Should the National Health Service use PSA to Screen for Prostate Cancer?
Sophie Merrick
Medical Student
Manchester Medical School
Dr Ananya Choudhury
Consultant Clinical Oncologist
The Christie, Manchester
Key Points:
• It has been suggested that PSA could be used as a screening tool for the identification of prostate cancer, the most common cancer occurring in men in the UK.
• The European Urological Association does not currently recommend that PSA should be used as a screening tool for prostate cancer.
• If PSA were to be implemented as a screening tool it is unclear what the recommended cut off value should be, due to effects on sensitivity and specificity of the test.
• PSA cannot differentiate between high and low grade tumours, and therefore is a poor predictor of outcome.
• It is unclear how often patients should be tested in a potential screening programme as the lead times differ for aggressive and slow growing cancers.
• Two large scale randomised control trials failed to provide conclusive evidence as to whether PSA screening would lead to a reduction in mortality.
• It is likely that PSA screening would lead to significant over diagnosis and subsequent over treatment of patients resulting in significant physical and psychological consequences for the individual, as well as cost implications for society.
Abstract
It has been suggested that Prostate Specific Antigen (PSA) may be an appropriate screening tool for the identification of prostate cancer, however there is debate surrounding the validity of this test. Two large scale randomised control trials failed to provide conclusive evidence as to whether PSA screening would lead to a reduction in mortality. The trials did raise concern over the volume of patients that would be over diagnosed and subsequently over treated if a screening programme were to be implemented. The treatment of these patients would be complex; resulting in unnecessary adverse physical and psychological consequences, in addition to significant cost implications for society.
Introduction
The NHS has nationwide screening programmes for breast, cervical, and bowel cancer, all of which have been shown to reduce mortality (Gøtzsche and Nielsen 2009; Forbes at al 2002; Hewitson et al 2007). Given the success of these programmes, and that prostate cancer is the most common cancer in men, it has been suggested that similar nationwide screening for prostate cancer should be introduced in the UK.
Prostate cancer is the most common cancer occurring in men in the UK, and the incidence is rising with rates almost tripling in the last 30 years (Cancer Research UK 2010). However this increase in incidence has not been reflected by a change in mortality rates, which stands at approximately 10,200 deaths per year in the UK. This is likely to be attributed to an increasing use of the Prostate Specific Antigen (PSA) test, which has lead to an increase in the identification of low risk localised prostate cancer (Brewster D et al 2000; Potosky et al 1995).
Many of the patients identified are thought to be over diagnosed, that is, if unidentified these patients would have died whilst suffering from prostate cancer, rather than dying of the prostate cancer itself (Selley at al 2000). In fact, 80% of 80-year-old men have been shown to have neoplastic change within the prostate gland. If all of these 80% were to be identified through screening the vast majority would certainly be over diagnosed and subsequently over treated (Sakr et al 1996).
PSA is a protein produced by cells of the prostate gland. Damage to these cells can lead to elevated levels of PSA. Not only can the PSA levels be abnormal in men with prostate cancer, PSA levels can also be elevated in benign conditions such as bacterial prostatitis and benign prostatic hypertrophy (BPH). In addition, some men with prostate cancer may have a normal PSA level.
The American Urological Association recommends annual PSA screening for men aged 40 or over, if they have a life expectancy of at least ten years. However, the European Urological Association, along with urological bodies in Japan and Australia, have deduced that there is insufficient evidence from current research to recommend the PSA test to be used as a screening tool for the identification of Prostate Cancer (Stark et al 2009). This decision has been supported by Cochrane Review (Ilic et al 2006). However, requests for the PSA test in primary care in the UK have risen substantially in the last year (Stark et al 2009).
The UK National Screening Committee outlines clear criteria for the use of screening, primarily:
1) the condition should be common
2) the test should be a validated screening test with a suitable cut off level defined
3) There should be effective treatment for those identified, and this should be an improvement on that offered if the disease were not identified through screening
4) the screening programme should be supported by evidence from high quality randomised control trials in that it will reduce morbidity and mortality
5) the benefit from the screening programme should outweigh the physical and psychological harm, as well as being cost effective (The Scottish Government 2010).
