Objective: To examine critically the clinical presentation, pathological stage and outcome in patients selected for radical prostatectomy combining data from three centres where the operation has been carried out routinely for more than 5 years. Comparisons were made between impalpable tumours presenting at transurethral resection of the prostate (TURP) for clinically benign disease, tumours diagnosed at needle biopsy performed because the serum prostate-specific antigen (PSA) was elevated, and palpable, clinically localized cancer detected by digital rectal examination (DRE).
Patients and methods: Clinical and pathological findings recorded in the hospital notes of 183 patients who had undergone exploration for radical prostatectomy at St Bartholomew's Hospital, London, Southmead Hospital, Bristol, and the Royal Infirmary, Stirling, between 1987 and 1994 were transcribed onto a proforma and analysed. Patients were categorized by clinical stage and the relationships between clinical presentation, serum PSA level, pathological stage, tumour grade and outcome were examined.
Results: The pathological extent of clinically unsuspected cancer identified at TURP was highly variable. Well-differentiated tumours occupying < 5% of the TURP specimen were generally found to be less extensive at subsequent radical prostatectomy than either impalpable malignancy diagnosed by needle biopsy performed because PSA levels were raised or palpable tumours associated with a unilateral abnormality on DRE. Unsuspected tumours diagnosed at TURP that were less than well differentiated or occupied > 5% of the surgical specimen were more commonly associated with extra-prostatic invasion or metastatic disease than were tumours detected by raised PSA levels or an abnormal DRE. Serum PSA did not reliably predict either clinical stage or pathological extent, but no patient with nodal metastases had a PSA level of < 12 ng/mL. Similarly, pathological stage could not be predicted confidently from the whole-tumour grade ascertained after surgery. The extent of malignancy in patients referred because of lower urinary tract symptoms was not significantly different from that among patients referred specifically because the PSA level was raised or the DRE was abnormal. However, there was a trend for patients found to have cancer in specific screening programmes to have a malignancy that was less extensive and therefore more frequently confined to the gland than in those not identified by screening.
Conclusion: Among patients with unsuspected malignancy diagnosed at TURP, those with well differentiated tumours in < 5% of the specimen had significantly less advanced disease than men found to have less differentiated or more extensive malignancy in the resected specimen. Impalpable cancer associated with a raised PSA level diagnosed by needle biopsy represented an intermediate group comparable to those with unilateral palpable malignancy. This suggests that needle biopsies may be worthwhile in some patients with an apparently benign prostate and raised PSA level with a view to identifying clinically significant but potentially curable cancer. Screening may detect less extensive tumours, but the natural history of these cancers is unknown and observed survival will be subject to lead-time and length bias when compared to symptomatic disease. Prospective randomized studies of both screening and treatment modalities, including surgery and radiotherapy, are required to define the impact of radical prostatectomy on disease-specific survival in men with early stage prostate cancer.