For asymptomatic men with an initial total PSA above 3.0 ng/mL, does measuring free-to-total PSA percentage improve relative specificity without compromising prostate cancer or high-grade prostate cancer detection, when compared with a single total PSA result above 3.0 ng/mL? (PICOi question 6.2b)
For asymptomatic men with an initial total PSA above 3.0 ng/mL, does measuring PSA velocity improve relative specificity without compromising prostate cancer or high-grade prostate cancer detection, when compared with a single total PSA result above 3.0 ng/mL? (PICOi question 6.2b)
For asymptomatic men with an initial total PSA above 3.0 ng/mL, does measuring the Prostate Health Index (PHI) improve relative specificity without compromising prostate cancer or high-grade prostate cancer detection, when compared with a single elevated total PSA result above 3.0 ng/mL? (PICOi question 6.3b)
For asymptomatic men with initial total PSA above 3.0 ng/mL, does repeating the total PSA test and using an initial and repeat total PSA above 3.0 ng/mL as the indication for biopsy, improve relative specificity without compromising prostate cancer or high-grade prostate cancer detection, when compared with a single total PSA result above 3.0 ng/mL as the indication for biopsy? (PICOi question 6.4)
Guidelines developed in partnership with
Prostate Cancer Foundation of Australia and Cancer Council Australia PSA Testing Guidelines Expert Advisory Panel. Clinical practice guidelines PSA Testing and Early Management of Test-Detected Prostate Cancer. Sydney: Cancer Council Australia. [Version URL: http://wiki.cancer.org.au/australiawiki/index.php?oldid=122830, cited 2017 Nov 18]. Available from: http://wiki.cancer.org.au/australia/Guidelines:PSA_Testing/Testing_with_variants_of_PSA_or_repeat_PSA_testing_to_improve_specificity_after_an_initial_elevated_total_PSA.
National Health and Medical Research Council
In approving the guidelines (recommendations), NHMRC considers that they meet the NHMRC standard for clinical practice guidelines. This approval is valid for a period of five years. NHMRC is satisfied that the guidelines (recommendations) are systematically derived, based on the identification and synthesis of the best available scientific evidence, and developed for health professionals practising in an Australian health care setting.This publication reflects the views of the authors and not necessarily the views of the Australian Government.
- 1 Background
- 2 Evidence
- 3 Evidence summary and recommendations
- 4 Footnote
- 5 References
- 6 Discussion
A total PSA threshold of 4.0 ng/mL has traditionally been used as the criterion for prostate biopsy. The current trend towards the use of lower total PSA thresholds (e.g. 3.0 ng/mL), in place of 4.0 ng/mL or thresholds based on age-related normal values, has the potential to increase the number of prostate biopsies performed.
In asymptomatic men without a diagnosis of prostate cancer, a single total PSA test result above 3.0 ng/mL identifies three-to-four times as many men who do not have prostate cancer on biopsy as it does men who do have prostate cancer (positive predictive value [PPV] of 20–25%). Consequently, there has been increasing interest in developing strategies to reduce the number of unnecessary biopsies as this reduces the risk of complications of biopsy, discomfort and cost. While improvements in PSA testing specificity may reduce unnecessary biopsies, ideally such strategies would not materially reduce the sensitivity of PSA testing to presence of prostate cancer. Our analysis is based on the following assumptions:
- A reduction in sensitivity of less than 10% is acceptable.
- It is desirable that, for every cancer missed, at least 3–4 unnecessary biopsies are avoided.
These systematic reviews focused on tests that improved specificity for men with a total PSA level above 3.0 ng/mL. Because of the analytical and biological variability of total PSA, including the chronological rise in PSA in men in their sixties, this review focused on studies that used total PSA thresholds between 2.0 and 4.0 ng/mL or age-specific thresholds. Restricting the evidence to studies that used a total PSA threshold of 3.0 ng/mL would have limited the evidence and would not have taken into account analytical variation in the total PSA test over the last two decades.