Criteria 1 is fulfilled, however in relation to 2, 3, 4 & 5 there are many concerns over the use of PSA as a screening test. The UK Screening Committee do not currently recommend the use of PSA as a screening tool but aims to review the evidence for this decision in 2010/2011. This article aims to explore whether PSA is a valid screening tool, and the effects of treating this largely low risk subgroup of patients identified with localised prostate cancer.
Is PSA a suitable test for screening?
An ideal screening test for the identification of prostate cancer would be highly sensitive and highly specific. A highly sensitive test would identify a large number of patients who had prostate cancer, with a high rate of true positives. A highly specific test would exclude a large number of patients who did not have prostate cancer, and result in a low rate of false positives. The sensitivity of PSA for prostate cancer is 75% when a 4µg/ml cut off is used, and the specificity is 60% if benign prostatic hypertrophy has been excluded (Wilson and Crawford 2004). This specificity results in approximately 2 in every 3 men who have an elevated PSA not having prostate cancer.
One of the main problems with the use of PSA as a screening tool is there is much debate over which threshold to use. As mentioned previously, Criteria 2 of the UK National Screening Committee states that there must be a clear cut off value.
Since the 1990’s 4µg/ml has generally been the accepted threshold, however high grade prostate cancers have been found below this threshold (Ilic and Green 2009; Holmstr?m B et al 2009). This can lead to false reassurance whereby even when patients do become symptomatic they present to primary care later than if they had not received this test result.
Using a PSA cut off value of 1µg/ml increases the sensitivity and allows clinicians to confidently exclude prostate cancer. Patients are likely to find this reassuring. However, this value reduces specificity leading to a large number of false positives. Furthermore, a lower threshold will to lead to an even higher level of over diagnosis leading to further investigations, possible unnecessary treatment and a huge psychological burden for the patient to deal with, all negatively affecting the patient’s quality of life (Stark et al 2009).
Furthermore the PSA test cannot differentiate between low and high grade tumours, and therefore the PSA value is a poor predictor of outcome as many men with low grade tumours would die with the prostate cancer, rather than of the prostate cancer itself (Stark et al 2009; Ilic & Green 2009).
A certain amount of lead time is required for cancers to be identified and treated before they become more serious. In breast cancer this is approximately 2-4 years (Pashayan et al 2009), in cervical cancer approximately 8-9 years (Hawkes et al), and in bowel cancer approximately 10 years (Heiken et al 2006). However, the lead time gained by use of the PSA is 11-30 years. This results in men living with worry and any adverse effects of investigation and treatment for far longer than in other diseases that are screened in the UK (Pashayan et al 2009). Furthermore a diagnosis significantly in advance of onset may result in lead time bias. This could result in an apparently prolonged survival, when in fact there is actually no effect on the outcome of the prostate cancer.
Another issue with PSA screening is how often patients should be tested. The European Randomised Study of Screening for Prostate Cancer screened participants every four years (Schr?der F et al 2009). However, the lead time of aggressive cancers, with the poorest prognosis associated, is likely to be much shorter. Therefore these cancers could be missed during a screening programme of this frequency. This can result in length time bias whereby the screening programme will be far more effective at picking up slow growing cancers that have a longer lead time, and these cancers are therefore likely to be over represented. In this case an apparent overall improvement in survival rates would be seen with screening, although the test may not have affected survival outcomes in any way.
Will the use of PSA screening decrease morbidity and mortality?
Criteria 4 outlined by the UK National Screening Committee is that a screening programme should result in an overall reduction in morbidity and mortality, and benefits should outweigh the costs. The best information on the use of PSA as a screening tool comes from two randomised control trials.
