Subject Review. Prostate-Specific Antigen: Critical Issues for the Practicing Physician

Transcription

1 Subject Review Prostate-Specific Antigen: Critical Issues for the Practicing Physician HERBERT C. RUCKLE, M.D.,* GEORGE G. KLEE, M.D., PH.D., AND JOSEPH E. OESTERLING, M.D. Background: Serum prostate-specific antigen (PSA), when used in combination with existing detection methods, improves the clinician's ability to detect early and potentially curable prostate cancer. Findings: This report describes clinically important issues about use of the serum PSA concentration for detecting early prostate cancer. Other PSA-related factors PSA density, PSA velocity, and age-specific reference ranges seem to enhance the ability of clinicians to distinguish benign prostatic conditions from early prostate cancer. Because digital rectal examination only minimally affects the serum PSA concentration, delaying a determination after this examination is unnecessary. Finasteride therapy for benign prostatic hyperplasia Prostate-specific antigen (PSA) is a kallikrein-like, serine protease that, for all practicality, is manufactured by and is specific for prostatic tissue.1 PSA is produced by the epithelial cells that line the acini and ducts of the prostate gland and is secreted into the prostatic ductal system, where it is present in high concentrations and serves to catalyze the liquefaction of the seminal coagulum after ejaculation.2-4 PSA is usually detected at low concentrations in the serum; only when the microscopic- structure of the prostate gland has been disrupted (as with prostate cancer, benign prostatic hyperplasia [BPH], acute prostatitis, and prostate biopsy) can PSA freely diffuse from the prostatic acini and ducts into the stroma and enter the general circulation through the lymphatic system and capillaries. An increase in the serum PSA concentration can lead to the diagnosis of early, clinically significant and potentially curable prostate cancers that would not have been detectable by other clinical methods. PSA AND THE EARLY DETECTION OF PROSTATE CANCER Because prostate cancer is endemic in elderly men, many more men die with prostate cancer than of it. Despite the From the Department of Urology (H.C.R., J.E.O.) and Division of Metabolic and Hématologie Biochemistry (G.G.K.), Mayo Clinic Rochester, Rochester, Minnesota. *Current address: Lorna Linda University Medical Center, Lorna Linda, California. Address reprint requests to Dr. J. E. Oesterling, Department of Urology, Mayo Clinic Rochester, 200 First Street SW, Rochester, MN should be initiated only after the prostate has been evaluated for cancer because this 5oc-reductase inhibitor lowers the serum PSA value by approximately 50%; however, reassessment of the prostate for cancer is necessary if the PSA level fails to decrease as expected or increases to more than 2 ng/ml during finasteride treatment. Conclusion: Currently, PSA is the most important, accurate, and clinically useful tumor marker for prostate cancer. (Mayo Clin Proc 1994; 69:59-68) BPH = benign prostatic hyperplasia; DRE = digital rectal examination; PSA = prostate-specific antigen; TRUS = transrectal ultrasonography abundance of clinically insignificant cancers in elderly men, the American Cancer Society projected that 35,000 men in the United States would die of prostate cancer in In these men, the cancer is clinically significant, but it may not cause death if it is diagnosed and effectively treated when it is truly organ-confined. Assuming that at some point all cancers are organ-confined and amenable to curative therapy is logical. Patients with prostate-confined tumors who are treated definitively with radical prostatectomy have survival rates that correspond with those of age-matched control subjects without prostate cancer.6 Because the cancers that will be clinically significant cannot be predicted, clinicians are obligated to provide the earliest possible diagnosis and treatment of prostate cancer in men who are at risk of death from this disease. Admittedly, the benefit of early detection of prostate cancer is controversial.7 Nonetheless, until appropriate studies are conducted to determine both the population at risk and the individual patients who will fare as well or better without early detection and treatment, prostate cancer in men with a life expectancy of at least 10 more years should be considered potentially lifethreatening and treated definitively. With the increasing emphasis on the early detection and treatment of prostate cancer, PSA has an important and evolving role in the detection of organ-confined and potentially curable prostate cancer. Routine use of PSA determination as a screening tool for prostate cancer has not been standardized. Several recent studies have shown that serum PSA, when used as a screening test, increases the detection rate of prostate cancer and, in Mayo Clin Proc 1994; 69: Mayo Foundation for Medical Education and Research

2 60 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES Mayo Clin Proc, January 1994, Vol 69 fact, is a better screening test than digital rectal examination (DRE). Catalona and associates 8 measured the serum PSA level in 1,653 asymptomatic men 50 years of age or older, and the overall detection rate of prostate cancer was 2.2%. When Brawer and associates 9 analyzed the serum PSA concentrations in 1,249 asymptomatic men 50 years of age or older who had no family history of prostate cancer, they detected cancer in 2.6%. Labrie and coworkers 10 used serum PSA concentrations to screen 1,002 men 45 to 80 years of age who were randomly chosen from the electoral rolls of and around Quebec City; using a cutoff point of 3.0 ng/ml (Tandem-R PSA assay, Hybritech, Inc., San Diego, California), they detected prostate cancer in 4.6%. These results are a considerable improvement over the cancer detection rate of 1.3 to 1.7% in asymptomatic men screened with only DRE. 11 In the studies by Catalona and associates 8 and by Brawer and colleagues, 9 the rates of undetected malignant tumors would have been 32% and 38%, respectively, if only DRE had been used to assess the men for prostate cancer. Thus, these largescale studies clearly show that screening with serum PSA can increase the detection rate of prostate cancer. Because PSA is specific for prostatic tissue and not prostate cancer, it is not a "perfect" tumor marker for screening for prostate cancer. Catalona and coworkers 8 assessed a comparison group of 300 men who had various