Prostate Cancer Proteomics

Introduction

Prostate cancer is the most frequently diagnosed tumor in North American and European men. During their lifetime, about 1 in 11 men will be told that they have prostate cancer and 3% will die as a direct result of the disease. Of men given a diagnosis of cancer, nearly 25% will be told they have prostate cancer. About 300,000 new cases are diagnosed in the EU every year.

Proteomics offers the hope of biomarkers that can be used to accurately

Screen and diagnose the disease

Biomarker Needs in the Clinic

The men diagnosed with prostate cancer are the tip of the iceberg, as men almost inevitably develop the disease if they live long enough (Franks, 1954). At age 70, about 40% of men dying of other causes have histopathological evidence of prostate cancer at autopsy, and by age 100 this figure rises toward 100%. Consequently, it has been calculated that there are about 8 million North American men living with prostate cancer, the majority being unaware of the condition. On this basis, most men with prostate cancer do not need treatment and may be better off without the diagnosis.

Serum levels of prostate specific antigen (PSA) are used to screen for prostate cancer. However, the serum test has poor sensitivity and specificity, and most of the cancers detected do not need to be treated. For every 100 men with prostate cancer detected by screening and treated by radical therapy, about 2 lives are saved (Schroder et al., 2009). Although these figures are excellent for the 2%, the other 98% will suffer varying degrees of impotence and incontinence, and it is a matter of debate whether the harm outweighs the benefit, particularly as a similar study showed no survival benefit (Andriole et al., 2009).

It has been calculated that if 1 million asymptomatic men agreed to have a PSA test, 90,000 will then undergo needle biopsy of the prostate and 20,000 of these will be diagnosed to have prostate cancer. A total of 10,000 of these men might choose to have a radical prostectomy, which will kill about 10. A total of 300 of these men will be left with severe incontinence and 4,000 will be impotent (Frankel et al., 2003).

Screening has the potential to save lives by detecting prostate cancer before it has spread. However, PSA screening detects cancers that will not affect survival and detects cancers early so men live longer with the disease. The ideal screening test would have low rates of false positives and negatives and result in a demonstrable improvement in survival.

Most cancers present in the PSA range of 4–10 ng/mL of serum, but about 70% of men with PSA levels in this range have benign prostatic hyperplasia, a condition that is not life threatening, but causes local symptoms that are similar to those of prostate cancer. The upper limit of normal value for PSA of 4 ng/mL of serum misses many men with prostate cancer. Over 5% of men with PSA levels of less than 0.5 ng/mL had cancer on biopsy, rising to approximately 30% of men having serum levels of 3–4 ng/mL (Thompson et al., 2004). PSA screening is more prevalent in the United States, and consequently, the proportion of men diagnosed is approximately double that in Europe, and yet survival rates are similar in the United States and Europe. Proteomics has the potential to identify a better marker than PSA and perhaps a cancer-specific marker that could differentiate BPH from cancer. Proteomics may also identify markers that, unlike PSA, differentiate the cancers that will reduce survival.

Prostate cancer that on clinical grounds appears to be restricted to the prostate can be treated with radical radiotherapy or prostatectomy. There is only one randomized study that has compared surveillance with radical prostatectomy. A total of 695 men with localized prostate cancer were randomly assigned to watchful waiting (348 men) or radical prostatectomy (347 men). During a median of 10.8 years follow-up, 68/348 (19.5%) of the men in the watchful waiting group died of prostate cancer compared to 47/347 (13.5%) of those in the surgical group. The difference was 5.4% (95% confidence interval 0.2–11.1%) with little or no further increase in benefit 10 years or more after surgery (Bill-Axelson et al., 2008). Again, it is clear that the benefit of active management of the disease is small and that over 90% of the men treated experience the side effects without any gain in survival. Identification of men with localized disease who will benefit from radical therapy would be a major step forward in the management of prostate cancer.

When prostate cancer spreads outside the primary organ, the staple treatment is hormone therapy—medical or surgical castration. Hormone therapy is effective in the majority of men, but most men relapse with castrate-resistant disease within 2 years. Hormone therapy was introduced by Charles Huggins in the 1940s, and there has been no major change in therapy since. Consequently, survival figures for men with metastatic prostate cancer have altered little for over 50 years. Proteomics might identify new targets for the treatment of metastatic disease.

Sources of Material for Proteomic Studies of Prostate Cancer

The prostate is situated at the base of the bladder surrounding the urethra. As a consequence of this anatomical location there are many potential sources of proteins derived from prostate cancer, including urine, expressed prostatic fluid, ejaculate, serum, or plasma and the tissue itself. Each of these sources has advantages and limitations as a potential source of biomarker.

Collection of urine is the noninvasive method of obtaining samples of prostate proteins (Jamaspishvili et al., 2010). But urine is particularly difficult to standardize and protein preservation in urine is problematic. Twenty-four-hour collections may be regarded as most representative, but many proteins are degraded over this period either as a result of oxidation, changes in pH, or bacteriological breakdown. Another sample of urine frequently used for analytical assessment is an early morning midstream urine, which under certain circumstances can be frozen rapidly to preserve the protein content for future analysis. Urine collection following prostatic massage may increase the yield of prostate-derived proteins. Urine contains over 1,500 proteins (Adachi et al., 2006).