The results from an American randomised control trial showed that there was no benefit from screening at 7 years follow up, with similar results at 10 years. The incidence of deaths per 10,000 was 50 and 44 deaths in the screened and control group respectively. However, in this study of 38,350 men in the comparison group who were not screened, half sought PSA screening elsewhere, contaminating the results. The lack of benefit in the screening group could be due to this contamination, or a number of other reasons such as improvements in treatment, too high a threshold for the PSA test or insufficient follow up time (Andriole et al 2009).
The European Randomised Study of Screening for Prostate Cancer reported a 20% relative reduction in the risk of prostate cancer mortality with screening taking place every four years. This benefit was only seen in those aged between 55-69 years at randomisation. It concluded that 1410 men must be screened, and 48 cases of prostate cancer treated to prevent one death. If it were assumed that a man who was not screened would die on average five years after his diagnosis of prostate cancer, this would result in an average gain of 2.6 days per man screened. However, this figure could be even lower as many men live much longer than five years after their diagnosis (Stark et al 2009).
Although the number needed to treat (NNT) is of a similar value to that for mammography for breast cancer screening and fecal occult blood sampling for colorectal cancer screening, over diagnosis and over treatment were significantly higher in the PSA prostate cancer screened group (Catalona et al 1991; Denis et al 1995). In fact the rate of over diagnosis in the European study was estimated to be as high as 50% (Draisma et al 2003).
Therefore both randomised control trials provide inconclusive evidence as to whether PSA screening for prostate cancer will lead to a reduction in mortality. More research is needed from large randomised control trials into over diagnosis and over treatment of patients, and the physical and psychological impact of this. In addition, cost-benefit analysis should be undertaken.
How would these patients be treated?
If PSA screening were initiated, it would aim to pick up prostate cancer at an early stage. There are four main treatment options available for patients with low risk prostate cancer: active surveillance, surgery (radical prostatectomy), external beam radiotherapy and low-dose rate brachytherapy (Merrick et al 2009). Increasingly patients are ‘suitable for all options’ meaning that they can pick which treatment they would prefer to receive having had all treatment options discussed.
Current research suggests that surgery, external beam radiotherapy and low-dose rate brachytherapy all have equal survival outcomes when treating low-risk prostate cancer, with 10 year overall survival rates of approximately 60% and cancer-specific survival rates of approximately 80% (Cancer Advice 2010; Cancer Research UK 2010). A large randomised trial called PROTECT is currently underway which aims to evaluate all of these treatments (University of Bristol 2010).
Patients who choose the option of active surveillance are monitored closely by testing serum PSA levels and repeating prostate biopsies if necessary. Results of these tests provide evidence supporting the continuation of the active surveillance programme or proceeding to curative treatment. Active surveillance could prevent the unnecessary debilitating side effects of treatment such as impotence and incontinence.
However, many men with a diagnosis of cancer choose to have the affected organ removed completely. For these men, radical prostatectomy may be a better option. Impotence occurs in up to 50% of men under the age of 60 who have had nerve sparing prostatectomy, increasing up to 80% of men over the age of 70 who have not had nerve sparing surgery (Macmillan Cancer Support 2010). In addition most men have some level of incontinence after surgery but this usually improves over time.
External beam conformal radiotherapy is now delivered in the form of Intensity Modulated Radiotherapy (IMRT) at many cancer centres, which reduces adverse side effects of treatment (Prostate Cancer Foundation 2010). It is delivered on consecutive weekdays on an outpatient basis for around 7 weeks, and some patients have to travel a considerable distance to receive this treatment.
The main side effects of external beam radiotherapy include tiredness, lethargy and “sunburn-like” irritation to the area being treated. The radiation field will encompass some of the bladder and rectum due to their proximity to the prostate gland, therefore, around 45% of patients experience increased urinary frequency and dysuria in the first six months, but the majority of these resolve by one year. In addition around 5% of men experience bowel symptoms such as increased frequency and urgency, proctitis, tenesmus, diarrhoea and bleeding and mucus per rectum (Prostate Cancer Foundation 2010). There is a 30-40% risk of impotence, although this is lower than with radical prostatectomy (The Christie 2010).
