British Journal of Urology (1991), 68, 263-265
01991 British Journal of Urology
Screening for Prostate Cancer. Comparison of Transrectal Ultrasound, Prostate Specific Antigen and Rectal Examination P. PERRIN, J. H. MAQUET, G.BRINGEON and M. DEVONEC Department of Urology, HGpital de I'Antiquaille, L yon, France
Summary-In seeking to define the relative value of transrectal ultrasound (TRUS), prostate specific antigen (PSA) and digital rectal examination (DRE) in the diagnosis of prostatic cancer, 863 patients were studied and the findings compared. DRE detected malignancy in 0.3% of the population of asymptomatic "normal" men undergoing routine health screening, and in 1.6% of patients who consulted their General Practitioner for one reason or another. In patients who attended our out-patient department with a variety of urological symptoms (not necessarily prostatic), TRUS suggested malignancy in 2% of those glands which were pronounced normal on DRE. Significantly elevated PSA detected malignancy in 0.3%of the patients undergoing routine health screening. (Although this figure equals the pick-up rate by DRE in this group, they were not necessarily the same patients). When these 3 investigations are summated, the pick-up rate is twice as high as when a single parameter is used.
The natural history of prostatic cancer is unknown. McNeal et al. (1966) suggested that the capacity of prostate cancer to metastasise was a function of its volume. Even if tumour volume correlates well with the Gleason grade, the degree of capsular or seminal vesicle penetration, or even the extent of metastases, tumour volume per se is not an indicator of likely progression. George (1989) showed that 16% of T2-T3 tumours did not progress during a 7-year follow-up, and only 13% metastasised. This lack of a predictor of the: biological behaviour of prostate cancer makes any screening programme irrelevant unless a randonnised study, including a control group of patients not subject to screening, has been carried out. Moreover, if we are unable to define the benefit of a screening programme for the population at risk (men over 50), the sensitivity of the most common and noninvasive diagnostic tests for this disease must be better defined. Accepted for publication 21 September 1990
Patients and Methods We have studied the 3 most common investigations : digital rectal examination, transrectal ultrasound and prostate specific antigen. The value of digital rectal examination Two separate studies were undertaken. The first group (group la) was studied over a 6month period and concerned a population of 863 asymptomatic men between 50 and 60 years of age who were attending for a routine quinquennial health check. The second study (group 1b) was conducted on a population of men consulting their General Practitioner for a multitude of symptoms, not necessarily urological. The purpose of the study was explained to them: only 70% (370 patients) agreed to submit to DRE. The value of transrectal ultrasound From October 1988 to April 1989,481 new patients attended our out-patient clinic. Their average age
263
264
BRITISH JOURNAL OF UROLOGY
was 67.2 years. All of these patients (group 2) underwent a rectal examination and a transrectal ultrasound scan of the prostate at the time of their consultation. The study was done prospectively. Those who had abnormalities either on DRE or on TRUS were submitted to ultrasound-guided transperineal prostatic biopsy.
Theplace of PSA PSA was evaluated as a screening procedure on a population of 863 men aged 50 to 60 years (group 3). Those with an elevated value were re-examined by a urologist and if the DRE was abnormal prostatic biopsy was carried out.
Table 3 Results of Digital Rectal Examination/Transrectal Ultrasound in 481 Patients
No. Benign benign
tumours
Positive predictive value (PPV)%
231 215 135 127 17 15 98 57 115 72
16 8 2 57 59
7 6 12 58 51
% DRE-/TRUSDRE-/TRUS+ DRE+/TRUSDRE+/TRUS+ DRE
+
93 94 88 58 62
No.of
malignancy was found in 3, giving a predictive value of 8% and an overall detection rate of 0.3% (3/863).
