- Email: [email protected]
Impact of Multiparametric Endorectal Coil Prostate Magnetic Resonance Imaging on Disease Reclassification Among Active Surveillance Candidates: A Prospective Cohort Study
Impact of Multiparametric Endorectal Coil Prostate Magnetic Resonance Imaging on Disease Reclassification Among Active Surveillance Candidates: A Prospective Cohort Study David Margel,* Stanley A. Yap, Nathan Lawrentschuk, Laurence Klotz, Masoom Haider, Karen Hersey, Antonio Finelli, Alexandre Zlotta, John Trachtenberg and Neil Fleshner† From the Division of Urology, Department of Surgical Oncology (DM, SAY, NL, KH, AF, AZ, JT, NF) and Department of Medical Imaging (MH), Princess Margaret Hospital, University Health Network and Sunnybrook Health Science Centre (LK), Toronto, Ontario, Canada
Purpose: We report magnetic resonance imaging findings among unselected men with low risk prostate cancer before active surveillance. Materials and Methods: We prospectively enrolled men with low grade, low risk, localized prostate cancer. All patients underwent multiparametric endorectal coil magnetic resonance imaging and were offered confirmatory biopsy within 1 year of imaging. The primary outcome was the impact of magnetic resonance imaging on identifying patients who were reclassified by confirmatory biopsy as no longer fulfilling active surveillance criteria. We further identified clinical parameters associated with reclassification. The cohort was stratified as patients with 1) normal magnetic resonance imaging, 2) cancer on magnetic resonance imaging concordant with initial biopsy (less than 1 cm) and 3) cancer on magnetic resonance imaging larger than 1 cm. We performed univariate analysis to assess differences in clinical parameters among the groups. Results: Magnetic resonance imaging did not detect cancer in 23 cases (38%) while magnetic resonance imaging and initial biopsy were concordant in 24 (40%). Magnetic resonance imaging detected a 1 cm or larger lesion in 13 patients (22%). Of the cases 18 (32.14%) were reclassified. When no cancer was identified on magnetic resonance imaging, only 2 cases (3.5%) were reclassified. The positive and negative predictive values for magnetic resonance imaging predicting reclassification were 83% (95% CI 73–93) and 81% (95% CI 71–91), respectively. Prostate specific antigen density was increased in patients with lesions larger than 1 cm on magnetic resonance imaging compared to those with no cancer on imaging (median 0.15 vs 0.07 ng/ml/cc, p ⫽ 0.016). Conclusions: Magnetic resonance imaging appears to have a high yield for predicting reclassification among men who elect active surveillance. Upon confirmation of our results magnetic resonance imaging may be used to better select and guide patients before active surveillance.
Abbreviations and Acronyms ADC ⫽ apparent diffusion coefficient AS ⫽ active surveillance DCE ⫽ dynamic contrast enhanced DWI ⫽ diffusion-weighted imaging MRI ⫽ magnetic resonance imaging PCa ⫽ prostate cancer PSA ⫽ prostate specific antigen TRUS ⫽ transrectal ultrasound Submitted for publication August 15, 2011. Study received institutional review board approval. Supported by a Prostate Cancer Canada grant. Supplementary material for this article can be obtained at http://www.prostatecentre.ca. * Correspondence: UroOncology Fellowship Office, 610 University Ave., Toronto, Ontario, Canada, M5G 2M9 (telephone: 416-946-2000; FAX: 416-598-9997; e-mail: [email protected]). † Financial interest and/or other relationship with Merck, Novartis, Pfizer, AstraZeneca, GlaxoSmithKline and Sanofi-Aventis.
Key Words: prostate, prostatic neoplasms, magnetic resonance imaging, biopsy, diagnosis CURRENTLY most patients with PCa are diagnosed with favorable risk disease,1 which poses little risk to longevity.2 Traditional treatments such
as surgery or radiation are associated with significant adverse events and can negatively affect the quality of life of patients and their families.3 AS has
0022-5347/12/1874-1247/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 1247-1252, April 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.11.112
AND
RESEARCH, INC.
www.jurology.com
1247
1248
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
the potential to decrease the overtreatment of patients who harbor clinically insignificant disease while offering curative treatment to those in whom disease is reclassified as higher risk after an observation period and repeat biopsy.4 – 6 Large cohorts showed a good outcome with this treatment paradigm5,7 A troubling aspect of AS is that a subset of patients diagnosed with low risk disease appears to be under sampled and in fact harbor larger, often higher grade cancer. Increasing evidence suggests that early failure among men on AS is due to underappreciation of the amount of disease at baseline.8 –10 Fear of under sampling is a potential impediment to wider scale adoption of AS. Although TRUS biopsy remains the gold standard for diagnosing prostate cancer, MRI currently has the highest accuracy among radiological modalities to detect and stage prostate tumors.11,12 However, little is known of MRI among patients with PCa at low risk. We report our experience with prospective MRI in an unselected, consecutive group of men with low risk PCa before enrollment in an AS protocol.
PATIENTS AND METHODS General We analyzed patients initially screened by MRI to determine eligibility to enroll in a vitamin intervention trial (ClinicalTrials.gov NCT00744549). It is a phase IIa, placebo controlled, randomized trial of the combination of vitamin E, lycopene and selenium. The aim of the study is to determine the impact of this combination on MRI detected PCa among men on AS. As a byproduct, we present MRI results in these men.
