Commentary: controversies in NICE guidance on prostate cancerBMJ 2008; 336 doi: https://doi.org/10.1136/bmj.39501.621690.AD (Published 13 March 2008) Cite this as: BMJ 2008;336:612
- Timothy J Wilt, professor of medicine1
- 1Minneapolis VA Center for Chronic Disease Outcomes Research, 1 Veterans Drive (111-0), Minneapolis, MN 55417
The NICE guidelines on prostate cancer provide comprehensive advice on best practice for diagnosis and treatment of prostate cancer. They are based on systematic reviews of the evidence, incorporate multidisciplinary opinions, and try to balance the values of healthcare providers and patients for various outcomes while emphasising a patient centred approach. Their conclusions are generally consistent with other reviews and guidelines evaluating similar evidence.1 2 3 4 5 If followed, these recommendations will likely improve prostate cancer outcomes while reducing unnecessary, ineffective, harmful, and costly care.
Paucity of randomised controlled trials
Although NICE’s recommendations are generally appropriate, any guidelines on prostate cancer are going to be hampered by the lack of high quality information available. The paucity of randomised trials limits the quality of data used for informed decision making, particularly regarding detection and treatment of localised disease. Even where randomised trials have shown benefits, the absolute magnitude of benefit is generally small, requires many years to accrue, and must be weighed against accompanying harms and costs. Few studies enrolled men with disease that had been detected as a result of measuring the level of prostate specific antigen (PSA) or were adequately powered to assess survival outcomes. Individual patients and providers may place different values on a particular outcome or have clinical and tumour characteristics that have not been adequately assessed by the trials providing the evidence used for drafting guidelines. Thus, it is difficult to mandate rigid adherence to these recommendations or to create national practice standards.
Limitations that may reduce uptake by clinicians
The full NICE guideline is wide ranging and can serve as a single source of information on prostate cancer. However, an individual clinician or patient will find that many sections are not pertinent to their situation. The broad scope makes the guideline more cumbersome than targeted recommendations aimed at specific points in the diagnosis and treatment pathway. Recommendations are made without adequate documentation as to their confidence, strength of evidence, magnitude of harms and benefits or methods for successful implementation and evaluation. The reader is required to spend considerable additional effort to determine the evidence base for these recommendations or accept or reject them as written.
These limitations can reduce clinicians’ acceptance and implementation.6 For example, some detection, treatment, and monitoring recommendations are very prescriptive despite lack of evidence supporting this level of detail. In contrast, recommendations not to routinely use combined androgen suppression or bisphosphonates for metastatic disease are brief and relatively underemphasised, though information supporting NICE recommendations exists from many large, high quality randomised studies. Clinicians are encouraged to counsel patients about the risks and benefits of prostate cancer screening as well as different treatment options. The full guideline notes that patients “should be adequately informed about the effects of prostate cancer and the treatment options on their sexual function, physical appearance, continence and other aspects of masculinity. Healthcare professionals should support men and their partners or carers to make treatment decisions taking into account the effects on quality of life as well as survival.” Although links to additional information are included, quantitative outcome data due to these effects and user friendly information tools to guide providers and patients in decision making are not immediately available.
Basing treatment recommendations on patient and tumour characteristics
For men found to have localised prostate cancer, NICE recommends that providers counsel men about several options and make treatment recommendations based on patient and tumour characteristics. NICE guidelines list watchful waiting as an option for low and intermediate risk disease. This is a step forward given the data on the natural course of the disease and the findings from randomised treatment trials.1 7 Active surveillance is recommended for low risk disease and radical prostatectomy or radiation therapy for high risk disease. However, consensus for these recommendations is lacking, little evaluation has been done for any active surveillance protocol, and little evidence shows superiority of one treatment option over another on the basis of disease risk status.
The role of PSA testing in deciding when to perform biopsy
NICE’s recommendations that the serum PSA level alone should not automatically lead to a prostate biopsy are sound. The aim of prostate biopsy is to detect prostate cancers that have the potential for causing harm and that can effectively be treated, rather than finding each and every cancer. No randomised trial has shown that PSA testing reduces overall or disease specific mortality. No PSA value accurately rules in or out prostate cancer or high grade disease. Widespread PSA testing and referral for prostate biopsy of all men with values greater than an arbitrary threshold of 4 ng/ml has labelled many men “abnormal,” increased the number having a biopsy, and resulted in detection of a large reservoir of clinically unimportant “pseudo-tumours.”5
Incorporating a patient’s age, race, family history of prostate cancer, digital rectal examination, and past biopsy results can predict the likelihood of a biopsy specimen proving to be positive for any cancer and high grade cancer.4 However, data are not available to determine the impact these indices have on patient anxiety or decision making, number of biopsies performed, and frequency of serendipitously detected pseudo-tumours5 versus tumours missed that would be clinically important but potentially effectively treated. Raising the PSA threshold for prostate biopsy (for example to >10 ng/ml) is more likely to reduce unnecessary diagnostic and treatment interventions, although missing clinically important but treatable disease may result.5
The use of emerging technologies such as cryotherapy, high intensity focus ultrasonography, photon beam radiation therapy, and robotic radical prostatectomy has increased in many countries. However, a comprehensive review found that no randomised trials had evaluated effectiveness. Few data exist on long term comparative effectiveness and harms.1 NICE is wise to caution against early adoption of these unproved therapies.
Does the structure and process of health services affect patient outcomes?
The guideline did not fully discuss the evidence for any effects of the structural and process measures of care on outcomes. Surgeons and hospitals with high annual volumes of radical prostatectomy have better outcomes.1 Additional structural and process measures to define quality of care have been proposed. Whether improved outcomes for prostate cancer will result is not yet known but needs evaluation.
Dissemination and implementation
The next steps for improving care for men with prostate cancer should focus on effective dissemination and implementation of the NICE guidelines as well as evaluation of subsequent practice patterns, healthcare costs, and patient centred outcomes. Development of effective and readily usable patient and provider tools are needed for targeting unique steps in the diagnosis and treatment pathway. NICE should create, use, and link understandable levels of strength of evidence, confidence in recommendation, and magnitude of risk and benefits with different options that incorporate patient and tumour characteristics as well as preferences. In the few instances where evidence is very strong and confidence high, performance standards and practice incentives might be introduced. Unfortunately, for most situations related to detection and treatment of prostate cancer, more definitive recommendations will require completion of randomised trials.
Competing interests: None declared.