The fourth option, low-dose rate (LDR) brachytherapy involves the insertion of radioactive iodine (I125) seeds directly into the prostate gland. Brachytherapy tends to cause more urinary symptoms but fewer bowel symptoms when compared with external beam radiotherapy. Urinary symptoms requiring medication affect over 70% of men at 6 months, but this drops to 10% at 2 years (Prostate Cancer Foundation 2010). Urinary side effects include: frequency and urgency of micturition, poor flow, dysuria, haematuria and discomfort. There is a 40-50% risk of impotence in men under 60 years, and a reduction in ejaculate volume (The Christie 2010).
Not only do all of these treatments have considerable side effects and risks, they also present the patient with a very difficult decision as to which treatment they should receive. Therefore, if estimates of over diagnosis of 50% in the European Randomised Control Trial are correct, large numbers of men may be subject to unnecessary adverse physical and psychological consequences.
Conclusion
This review supports the UK National Screening Committee in its decision not to initiate prostate cancer screening in the UK. Although PSA can be a useful clinical aid, it cannot be described as a validated screening tool. It is neither sensitive, nor specific, and no clear cut off level has been defined. Furthermore it is unable to differentiate between indolent and life threatening cancers and therefore, provides no clear direction as to how a positive result should be dealt with.
There is no substantial evidence to support the use of PSA screening to reduce morbidity and mortality. However, research is highly suggestive that if PSA screening were implemented, it could lead to extremely high rates of over diagnosis and over treatment resulting in unnecessary physical and psychological adverse events for a large number of patients. In addition, over treatment would have huge cost implications for the NHS, and an accurate cost-benefit analysis would be required for screening to be considered ‘value for money’.
As an aside, it is recommended that clinicians, particularly in primary care, take more time to evaluate their decision to perform a PSA test. A PSA test should be considered in the same way that endoscopy would be considered for bowel cancer investigation, as an elevated result clearly has significant implications. Therefore it is suggested that clinicians ordering this test should counsel the patient in making an informed decision, with the patient understanding the limits of the use of the PSA test.
Recommended further research:
• Large randomised control trials should assess the impact of over diagnosis and therefore over treatment of patients identified with an elevated PSA through screening.
• The physical and psychological implications of diagnosis and treatment should be evaluated, particularly in those for whom the cancer would never have proved fatal.
• Cost benefit analysis of a screening programme within the National Health Service should be evaluated.
Key References
Andriole, G et al. Morality Results from a Randomized Prostate Cancer Screening Trial. New England Journal of Medicine. 2009; 360:1310-1319
Ilic D, O’Connor, Green S, Wilt T. Screening for prostate cancer. Cochrane Database of Systematic Reviews. 2006, Issue 3. Art. No.: CD004720. DOI: 10.1002/14651858. CD004720.pub2
Schr?der F et al. Screening and Prostate-Cancer Mortality in a Randomized European Study. New England Journal of Medicine. 2009; 360:1320-1328
Stark J, et al. Screening for prostate cancer remains controversial. BMJ. 2009; 339:3601-3609
The Scottish Government. Screening Criteria set by the UK National Screening Committee. Available from: http://www.scotland.gov.uk/Topics/Health/health/screening/criteria [Accessed 12.03.10]
References
Andriole, G et al. Morality Results from a Randomized Prostate Cancer Screening Trial. New England Journal of Medicine. 2009; 360:1310-1319
Brewster D, et al. Rising incidence of prostate cancer in Scotland: increased risk or increased detection? BJU International. 2000; 85:463-473
Cancer Advice. Treatment of Prostate Cancer. Available from: http://www.canceradvice.co.uk/prostate-cancer/treatment/ [Accessed 09.03.10]
Cancer Research UK. Prostate Cancer. Available from: http://info.cancerresearchuk.org/cancerstats/types/prostate/ [Accessed 09.03.10]
Catalona W et al. Measurement of prostate specific antigen in serum as a screening test for prostate cancer. New England Journal of Medicine. 1991; 324:1156-1161
Denis L et al. Report of the consensus workshop on screening and global strategy for prostate cancer. Cancer. 1995; 75:1187-1207
Draisma G et al. Lead times and overdetection due to prostate specific antigen screening: estimates from the European Randomized Study of Screening for Prostate Cancer. Journal of the National Cancer Institute. 2003; 95:868-878
Forbes CA, Jepson RG, Martin-Hirsch PPL. Interventions targeted at women to encourage the uptake of cervical screening. Cochrane Database of Systematic Reviews 2002, Issue 3. Art. No.: CD002834. DOI: 10.1002/14651858.CD002834.