Patient Population Upon receiving approval from our institutional review board we enrolled 60 consecutive patients who elected AS for recently diagnosed PCa. All patients provided informed consent. For study inclusion patients had to have low grade, low risk, localized PCa (clinical stage T1c-T2a, Gleason score 6 or less, no Gleason pattern 4 and serum PSA 10 ng/ml or less) and be considered suitable candidates for AS. No more than 3 cores on the entry biopsy could be positive for PCa and all biopsy cores had to have less than 50% involvement. Eligible candidates had to be between ages 50 and 80 years, and have a life expectancy of more than 5 years. Initial biopsy, which had to have at least 10 cores, was performed at Princess Margaret Hospital in 42 patients (75%). Patients were excluded from analysis if they had previously undergone a prostate related surgery or procedure since that may interfere with MRI analysis.
care, Milwaukee, Wisconsin) using a 4-channel phased array surface coil coupled to an endorectal coil. T2weighted MRI was done with 2 phases acquired before injection with 20 ml gadopentate-diethylenetetramine pentaacetic acid contrast agent at 4 ml per second, followed by a 20 ml saline flush using a Spectris MR power injection system (Medrad®). A total of 26 slices were obtained per phase. All MRI data sets were obtained at identical slice locations with a 3 mm slice thickness and no intersection gap. Image interpretation. Images were reviewed by a single experienced uroradiologist. All MRIs were done at least 6 weeks from biopsy to minimize artifact. Diagnostic features for malignancy were a low T2 signal in the peripheral zone, a relatively low ADC calculated from DWI, early enhancement and washout on DCE MRI. For the transitional zone a poorly defined nodule that distorted the normal architecture and had concordant abnormalities on DWI and DCE MRI was considered suspicious for malignancy.13–15
Confirmatory Biopsy Confirmatory biopsy was offered to all study participants and done within 1 year of diagnosis according to our AS protocol. Patients with MRI that showed an MRI lesion greater than 1 cm in any dimension were offered immediate confirmatory biopsy. We used a threshold of 1 cm since it represents a sphere of 0.5 cc, which is the Epstein criterion for clinically significant PCa.16 We used a clinically relevant cutoff to avoid unnecessary biopsy while still detecting most clinically relevant lesions. Prostate biopsies were TRUS guided and obtained within a year of PCa diagnosis so that correlation was possible between biopsy and imaging findings. Confirmatory biopsy consisted of a minimum 12 cores. It was done at Princess Margaret Hospital in all cases. In patients with a lesion on MRI additional cores were taken via a previously reported technique.10 Briefly, abnormal sites on MRI were annotated into zones and provided to the radiologist performing TRUS biopsy to target the biopsy. Magnetic resonance images appeared on the screen adjacent to the ultrasound, allowing real-time comparison of methods and zones of interest. All entry and secondary prostate biopsies were reviewed by dedicated uropathologists at our institution.
Study End Points The primary outcome was MRI accuracy for identifying patients reclassified as no longer fulfilling AS criteria by confirmatory biopsy. Reclassification occurred if confirmatory biopsy revealed any Gleason 7 or greater, 3 or more cores positive for PCa, or a highest tumor volume in any core of greater than 50%. We further identified any clinical parameters, such as patient age, prostate volume, PSA, PSA density, number of positive cores and percent of cancer in a core, associated with a lesion larger than 1 cm on MRI.
Magnetic Resonance Imaging
Statistical Analysis
Protocol. Axial-oblique, fast spin-echo, T2-weighted, echo planar DWI, and multi and DCE MRI were done on a 1.5 Tesla EchoSpeed or Excite HD MRI system (GE Health-
For analysis purposes the study population was stratified into 3 groups, including group 1—normal MRI, group 2— cancer characteristics on MRI concordant with initial
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
biopsy, defined as a lesion less than 1 cm in all dimensions, and group 3—an MRI lesion larger than 1 cm. Descriptive statistics were used to describe the clinical parameters of each group, such as age, PSA, prostate volume (calculated based on TRUS data), PSA density (calculated as PSA/prostate volume) and confirmatory biopsy data. Univariate analysis using parametric (1-way ANOVA) and nonparametric (the Kruskal-Wallis test) was done to assess differences in these clinical parameters among groups. When a test revealed a significant value, we used a post hoc test with the Bonferroni correction to determine statistically significant differences among the groups. The accuracy of MRI to predict reclassification on confirmatory biopsy was assessed with a contingency table. We calculated sensitivity, specificity, and positive and negative predictive values along with the corresponding 95% CIs.
RESULTS Figure 1 shows the study design. The cohort consisted of 60 consecutive patients. MRI did not detect cancer in 23 patients (38%) while MRI and initial biopsy were concordant in 24 (40%). MRI detected a larger than 1 cm lesion in 13 patients (22%). Data on confirmatory biopsy were missing in 4 patients, of whom 3 refused and 1 was lost to followup. A total of 56 patients were included in further analysis. Overall 18 of the 56 patients (32.14%) in our cohort were reclassified. The reclassification rate by confirmatory biopsy was significantly higher among those with a large lesion on MRI compared to that of any other group (17.85% vs 10.71% and 3.5%, respectively, p ⫽ 0.02).
1249
The specificity and sensitivity of MRI compared to confirmatory biopsy results were 95% (95% CI 82– 99) and 55% (95% CI 43– 67), respectively. Falsepositive MRI was observed in 2 patients (3.5%) while 8 (14.28%) had a false-negative prediction. Calculated positive and negative predictive values were 83% (95% CI 73–93) and 81% (95% CI 71–91), respectively. The table lists baseline clinical parameters stratified by MRI findings. On univariate analysis PSA density was increased in patients with a lesion larger than 1 cm on MRI compared to those with normal MRI (0.15 vs 0.07 ng/ml/cc, p ⫽ 0.018). The table also shows differences in the confirmatory biopsy procedure and results. Time from initial biopsy was shorter for patients with a large MRI lesion compared to the 2 other groups (1.8 vs 10.3 and 9.6 months, respectively, p ⫽ 0.02). The mean ⫾ SD number of cores taken on confirmatory biopsy was greater in each group with an MRI detected lesion compared to the normal MRI group (16.2 ⫾ 0.9 and 15.3 ⫾ 1.8, respectively, vs 12.3 ⫾ 2, p ⫽ 0.041). TRUS identified a suspicious lesion in 5 patients and these lesions were also biopsied. TRUS identified lesions were concordant with MRI in 2 cases in which the lesions were cancerous. Other TRUS lesions were benign. MRI findings were anatomically concordant with confirmatory biopsy in 92% of MRI identified lesions. Of lesions larger than 1 cm 55% were located in the anterior zone while the others were in the peripheral (27%) or the transitional (27%) zone.
