New circulating biomarkers for prostate cancer

Prostate Cancer and Prostatic Diseases (2008) 11, 112–120 & 2008 Nature Publishing Group All rights reserved 1365-7852/08 $30.00 www.nature.com/pcan REVIEW New circulating biomarkers for prostate cancer K Bensalah, Y Lotan, JA Karam and SF Shariat Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX, USA The introduction of prostate-specific antigen (PSA) revolutionized prostate cancer (PCa) screening and ushered the PSA era. However, its use as a screening tool remains controversial and changes in the epidemiology of PCa have strongly limited its prognostic role. Therefore, we need novel approaches to improve our ability to detect PCa and foretell the course of the disease. To improve the specificity of total PSA, several approaches based on PSA derivatives have been investigated such as age-specific values, PSA density (PSAD), PSAD of the transition zone, PSA velocity and assessment of various isoforms of PSA. With recent advances in biotechnology such as highthroughput molecular analyses, many potential blood biomarkers have been identified and are currently under investigation. Given the plethora of candidate PCa biomarkers, we have chosen to discuss a select group of candidate blood-based biomarkers including human glandular kallikrein, early prostate cancer antigens, insulin-like growth factor-I (IGF-I) and its binding proteins (IGFBP2 and IGFBP-3), urokinase plasminogen activation system, transforming growth factor-b1, interleukin-6, chromogranin A, prostate secretory protein, prostate-specific membrane antigen, PCa-specific autoantibodies and a-methylacyl-CoA racemase. While these and other markers have shown promise in early phase studies, no single biomarker is likely to have the appropriate degree of certainty to dictate treatment decisions. Consequently, the future of cancer prognosis may rely on small panels of markers that can accurately predict PCa presence, stage, metastasis, and serve as prognosticators, targets and/or surrogate end points of disease progression and response to therapy. Prostate Cancer and Prostatic Diseases (2008) 11, 112–120; doi:10.1038/sj.pcan.4501026; published online 13 November 2007 Keywords: biomarker; prostate-specific antigen; prognosis; detection Introduction Prostate cancer (PCa) is the most frequently diagnosed cancer among men in the United States, with a lifetime prevalence of one in six men.1 PCa displays a range of clinical behavior, from a slow-growing tumor of no clinical significance to aggressively metastatic and lethal disease. The introduction of prostate-specific antigen (PSA) revolutionized PCa screening and ushered the PSA era. This has resulted in earlier PCa detection and an increase in incidence. However, its use as a screening tool remains controversial due to questions regarding survival benefit, cost effectiveness and clinical factors such as the optimal age and total PSA at which to recommend biopsy.2 Recently, Thompson et al.3 showed that there is no true PSA cutoff point for identifying PCa risk in that there are a significant number of men with PSA values o4.0 ng ml 1 who actually have PCa. In addition to the controversy surrounding PSA-based screening, there is considerable debate whether PSA Correspondence: Dr SF Shariat, Department of Urology, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9110, USA. E-mail: Received 30 July 2007; revised 1 October 2007; accepted 16 October 2007; published online 13 November 2007 remains an important prognostic marker among men with PCa.4 The first and most important concern regarding the use of PSA is the lack of cancer specificity. A rise of the PSA level can reflect the presence of cancerous cells but can also be related to nonmalignant disorders such as benign prostatic hyperplasia (BPH), infection or chronic inflammation. All types of prostatic cells, whether normal, hyperplastic or cancerous, synthesize PSA, with highest levels found in the transitional zone of the prostate. Neoplastic cells produce lower levels of PSA compared to BPH cells but deliver a greater amount of PSA into the blood stream, presumably because of the disordered architecture of PCa. Therefore, it has been suggested that total PSA should be considered as a significant marker of BPH-related prostate volume, growth and outcome rather than a reliable marker of PCa.5–7 Changes in the epidemiology of PCa have strongly limited the correlation between total PSA and the stage of PCa.8–10 PSA can no longer be considered as a classical tumor marker whose levels are directly correlated with increasing stage of the disease. Moreover, the relationship with tumor grade remains unclear since it has been suggested that total PSA expression decreases with higher Gleason scores.11,12 Classical prognostic tools based on pretreatment PSA levels, such as Partin tables13 and Kattan nomograms14–16 have become less reliable Novel blood biomarkers for prostate cancer K Bensalah et al because of the continuous shift toward earlier stages of the disease at diagnosis. Consequently, the upper limit of normality of 4.0 ng ml 1, that was set in initial studies, may no longer be optimal since it has been proven that a significant number of cancers remained undetected in patients with PSA levels below this cutoff point.3,17 Initial reports suggested that men with a slightly ‘abnormal’ value (4.0–9.9 ng ml 1) had a 22% chance of having PCa, and those with a significant rise of 410.0 ng ml 1 had a 67% risk.18 However, it became quickly clear that these relative risks were in context of a sextant biopsy technique fraught with a high falsenegative detection rate, and that the risk of harboring PCa in men with a PSA level above 4.0 was approximately twice as likely, in the 40–50% range.19 Multiple studies have now demonstrated that significant numbers of men with ‘normal’ PSA values harbor PCa. The most definitive was the Prostate Cancer Prevention Trial, which determined that there is no PSA level below which PCa can be ruled out, and no cutoff above which PCa can be assured.20 As a result, it is now clear that men with PSA values below the artificial 4.0 cutoff are inaccurately categorized as being normal, and those above the cutoff are mischaracterized as being abnormal. Moreover, PSA levels in men that have undergone prior treatment for PCa are completely independent of the reference ranges in widespread laboratory use, making such references and thresholds even more meaningless in this setting. Therefore, clinicians and researchers are still on a quest for novel approaches to improve our ability to detect PCa and foretell the course of the disease. New therapeutics, such as chemoprevention, gene therapy and adjuvant therapies, will need a more reliable set of markers for their development. Because the most useful clinical biomarkers will likely be those that can be assayed from blood, there is much interest in proteomics, which ha (...truncated)