Objective: To characterize prostate cancer in men undergoing radical prostatectomy who have a prostate-specific antigen (PSA) level of < 4.0 ng/mL, hypothesizing that a low PSA is not caused by diminished tumour production of PSA, nor does it signify clinically insignificant disease.
Patients and methods: Seventy-nine men (mean age 59.3 years, range 43-77) with a PSA level of < 4.0 ng/mL were identified from 702 who had a radical prostatectomy between 1994 and 2000. Demographic and clinical data were analysed; pathological specimens were evaluated by routine haematoxylin and eosin staining and immunohistochemistry with anti-PSA antibody, for both pathological staging and grading, and for the presence of PSA production. Tumours were classified as 'clinically insignificant' if the tumour volume was < 0.5 mL and the Gleason score < 7.
Results: The mean (SD, range) preoperative PSA level was 3.04 (0.85, 0.8-3.8) ng/mL Indications for biopsy included an abnormal digital rectal examination (61%), a PSA velocity of > 0.75 ng/mL/year (12%), a strong family history of prostate cancer (3%), obstructive urinary symptoms (2%), or no obvious indication (23%). Thirty-eight (48%) tumours were clinically insignificant. Of 41 clinically significant cancers, 13 had a final Gleason score of > or = 7, 20 had extraprostatic extension and 11 had a tumour volume of > or = 10 mL Of the 79 prostate cancer specimens 78 stained strongly for PSA; the exception was a Gleason 9 tumour. With a mean (range) follow-up of 3.5 (0.18-6) years only one patient had a biochemical recurrence (PSA > or = 0.1 ng/mL).
Conclusions: Most prostate cancers in men with a PSA level of < 4.0 ng/mL are clinically significant and PSA-producing. Many of these tumours are high-grade, high-volume and extraprostatic. We are currently exploring factors to explain why serum PSA is not elevated in these men, including tumour location, pattern of invasion and microvessel density.