- Email: [email protected]
CONTEMPORARY IMPACT OF TRANSRECTAL ULTRASOUND LESIONS FOR PROSTATE CANCER DETECTION
0022-5347/04/1722-0512/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 172, 512–514, August 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000131621.61732.6b
CONTEMPORARY IMPACT OF TRANSRECTAL ULTRASOUND LESIONS FOR PROSTATE CANCER DETECTION RAHMI ONUR, PETER J. LITTRUP, J. EDSON PONTES
AND
FERNANDO J. BIANCO, JR.*
From the Departments of Urology and Radiology (PL), Wayne State University School of Medicine and The Prostate Program of The Barbara Ann Karmanos Cancer Institute, Detroit, Michigan
ABSTRACT
Purpose: Transrectal ultrasound (TRUS) guided systematic biopsy of the prostate is the gold standard diagnostic modality for prostate cancer. Consequently, the value of discrete hypoechoic lesions on TRUS lesions considered suspicious for cancer deserves meticulous reevaluation, specifically in the prostate specific antigen era when the majority of tumors diagnosed are nonpalpable. We studied whether the predictability of a biopsy core changes if the tissue comes from an isoechoic vs hypoechoic lesion. Materials and Methods: Prospective data were collected on 3,912 consecutive patients referred to our medical center between 1993 and 1999 for biopsy of the prostate. A sextant technique (apex, mid gland and base) with an additional core biopsy from the transitional zone was used. If a hypoechoic lesion was identified, the biopsy was taken from the lesion. Correlation between hypoechoic lesions, isoechoic areas and cancer detection for each core was performed. Results: A total of 31,296 cores were obtained from the cohort. Overall 2,642 (68%) cores had at least 1 hypoechoic lesion ultrasonographically. Cancer was detected in 675 (25.5%) and 323 (25.4%) patients with or without hypoechoic lesions (p ⫽ 0.97). The per core cancer detection was fairly uniform and averaged 9.3% and 10.4% for hypoechoic and isoechoic areas, respectively. The difference was not statistically significant (p ⫽ 0.3). Gleason scores were less than 7, 7 and greater than 7 in 46%, 34% and 20% of cases, respectively. Conclusions: Despite the higher prevalence of cancers discovered in prostates with hypoechoic areas, the hypoechoic lesion itself was not associated with increased cancer prevalence compared with biopsy cores from isoechoic areas. For impalpable tumors TRUS findings are not contributory for staging. KEY WORDS: prostate-specific antigen, ultrasonography, biopsy, prostate neoplasms
Transrectal ultrasound (TRUS) guided systematic biopsy of the prostate is the gold standard diagnostic modality for prostate cancer.1, 2 Sonographic appearance of prostate cancer (CaP) was originally described as a hyperechoic lesion.2 Nevertheless, with the advent of high frequency transducers and improved instrumentation, CaP was recognized as appearing hypoechoic, hyperechoic, isoechoic or mixed on ultrasonography.3, 4 In the last 2 decades smaller cancers were considered to show less echogenicity on ultrasonography (ie hypoechoic) compared with larger tumors and, thus, TRUS had been used to screen for early nonpalpable cancer.5, 6 However, widespread use of prostate specific antigen (PSA) has led to an unparallel shift in stage and volume.7, 8 Consequently, the value of discrete hypoechoic lesions on transrectal ultrasound considered highly suspicious for cancer deserves meticulous reevaluation, specifically when the majority of tumors diagnosed are nonpalpable.9 PSA levels may increase the positive predictive value of hypoechoic lesions.10 However, not all hypoechoic regions represent cancer and, conversely, not all cancers are hypoechoic, with as many as 50% of clinically significant prostate cancers not being purely hypoechoic and 37% of all diagnosed cancers in a series being isoechoic.9 Recently, the clinical significance of ultrasonographic distinctions for CaP has come under scrutiny.11 However, studies assessing the exact prediction of a biopsy core, whether it comes from the hypo-
echoic lesion or not, are lacking. We studied the predictability of a biopsy core from each side of the prostate at the base, mid gland, apex and transitional zone (TZ) to determine the contemporary predictive value of a hypoechoic lesion identified in those areas. MATERIALS AND METHODS