Expressed prostatic fluid can be obtained from most men by a massaging the prostate with a finger inserted up the rectum. The yield of fluid varies widely depending on the patient, the experience of the operator, and his or her technique. Ejaculate manually expressed by the patient should be fairly standard. However, the majority of men with prostate cancer are over 65 years of age. The more elderly the man, the less willing or able he may be to provide an ejaculate. Furthermore, the ejaculate and the expressed prostatic fluids are complex mixtures of proteins. Seminal fluid contains over 900 proteins (Pilch and Mann, 2006).

Serum or plasma are the usual source of biomarkers, and there are a number of examples of clinically useful markers. PSA, despite its limitations for screening and diagnosis, is a clinically useful and standard marker of relapse following a diagnosis of prostate cancer. Rising serum PSA levels following radical treatment of early disease are a reliable marker of local recurrence or the growth of metastatic disease.

Tissue is a challenging source of biomarker because histopathological confirmation is needed for the presence of cancer, and the relative contributions of cancer, normal epithelium, nerve, blood vessel, muscle, and connective tissue will differ in every biopsy, even from the same patient.

In addition to these direct sources of prostate-derived proteins, many proteomic studies have used cancer cell lines. This material is at least one step removed from any tissue or fluid directly obtained from a patient, but does provide a relatively homogeneous tissue source. Nevertheless, any data from human cancer cell lines must be treated with great caution unless there is unequivocal evidence of the origin of the cell line. There is no evidence that the two most frequently used “prostate” cancer cell lines, DU145 and PC3, are derived from prostate cancer, and despite their putative origin from castrate-resistant disease, these cell lines do not express characteristics of this stage of prostate cancer such as PSA and androgen receptor expression. Many proteomic studies have been misled as to the origin of human cancer cell lines. For example, there are at least three articles using Chang liver cells to identify liver-specific proteins, but the authors and the editor of Proteomics seem unaware that Chang liver cells are, in fact, the cervical cancer cell line HeLa (Sui et al., 2008; Tan et al., 2008; Xu et al., 2007).

Proteomic Platforms

Global isolation, detection, and quantitation of proteins are more difficult to achieve than such procedures for DNA and RNA. Simple, relatively cheap and quantitative methods for global analysis of proteins are not yet available. The complexities of posttranslational modification and its effects on function and metabolism need to be taken into account. However, once clinically relevant markers are available, clinical biochemistry has a wealth of tools to accurately measure individual proteins in almost any fluid. But the relevant markers have yet to be discovered, at least using proteomic methods.

Classical proteomics uses 2D gels with a variety of analytical methods to detect peptides. These approaches have become increasingly sophisticated and amenable to quantitation, but have yet to yield any clinically relevant biomarkers for prostate cancer. The technical skill and experience required for the classical methods contrast with the ease of use of surface enhanced laser desorption ionization-time of flight (SELDI-TOF). Using this technique, biological fluids are applied to various surfaces, dried and placed in a machine that provides a series of peaks that can be compared between samples. Supposedly reproducible and clinically relevant findings were claimed to be produced, but were eventually shown to be neither reproducible nor clinically relevant. Reproducibility was poor in the absence of stringent control of sample collection, separation, and storage. Currently there is little interest in pursuing this approach (Goo and Goodlett, 2010).

A swathe of new techniques is being used to identify proteins, protein interactions, protein metabolism, and posttranslational modification using a variety of platforms, both new and old. However, for every article demonstrating a new and clinically relevant biomarker such as PSA, CA-125, or β-HCG, there are probably at least 10,000 articles containing negative results or positive results that are either not reproducible or of no clinical value.

Prostate Cancer Protoemics—The Future

There is flurry of activity related to the detection of sarcosine in urine for prostate cancer diagnosis. Using various types of sample, Sreekumar et al. (2009) demonstrated that sarcosine levels increased greatly during prostate cancer progression to metastasis and that sarcosine can be detected and measured in urine samples. As is typical in the biomarker field, the first independent study attempting to replicate this finding failed, concluding that sarcosine levels in urine are not diagnostic or associated with stage or grade of prostate cancer (Jentzmik et al., 2010). As is also typical in the biomarker field, the two studies are not directly comparable due to differences in a number of variables such as the assay methodology and the cohort of patients, and therefore further studies will be needed to resolve the contradictions. In the biomarker field, it seems that hype, hope, and optimism are essential prerequisites, but the grim truth is almost but not total universal disappointment when attempts are made to replicate positive results with claims of clinical benefit.

Eventually further markers for prostate cancer will be identified. They will be small in number and help to answer specific clinical questions relating to the management of the patients. These markers will be analyzed individually or in small groups using various methodologies developed by clinical biochemists and will contribute to a higher standard of care for men with prostate cancer.