As far as the laws of mathematics refer to reality, they are not certain, and as far as they are certain, they do not refer to reality. -- Albert Einstein
Dr. DeAntoni has carefully reviewed the literature on age-specific reference ranges for prostate-specific antigen (PSA) in the diagnosis of prostate cancer and the controversy surrounding their use. Key to understanding of this debate are two fundamental concepts: (1) the definition of "clinically significant prostate cancer" and (2) the use of sensitivity and specificity, which is frequently obscured by the surrounding rhetoric. The assumption that all readers uniformly interpret the meaning of clinically significant prostate cancer and wish to achieve the same results by manipulating sensitivity and specificity is probably incorrect.
How to Define Clinically Significant Prostate Cancer?
The definition of "clinically significant prostate cancer" is currently imprecise. Some prostate cancers are particularly virulent and pose a significant risk of death. At the other extreme, many men will die of other causes while harboring occult microscopic disease. Most men with clinically detectable prostate cancer fall somewhere in between. The categorization of clinically significant prostate cancer in these individuals is less clear.
My definition of a biologically or clinically significant tumor is one that poses a significant risk to the health of an individual. The two variables that affect this definition are the aggressiveness of the tumor and the life expectancy (period of risk) of the individual. Even a relatively slow-growing neoplasm may be a significant threat if the man is young, otherwise healthy, and has a long life expectancy.
Sensitivity and Specificity of PSA
It must be recalled that the sensitivity and specificity of a test cannot be improved simultaneously. The sensitivity and specificity of an assay can be depicted by receiver-operating-characteristic (ROC) analysis. The PSA assay, when used for detecting prostate cancer, performs relatively poorly by this yardstick. There is some hope that a new assay that measures free PSA will result in a net improvement in both sensitivity and specificity. However, as Dr. DeAntoni points out, the free/total PSA ratio has engendered its own controversy, which shows no sign of abating.
Despite less-than-optimal performance characteristics, the measurement of PSA has been demonstrated to improve the detection of prostate cancer and to enhance the recognition of disease recurrence. The clinical utility of this assay is obvious. Nevertheless, controversy continues over its use in prostate cancer screening.[2,3]
Age-Specific Reference Ranges
The arguments for age-specific reference ranges are thoughtfully discussed by Dr. DeAntoni. However, it must be remembered that, in the manipulation of sensitivity and specificity, you can't have your cake and eat it too. Lowering the "normal" PSA threshold in younger men will increase the detection of prostate cancer but will also increase the number of biopsies performed in men who do not harbor this disease. Raising the "normal" PSA threshold in older men will decrease the number of biopsies but will miss prostate cancers in some men.
Is this a good trade-off? There is evidence suggesting that a normal reference range of 4 ng/mL may be too conservative. The Physicians' Health Study demonstrated that men with a PSA value between 2.0 and 3.0 ng/mL had a 5.5-fold higher relative risk of developing prostate cancer than men with PSA levels less than 1.0 ng/mL.
With our current state of knowledge, it appears rational to increase the detection of prostate cancer in younger men who are at a greater risk of dying of their disease because of their longer life expectancy. Although the determination of life expectancy is imprecise, urologists have demonstrated a remarkably good record of selecting appropriate candidates for radical prostatectomy. A review of the literature showed that the risk of dying of intercurrent disease was lower in patients who underwent radical prostatectomy than in men managed with radiation therapy or observation only.
An Important Question
Dr. DeAntoni poses an important question: Can age-specific reference ranges for PSA improve the detection of clinically significant prostate cancer? The answer is relative, not absolute, varying with the responder's perspective. It is increasingly evident that lowering PSA values in younger men will increase the detection of biologically important, clinically significant, or life-threatening disease (depending on the nomenclature that one finds most appealing).
The benefit to the other end of the clinical spectrum is more obscure. Use of age-specific reference ranges decreases the rate of detection of prostate cancer in older men. In many men, this is appropriate because they will succumb to intervening illnesses. However, not all 70-year-old men are identical, and some of these men will benefit from early detection and therapy.
A Better Question?
Thus, the posed question, while important, ignores an important part of the diagnostic spectrum. A better question may be, when should screening for PSA stop? Most urologists would agree that there is little need for PSA screening of asymptomatic men who are 90 years of age. However, the age at which prostate cancer screening should cease in asymptomatic individuals has received little attention and is not defined. Recommending PSA screening without suggesting an upper age limit is doing our patients a disservice.
1. Marley GM, Miller CM, Kattan MW, et al: Free and complexed prostate-specific
antigen serum ratios to predict probability of primary prostate cancer
and benign prostatic hyperplasia. Urology 48(6A):16-22, 1996.
2. Collins MM, Barry MJ: Controversies in prostate cancer screening:
Analogies to the early lung cancer screening debate. JAMA 276:1976-1979,
3. Crawford ED: Prostate cancer: To screen or not to screen? Oncology
4. Gann PH, Hennekens CH, Stampfer MJ: A prospective evaluation of plasma
prostate-specific antigen for detection of prostatic cancer. JAMA 273:289-294,
5. Aus G, Hugosson J, Norlen L: Long-term survival and mortality in
prostate cancer treated with noncurative intent. J Urol 154:460-465, 1995.
6. Adofsson J, Steineck G, Whitmore WF Jr: Recent results of management
of palpable clinically localized prostate cancer. Cancer 72:310-322, 1993.