clinical conditions suggestive of prostate cancer and who had undergone prostate biopsy; of these men, 235 (78%) had a serum PSA determination before biopsy. Of these 235 men, 61 (26%) had biopsy results positive for prostate cancer; 13 of those with prostate cancer (21%) had a serum PSA value within the reference range (0.0 to 3.9 ng/ml; Tandem-R PSA assay). Thus, the cancers would have been missed if the decision to perform biopsy had been based on only the serum PSA value. Cooner and associates 12 assessed 1,807 men 50 to 89 years of age who came to their urology practice for a prostatic examination; using the triad of serum PSA testing, DRE, and transrectal ultrasonography (TRUS), they found prostate cancer in 263 men (15%). Of note, the detection rate of 15% was in a group of patients who requested a medical examination and not in a cohort of men being screened for prostate cancer. Nevertheless, of the 263 patients diagnosed with cancer, 52 (20%) had a normal PSA value with the Tandem-R PSA assay. Labrie and colleagues 10 assessed all their patients with serum PSA testing, DRE, and TRUS and found that of the 57 men with biopsy-proven cancer, 16 (28%) had a serum PSA value within the reference range. In a study of 2,425 men, Babaian and coworkers 13 used DRE, determination of serum PSA, and TRUS and found prostate cancer in 88 (4%). Of these 88 patients, 29 (33%) had a serum PSA level in the reference range. Thus, in these studies, if only the serum PSA value had been used for evaluation, 20 to 33% of the cancers would have been overlooked. Furthermore, these data demonstrate that although the serum PSA value can detect more cancers than the DRE, not all prostate cancers cause an increase in the serum PSA concentration. PSA and DRE each detect certain cancers, and these are not necessarily the same tumors. Neither PSA nor DRE has adequate sensitivity or specificity to be the "ideal" screening test for early prostate cancer. The distinction must be made between screening for prostate cancer and early detection of asymptomatic prostate cancer. Screening is an active program to assess all men, whereas early detection relates to the assessment of men who are examined by a physician. Early detection guidelines are designed to help patients and physicians use the available diagnostic methods most effectively to detect prostate cancer at a curable stage. 14 Both the American Cancer Society and the American Urological Association currently recommend that all men 50 years of age or older undergo an annual examination for detection of early prostate cancer This assessment should consist of a DRE and a serum PSA determination; TRUS is reserved for patients who have abnormalities detected on DRE or an increased serum PSA level. The annual examination for asymptomatic men beginning at age 50 years is for those with "average risk." Men at "high risk" for the development of prostate cancer (black men, men with a family history of prostate cancer, 16 and perhaps men who have had a vasectomy 17 ) should undergo annual examinations beginning at age 40 years. If a man is to benefit from the early detection of asymptomatic prostate cancer, he should have a life expectancy of at least 10 more years. 18 PSA DENSITY For further refinement of the ability of the PSA value to distinguish small-volume, organ-confined prostate cancer from BPH (an ideal condition for curative therapy), Benson and associates described a new factor PSA density that correlates the serum PSA concentration with the prostatic volume. It is defined as the quotient of the serum PSA concentration divided by the volume of the prostate gland, as determined by TRUS. In a preliminary study of 61 patients, these investigators found that PSA density was useful for distinguishing BPH from cancer. 19 The mean PSA density for the 41 patients with clinically localized prostate cancer was 0.58, whereas the mean PSA density for the 20 men with BPH was 0.04 (P<0.0001). In that study, all 33 patients whose PSA density was more than 0.12 had prostate cancer; only 2 patients with prostate cancer had a PSA density of less than Of the 12 patients with prostate cancer and a normal serum PSA value, 10 (83%) had an increased PSA density. Of the 20 patients with BPH, 19 (95%) had a PSA density of less than 0.1. The highest value in the BPH group was

3 Mayo Clin Proc, January 1994, Vol 69 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES 61 Benson and colleagues 20 subsequently assessed 533 men by using serum PSA determinations, DRE, and TRUS. In this larger group of patients, the investigators found that the mean PSA density was 0.19 ± 0.10 for the patients without clinical evidence of prostate cancer and 0.30 ±0.15 for the group with prostate cancer (P< ). Using these data, Benson and coworkers developed probability curves that allow physicians to estimate the probability that a specific patient has prostate cancer on the basis of the PSA density. The commonly used PSA density cutoff point of 0.15 was derived from these curves. On the curve, if all 533 patients are considered and a cancer prevalence of 18.4% is assumed, a PSA density of 0.15 corresponds with a 12% probability of finding prostate cancer, and the chance of finding prostate cancer increases as the PSA density increases (Fig. 1). Andriole and associates 21 showed that the PSA density was useful for further assessment of patients with a persistently increased PSA level and normal results of biopsy. In their study of such patients, men with a PSA density of more than 0.15 had an 82% probability that the subsequent biopsy would reveal prostate cancer. Using a variation of PSA density, Clements and colleagues 22 retrospectively analyzed data from a group of patients with confirmed BPH, localized prostate cancer (M0), or metastatic prostate cancer (Ml). For the patients with BPH, these investigators found a highly significant correlation between log 10 PSA and prostatic volume. When this PSA-volume regression pattern for BPH was used as a reference standard, all 17 patients with Ml disease and 19 of the 23 patients with M0 disease (83%) were distinguished from those with BPH. The merit of PSA density seems to be in distinguishing men with normal findings on a DRE who have a high probability of having prostate cancer from men with a lower probability who have PSA values slightly above normal (4.0 to 10.0 ng/ml). By correlation of the serum PSA level with the volume of the prostate gland, PSA density may facilitate greater selectivity of patients who have mild increases in the PSA level, but normal findings on other evaluations (DRE and TRUS), to undergo prostatic biopsy. Despite the ability of PSA density to indicate inapparent but clinically significant cancers in some patients, the following factors may affect its accuracy to detect cancer: its variability with age, 23 the approximately 10% error in the measurements of prostatic volume with TRUS, 24 and the approximately threefold difference in the ratio of epithelium to stroma between prostates. 