1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Gleason Score Determination with Transrectal Ultrasound-Magnetic Resonance Imaging Fusion Guided Prostate Biopsiesd Are We Gaining in Accuracy? de ric Beuvon,* Arnaud Lefe vre,* Camille Lanz,* Franc¸ois Cornud,* Fre Paul Legmann,* Marc Zerbib* and Nicolas Barry Delongchamps†,‡ From the Department of Urology (CL, MZ, NBD), Radiology (FC, AL, PL) and Pathology (FB), Cochin Hospital, and Inserm U1154, Institut Necker-Enfants Malades (NBD), Paris Descartes University, Paris, France
Purpose: We evaluated the accuracy of prostate magnetic resonance imagingtransrectal ultrasound targeted biopsy for Gleason score determination. Materials and Methods: We selected 125 consecutive patients treated with radical prostatectomy for a clinically localized prostate cancer diagnosed on magnetic resonance imaging-transrectal ultrasound targeted biopsy and/or systematic biopsy. On multiparametric magnetic resonance imaging each suspicious area was graded according to PI-RADSÔ score. A correlation analysis between multiparametric magnetic resonance imaging and pathological findings was performed. Factors associated with determining the accuracy of Gleason score on targeted biopsy were statistically assessed. Results: Pathological analysis of radical prostatectomy specimens detected 230 tumor foci. Multiparametric magnetic resonance imaging detected 151 suspicious areas. Of these areas targeted biopsy showed 126 cancer foci in 115 patients, and detected the index lesion in all of them. The primary Gleason grade, secondary Gleason grade and Gleason score of the 126 individual tumors were determined accurately in 114 (90%), 75 (59%) and 85 (67%) cases, respectively. Maximal Gleason score was determined accurately in 80 (70%) patients. Gleason score determination accuracy on targeted biopsy was significantly higher for low Gleason and high PI-RADS score tumors. Conclusions: Magnetic resonance imaging-transrectal ultrasound targeted biopsy allowed for an accurate estimation of Gleason score in more than two-thirds of patients. Gleason score misclassification was mostly due to a lack of accuracy in the determination of the secondary Gleason grade. Key Words: prostatic neoplasms, biopsy, magnetic resonance imaging, neoplasm grading
GLEASON score is one of the most relevant pretreatment criteria for patient counseling and decision making. Its reliability is crucial when proposing less aggressive treatment alternatives such as active surveillance, focal treatment or brachytherapy. Multiple reports have
highlighted that its determination by systematic biopsies was achieved with a limited accuracy, with underestimation reported in approximately a third of patients.1e4 These studies also suggested great variability in Gleason score underestimation, especially in cases of Gleason score
0022-5347/16/1951-0001/0 THE JOURNAL OF UROLOGY® Ó 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
Dochead: Adult Urology
http://dx.doi.org/10.1016/j.juro.2015.07.021 Vol. 195, 1-6, January 2016 Printed in U.S.A.
Abbreviations and Acronyms DCE ¼ dynamic contrast enhanced DW ¼ diffusion weighted mp ¼ multiparametric MRI ¼ magnetic resonance imaging RP ¼ radical prostatectomy T2W ¼ T2-weighted TB ¼ targeted biopsy TRUS ¼ transrectal ultrasound Accepted for publication July 2, 2015. The corresponding author certifies that, when applicable, a statement(s) has been included in the manuscript documenting institutional review board, ethics committee or ethical review board study approval; principles of Helsinki Declaration were followed in lieu of formal ethics committee approval; institutional animal care and use committee approval; all human subjects provided written informed consent with guarantees of confidentiality; IRB approved protocol number; animal approved project number. * No direct or indirect commercial incentive associated with publishing this article. † Correspondence: Service d’urologie, h^opital Cochin, 27, rue du Faubourg Saint Jacques, 75014 Paris (telephone: 01 58 41 27 64; FAX: 01 58 41 27 85). ‡ Financial interest and/or other relationship with Koelis.
www.jurology.com
FLA 5.4.0 DTD JURO12755_proof 29 September 2015 1:32 pm EO: JU-15-831
j
1
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
ACCURACY OF GLEASON SCORE DETERMINATION
2
115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171
3þ3 cancer on biopsy, which can be upgraded to Gleason score 3þ4 on RP specimen in 20% to 66% of cases.1 In cases of Gleason score 3þ4 cancer on biopsy, upgrading to 4þ3 cancer on RP specimen was reported in approximately 25% of cases.1 Several authors attempted to address this issue by proposing transrectal5 or transperineal6 saturation biopsies, considering that increased tissue sampling and more exhaustive mapping of the prostate would improve the accuracy of Gleason score determination. However, even the most invasive saturation biopsy protocols using more than 50 template guided (5 mm grid) transperineal cores still misclassify Gleason score in almost a third of the cases.7 The precise knowledge of tumor location using imaging represents, at least theoretically, the best way to accurately target suspicious foci and, thus, determine Gleason score with optimal accuracy. Multiparametric MRI, combining T2W, DW and DCE MRI, is now an established imaging modality to localize prostatic tumor foci larger than 0.2 cm3.8e11 PI-RADS was developed in 2012 for the detection of clinically significant tumor foci. When MRI is performed before biopsy, suspicious areas can be targeted to take additional cores during standard biopsy. Also, software based MRI-TRUS image registration systems now enable the accurate performance of these magnetic resonance targeted biopsies.12e14 Another potential advantage of MRI-TRUS fusion TB suggested in the recent literature is the improved characterization of the index tumor.15,16 Detecting appropriately the more aggressive foci in the prostate should hypothetically improve the accuracy in Gleason score determination. In this setting Gleason score 7 may receive more attention. Studies comparing biopsy and true (on RP specimen) Gleason score use a binary stratification of Gleason score 7, 3þ4 vs 4þ3. Although the poorer prognosis of Gleason score 4þ3 vs 3þ4 has been demonstrated in several studies,17,18 it may be more relevant to consider that the percentage of grade 4 is a more useful discriminant tool to evaluate tumor aggressiveness. In this study we evaluated the accuracy of MRI-TRUS fusion TB for Gleason score determination, incorporating the percentage of grade 4 as a continuous pattern in Gleason score 7 tumors.