Prospective data were collected on 3,912 consecutive patients referred to our medical center between January 1993 and December 1999 for initial TRUS biopsy of the prostate. Biopsy indications included abnormal digital rectal examination, increased total PSA level or velocity, low percent free PSA and increased PSA density. Digital rectal examination (DRE) was regarded as abnormal if a nodule or hard consistency was palpated. Patients were medicated with a fluoroquinolone for 3 days starting the day before the procedure. Men at risk for heart valve disease or history of valve surgery received additional amoxicillin or erythromycin. A cleansing enema was administered the morning of the biopsy. After informed consent was obtained from each patient, DRE was performed followed by TRUS. The gland was carefully examined to ensure that all areas of the gland were optimally evaluated. The sagittal and axial planes were examined. Echogenic patterns were characterized by location (base, mid gland or apex) and side, and prostatic volume was measured. Biopsies were performed bilaterally in the axial projection using the sextant technique with an additional core from the TZ. Every biopsy core from the base, mid gland, apex and TZ on the right and left sides was placed into a container and labeled accordingly. DRE, serum PSA, area of
Accepted for publication March 19, 2004. * Correspondence: Department of Urology, Memorial SloanKettering Cancer Center, 1275 York Ave., New York, New York 10021 (telephone: 646-523-1051; FAX: 212-988-0683; e-mail: [email protected]). 512
TRANSRECTAL ULTRASOUND BIOPSY CORE POSITIVE PREDICTIVE VALUE
513
the prostate and isoechoic or hypoechoic origin of each core were meticulously recorded into a relational database. Every positive core was graded using the Gleason system and a score was assigned. Chi-square tests were used to assess the statistical comparison of these nonparametric variables. The preoperative levels were stratified into 5 categories, and the correlation with presence of cancer and prostate echogenicity was also assessed. RESULTS
A total of 31,296 biopsy cores were obtained from 3,912 patients. Median patient age and PSA were 67 years (range 37 to 95) and 6.6 ng/ml (range 0.9 to 51). Overall, cancer was detected in 998 (25.5%) patients. At least 1 hypoechoic lesion was seen in 2,642 (68%) patients. Exclusive lesions in the right or left lobes were detected in 722 (19%) and 702 (18%) patients, respectively, and 1,218 (31%) had hypoechoic lesions in both lobes. However, even if bilateral hypoechogenic lesions were seen, cancer detection was not different than for those cases with no hypoechogenicity (312 of 1,218 [25.6%] vs 232 of 1,270 [25.4%], p ⫽ 0.9). A detailed description showing the cancer detection rates for each core by side and prostate zone is provided in the table. Hypoechoic lesions at the prostate base harbored cancer in 9.3% of all cores, whereas tumors were detected in isoechoic areas 10.4% of the time, which was not statistically different (p ⫽ 0.3). This pattern was similar in the other prostatic areas as biopsies directed to hypoechoic areas were not able to provide better detection rates compared with random cores of the gland. The specific probability of finding cancer in a hypoechoic area within the right or left lobe was 18% (131 of 722) and 19% (134 of 702) of patients, respectively (p ⫽ 0.7), which was similar to yields from isoechoic area (fig. 1). The overall likelihood of detecting cancer in patients with or without hypoechoic lesions in the gland was similar (25.5% [675 of 2,642] vs 25.4% [323 of 1,270], respectively, p ⫽ 0.8). Figure 2 shows the cancer detection rates stratified by PSA ranges. Although there was a positive correlation between PSA level and cancer detection rate, examination of hypoechoic and isoechoic biopsies revealed no significant difference in sensitivity of cancer detection. DISCUSSION
Transrectal ultrasonography guided biopsy of the prostate in men with abnormal physical examination and/or laboratory findings is the mainstay of diagnosing contemporary CaP.1 Traditionally, the TNM staging system has incorporated TRUS findings to differentiate T1 from T2 CaP. However, the limitations of TRUS become apparent from the lack
Cancer detection by cores performed in different areas of the prostate Each Zone/Side (3,912 cores)
No. TRUS (%) Hypoechoic (⫹CaP)
Isoechoic (⫹CaP)
p Value
Base: 220 (9.3) 569 (10.4) 0.3 Lt 117 (10.2) 284 (10.3) 0.9 Rt 103 (8.4) 285 (10.6) 0.2 Middle: 238 (10.3) 535 (9.7) 0.4 Lt 120 (10.7) 270 (9.7) 0.3 Rt 118 (9.9) 265 (9.7) 0.8 Apex: 137 (9.4) 582 (9.2) 0.8 Lt 67 (9.0) 299 (9.4) 0.8 Rt 70 (9.7) 283 (8.9) 0.5 TZ: 7 (1.4) 63 (0.9) 0.2 Lt 5 (1.9) 29 (0.8) 0.1 Rt 2 (0.9) 34 (0.9) 0.6 Every core from the peripheral zone carried an average of 9.8% likelihood of positivity for cancer, irrespective of echogenicity, and no significant differences were observed.
