- Email: [email protected]
PD61-08 DOES GLEASON SCORE AT THE SITE OF POSITIVE SURGICAL MARGIN PREDICT RECURRENCE FOLLOWING RADICAL PROSTATECTOMY?
THE JOURNAL OF UROLOGYâ
e1190
inter-reader agreement and diagnostic performance of qualitative descriptors versus a 5-point Likert scale for determination of EPE. METHODS: This was an IRB-approved, HIPAA-compliant, single-center, retrospective study with 3 experienced and 2 non-experienced readers. Men who underwent mpMRI of the prostate followed by radical prostatectomy between Nov/2015 and Jul/2016 were eligible. Whole-mount prostatectomy specimen processed with a 3D-printed, patient-specific mold for precise anatomical registration was the standard of reference. Reviewers chose one or more of 8 qualitative descriptors (e.g., capsular bulging, irregular margin) and, after a washout period, assigned a Likert score for the likelihood of EPE: 1, highly unlikely; 2, unlikely; 3, indeterminate; 4, likely; 5, highly likely. Reproducibility among reviewers was assessed with weighted kappa statistics (<0,no agreement; 0-0.20 slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.811 almost perfect). Cochran-Armitage Trend test was used to test the association bewteen pathology-proven EPE and MRI-based Likert score. RESULTS: Eighty men met eligibility criteria; mean age: 64 years, PSA: 8.0 ng/mL; prostate volume: 39.9 cc; Histologic index lesion size: 229 mm; Gleason score 3+4, 62.5%; 4+3, 37.5%. EPE was present in 40(50%) men. Qualitative descriptors had inconsistent reproducibility (kappa as low as 0.33 for experienced and -0.02 for inexperienced readers) and poor accuracy (as low as 0.41). Agreement was moderate for experienced (k¼0.56) and non-experienced (0.46) readers with the Likert scale. There was significant increase in the likelihood of EPE with higher Likert scores (Figure 1). CONCLUSIONS: A 5-point Likert scale improves inter-reader reproducibility and the diagnostic performance of mpMRI compared to qualitative descriptors of EPE, facilitating informed decision making, treatment planning and patient counseling.
Vol. 197, No. 4S, Supplement, Monday, May 15, 2017
RESULTS: The sensitivity and specificity to diagnose seminal vesicle invasion (SVI) on mpMRI were 43.8% and 95.4%, respectively. The negative predictive value was 78.9%. The sensitivity and specificity to diagnose extracapsular invasion (ECE) were 54.5% and 80.5%, respectively. The overall sensitivity and specificity of diagnosing pathological T3 or higher were 52.6% and 82.1%, respectively. Non-organconfined disease determined by mpMRI was significantly associated with positive surgical margin and pathological T3 disease on multivariate analysis. Preoperative adverse findings on mpMRI were significantly associated with performance of the non-nerve-sparing technique. CONCLUSIONS: The mpMRI did not show outstanding diagnostic accuracy relative to our expectations in predicting SVI or ECE preoperatively. However, adverse findings on preoperative mpMRI were significantly related with worse postoperative pathological outcomes as well as postoperative biochemical recurrence.
Source of Funding: None
PD61-08 Source of Funding: This investigation was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UL1TR001105.
PD61-07 ACCURACY OF PREOPERATIVE MULTIPARAMETRIC MAGNETIC RESONANCE IMAGING FOR PREDICTION OF UNFAVORABLE PATHOLOGY IN PATIENTS WITH LOCALIZED PROSTATE CANCER UNDERGOING RADICAL PROSTATECTOMY Hakmin Lee, Ohseong Kwon*, Seoul, Korea, Republic of; Sangchul Lee, Sung Kyu Hong, Seok-Soo Byun, Seongnam, Korea, Republic of; Hwang Gyun Jeon, Byong Chang Jeong, Seong Il Seo, Seong Soo Jeon, Han-Yong Choi, Hyun Moo Lee, Seoul, Korea, Republic of INTRODUCTION AND OBJECTIVES: We investigated the accuracy of multi-parametric MRI (mpMRI) for preoperative staging and its influence on the determination of neurovascular bundle sparing and disease prognosis in patients with localized prostate cancer. METHODS: We reviewed 1,045 patients who underwent radical prostatectomy with preoperative mpMRI at a single institution. Clinical local stages determined from mpMRI were correlated with preoperative and postoperative pathological outcomes.
