Prostate Cancer

Prostate cancer

Prostate cancer is one of the most frequently diagnosed cancers in men worldwide (Torre et al., 2015). Johannsen et al. (2005) conducted a pilot study to evaluate whether a new technique of malignant fluid hyperthermia can be used for minimally invasive treatment of prostate cancer. This study presents the first clinical application of interstitial hyperthermia using magnetic nanoparticles in locally recurrent prostate cancer. A special edition of the International Journal of Hyperthermia (Hurwitz, 2010) includes seven outstanding reviews as collective efforts of biologists, physicists, engineers, and clinicians committed to advancing thermal medicine using various hyperthermia techniques in regard to prostate cancer and thermal medicine. From a clinical standpoint, prostate anatomy allows for several heating approaches, including noninvasive, transrectal, transurethral, and transperineal techniques.

B. Pathology and Diagnosis of Prostate Cancer

Almost all prostate tumors are classified as adenocarcinomas (i.e., they arise from the glandular epithelial cells), and occur most commonly in the peripheral zone of the gland. Accordingly, they can often be felt by the physician during digital rectal examination. A unique feature of human prostate cancer is the high frequency of small, latent tumors in older men. A clear relationship between these occult tumors and those which become clinically apparent has not been established, although it is commonly assumed that the latter evolve from the former as a consequence of additional genetic mutations.

Generally, prostate cancer in its early stages is asymptomatic. Enlargement of the prostate gland, (benign prostatic hyperplasia or BPH) commonly begins after the age of 45, ultimately leading to urinary tract symptoms (difficult and frequent urination). Many cases of prostate cancer are diagnosed as a result of digital rectal examination performed when a man visits his physician for relief of these symptoms. (Suspicious lesions on examination may be confirmed by transrectal ultrasound, followed by a biopsy of the gland.) In recent years, the prostate-specific antigen (PSA) test has come into widespread use. This test is not specific for prostate cancer, however, and gives an abnormal result if there is any increased tissue growth in the gland, such as occurs in BPH. Because of its sensitivity, the PSA test can lead to the diagnosis of very early, microscopic tumors. Although such lesions might never progress to clinical disease, surgical removal carries a risk of major complications (notably incontinence and/or impotence), leading to controversy regarding the proper use of PSA as a screening test for early prostate cancer [3].

a. Obesity.

Prostate cancer is sometimes considered a male counterpart to breast cancer in women, for which there is clear evidence of a positive association with obesity, especially in postmenopausal cases. However, evidence for a similar association of adult obesity with prostate cancer is limited. Although a few epidemiological studies reported a significant positive association [28, 49, 93], most studies found no clear relationship of measures of obesity to prostate cancer risk [35, 37, 51, 59, 60, 65–67, 94–96]. Even when limited to prospective cohort investigations, the findings have been inconsistent, and some studies even showed inverse associations [e.g., 97–100]. The influence of obesity at younger ages is also unclear. One study suggested that childhood obesity may be protective against adult prostate cancer [99], but another study found that obesity at age 20 years was associated with an increased risk [97].

The basis for an association between obesity and prostate cancer could involve endocrine factors, because adult obesity in men has been associated with decreased circulating levels of testosterone and increased levels of estrogen [101]. This mechanism would suggest an inverse rather than a direct association between obesity and this cancer.

Prostate cancer

Cancer of the prostate is the second most common cancer diagnosed worldwide (after breast cancer), with 889,102 cases diagnosed in 2008 (Chisholm, McCabe, Wootten, & Abbott, 2012). Since the survival rate of prostate cancer is high, quality of life issues have become very important (Ferlay et al., 2010). Men diagnosed with prostate cancer frequently experience erectile dysfunction, with rates ranging from 60%–85% in men treated for localized prostate cancer (Penson et al., 2005; Potosky et al., 2004). That poses challenges to the man’s social, emotional, and physical well-being (Manne, Badr, Zaider, Nelson, & Kissane, 2010; Weber & Sherwill-Navarro, 2005), and may remain a problem up to 5 years posttreatment (Penson et al., 2005).

Background and the Benchmark Data

According to a 2013 report from the American Cancer Society, prostate cancer is the most common type of cancer in their top 10 list of cancers, with more than 238,000 new cases expected in the United States in 2013. The next most common cancers are breast cancer and lung cancer (www.cancer.org/research/cancerfactsfigures/cancerfactsfigures/cancer-facts-figures-2013, www.cancer.gov/cancertopics/types/prostate).