Men with only slightly elevated levels are less likely to have prostate cancer and could benefit from attempts to improve specificity without compromising sensitivity, whereas men with higher PSA levels are more likely to have prostate cancer and for such men attempts to reduce unnecessary biopsies could compromise the effectiveness of the recommended PSA testing strategy. As a result, studies using a single total PSA threshold were restricted to those whose participants had a total PSA ≤ 5.5 ng/mL unless there were analyses for older men (who are more likely not to have prostate cancer despite a total PSA > 5.5 ng/mL). The majority of studies that included men with total PSA levels above 3.0 ng/mL threshold also included men with levels up to 10.0 ng/mL. Accordingly, these studies were excluded unless they provided subgroup analysis for men with total PSA less than or equal to 5.5 ng/mL. However, an exception to this principle was that studies of repeat PSA were included if their focus of investigation was the threshold for repeat PSA, and the initial PSA was a lesser concern.
To reduce the potential for bias, studies were restricted to those in which all participants underwent biopsy and there were clear indications for biopsy which included a specified total PSA threshold.
Free-to-total PSA percentage
Lowering the total PSA threshold to 3.0 ng/mL (compared with 4.0 ng/mL) will result in an increase in sensitivity and a fall in specificity. In principle, free-to-total PSA% can then be used to improve specificity. As the ratio of false positive to true positive biopsies with total PSA alone is typically three or four to one, a combined strategy with free-to-total PSA% should improve the efficiency of testing by removing more than three or four false positive biopsies for the loss of one true positive cancer detected.
More formal analysis of PSA dynamics, such as PSA velocity, PSA doubling time or PSA change require at least three or four total PSA measurements separated by several months. For men with total PSA levels already above the threshold, the delay in obtaining these PSA dynamic parameters may cause both anxiety and the possibility that the cancer will spread during that period.
Prostate Health Index
Criteria for biopsy have been proposed based on PHIii thresholds. A given PHI threshold might not be exceeded in a situation where pro2PSA is low and/or free PSA is high, despite a total PSA value greater than 3.0 ng/mL. Therefore, combining a total PSA threshold of 3.0 ng/mL with PHI might avoid unnecessary biopsies without significantly reducing the rate of detection of prostate cancer. PHI is a relatively new test and most PHI studies have been performed retrospectively. Furthermore, the ability of the PHI test to offset the decrease in total PSA specificity with increasing age is not understood.
Repeated total PSA
Given the current focus on total PSA above a given threshold as the criterion for referral or biopsy, men will often be referred as soon as total PSA is above the threshold, regardless of the possibility that such elevation may represent a transient rise from a lower baseline. Day-to-day biological variability of 15% in a man’s PSA level also means that for a man with an average level of 3.0 ng/mL the levels on consecutive days can be as high as 3.9 ng/mL (upper 95th percentile) or as low as 2.1 ng/mL (lower 95th percentile). It has therefore been suggested that elevated total PSA should be confirmed by a repeat test within several weeks. Should the repeat total PSA be below the total PSA threshold, biopsy might be avoided and cancer detection unaffected.
The search strategy, inclusion and exclusion criteria, and quality assessment are described in detail in the Technical report.
Free-to-total PSA percentage
A total of 14 studies met the inclusion criteria for this systematic review.
Twelve diagnostic accuracy studies were identified that compared the diagnostic performance of the free-to-total PSA% test with that of total PSA alone in men with total PSA levels above the threshold for biopsy but below 5.5 ng/mL or an age-specific threshold. All were assessed to be at risk of bias.iii
These studies found that lowering the free-to-total PSA% threshold gradually lowered sensitivity and improved specificity. Eight studies used a free-to-total PSA% threshold that retained a sensitivity of over 90% compared to total PSA alone. For men with a total PSA less than 5.5 ng/mL, using free-to-total PSA% thresholds of 25–31% reduced the number of unnecessary biopsies by 3.8, 4.0, 6.0, 8.0, 9.7 or 12.5 for each cancer missed. This variation may have been due to standardisation issues with both total PSA and free-to-total PSA% during the period 1997–2006.
Older men will more often have higher total PSA levels (> 4.0 ng/mL), without the presence of prostate cancer. Two studies examined the use of free-to-total PSA% for men aged over 69 years with a total PSA of 4.0–10.0 ng/mL. In one study, a free-to-total PSA% threshold of 22% resulted in over 96% sensitivity and the avoidance of at least 32 unnecessary biopsies for each cancer missed. In the other study, the use of a free-to-total PSA% threshold of > 25% resulted in much lower improvement of 4.4 unnecessary biopsies avoided for each cancer missed. The cancer detection rate in this study was 44%, so it is likely to represent a high-risk cohort. This may account for the reduced ratio of the unnecessary biopsies avoided to cancers missed.