DOES GLEASON SCORE AT THE SITE OF POSITIVE SURGICAL MARGIN PREDICT RECURRENCE FOLLOWING RADICAL PROSTATECTOMY? Goran Rac*, Lawrence Dagrosa, Laura Spruill, Thomas Keane, Charleston, SC INTRODUCTION AND OBJECTIVES: Multiple pathologic features have been shown to predict biochemical recurrence (BCR) in patients after radical prostatectomy (RP) for prostate cancer. While positive surgical margins (PSM) have been shown to increase the likelihood of BCR, little data exists on the clinical significance of the tumor Gleason grade at the site of PSM. This study aims to assess if the Gleason grade at the PSM is predictive of BCR, and whether its predictive value differs from that of other commonly referenced risk factors. METHODS: We performed a retrospective review of a prospectively maintained database of all patients who underwent RP at our institution from 2009 to 2015. We identified 403 patients, 58 (14.4%) of whom were noted to have PSM. These cases were reviewed by an attending Pathologist who assigned a Gleason grade (3, 4 or 5) to the tumor at the site of PSM. The predictive value for BCR was compared to that of final pathology Gleason score and presence of PSM alone. RESULTS: We found that 34.5% (20/58) of patients with PSM had BCR, which was greater than the overall BCR rate of 19.9% (80/ 403) (p < 0.0001). Patients with Gleason 4+ disease at the PSM had a higher BCR rate (57.9%, 11/19) compared to those with Gleason 3 (23.1%, 9/39, p ¼ 0.009) and those with a negative margin (17.4%, 60/ 345, p < 0.0001). Interestingly, patients with Gleason 3 at the PSM did not have a significantly higher BCR rate than those with a negative
THE JOURNAL OF UROLOGYâ
Vol. 197, No. 4S, Supplement, Monday, May 15, 2017
margin. Gleason grade at the PSM was an independent predictor of BCR compared to presence of PSM alone. In patients with Gleason sum 7 disease on traditional final pathology, those with Gleason 3 at the PSM had a lower BCR rate (25.0%, 8/32) compared to those with Gleason 4+ (58.8%, 10/17, p ¼ 0.02). CONCLUSIONS: Our data suggests that Gleason score at the site of PSM independently predicts BCR in patients following RP with accuracy similar to traditional pathologic staging. However, in patients with a PSM and Gleason 7 on traditional pathologic staging, the presence of Gleason 4 or tertiary 5 disease at the margin can serve as an independent predictor of BCR, relative to patients with Gleason 3 at the margin. Routine reporting of the Gleason score at the site of a positive surgical margin may aid in postoperative risk stratification following RP. Source of Funding: none
PD61-09 INDEX TUMOR VOLUME ON MRI AS A PREDICTOR OF PATHOLOGIC OUTCOMES FOLLOWING RADICAL PROSTATECTOMY Dordaneh Sugano*, Abhinav Sidana, Brian Calio, Sonia Gaur, Amit Jain, Mahir Maruf, Maria Merino, Peter Choyke, Baris Turkbey, Bradford Wood, Peter Pinto, Bethesda, MD INTRODUCTION AND OBJECTIVES: Tumor volume measured on radical prostatectomy (RP) specimen has been shown to be associated with adverse pathologic and oncologic outcomes; however, it is difficult to calculate and cannot contribute to preoperative decision making. Advances in imaging technology may facilitate the prediction of prostate cancer outcomes prior to surgery. In this study, we evaluated the predictive value of the index tumor volume (ITV) calculated from prostate MRI in analyzing adverse pathologic outcomes following RP in a higher risk cohort. METHODS: Clinical and pathologic data from a prospectively maintained, single-institution database were analyzed for patients who underwent 3T MRI prior to RP (without prior therapy) between 2007 and 2016, with an index tumor defined as a T2-visible lesion with the longest diameter. ITV was calculated from T2W MRI by multiplying length by width by depth by 0.52 to generate the volume of an ellipse. Adverse pathologic outcomes were determined on whole mount RP specimens, and defined as positive margins (PM), extracapsular extension (ECE), positive lymph nodes (LNI), and seminal vesicle invasion (SVI). Logistic regression was used to assess associations of clinical, imaging, and histopathological variables with adverse pathologic features. Receiver operating characteristic curves were used to characterize and compare ITV performance with Partin tables. RESULTS: In our study period, 464 patients met our inclusion criteria. In our cohort, median age and PSA were 60 years (IQR 10) and 6.21 ng/ml (IQR 6), and 24.4% were 00 high risk00 (Gleason 8-10) on biopsy . 