Viva Practice for the FRCS(Urol) and Postgraduate Urology Examinations, 2nd ed.

Chapter 1 Prostate Cancer

Hashim U Ahmed, Jane Boddy, Alan Doherty, Lyndon Gommersall and Manit Arya

Prostate-Specific Antigen (Psa) And Prostate Cancer Detection

Q. A 66-year-old man with a PSA of 5.3 is referred to you. Testing was performed during a routine health check with his GP. He has mild lower urinary tract symptoms but is otherwise fit and well. Examination revealed a small benign feeling prostate. How would you assess this man?

A. A full urological history should be taken with an emphasis on lower urinary tract symptoms (LUTS), age, race, family history and history of urinary infections. If advanced disease is suspected then a history of bone pain, leg swelling, anorexia, weight loss, coagulopathy and new onset peripheral neurology is important. A general urology examination should be performed along with a digital rectal examination. If there is high-quality multiparametric MRI (mpMRI) available in my unit, then I would offer an mpMRI prior to biopsy. I would then offer a transrectal ultrasound (TRUS) guided systematic prostate biopsy with targeting of any suspicious areas, quoting a 1%-2% risk of severe sepsis. Almost all patients experience some bleeding (rectal bleeding/haematuria/haematospermia) following this procedure which then resolves. Treatment is dependent on the grade and stage of the tumour as well as any co-morbidity.

Q. What is PSA?

A. PSA is a 34 kD serine protease. This glycoprotein was first discovered in 1970. Also known as human kallikrein 3 (HK3) it has 261 amino acids. HK2 and HK1 also exist. It is encoded by a gene on chromosome 19. It is secreted uniquely by prostatic ductal epithelial cells and its biological effect is to liquefy the seminal coagulum within the ejaculate. It is synthesised as pre-pro-PSA which is converted to pro-PSA and then PSA. PSA exists in three forms in serum:

1. Free - Unbound with a half-life of between 2 and 3 hours.

2. Bound to alpha-1 antichymotripsin (ACT - a serine protease inhibitor or serpin), this form has some epitopes exposed which affects free to total PSA measurement. It has a half-life of 4-5 days.

3. Bound to alpha-2 macroglobulin (AMG - all five PSA epitopes are covered making this more difficult to quantitate) with a half-life of 4-5 days.

Overall PSA has a half-life of between 2 and 3 days. Total PSA There fore equates to:

Total PSA = Free PSA + Complexed PSA (i.e. ACT bound PSA but not including AMG bound PSA)

The measurement of PSA is now automated using a monoclonal antibody assay technique and commercially produced antibodies.

Q. What is a normal PSA?

A. There is no PSA cut-off that completely predicts the absence of prostate cancer. However in clinical practice a normal PSA can be defined as either being lower than an absolute figure (generally considered as 3 or 4 ng/mL) or as a range of values related to the patient’s age. Oesterling described the most commonly used age-specific reference ranges for PSA [1]. This is reproduced in Table 1.1. More recently Sun et al. have reported on 12,078 patients who have been retrospectively reviewed following prostate biopsy [2]. In this study receiver operative characteristics (ROC) analysis of the PSA results demonstrated a normal PSA cut-off as 2.3 ng/mL. An analysis of the Swedish Cancer Registry data showed that a very low PSA at the age of 50 years (usually less than 1.2 ng/mL) confers a very low (about 1%) probability of cancer-specific mortality over 15 years.

PSA is prostate specific but not prostate cancer specific. It is also elevated in benign prostatic hyperplasia, prostatitis, catheterisation and other non-malignant conditions.

Note: The threshold of 4 ng/mL was defined from the use of ROC analysis, a technique used in radar detection of enemy aircraft during the Second World War. From the 1970s this approach was applied to medical testing in general. For a range of values this technique delineated when a given signal is more likely to be enemy aircraft than radar artefact. When applied to PSA readings it showed that PSA can distinguish between benign disease or malignancy at a certain threshold.

The original clinical work on PSA was performed by Catalona et al. in the early 1990s [3]. This study investigated men invited to attend for a PSA blood test. They were stratified into three groups according to their PSA. The first group had a PSA between 0 and 4 ng/mL, the second between 4 and 10 and the third over 10. Each patient with a PSA greater than 4 ng/mL was offered a prostate biopsy. Prostate cancer was detected in 26% of the patients with a PSA between 4 and 10 ng/mL and in 53% of patients with a PSA of greater than 10 ng/mL. More recently the prostate cancer prevention trial (PCPT) gives further insight into the percentage of positive biopsies for a ‘normal’ PSA [4]. These data are reproduced in Table 1.2. This landmark study reveals the high proportion of patients with prostate cancer with low PSA readings. A criticism of the PCPT is the overdiagnosis of prostate cancer, i.e. a man’s lifetime risk of developing prostate cancer is approximately 16% but in the PCPT trial, prostate cancer was detected in 24.4% of patients analysed.