The use of very low free-to-total PSA% thresholds improved specificity but compromised sensitivity to an unacceptable degree. For example, the use of free-to-total PSA < 10% as a threshold for biopsy resulted in failure to detect 70–90% of cancers in men with total PSA ranging from 2.0–4.0 ng/mL in two studies.
One diagnostic accuracy study at risk of biasiii was identified that compared the diagnostic performance of PSA velocity with that of total PSA alone in men with total PSA levels above the threshold for biopsy but below 5.5 ng/mL. Other studies were excluded because they did not use the recommended protocols for calculating PSA velocity.
The addition of PSA velocity to total PSA did not appear to improve diagnostic performance for men with a total PSA of 2.5–4.0 ng/mL. The single included study found that, for these men, the area under the receiver–operator curve for PSA velocity was significantly less than that for total PSA, which was, in turn, significantly less than that for free-to-total PSA%. Also, using a PSA velocity threshold that missed 20% of cancers (80% relative sensitivity), only approximately 27% of unnecessary biopsies (27% relative specificity) would have been avoided.
Prostate Health Index
No diagnostic accuracy studies were identified that compared the diagnostic performance of PHI with that of total PSA alone in men with total PSA levels above the threshold for biopsy but below 5.5 ng/mL.
Repeated total PSA
Two diagnostic accuracy studies at risk of biasiii were identified that compared the diagnostic performance of repeat total PSA with that of a single total PSA alone in men with total PSA levels above the threshold for biopsy but below 5.5 ng/mL. Both studies found that if the total PSA was lower or normalised on the second measurement, the number of negative biopsies could be reduced. The larger study found that if men were not biopsied because their total PSA had normalised to < 3.0 ng/mL, 8.6% of all cancer and 4% of higher-grade cancer would have been missed. If men did not undergo prostate biopsy because their total PSA fell by 30%, 5.9% of cancers would have been missed. In this study the ratio of avoided unnecessary biopsies to missed cancers was 4.99 if prostate biopsy was restricted to men with PSA levels that did not normalise (fall to below 3.0 ng/mL) or whose total PSA levels did not drop at least 30%. The smaller study using age-specific PSA thresholds found that referring for biopsy only those with total PSA levels that remained elevated, missed 6.0% of cancers and avoided 3.2 unnecessary biopsies for each cancer missed.
Evidence summary and recommendations
| Free-to-total PSA
In populations of men without a diagnosis of prostate cancer or symptoms that suggest prostate cancer, and with total PSA levels of 3.0–4.0 ng/mL, using a free-to-total PSA threshold of 26% as an indication for biopsy missed 7.4% of cancers, with 12.5 false positives avoided per each cancer missed.
In populations of men without a diagnosis of prostate cancer or symptoms that suggest prostate cancer, and total PSA levels between 2.0 and 4.0 ng/mL, using free-to-total PSA thresholds from 25% to 31% as indications for biopsy maintained a sensitivity of at least 90%, with 3.8–12.5 false positives avoided per cancer missed.
In populations of men aged over 69 years without a diagnosis of prostate cancer or symptoms that suggest prostate cancer, with a total PSA of 4.0–10.0 ng/mL and a cancer detection rate of 15%, using a free-to-total PSA threshold of 22% as an indication for biopsy maintained over 90% sensitivity and avoided 32 false positives per missed cancer.
There is very little evidence for whether free-to-total PSA% improves specificity in men aged under 50 years. Studies that reported free-to-total PSA% thresholds with acceptable sensitivity either did not include men under 50, or included only a small proportion.
|III-2||, , , , , , , , , , , , |
| PSA velocity
In a single level III-2 study, the use of PSA velocity to increase the specificity at PSA levels in the range of 2.5–4.0 ng/mL reduced sensitivity to an unacceptable degree.
| Prostate Health Index
There was no evidence for whether or not PHI testing improves the specificity of PSA testing in men with an elevated PSA up to 5.5 ng/mL, compared with PSA alone.
| Repeated total PSA
In men with an initial total PSA ≥ 3.0 ng/mL who underwent a second total PSA test within 1–3 months after the initial test, referring to biopsy only those men whose total PSA failed to normalise or reduce by 30% on the repeat total PSA test missed 8.6% and 5.9% of cancers, respectively, and avoided 4.99 unnecessary biopsies per cancer missed.