15.6% of patients had PM, 23.5% ECE, 6.3% LNI, and 6.5% SVI. Patients with adverse outcomes were found to have larger median ITV (PM: 1.236cc vs 0.832cc, p¼0.045; ECE: 1.388cc vs 0.771cc, p<0.001; LNI: 2.750cc vs 0.801cc, p<0.001; SVI: 2.269cc vs 0.806cc, p<0.001). On multivariate analysis, ITV was found to be an independent predictor of ECE (OR: 1.211, p¼0.005), LNI (OR: 1.366, p<0.001), and SVI (OR: 1.305, p¼0.002), but not PM (OR: 1.052, p¼0.300). ITV alone and ITV+PSA were found to have predictive ability comparable to that of Partin tables (ECE: ITVAUC: 0.660 vs. ITV+PSAAUC:0.721 vs. PartinAUC: 0.717, LNI: ITVAUC: 0.802 vs. ITV+PSAAUC:0.881 vs. PartinAUC: 0.873, SVI: ITVAUC: 0.749 vs. ITV+PSAAUC:0.762 vs. PartinAUC: 0.806). CONCLUSIONS: We demonstrate that Index Tumor Volume measured on T2W MRI is an independent predictor of ECE, LNI, and SVI following RP. We believe this easily calculated preoperative marker provides additional prognostic information, particularly in higher risk cohorts. Source of Funding: This research was supported by the Intramural Research Program of the National Cancer Institute, NIH
e1191
PD61-10 DOES THE NUMBER OF GLEASON 6 POSITIVE CORES ON PROSTATE BIOPSY PREDICT ADVERSE PATHOLOGY AT RADICAL PROSTATECTOMY IN PATIENTS WHO ARE CANDIDATES FOR ACTIVE SURVEILLANCE? François AUDENET*, Emily VERTOSICK, Samson FINE, New York, NY; Rafael SANCHEZ-SALAS, Marc GALIANO, Eric BARRET, Xavier CATHELINEAU, François ROZET, Paris, France; James EASTHAM, Peter SCARDINO, Karim TOUIJER, New York, NY INTRODUCTION AND OBJECTIVES: Patients with low-risk prostate cancer (PCa) at diagnosis are at risk for upgrading or upstaging on radical prostatectomy (RP). The possibility of occult aggressive disease is a concern for active surveillance (AS). Accordingly, AS is commonly restricted to men with 3 or fewer positive cores. The goal of this study was to model the relationship between the number of positive cores and the risk of adverse pathology in order to determine a threshold number of cores for AS. METHODS: We identified a cohort of 1,820 men who underwent RP at Memorial Sloan Kettering Cancer Center (MSKCC) between January 2000 and August 2016, and had low risk PCa. A comparable cohort of 1,469 French patients treated between December 2004 and November 2012 was identified. Adverse pathology was defined as Gleason score 4+3, seminal vesicle invasion or lymph node involvement. The association between number of biopsy cores and the risk of adverse pathology was analyzed using locally weighted scatterplot smoothing. RESULTS: In the MSKCC cohort, 171 (9.4%) patients had adverse pathology at RP, compared to 48 (3.3%) in the French cohort. There was a small increase in the risk of adverse pathology with the number of positive cores: patients with 1 positive core had a 4% risk which increased to 8% for patients with 12 positive cores. The increase in risk was smooth, with no discontinuities suggesting an obvious threshold for AS eligibility (Figure 1). There were some differences between cohorts, with the change in risk with increasing cores being flatter in MSKCC cohort. However, even in the French cohort, there were not large differences in risk by number of cores. CONCLUSIONS: There is no number of positive cores threshold associated with a sharp increase in the risk of adverse pathology. Consequently, patients should not be excluded from AS solely based on the number of positive cores.
