Eight Misconceptions about Prostate-Specific Antigen

Prostate-specific antigen (PSA) was discovered in the 1970s, with assays to detect PSA in blood approved by the US Food and Drug Administration in 1986. Almost from the very beginning, its role as a prostate cancer marker has been severely misunderstood. One of the most well-known misconceptions is that prostate cancer detectable by elevated PSA in asymptomatic men requires treatment, irrespective of stage and grade. Accordingly, when PSA testing to detect prostate cancer began in the United States in the late 1980s and early 1990s, most men with elevated PSA were biopsied and nearly all cancers detected were treated. It is estimated that in its first 20 years of use, PSA screening led to >1 million Americans suffering harm from radiotherapy or surgery to treat a cancer that would never become apparent before they died of another cause (1).

More than 35 years later, PSA continues to be misunderstood. Here, we try to correct 8 common contemporary misconceptions about PSA. In brief, PSA may have a bad reputation, but it has remarkable properties—when used correctly—that make the clinical biochemistry of prostate cancer the envy of all other solid tumors.

Misconception 1: PSA Is a Molecule Detected in Blood

A PSA measurement such as 1.2 ng/mL suggests there is 1.2 ng of a molecule called PSA in each milliliter of blood. But what is measured by the PSA test is actually a variety of different molecules (2). As PSA is a serine proteinase (3), and there is huge excess of proteinase inhibitors in blood, most PSA circulates covalently bound to alpha 1-antichymotrypsin (“complexed PSA”) (4). Unbound PSA is known as “free” PSA, but actually comes in multiple distinct, noncatalytic proteoforms, such as the intact (single-chain) free PSA and “nicked” PSA, which is internally cleaved at Lys145 and/or Lys146 (5). Depending on the technique used, some PSA assays also measure human kallikrein-related peptidase 2 (hK2) (6), which has an 80% identical amino-acid sequence to PSA and the ability to convert the pro-PSA zymogen to the catalytically active enzyme (7). Assays to measure total PSA use monoclonal antibodies binding to nonoverlapping antigenic epitopes available on both free PSA and PSA bound to alpha 1-antichymotrypsin to enable equimolar detection of free and complex PSA. PSA measured by an assay typically comprises 65%–90% complexed PSA, 10%–35% free PSA, and 1%–2% hK2. Free PSA comprises roughly 50:50 nicked and intact forms (8).

Misconception 2: PSA in Blood Has a Half-Life of 2 Days

A highly influential paper published in 1987 (9) included an estimate that PSA was eliminated from blood with a half-life of a couple of days after radical prostatectomy. However, inferences based on the large difference in size of free compared to complexed PSA indicate that free roughly 28-kDa PSA—but not the roughly 90-kDa complexed PSA—is amenable to rapid elimination by glomerular filtration with an estimated half-life of <24 hours (10). Evidence of rapid elimination of free PSA (and hK2)—but not complexed PSA—by the kidneys has also been documented following the receipt of kidney transplants in men on hemodialysis due to chronic kidney disease (11). By contrast, the elimination rate of complexed PSA is very slow (10) with an estimated half-life of about 1 week and not significantly different from that of total PSA (12).

Misconception 3: PSA Is Either Elevated or Normal

Many laboratories report that PSA either is or is not “in the normal range,” typically using a cut-point of 4 ng/mL. Perhaps, as a result, physicians will tell patients “there is nothing to worry about” and patients will say “my test came back negative.” This is to treat PSA like many other blood markers where values within the normal range are essentially equivalent. For instance, there is no clinically relevant difference between a hemoglobin of 15 vs 13 g/dL, or between 6000 vs 8000 white blood cells per μL. Conversely, PSA is strongly associated with risk of clinically significant prostate cancer along a continuum, where the higher the PSA, the higher the risk, irrespective of PSA level. This means that risk varies greatly between similarly aged men with PSA within the “normal” (<4.0 ng/mL) range. A 60-year-old white man with PSA of 3.5 ng/mL has a 4% chance of high-grade prostate cancer being detected on biopsy vs only 1% if his PSA was 0.5 ng/mL (13); the risk of death from prostate cancer over the next 25 years is about 50-fold higher for a PSA of 3.5 vs 1 (14).