Table 1.1 Oesterling’s age-specific reference ranges

Source: Adapted from Oesterling JE et al. JAMA 1993; 270: 860-864.

Table 1.2 Prevalence of prostate cancer among men with a prostate-specific antigen level less than 4 ng/mL

Source: Adapted from Thompson IM et al. N Engl J Med 2004; 350: 2239-2246.

Gerstenbluth et al. reported on the positive predictive value of PSA. A PSA of greater than 20 ng/mL relates to an 87% chance of prostate cancer being detected on biopsy. Table 1.3 reports the positive predictive value for PSA ranges from this same paper [5].

Table 1.3 Positive predictive value for various PSA ranges

Source: Adapted from Gerstenbluth RE et al.J Urol 2002; 168: 1990-1993.

Q. What do you know about PSA velocity (PSAV)?

A. PSA is not a specific test. To improve the accuracy of detection of prostate cancer many adaptations of PSA have been described (Table 1.4). PSAV is defined as the rise in PSA per year in ng/mL/y. A rise in PSA of greater than 0.75 ng/mL/y is associated with an increased risk of prostate cancer [6]. Sun et al. use a PSAV cut-off of 0.6 ng/mL/y [2]. Ideally at least three PSAs are required to produce a PSAV reading due to the variability of the test. This is calculated with the following equation:

PSAV = 0.5x(PSA2 — PSAl/time 1in years + PSA3 — PSA2/time 2 in years)

Table 1.4 PSA derivatives and normal values

Q. What do you know about PSA doubling time (PSADT)?

A. PSADT is defined as the length of time that a patient’s PSA takes to double in months or years. PSADT is useful in patients under surveillance for high PSA readings and negative biopsies, active surveillance for low-risk disease or patients with a rising PSA following radical treatment. It is calculated using regression analysis of the PSA tests recorded. The Marsden experience with active surveillance uses a PSADT of less than 2 years. The data on active monitoring provide useful information on how to follow patients with a high PSA who have negative biopsies. Following radical prostatectomy biochemical progression can be due to local recurrence, lymph node positive disease or metastatic disease. Calculation of the PSA doubling time can elucidate between these two scenarios. In one study a PSADT <4.3 months suggested metastatic disease and a PSADT of >11.7 months suggested local recurrence [7]. This is important in deciding whether adjuvant local treatment is appropriate. Similar data exist for post radiotherapy biochemical progression.

Q. What do you know about PSA density (PSAD)?

A. PSAD is defined by the serum PSA per millilitre of prostate tissue. Prostate volume can be measured using the ellipsoid volume formula:

Prostate volume = Height x Width x Length x 0.52

A PSAD of >0.15 ng/mL/mL of prostate tissue is more likely to lead to a diagnosis of prostate cancer [8]. Other authors have not proven the utility of PSAD. This may relate to the inability to accurately measure prostate volume and the considerable variability of PSA production from benign prostate epithelium. PSAD when applied to equivocal mpMRI findings of the prostate was much more useful in reducing the false-negative rate of mpMRI.

The PSA transitional zone density (PSATZD) can also be quantified. This is defined as the amount of PSA per millilitre of transitional zone tissue. Djavan et al. described a normal value of 0.35 ng/mL and published a PPV of 74% for PSA levels less than 10 ng/mL [9]. This result has not been replicated.

Q. What do you know about free to total PSA (f/T PSA)?

A. f/T PSA is defined as the percentage free PSA compared to the total PSA in serum. PSA exists in serum as either free PSA or bound to ACT and AMG. Prostate cancer has a significantly lower f/T PSA value compared with BPH in the PSA range 4-10 ng/mL [10]. A cut-off of greater than 25% suggests benign disease. Despite several early meta-analyses which showed the clinical utility of f/T PSA had not been clearly defined, recent studies in developing nomograms have shown that it can be predictive.