FIG. 1. Association among side of gland, hypoechoic (Hypo) or isoechoic (Iso) lesions and cancer detection (⫹PcA). Detection rate by core was similar irrespective of echogenic pattern.
FIG. 2. Positive correlation between cancer detection (⫹PcA) and increased PSA levels. No significant differences were observed between biopsies taken from hypoechoic (Hypo) vs isoechoic (Iso) lesions.
of pathological correlation, questioning the value of TRUS lesions.11 Earlier ultrasonic categorization of CaP described these tumors as more hypoechoic than normal prostate However, when they enlarged, invaded other structures and developed calcifications, they became either hyperechoic, isoechoic, hypoechoic or mixed.4 Nevertheless, in an era when tumors were notoriously larger, ultrasonographic evaluation of small cancers revealed that up to 40% were from isoechoic areas.12, 13 Ellis et al evaluated the diagnostic accuracy of PSA, DRE and TRUS for the diagnosis of CaP.14 Although they found hypoechoic sectors more than twice as likely as isoechoic sectors of the prostate to contain malignancy on biopsy, 38% of the cancers detected in their series were from isoechoic areas. Overall, only 17% of all hypoechoic sectors contained carcinoma on biopsy and if only those lesions were sampled they would have missed the diagnosis in 25% of the cases. Dyke et al studied whether there was a staging difference between hypoechoic nodule directed and randomly taken TRUS guided biopsies, and noted that random biopsy results did not alter staging.15 In addition, random biopsy was responsible for an increased cancer yield of just 3%. Their conclusion stressed that lower grade tumors were not sensitive to TRUS unlike high grade lesions. Ohori et al examined 986 consecutive patients, and in 51% of their 241 cancer cases an ultrasonographic lesion was observed.11 However, for impalpable cancers ultrasound results provided no additional information regarding prognosis or pathological stage. In our study we provided detailed analysis per biopsy core, and found no relation between echogenicity pattern and CaP
514
TRANSRECTAL ULTRASOUND BIOPSY CORE POSITIVE PREDICTIVE VALUE
detection. The average discovery rate for a single core in the peripheral zone was between 9% and 10%, and the echogenic pattern source of the core made no difference. The incidence of CaP from isoechoic areas in our study (33%) is similar to the quoted literature of 32% to 42% of CaP being from isoechoic areas.16, 17 We demonstrated that a biopsy from a hypoechoic lesion was as sensitive as that of an isoechoic region on a per core basis. In general, prostates with hypoechoic lesions tended to have cancers but the lesion itself may not contain the tumor. Analysis of the 4 different compartments of the prostate revealed similar cancer detection rates. Therefore, from a given area, adjacent isoechoic zones should always be sampled, because if only hypoechoic lesions are sampled, significant disease could be missed. Thus, our results suggest the use of randomized biopsy strategy, regardless of the echogenicity pattern of TRUS, especially for nonpalpable lesions. To improve the diagnostic sensitivity of TRUS in early detection of prostate cancer, the use of different thresholds and PSA ranges has been recommended.18 We evaluated the rate of positive biopsy results of isoechoic or hypoechoic regions at different PSA levels. Although cancer detection rate improved with increasing levels of PSA, sensitivity of isoechoic or hypoechoic lesions to detect cancer was not different. Ito et al reported a positive predictive value and a negative predictive value for hypoechoic regions of 86% and 67%, respectively, in patients with serum PSA greater than 10 ng/ml.10 However, when PSA was 4.0 ng/ml or less positive and negative predictive values for such lesions ware 9.1% and 97.6%, respectively. Finally, they found 33% of isoechoic areas harboring cancer in patients with PSA greater than 10 ng/ml. Our data mirror these detection rates for isoechoic areas with PSA levels of more than 10 ng/ml. In all PSA ranges the hypoechoic areas were not able to provide better cancer detection rates than isoechoic biopsy samples. Indeed, a single set of sextant biopsies may miss clinically detectable cancer in 15% to 34% of men,19, 20 which is a limitation of our study. We evaluated the initial core by prostate area, which tends to be slightly medial. Likely, some tumors might be missed because previous knowledge and publications imply a higher yield for lateral cores. However, our aim was to determine the cancer significance of the hypoechoic lesions observed on TRUS in a large and contemporary series of patients, and we found that sensitivities were no different than those from isoechoic areas. CONCLUSIONS