Figure 1. Study schematic shows 60 patients who underwent MRI and 3 predefined groups, including 1—normal MRI, 2— cancer characteristics on MRI concordant with initial biopsy (Bx) and 3—lesion on MRI more than 1 cm in any dimension.
1250
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
Clinical characteristics and confirmatory biopsy data
Mean ⫾ SD age Median ng/ml PSA (range) Median cc prostate vol (range) Median No. biopsy 1 pos cores (range) Mean ⫾ SD % largest biopsy 1 core PCa Median ng/ml/cc PSA density (range) Confirmatory biopsy data: Median mos from biopsy 1 (range) Mean ⫾ SD No. biopsy 2 cores % TRUS identified Ca (No. pts) % MRI-biopsy anatomical concordance Mean ⫾ SD % biopsy 2 Gleason 7 or greater (No. pts)
MRI Greater Than 1 cm PCa
MRI Concordant With Initial Biopsy
Normal MRI
p Value
62.6 ⫾ 7 5.94 (1.7–10) 37.5 (22–67) 2 (1–3) 12.1 ⫾ 7.1 0.15 (0.6–0.05)
63.5 ⫾ 6 4.4 (1.1–9.1) 37.5 (19–76) 2 (1–3) 15.3 ⫾ 8.2 0.13 (0.22–0.03)
64 ⫾ 8.2 4.1 (1.1–9.9) 41 (21–83) 2 (1–3) 13.6 ⫾ 5.1 0.07 (0.21–0.03)
0.6 0.07 0.09 0.5 0.3 0.04
1.8 (1.2–2.4) 16.2 ⫾ 0.9 5.3 (3) 100 9 (5)
10.3 (6–12) 15.3 ⫾ 1.8 1.7 (1) 83 5.3 (3)
9.6 (7–12) 12.3 ⫾ 0.7 1.7 (1) Not recorded 1.7 (1)
0.02 0.041 0.052 0.2 0.047
Gleason 7 cancers were identified in 9 patients, including 5 with large MRI detectable lesions and only 1 with normal prostate MRI (p ⫽ 0.047). Figure 2 shows a large anterior tumor that was missed on initial diagnosis but captured well on MRI.
DISCUSSION The most important finding of our study is that MRI may reveal an unrecognized significant lesion in up to 22% of patients and biopsy of these areas reclassified 17.85% of cases as no longer fulfilling conservative criteria for AS. Also, when no cancerous lesion was identified on MRI, the chance of reclassification on repeat biopsy was extremely low at 3.5%. To our knowledge we are the first to report the positive and negative predictive values of MRI to predict reclassification (83%, 95% CI 73–93 and 81%, 95% CI 71–91, respectively) in an unselected group of patients before AS. Several earlier studies examined MRI in patients on AS. Cabrera et al retrospectively reported on 92 men with biopsy proven PCa.17 They found no association among clinical stage, Gleason, PSA and apparent tumor on endorectal MRI. In another retrospective analysis van As et al noted that low ADC was associated with adverse histology on repeat biopsy.18 Tumor ADC correlated with maximum core
Figure 2. MRI reveals large anterior tumor missed on initial diagnosis. Red dotted outline indicates tumor.
involvement, the percent of positive cores, initial PSA and the free-to-total PSA ratio. Our findings are also supported by those of Fradet et al.19 They reported that in the context of AS a lesion suggesting cancer on MRI may confer a threefold increased risk of overall cancer progression. deSouza et al prospectively assessed 44 consecutive patients with clinically localized PCa by performing diffusion-weighted MRI in addition to standard T2-weighted MRI.20 Water diffusion in PCa was different between 26 patients with low risk PCa and 18 with intermediate or high risk disease. Thus, ADCs offer the potential to differentiate indolent from aggressive PCa. However, that study included patients who did not meet AS criteria. To our knowledge we are the first to report MRI findings done prospectively for all patients eligible for AS. The site of missed tumors is another factor to consider. In our study most missed large tumors were located in the anterior (55%) or the transitional (27%) zone. This was also documented in other series.21,22 MRI seems to be a valuable tool to detect these hidden cancers. The optimal MRI protocol before AS is not known. We used a multiparametric combination of T2weighted, echo planar, DWI, and multi and DCE MRI. We believe that any single technique would lack the sensitivity to detect PCa before AS since most patients harbor low volume cancer. Ploussard et al recently reported a retrospective analysis of 96 patients who met AS criteria and underwent T2weighted MRI.23 They concluded that MRI does not improve the prediction of reclassification. We believe that these discrepancies may have been due to the use of multiparametric MRI. Larger studies are needed to determine the best, most costeffective MRI protocol. Some AS candidates had normal MRIs. In our series only 2 such cases were reclassified. Cabrera et al concluded that MRI does not appear to have prognostic value for AS.17 However, they did not use DWI. Upon confirmation of our results patients with
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
normal MRI may be spared the pain and complications associated with repeat biopsy at 1 year. Our results should be viewed in the context of a tertiary referral center where there is expertise in prostate MRI and an interest in AS. We have a radiology team dedicated to prostate imaging that has experience with interpreting prostate MRI results, enabling true guidance for subsequent biopsies. Since MRI readings are observer dependent, these results may take time to duplicate at less experienced centers. A potential limitation is the comparison of targeted biopsy with TRUS guided biopsy. It is expected that biopsy to a suspicious lesion would yield a higher percent of positive cores and of cancer in each core. However, the clinical usefulness of MRI in this setting is to detect unexpected lesions and aid in targeting a biopsy. Thus, we believe that such a comparison is appropriate. Ideally additional random biopsies in the group with normal prostate MRI would be done to compensate for additional sampling. However, prior data suggest that the yield of this technique is minimal above the traditional 12 to 16 cores.24 Time to confirmatory biopsy was shorter among patients with a large MRI lesion. We thought that it was unethical not to biopsy patients with lesions greater than 1 cm yet almost unethical to rebiopsy patients with normal prostate MRI in light of increasing biopsy complications.25 However, since PCa grows slowly, we believe that a 1-year delay almost
1251
certainly captures missed initial lesions and not progression during followup. Also, not all initial biopsies were done locally. However, this percent did not differ among the study groups and, thus, should not have affected our results. Finally, our cohort was rather small. Further confirmation is needed in a larger, multicenter study. Another potential concern with our study is the validity of a 1 cm threshold for repeat biopsy. One cm represents a sphere of 0.5 cc, which is the Epstein criterion for clinically significant PCa.16 We specifically did not choose a 0.5 cm cutoff, considering the highly prevalent screening protocols and overtreatment. We believe that the purpose of extended diagnostic strategies would be to decrease the risk of misclassification and preclude the need for rebiopsy. Future study must address the criteria that trigger biopsy, including refined dimension thresholds and other potential MRI characteristics.
