended. Plasma should be separated by centrifugation within a short time of collection, preferably within 3 hours. The blood draw should take place prior to digital examination of the prostate gland. Table 3_Interpretation of Prostate-Specific Antigen Quotients 32 Probability in Percent that a Carcinoma is Present Free PSA (%) (PSA values between 4 and 10 µg/l, age 60 69) % % % >25 12% drawing blood can lead to increased PSA values. Elevated values decrease with a biological half-life of 2 to 3 days. Certain medications, including antiandrogens, inhibit progression of BPH and can lead to a decrease in PSA concentration. Further errors can result from the analytic measuring procedure. This error distribution is different for PSA and PSA quotient measurements as examined and presented in Table 5. Characteristics of the Roche Diagnostics Elecsys test are given in Table 4 (sensitivity and specificity) and Table 5 (imprecision of free and total PSA). Sample Storage for Measuring Total PSA and Stability of Free PSA A sample intended to measure total PSA, when stored at 20 C to 25 C, will see a 25% decrease in 1 week. It will remain stable for 5 days when stored at 4 C to 8 C. Free PSA is accurate for 1 day when a sample is stored at 4 C. Preanalytics When determining free PSA, plasma should be separated from cells within 3 hours of the blood being drawn. Interferences The following are examples of factors that will not interfere with the measurement of total PSA below specific levels (EIAII by Roche 28 ): No interference by hemoglobin up to 1 g/dl. No interference by bilirubin up to 60 mg/dl. No interference by triglycerides up to 2,000 mg/dl. Different medicines tested by the manufacturer did not influence results. Error Sources Palpation of the prostate gland, especially prostate massage, or surgical manipulation of the prostate gland before Gaussian Error Calculation/Analytic Measurement Error The coefficient of variation (CV) is the standard deviation divided by the mean value from a set of data. This is a close approximation of the mean relative error, which is the error divided by the measured value. In physics, it is common practice to calculate (or estimate) the measurement error of a product or quotient of 2 measured quantities, according to the Gaussian error propagation equation, which utilizes the Pythagorean Theorem: w/w = ( x/x) 2 + ( y/y) 2 Table 4_Characteristics of the Roche Diagnostics Elecsys Test Total PSA (µg/l) Free PSA/Total PSA Sensitivity Specificity % 63% % 82% < % 49% Table 5_Imprecision of Free and Total PSA (Roche Diagnostics Elecsys Test) Imprecision Lower Detection Limit Range Total PSA <3.7% ng/ml Free PSA <5.5% <0.01 ng/ml August 2009 Volume 40 Number 8 LABMEDICINE 489

4 Measurement of the day-to-day precision for free PSA yields a coefficient of variation (CV) of 1.71% for free PSA at 1.3 ng/ml, and a CV of 2.5% for free PSA at 1.6 ng/ml. For total PSA, a CV of 4.94% is found at 43.8 ng/ml and a CV of 4.67% is found at 44.5 ng/ml (data not shown). At free PSA values <5 ng/ml, the CV is approximately 2.5%. For total PSA, the CV is approximately 4.21% at the levels tested (see Table 6). The calculated error of the product of total PSA and free PSA is 4.89%. The calculated error of the free-psa-to total-psa quotient is 3.38%. Thus, the CV of the PSA quotient should be better than the error of total PSA alone (3.38%<4.21%). Practical CV Measurements Daily measurements of free PSA and total PSA, as well as calculated PSA quotients, are presented in Table 6. Free PSA has a CV of 2.5%, total PSA has a CV of 4.21%, and the free-psa-to-total-psa quotient has a CV of 2.94% (again, clearly less than the measurement error of the total PSA alone). Using controls of less than 5 ng/ml for free PSA and total PSA yields an even better CV for the free-psato-total-psa quotient than the theoretical value derived by Gaussian error calculations (see above 2.94% versus 3.38%). The errors for total PSA and for free PSA are presumably in the same direction, and consequently cancel out each other in the free-psa-to-total-psa quotient. The total PSA measurement