25 Thus, additional studies will be necessary to determine the actual role of PSA density in the early diagnosis of prostate cancer. PSA VELOCITY Another concept that may improve the ability of the serum PSA concentration to detect early prostate cancer is monitorgs 100 g βο ö PSA density Fig. 1. Probability of prostate cancer as function of prostatespecific antigen (PSA) density. For PSA density of 0.15, risk of prostate cancer is approximately 12%. (From Benson and associates. 20 By permission of the American Urological Association.) ing the change in the PSA value over time. Carter and coworkers 26 defined the term "PSA velocity" as the rate of change in the serum PSA level and appraised this concept in a preliminary, retrospective study. They reviewed the medical records of 54 men who had been followed up in the Baltimore Longitudinal Study of Aging for a minimum of 7 years before no prostate disease, BPH, or prostate cancer was diagnosed. The control group consisted of 16 men with no prostate disease, the second group included 20 men who underwent a simple prostatectomy and had a histologie diagnosis of BPH, and the third group comprised 18 patients with biopsy-proven prostate cancer. In all men, a complete history was elicited, a physical examination was performed, and laboratory studies were done every 2 to 2 l h years. In their study, Carter and associates 26 demonstrated that the rate of change was more useful for detecting prostate cancer than was the actual serum PSA level at a specific time. By using PSA velocity, they found that prostate cancer could be predicted years before it could be diagnosed by other clinical methods. At 5 years before diagnosis, when serum PSA levels did not differ significantly between the patients with BPH and those who would be diagnosed with cancer, the PSA velocity was significantly higher in the group eventually diagnosed with prostate cancer than in the control group (P<0.01) and the BPH group (P<0.01) (Fig. 2). When a PSA velocity of more than 0.75 ng/ml per year was used as the criterion, the specificity for distinguishing the men with prostate cancer from the BPH group and the control subjects was 90% and 100%, respectively. These values are substantially better than the 60% specificity observed for the serum PSA concentration. The sensitivity of the PSA velocity, however, was not significantly better than that of the serum PSA concentration.

4 62 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES Mayo Clin Proc, January 1994, Vol 69 Fig. 2. Serum prostate-specific antigen (PSA) concentration as function of time before diagnosis of prostatic condition. Note substantial increase in serum value years before local or regional cancer or metastatic cancer was diagnosed. BPH=benign prostatic hyperplasia. (From Carter and associates.26 By permission of the American Medical Association.) Oesterling and colleagues27 examined the rate of change in serum PSA levels in 415 healthy men with no clinical evidence of prostate cancer who had been monitored for at least 12 months. In their prospective, community-based investigation, they determined the upper limit of the reference range for PSA velocity to be 0.8 ng/ml per year. This rate corresponds with a 70% change during 1 year because the median serum PSA concentration for the cohort was 1.3 ng/ml. These findings confirm the cutoff point of 0.75 ng/ ml per year established by Carter and coworkers26 from the Baltimore Longitudinal Study of Aging. These important observations on the rate of change in the serum PSA concentration obviously have much potential as more patients return for annual serum PSA determinations. Thus, in a 65-year-old man, a serum PSA level that increases from 2.3 to 3.6 ng/ml during 1 year indicates the need for further evaluation, which may reveal prostate cancer even though the serum PSA level is within the reference range. These preliminary data indicate that determining the rate of change of the serum PSA value may improve the ability to distinguish men with a low serum PSA value and early, potentially curable prostate cancer from those who have BPH. As outlined, we believe that a patient with a serum PSA level within the reference range who has an increase of 0.8 ng/ml per year or more, as assessed by a minimum of three determinations, should undergo further evaluation of the prostate gland for the possibility of malignant disease. AGE-SPECIFIC PSA REFERENCE RANGES The PSA reference range (0.0 to 4.0 ng/ml) that physicians currently use to determine a man's risk of harboring asymptomatic prostate cancer does not account for age differences or variations in prostatic volumes. The serum PSA concentration of a 76-year-old man with normal findings on DRE may necessitate a different reference range for accurate interpretation than that needed to interpret the PSA value of an asymptomatic 51-year-old man. The reason is that the PSA in the serum originates from the prostate gland, and the prostate of younger men (40 to 59 years of age) is not the same as that of older men (60 to 79 years of age). With advancing age, the prostate gland undergoes numerous microscopic and macroscopic changes that may affect the quantity of PSA entering the systemic circulation. In an effort to determine the relationship between patient age and serum PSA concentration, Oesterling and associates23 performed a prospective study of 471 men in whom the serum PSA determination (Tandem-R PSA assay), DRE, and TRUS disclosed no clinical evidence of prostate cancer. For these men, the prostatic volume also was calculated from the measurements obtained by TRUS. Nomograms that portrayed the distribution of PSA concentrations and prostatic volumes as a function of age were generated. These investigators found that the median serum PSA concentration increased with each decade of age. No men 40 to 49 years old had serum PSA concentrations above the currently used reference range, whereas 13 men 70 to 79 years