CONCLUSIONS This study highlights several important findings. MRI appears to have the potential to identify early and noninvasively candidates for surveillance who harbor occult adverse disease. In this series significant PCa was present in 17.58% of those eligible for AS. MRI can detect and accurately direct biopsy to these hidden cancers. PSA density may be used to help clinicians select cases appropriate for MRI. Upon confirmation MRI may be used to better select and guide patients before AS.
REFERENCES 1. Detailed guide: prostate cancer—what are the key statistics about prostate cancer? American Cancer Society 2009. Available at: http://www.cancer. org/docroot/CRI/content/CRI_2_4_1X_What_are_ the_key_statistics_for_prostate_cancer_36.asp. Accessed November 2011. 2. Jemal A, Siegel R, Ward E et al: Cancer statistics, 2008. CA Cancer J Clin 2008; 58: 71. 3. Sanda MG, Dunn RL, Michalski J et al: Quality of life and satisfaction with outcome among prostate-cancer survivors. N Engl J Med. 2008; 358: 1250. 4. Warlick C, Trock BJ, Landis P et al: Delayed versus immediate surgical intervention and prostate cancer outcome. J Natl Cancer Inst 2006; 98: 355. 5. Klotz L, Zhang L, Lam A et al: Clinical results of long-term follow-up of a large, active surveillance cohort with localized prostate cancer. J Clin Oncol 2010; 28: 126. 6. NCCN Guidelines® and Clinical Resources. Updated Prostate Cancer Guidelines, vs 4.2011. National Comprehensive Cancer Network. Available
at http://www.nccn.org/professionals/physician_ gls/f_guidelines.asp#site. Accessed November 2011. 7. Stattin P, Holmberg E, Johansson JE et al: Outcomes in localized prostate cancer: National Prostate Cancer Register of Sweden follow-up study. J Natl Cancer Inst 2010; 102: 950. 8. Klotz L: Active surveillance for prostate cancer: patient selection and management. Curr Oncol, suppl., 2010; 2: S11.
12. Weinreb JC, Blume JD, Coakley FV et al: Prostate cancer: sextant localization at MR imaging and MR spectroscopic imaging before prostatectomy— results of ACRIN prospective multi-institutional clinicopathologic study. Radiology 2009; 251: 122. 13. Haider MA, van der Kwast TH, Tanguay J et al: Combined T2-weighted and diffusion-weighted MRI for localization of prostate cancer. AJR Am J Roentgenol 2007; 189: 323.
9. Bott SR, Young MP, Kellett MJ et al: Anterior prostate cancer: is it more difficult to diagnose? BJU Int 2002; 89: 886.
14. Akin O, Sala E, Moskowitz CS et al: Transition zone prostate cancers: features, detection, localization, and staging at endorectal MR imaging. Radiology 2006; 239: 784.
10. Lawrentschuk N, Haider MA, Daljeet N et al: ‘Prostatic evasive anterior tumours’: the role of magnetic resonance imaging. BJU Int 2010; 105: 1231.
15. Futterer JJ, Heijmink SW, Scheenen TW et al: Prostate cancer localization with dynamic contrast-enhanced MR imaging and proton MR spectroscopic imaging. Radiology 2006; 241: 449.
11. Kurhanewicz J, Vigneron D, Carroll P et al: Multiparametric magnetic resonance imaging in prostate cancer: present and future. Curr Opin Urol 2008; 18: 71.
16. Epstein JI, Walsh PC, Carmichael M et al: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 1994; 271: 368.
1252
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
17. Cabrera AR, Coakley FV, Westphalen AC et al: Prostate cancer: is in apparent tumor at endorectal MR and MR spectroscopic imaging a favorable prognostic finding in patients who select active surveillance? Radiology 2008; 247: 444.
20. deSouza NM, Riches SF, Vanas NJ et al: Diffusion-weighted magnetic resonance imaging: a potential noninvasive marker of tumour aggressiveness in localized prostate cancer. Clin Radiol 2008; 63: 774.
18. van As NJ, de Souza NM, Riches SF et al: A study of diffusion-weighted magnetic resonance imaging in men with untreated localised prostate cancer on active surveillance. Eur Urol 2008; 56: 981.
21. Hambrock T, Somford DM, Hoeks C et al: Magnetic resonance imaging guided prostate biopsy in men with repeat negative biopsies and increased prostate specific antigen. J Urol 2009; 183: 520.
19. Fradet V, Kurhanewicz J, Cowan JE et al: Prostate cancer managed with active surveillance: role of anatomic MR imaging and MR spectroscopic imaging. Radiology 2010; 256: 176.
22. Lemaitre L, Puech P, Poncelet E et al: Dynamic contrast-enhanced MRI of anterior prostate cancer: morphometric assessment and correlation with radical prostatectomy findings. Eur Radiol 2009; 19: 470.