Misconception 4: PSA is a Diagnostic Marker for Prostate Cancer

PSA is sometimes treated like a pregnancy test: you could test every day and it would be negative, and then the patient becomes pregnant and the test would turn positive. Most prostate cancers, on the other hand, grow very slowly. Autopsy studies show that 10% to 20% of men who die of causes unrelated to cancer in their 40s have cancer detectable in their prostate (15). PSA similarly shows changes decades before prostate cancer causes symptoms and death. Men aged 45–50 who have PSA levels in the top 10% of values for their age group (>1.3 ng/mL), have a roughly 10 times greater risk of prostate cancer metastasis and death over the next 25 years compared to those with a PSA below the median (<0.6 ng/mL) (16). Hence, the clinical value of PSA is not so much a diagnostic marker for prostate cancer at biopsy, as a prognostic marker for the long-term risk of prostate cancer morbidity and mortality. Hence, diagnostic work-up for a man with elevated PSA should not primarily be based on risk of finding evidence of prostate cancer. This is particularly because autopsy studies show that presence of prostate cancer in the prostate is a near ubiquitous finding in older men (17). Instead, decision-making should be based on the risk of suffering symptoms of and dying from prostate cancer over the next decades.

Misconception 5: PSA Does Not Have High Sensitivity or Specificity for Prostate Cancer

It is routine to denigrate the PSA test as having poor operating characteristics. Researchers often conclude that “better markers are needed” (18); patient decision aids about PSA screening typically include language such as “PSA tests are unreliable and can suggest prostate cancer when no cancer exists … Furthermore, around 1 in 7 of those with normal PSA levels may have prostate cancer (a false-negative result), so many cases may be missed” (19). But prostate cancer is not in itself a problem as most men will have evidence of prostate cancer in their prostate if they live long enough. Accordingly, the important question is not the sensitivity and specificity of PSA for any prostate cancer at biopsy, irrespective of grade, but rather the association between PSA and long-term risk of metastasis and death. In that respect, PSA has excellent properties. For instance, a single PSA at age 60 has an area under the curve (AUC) of 0.90 for predicting cancer-specific death within 25 years (14); by way of contrast, one model for predicting a diagnosis of breast cancer within 5 years has an AUC of 0.58, which rises to 0.62 when adding a panel of single nucleotide polymorphisms (20). It is also true that PSA can be elevated for a host of reasons other than prostate cancer, after all, PSA stands for “prostate-specific” not “prostate cancer-specific” antigen. Most guidelines therefore recommend additional work-up, rather than performing a prostate biopsy, for men with higher PSAs. This work-up might include evaluation of benign disease and use of secondary tests with higher specificity, whether imaging or biomarkers (21).

Misconception 6: What Matters is Not so Much the Absolute Level of PSA, but Change in PSA

Cancer is a growth process and so we are often more interested in change than in the absolute size of tumor measures or markers, the concept of tumor progression being one example. This is exactly what we do in prostate cancer with an important caveat: PSA is only a tumor marker in men without a prostate. In a man with a prostate, the level of PSA in blood depends on benign prostate epithelium and stromal cells, as well as prostate cancer. In a man who has been treated with radical prostatectomy for prostate cancer, PSA in blood generally depends only on cancer burden. As a result, changes in PSA given by metrics such as PSA doubling time are routine for prognosis and response to treatment in the case of advanced prostate cancer, but not of value for diagnosis and prognosis of localized disease (22). For instance, there is no evidence to support the suggestion that men with low PSA should be subject to biopsy if they have a high PSA velocity (23).

