The Journal of Urology
Volume 178, Issue 5 , Pages 1929-1932, November 2007

Current Age and Race Adjusted Prostate Specific Antigen Threshold Values Delay Diagnosis of High Grade Prostate Cancer

Department of Urology, University of Texas Health Sciences Center-San Antonio, San Antonio, Texas

Received 8 April 2007 published online 14 September 2007.

Article Outline

Purpose

Age specific prostate specific antigen ranges have been advocated to increase the predictive value of prostate specific antigen based on increases that occur with aging. We suggest that prostate specific antigen is not a dichotomous biomarker and age specific reference ranges delays the diagnosis of high grade prostate cancer in older and black American men.

Materials and Methods

Using the Prostate Cancer Prevention Trial risk calculator we evaluated the impact of age on the risk of high grade prostate cancer in white and black men. We created a hypothetical population of men by standardizing risk variables, including negative family history, normal digital rectal examination and no history of negative biopsy. Results were compared in the 2 populations using 5-year age increments from ages 55 to 75 years and increasing prostate specific antigen.

Results

Increasing age was associated with a higher risk of high grade prostate cancer in white and black men. The risk of high grade prostate cancer in a black man was twice that in a white man with the same prostate specific antigen at all prostate specific antigen values.

Conclusions

Age specific and race specific prostate specific antigen ranges are flawed. Many patients who would not be considered for biopsy based on these prostate specific antigen ranges are at significant risk for high grade prostate cancer. The risk of high grade prostate cancer in black men is twice that in white men. Risk assessment in black men and older men is necessary to diagnose high grade prostate cancer when treatment can be effective.

Key Words: prostate, prostatic neoplasms, prostate-specific antigen, African Americans, European continental ancestry group

Abbreviations and Acronyms: CAP, prostate cancer, DRE, digital rectal examination, PCPT, Prostate Cancer Prevention Trial, PSA, prostate specific antigen

 

In the last 2 decades PSA has been extensively used for the early detection of CAP. More than half of American men older than 50 years undergo regular PSA screening. Since the mid 1980s, PSA has been used predominantly as a dichotomous biomarker (normal or abnormal) with a threshold value of 4.0 ng/ml prompting prostate biopsy. There is a long history of attempts to define optimal PSA cutoffs to increase its performance characteristics. Recognizing that PSA and its variability increases with increasing age, the first suggestions focused on age specific PSA threshold values. Oesterling1 and Anderson2 et al noted that the upper limit of normal or PSA cutoff achieving 95% specificity would increase with age. Therefore, to maintain equitable 95% specificity or an equivalent 5% false-positive rate across different ages the PSA cutoff for recommending biopsy must necessarily increase with age. Based on cohorts of 537 patients with cancer-free status and benign prostatic hyperplasia,1 and 1,716 cancer-free, symptom-free, healthy volunteers who were 40 to 79 years old2 they derived proposed reference ranges (table 1). Relative to using a uniform PSA cutoff of 4.0 or 2.5 ng/ml for all ages, the age specific PSA cutoffs would result in higher sensitivity and higher false-positive rates for younger men and lower sensitivity for older men. In the time since the study by Oesterling et al1 other groups have suggested that alternate age specific reference ranges should be applied.3, 4, 5, 6

Table 1.
ReferencesAge Group PSA Cutoff (ng/ml)
40–4950–5960–6970–79
Oesterling et al12.53.54.56.5
Anderson et al21.52.54.57.5
Present series:
White2.53.53.53.5
Black2.04.04/55/5

Specificity 95% for Oesterling1 and Anderson2 et al, and sensitivity 95% for present series.

Morgan et al reported that most of these ranges were derived from white populations.7 Based on a retrospective analysis of 3,174 white and 2,084 black American men who underwent prostate biopsy for various causes the investigators arrived at cutoffs based on maintaining 95% sensitivity across age and race groups (table 1). They found that serum PSA in black men was significantly higher than in white men and the value correlated directly with age. Morgan et al suggested that in this high risk population 95% sensitivity, as opposed to the alternative approach of 95% specificity, may be used with acceptable decrements in specificity.