Prostate cancer detection has shifted to earlier stages of the disease with the widespread use of PSA screening. Clinically, this prevents the degenerative changes that appear as hypoechoic areas on TRUS examination. Prostates with hypoechoic lesions are likely to harbor CaP but the lesion itself is as likely to be the source of cancer as the next adjacent area. Therefore, echogenicity has little role in determining the stage of operable tumors. REFERENCES
1. Hodge, K. K., McNeal, J. E., Terris, M. K. and Stamey, T. A.: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol, 142: 71, 1989
2. Lippman, H. R., Ghiatas, A. A. and Sarosdy, M. F.: Systematic transrectal ultrasound guided prostate biopsy after negative digitally directed prostate biopsy. J Urol, 147: 827, 1992 3. Tzai, T. S., Lin, J. S., Yeh, Y. C. and Chow, N. H.: The role of transrectal ultrasonography on the palpable and impalpable abnormal prostate. Eur Urol, 27: 142, 1995 4. Muldoon, L. and Resnick, M. I.: Results of ultrasonography of the prostate. Urol Clin North Am, 16: 693, 1989 5. Rifkin, M. D. and Choi, H.: Implications of small, peripheral hypoechoic lesions in endorectal US of the prostate. Radiology, 166: 619, 1988 6. Rifkin, M. D., Friedland, G. W. and Shortliffe, L.: Prostatic evaluation by transrectal endosonography: detection of carcinoma. Radiology, 158: 85, 1986 7. Bianco, F. J., Jr., Wood, D. P., Jr., Grignon, D. J., Sakr, W. A., Pontes, J. E. and Powell, I. J.: Prostate cancer stage shift has eliminated the gap in disease-free survival in black and white American men after radical prostatectomy. J Urol, 168: 479, 2002 8. Schwartz, K. L., Grignon, D. J., Sakr, W. A. and Wood, D. P., Jr.: Prostate cancer histologic trends in the metropolitan Detroit area, 1982 to 1996. Urology, 53: 769, 1999 9. Vo, T., Rifkin, M. D. and Peters, T. L.: Should ultrasound criteria of the prostate be redefined to better evaluate when and where to biopsy. Ultrasound Q, 17: 171, 2001 10. Ito, K., Ichinose, Y., Kubota, Y., Imai, K. and Yamanaka, H.: Clinicopathological features of prostate cancer detected by transrectal ultrasonography-guided systematic six-sextant biopsy. Int J Urol, 4: 474, 1997 11. Ohori, M., Kattan, M. W., Utsunomiya, T., Suyama, K., Scardino, P. T. and Wheeler, T. M.: Do impalpable stage T1c prostate cancers visible on ultrasound differ from those not visible? J Urol, 169: 964, 2003 12. Dahnert, W. F., Hamper, U. M., Eggleston, J. C., Walsh, P. C. and Sanders, R. C.: Prostatic evaluation by transrectal sonography with histopathologic correlation: the echopenic appearance of early carcinoma. Radiology, 158: 97, 1986 13. Shinohara, K., Wheeler, T. M. and Scardino, P. T.: The appearance of prostate cancer on transrectal ultrasonography: correlation of imaging and pathological examinations. J Urol, 142: 76, 1989 14. Ellis, W. J., Chetner, M. P., Preston, S. D. and Brawer, M. K.: Diagnosis of prostatic carcinoma: the yield of serum prostate specific antigen, digital rectal examination and transrectal ultrasonography. J Urol, 152: 1520, 1994 15. Dyke, C. H., Toi, A. and Sweet, J. M.: Value of random USguided transrectal prostate biopsy. Radiology, 176: 345, 1990 16. Ellis, W. J. and Brawer, M. K.: The significance of isoechoic prostatic carcinoma. J Urol, 152: 2304, 1994 17. Carter, H. B., Hamper, U. M., Sheth, S., Sanders, R. C., Epstein, J. I. and Walsh, P. C.: Evaluation of transrectal ultrasound in the early detection of prostate cancer. J Urol, 142: 1008, 1989 18. Fleshner, N. E., O’Sullivan, M., Premdass, C. and Fair, W. R.: Clinical significance of small (less than 0.2 cm3) hypoechoic lesions in men with normal digital rectal examinations and prostate-specific antigen levels less than 10 ng/ml. Urology, 53: 356, 1999 19. Roehrborn, C. G., Pickens, G. J. and Sanders, J. S.: Diagnostic yield of repeated transrectal ultrasound-guided biopsies stratified by specific histopathologic diagnoses and prostate specific antigen levels. Urology, 47: 347, 1996 20. Keetch, D. W., Catalona, W. J. and Smith, D. S.: Serial prostatic biopsies in men with persistently elevated serum prostate specific antigen values. J Urol, 151: 1571, 1994
Vol. 172, 512–514, August 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000131621.61732.6b