old (19%) had serum PSA values of more than 4.0 ng/ml. The serum PSA concentration as a function of age for all 471 men and the curve corresponding with the 95th percentile are shown in Figure 3. For the entire age range (40 to 79 years), the serum PSA levels correlated directly with age (r = 0.43; P<0.0001). From these data, age-specific reference ranges were established for each 10-year age-group. With use of the 95th percentile, the appropriate reference range is 0.0 to 2.5 ng/ml for men 40 to 49 years old, 0.0 to 3.5 ng/ml for men 50 to 59 years old, 0.0 to 4.5 ng/ml for men 60 to 69 years old, and 0.0 to 6.5 ng/ml for men 70 to 79 years old (Table 1). For these patients, the median prostatic volume also increased with age; for the entire age range, prostatic volume correlated directly with age (r = 0.43; P<0.001). In addition, the serum PSA value correlated well with prostatic volume (r = 0.55; F<0.0001), and overall the serum PSA concentration was higher for men with large prostates than for those with small prostates. Regression analysis suggested that 30% of the variance in the PSA level is related to prostatic

5 Mayo Clin Proc, January 1994, Vol 69 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES *-"» 8 _l E α> * ' c a 6 < CO Q. E 4 3 ^ Φ CO 2 U 40 -^ Λ-lJ ' *± *- W.. T a -ΓΤι ι 50. ; > * ^^ ^f^ ^ί ^^^ ^^^ ^^^. >^ ' * <^r^ ^^^^ ' ^^^^ * ^^^ * ^^^^' ^ ^ L^^ 1^^^^ ^-*"***^ * ^ -*""" ; ^10 "ilffi 1 ^^ ^^ ^^P i r" 1 "i ' ' ~~*~*"* 5 Percentile &S^-ä~m. 1 i JV Age (yr) 75 Fig. 3. Serum prostate-specific antigen (PSA) concentration as function of age. Scattergram of individual serum PSA values for 471 men; nomogram demonstrates 2.5th, 5th, 10th, 25th, 50th, 75th, 90th, 95th, and 97.5th percentiles for serum PSA on basis of age. (From Oesterling and associates. 23 By permission of the American Medical Association.) volume (/><0.001). When controlling for prostate volume, however, the investigators found that age by itself explained 5% of the variance of the PSA value (P<0.01). Thus, the median PSA density value increased moderately with each decade of life; for the entire age range, the PSA density correlated moderately with age (r = 0.25; P<0.0001). This increase in the serum PSA concentration with advancing age in the absence of clinically detectable prostate cancer has been observed in other investigations. In a study of 5,220 patients with such traits, Dalkin and colleagues 28 found the upper limit of normal (95th percentile) to be 3.7 ng/ml for men 50 to 54 years of age, 4.0 ng/ml for men 55 to 59 years of age, 5.4 ng/ml for men 60 to 64 years of age, 6.2 ng/ml for men 65 to 69 years of age, and 6.6 ng/ml for men between 70 and 74 years of age, inclusive. Crawford and coworkers, 29 analyzing the data obtained during Prostate Cancer Awareness Week in 1992, observed a similar phenomenon. In another well-controlled, community-based investigation of almost 500 men with no evidence of prostate cancer, the serum PSA concentration was found to increase with each advancing decade of life. 30 The mean serum PSA value was 1.5 ng/ml for men 40 to 49 years of age, 1.6 ng/ ml for men 50 to 59 years of age, 3.1 ng/ml for men 60 to 69 years of age, and 3.3 ng/ml for men 70 to 79 years of age. When these investigations are analyzed collectively, the implication is that the "clinically appropriate" reference range for serum PSA depends on the patient's age. Elderly men have higher serum PSA values than do young men in the absence of clinical prostate cancer. The main reason for the increase in serum PSA level with advancing age is the concomitant increase in prostatic size; however, an enlarging prostate gland does not completely explain such an increase. Other factors that may contribute to the increase in serum PSA concentration in elderly men include subclinical prostatitis, prostatic ischemia or infarction, the presence of undetected prostate cancer, and the possibility of the prostate "leaking" PSA. Because the serum PSA concentration increases at a slightly greater rate than does the prostatic volume over the entire age range, the PSA density cutoff point of 0.15 may also necessitate adjustment for age. The Table 1. Age-Specific Reference Ranges* for Serum Prostate-Specific Antigent Age (yr) Serum PSA (ng/ml) *Upper limit defined as the 95th percentile. tpsa = prostate-specific antigen. Modified from Oesterling and associates. 23 American Medical Association. PSA density By permission of the

6 64 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES Mayo Clin Proc, January 1994, Vol 69 age-specific reference ranges for PSA density as suggested by Oesterling and associates 23 are listed in Table 1. The use of age-specific reference ranges has the potential to make serum PSA a more sensitive tumor marker for men younger than 60 years and a more specific tumor marker for men older than 60 years. For the original study population of 471 men, if the newly developed age-specific reference ranges (Table 1) were used rather than the 0.0 to 4.0 ng/ml range, 3 times as many men younger than 60 years would be considered to have an abnormal value. 23 With the reference range of 0.0 to 2.5 ng/ml, a 44-year-old man with a serum PSA value of 2.7 ng/ml would be assessed further with TRUS and prostate biopsy. This examination may lead to the diagnosis of prostate cancer that otherwise would not have been detected. Indeed, young men with a life expectancy of 25 to 30 additional years should be assessed the most thoroughly. Alternatively, for older men, the currently recommended reference range of 0.0 to 4.0 ng/ml may be too stringent. With advancing age, the increase in prostatic volume and other nonspecific factors cause the serum PSA concentration to increase in the absence of clinically identifiable prostate cancer. Therefore, with the goal being to detect clinically significant cancers, having a higher reference range is appropriate. If the age-specific reference ranges were used rather than the 0.0 to 4.0 ng/ml range for men 60 to 79 years old, as suggested in the investigation by Oesterling and colleagues, 23 fewer patients would undergo TRUS and prostate biopsy (9% versus 15%, respectively). Thus, with the age-specific reference ranges, PSA determination should be a more selective tumor marker such that (1) clinically significant prostate cancers could be detected at an early, curable stage in men who are most likely to receive the greatest benefit from definitive therapy and (2) unnecessary diagnostic procedures would