Results Table 1 shows that in the asymptomatic population the DRE was abnormal in 20% but the malignant pick-up rate was 0.34%. In group l b (Table 2), 59 of 320 DREs (16%) were abnormal and histological confirmation of malignancy was obtained in 6, giving a pick-up rate of 1.6%. Table 3 shows that an abnormal DRE (in the out-patient clinic population) had a predictive value of 51%, but when an abnormal DRE was corroborated by abnormal ultrasonography the predictive value rose to 58%. With regard to the value of PSA estimation, of the 38 patients with a raised value, biopsy proven Table 1 Prostate Findings in 863 Asymptomatic Patients Digital rectal examination
No.
%
Normal B.P.H.* Firm Nodular
697 105 37 24
80 12 4
3
No. of biopsies
No. of tumours
0 10 6 5
0 1**
0
2t
* B.P.H. = benign prostatic hyperplasia. Patients’ ages: ** 55 years; t 58 and 60 years. Table 2 Prostate Findings in 370 Patients Consulting their General Practitioner No. of
Digital rectal examination
NO.
%
tumours
Normal Suspicious
31 1 59
84 16
6*
Patients’ages: * 61,63,67,70,72, and 76 years.
Discussion From our first cohort (Group la), it would appear that the percentage of abnormal prostates on DRE in the population screened was 20% but the detection rate for prostatic cancer only 0.34%. Faul (1982) examined 1000 “normal” men over the age of 45 and found abnormal prostates on DRE in 16% but his detection rate for prostatic carcinoma was 2%. This difference in pick-up may be explained by the fact that in our series the upper age limit was 60. Thompson et al. (1984) had a detection rate of 0.55% in a population between 40 and 70 years old. Our low detection rate in this group (note the upper age limit of 60) implies that simple DRE as a screening technique is not justified. In group l b (Table 2), 59 rectal examinations were abnormal (16%) and 6 localised carcinomas were found on biopsy, giving a detection rate of 1.6%.Such a figure (1.3%)had already been reported by Lee et al. (1988) in a self-referred population of 784 men and a figure of 1.8% was reported by Chodack and Schornberg (unpublished data) in 21 39 patients undergoing routine screening. In all of these series the DRE was undertaken by General Practitioners as part of their routine physical examination. In Group 2 (DRE TRUS) the results show that when the DRE was abnormal, ultrasound guided biopsy was positive in 50%, giving a detection rate of 12%. By contrast, when the DRE was normal a hypoechoic lesion had a positive predictive value of 6% with a detection rate of 1.6%. This very low rate makes any massive screening programme based on ultrasonography an expensive waste of time.
+
265
SCREENING FOR PROSTATE CANCER
When both DRE and TRUS were positive, the predictive value of a hypoechoic lesion was 58%, a figure to be compared with 43.2% quoted by Cooner (unpublished data) and 62% quoted b y Lee et al. (1989) (Tables 4 and 5). In the present series there were 24 Tl-T2 tumours but only 8 of these were detected by ultrasound, i.e. two-thirds of these early carcinomas were normal according to DRE and TRUS. Table 6 shows that 4.4% of the men istudied had an abnormal PSA value. A similar figure was Table 4 Results of DRE and Ulstrasound Guided Biopsy According to Cooner (unpublished data)
+
DRE DRE Total
No. of patients
No. of patients with cancer
Positive predictite value ( P P V ) ;{
Detection rate %
565 1242 1807
203 60 263
43.2 16.4 31.5
35.9 4.8 14.6
Table 5 Results of DRE and Ultrasound Guided Biopsy According to Lee et al. (1989)
+
DRE DRE Total
Biopsy
Number of patients with cancer
Positive predictive value (PPV) %
114 104 256
71 33 104
62 32 40
Table 6 Serum PSA and Detection of Prostate Cancer
DRE
No. of patients
PSA>4
%
Normal B.P.H. Firm Nodular Total
697 105 37 24 863
23 9 3 2 38
3.2 8.5 8 8 4.4
Patients’ ages: * 58 years;
t 58 and 60 years.