CONTEMPORARY IMPACT OF TRANSRECTAL ULTRASOUND LESIONS FOR PROSTATE CANCER DETECTION RAHMI ONUR, PETER J. LITTRUP, J. EDSON PONTES
AND
FERNANDO J. BIANCO, JR.*
From the Departments of Urology and Radiology (PL), Wayne State University School of Medicine and The Prostate Program of The Barbara Ann Karmanos Cancer Institute, Detroit, Michigan
ABSTRACT
Purpose: Transrectal ultrasound (TRUS) guided systematic biopsy of the prostate is the gold standard diagnostic modality for prostate cancer. Consequently, the value of discrete hypoechoic lesions on TRUS lesions considered suspicious for cancer deserves meticulous reevaluation, specifically in the prostate specific antigen era when the majority of tumors diagnosed are nonpalpable. We studied whether the predictability of a biopsy core changes if the tissue comes from an isoechoic vs hypoechoic lesion. Materials and Methods: Prospective data were collected on 3,912 consecutive patients referred to our medical center between 1993 and 1999 for biopsy of the prostate. A sextant technique (apex, mid gland and base) with an additional core biopsy from the transitional zone was used. If a hypoechoic lesion was identified, the biopsy was taken from the lesion. Correlation between hypoechoic lesions, isoechoic areas and cancer detection for each core was performed. Results: A total of 31,296 cores were obtained from the cohort. Overall 2,642 (68%) cores had at least 1 hypoechoic lesion ultrasonographically. Cancer was detected in 675 (25.5%) and 323 (25.4%) patients with or without hypoechoic lesions (p ⫽ 0.97). The per core cancer detection was fairly uniform and averaged 9.3% and 10.4% for hypoechoic and isoechoic areas, respectively. The difference was not statistically significant (p ⫽ 0.3). Gleason scores were less than 7, 7 and greater than 7 in 46%, 34% and 20% of cases, respectively. Conclusions: Despite the higher prevalence of cancers discovered in prostates with hypoechoic areas, the hypoechoic lesion itself was not associated with increased cancer prevalence compared with biopsy cores from isoechoic areas. For impalpable tumors TRUS findings are not contributory for staging. KEY WORDS: prostate-specific antigen, ultrasonography, biopsy, prostate neoplasms
Transrectal ultrasound (TRUS) guided systematic biopsy of the prostate is the gold standard diagnostic modality for prostate cancer.1, 2 Sonographic appearance of prostate cancer (CaP) was originally described as a hyperechoic lesion.2 Nevertheless, with the advent of high frequency transducers and improved instrumentation, CaP was recognized as appearing hypoechoic, hyperechoic, isoechoic or mixed on ultrasonography.3, 4 In the last 2 decades smaller cancers were considered to show less echogenicity on ultrasonography (ie hypoechoic) compared with larger tumors and, thus, TRUS had been used to screen for early nonpalpable cancer.5, 6 However, widespread use of prostate specific antigen (PSA) has led to an unparallel shift in stage and volume.7, 8 Consequently, the value of discrete hypoechoic lesions on transrectal ultrasound considered highly suspicious for cancer deserves meticulous reevaluation, specifically when the majority of tumors diagnosed are nonpalpable.9 PSA levels may increase the positive predictive value of hypoechoic lesions.10 However, not all hypoechoic regions represent cancer and, conversely, not all cancers are hypoechoic, with as many as 50% of clinically significant prostate cancers not being purely hypoechoic and 37% of all diagnosed cancers in a series being isoechoic.9 Recently, the clinical significance of ultrasonographic distinctions for CaP has come under scrutiny.11 However, studies assessing the exact prediction of a biopsy core, whether it comes from the hypo-
echoic lesion or not, are lacking. We studied the predictability of a biopsy core from each side of the prostate at the base, mid gland, apex and transitional zone (TZ) to determine the contemporary predictive value of a hypoechoic lesion identified in those areas. MATERIALS AND METHODS