not be performed routinely in men who are unlikely to harbor a life-threatening prostatic malignant tumor or to benefit from therapy. Because of the newly discovered relationship between prostatic volume and serum PSA and between serum PSA and patient age, the issue of prostatic size and its influence on the serum PSA concentration can now be accounted for primarily by using age-specific reference ranges to determine who has a normal or an abnormal serum PSA concentration. If age-specific reference ranges are used to make clinical decisions, determination of PSA density may be unnecessary for early detection of prostate cancer. In the community-based study by Oesterling and coworkers, 23 only seven patients (1%) had an increased PSA density value who did not have an increased serum PSA level on the basis of the age-specific reference ranges for both variables. Subsequent sextant biopsy of the prostate gland in these patients revealed no adenocarcinoma of the prostate. In addition, the overall correlation between serum PSA and PSA density, based on age-specific reference ranges, was 96%. Thus, on the basis of these preliminary data, when age-specific reference ranges are used for serum PSA, determination of PSA density may not provide much additional clinical information over that obtained from only the serum PSA value. In light of these new findings, the diagnostic algorithm presented in Figure 4 is now used to assess men at risk for the development of prostate cancer. 31 In simple terms, if the serum PSA level is less than or equal to the age-specific reference range (Table 1) and the findings on DRE are unremarkable, the patient should undergo annual examinations; the prostate gland should be assessed for the development of palpable abnormalities, and the serum PSA concentration should be monitored for "rate of change." If the serum PSA level is greater than the age-specific reference range (Table 1) and the findings on DRE are normal, TRUS should be performed. After the echogenicity of the gland has been assessed, a biopsy specimen of any visible lesion should be obtained. In addition, a systematic, sextant biopsy of the remaining prostatic tissue should be done. If none of these cores contains tissue from the transition zone, two additional specimens (one from each side) from the anterior part of the prostate should be obtained for a complete sample. Ifthe findings on DRE are abnormal, irrespective of the serum PSA level, the patient should undergo TRUS. With ultrasound guidance, biopsy specimens of the palpable abnormality and of all hypoechoic lesions should be obtained. In addition, a systematic, sextant biopsy of the remaining prostate gland should be performed. This approach should enable practicing clinicians to detect more clinically significant prostate cancers at an early, potentially curable stage. Nonetheless, as more is learned about the PSA s Age-specific range* > Age-specific range" Any value DRE Diagnostic Action Negative -* Annual PSA and DRE Negative Positive a years: ng/ml; years: ng/ml years: ng/ml; years: ng/ml TRUS: Biopsy visible lesions; sextant biopsy of remaining prostate, with two cores containing transition zone -* TRUS: Biopsy palpable and visible lesions; sextant biopsy of remaining prostate Fig. 4. Diagnostic algorithm for using age-specific prostate-specific antigen (PSA) reference ranges and digital rectal examination (DRE) to detect clinically significant prostate cancers at early, curable stage. TRUS = transrectal ultrasonography. (From Oesterling and associates. 31 By permission of Saunders.)

7 Mayo Clin Proc, January 1994, Vol 69 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES 65 tumor marker PSA and about prostate cancer, modifying this diagnostic approach may be necessary. EFFECT OF DRE ON THE SERUM PSA CONCENTRATION Physicians, both urologists and nonurologists, speculate that performing a DRE before measuring the serum PSA concentration iatrogenically increases the PSA value and thus leads to needless further evaluation of the prostate. Therefore, to avoid a false-positive result, many physicians wait 2 to 3 weeks after a DRE of the prostate gland with unexpectedly suspicious findings before they measure the serum PSA value. Several studies, however, have not supported the necessity of delaying the serum PSA determination after a DRE. Brawer and associates 32 measured the serum PSA concentration in 26 men both before and after DRE and found that the PSA value did not change significantly. A possible criticism of their method is that the PSA levels were determined only 5 and 30 minutes after the DRE, and thus the PSA may not have had enough time to diffuse across the physiologic barriers between the prostatic ductal system and the capillaries (basal cell layer, basement membrane, stromal tissue, capillary basement membrane, and capillary endothelial cells). Stamey 33 studied the effect of prostatic massage (a more vigorous manipulation than the prostate is subjected to during a routine DRE) on the serum PSA concentration in 16 patients. He noted a mean 1.92-fold increase in the serum PSA value (P = 0.05), as determined with the Pros-Check PSA assay (Yang Laboratories, Bellevue, Washington). This polyclonal assay has the theoretic potential to recognize several epitopes and thereby-be more sensitive than a monoclonal assay. Nevertheless, neither Brawer and associates nor Stamey examined a large group of patients or had a control group in their investigation. In a multicenter trial, Crawford and colleagues 34 measured the serum PSA level before and after DRE in 2,754 healthy men 40 years of age or older who were being screened for prostate cancer. The patients were classified into four groups on the basis of their initial serum PSA levels. In the groups with the lowest PSA levels (0.1 to 4.0 ng/ml and 4.1 to 10.0 ng/ml), no significant difference was noted between the pre-dre and the post-dre PSA levels. In the group whose initial PSA level was 10.1 to 20.0 ng/ml, the trend was toward a statistically significant increase (P = 0.053), and in the group whose initial PSA value was more than 20.0 ng/ml, a statistically significant increase was noted in the PSA level (P = 0.01). Thus, although the two groups with the highest PSA levels had statistically significant increases, these groups would require prostatic evaluation regardless of the second PSA level. In the two groups whose PSA values were less than 10.0 ng/ml, no statistically significant increase was noted in