No. of patients with cancer
1* 2t 3
reported by Myrtle (unpublished data). In this group only 3 patients were found to have carcinoma on biopsy and they all had abnormal findings on DRE. In other words, PSA plays no part in a screening programme. Digital rectal examination is currently as useful an examination as any in the detection of asymptomatic prostatic cancer. Transrectal ultrasonography should not be used as a screening tool but it does double the detection rate by picking up cancer in clinically normal glands. In the further elucidation of the clinically abnormal gland its value is unquestioned, but this is not the same as screening. As a screening tool PSA estimation alone is valueless.
Acknowledgement TheauthorsthankMr J. P. Williamsforrewritingthemanuscript and for his friendly support.
References Faul, P. (1982). Experience with German annual preventive check-up examination. In Prostate Cancer. International Perspectives in Urology, ed. Jacobi, G . H. and Hohenfellner, R. Volume 3, Pp. 55-67. Baltimore: Williams and Wilkins. George, N. J. R. (1989). Natural history of localised prostatic cancer managed by conservative therapy alone. Lancet, I, 494-491. Lee, F., Littrup, P. J., Torp-Pedersen,S. T. et al. (1988). Prostate cancer: comparison of transrectal US and digital rectal examination for screening. Radiology, 168,389-394. Lee, F., Torp-Pedersen, S. T., Littrup, P. J. et al. (1989). Hypoechoic lesions of the prostate :clinical relevance of tumor size, digital rectal examination, and prostate-specific antigen. Radiology, 110,29-32. McNeal, J. E., Kindrachuk, R. A., Freiha, F. S. et al. (1986). Patterns of progression in prostate cancer. Lancet, II,60-63. Thompson, I. M., Emst, J. J., Gangai, M. P. etal. (1984). Adenocarcinoma of the prostate: results of routine urological screening. J . Urol., 132,690-692.
The Authors P. Perrin, MD, Professor of Urology. J. H. Maquet, MD, Senior Resident. G. Bringeon, MD, Senior Resident. M. Devonec, MD, Assistant Professor of Urology. Requests for reprints to: P. Perrin, Service d’urologie, HBpital de I’Antiquaille, 1 rue de I’Antiquaille, 69321 Lyon Cedex 05, France.
01991 British Journal of Urology
Screening for Prostate Cancer. Comparison of Transrectal Ultrasound, Prostate Specific Antigen and Rectal Examination P. PERRIN, J. H. MAQUET, G.BRINGEON and M. DEVONEC Department of Urology, HGpital de I'Antiquaille, L yon, France
Summary-In seeking to define the relative value of transrectal ultrasound (TRUS), prostate specific antigen (PSA) and digital rectal examination (DRE) in the diagnosis of prostatic cancer, 863 patients were studied and the findings compared. DRE detected malignancy in 0.3% of the population of asymptomatic "normal" men undergoing routine health screening, and in 1.6% of patients who consulted their General Practitioner for one reason or another. In patients who attended our out-patient department with a variety of urological symptoms (not necessarily prostatic), TRUS suggested malignancy in 2% of those glands which were pronounced normal on DRE. Significantly elevated PSA detected malignancy in 0.3%of the patients undergoing routine health screening. (Although this figure equals the pick-up rate by DRE in this group, they were not necessarily the same patients). When these 3 investigations are summated, the pick-up rate is twice as high as when a single parameter is used.