error depends on the characteristics of the assay employed. By contrast, free-psa-to-total-psa quotients have Table 6_Daily Measurements of Free PSA and Total PSA in Control Subjects and Calculated Free-PSA-to-Total- PSA Quotient (Roche Diagnostics Elecsys Test) Total PSA <5 ng/ml or PSA = 5 ng/ml Free PSA Total PSA Free PSA/PSA , Mean SD CV % a measurement error considerably smaller than that of total PSA alone. The magnitude of this error is likely to be independent of the assay used. This observation holds true both for the theoretical derivation of the measurement error for the PSA quotient using the Gaussian error calculation and for the practical evaluation of the deviations from standardized control serum. The current study demonstrates that the PSA quotient should be given more attention in clinical practice. Compared with the measurement of total PSA alone, the PSA quotient is clearly associated with a much smaller measurement error (2.94% for the PSA quotient compared with 4.21% for total PSA alone), especially in the clinically relevant lower PSA range (total PSA <4.0 ng/ml). Further studies are warranted on the practical relevance of the PSA quotient. PSA Screening Pros and Cons Annual screening is recommended for men over 50, though some contend that PSA should not be measured in all men above 50. Some argue that untargeted screening does more harm than good, although no one denies that more prostate carcinomas are recognized with screening. 29 Opponents of untargeted PSA screening criticize those who favor it, claiming that surgical excision of prostate carcinoma is not always advantageous for the patient. 30 This assertion may not sound logical; however, it has not been proven that all prostate carcinomas should be treated. Finding an elevated PSA value may result in needless anxiety for the patient (and sometimes even for the doctor) and may result in unnecessary surgery with resultant impotence and urinary incompetence. The percentage of patients with asymptomatic prostate carcinoma who would develop clinically significant disease is unknown. Frankel and colleagues dispute the merits of broad screening because evidence from natural history is unhelpful since men are more likely to die with, rather than from, prostate cancer. 31 An emotional debate persists among experts. 32 Although we do not presume to resolve this controversy, we seek to elucidate the role that PSA may have in the diagnosis of this epidemiologically and economically important disease. Summary Prostate-specific antigen is a glycoprotein that is produced primarily by the prostate. It is a highly tissue-specific marker in serum, but it only has minimal disease specificity. To increase this specificity, various measures have been derived using the PSA concentration, such as PSA density, PSA velocity, and, in particular, PSA quotient. This quotient is calculated by dividing the concentration of free PSA by the total PSA concentration. In healthy patients, more than 23% of the total PSA content is free. In most patients with prostate carcinoma, less than 23% of the PSA exists in the free form. Although diagnostic improvements have been made through the use of PSA density, PSA velocity, and the PSA quotient, the diagnosis of prostate carcinoma by serum analysis is, nevertheless, accompanied by uncertainties. A definitive diagnosis of prostate carcinoma can be made only after tumor biopsy and histopathologic examination. LM Downloaded 490 from LABMEDICINE Volume Number 8 August 2009 labmedicine.com