23. Ploussard G, Xylinas E, Durand X et al: Magnetic resonance imaging does not improve the prediction of misclassification of prostate cancer patients eligible for active surveillance when the most stringent selection criteria are based on the saturation biopsy scheme. BJU Int 2011; 108: 513. 24. Neill MG, Toi A, Lockwood GA et al: Systematic lateral prostate biopsy—are the benefits worth the costs? J Urol 2008; 179: 1321. 25. Nam RK, Saskin R, Lee Y et al: Increasing hospital admission rates for urological complications after transrectal ultrasound guided prostate biopsy. J Urol 2010; 183: 963.
Purpose: We report magnetic resonance imaging findings among unselected men with low risk prostate cancer before active surveillance. Materials and Methods: We prospectively enrolled men with low grade, low risk, localized prostate cancer. All patients underwent multiparametric endorectal coil magnetic resonance imaging and were offered confirmatory biopsy within 1 year of imaging. The primary outcome was the impact of magnetic resonance imaging on identifying patients who were reclassified by confirmatory biopsy as no longer fulfilling active surveillance criteria. We further identified clinical parameters associated with reclassification. The cohort was stratified as patients with 1) normal magnetic resonance imaging, 2) cancer on magnetic resonance imaging concordant with initial biopsy (less than 1 cm) and 3) cancer on magnetic resonance imaging larger than 1 cm. We performed univariate analysis to assess differences in clinical parameters among the groups. Results: Magnetic resonance imaging did not detect cancer in 23 cases (38%) while magnetic resonance imaging and initial biopsy were concordant in 24 (40%). Magnetic resonance imaging detected a 1 cm or larger lesion in 13 patients (22%). Of the cases 18 (32.14%) were reclassified. When no cancer was identified on magnetic resonance imaging, only 2 cases (3.5%) were reclassified. The positive and negative predictive values for magnetic resonance imaging predicting reclassification were 83% (95% CI 73–93) and 81% (95% CI 71–91), respectively. Prostate specific antigen density was increased in patients with lesions larger than 1 cm on magnetic resonance imaging compared to those with no cancer on imaging (median 0.15 vs 0.07 ng/ml/cc, p ⫽ 0.016). Conclusions: Magnetic resonance imaging appears to have a high yield for predicting reclassification among men who elect active surveillance. Upon confirmation of our results magnetic resonance imaging may be used to better select and guide patients before active surveillance.
Abbreviations and Acronyms ADC ⫽ apparent diffusion coefficient AS ⫽ active surveillance DCE ⫽ dynamic contrast enhanced DWI ⫽ diffusion-weighted imaging MRI ⫽ magnetic resonance imaging PCa ⫽ prostate cancer PSA ⫽ prostate specific antigen TRUS ⫽ transrectal ultrasound Submitted for publication August 15, 2011. Study received institutional review board approval. Supported by a Prostate Cancer Canada grant. Supplementary material for this article can be obtained at http://www.prostatecentre.ca. * Correspondence: UroOncology Fellowship Office, 610 University Ave., Toronto, Ontario, Canada, M5G 2M9 (telephone: 416-946-2000; FAX: 416-598-9997; e-mail: [email protected]). † Financial interest and/or other relationship with Merck, Novartis, Pfizer, AstraZeneca, GlaxoSmithKline and Sanofi-Aventis.
Key Words: prostate, prostatic neoplasms, magnetic resonance imaging, biopsy, diagnosis CURRENTLY most patients with PCa are diagnosed with favorable risk disease,1 which poses little risk to longevity.2 Traditional treatments such
as surgery or radiation are associated with significant adverse events and can negatively affect the quality of life of patients and their families.3 AS has
0022-5347/12/1874-1247/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 187, 1247-1252, April 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.11.112
AND
RESEARCH, INC.
www.jurology.com
1247
1248
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
the potential to decrease the overtreatment of patients who harbor clinically insignificant disease while offering curative treatment to those in whom disease is reclassified as higher risk after an observation period and repeat biopsy.4 – 6 Large cohorts showed a good outcome with this treatment paradigm5,7 A troubling aspect of AS is that a subset of patients diagnosed with low risk disease appears to be under sampled and in fact harbor larger, often higher grade cancer. Increasing evidence suggests that early failure among men on AS is due to underappreciation of the amount of disease at baseline.8 –10 Fear of under sampling is a potential impediment to wider scale adoption of AS. Although TRUS biopsy remains the gold standard for diagnosing prostate cancer, MRI currently has the highest accuracy among radiological modalities to detect and stage prostate tumors.11,12 However, little is known of MRI among patients with PCa at low risk. We report our experience with prospective MRI in an unselected, consecutive group of men with low risk PCa before enrollment in an AS protocol.
PATIENTS AND METHODS General We analyzed patients initially screened by MRI to determine eligibility to enroll in a vitamin intervention trial (ClinicalTrials.gov NCT00744549). It is a phase IIa, placebo controlled, randomized trial of the combination of vitamin E, lycopene and selenium. The aim of the study is to determine the impact of this combination on MRI detected PCa among men on AS. As a byproduct, we present MRI results in these men.
Patient Population Upon receiving approval from our institutional review board we enrolled 60 consecutive patients who elected AS for recently diagnosed PCa. All patients provided informed consent. For study inclusion patients had to have low grade, low risk, localized PCa (clinical stage T1c-T2a, Gleason score 6 or less, no Gleason pattern 4 and serum PSA 10 ng/ml or less) and be considered suitable candidates for AS. No more than 3 cores on the entry biopsy could be positive for PCa and all biopsy cores had to have less than 50% involvement. Eligible candidates had to be between ages 50 and 80 years, and have a life expectancy of more than 5 years. Initial biopsy, which had to have at least 10 cores, was performed at Princess Margaret Hospital in 42 patients (75%). Patients were excluded from analysis if they had previously undergone a prostate related surgery or procedure since that may interfere with MRI analysis.
