In most developed countries, the most common mode of presentation of prostate cancer is that of an asymptomatic man with a raised serum prostate-specific antigen (PSA), detected at opportunistic testing. There has been considerable debate about the potential merits of a PSA-based screening programme, but there is little evidence at present to suggest that screening meets the World Health Organization (WHO) criteria for screening. Two large, multicentre trials of prostate cancer screening, based on serum PSA assay and digital rectal examination, and involving more than 100 000 men, are in progress; results are not expected until 2008. Asymptomatic men may also present as a result of diligent use of the digital rectal examination in the primary or secondary care setting; this is the primary means of diagnosing the large number of cancers in men whose serum PSA lies within the age-specific ‘normal’ range.

Symptomatic men may present with symptoms from either the primary prostate cancer or from metastases. Local symptoms include those of bladder outlet and ureteric obstruction, haemospermia, impotence and reduced ejaculatory volumes. Investigation of lower urinary tract symptoms and erectile dysfunction is discussed elsewhere in this book and will not be discussed further here; it is important, however, that prostate malignancy is not forgotten as a potential cause.

PSA is a 28.5 kDa glycoprotein containing 237 amino acids; the mature protein is cleaved from a 261 amino acid precursor molecule. It is a serine protease belonging to the kallikrein protease family and is also classified as human kallikrein 3 (hK3). Transcription of the PSA gene is under the control of an androgen-responsive element (ARE) within its promoter; expression therefore increases when prostate cells are treated with androgens or become more sensitive to androgens. Highlevel expression of PSA is exclusive to the prostate epithelium; lower levels of expression are seen in the placenta, some breast carcinomas and the male peri-anal glands.

Healthy prostate glands sequester PSA in an inactive state within the glandular lumina, where it may be activated on cleavage by human kallikrein 2 (hK2). PSA is thought to be responsible for liquefying semen; it is present in the semen at high concentration (0.5-2 g/l – about 1 000 000-fold the normal serum concentration), with the ability to digest seminogelins, the proteins which stabilize the seminal clot. PSA in the seminal fluid is predominantly ‘free PSA’, although a small proportion is in an inactive complex with protein C inhibitor.

Both normal and malignant prostate tissue secretes a large amount of PSA, although the rate of production of ‘normal’ (i.e. assay-detectable) PSA can decline as a result of cellular de-differentiation in advanced malignancy. Development of a focus of prostate cancer leads to increased production of PSA by individual cells as well as disruption of the normal glandular architecture with consequent increased leakage of PSA into the circulation. Serum PSA can therefore be used as a prostate cancer tumour marker. It is neither a particularly sensitive, nor specific, tumour marker because benign diseases of the prostate (for example, bacterial prostatitis or benign hyperplasia) may also result in elevated serum PSA.

Within the circulation, PSA is almost completely complexed to protease inhibitors. The predominant complex-forming protein is α₁-antichymotrypsin (ACT), with smaller amounts being bound to α₁-protease inhibitor and α₂-macroglobulin (A2M). Between 5% and 40% of serum PSA is free and it is believed that this represents an enzymatically inactive form, either pro-PSA or nicked PSA. Complex formation with ACT dramatically increases the half-life of PSA (from a couple of hours for the free form to 2 or 3 days for the PSA–ACT complex). Conversely the PSA–A2M complex is metabolized within minutes; this complex probably represents the major route of metabolism for serum PSA. PSA does not appear to have any enzymatic function within the circulation.

Serum immunoassay for PSA has become a standard diagnostic test for prostate cancer. Because of the confounding effect of PSA produced from benign prostatic tissue, a ‘cut-off’ level of serum PSA has to be used, accepting that this will result in a number of false-positive and false-negative results. There has been considerable debate about the appropriate cut-off. Initially a cut-off of 4 ng/ml was commonly used as the trigger for further investigation; this has been superseded by age-specific cut-offs due to a significant false-negative rate using the higher level. Some studies have suggested that as many as 25% of men with prostate cancer have a serum PSA measuring less than 4 ng/ml. Reducing this cut-off to 2.5 ng/ ml improves the sensitivity of testing, but with an inevitable reduction in specificity and a consequent increase in the number of negative prostatic biopsies. Many groups have attempted to improve the performance of PSA by the use of modifications such as prostatic PSA density, transition zone PSA density, PSA velocity and ratio of free:total PSA.

Serum PSA velocity is perhaps the current favourite method of improving the diagnostic performance of PSA. Data from the USA, where a large proportion of men undergo yearly measurement of PSA, suggests that the rate of change of PSA is a more accurate predictor of a positive prostate biopsy than absolute level of PSA. This seems particularly evident in the detection of high-grade adenocarcinomas – prostate cancers that are most threatening to the patient.

PSA is also central to staging and prognostication in men diagnosed with prostate cancer. Using validated tables and nomograms, the combination of PSA, Gleason grade and clinical stage enables reliable estimation of the likelihood of capsular penetration, involvement of pelvic lymph nodes and bone metastases. These probabilities are used, with an assessment of comorbidity and life expectancy, to select the appropriate management option for each patient. Many clinicians will use these variables to determine the need for staging using imaging techniques such as isotope bone scanning or pelvic magnetic resonance imaging (MRI).

Once treatment or active surveillance has commenced, PSA is a central element of patient monitoring and determines most of the changes in management. Several studies have shown that changes in serum PSA pre-date clinical events by months or years. Men who have an undetectable (after surgery) or stable (after radiotherapy) serum PSA are frequently cured. Small increases in PSA levels can give a very early indication of failure of radical treatment, allowing the opportunity for salvage treatment before the development of macroscopic disease or symptoms.

The ‘gold standard’ for diagnosis of prostate cancer is histological assessment of biopsied tissue. This is most commonly in the form of needle-core biopsies obtained using a transrectal ultrasound scanning (TRUSS) guide according to a standard biopsy protocol; frequently, however, prostate cancer is an incidental diagnosis in chippings from a transurethral resection of the prostate (TURP), in a specimen from a radical cystoprostatectomy or in percutaneous, laparoscopic or open biopsies from distant lymph nodes or organs.

In a standard TRUSS-guided prostate biopsy, at least 10 areas are routinely sampled, along with any areas which are found to be hypo-echoic, or otherwise abnormal on TRUSS. A typical ultrasound machine used for TRUSS-guided biopsy is shown in 5.1; a typical ‘rapid-fire’ biopsy gun is seen in 5.2. As well as its use in the initial diagnosis of prostate cancer, histological analysis of TRUSS-guided biopsy specimens can be used to monitor grade progression in men being managed by active surveillance and to confirm local recurrence of prostate cancer after radical radiotherapy.