Gøtzsche PC, Nielsen M. Screening for breast cancer with mammography. Cochrane Database of Systematic Reviews 2009, Issue 4. Art. No.: CD001877. DOI: 10.1002/14651858.CD001877.pub3.
Hawkes, A et al. Cervical Cancer Screening. American College of Preventative Medicine Practice Policy Statement.
Heiken et al. Appropriateness criteria colorectal cancer screening. American College Radiology. 2006;1-7
Hewitson P, Glasziou PP, Irwig L, Towler B, Watson E. Screening for colorectal cancer using the faecal occult blood test, Hemoccult. Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No.: CD001216. DOI: 10.1002/14651858.CD001216.pub2.
Holmstr?m B et al. Prostate specific antigen for early detection of prostate cancer: longitudinal study. BMJ. 2009; 339;3537
Ilic D & Green S. Prostate specific antigen for detecting early prostate cancer. BMJ. 2009; 339:3572-3573
Macmillan Cancer Support. Surgery for early prostate cancer. Available from: http://www.cancerbackup.org.uk/Cancertype/Prostate/Treatmentforearlyprostatecancer/Surgery [Accessed 09.03.10]
Merrick S et al. Prostate Cancer. GM2 Oncology. 2009; 39(12):7-11
Pashayan N et al. Mean soujourn time, overdiagnosis, and reduction in advanced stage prostate cancer due to screening with PSA: implications of sojourn time on screening. Br J Cancer. 2009; 100:1198-204
Potosky A, et al. The Role of Increasing Detection in the Rising Incidence of Prostate Cancer. JAMA. 1995; 273:548-552
Prostate Cancer Foundation. Prostate Cancer info. Available from: http://www.prostatecancerfoundation.org/site/c.itIWK2OSG/b.1419647/k.B5CA/Bowel_Dysfunction.htm [Accessed 09.03.10]
Sakr WA, et al. Age and Racial distribution of prostatic intraepithelial neoplasia. Eur Urol. 1996; 30(2):138-44
Schr?der F et al. Screening and Prostate-Cancer Mortality in a Randomized European Study. New England Journal of Medicine. 2009; 360:1320-1328
Selley S, et al. Diagnosis, management and screening of early localised prostate cancer. BJU International. 2000; 85(4):463-72
Stark J, et al. Screening for prostate cancer remains controversial. BMJ. 2009; 339:3601-3609
Ilic D, O'Connor D, Green S, Wilt T. Screening for prostate cancer. Cochrane Database of Systematic Reviews 2006, Issue 3. Art. No.: CD004720. DOI: 10.1002/14651858.CD004720.pub2.
The Christie. Prostate Brachytherapy. Available from: http://www.christie.nhs.uk/patients/booklets/text/prostate_brachytherapy/default.aspx#options [Accessed 09.03.10]
The Scottish Government. Screening Criteria set by the UK National Screening Committee. Available from: http://www.scotland.gov.uk/Topics/Health/health/screening/criteria [Accessed 12.03.10]
Uiversity of Bristol. ProtecT study. Available from: http://www.epi.bris.ac.uk/protect/ [Accessed 09.03.10]
Wilson, S & Crawford, D. Screening for Prostate Cancer. Clinical Genitourinary Cancer. 2004;3:1558-7673
Copyright © Priory Lodge Education 2010
First Published June 2010
All pages copyright ©Priory Lodge Education Ltd 1994-