care, Milwaukee, Wisconsin) using a 4-channel phased array surface coil coupled to an endorectal coil. T2weighted MRI was done with 2 phases acquired before injection with 20 ml gadopentate-diethylenetetramine pentaacetic acid contrast agent at 4 ml per second, followed by a 20 ml saline flush using a Spectris MR power injection system (Medrad®). A total of 26 slices were obtained per phase. All MRI data sets were obtained at identical slice locations with a 3 mm slice thickness and no intersection gap. Image interpretation. Images were reviewed by a single experienced uroradiologist. All MRIs were done at least 6 weeks from biopsy to minimize artifact. Diagnostic features for malignancy were a low T2 signal in the peripheral zone, a relatively low ADC calculated from DWI, early enhancement and washout on DCE MRI. For the transitional zone a poorly defined nodule that distorted the normal architecture and had concordant abnormalities on DWI and DCE MRI was considered suspicious for malignancy.13–15
Confirmatory Biopsy Confirmatory biopsy was offered to all study participants and done within 1 year of diagnosis according to our AS protocol. Patients with MRI that showed an MRI lesion greater than 1 cm in any dimension were offered immediate confirmatory biopsy. We used a threshold of 1 cm since it represents a sphere of 0.5 cc, which is the Epstein criterion for clinically significant PCa.16 We used a clinically relevant cutoff to avoid unnecessary biopsy while still detecting most clinically relevant lesions. Prostate biopsies were TRUS guided and obtained within a year of PCa diagnosis so that correlation was possible between biopsy and imaging findings. Confirmatory biopsy consisted of a minimum 12 cores. It was done at Princess Margaret Hospital in all cases. In patients with a lesion on MRI additional cores were taken via a previously reported technique.10 Briefly, abnormal sites on MRI were annotated into zones and provided to the radiologist performing TRUS biopsy to target the biopsy. Magnetic resonance images appeared on the screen adjacent to the ultrasound, allowing real-time comparison of methods and zones of interest. All entry and secondary prostate biopsies were reviewed by dedicated uropathologists at our institution.
Study End Points The primary outcome was MRI accuracy for identifying patients reclassified as no longer fulfilling AS criteria by confirmatory biopsy. Reclassification occurred if confirmatory biopsy revealed any Gleason 7 or greater, 3 or more cores positive for PCa, or a highest tumor volume in any core of greater than 50%. We further identified any clinical parameters, such as patient age, prostate volume, PSA, PSA density, number of positive cores and percent of cancer in a core, associated with a lesion larger than 1 cm on MRI.
Magnetic Resonance Imaging
Statistical Analysis
Protocol. Axial-oblique, fast spin-echo, T2-weighted, echo planar DWI, and multi and DCE MRI were done on a 1.5 Tesla EchoSpeed or Excite HD MRI system (GE Health-
For analysis purposes the study population was stratified into 3 groups, including group 1—normal MRI, group 2— cancer characteristics on MRI concordant with initial
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
biopsy, defined as a lesion less than 1 cm in all dimensions, and group 3—an MRI lesion larger than 1 cm. Descriptive statistics were used to describe the clinical parameters of each group, such as age, PSA, prostate volume (calculated based on TRUS data), PSA density (calculated as PSA/prostate volume) and confirmatory biopsy data. Univariate analysis using parametric (1-way ANOVA) and nonparametric (the Kruskal-Wallis test) was done to assess differences in these clinical parameters among groups. When a test revealed a significant value, we used a post hoc test with the Bonferroni correction to determine statistically significant differences among the groups. The accuracy of MRI to predict reclassification on confirmatory biopsy was assessed with a contingency table. We calculated sensitivity, specificity, and positive and negative predictive values along with the corresponding 95% CIs.
RESULTS Figure 1 shows the study design. The cohort consisted of 60 consecutive patients. MRI did not detect cancer in 23 patients (38%) while MRI and initial biopsy were concordant in 24 (40%). MRI detected a larger than 1 cm lesion in 13 patients (22%). Data on confirmatory biopsy were missing in 4 patients, of whom 3 refused and 1 was lost to followup. A total of 56 patients were included in further analysis. Overall 18 of the 56 patients (32.14%) in our cohort were reclassified. The reclassification rate by confirmatory biopsy was significantly higher among those with a large lesion on MRI compared to that of any other group (17.85% vs 10.71% and 3.5%, respectively, p ⫽ 0.02).
1249
The specificity and sensitivity of MRI compared to confirmatory biopsy results were 95% (95% CI 82– 99) and 55% (95% CI 43– 67), respectively. Falsepositive MRI was observed in 2 patients (3.5%) while 8 (14.28%) had a false-negative prediction. Calculated positive and negative predictive values were 83% (95% CI 73–93) and 81% (95% CI 71–91), respectively. The table lists baseline clinical parameters stratified by MRI findings. On univariate analysis PSA density was increased in patients with a lesion larger than 1 cm on MRI compared to those with normal MRI (0.15 vs 0.07 ng/ml/cc, p ⫽ 0.018). The table also shows differences in the confirmatory biopsy procedure and results. Time from initial biopsy was shorter for patients with a large MRI lesion compared to the 2 other groups (1.8 vs 10.3 and 9.6 months, respectively, p ⫽ 0.02). The mean ⫾ SD number of cores taken on confirmatory biopsy was greater in each group with an MRI detected lesion compared to the normal MRI group (16.2 ⫾ 0.9 and 15.3 ⫾ 1.8, respectively, vs 12.3 ⫾ 2, p ⫽ 0.041). TRUS identified a suspicious lesion in 5 patients and these lesions were also biopsied. TRUS identified lesions were concordant with MRI in 2 cases in which the lesions were cancerous. Other TRUS lesions were benign. MRI findings were anatomically concordant with confirmatory biopsy in 92% of MRI identified lesions. Of lesions larger than 1 cm 55% were located in the anterior zone while the others were in the peripheral (27%) or the transitional (27%) zone.
Figure 1. Study schematic shows 60 patients who underwent MRI and 3 predefined groups, including 1—normal MRI, 2— cancer characteristics on MRI concordant with initial biopsy (Bx) and 3—lesion on MRI more than 1 cm in any dimension.