Similar studies of the validation of age specific PSA reference ranges have been performed in diverse populations, such as Arab, Chinese, Japanese and Jordanian men.8, 9, 10 To date these age and race specific PSA reference ranges have been commonly used to recommend prostate biopsy. For example, if a 70-year-old white man and a 55-year-old white man presented with PSA 6 ng/ml, it is likely that the 70-year-old man would be told that his PSA was normal for age, while the 55-year-old would be told that his PSA was abnormal and he would be offered biopsy. We quantified how current age and race adjusted PSA threshold values may delay the diagnosis of high grade CAP by applying the recently validated PCPT risk calculator.11

Back to Article Outline

Methods and Materials 

We used the on-line CAP risk calculator developed from the PCPT11 to assess the probability of high grade (Gleason grade 7 or greater) CAP in black and white American men. The calculator is based on logistic regression analysis of the risk of CAP and high grade disease associated with age at biopsy, race (black vs not black), family history of CAP (father, brother or son), PSA, DRE and previous prostate biopsy. Risk equations were created from the logistic regression models.

We used the PCPT risk calculator to determine the risk of high grade CAP in white and black American men. To achieve this we created 2 hypothetical populations of men (1 white and 1 black) by standardizing other variables, such as negative family history of CAP, normal DRE and no history of negative biopsy. We selected 5-year age ranges beginning at age 55 years, representing the lower PCPT age limit, and ending at age 75 years, as the life expectancy of a man in the United States approaches 10 years. We used increasing PSA of 0.5 to 10 ng/ml in 0.5 ng/ml increments. We applied the PCPT risk calculator to these populations across several PSA and age ranges to calculate the risk of high grade CAP in the white and black populations.

Back to Article Outline

Results 

Table 2 lists the results of the PCPT risk calculator for the risk of high grade CAP in an individual. Underlined numbers illustrate the PSA threshold values at which biopsy would be recommended using the age adjusted PSA thresholds of Oesterling et al.1 In the white population a 55-year-old man would be recommended to undergo biopsy when the risk of high grade disease was 5%. Conversely a black man with identical risk factors would be recommended to undergo biopsy when the risk was 12%. For a 70-year-old white man the most commonly used age adjusted threshold PSA would require him to be at 16% risk for high grade disease before biopsy would be recommended. For a similar 70-year-old black man this risk would have to be 33% before biopsy would be recommended. Thus, the use of age specific cutoffs implies that the risk of high grade CAP would exceed 5% and on average lie near 10% for white men, and exceed 12% and on average lie near 22% for black men. The risk of high grade disease in black men is double or more compared to that in white men at every age specific recommended cutoff.

Table 2. Risk of biopsy detected high grade CAP in men with normal DRE, no family history of CAP and no previous prostate biopsy
AgePSA (ng/ml)
0.51.01.52.02.53.03.54.04.55.05.56.06.5
White:
55012234567891011
600123456789101112
6511234678910111314
70123456891012131416
751235679101213151618
Black:
55134681012141618202223
60135791214161820222426
651468111316182123252729
7024710121518212326283033
7525811141720232629313436

Underlined values indicate threshold PSA at which biopsy would be recommended.1

Back to Article Outline

Discussion 

Shortly after widespread PSA testing began in the United States age adjusted PSA reference ranges were recommended for CAP screening. These recommendations seemed reasonable, given the state of knowledge at the time. 1) CAP was unlikely at levels below 4.0 ng/ml, 2) PSA increased as a man aged and 3) older men may be less likely to benefit from CAP detection and treatment. With the understanding that black men were at greater risk for CAP different threshold PSA levels were developed specifically for this group of men. American College of Physicians guidelines suggest that patients should be referred to a urologist if an abnormality is noted on DRE, or if PSA is more than 4 ng/ml.12 These guidelines also recommend that there may be little benefit in screening men who are older than 69 years. Age specific reference ranges have been proposed and practiced by several groups within the United States and globally.13

We now understand that PSA is not normal or abnormal. Rather, PSA reflects a continuum of CAP risk detected by biopsy14 and it is a superior biomarker for high grade CAP (the most lethal form of the disease) instead of a test for CAP in general.15 Other risk factors, including age, race, family history of CAP, rectal examination and prostate biopsy history, independently contribute to the risk of cancer and high grade disease.11 Therefore, against the backdrop of the often used age adjusted PSA values for referral to prostate biopsy we determined how often men with high grade CAP might be told that biopsy was unnecessary.