Prospective data were collected on 3,912 consecutive patients referred to our medical center between January 1993 and December 1999 for initial TRUS biopsy of the prostate. Biopsy indications included abnormal digital rectal examination, increased total PSA level or velocity, low percent free PSA and increased PSA density. Digital rectal examination (DRE) was regarded as abnormal if a nodule or hard consistency was palpated. Patients were medicated with a fluoroquinolone for 3 days starting the day before the procedure. Men at risk for heart valve disease or history of valve surgery received additional amoxicillin or erythromycin. A cleansing enema was administered the morning of the biopsy. After informed consent was obtained from each patient, DRE was performed followed by TRUS. The gland was carefully examined to ensure that all areas of the gland were optimally evaluated. The sagittal and axial planes were examined. Echogenic patterns were characterized by location (base, mid gland or apex) and side, and prostatic volume was measured. Biopsies were performed bilaterally in the axial projection using the sextant technique with an additional core from the TZ. Every biopsy core from the base, mid gland, apex and TZ on the right and left sides was placed into a container and labeled accordingly. DRE, serum PSA, area of
Accepted for publication March 19, 2004. * Correspondence: Department of Urology, Memorial SloanKettering Cancer Center, 1275 York Ave., New York, New York 10021 (telephone: 646-523-1051; FAX: 212-988-0683; e-mail: [email protected]). 512
TRANSRECTAL ULTRASOUND BIOPSY CORE POSITIVE PREDICTIVE VALUE
513
the prostate and isoechoic or hypoechoic origin of each core were meticulously recorded into a relational database. Every positive core was graded using the Gleason system and a score was assigned. Chi-square tests were used to assess the statistical comparison of these nonparametric variables. The preoperative levels were stratified into 5 categories, and the correlation with presence of cancer and prostate echogenicity was also assessed. RESULTS
A total of 31,296 biopsy cores were obtained from 3,912 patients. Median patient age and PSA were 67 years (range 37 to 95) and 6.6 ng/ml (range 0.9 to 51). Overall, cancer was detected in 998 (25.5%) patients. At least 1 hypoechoic lesion was seen in 2,642 (68%) patients. Exclusive lesions in the right or left lobes were detected in 722 (19%) and 702 (18%) patients, respectively, and 1,218 (31%) had hypoechoic lesions in both lobes. However, even if bilateral hypoechogenic lesions were seen, cancer detection was not different than for those cases with no hypoechogenicity (312 of 1,218 [25.6%] vs 232 of 1,270 [25.4%], p ⫽ 0.9). A detailed description showing the cancer detection rates for each core by side and prostate zone is provided in the table. Hypoechoic lesions at the prostate base harbored cancer in 9.3% of all cores, whereas tumors were detected in isoechoic areas 10.4% of the time, which was not statistically different (p ⫽ 0.3). This pattern was similar in the other prostatic areas as biopsies directed to hypoechoic areas were not able to provide better detection rates compared with random cores of the gland. The specific probability of finding cancer in a hypoechoic area within the right or left lobe was 18% (131 of 722) and 19% (134 of 702) of patients, respectively (p ⫽ 0.7), which was similar to yields from isoechoic area (fig. 1). The overall likelihood of detecting cancer in patients with or without hypoechoic lesions in the gland was similar (25.5% [675 of 2,642] vs 25.4% [323 of 1,270], respectively, p ⫽ 0.8). Figure 2 shows the cancer detection rates stratified by PSA ranges. Although there was a positive correlation between PSA level and cancer detection rate, examination of hypoechoic and isoechoic biopsies revealed no significant difference in sensitivity of cancer detection. DISCUSSION
Transrectal ultrasonography guided biopsy of the prostate in men with abnormal physical examination and/or laboratory findings is the mainstay of diagnosing contemporary CaP.1 Traditionally, the TNM staging system has incorporated TRUS findings to differentiate T1 from T2 CaP. However, the limitations of TRUS become apparent from the lack
Cancer detection by cores performed in different areas of the prostate Each Zone/Side (3,912 cores)
No. TRUS (%) Hypoechoic (⫹CaP)
Isoechoic (⫹CaP)
p Value
Base: 220 (9.3) 569 (10.4) 0.3 Lt 117 (10.2) 284 (10.3) 0.9 Rt 103 (8.4) 285 (10.6) 0.2 Middle: 238 (10.3) 535 (9.7) 0.4 Lt 120 (10.7) 270 (9.7) 0.3 Rt 118 (9.9) 265 (9.7) 0.8 Apex: 137 (9.4) 582 (9.2) 0.8 Lt 67 (9.0) 299 (9.4) 0.8 Rt 70 (9.7) 283 (8.9) 0.5 TZ: 7 (1.4) 63 (0.9) 0.2 Lt 5 (1.9) 29 (0.8) 0.1 Rt 2 (0.9) 34 (0.9) 0.6 Every core from the peripheral zone carried an average of 9.8% likelihood of positivity for cancer, irrespective of echogenicity, and no significant differences were observed.