The use of an age-specific threshold, and referring to biopsy only those whose total PSA did not normalise on repeat total PSA, missed 6% of cancers and resulted in a ratio of unnecessary biopsies to missed cancers of 3.20.
For men aged 50–69 years with initial total PSA greater than 3.0 ng/mL, offer repeat PSA within 1–3 months.
For those with initial total PSA greater than 3.0 ng/mL and up to 5.5 ng/mL, measure free-to-total PSA percentage at the same time as repeating the total PSA.
For men aged 50–69 years with initial total PSA greater than 3.0 ng/mL who have undergone repeat total PSA and free-to-total PSA percentage tests at follow-up 1–3 months later, offer prostate biopsy:
For men aged 50–69 years with a previous total PSA test result greater than 3.0 ng/mL who are not offered prostate biopsy (or do not accept prostate biopsy when offered) after follow-up PSA testing, explain that there is a small chance of missing a significant cancer and advise them to return for PSA testing within 2 years.
|Measurement of PSA velocity is not recommended to increase specificity of a total PSA test result of 3.0 ng/mL or greater.||D|
Do not use the PHI test to increase specificity of a total PSA test result of 3.0 ng/mL or greater, except in the context of research conducted to assess its utility for this purpose.
Health system implications
Free-to-total PSA% is in common usage when total PSA levels are elevated. The free-to-total PSA% decision thresholds used are either < 10% or < 25%.
Implementation of these recommendations would not require changes in the way care is currently organised.
Misuse or new safety concerns from these recommendations are not envisaged. An increase in litigation alleging malpractice is possible if the Evidence-based and Consensus-based recommendations relating to total PSA and free-to-total PSA are not followed in practice. This potential legal risk will be mitigated by robust efforts to ensure that knowledge of the guideline is disseminated to all relevant health practitioners and the development of aids that will assist them in practising according to the guideline. The Consensus-based recommendations relating to the PSA velocity and PHI tests could mitigate risk of litigation for practitioners who practice in accordance with the evidence with respect to these tests.
Offering a repeat total PSA test and free-to-total PSA% test if total PSA is greater than 3.0 ng/mL will increase the number of PSA estimations and reduce the number of biopsies.
The measurement of free-to-total PSA% is reimbursable in Australia and extensively used. These recommendations should increase appropriateness of existing use.
Barriers to implementation
There are no apparent barriers to the implementation of the recommendations regarding repeat total PSA tests or free-to-total PSA% tests.
i Clinical questions were translated into the PICO framework: population, intervention (or exposure), comparator and outcome (see Appendix 3).
ii PHI is calculated using the formula: ([-2]proPSA/free PSA) × √ total PSA.
iii The tool for assessing risk of bias for this type of research question classified studies as being ‘at risk’ or ‘not at risk’ (see Technical report).
- Bokhorst LP, Zhu X, Bul M, Bangma CH, Schröder FH, Roobol MJ. Positive predictive value of prostate biopsy indicated by prostate-specific-antigen-based prostate cancer screening: trends over time in a European randomized trial*. BJU Int 2012 Dec;110(11):1654-60 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/23043563.
- Hugosson J, Aus G, Bergdahl S, Fernlund P, Frösing R, Lodding P, et al. Population-based screening for prostate cancer by measuring free and total serum prostate-specific antigen in Sweden. BJU Int 2003 Dec;92 Suppl 2:39-43 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/14983953.
- Catalona WJ, Partin AW, Finlay JA, Chan DW, Rittenhouse HG, Wolfert RL, et al. Use of percentage of free prostate-specific antigen to identify men at high risk of prostate cancer when PSA levels are 2.51 to 4 ng/mL and digital rectal examination is not suspicious for prostate cancer: an alternative model. Urology 1999 Aug;54(2):220-4 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/10443714.
- Egawa S, Suyama K, Matsumoto K, Kuwao S, Baba S. Prospective evaluation of prostate cancer detection by prostate specific antigen related parameters: comparison in serum and plasma samples. J Urol 2002 Jan;167(1):97-102 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/11743284.