This analysis clearly demonstrates that age adjusted PSA values can dramatically increase the likelihood that a man with high grade CAP, especially a black man, would not receive a recommendation for prostate biopsy and there would be a consequent delay in diagnosis of the disease. The current standard of 10 to 12 core biopsies as opposed to the sextant biopsies of the PCPT would most likely increase the risk that high grade cancers would be missed using age specific PSA reference ranges, especially in older men.

Additional observations are important. As PSA increases, the risk that the disease is no longer organ confined also increases.16 As a result, the delay resulting from waiting for higher PSA in these men could significantly compromise treatment effectiveness by the time that it is finally diagnosed. Two pieces of data illustrate this risk. 1) In the United States the life expectancy of a 75-year-old white man and a 75-year-old black man is 10.5 vs 9.9 years.17 2) The 10-year mortality rate for high grade CAP approaches 50%. The combination of these observations coupled with a delay in the diagnosis of high grade disease with age adjusted PSA cutoffs may have seriously compromised the impact of early detection of CAP in the United States in the last 2 decades.

The purpose of the PCPT risk calculator is to incorporate commonly available and validated markers of risk into a calculation of risk that is specific for a given patient. The risk calculator was developed in an extremely well characterized cohort of 5,519 men and it was subsequently validated in 3 external populations.18 As such, this estimate of the true risk of cancer at biopsy is far more accurate than other estimates, which are severely biased by a lack of ascertainment at lower PSA levels. We anticipate that other risk factors/biomarkers will ultimately be incorporated. One such biomarker, PSA velocity, was examined in our initial study. Alone it highly correlated with the risk of cancer and high grade cancer. When examined against the other risk factors, PSA velocity did not contribute independent information to an estimate of the cancer risk in an individual.

It is possible that low grade, low volume CAP would increase if the PCPT risk calculator were widely used to predict the risk of CAP. Concurrently in men at higher risk the obverse of this concern is that the calculator may lead to an early diagnosis of high grade and, thus, high risk CAP. The latter of these options is supported by recent studies suggesting that a decreased risk of mortality is associated with active treatment for low and intermediate risk CAP.19

The use of the risk calculator is a different approach to CAP screening with PSA and DRE. However, it is an approach that has widely used analogies in other areas of clinical medicine. For example, in cardiology decisions for preventive or therapeutic interventions are often based on the risk of cardiovascular disease in an individual within a given interval. The 10-year cardiac event risk calculator, which was developed from data from the Framingham study, is widely used.20 However, the risk calculator does not simplify clinical recommendations for biopsy, although it helps individualize patient decision making. As a result, a single threshold of risk at which biopsy would be recommended does not exist. The level of risk at which biopsy is elected by a patient would be based on the degree of concern regarding the CAP risk and the degree of aversion to medical interventions (biopsy and possible treatment) as well as on comorbidities and life expectancy.

Ultimately the goal of CAP early detection is clearly not detection, but a decrease in the risk of the morbidity and mortality of the disease. Of all prevalent CAPs the detection of those that are indolent and pose no risk to the patient would add little respite from the burden of this disease. Similarly and as the point of this study, the detection of biologically consequential tumors at such a late stage that treatment is ineffective would be unlikely to benefit men at risk. Our observation suggests that as a man ages and as his risk of high grade disease increases for the same PSA level, the clinician cannot rely on PSA alone but must constantly reevaluate the confluence of considerations related to the health priorities of an individual, his overall health, and his risk of aggressive and potentially lethal CAP.

Back to Article Outline

Conclusions 

These data suggest that applying age and race adjusted PSA reference ranges in men who have chosen to participate in PSA screening may dramatically increase the risk of a delayed diagnosis of high grade disease, a delay that is exacerbated in black men. Instead of previously published age and race specific cutoffs we recommend the use of the online PCPT CAP risk calculator, which integrates the currently established risk factors of age, race, PSA, family history and history of biopsy into a validated risk score of cancer and high grade disease. This calculator facilitates individualized decision making between patients and physicians with regard to the level of cancer risk when biopsy may be recommended.