In this study, we will investigate a set of prostate cancer data to see whether statistical methods and machine-learning tools can help identify the genes that are related to this specific disease. The data comprises 102 patients (52 cancer, 50 normal) and 6,033 genes. The original data were collected and analyzed by a team of 15 scientists from a dozen institutions, including Harvard Medical School, Whitehead Institute/Massachusetts Institute of Technology, and Bristol-Myers Squibb Inc., Princeton (Singh et al., 2002).

Efron and colleagues (Efron and Zhang, 2011; Efron, 2010, 2011) also discussed this set of data in the context of Benjamini-Hochberg FDR (false discovery rate) and Bayesian analysis. We are very grateful that Dr. Efron emailed us the data he used in his papers. The data are in the Dap structure (http://fossies.org/dox/dap-3.8/classes.html) with a size of 11.5 MB. A glimpse of the data follows (here, we show only the first and the latter three lines of the file):

In order to facilitate the analysis that will be carried out by SAS, STATISTICA, R, and other software packages, we use the following SAS code to convert the Dap data:

To run the SAS code as is, the user needs to create a new folder called “Prostate_Cancer_data” in the C:drive, deposit the raw data in C:\Prostate_Cancer_data, and then run the above code in SAS Editor window. The output data file will be in the folder “C:\Prostate_Cancer_data,” and will be called “dapout.sas7bdat.”

After the conversion, there will be 102 rows, representing n=102 patients in the study. The first column of the data will be “_target” (1=cancer, 0=normal patient), the second column will be “PatientID” and the remaining columns will be gene1–gene6033. One goal of the study is to find the genes that are related to the disease.

Due to the amount of the data and the intricate process of data conversion, it will take a few minutes to complete the run, so be patient.

Prostate cancer.

Prostate cancer is the most common nonskin cancer in males. The greatest risk factor is age, with more than 75 percent of new diagnoses occurring in men over 65. The autopsy prevalence of microscopic or occult prostate cancers is rare in men who were in their 40s, but approximately 40 to 50% of men 60 to 70 years of age have occult prostate cancer and 80% by age 80 (Sakr et al., 1994). The prevalence of prostate cancer has been reported for the 2,950 men (age range, 62–91 years) who were assigned to the placebo group in the Prostate Cancer Prevention Trial. Men who never had a PSA level of > 4.0 ng/mL or an abnormal digital rectal examination after being in the study for seven years underwent a prostate biopsy (Thompson et al., 2004). Prostate cancer was diagnosed in 15.2% of the men, and 14.9% had a Gleason score of 7 or higher. The prevalence of prostate cancer for men with PSA levels <0.5 ng/mL was 6.6%, 10.1% for values of 0.6 to 1.0 ng/ml; 17% for values of 1.1 to 2.0 ng/mL; 23.9% for values 2.1 to 3.0 ng/mL; and 26.9% for values of 3.1 to 4.0 ng/mL. High-grade cancers (7 or greater) increased from 12.5% with the lowest PSA levels to 25% in the group with PSA of 3.1 to 4.0 ng/mL. Most of the occult cancers never become clinical cancers. However, it is not known whether TRT will increase this risk.

Two prospective cohort studies and 10 nested case-control studies have correlated T levels with future development of prostate cancer (Rhoden et al., 2004). None of these studies found a positive correlation between total T or bioavailable T (four studies) levels and future prostate cancer; however, the nested case-control study reported by Gann and colleagues found a positive relationship after T was adjusted for the SHBG level. Although it is reasonable to adjust for SHBG, this study found that high total T levels and low SHBG levels were correlated with future prostate cancer. Total T levels usually increase with increasing SHBG. As suggested by Hsing, it is desirable to correlate prostate cancer with measures of androgen action, rather than a single blood level at some point in the past, but such studies are lacking.

The influence of T on prostate carcinogenesis and other prostate outcomes remains poorly defined. Some animal studies suggest that T may be a weak carcinogen in susceptible animals. Most studies indicate that T can act as tumor promoter at normal physiologic levels (Bosland, 2000; Stanbrough et al., 2001). The direct relevance of these studies to humans is uncertain. Despite the lack of evidence implicating androgens in carcinogenesis, it is clear that prostate cancer rarely, if ever, develops in an environment devoid of androgens, and the majority of prostate carcinomas require androgens for growth. Androgen ablation causes regression of metastatic prostate carcinoma; however, cancer cells that survive androgen deprivation ultimately proliferate, causing relapse and androgen-independent disease.