- Kobayashi T, Kawahara T, Nishizawa K, Ogura K, Mitsumori K, Ide Y. Volume-adjusted prostate-specific antigen (PSA) variables in detecting impalpable prostate cancer in men with PSA levels of 2-4 ng/mL: transabdominal measurement makes a significant contribution. BJU Int 2005 Jun;95(9):1245-8 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/15892809.
- Lodding P, Aus G, Bergdahl S, Frösing R, Lilja H, Pihl CG, et al. Characteristics of screening detected prostate cancer in men 50 to 66 years old with 3 to 4 ng./ml. Prostate specific antigen. J Urol 1998 Mar;159(3):899-903 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/9474178.
- Okihara K, Fritsche HA, Ayala A, Johnston DA, Allard WJ, Babaian RJ. Can complexed prostate specific antigen and prostatic volume enhance prostate cancer detection in men with total prostate specific antigen between 2.5 and 4.0 ng./ml. J Urol 2001 Jun;165(6 Pt 1):1930-6 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/11371884.
- Raaijmakers R, Blijenberg BG, Finlay JA, Rittenhouse HG, Wildhagen MF, Roobol MJ, et al. Prostate cancer detection in the prostate specific antigen range of 2.0 to 3.9 ng/ml: value of percent free prostate specific antigen on tumor detection and tumor aggressiveness. J Urol 2004 Jun;171(6 Pt 1):2245-9 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/15126795.
- Reissigl A, Klocker H, Pointner J, Fink K, Horninger W, Ennemoser O, et al. Usefulness of the ratio free/total prostate-specific antigen in addition to total PSA levels in prostate cancer screening. Urology 1996 Dec;48(6A Suppl):62-6 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/8973702.
- Reissigl A, Horninger W, Fink K, Klocker H, Bartsch G. Prostate carcinoma screening in the county of Tyrol, Austria: experience and results. Cancer 1997 Nov 1;80(9):1818-29 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/9351555.
- Roehl KA, Antenor JA, Catalona WJ. Robustness of free prostate specific antigen measurements to reduce unnecessary biopsies in the 2.6 to 4.0 ng./ml. range. J Urol 2002 Sep;168(3):922-5 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/12187191.
- Safarinejad MR. Population-based screening for prostate cancer by measuring free and total serum prostate-specific antigen in Iran. Ann Oncol 2006 Jul;17(7):1166-71 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/16684791.
- Kravchick S, Peled R, Dorfman D, Agulansky L, Ben-Dor D, Cytron S. Predictive criteria for prostate cancer detection in men with serum PSA concentration of 2.0 to 4.0 ng/mL. Urology 2005 Sep;66(3):542-6 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/16140074.
- Luboldt HJ, Swoboda A, Börgermann C, Fornara P, Rübben H, Early Detection Project Group of the German Society of Urology. Clinical usefulness of free PSA in early detection of prostate cancer. Onkologie 2001 Feb;24(1):33-7 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/11441278.
- Catalona WJ, Partin AW, Slawin KM, Brawer MK, Flanigan RC, Patel A, et al. Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA 1998 May 20;279(19):1542-7 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/9605898.
- Djavan B, Zlotta A, Kratzik C, Remzi M, Seitz C, Schulman CC, et al. PSA, PSA density, PSA density of transition zone, free/total PSA ratio, and PSA velocity for early detection of prostate cancer in men with serum PSA 2.5 to 4.0 ng/mL. Urology 1999 Sep;54(3):517-22 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/10475364.
- Rosario DJ, Lane JA, Metcalfe C, Catto JW, Dedman D, Donovan JL, et al. Contribution of a single repeat PSA test to prostate cancer risk assessment: experience from the ProtecT study. Eur Urol 2008 Apr;53(4):777-84 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/18079051.
- Boddy JL, Pike DJ, Al-Hayek S, Shaida N, Malone PR. An elevated PSA, which normalizes, does not exclude the presence of prostate cancer. Prostate Cancer Prostatic Dis 2005;8(4):349-52 Abstract available at http://www.ncbi.nlm.nih.gov/pubmed/16010283.
- The Royal College of Pathologists of Australasia. Prostate specific antigen testing: Age-related interpretation in early prostate cancer detection. Revised position statement.; 2011 [cited 2014 Nov 20] Available from: http://www.rcpa.edu.au/getattachment/37efcb2a-0844-4250-b9e7-a53a26eeafec/Prostate-Specific-Antigen-Testing-Age-related-inte.aspx.