Source of Funding: None
e1190
inter-reader agreement and diagnostic performance of qualitative descriptors versus a 5-point Likert scale for determination of EPE. METHODS: This was an IRB-approved, HIPAA-compliant, single-center, retrospective study with 3 experienced and 2 non-experienced readers. Men who underwent mpMRI of the prostate followed by radical prostatectomy between Nov/2015 and Jul/2016 were eligible. Whole-mount prostatectomy specimen processed with a 3D-printed, patient-specific mold for precise anatomical registration was the standard of reference. Reviewers chose one or more of 8 qualitative descriptors (e.g., capsular bulging, irregular margin) and, after a washout period, assigned a Likert score for the likelihood of EPE: 1, highly unlikely; 2, unlikely; 3, indeterminate; 4, likely; 5, highly likely. Reproducibility among reviewers was assessed with weighted kappa statistics (<0,no agreement; 0-0.20 slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.811 almost perfect). Cochran-Armitage Trend test was used to test the association bewteen pathology-proven EPE and MRI-based Likert score. RESULTS: Eighty men met eligibility criteria; mean age: 64 years, PSA: 8.0 ng/mL; prostate volume: 39.9 cc; Histologic index lesion size: 229 mm; Gleason score 3+4, 62.5%; 4+3, 37.5%. EPE was present in 40(50%) men. Qualitative descriptors had inconsistent reproducibility (kappa as low as 0.33 for experienced and -0.02 for inexperienced readers) and poor accuracy (as low as 0.41). Agreement was moderate for experienced (k¼0.56) and non-experienced (0.46) readers with the Likert scale. There was significant increase in the likelihood of EPE with higher Likert scores (Figure 1). CONCLUSIONS: A 5-point Likert scale improves inter-reader reproducibility and the diagnostic performance of mpMRI compared to qualitative descriptors of EPE, facilitating informed decision making, treatment planning and patient counseling.
Vol. 197, No. 4S, Supplement, Monday, May 15, 2017
RESULTS: The sensitivity and specificity to diagnose seminal vesicle invasion (SVI) on mpMRI were 43.8% and 95.4%, respectively. The negative predictive value was 78.9%. The sensitivity and specificity to diagnose extracapsular invasion (ECE) were 54.5% and 80.5%, respectively. The overall sensitivity and specificity of diagnosing pathological T3 or higher were 52.6% and 82.1%, respectively. Non-organconfined disease determined by mpMRI was significantly associated with positive surgical margin and pathological T3 disease on multivariate analysis. Preoperative adverse findings on mpMRI were significantly associated with performance of the non-nerve-sparing technique. CONCLUSIONS: The mpMRI did not show outstanding diagnostic accuracy relative to our expectations in predicting SVI or ECE preoperatively. However, adverse findings on preoperative mpMRI were significantly related with worse postoperative pathological outcomes as well as postoperative biochemical recurrence.
Source of Funding: None
PD61-08 Source of Funding: This investigation was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UL1TR001105.
PD61-07 ACCURACY OF PREOPERATIVE MULTIPARAMETRIC MAGNETIC RESONANCE IMAGING FOR PREDICTION OF UNFAVORABLE PATHOLOGY IN PATIENTS WITH LOCALIZED PROSTATE CANCER UNDERGOING RADICAL PROSTATECTOMY Hakmin Lee, Ohseong Kwon*, Seoul, Korea, Republic of; Sangchul Lee, Sung Kyu Hong, Seok-Soo Byun, Seongnam, Korea, Republic of; Hwang Gyun Jeon, Byong Chang Jeong, Seong Il Seo, Seong Soo Jeon, Han-Yong Choi, Hyun Moo Lee, Seoul, Korea, Republic of INTRODUCTION AND OBJECTIVES: We investigated the accuracy of multi-parametric MRI (mpMRI) for preoperative staging and its influence on the determination of neurovascular bundle sparing and disease prognosis in patients with localized prostate cancer. METHODS: We reviewed 1,045 patients who underwent radical prostatectomy with preoperative mpMRI at a single institution. Clinical local stages determined from mpMRI were correlated with preoperative and postoperative pathological outcomes.