Q. What do you know about supersensitive PSA (sPSA)?

A. sPSA enables PSA to be detected to a threshold of 0.003 ng/mL. Following radical prostatectomy 10%-40% of patients will develop biochemical relapse. An undetectable PSA is defined as <0.01 ng/mL. In a study of 200 patients following radical prostatectomy if the PSA nadir was <0.01 then biochemical progression occurred in only 3% of patients compared to 75% if this level was not attained [11]. This assay There fore allows the early detection of biochemical relapse after radical prostatectomy and can expedite the use of secondary interventions.

Q. Does digital rectal examination (DRE) change the PSA results?

A. Many factors alter the serum PSA level. In a series of 202 patients the DRE significantly changed the PSA by only 0.26 ng/mL. It was concluded that PSA elevation due to DRE was clinically insignificant [12].

Q. What is uPM3 or PCA3?

A. The uPM3 test detects prostate cancer antigen 3 (PCA3), a non-coding segment of mRNA.

It is a gene specifically produced by prostate epithelial cells 60-100 times more in prostate cancer than benign prostatic disease [13]. A uPM3 or PCA3 test is performed by collecting the first 20-30 mL of urine voided following vigorous prostatic massage. The ratio of PCA3 to PSA mRNA can then be calculated. The test claims to have a specificity of 70%. It is sometimes used in men with a negative TRUS biopsy prior to further biopsy in the event of persistently elevated serum PSA.

Q. Is screening for prostate cancer justified?

A. Prostate cancer screening has a considerable evidence base but there is controversy about offering it within healthcare systems. There fore it is not currently undertaken in the United Kingdom or Europe as a whole. In the United States, the U.S. Preventative Services Taskforce issued guidelines in 2012 advocating against wholesale PSA screening. Recently, in 2017, these guidelines were changed to advocate discussing PSA testing with men between the ages of 55 and 69 years with appropriate informed decision making with the patient. In the United Kingdom a policy of early detection through case finding is recommended via the prostate cancer risk management programme (www.cancerscreening.nhs.uk/prostate). This aims to ensure concerned individuals receive clear and balanced information about the advantages and disadvantages of the PSA blood test and treatment of prostate cancer before receiving the test.

Prostate cancer screening is controversial and does not fulfil all the Wilson and Junger’s (World Health Organisation [WHO]) criteria for a good screening programme (Table 1.5). Prostate cancer is an important health problem with a 16% lifetime risk of developing it and a 3% risk of dying from prostate cancer. In the United Kingdom in 2014 the incidence was over 46,000 cases (Office of National Statistics data). The disease has a long latent period often allowing the patient to die of other causes rather than prostate cancer. PSA is an acceptable screening test for patients but lacks sensitivity and specificity. Digital rectal examination, TRUS and prostate biopsy are invasive which would be required by a large group of men in a screened population.

The European Screening trial and the prostate, lung, colorectal and ovarian (PLCO) trial from the United States have differed in their conclusions on the benefit of screening reducing mortality with the European Screening study showing a survival benefit using screening and the PLCO study showing no survival benefit. It is now accepted that the PLCO had significant contamination in that many men in the control arm had PSA testing anyway. Estimates put this at well over 50% and possibly even 70%-90%.

Although localised prostate cancer can be treated with surgery, radiation or more novel techniques, some recent randomised controlled trial (RCT) data have drawn light on the small benefit on reduction in metastases and mortality using treatment compared to active monitoring or watchful waiting.

However, one of the first groups to report was the Scandinavian Prostate Cancer Group - this study has reported a number of published papers on the success of radical prostatectomy versus watchful waiting. The group has reported that radical prostatectomy significantly improved local progression, metastasis and prostate cancer death (both prostate cancer specific mortality and importantly overall mortality) at a median follow-up of just over 8 years [14]. The latest article in the series published the results at approximately 13 years median follow-up and demonstrated that the significant overall survival benefit was more notable in men less than 65 years of age and also that radical prostatectomy reduces metastatic disease burden, androgen deprivation therapy and palliative treatments across all age groups as compared with watchful waiting [15].

In contrast in 2012, Wilt et al. showed that in men randomised between radical prostatectomy and watchful waiting diagnosed early in the PSA screening era of the United States, there was no overall survival or cancer-specific survival difference between the two arms. Subgroup analyses showed that there was a benefit in treating high-risk disease and probably in intermediate-risk disease but not in low-risk disease [16].

In 2017, Hamdy et al. reported on the ProtecT (prostate testing for cancer and treatment) trial which randomised men between radiotherapy, surgery and active monitoring [17]. They showed no survival difference at 10 years between these three arms which recruited from men diagnosed in the screening arm of the UK screening study called CaP. There were no differences in subgroups in this study. They showed a slight reduction in metastases in favour of treatment but there is criticism that ProtecT did not carry out modern active surveillance which might have not even showed this difference.

Radical treatment can obviously also have side effects with surgical and radiotherapy series reporting significant problems with erectile dysfunction (ED), incontinence, voiding LUTS and rectal toxicity [18]. The multidisciplinary team framework and recent National Institute for Health and Care Excellence (NICE) guidelines solidify the indication for treatment at all stages of prostate cancer. Without definitive evidence that prostate cancer screening decreases prostate cancer-specific mortality the cost per year of life saved is difficult to quantify. Overall it seems unlikely that such a programme will exist in the near future.

Table 1.5 Wilson and Junger criteria for a screening programme

Source: Adapted from Wilson JMG and Junger G.J R Coll Gen Pract 1968; 16: 318 [49].

In terms of the evidence for prostate cancer screening, several studies are now quoted:

1. The Quebec study - In the Quebec study patients were randomised to screened and non- screened populations. Initial reports showed a 70% decrease in prostate cancer death rates in the screened population. However this trial has been widely criticised due to crosscontamination of the patient groups. Further analysis on an intention to screen basis has shown no difference in mortality.

2. The Tyrol study - The Tyrol study is a natural experiment comparing two areas of Austria. In the Tyrol free PSA testing was introduced and a 70% uptake achieved. This resulted in a 44% decrease in prostate cancer mortality in 2000. However this effect was too rapid to explain this outcome. To elucidate the efficacy of localised disease treatment would take a far longer period of time and There fore these results probably represent aggressive treatment of locally advanced and metastatic disease.

3. The Seattle and Connecticut study - Seattle and Connecticut have very disparate socioeconomic populations. Seattle is prosperous and has more PSA screening and aggressive treatment of prostate cancer than in Connecticut. However no difference in prostate cancer mortality was demonstrated in these two populations.

4. European Randomised Study of Screening for Prostate Cancer (ERSPC) - This was conducted in several countries and has reported several times. The most recent was after 13 years of median follow-up and showed that one prostate cancer death was averted per 781 men invited for screening and that 27 screened men would have to be treated to prevent 1 death.

5. The U.S. Prostate, Lung, Colon and Ovary (PLCO) trial - This showed no mortality difference between systematic PSA testing and opportunistic PSA testing although recent data show that PSA testing in the control arm was likely to be as high as 90%.

Q. What do you know about the Prostate Cancer Prevention Trial (PCPT)?

A. The PCPT reported on 18,882 men randomised to receive prostate cancer chemoprevention with finasteride 5 mg od or placebo for 7 years [19]. The entry criterion was simply a PSA of less than 3 ng/mL. Prostate biopsies were performed if an abnormal DRE or PSA >4 ng/mL was found. All patients were offered an end-of-study biopsy. The study closed 15 months early due to significant results being achieved. It was then published in the New England Journal of Medicine in 2003. The study found a relative risk reduction in prostate cancer of 24.8% in the finasteride group. However a higher incidence of high-grade cancer (Gleason score 7, 8, 9 or 10) was detected in the finasteride arm (6.4%) compared to the placebo arm (5.1%). This worrying effect has been discussed at length in the urological literature. The main issues are first that cancer was detected in four times the number of patients than expected, and second that cancers were low or intermediate grade. The possible explanations of these findings have been discussed by Grover et al. among others [20]. They concluded that Finasteride may induce histological changes that mimic those seen in high-grade cancer. High-grade tumours are resistant to androgen deprivation and are unaffected by finasteride. Treatment created an environment that promotes the growth of high-grade tumours. These results may be an artefact due to increased detection (finasteride reduces prostate volume).

The REDUCE (Reduction by Dutasteride of Prostate Cancer Events) trial was an international, multicentre, double-blind, placebo-controlled chemoprevention study using the dual inhibitor of 5 alpha reductase subtypes I and II, dutasteride. In this trial patients were enrolled with a negative prostate biopsy within 6 months of entry and a PSA of between 2.5-10 ng/mL (50-60 years) and 3-10 ng/mL (>60 years). Repeat biopsy was performed at 2 and 4 years. The trial showed a relative risk reduction in detected cancer on biopsy with dutasteride of 22.8% over the 4-year study period [21]. However, once again there were some concerns in the numbers of higher-grade cancers detected in the dutasteride group in years 3-4 of the study.