Misconception 7: The Benefits of PSA Screening Are Controversial

PSA screening is often described as “controversial.” However, the reason why it is controversial is often misstated. Despite assertions to the contrary (19), this is no ambiguity as to whether PSA screening reduces prostate cancer-specific mortality. Clear evidence from randomized trials (24) is supported by the observation that prostate cancer mortality in the US has fallen by about half since the introduction of PSA, an effect that cannot be attributed to changes in treatment (25). The controversy about PSA screening is therefore not whether it has any benefit in terms of reduced cancer mortality, but about whether these benefits offset the harms of overdiagnosis and overtreatment. Our view is that the controversy needs to be reframed. PSA screening is not a unitary intervention, such as 325 mg of Aspirin or 20 mg of Tamoxifen; rather, it can vary in terms of the age range, criteria for biopsy, and eligibility for treatment. For many years, PSA screening involved a focus on older men, routine biopsy for men with elevated PSA, and aggressive treatment for almost all cancers found (26). This approach likely did more harm than good. Conversely, an approach that focuses PSA screening on younger men, involves biopsy only for those at high risk on a secondary test, and restricts treatment to high-grade tumors, would have a far better benefit-to-harm ratio (26).

Misconception 8: Population-based PSA Screening Would Dramatically Increase the Number of PSA Tests, Prostate Cancer Diagnoses, and Treatments

Many countries offer national programs of cancer screening, where colonoscopies or mammograms are provided by a government body with invitations sent to all eligible members of the population. National policy-making bodies have typically concluded that the evidence on PSA screening is insufficient to warrant such a program and recommend a shared decision-making approach instead. For instance, the recommendation of the Australian Health Ministers’ Advisory Council is “The PSA test is not suitable for population screening … We encourage men to speak to their doctor so they can make an informed choice about prostate cancer testing” (27); similarly, instead of a screening program, the UK has an “informed choice programme,” in which a patient can get a PSA if they are aged 50 or over and request a test after a discussion with a primary care physician (19). Such policies have paradoxically led to very high rates of PSA screening, particularly in older men, who are the least likely to benefit and most likely to be harmed by PSA testing (28). Implementation of a comprehensive, risk-adapted national program of PSA screening would reduce the number of PSA tests, and consequent overdiagnosis and overtreatment, in comparison to current trends (28).

Concluding Remarks

The development of the PSA test should be seen as one of the crowning achievements of clinical chemistry in oncology. Gross misuse of this test, and widespread misunderstanding of its properties, have severely tainted its reputation. Better use based on best evidence will ensure that PSA can reduce cancer-related suffering without causing undue harm.

Author Contributions

The corresponding author takes full responsibility that all authors on this publication have met the following required criteria of eligibility for authorship: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Nobody who qualifies for authorship has been omitted from the list.

Authors’ Disclosures or Potential Conflicts of Interest

Upon manuscript submission, all authors completed the author disclosure form.

Research Funding

This work was supported in part by grant funding from: The National Institutes of Health/National Cancer Institute (NIH/NCI) with a Cancer Center Support Grant to Memorial Sloan Kettering Cancer Center [P30 CA008748]; NIH/NCI with a SPORE grant in Prostate Cancer to Dr. H. Scher [P50 CA092629]; a Swedish Cancer Society project grant to H. Lilja [Cancerfonden 20 1354 PjF]. This work was also supported by philanthropic funding from The Sidney Kimmel Center for Prostate and Urologic Cancers and David H. Koch through the Prostate Cancer Foundation.

Disclosures

H. Lilja is named on a patent for intact PSA assays, is a member of the Scientific Advisory Board of Fujirebio Diagnostics Inc., owns stock in Diaprost AB and in Acousort AB. A.J. Vickers and H. Lilja are named on a patent for a statistical method to detect prostate cancer that has been licensed to and commercialized as the 4Kscore test by OPKO Health and receive royalties from sales of this test. A.J. Vickers has stock options and H. Lilja has stock in OPKO Health.

References