Back to Article Outline

References 

  1. 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;270:860
  2. Anderson JR, Strickland D, Corbin D, Byrnes JA, Zweiback E. Age-specific reference ranges for serum prostate-specific antigen. Urology. 1995;46:54
  3. Catalona WJ, Hudson MA, Scardino PT, Richie JP, Ahmann FR, Flanigan RC, et al. Selection of optimal prostate specific antigen cutoffs for early detection of prostate cancer: receiver operating characteristic curves. J Urol. 1994;152:2037
  4. Dalkin BL, Ahmann FR, Kopp JB. Prostate specific antigen levels in men older than 50 years without clinical evidence of prostatic carcinoma. J Urol. 1993;150:1837
  5. DeAntoni EP, Crawford ED, Oesterling JE, Ross CA, Berger ER, McLeod DG, et al. Age- and race-specific reference ranges for prostate-specific antigen from a large community-based study. Urology. 1996;48:234
  6. Partin AW, Criley SR, Subong EN, Zincke H, Walsh PC, Oesterling JE. Standard versus age-specific prostate specific antigen reference ranges among men with clinically localized prostate cancer: a pathological analysis. J Urol. 1996;155:1336
  7. Morgan TO, Jacobsen SJ, McCarthy WF, Jacobson DJ, McLeod DG, Moul JW. Age-specific reference ranges for prostate-specific antigen in black men. N Engl J Med. 1996;335:304
  8. Kehinde EO, Mojiminiyi OA, Sheikh M, Al-Awadi KA, Daar AS, Al-Hunayan A, et al. Age-specific reference levels of serum prostate-specific antigen and prostate volume in healthy Arab men. BJU Int. 2005;96:308
  9. Wu TT, Huang JK. The clinical usefulness of prostate-specific antigen (PSA) level and age-specific PSA reference ranges for detecting prostate cancer in Chinese. Urol Int. 2004;72:208
  10. Battikhi MN. Age-specific reference ranges for prostate-specific antigen (PSA) in Jordanian patients. Prostate Cancer Prostatic Dis. 2003;6:256
  11. Thompson IM, Ankerst DP, Chi C, Goodman PJ, Tangen CM, Lucia MS, et al. Assessing prostate cancer risk: results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2006;98:529
  12. American College of Physicians. Screening for prostate cancer. Ann Intern Med. 1997;126:480
  13. Moul JW, Sun L, Hotaling JM, Fitzsimons NJ, Polascik TJ, Robertson CN, et al. Age adjusted prostate specific antigen and prostate specific antigen velocity cut points in prostate cancer screening. J Urol. 2007;177:499
  14. Thompson IM, Pauler DK, Goodman PJ, Tangen CM, Lucia MS, Parnes HL, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level < or =4.0 ng per milliliter. N Engl J Med. 2004;350:2239
  15. Thompson IM, Ankerst DP, Chi C, Lucia MS, Goodman PJ, Crowley JJ, et al. Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA. 2005;294:66
  16. Loeb S, Roehl KA, Yu X, Antenor JA, Han M, Gashti SN, et al. Lymphovascular invasion in radical prostatectomy specimens: prediction of adverse pathologic features and biochemical progression. Urology. 2006;68:99
  17. Mortality Rates for African Americans and Caucasian Americans. Available at http://www.cdc.gov/nchs/data/hus/hus06.pdf#027. Accessed March 1, 2007.
  18. Parekh DJ, Ankerst DP, Higgins BA, Hernandez J, Canby-Hagino E, Brand T, et al. External validation of the Prostate Cancer Prevention Trial risk calculator in a screened population. Urology. 2006;68:1152
  19. Wong YN, Mitra N, Hudes G, Localio R, Schwartz JS, Wan F, et al. Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA. 2006;296:2683
  20. Wilson PW, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:1837

 Supported by National Cancer Institute and Early Detection Research Network Grant U01-CA86402 and a University of Texas Health Sciences Center-San Antonio Institute for Integration of Medicine and Science Mentored Career Development Award (DJP).

PII: S0022-5347(07)01781-8

doi:10.1016/j.juro.2007.07.045

The Journal of Urology
Volume 178, Issue 5 , Pages 1929-1932, November 2007

Article Tools