FIG. 1. Association among side of gland, hypoechoic (Hypo) or isoechoic (Iso) lesions and cancer detection (⫹PcA). Detection rate by core was similar irrespective of echogenic pattern.
FIG. 2. Positive correlation between cancer detection (⫹PcA) and increased PSA levels. No significant differences were observed between biopsies taken from hypoechoic (Hypo) vs isoechoic (Iso) lesions.
of pathological correlation, questioning the value of TRUS lesions.11 Earlier ultrasonic categorization of CaP described these tumors as more hypoechoic than normal prostate However, when they enlarged, invaded other structures and developed calcifications, they became either hyperechoic, isoechoic, hypoechoic or mixed.4 Nevertheless, in an era when tumors were notoriously larger, ultrasonographic evaluation of small cancers revealed that up to 40% were from isoechoic areas.12, 13 Ellis et al evaluated the diagnostic accuracy of PSA, DRE and TRUS for the diagnosis of CaP.14 Although they found hypoechoic sectors more than twice as likely as isoechoic sectors of the prostate to contain malignancy on biopsy, 38% of the cancers detected in their series were from isoechoic areas. Overall, only 17% of all hypoechoic sectors contained carcinoma on biopsy and if only those lesions were sampled they would have missed the diagnosis in 25% of the cases. Dyke et al studied whether there was a staging difference between hypoechoic nodule directed and randomly taken TRUS guided biopsies, and noted that random biopsy results did not alter staging.15 In addition, random biopsy was responsible for an increased cancer yield of just 3%. Their conclusion stressed that lower grade tumors were not sensitive to TRUS unlike high grade lesions. Ohori et al examined 986 consecutive patients, and in 51% of their 241 cancer cases an ultrasonographic lesion was observed.11 However, for impalpable cancers ultrasound results provided no additional information regarding prognosis or pathological stage. In our study we provided detailed analysis per biopsy core, and found no relation between echogenicity pattern and CaP
514
TRANSRECTAL ULTRASOUND BIOPSY CORE POSITIVE PREDICTIVE VALUE
detection. The average discovery rate for a single core in the peripheral zone was between 9% and 10%, and the echogenic pattern source of the core made no difference. The incidence of CaP from isoechoic areas in our study (33%) is similar to the quoted literature of 32% to 42% of CaP being from isoechoic areas.16, 17 We demonstrated that a biopsy from a hypoechoic lesion was as sensitive as that of an isoechoic region on a per core basis. In general, prostates with hypoechoic lesions tended to have cancers but the lesion itself may not contain the tumor. Analysis of the 4 different compartments of the prostate revealed similar cancer detection rates. Therefore, from a given area, adjacent isoechoic zones should always be sampled, because if only hypoechoic lesions are sampled, significant disease could be missed. Thus, our results suggest the use of randomized biopsy strategy, regardless of the echogenicity pattern of TRUS, especially for nonpalpable lesions. To improve the diagnostic sensitivity of TRUS in early detection of prostate cancer, the use of different thresholds and PSA ranges has been recommended.18 We evaluated the rate of positive biopsy results of isoechoic or hypoechoic regions at different PSA levels. Although cancer detection rate improved with increasing levels of PSA, sensitivity of isoechoic or hypoechoic lesions to detect cancer was not different. Ito et al reported a positive predictive value and a negative predictive value for hypoechoic regions of 86% and 67%, respectively, in patients with serum PSA greater than 10 ng/ml.10 However, when PSA was 4.0 ng/ml or less positive and negative predictive values for such lesions ware 9.1% and 97.6%, respectively. Finally, they found 33% of isoechoic areas harboring cancer in patients with PSA greater than 10 ng/ml. Our data mirror these detection rates for isoechoic areas with PSA levels of more than 10 ng/ml. In all PSA ranges the hypoechoic areas were not able to provide better cancer detection rates than isoechoic biopsy samples. Indeed, a single set of sextant biopsies may miss clinically detectable cancer in 15% to 34% of men,19, 20 which is a limitation of our study. We evaluated the initial core by prostate area, which tends to be slightly medial. Likely, some tumors might be missed because previous knowledge and publications imply a higher yield for lateral cores. However, our aim was to determine the cancer significance of the hypoechoic lesions observed on TRUS in a large and contemporary series of patients, and we found that sensitivities were no different than those from isoechoic areas. CONCLUSIONS