The natural history of prostatic cancer is unknown. McNeal et al. (1966) suggested that the capacity of prostate cancer to metastasise was a function of its volume. Even if tumour volume correlates well with the Gleason grade, the degree of capsular or seminal vesicle penetration, or even the extent of metastases, tumour volume per se is not an indicator of likely progression. George (1989) showed that 16% of T2-T3 tumours did not progress during a 7-year follow-up, and only 13% metastasised. This lack of a predictor of the: biological behaviour of prostate cancer makes any screening programme irrelevant unless a randonnised study, including a control group of patients not subject to screening, has been carried out. Moreover, if we are unable to define the benefit of a screening programme for the population at risk (men over 50), the sensitivity of the most common and noninvasive diagnostic tests for this disease must be better defined. Accepted for publication 21 September 1990
Patients and Methods We have studied the 3 most common investigations : digital rectal examination, transrectal ultrasound and prostate specific antigen. The value of digital rectal examination Two separate studies were undertaken. The first group (group la) was studied over a 6month period and concerned a population of 863 asymptomatic men between 50 and 60 years of age who were attending for a routine quinquennial health check. The second study (group 1b) was conducted on a population of men consulting their General Practitioner for a multitude of symptoms, not necessarily urological. The purpose of the study was explained to them: only 70% (370 patients) agreed to submit to DRE. The value of transrectal ultrasound From October 1988 to April 1989,481 new patients attended our out-patient clinic. Their average age
263
264
BRITISH JOURNAL OF UROLOGY
was 67.2 years. All of these patients (group 2) underwent a rectal examination and a transrectal ultrasound scan of the prostate at the time of their consultation. The study was done prospectively. Those who had abnormalities either on DRE or on TRUS were submitted to ultrasound-guided transperineal prostatic biopsy.
Theplace of PSA PSA was evaluated as a screening procedure on a population of 863 men aged 50 to 60 years (group 3). Those with an elevated value were re-examined by a urologist and if the DRE was abnormal prostatic biopsy was carried out.
Table 3 Results of Digital Rectal Examination/Transrectal Ultrasound in 481 Patients
No. Benign benign
tumours
Positive predictive value (PPV)%
231 215 135 127 17 15 98 57 115 72
16 8 2 57 59
7 6 12 58 51
% DRE-/TRUSDRE-/TRUS+ DRE+/TRUSDRE+/TRUS+ DRE
+
93 94 88 58 62
No.of
malignancy was found in 3, giving a predictive value of 8% and an overall detection rate of 0.3% (3/863).
Results Table 1 shows that in the asymptomatic population the DRE was abnormal in 20% but the malignant pick-up rate was 0.34%. In group l b (Table 2), 59 of 320 DREs (16%) were abnormal and histological confirmation of malignancy was obtained in 6, giving a pick-up rate of 1.6%. Table 3 shows that an abnormal DRE (in the out-patient clinic population) had a predictive value of 51%, but when an abnormal DRE was corroborated by abnormal ultrasonography the predictive value rose to 58%. With regard to the value of PSA estimation, of the 38 patients with a raised value, biopsy proven Table 1 Prostate Findings in 863 Asymptomatic Patients Digital rectal examination
No.
%
Normal B.P.H.* Firm Nodular
697 105 37 24
80 12 4
3
No. of biopsies
No. of tumours
0 10 6 5
0 1**
0
2t
* B.P.H. = benign prostatic hyperplasia. Patients’ ages: ** 55 years; t 58 and 60 years. Table 2 Prostate Findings in 370 Patients Consulting their General Practitioner No. of
Digital rectal examination
NO.
%
tumours
Normal Suspicious
31 1 59
84 16
6*
Patients’ages: * 61,63,67,70,72, and 76 years.
Discussion From our first cohort (Group la), it would appear that the percentage of abnormal prostates on DRE in the population screened was 20% but the detection rate for prostatic cancer only 0.34%. Faul (1982) examined 1000 “normal” men over the age of 45 and found abnormal prostates on DRE in 16% but his detection rate for prostatic carcinoma was 2%. This difference in pick-up may be explained by the fact that in our series the upper age limit was 60. Thompson et al. (1984) had a detection rate of 0.55% in a population between 40 and 70 years old. Our low detection rate in this group (note the upper age limit of 60) implies that simple DRE as a screening technique is not justified. In group l b (Table 2), 59 rectal examinations were abnormal (16%) and 6 localised carcinomas were found on biopsy, giving a detection rate of 1.6%.Such a figure (1.3%)had already been reported by Lee et al. (1988) in a self-referred population of 784 men and a figure of 1.8% was reported by Chodack and Schornberg (unpublished data) in 21 39 patients undergoing routine screening. In all of these series the DRE was undertaken by General Practitioners as part of their routine physical examination. In Group 2 (DRE TRUS) the results show that when the DRE was abnormal, ultrasound guided biopsy was positive in 50%, giving a detection rate of 12%. By contrast, when the DRE was normal a hypoechoic lesion had a positive predictive value of 6% with a detection rate of 1.6%. This very low rate makes any massive screening programme based on ultrasonography an expensive waste of time.