1250
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
Clinical characteristics and confirmatory biopsy data
Mean ⫾ SD age Median ng/ml PSA (range) Median cc prostate vol (range) Median No. biopsy 1 pos cores (range) Mean ⫾ SD % largest biopsy 1 core PCa Median ng/ml/cc PSA density (range) Confirmatory biopsy data: Median mos from biopsy 1 (range) Mean ⫾ SD No. biopsy 2 cores % TRUS identified Ca (No. pts) % MRI-biopsy anatomical concordance Mean ⫾ SD % biopsy 2 Gleason 7 or greater (No. pts)
MRI Greater Than 1 cm PCa
MRI Concordant With Initial Biopsy
Normal MRI
p Value
62.6 ⫾ 7 5.94 (1.7–10) 37.5 (22–67) 2 (1–3) 12.1 ⫾ 7.1 0.15 (0.6–0.05)
63.5 ⫾ 6 4.4 (1.1–9.1) 37.5 (19–76) 2 (1–3) 15.3 ⫾ 8.2 0.13 (0.22–0.03)
64 ⫾ 8.2 4.1 (1.1–9.9) 41 (21–83) 2 (1–3) 13.6 ⫾ 5.1 0.07 (0.21–0.03)
0.6 0.07 0.09 0.5 0.3 0.04
1.8 (1.2–2.4) 16.2 ⫾ 0.9 5.3 (3) 100 9 (5)
10.3 (6–12) 15.3 ⫾ 1.8 1.7 (1) 83 5.3 (3)
9.6 (7–12) 12.3 ⫾ 0.7 1.7 (1) Not recorded 1.7 (1)
0.02 0.041 0.052 0.2 0.047
Gleason 7 cancers were identified in 9 patients, including 5 with large MRI detectable lesions and only 1 with normal prostate MRI (p ⫽ 0.047). Figure 2 shows a large anterior tumor that was missed on initial diagnosis but captured well on MRI.
DISCUSSION The most important finding of our study is that MRI may reveal an unrecognized significant lesion in up to 22% of patients and biopsy of these areas reclassified 17.85% of cases as no longer fulfilling conservative criteria for AS. Also, when no cancerous lesion was identified on MRI, the chance of reclassification on repeat biopsy was extremely low at 3.5%. To our knowledge we are the first to report the positive and negative predictive values of MRI to predict reclassification (83%, 95% CI 73–93 and 81%, 95% CI 71–91, respectively) in an unselected group of patients before AS. Several earlier studies examined MRI in patients on AS. Cabrera et al retrospectively reported on 92 men with biopsy proven PCa.17 They found no association among clinical stage, Gleason, PSA and apparent tumor on endorectal MRI. In another retrospective analysis van As et al noted that low ADC was associated with adverse histology on repeat biopsy.18 Tumor ADC correlated with maximum core
Figure 2. MRI reveals large anterior tumor missed on initial diagnosis. Red dotted outline indicates tumor.
involvement, the percent of positive cores, initial PSA and the free-to-total PSA ratio. Our findings are also supported by those of Fradet et al.19 They reported that in the context of AS a lesion suggesting cancer on MRI may confer a threefold increased risk of overall cancer progression. deSouza et al prospectively assessed 44 consecutive patients with clinically localized PCa by performing diffusion-weighted MRI in addition to standard T2-weighted MRI.20 Water diffusion in PCa was different between 26 patients with low risk PCa and 18 with intermediate or high risk disease. Thus, ADCs offer the potential to differentiate indolent from aggressive PCa. However, that study included patients who did not meet AS criteria. To our knowledge we are the first to report MRI findings done prospectively for all patients eligible for AS. The site of missed tumors is another factor to consider. In our study most missed large tumors were located in the anterior (55%) or the transitional (27%) zone. This was also documented in other series.21,22 MRI seems to be a valuable tool to detect these hidden cancers. The optimal MRI protocol before AS is not known. We used a multiparametric combination of T2weighted, echo planar, DWI, and multi and DCE MRI. We believe that any single technique would lack the sensitivity to detect PCa before AS since most patients harbor low volume cancer. Ploussard et al recently reported a retrospective analysis of 96 patients who met AS criteria and underwent T2weighted MRI.23 They concluded that MRI does not improve the prediction of reclassification. We believe that these discrepancies may have been due to the use of multiparametric MRI. Larger studies are needed to determine the best, most costeffective MRI protocol. Some AS candidates had normal MRIs. In our series only 2 such cases were reclassified. Cabrera et al concluded that MRI does not appear to have prognostic value for AS.17 However, they did not use DWI. Upon confirmation of our results patients with
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
normal MRI may be spared the pain and complications associated with repeat biopsy at 1 year. Our results should be viewed in the context of a tertiary referral center where there is expertise in prostate MRI and an interest in AS. We have a radiology team dedicated to prostate imaging that has experience with interpreting prostate MRI results, enabling true guidance for subsequent biopsies. Since MRI readings are observer dependent, these results may take time to duplicate at less experienced centers. A potential limitation is the comparison of targeted biopsy with TRUS guided biopsy. It is expected that biopsy to a suspicious lesion would yield a higher percent of positive cores and of cancer in each core. However, the clinical usefulness of MRI in this setting is to detect unexpected lesions and aid in targeting a biopsy. Thus, we believe that such a comparison is appropriate. Ideally additional random biopsies in the group with normal prostate MRI would be done to compensate for additional sampling. However, prior data suggest that the yield of this technique is minimal above the traditional 12 to 16 cores.24 Time to confirmatory biopsy was shorter among patients with a large MRI lesion. We thought that it was unethical not to biopsy patients with lesions greater than 1 cm yet almost unethical to rebiopsy patients with normal prostate MRI in light of increasing biopsy complications.25 However, since PCa grows slowly, we believe that a 1-year delay almost
1251
certainly captures missed initial lesions and not progression during followup. Also, not all initial biopsies were done locally. However, this percent did not differ among the study groups and, thus, should not have affected our results. Finally, our cohort was rather small. Further confirmation is needed in a larger, multicenter study. Another potential concern with our study is the validity of a 1 cm threshold for repeat biopsy. One cm represents a sphere of 0.5 cc, which is the Epstein criterion for clinically significant PCa.16 We specifically did not choose a 0.5 cm cutoff, considering the highly prevalent screening protocols and overtreatment. We believe that the purpose of extended diagnostic strategies would be to decrease the risk of misclassification and preclude the need for rebiopsy. Future study must address the criteria that trigger biopsy, including refined dimension thresholds and other potential MRI characteristics.