the serum PSA concentration after DRE; as a result, these investigators concluded that management was unaffected. Therefore, in their study, the increase in the serum PSA level caused by DRE was not clinically significant. In a similar study by Thomson and Clejan 35 that compared the PSA values before and after DRE in 2,736 men screened for prostate cancer, the PSA value increased from less than 4.0 ng/ml to more than 4.0 ng/ml in fewer than 2% of the patients, and for most men the serum PSA level remained at less than 5.0 ng/ml. The investigators' conclusion was similar to that of the others: DRE causes a small and clinically insignificant increase in the serum PSA concentration, and this outcome should be of limited concern to clinicians. In a prospective, randomized, controlled study that assessed the effect of DRE on the serum PSA level, Chybowski and coworkers 36 analyzed data from 71 study patients and 72 control subjects. In the study patients, the serum PSA value was determined with the Tandem-R PSA assay before DRE and from 2 to 30 hours after DRE; in the control group, the two serum PSA values were measured 2 to 30 hours apart without an intervening DRE. The median change in the serum PSA concentration was 0.4 ng/ml in the study group and -0.1 ng/ml in the control group (P< ) (Fig. 5). The increased post-dre PSA value (above the cutoff point of 4.0 ng/ml) resulted in two patients in the study group (3%) and one patient in the control cohort (1%) undergoing TRUS that they would not have had after the first determination; one man from each group underwent prostate biopsy when the serum PSA value was increased on the second determination (above the cutoff point of 10.0 ng/ in serum PSA, ng/ml 8 8, - o> C o -8 i 1 * I * 1 Study n = 71 Group m i Control n = 72 Ν ^ Fig. 5. Effect of digital rectal examination (DRE) on serum prostate-specific antigen (PSA) concentration for men in study group (order of tests: PSA, DRE, PSA) and men in control group (order of tests: PSA, PSA, DRE). (Data from Chybowski and associates. 36 )

8 66 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES Mayo Clin Proc, January 1994, Vol 69 ml). This study shows that the frequency of an increased PSA level as a result of DRE, and thus a change in the clinical management of a patient, equals that of routine variation in the results of laboratory assay (control cohort). In summary, the described studies show that the iatrogenically induced increase (false-positive result) in the serum PSA concentration after DRE is small and clinically insignificant. On the basis of these data, clinicians can be confident that a serum PSA value determined after a diagnostic, routine DRE is accurate and reliable; an increased PSA value in this setting is not indicative of a false-positive finding. EFFECT OF FINASTERIDE ON THE PSA CONCENTRATION Finasteride (Proscar) is a 5a-reductase inhibitor that has been approved by the Food and Drug Administration for the treatment of bladder outlet obstruction due to BPH. Finasteride competitively inhibits the conversion of testosterone to dihydrotestosterone, a factor that leads to a median decrease of 20% in the size of the prostate gland after 6 months of treatment. 37 The recommended oral dosage is 5 mg/day. 38 In a multicenter, double-blind, placebo-controlled study of 895 men, Gormley and associates 39 showed that patients treated with finasteride had significant objective and subjective improvement in voiding. At 12 months, the median serum PSA level had decreased by 50% in the 297 men receiving 5 mg of finasteride daily and by 48% in those taking 1 mg daily; however, the median serum PSA value of the men in the placebo group did not change significantly (Fig. 6). Because the patient population treated for BPH is the same population in which prostate cancer is also prevalent, concern exists that finasteride may iatrogenically depress the serum PSA value, an outcome that would mask and delay the diagnosis of early and potentially curable prostate cancer. Guess and coworkers 40 further analyzed the PSA data from some patients in the study by Gormley and colleagues 39 and found that 72% of 212 patients had a serum PSA value of less than 4.0 ng/ml and that 93% had a serum PSA level of less than 10.0 ng/ml at the time of initiation of finasteride therapy. After 12 months of finasteride therapy, the serum PSA concentration was less than 2.0 ng/ml in 75% of the patients and less than 5.0 ng/ml in 95%. The entire distribution of PSA values was shifted downward in a proportionate manner (Fig. 7). The Pearson correlation coefficients (placebo group: r = 0.91, 95% confidence interval = 0.86 to 0.94; finasteride group: r = 0.80, 95% confidence interval = 0.74 to 0.85) showed that the pretreatment and posttreatment PSA values were extremely well correlated. Guess and associates 4 ' further analyzed the available data from the North American phase III clinical trial of finasteride for the treatment of symptomatic BPH and defined PSA-monitoring criteria applicable to men receiving v c " \ O Placebo Finasteride, 5 mg Months after initiating finasteride therapy Fig. 6. Effect of finasteride therapy on serum prostate-specific antigen (PSA) value. Within 3 months after initiation of therapy, PSA level decreased to 50% of pretreatment value. (Data from Gormley and associates. 3 '') finasteride. They found that the serum PSA concentration for men treated with a 5-mg daily dose of finasteride for 12 months was shifted downward 50% from the pretreatment value (Fig. 7). Thus, for patients taking 5 mg of finasteride daily, a serum PSA value of 3.0 ng/ml corresponded with a level of 6.0 ng/ml in untreated men, and a serum PSA value of 8.0 ng/ml in the finasteride-treated patients equaled a level of 16.0 ng/ml in the untreated group. Similarly, the upper limit of the reference range for PSA in finasteridetreated patients is halfthat for untreated men. On the basis of the work of Guess and coworkers, 41 age-specific reference ranges are as applicable to finasteride-treated patients as they are to untreated men; the only difference is that the finasteride-adjusted, age-specific reference ranges are decreased by 50%. The usefulness of the serum PSA value as an aid in detecting early prostate cancer in finasteride-treated men is currently unknown. To date, 13 patients receiving finasteride therapy for BPH in the North American phase III clinical trial of finasteride have been diagnosed with prostate cancer. 41 The method of diagnosis was DRE in seven patients (54%), increased PSA value in four (31%), hemospermia in one (8%), and abdominal ultrasonography in one (8%). The median percentage change in the PSA value from baseline to the time of diagnosis was -30% (range, -64% to +12%). At the time of diagnosis, 12 of 13 patients (92%) had a serum PSA concentration that was more than the finasteride-adjusted, age-specific reference range. The minimal duration of finasteride therapy until prostate cancer was diagnosed was 346 days (mean, 562; range, 346 to 1,014). The mean change in the serum PSA level from 6 months (nadir PSA level) to the time of diagnosis of cancer was