Prostate cancer detection has shifted to earlier stages of the disease with the widespread use of PSA screening. Clinically, this prevents the degenerative changes that appear as hypoechoic areas on TRUS examination. Prostates with hypoechoic lesions are likely to harbor CaP but the lesion itself is as likely to be the source of cancer as the next adjacent area. Therefore, echogenicity has little role in determining the stage of operable tumors. REFERENCES
1. Hodge, K. K., McNeal, J. E., Terris, M. K. and Stamey, T. A.: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol, 142: 71, 1989
2. Lippman, H. R., Ghiatas, A. A. and Sarosdy, M. F.: Systematic transrectal ultrasound guided prostate biopsy after negative digitally directed prostate biopsy. J Urol, 147: 827, 1992 3. Tzai, T. S., Lin, J. S., Yeh, Y. C. and Chow, N. H.: The role of transrectal ultrasonography on the palpable and impalpable abnormal prostate. Eur Urol, 27: 142, 1995 4. Muldoon, L. and Resnick, M. I.: Results of ultrasonography of the prostate. Urol Clin North Am, 16: 693, 1989 5. Rifkin, M. D. and Choi, H.: Implications of small, peripheral hypoechoic lesions in endorectal US of the prostate. Radiology, 166: 619, 1988 6. Rifkin, M. D., Friedland, G. W. and Shortliffe, L.: Prostatic evaluation by transrectal endosonography: detection of carcinoma. Radiology, 158: 85, 1986 7. Bianco, F. J., Jr., Wood, D. P., Jr., Grignon, D. J., Sakr, W. A., Pontes, J. E. and Powell, I. J.: Prostate cancer stage shift has eliminated the gap in disease-free survival in black and white American men after radical prostatectomy. J Urol, 168: 479, 2002 8. Schwartz, K. L., Grignon, D. J., Sakr, W. A. and Wood, D. P., Jr.: Prostate cancer histologic trends in the metropolitan Detroit area, 1982 to 1996. Urology, 53: 769, 1999 9. Vo, T., Rifkin, M. D. and Peters, T. L.: Should ultrasound criteria of the prostate be redefined to better evaluate when and where to biopsy. Ultrasound Q, 17: 171, 2001 10. Ito, K., Ichinose, Y., Kubota, Y., Imai, K. and Yamanaka, H.: Clinicopathological features of prostate cancer detected by transrectal ultrasonography-guided systematic six-sextant biopsy. Int J Urol, 4: 474, 1997 11. Ohori, M., Kattan, M. W., Utsunomiya, T., Suyama, K., Scardino, P. T. and Wheeler, T. M.: Do impalpable stage T1c prostate cancers visible on ultrasound differ from those not visible? J Urol, 169: 964, 2003 12. Dahnert, W. F., Hamper, U. M., Eggleston, J. C., Walsh, P. C. and Sanders, R. C.: Prostatic evaluation by transrectal sonography with histopathologic correlation: the echopenic appearance of early carcinoma. Radiology, 158: 97, 1986 13. Shinohara, K., Wheeler, T. M. and Scardino, P. T.: The appearance of prostate cancer on transrectal ultrasonography: correlation of imaging and pathological examinations. J Urol, 142: 76, 1989 14. Ellis, W. J., Chetner, M. P., Preston, S. D. and Brawer, M. K.: Diagnosis of prostatic carcinoma: the yield of serum prostate specific antigen, digital rectal examination and transrectal ultrasonography. J Urol, 152: 1520, 1994 15. Dyke, C. H., Toi, A. and Sweet, J. M.: Value of random USguided transrectal prostate biopsy. Radiology, 176: 345, 1990 16. Ellis, W. J. and Brawer, M. K.: The significance of isoechoic prostatic carcinoma. J Urol, 152: 2304, 1994 17. Carter, H. B., Hamper, U. M., Sheth, S., Sanders, R. C., Epstein, J. I. and Walsh, P. C.: Evaluation of transrectal ultrasound in the early detection of prostate cancer. J Urol, 142: 1008, 1989 18. Fleshner, N. E., O’Sullivan, M., Premdass, C. and Fair, W. R.: Clinical significance of small (less than 0.2 cm3) hypoechoic lesions in men with normal digital rectal examinations and prostate-specific antigen levels less than 10 ng/ml. Urology, 53: 356, 1999 19. Roehrborn, C. G., Pickens, G. J. and Sanders, J. S.: Diagnostic yield of repeated transrectal ultrasound-guided biopsies stratified by specific histopathologic diagnoses and prostate specific antigen levels. Urology, 47: 347, 1996 20. Keetch, D. W., Catalona, W. J. and Smith, D. S.: Serial prostatic biopsies in men with persistently elevated serum prostate specific antigen values. J Urol, 151: 1571, 1994