DOES GLEASON SCORE AT THE SITE OF POSITIVE SURGICAL MARGIN PREDICT RECURRENCE FOLLOWING RADICAL PROSTATECTOMY? Goran Rac*, Lawrence Dagrosa, Laura Spruill, Thomas Keane, Charleston, SC INTRODUCTION AND OBJECTIVES: Multiple pathologic features have been shown to predict biochemical recurrence (BCR) in patients after radical prostatectomy (RP) for prostate cancer. While positive surgical margins (PSM) have been shown to increase the likelihood of BCR, little data exists on the clinical significance of the tumor Gleason grade at the site of PSM. This study aims to assess if the Gleason grade at the PSM is predictive of BCR, and whether its predictive value differs from that of other commonly referenced risk factors. METHODS: We performed a retrospective review of a prospectively maintained database of all patients who underwent RP at our institution from 2009 to 2015. We identified 403 patients, 58 (14.4%) of whom were noted to have PSM. These cases were reviewed by an attending Pathologist who assigned a Gleason grade (3, 4 or 5) to the tumor at the site of PSM. The predictive value for BCR was compared to that of final pathology Gleason score and presence of PSM alone. RESULTS: We found that 34.5% (20/58) of patients with PSM had BCR, which was greater than the overall BCR rate of 19.9% (80/ 403) (p < 0.0001). Patients with Gleason 4+ disease at the PSM had a higher BCR rate (57.9%, 11/19) compared to those with Gleason 3 (23.1%, 9/39, p ¼ 0.009) and those with a negative margin (17.4%, 60/ 345, p < 0.0001). Interestingly, patients with Gleason 3 at the PSM did not have a significantly higher BCR rate than those with a negative
THE JOURNAL OF UROLOGYâ
Vol. 197, No. 4S, Supplement, Monday, May 15, 2017
margin. Gleason grade at the PSM was an independent predictor of BCR compared to presence of PSM alone. In patients with Gleason sum 7 disease on traditional final pathology, those with Gleason 3 at the PSM had a lower BCR rate (25.0%, 8/32) compared to those with Gleason 4+ (58.8%, 10/17, p ¼ 0.02). CONCLUSIONS: Our data suggests that Gleason score at the site of PSM independently predicts BCR in patients following RP with accuracy similar to traditional pathologic staging. However, in patients with a PSM and Gleason 7 on traditional pathologic staging, the presence of Gleason 4 or tertiary 5 disease at the margin can serve as an independent predictor of BCR, relative to patients with Gleason 3 at the margin. Routine reporting of the Gleason score at the site of a positive surgical margin may aid in postoperative risk stratification following RP. Source of Funding: none
PD61-09 INDEX TUMOR VOLUME ON MRI AS A PREDICTOR OF PATHOLOGIC OUTCOMES FOLLOWING RADICAL PROSTATECTOMY Dordaneh Sugano*, Abhinav Sidana, Brian Calio, Sonia Gaur, Amit Jain, Mahir Maruf, Maria Merino, Peter Choyke, Baris Turkbey, Bradford Wood, Peter Pinto, Bethesda, MD INTRODUCTION AND OBJECTIVES: Tumor volume measured on radical prostatectomy (RP) specimen has been shown to be associated with adverse pathologic and oncologic outcomes; however, it is difficult to calculate and cannot contribute to preoperative decision making. Advances in imaging technology may facilitate the prediction of prostate cancer outcomes prior to surgery. In this study, we evaluated the predictive value of the index tumor volume (ITV) calculated from prostate MRI in analyzing adverse pathologic outcomes following RP in a higher risk cohort. METHODS: Clinical and pathologic data from a prospectively maintained, single-institution database were analyzed for patients who underwent 3T MRI prior to RP (without prior therapy) between 2007 and 2016, with an index tumor defined as a T2-visible lesion with the longest diameter. ITV was calculated from T2W MRI by multiplying length by width by depth by 0.52 to generate the volume of an ellipse. Adverse pathologic outcomes were determined on whole mount RP specimens, and defined as positive margins (PM), extracapsular extension (ECE), positive lymph nodes (LNI), and seminal vesicle invasion (SVI). Logistic regression was used to assess associations of clinical, imaging, and histopathological variables with adverse pathologic features. Receiver operating characteristic curves were used to characterize and compare ITV performance with Partin tables. RESULTS: In our study period, 464 patients met our inclusion criteria. In our cohort, median age and PSA were 60 years (IQR 10) and 6.21 ng/ml (IQR 6), and 24.4% were 00 high risk00 (Gleason 8-10) on biopsy . 