Evidence based medicine suggests that evidence of effectiveness should accumulate, preferably from randomised controlled trials, before treatments for any condition become widely used. The case of localised prostate cancer shows how difficult this can be in practice. The suitability of population screening for localised prostate cancer has been debated,1,2 with particular concerns about the comparative effectiveness of the main treatments for the disease: radical prostatectomy, radical radiotherapy, and conservative management (also known as watchful waiting or surveillance).3,4 Systematic reviews show that published evidence is limited to two seriously flawed randomised controlled trials and a range of observational studies with biases relating to patient selection, variable treatment techniques, outcome assessments, and methods of data analysis.3 These studies show that 10 year survival is good and overlaps for the three treatments, being 85-90% for radical prostatectomy, 65-90% for radical radiotherapy, and 70-90% for conservative management.3 Although some studies indicate a survival advantage of radical treatments in some patients, this advantage is small and uncertain given the particular study designs. Furthermore, quality of life may be worse among those receiving radical treatments because of resulting complications.5 For example, after radical prostatectomy up to 3% of patients may be totally incontinent, with up to 60% “dribbling” urine, and 20-80% impotent, while after radical radiotherapy up to 36% may have damage to adjacent organs, 10% incontinence, and 40% impotence.3 Morbidity from conservative management relates to symptoms (and hormonal treatment if required) if the disease progresses.
Subjects, methods, and results
A postal questionnaire survey of practising consultant urologists registered with the British Association of Urological Surgeons was conducted exploring their treatment preferences for various clinical case vignettes. General surgeons, trainees, and those with paediatric caseloads were excluded. A total of 244 consultant urologists replied (response rate 60%). Urologists had a mean of 14.1 years’ experience (range 2-30 years), and 130 of them managed 100 patients or more with prostate cancer. Expertise in performing radical prostatectomy was restricted to comparatively few urologists—98 reported having ever performed the procedure and only 12 (14%) that they performed 20 or more operations per year. The table shows that radical treatments were the first choice treatment for all hypothetical patients with apparently localised disease under the age of 70 years (cases 1, 2, 3, and 5 in table), irrespective of mode of presentation, prostate specific antigen concentration, and grade of tumour. Radical prostatectomy was preferred for the man of 55 (case 1), radical radiotherapy for the man of 69 (case 3). Conservative management was first choice for the majority of urologists for the man of 75 (case 4). For the youngest men (cases 1 and 5) only a few urologists selected conservative management, rising to 27% for the man aged 69.
Comment
Although reliable evidence supporting radical treatments is scarce, British urologists seem to favour them for all patients under 70. These findings may be open to misinterpretation, representing what urologists say they do rather than what they do, although hospital episode statistics confirm that numbers of radical prostatectomies have doubled nearly every year between 1990-1 and 1994-5 in the United Kingdom (hospital episode statistics, 1989-90 to 1994-5).
Clearly, evidence is needed from randomised controlled trials, but such studies have proved difficult because of perceptions that patients are reluctant to accept conservative management. New methodological approaches are required urgently to investigate this issue and to bridge the gap between clinical practice and the need to acquire evidence. Such approaches need to retain the essential principle of randomisation while incorporating more fully patients’ perspectives and preferences. Without this, the increasing availability of radical treatments, rising rate of detection of localised prostate cancer in younger men, the concerns of men about harbouring an untreated malignancy, and the desire of clinicians to cure patients if at all possible will combine to ensure that the situation cannot resolve spontaneously. Trials undoubtedly need to be mounted, and until more evidence accumulates, patients and urologists should use the information available from recent systematic reviews3,4 to reach shared decisions about treating localised prostate cancer—information that highlights uncertainties about the potential effects of such treatments on survival and quality of life.
Table.
First choice of treatment of consultant urologists for clinical vignettes. Values are numbers (percentages) of consultants
| Case No | Presenting characteristics* | No of consultants responding | Radical prostatectomy | Radical radiotherapy | Brachytherapy | Hormone (medical) | Hormone (surgical) | Surveillance only | Surveillance and hormone |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Age 55, incidental, prostate specific antigen 7 ng/ml, transrectal ultrasound and magnetic resonance imaging results suggest localised tumour, Gleason score 3 | 211 | 136 (64) | 49 (23) | 1 (0.5) | 3 (1) | 0 | 16 (8) | 6 (3) |
| 2 | Age 62, bladder outlet obstruction, prostate specific antigen 15 ng/ml, Gleason score 5 from chippings | 202 | 80 (40) | 82 (41) | 0 | 7 (3) | 3 (1) | 10 (5) | 20 (10) |
| 3 | Age 69, incidental, prostate specific antigen 16 ng/ml, transrectal ultrasound and magnetic resonance imaging results suggest localised tumour, Gleason score 3 | 206 | 50 (24) | 85 (41) | 2 (1) | 13 (6) | 0 | 27 (13) | 29 (14) |
| 4 | Age 75, incidental, prostate specific antigen 20 ng/ml, transrectal ultrasound and magnetic resonance imaging results suggest localised tumour, Gleason score 3 | 218 | 0 | 51 (23) | 0 | 22 (10) | 1 (0.5) | 52 (24) | 92 (42) |
| 5 | Age 58, poorly differentiated localised tumour, prostate specific antigen 17 ng/ml, Gleason score 8 | 202 | 67 (33) | 102 (50) | 2 (1) | 23 (11) | 1 (0.5) | 1 (0.5) | 6 (3) |
Normal values of prostate specific antigen are around 4 ng/ml. In Gleason grading system sections of tumour are graded from 1 (least aggressive) to 5 (most aggressive). The two highest grades from each tumour are added to give a score ranging from 2-10. Scores of 7 and above indicate worse prognosis than lower scores.
Acknowledgments
The list of urological surgeons was kindly provided by the British Association of Urological Surgeons. The views expressed in this paper are those of the authors and not necessarily those of the Standing Group on Health Technology, the HTA Commissioning Board, the HTA Panel members, or the Department of Health.
Editorial by Emberton
Footnotes
Funding: The systematic review and questionnaire survey were supported by a grant from the NHS health technology assessment programme.
Conflict of interest: None.
References
- 1.Stewart-Brown S, Farmer A. Screening could seriously damage your health. BMJ. 1997;314:533. doi: 10.1136/bmj.314.7080.533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Woolf SH. Should we screen for prostate cancer? BMJ. 1997;314:989–990. doi: 10.1136/bmj.314.7086.989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Selley S, Donovan JL, Faulkner A, Coast J, Gillatt D. Diagnosis, management and screening of early localised prostate cancer. Health Technol Assess 1997;1(2). [PubMed]
- 4.Chamberlain J, Melia J, Moss S, Brown J. The diagnosis, management, treatment and costs of prostate cancer in England and Wales. Health Technol Assess 1997;1(3). [PubMed]
- 5.Litwin MS, Hays RD, Fink A, Ganz PA. Quality of Life outcomes in men treated for localized prostate cancer. JAMA. 1995;273:129–135. doi: 10.1001/jama.273.2.129. [DOI] [PubMed] [Google Scholar]