PROSTATE CANCER SCREENING WITHIN A PROSTATE SPECIFIC ANTIGEN RANGE OF 3 TO 3.9 NG./ML.: A COMPARISON OF DIGITAL RECTAL EXAMINATION AND FREE PROSTATE SPECIFIC ANTIGEN AS SUPPLEMENTAL SCREENING TESTS T. MA ¨ KINEN, T. L. J. TAMMELA, M. HAKAMA,* U.-H. STENMAN, S. RANNIKKO,* J. ARO,* H. JUUSELA,* L. MA ¨ A ¨ TTA ¨ NEN* AND A. AUVINEN* From the Finnish Cancer Registry, Departments of Clinical Chemistry and Urology, Helsinki University Central Hospital and Department of Surgery, Helsinki City Hospital, Helsinki, Department of Urology, Tampere University Hospital and Medical School and Tampere School of Public Health, University of Tampere, Tampere and Department of Surgery, Jorvi Hospital, Espoo, Finland ABSTRACT Purpose: Performing biopsy in all men with a serum prostate specific antigen (PSA) of 3 to 3.9 ng./ml. increases the sensitivity of prostate cancer screening compared with a PSA cutoff of 4 ng./ml. but decreases specificity and may contribute to over diagnosis. Therefore, we evaluated the detection rate and specificity attributable to digital rectal examination and percent free PSA within the PSA range of 3 to 3.9 ng./ml. Materials and Methods: Serum PSA was determined in 20,716 participants in the Finnish population based screening trial. Supplementary digital rectal examination was offered to men with a PSA of 3 to 3.9 ng./ml. during 1996 to 1998 (protocol 1). Those with a suspicious digital rectal examination finding were referred for biopsy. The screening algorithm was modified by substituting percent free PSA for digital rectal examination with a cutoff of 16% as a biopsy criterion in 1999 (protocol 2). In addition, biopsies were performed in all men with PSA 4 ng./ml. or greater. Results: A total of 23 cancers (2.9%) were detected by digital rectal examination among 801 men, while percent-free PSA resulted in the diagnosis of 13 cases (4.8%) among 270 men with a PSA of 3 to 3.9 ng./ml. The detection rate of tumors with a Gleason score of 5 or greater increased from 1.6% (13 of 801 cases) to 4.4% (12 of 270) in the modified screening program. The PSA cutoff of 3 ng./ml. alone showed 88.6% and 87.5% specificity in protocols 1 and 2 but specificity increased to 93.3% and 91.7% using digital rectal examination and percent free PSA, respec- tively. Conclusions: Using percent free PSA increased the detection rate of aggressive disease com- pared with digital rectal examination and provided higher specificity than PSA alone. KEY WORDS: prostate, prostatic neoplasms, prostate-specific antigen, mass screening, sensitivity and specificity Serum prostate specific antigen (PSA) testing is widely done for the early detection of prostate cancer. At a PSA cutoff of 4 ng./ml. prostate cancer is detected in 25% to 30% of men with positive screening results. 1 A fairly high detec- tion rate of 13% to 22% has been reported at a PSA of 3 to 3.9 ng./ml. but lowering the cutoff to 3 ng./ml. significantly in- creases the number of screening positive cases. 2, 3 A PSA cutoff of 3 ng./ml. exposes more men to the negative effects of screening, such as psychological distress caused by false-positive test results and over diagnosis, than screening with higher PSA cutoffs. This aggressive strategy also in- creases screening costs, which may be a major obstacle for screening as a public health policy. It is also unclear whether cancer diagnosed at PSA 3 to 3.9 ng./ml. would be detectable at a localized stage during subsequent screening. Further- more, the effectiveness of prostate cancer screening, namely mortality reduction, remains to be established. Thus, a more selective screening method that improves the specificity of screening without compromising sensitivity is urgently needed. Various strategies based on supplementary tests have been used to decrease the frequency of false-positive tests with lower cutoffs. 4–7 Our study was based on the Finnish prostate cancer screening trial, which is the largest component in the Euro- pean Randomized Study of Screening for Prostate Cancer. We compared the performance of digital rectal examination and percent free PSA at a PSA of 3 to 3.9 ng./ml. for prostate cancer screening. The main outcome measures reported are the prostate cancer detection rate, screening program speci- ficity and the histological characteristics of detected tumors. MATERIAL AND METHODS Subjects. The Finnish prostate cancer screening trial was initiated in May 1996. Our study was based on prevalence screening done in 1996 to 1999. A total of 59,973 men who were 55 to 67 years old were enrolled from the Population Register of Finland during years 1 to 3 up to the end of 1998 (protocol 1). Men diagnosed with prostate cancer before ran- domization were excluded from study. A total of 24,000 men were randomly allocated to the screening arm and the re- maining 35,973 comprised the control arm, which was not contacted. Those who had died, moved outside of the study area by the time of invitation or refused the use of their addresses for any purpose were also excluded from analysis. Accepted for publication May 18, 2001. Supported by Academy of Finland, Cancer Society of Pirkanmaa, Cancer Society of Finland, Helsingin Sanomat Centenarian Fund, Tampere University Hospital Research Fund, Europe Against Can- cer program and Beckman-Hybritech Corp. * Financial interest and/or other relationship with Hybritech Corp. 0022-5347/01/1664-1339/0 THE JOURNAL OF UROLOGY ® Vol. 166, 1339 –1342, October 2001 Copyright © 2001 by AMERICAN UROLOGICAL ASSOCIATION,INC. ® Printed in U.S.A. 1339