A. Incidence and Mortality Trends

Prostate cancer is a common cancer among men in many Western countries, and is the leading male incident cancer in the United States, where 180,400 new cases are projected for the year 2000 [4, 5]. Incidence trends in the United States show a rather slow increase over most of the last 50 years, with a rather striking increase between 1989 and 1992, attributable in large measure to the widespread adoption of the PSA screening test, which first became available in the early 1980s [6]. Since 1992, the incidence has declined, reflecting an end to the surge in cases due to the introduction of this new screening procedure [7], as well as, perhaps, to the more judicious application of PSA screening. Moreover, mortality from prostate cancer is low relative to its incidence. This is because prostate cancer is generally well controlled by treatment (surgery, radiation, and androgen ablation) and occurs at relatively late ages, so that even men who are not cured of the disease often die from other causes. Interestingly, a parallel increase in prostate cancer mortality did not occur during the period 1989–1992, presumably because most of the additional cases diagnosed would not otherwise have led to fatal outcomes.

A Incidence and Mortality Trends

Prostate cancer is a common cancer among men in many Western countries, and it is the leading male incident cancer in the United States, where 180,890 new cases are projected for the year 2016 [8–10]. Incidence trends in the United States show a rather slow increase over most of the past 50 years, with a striking increase between 1989 and 1992, attributable in large measure to the widespread adoption of the PSA screening test, which first became available in the early 1980s [11,12]. After 1992, the incidence declined until approximately 1995, remained stable until 2000, and then declined again after 2000 [13]. Specifically, from 2003 to 2012, rates decreased by 4% per year [10]. Moreover, mortality from prostate cancer is low relative to its incidence. This is because prostate cancer is generally well controlled by treatment (surgery, radiation, and androgen ablation) and occurs at relatively late ages so that even men who are not cured of the disease often die from other causes. Interestingly, a parallel increase in prostate cancer mortality did not occur during the period 1989–92, presumably because most of the additional cases diagnosed would not otherwise have led to fatal outcomes. Prostate cancer mortality rates decreased by 3.5% per year from 2003 to 2012, attributed to improvements in early detection and treatment [10].

Introduction

The prostate functions as a reproductive exocrine organ and functions to prevent urinary tract infections. However, the prostate is most noted for the frequency with which it is the origin of benign and malignant neoplasms and infectious diseases in aging men. The most important of these diseases are prostatitis, benign prostatic hyperplasia, and prostate cancer. This article will review the prostate in health and disease with an emphasis, where possible, on the gerontological aspects of prostate disease. (See Renal and Urinary Tract Function.)

B Pathology and Diagnosis of Prostate Cancer

Almost all prostate tumors are classified as adenocarcinomas (i.e., they arise from the glandular epithelial cells) and occur most commonly in the peripheral zone of the gland. Accordingly, they can often be felt by the physician during digital rectal examination. A unique feature of human prostate cancer is the high frequency of small, latent tumors in older men. A clear relationship between these occult tumors and those that become clinically apparent has not been established, although it is commonly assumed that the latter evolve from the former as a consequence of additional genetic mutations.

Generally, prostate cancer in its early stages is asymptomatic. Enlargement of the prostate gland (BPH) commonly begins after the age of 45 years, ultimately leading to urinary tract symptoms (difficult and frequent urination). Many cases of prostate cancer are diagnosed as a result of digital rectal examination performed when a man visits his physician for relief of these symptoms. (Suspicious lesions on examination may be confirmed by transrectal ultrasound, followed by a biopsy of the gland.) Since its approval by the U.S. Food and Drug Administration in 1986, the prostate-specific antigen (PSA) test has come into widespread use. This test is not specific for prostate cancer, however, and gives an abnormal result if there is any increased tissue growth in the gland, such as occurs in BPH. Because of its sensitivity, the PSA test can lead to the diagnosis of very early, microscopic tumors. Although such lesions might never progress to clinical disease, surgical removal carries a risk of major complications (notably incontinence and/or impotence), leading to controversy regarding the proper use of PSA as a screening test for early prostate cancer [3,4]. Indeed, in its review in 2011, the U.S. Preventive Services Task Force (USPSTF) concluded that use of PSA as a screening modality has not resulted in any decrease in prostate cancer mortality and that, to the contrary, its use has led to more harm than good [5]. Thus, in 2012, the USPSTF published a statement recommending against PSA-based screening for prostate cancer [6]. Based on national surveys in the United States, the proportion of men aged 50 years and older reporting PSA screening in the past year was 36.9% in 2005, 40.6% in 2008, and 37.8% in 2010, but after the 2012 USPSTF recommendation, declined to 30.8% in 2013 [7].