5 1. Balk S, Ko Y, Bubley G. Biology of prostate-specific antigen. Am J Clin Oncol. 2003;28: American Society of Andrology. What is the prostate and what is its function? In: American Society of Andrology Handbook of Andrology. Ch. 8. Schaumburg, IL: ASH; Laux D, Custis M, Cutis S. Forensic Detection of Semen III. Detection of PSA Using Membrane Based Tests: Sensitivity Issues with Regards to the Presence of PSA in Other Body Fluids. Midwestern Association of Forensic Scientists; Lilja H. Biology of prostate-specific antigen. Urology. 2003;62: Catalona W, Partin A, Slawin K, et al. Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: A prospective multicenter clinical trial. JAMA. 1998;279: Thompson I, Pauler D, Goodman P, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level < or =4.0 ng per millil. N Engl J Med. 2004;350: Lamerz R. PSA prostata spezifisches antigen. In: Thomas L, ed. Labor und Diagnose. Frankfurt am Main, Germany: TH-Books; 2000: Punglia R, Amico A, Catalona W, et al. Effect of verification bias on screening for prostate cancer by measurement of prostate-specific antigen. N Engl J Med. 2003;349: Crawford E, Schutz M, Clejan S, et al. The effect of digital rectal examination on prostate-specific antigen levels. JAMA. 1992;267: Chybowski F, Bergstralh E, Oesterling J. The effect of digital rectal examination on the serum prostate specific antigen concentration: Results of a randomized study. J Urol. 1992;148: Collins G, Martin P, Wynn-Davies A, et al. The effect of digital rectal examination on the serum prostate specific antigen concentration: Results of a randomized study. J Urol. 1997;157: Tarhan F, Orçun A, Küçükercan I, et al. Effect of prostatic massage on serum complexed prostate-specific antigen levels. Urology. 2005;66: Carter H. Prostate cancers in men with low PSA levels Must we find them? N Engl J Med. 2004;350: Mytrle J. Normal Levels of Prostate-Specific Antigen (PSA), Clinical Aspects of Prostate Cancer. Assessment of New Diagnostic and Management Procedures. Paper presented at: Prostate Cancer Working Group of the National Cancer Institute s Organ Systems Program. October 16 19, Prout s Neck, Me. New York: Elsevier; 1989: Carter H, Pearson J, Metter E, et al. Longitudinal evaluation of prostatespecific antigen levels in men with and without prostate disease. JAMA. 1992;267: Carter H. Assessing risk: Does this patient have prostate cancer? J Natl Cancer Inst. 2006;98: Nadler R, Humphrey P, Smith D, et al. Effect of inflammation and benign prostatic hyperplasia on elevated serum prostate specific antigen levels. J Urol. 1995;154: Herschman J, Smith D, Catalona W. Effect of ejaculation on serum total and free prostate-specific antigen concentrations. Urology. 1997;50: Bañez L, Hamilton R, Partin A, et al. Obesity-related plasma hemodilution and PSA concentration among men with prostate cancer. JAMA. 2004;350: Dreicer R. Why do obese men have lower PSA concentrations? Journal Watch (N Engl J Med). 2007;20: Catalona W, Richie J, Ahmann F, et al. Comparison of digital rectal examination and serum prostate specific antigen in the early detection of prostate cancer: Results of a multicenter clinical trial of 6,630 men. J Urol. 1994;151: D Amico A, Chen M, RoehL K, et al. Preoperative PSA velocity and the risk of death from prostate cancer after radical prostatectomy. N Engl J Med. 2004;351: Catalona W, Smith D, Ornstein D. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/ml and benign prostate examination. Enhancement of specificity with free PSA measurements. JAMA. 1997;277: Christensson A, Björk T, Nilsson O, et al. Serum prostate specific antigen complexed to alpha 1-antichymotrypsin as an indicator of prostate cancer. J Urol. 1993;150: Catalona W, Smith D, Wolfert R, et al. Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA. 1995;274: Chen Y, Luderer A, Thiel R, et al. Using proportions of free to total prostatespecific antigen, age, and total prostate-specific antigen to predict the probability of prostate cancer. Urology. 1996;47: Walsh P, DeWeese T, Eisenberger M. Localized prostate cancer. N Engl J Med. 2007;357: Roche Cobas product information. Elecsys Total-PSA and Free PSA Elecsys Gesamt-PSA und freies PSA. Roche Diagnostics. 2006; Catalona W, Smith D, Ratliff T, et al. Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med. 1991;324: Barry M. Prostate-specific antigen for early diagnosis of prostate cancer. N Engl J Med. 2001;344: Frankel S, Smith GD, Donovan J, et al. Screening for prostate cancer. Lancet. 2003;361: Woolf S. Screening for prostate cancer with prostate-specific antigen. An examination of the evidence. N Engl J Med. 1995;333: August 2009 Volume 40 Number 8 LABMEDICINE 491