+
265
SCREENING FOR PROSTATE CANCER
When both DRE and TRUS were positive, the predictive value of a hypoechoic lesion was 58%, a figure to be compared with 43.2% quoted by Cooner (unpublished data) and 62% quoted b y Lee et al. (1989) (Tables 4 and 5). In the present series there were 24 Tl-T2 tumours but only 8 of these were detected by ultrasound, i.e. two-thirds of these early carcinomas were normal according to DRE and TRUS. Table 6 shows that 4.4% of the men istudied had an abnormal PSA value. A similar figure was Table 4 Results of DRE and Ulstrasound Guided Biopsy According to Cooner (unpublished data)
+
DRE DRE Total
No. of patients
No. of patients with cancer
Positive predictite value ( P P V ) ;{
Detection rate %
565 1242 1807
203 60 263
43.2 16.4 31.5
35.9 4.8 14.6
Table 5 Results of DRE and Ultrasound Guided Biopsy According to Lee et al. (1989)
+
DRE DRE Total
Biopsy
Number of patients with cancer
Positive predictive value (PPV) %
114 104 256
71 33 104
62 32 40
Table 6 Serum PSA and Detection of Prostate Cancer
DRE
No. of patients
PSA>4
%
Normal B.P.H. Firm Nodular Total
697 105 37 24 863
23 9 3 2 38
3.2 8.5 8 8 4.4
Patients’ ages: * 58 years;
t 58 and 60 years.
No. of patients with cancer
1* 2t 3
reported by Myrtle (unpublished data). In this group only 3 patients were found to have carcinoma on biopsy and they all had abnormal findings on DRE. In other words, PSA plays no part in a screening programme. Digital rectal examination is currently as useful an examination as any in the detection of asymptomatic prostatic cancer. Transrectal ultrasonography should not be used as a screening tool but it does double the detection rate by picking up cancer in clinically normal glands. In the further elucidation of the clinically abnormal gland its value is unquestioned, but this is not the same as screening. As a screening tool PSA estimation alone is valueless.
Acknowledgement TheauthorsthankMr J. P. Williamsforrewritingthemanuscript and for his friendly support.
References Faul, P. (1982). Experience with German annual preventive check-up examination. In Prostate Cancer. International Perspectives in Urology, ed. Jacobi, G . H. and Hohenfellner, R. Volume 3, Pp. 55-67. Baltimore: Williams and Wilkins. George, N. J. R. (1989). Natural history of localised prostatic cancer managed by conservative therapy alone. Lancet, I, 494-491. Lee, F., Littrup, P. J., Torp-Pedersen,S. T. et al. (1988). Prostate cancer: comparison of transrectal US and digital rectal examination for screening. Radiology, 168,389-394. Lee, F., Torp-Pedersen, S. T., Littrup, P. J. et al. (1989). Hypoechoic lesions of the prostate :clinical relevance of tumor size, digital rectal examination, and prostate-specific antigen. Radiology, 110,29-32. McNeal, J. E., Kindrachuk, R. A., Freiha, F. S. et al. (1986). Patterns of progression in prostate cancer. Lancet, II,60-63. Thompson, I. M., Emst, J. J., Gangai, M. P. etal. (1984). Adenocarcinoma of the prostate: results of routine urological screening. J . Urol., 132,690-692.
The Authors P. Perrin, MD, Professor of Urology. J. H. Maquet, MD, Senior Resident. G. Bringeon, MD, Senior Resident. M. Devonec, MD, Assistant Professor of Urology. Requests for reprints to: P. Perrin, Service d’urologie, HBpital de I’Antiquaille, 1 rue de I’Antiquaille, 69321 Lyon Cedex 05, France.