15.6% of patients had PM, 23.5% ECE, 6.3% LNI, and 6.5% SVI. Patients with adverse outcomes were found to have larger median ITV (PM: 1.236cc vs 0.832cc, p¼0.045; ECE: 1.388cc vs 0.771cc, p<0.001; LNI: 2.750cc vs 0.801cc, p<0.001; SVI: 2.269cc vs 0.806cc, p<0.001). On multivariate analysis, ITV was found to be an independent predictor of ECE (OR: 1.211, p¼0.005), LNI (OR: 1.366, p<0.001), and SVI (OR: 1.305, p¼0.002), but not PM (OR: 1.052, p¼0.300). ITV alone and ITV+PSA were found to have predictive ability comparable to that of Partin tables (ECE: ITVAUC: 0.660 vs. ITV+PSAAUC:0.721 vs. PartinAUC: 0.717, LNI: ITVAUC: 0.802 vs. ITV+PSAAUC:0.881 vs. PartinAUC: 0.873, SVI: ITVAUC: 0.749 vs. ITV+PSAAUC:0.762 vs. PartinAUC: 0.806). CONCLUSIONS: We demonstrate that Index Tumor Volume measured on T2W MRI is an independent predictor of ECE, LNI, and SVI following RP. We believe this easily calculated preoperative marker provides additional prognostic information, particularly in higher risk cohorts. Source of Funding: This research was supported by the Intramural Research Program of the National Cancer Institute, NIH
e1191
PD61-10 DOES THE NUMBER OF GLEASON 6 POSITIVE CORES ON PROSTATE BIOPSY PREDICT ADVERSE PATHOLOGY AT RADICAL PROSTATECTOMY IN PATIENTS WHO ARE CANDIDATES FOR ACTIVE SURVEILLANCE? François AUDENET*, Emily VERTOSICK, Samson FINE, New York, NY; Rafael SANCHEZ-SALAS, Marc GALIANO, Eric BARRET, Xavier CATHELINEAU, François ROZET, Paris, France; James EASTHAM, Peter SCARDINO, Karim TOUIJER, New York, NY INTRODUCTION AND OBJECTIVES: Patients with low-risk prostate cancer (PCa) at diagnosis are at risk for upgrading or upstaging on radical prostatectomy (RP). The possibility of occult aggressive disease is a concern for active surveillance (AS). Accordingly, AS is commonly restricted to men with 3 or fewer positive cores. The goal of this study was to model the relationship between the number of positive cores and the risk of adverse pathology in order to determine a threshold number of cores for AS. METHODS: We identified a cohort of 1,820 men who underwent RP at Memorial Sloan Kettering Cancer Center (MSKCC) between January 2000 and August 2016, and had low risk PCa. A comparable cohort of 1,469 French patients treated between December 2004 and November 2012 was identified. Adverse pathology was defined as Gleason score 4+3, seminal vesicle invasion or lymph node involvement. The association between number of biopsy cores and the risk of adverse pathology was analyzed using locally weighted scatterplot smoothing. RESULTS: In the MSKCC cohort, 171 (9.4%) patients had adverse pathology at RP, compared to 48 (3.3%) in the French cohort. There was a small increase in the risk of adverse pathology with the number of positive cores: patients with 1 positive core had a 4% risk which increased to 8% for patients with 12 positive cores. The increase in risk was smooth, with no discontinuities suggesting an obvious threshold for AS eligibility (Figure 1). There were some differences between cohorts, with the change in risk with increasing cores being flatter in MSKCC cohort. However, even in the French cohort, there were not large differences in risk by number of cores. CONCLUSIONS: There is no number of positive cores threshold associated with a sharp increase in the risk of adverse pathology. Consequently, patients should not be excluded from AS solely based on the number of positive cores.
Source of Funding: None