9 Mayo Clin Proc, January 1994, Vol 69 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES K if Ï* -, ί 1 f 1 V!' 1 ' 0.0 C 2 4 S \.^ 10 ^. PSA (ng/ml) ' 15 Baseline Month 12 Fig. 7. Cumulative distribution of prostate-specific antigen (PSA) values at baseline (solid line) and at month 12 (dashed line) among men with benign prostatic hyperplasia but without known prostate cancer who received 5-mg daily dose of finasteride in North American phase III clinical trials. Note that PSA values for specific proportion of patients are shifted downward by 50% after 12 months of finasteride treatment. (From Guess and associates. 41 By permission of Saunders.) +11.5%. Thus, on the basis of this study of a small number of patients, the sensitivity of PSA for detecting prostate cancer does not seem compromised as long as the finasteride-induced decrease (50%) in the serum PSA concentration is considered. In a patient treated with finasteride who has less than a 50% decrease in the serum PSA value from the pretreatment level or who has a serum PSA concentration of more than the finasteride-adjusted, age-specific reference range while receiving the therapy, the physician should be cognizant of the possibility of prostate cancer. Immediately before initiating finasteride treatment of symptomatic BPH, the physician should determine the serum PSA level and perform a baseline DRE. After 6 months of finasteride therapy, a second serum PSA determination should be obtained. If the serum PSA value has not decreased by 50%, the patient should be examined for prostate cancer. Any patient taking finasteride whose 6-month serum PSA value has decreased by 50% or more should be monitored thereafter with an annual DRE and serum PSA determination. The finasteride-adjusted, age-specific reference ranges, which are 50% of the age-specific reference ranges for untreated men, should be considered when subsequent serum PSA values are interpreted (Fig. 7). Patient compliance with finasteride therapy can be monitored by measuring the serum dihydrotestosterone level, which should be suppressed by more than 60% if the patient is taking the recommended dose. 37 If a question exists about whether the patient is taking the medication (an increasing serum PSA concentration), compliance should be verified before an evaluation for prostate cancer is initiated. More data, however, are necessary before recommendations can be made about the use of PSA density and PSA velocity as aids in cancer detection and surveillance in patients who are receiving finasteride. In conclusion, the effect of finasteride therapy (5 mg daily) on the serum PSA concentration is to decrease the level to approximately 50% of the baseline value after 6 to 12 months of treatment. The age-specific PSA reference ranges for patients taking finasteride are 50% of the age-specific reference ranges for men who are not receiving this medication. Men who are beginning finasteride treatment should have a baseline DRE and determination of the serum PSA level. After 6 months of therapy, the serum PSA value should show a 50% decrease from the baseline level. If this decline does not occur, the patient should be examined for prostate cancer. Patients who receive finasteride treatment should have an annual DRE and serum PSA determination. If the serum PSA level becomes increased above the finasteride-adjusted, age-specific reference range or the DRE results become abnormal, the patient should be assessed for prostate cancer. More information is needed before PSA density and PSA velocity can be used for cancer detection in finasteride-treated patients. CONCLUSION Currently, PSA is the most important, accurate, and clinically useful tumor marker for prostate cancer. Regardless of the clinical situation, the serum PSA concentration is superior to the serum prostatic acid phosphatase level. Evaluation of the serum PSA concentration can lead to the diagnosis of clinically significant prostate cancer at an early and potentially curable stage, and in some patients, the PSA value may be the only indication of cancer. PSA density, PSA velocity, and age-specific reference ranges are all factors used in an attempt to enhance the ability to distinguish BPH from early, clinically significant prostate cancer. The serum PSA concentration determined after DRE is accurate and dependable; delaying a PSA determination after DRE is unnecessary. Finasteride treatment of BPH decreases the serum PSA value by approximately 50%; thus, the finasteride-adjusted, age-specific PSA reference ranges are 50% of the age-specific reference ranges for untreated men. The clinical significance of this phenomenon relative to use of the PSA concentration to detect early prostate cancer in these men is unknown. Although further studies are needed to address these issues, evaluation of the prostate gland is currently recommended if the serum PSA value does not decrease below 50% of the pretreatment level after therapy has been initiated or if it increases beyond the finasterideadjusted, age-specific reference range while the patient is receiving the medication.