CONCLUSIONS This study highlights several important findings. MRI appears to have the potential to identify early and noninvasively candidates for surveillance who harbor occult adverse disease. In this series significant PCa was present in 17.58% of those eligible for AS. MRI can detect and accurately direct biopsy to these hidden cancers. PSA density may be used to help clinicians select cases appropriate for MRI. Upon confirmation MRI may be used to better select and guide patients before AS.
REFERENCES 1. Detailed guide: prostate cancer—what are the key statistics about prostate cancer? American Cancer Society 2009. Available at: http://www.cancer. org/docroot/CRI/content/CRI_2_4_1X_What_are_ the_key_statistics_for_prostate_cancer_36.asp. Accessed November 2011. 2. Jemal A, Siegel R, Ward E et al: Cancer statistics, 2008. CA Cancer J Clin 2008; 58: 71. 3. Sanda MG, Dunn RL, Michalski J et al: Quality of life and satisfaction with outcome among prostate-cancer survivors. N Engl J Med. 2008; 358: 1250. 4. Warlick C, Trock BJ, Landis P et al: Delayed versus immediate surgical intervention and prostate cancer outcome. J Natl Cancer Inst 2006; 98: 355. 5. Klotz L, Zhang L, Lam A et al: Clinical results of long-term follow-up of a large, active surveillance cohort with localized prostate cancer. J Clin Oncol 2010; 28: 126. 6. NCCN Guidelines® and Clinical Resources. Updated Prostate Cancer Guidelines, vs 4.2011. National Comprehensive Cancer Network. Available
at http://www.nccn.org/professionals/physician_ gls/f_guidelines.asp#site. Accessed November 2011. 7. Stattin P, Holmberg E, Johansson JE et al: Outcomes in localized prostate cancer: National Prostate Cancer Register of Sweden follow-up study. J Natl Cancer Inst 2010; 102: 950. 8. Klotz L: Active surveillance for prostate cancer: patient selection and management. Curr Oncol, suppl., 2010; 2: S11.
12. Weinreb JC, Blume JD, Coakley FV et al: Prostate cancer: sextant localization at MR imaging and MR spectroscopic imaging before prostatectomy— results of ACRIN prospective multi-institutional clinicopathologic study. Radiology 2009; 251: 122. 13. Haider MA, van der Kwast TH, Tanguay J et al: Combined T2-weighted and diffusion-weighted MRI for localization of prostate cancer. AJR Am J Roentgenol 2007; 189: 323.
9. Bott SR, Young MP, Kellett MJ et al: Anterior prostate cancer: is it more difficult to diagnose? BJU Int 2002; 89: 886.
14. Akin O, Sala E, Moskowitz CS et al: Transition zone prostate cancers: features, detection, localization, and staging at endorectal MR imaging. Radiology 2006; 239: 784.
10. Lawrentschuk N, Haider MA, Daljeet N et al: ‘Prostatic evasive anterior tumours’: the role of magnetic resonance imaging. BJU Int 2010; 105: 1231.
15. Futterer JJ, Heijmink SW, Scheenen TW et al: Prostate cancer localization with dynamic contrast-enhanced MR imaging and proton MR spectroscopic imaging. Radiology 2006; 241: 449.
11. Kurhanewicz J, Vigneron D, Carroll P et al: Multiparametric magnetic resonance imaging in prostate cancer: present and future. Curr Opin Urol 2008; 18: 71.
16. Epstein JI, Walsh PC, Carmichael M et al: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 1994; 271: 368.
1252
PROSTATE MAGNETIC RESONANCE IMAGING AND DISEASE RECLASSIFICATION
17. Cabrera AR, Coakley FV, Westphalen AC et al: Prostate cancer: is in apparent tumor at endorectal MR and MR spectroscopic imaging a favorable prognostic finding in patients who select active surveillance? Radiology 2008; 247: 444.
20. deSouza NM, Riches SF, Vanas NJ et al: Diffusion-weighted magnetic resonance imaging: a potential noninvasive marker of tumour aggressiveness in localized prostate cancer. Clin Radiol 2008; 63: 774.
18. van As NJ, de Souza NM, Riches SF et al: A study of diffusion-weighted magnetic resonance imaging in men with untreated localised prostate cancer on active surveillance. Eur Urol 2008; 56: 981.
21. Hambrock T, Somford DM, Hoeks C et al: Magnetic resonance imaging guided prostate biopsy in men with repeat negative biopsies and increased prostate specific antigen. J Urol 2009; 183: 520.
19. Fradet V, Kurhanewicz J, Cowan JE et al: Prostate cancer managed with active surveillance: role of anatomic MR imaging and MR spectroscopic imaging. Radiology 2010; 256: 176.
22. Lemaitre L, Puech P, Poncelet E et al: Dynamic contrast-enhanced MRI of anterior prostate cancer: morphometric assessment and correlation with radical prostatectomy findings. Eur Radiol 2009; 19: 470.
23. Ploussard G, Xylinas E, Durand X et al: Magnetic resonance imaging does not improve the prediction of misclassification of prostate cancer patients eligible for active surveillance when the most stringent selection criteria are based on the saturation biopsy scheme. BJU Int 2011; 108: 513. 24. Neill MG, Toi A, Lockwood GA et al: Systematic lateral prostate biopsy—are the benefits worth the costs? J Urol 2008; 179: 1321. 25. Nam RK, Saskin R, Lee Y et al: Increasing hospital admission rates for urological complications after transrectal ultrasound guided prostate biopsy. J Urol 2010; 183: 963.