Table of Contents:
Prostate cancer is the most common malignant neoplasm among men in Germany. Early detection of the disease it the key point for a curative therapeutic approach. The screening for prostate cancer with Prostate-Specific Antigen (PSA) plays an important role for early detection of prostate cancer.
However, because of relatively low specifity and sensitivity of PSA there are limitations in screening for prostate cancer using this method. Especially in slightly elevated PSA levels (> 4 ng per milliliter and < 10 ng per milliliter) false positive results due to presence of benign prostatic diseases are common. Some new techniques were developed to improve the accuracy of PSA, e.g. PSA density (the PSA concentration divided by the volume of the gland).
Since the first description of free PSA by Lilja and Stenman 1991 many studies have shown clear evidence that the ratio of free to total PSA in patients with prostate malignancy is lower than in patients with benign prostatic diseases. On the other hand the appropriate cut off level for the ratio of free to total PSA and the PSA range in which the ratio gives the most diagnostic benefit is remaining still unclear.
The large number of available test methods and test kits for PSA screening makes this situation even more complicate.
The departments of urology and clinical chemistry of the Philipps-University Marburg use the PSA immulite third generation and freies PSA Immulite (DPC Biermann, Bad Nauheim, Germany) kits. The manufacturer provides no reference values for the test kits.
The aims of our study were: First, the determination of an optimal cut off point for the ratio of free to total PSA for the used test kits and second, the verification if the ratio of free to total PSA improves specifity and sensitivity in the diagnostic of prostate cancer in patients with slightly elevated PSA levels (> 4 ng per milliliter and < 10 ng per milliliter).
The study duration was one year (may 1997- may 1998). 454 consecutive male patients from the department of urology were included into the study. PSA and free PSA were measured and the ratio of free to total PSA was calculated. For 379 patients which received a volume measurement of the prostatic gland PSA density was also calculated. From these patients 39 had a diagnosis of prostate cancer, 230 a benign prostate hypertrophy. In 36 patients a chronic prostatitis was diagnosed, in 10 patients an acute prostatitis. An inconspicuously prostate was found in 64 men.
For PSA best specifity (76.9%) and sensitivity (76.8%) for detection of prostate cancer was reached with a cut off level of 4 ng per milliliter.
Ratio of free to total PSA in patients with slightly elevated PSA levels (> 4 ng per milliliter and < 10 ng per milliliter) showed an optimal cut off point at 20%. Sensitivity was 91.5%, but specifity only 40%. The area under the ROC (Receiver Operation Characteristic) curve was not significantly different compared to the PSA ROC curve. Measurement of free PSA and calculation of free to total PSA therefore can not be recommended for our patient group.
The area under the ROC curve of PSA density compared to PSA was significantly different (p < 0.05) with advantage for PSA density. Because of the costly volume measurement of the prostatic gland PSA density is not an appropriate method for screening for prostate cancer.