10 68 PROSTATE-SPECIFIC ANTIGEN: CRITICAL ISSUES Mayo Clin Proc, January 1994, Vol 69 REFERENCES 1. Wang MC, Valenzuela LA, Murphy GP, Chu TM. Purification of a human prostate specific antigen. Invest Urol 1979;17: Wang MC, Papsidero LD, Kuriyama M, Valenzuela LA, Murphy GP, Chu TM. Prostate antigen: a new potential marker for prostatic cancer. Prostate 1981;2: Sensabaugh GF. Isolation and characterization of a semen-specific protein from human seminal plasma: a potential new marker for semen identification. J Forensic Sci 1978;23: Lilja H. A kallikrein-like serine protease in prostatic fluid cleaves the predominant seminal vesicle protein. J Clin Invest 1985; 76: Boring CC, Squires TS, Tong T. Cancer statistics, CA Cancer JClin 1993;43: Jewett HJ, Bridge RW, Gray GF Jr, Shelley WM. The palpable nodule of prostatic cancer: results 15 years after radical excision. JAMA 1968;203: Johansson J-E, Adami H-O, Andersson S-O, Bergström R, Holmberg L, Krusemo UB. High 10-year survival rate in patients with early, untreated prostatic cancer. JAMA 1992;267: Catalona WJ, Smith DS, Ratliff TL, Dodds KM, Coplen DE, Yuan JJJ, et al. Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 1991; 324: Brawer MK, Chetner MP, Beatie J, Büchner DM, Vessella RL, Lange PH. Screening for prostatic carcinoma with prostate specific antigen. J Urol 1992;147: Labrie F, Dupont A, Suburu R, Cusan L, Tremblay M, Gomez J-L, et al. Serum prostate specific antigen as a pre-screening test for prostate cancer. J Urol 1992;147: Cupp MR, Oesterling JE. DRE, PSA and TRUS : a proposal for their efficient use in early detection of clinically significant prostate cancer. AUA Today 1992; 5(No. 2): Cooner WH, Mosley BR, Rutherford CL Jr, Beard JH, Pond HS, Terry WJ, et al. Prostate cancer detection in a clinical urological practice by ultrasonography, digital rectal examination and prostate specific antigen. J Urol 1990;143: Babaian RJ, Mettlin C, Kane R, Murphy GP, Lee F, Drago JR, et al. The relationship of prostate-specific antigen to digital rectal examination and transrectal ultrasonography: findings of the American Cancer Society National Prostate Cancer Detection Project. Cancer 1992;69: Mettlin C, Jones G, Averette H, Guseberg SB, Murphy GP. Defining and updating the American Cancer Society guidelines for the cancerrelated checkup: prostate and endometrial cancers. CA Cancer JClin 1993;43: AUA policy statement: early detection of prostate cancer and use of transrectal ultrasound. In: American Urological Association 1992 Policy Statement Book. Baltimore: Williams & Wilkins, 1992: Steinberg GD, Carter BS, Beaty TH, Childs B, Walsh PC. Family history and the risk of prostate cancer. Prostate 1990;17: Giovannucci E, Ascherio A, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. A prospective cohort study of vasectomy and prostate cancer in US men. JAMA 1993;269: Oesterling JE. Prostate-specific antigen: improving its ability to diagnose early prostate cancer [editorial]. JAMA 1992; 267: Benson MC, Whang IS, Pantuck A, Ring K, Kaplan SA, Olsson CA, et al. Prostate specific antigen density: a means of distinguishing benign prostatic hypertrophy and prostate cancer. J Urol 1992; 147: Benson MC, Whang IS, Olsson CA, McMahon DJ, Cooner WH. The use of prostate specific antigen density to enhance the predictive value of intermediate levels of serum prostate specific antigen. J Urol 1992; 147: Andriole GL, Telle WB, Coplen DE, Catalona WJ. PSA index (PSAI) as a predictor of prostate cancer (CaP) in men with persistent serum PSA elevation [abstract]. J Urol 1992; 147:387A 22. Clements R, Penney MD, Etherington RJ, Griffiths GJ, Hughes H, Peeling WB. Volume of normal prostate, of prostate cancer, and of benign prostatic hyperplasia: are correlations with prostatic specific antigen clinically useful? Prostate Suppl 1992; 4: Oesterling JE, Jacobsen SJ, Chute CG, Guess HA, Girman CJ, Panser LA, et al. Serum prostate-specific antigen in a community-based population of healthy men: establishment of age-specific reference ranges. JAMA 1993;270: Kleer E, King BF, Oesterling JE, Weaver AL, Lewis RW. Estimation of prostatic volume: comparison of transrectal ultrasonography and magnetic resonance imaging using various formulas with pathologic correlation. J Androl [in press] 25. Weber JP, Oesterling JE, Peters CA, Partin AW, Chan DW, Walsh PC. The influence of reversible androgen deprivation on serum prostate-specific antigen levels in men with benign prostatic hyperplasia. J Urol 1989; 141: Carter HB, Pearson JD, Metter EJ, Brant LJ, Chan DW, Andres R, et al. Longitudinal evaluation of prostate-specific antigen levels in men with and without prostate disease. JAMA 1992;267: Oesterling JE, Chute CG, Jacobsen SJ, Guess HA, Panser LA, Johnson CL, et al. Longitudinal changes in serum PSA (PSA velocity) in a community-based cohort of men [abstract]. J Urol 1993; 149:412A 28. Dalkin BL, Ahmann F, Southwick P, Bottaccini MR. Derivation of normal prostate specific antigen (PSA) level by age [abstract]. J Urol 1993; 149:413A 29. Crawford ED, DeAntoni EP, Stone NW, Blum DS Prostate Cancer Awareness Week: benefits and shortcomings of communitybased screening for prostate cancer [abstract]. J Urol 1993; 149:214A 30. Collins GN, Lee R-J, McKelvie GB, Rogers ACN, Hehir M. Relationship between prostate specific antigen, prostate volume and age in the benign prostate. BrJUrol 1993;71: Oesterling JE, Cooner WH, Jacobsen SJ, Guess HA, Lieber MM. The influence of patient age on the serum prostate-specific antigen concentration: an important clinical observation. Urol Clin North Am 1993 Nov; 20: Brawer MK, Schifman RB, Ahmann FR, Ahmann ME, Coulis KM. The effect of digital rectal examination on serum levels of prostaticspecific antigen. Arch Pathol Lab Med 1988;112: Stamey TA. Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. Monogr Urol 1989; 10: Crawford ED, Schutz MJ, Clejan S, Drago J, Resnick MI, Chodak GW, et al. The effect of digital rectal examination on prostatespecific antigen levels. JAMA 1992;267: Thomson RD, Clejan S. Digital rectal examination-associated alterations in serum prostate-specific antigen. Am J Clin Pathol 1992; 97: Chybowski FM, Bergstralh EJ, Oesterling JE. The effect of digital rectal examination on the serum prostate specific antigen concentration: results of a randomized study. J Urol 1992;148: McConnell JD. Current medical therapy for benign prostatic hyperplasia: the scientific basis and clinical efficacy of finastende and alpha-blockers. In: Walsh PC, Retik AB, Stamey TA, Vaughn ED Jr, editors. Campbell's Urology. 6th ed. Update no. 3. Philadelphia: Saunders, 1992: Food and Drug Administration. Proscar (finasteride) package insert May 39. Gormley GJ, Stoner E, Bruskewitz RC, Imperato-McGinley J, Walsh PC, McConnell JD, et al. The effect of finasteride in men with benign prostatic hyperplasia. N Engl J Med 1992; 327: Guess HA, Heyse JF, Gormley GJ. The effect of finasteride on prostate specific antigen in men with benign prostatic hyperplasia. Prostate 1993;22: Guess HA, Heyse JF, Gormley GJ, Stoner E, Oesterling JE. The effect of finasteride on serum PSA concentrations in men with benign prostatic hyperplasia: results from the North American phase III clinical trial. Urol Clin North Am 1993 Nov; 20: