NHMRC Public Consultations

Skip Navigation and go to Content
Visit NHMRC website

PSA Testing for Prostate Cancer in Asymptomatic Men: Information for Health Practitioners submission

This submission reflects the views of
Organisation Name: 
Urological Society of Australia and New Zealand
Personal Details
Specific Questions
1. What are the potential benefits of PSA testing?: 

Early detection: treatment is more likely to be potentially curative, rather than "simpler and more effective".


2. What are the potential harms of PSA testing?: 

Over-diagnosis: once again, the document inappropriately links diagnosis with treatment, with no comment as to the large number of men managed with surveillance rather than treatment.


3. How frequent are these benefits and harms?: 

We have significant concerns regarding the figures quoted in this section:

  1. Overall figures are misleading, since they vary considerably with patient age, medical history, family history and other factors. Although quoted figures are meant to refer to low-risk men aged 60, although it is not defined what "low risk" means. Recommend including figures for other risk groups as well.
  2. Mortality reduction & other figures are derived from specific studies – other studies suggest greater benefit – a range (2-5/1000) may be better
  3. The use of the word "many" in the context of over diagnosis & overtreatment is pejorative, and not reflective of the uncertainty surrounding these concepts – "some" would be better
  4. No account is taken of recent changes in practice which impact quoted figures:
  • Changes in biopsy technique, use of anaesthesia and analgesia to mitigate discomfort, increasing use of trans-perineal route
  • Increased application of active surveillance (~40% of low-risk men in Victorian data from the Prostate Cancer Registry) to minimise treatment-related harm – invalidates the quoted figures for treatment side-effects, hence quoting US data is inappropriate in Australia

4. What research has been done to study the effectiveness of PSA testing for prostate cancer?: 

Restricting the analysis of evidence to randomised controlled trials (RCTs) only does not allow an adequate assessment of the impact of PSA testing on prostate cancer outcomes. The RCTs are heterogeneous, and suffer from significant shortcomings in design that make them less likely to detect a true impact. Of the 6 studies quoted only the ERSPC and Goteborg are reasonable studies with major flaws in PLCO, Norkopping, Stockholm and Quebec all of which are well recognised in world literature. In particular PLCO compares an 85% screened group to a 52% screened group, hence it is NOT a study of testing vs no testing; Norkopping did not use modern biopsy techniques for its study nor did it included PSA in the first half of the study , both totally inappropriate to include in this analysis, as are the other 2 studies.

The RCTs also focus on population-wide screening, which is distinct from the current approach in Australia, which is usually opportunistic testing.

There are additional sources of evidence, especially from epidemiological studies which suggest diminishing rates of prostate cancer mortality and metastatic disease rates resulting from PSA testing.  

5. Does PSA testing in asymptomatic men reduce their risk of dying from prostate cancer?: 

The most reliable data on prostate cancer mortality is derived from the ERSPC and Goteborg trials, with other trials being too small or methodologically flawed. This show a reduced rate of dying from prostate cancer and whilst the absolute risk reduction is small, as not that many die from prostate cancer in the 10-15 year time frame, a point that we would accept, there is a significant relative risk reduction much more pronounced in younger men with no comorbidities in the Goteborg data which should be quantified.

Additional data from epidemiological studies (see above) that has not been considered.


6. Does PSA testing in asymptomatic men reduce their overall risk of dying?: 

Overall survival is a less meaningful endpoint than cancer specific survival in the screening studies. Certainly the power to detect any impact on overall survival was limited, and in the older population that is at risk for prostate cancer, death from competing causes is common. As such the aim of testing for a disease should be reduction of dying from that disease, given the inability to control for other disease states. 

7. Does PSA testing in asymptomatic men reduce their risk of having metastases present at diagnosis of prostate cancer?: 

This is a very important endpoint for PSA testing, since even if men do not die from their prostate cancer, if they develop metastaic disease and need androgen deprivation therapy, this causes significant suffering and impact on quality of life. This reduction should be quantified for consistency, as these guidelines have quantified harms it would seem appropriate that they also quantify benefits. 

8. Does PSA testing in asymptomatic men affect the quality of life of men who are diagnosed with prostate cancer?: 

This section is misleading - PSA testing does impact the quality of life of men diagnosed with prostate cancer, by reducing the risk of metastatic disease and consequent androgen deprivation therapy. The highest cost of prostate cancer care is in the final year of life with men with prostates left in situ suffering morbidity of local growth viz., need for prostate surgery, bleeding, obstructive renal failure requiring stents or nephrostomy, incontinence, permanent catheterisation – all well reported in the literature. It also fails to acknowledge the satisfactory QOL of men on AS. The reports highlights and quantify the risks of incontinence and erectile dysfunction (many of whom have ED pre treatment which is not mentioned) should men choose therapy but it fails to mention or quantify the harms that may occur if men are diagnosed with advanced disease.

10. How accurate is a PSA test?: 

The "accuracy" of the PSA test is variable, depending on factors such as patient age, family history, degree and rate of elevation, DRE findings etc. The overall figure of 7/10 is potentially misleading without taking the above into account.

Currently, additional measures such as PSA kinetics and isoforms, as well as new biomarkers and imaging findings are being assessed in an effort to improve the selection of men for prostate biopsy.

Again the use of the word "many" in the context of "slow-growing" cancers is pejorative, and would be better to replace with "some" – the future behaviour of a particular man’s prostate cancer is often not predictable at the time of diagnosis, and the proportion of so-called "insignificant" or non life-threatening cancers varies depending on a range of factors such as PSA level, DRE findings, age, family history etc.


11. Are there any other screening tests available for prostate cancer?: 

It is important to stress that DRE should be performed along with PSA-testing, since a proportion of prostate cancers may present with a normal PSA but abnormal findings on DRE.


13. What are normal and abnormal PSA test results?: 

The impact of pharmacological treatment with 5-alpha reductase inhibitors on lowering PSA levels (approximatly 50%) is worth including


15. What happens if a man receives an abnormal PSA test result?: 

The information presented regarding prostate biopsy over-emphasises the negative impact of biopsies. While it is reasonable to list them, it is important to point out:

  1. False negative rates from current biopsy approaches are low, around 5-10%.
  2. Most adverse effects of biopsies are mild and self-limited, serious complications are rare

16. If a man receives a diagnosis of prostate cancer after an abnormal PSA test, what choices does he have?: 

The summary of treatment options comes across as confusing. Watchful waiting and active surveillance are different in their intent, and applied to different patient populations. Curative local treatment usually entails either radical prostatectomy or radiotherapy. Androgen deprivation therapy is usually a palliative treatment for patients with metastatic disease.

The side-effects of treatments are characterised as being "common" – although they vary significantly depending on many factors. It would be more appropriate to state "known side-effects"


17. If a man decides not to have a PSA test what risks should he and his family be aware of?: 

"In the unlikely event that prostate cancer is present" – makes the unwarranted presupposition that the man has a low risk of prostate cancer – better replaced with "If prostate cancer is present"


General Comments
18. Considering the Information Document is for Health Practitioners, do you have any other comments?: 

USANZ would suggest keeping in mind the following caveats:

  • Prostate cancer remains a common condition, and as such continues to cause a large number of deaths. Currently, PSA testing and early detection provides the main means of preventing these deaths. 
  •  Prostate cancer also causes significant symptoms and complications in many men even when it is not fatal for them – the focus on death can be misleading because prostate cancer occurs in older men, who have competing causes of death. Published trials clearly show that PSA testing reduces the occurrence of metastatic prostate cancer.
  • Some men are at a higher risk of developing and dying from prostate cancer (e.g. those with a family history of prostate cancer) – if they are discouraged from PSA testing, they may miss a critical opportunity for early detection and treatment.
  • The guidelines refer to asymptomatic men only, and it should be remembered that PSA testing has a well-defined role in the assessment of men with urinary symptoms

Page reviewed: 4 March, 2014