PATIENTS AND METHODS Patient Inclusion We recently conducted a correlation analysis between MRI-TRUS TB and RP specimens in 125 consecutive patients, and described the characteristics of tumor foci missed with a TB strategy alone.16 These patients form Dochead: Adult Urology
the basis of the present study. The study was approved by the institutional review board, which issued a waiver of informed consent for the review of clinical, biological, histological and MRI data.
Magnetic Resonance Imaging MRI was performed using a 1.5T scanner with integrated endorectal and pelvic phased array coils. The endorectal coil was inserted and inflated with air to a volume of approximately 80 to 100 ml. The diffusion weighted and DCE images had the same orientation as the transverse T2W images. DCE images were evaluated qualitatively on a cine loop of subtracted gradient echo images, and regions of interest were drawn on foci showing an early and focal enhancement to display the curve type of the kinetics of gadolinium. A typical suspicious lesion was well circumscribed and of low signal intensity on T2W imaging, showing restricted diffusion and early and intense enhancement with rapid washout on DCE imaging. Each suspicious area was further characterized according to the European Society of Urogenital Radiology PI-RADS Likert-like global score, as score 1dclinically significant disease highly unlikely to be present, score 2dclinically significant cancer unlikely to be present, score 3dpresence of clinically significant cancer is equivocal, score 4dclinically significant cancer likely present, and score 5dclinically significant cancer highly likely present.19
Prostate Biopsies All patients underwent 10 to 12-core random systematic biopsies. At least 2 additional TBs were performed in the suspicious areas detected on mp-MRI using an elastic MRI-TRUS image registration system (KoelisÒ).
Histological Evaluation All biopsy cores were individually labeled. The number of cores involved with cancer, the total length of tissue sampled, total length of cancer detected and Gleason score were determined. The overall percentage of Gleason 4 in each target was calculated by averaging the percentage of grade 4 in each targeted core. After radical prostatectomy the glands were cut into 4 mm sections perpendicular to the posterior plane, including apex and base, according to a modified Stanford protocol. Paraffin embedded blocks were cut to produce 5 mm whole mount sections, and were stained with hematoxylin and eosin. All slides were digitized with a high resolution scanner (Hamamatsu, Japan). The total number of tumor foci and their locations were recorded. If the distance between 2 tumor foci was greater than 4.5 mm, they were considered separate.20 Tumor foci were graded according to the modified Gleason grading system and the percentage of grade 4 was recorded.21 Pathological stage was determined according to the 2002 TNM classification.22 Tumor volume was calculated by computerized planimetry. Tumor foci were considered clinically insignificant if they were organ confined with a volume less than 0.5 cc and Gleason score 6 or less.
Measurements and Statistical Analysis Each pathological whole mount section previously digitized was matched to the corresponding MRI on the basis
FLA 5.4.0 DTD JURO12755_proof 29 September 2015 1:32 pm EO: JU-15-831
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228
ACCURACY OF GLEASON SCORE DETERMINATION
of the location of the ejaculatory ducts, diameter of the 229 prostate and approximate distance from the base or apex. 230 The limit between anterior and posterior gland was 231 defined as the transverse section running through the 232 urethra. A correlation analysis between mp-MRI findings 233 and TB results and whole mount pathological analysis 234 was then performed for each individual tumor focus. 235 This analysis was conducted by 2 radiologists and 236 1 pathologist who interpreted images by consensus. 237 Population description was performed using proportions 238 for categorical data, mean and standard error of the mean 239 for normally distributed data, and median and range for other quantitative data. Factors associated with the ac240 curacy of Gleason score determination on TB were 241 assessed with stepwise logistic regression. Analyses were 242 conducted using MedCalcÒ software with p
Gleason Score Determination with Transrectal Ultrasound-Magnetic Resonance Imaging Fusion Guided Prostate Biopsiesd Are We Gaining in Accuracy? de ric Beuvon,* Arnaud Lefe vre,* Camille Lanz,* Franc¸ois Cornud,* Fre Paul Legmann,* Marc Zerbib* and Nicolas Barry Delongchamps†,‡ From the Department of Urology (CL, MZ, NBD), Radiology (FC, AL, PL) and Pathology (FB), Cochin Hospital, and Inserm U1154, Institut Necker-Enfants Malades (NBD), Paris Descartes University, Paris, France
Purpose: We evaluated the accuracy of prostate magnetic resonance imagingtransrectal ultrasound targeted biopsy for Gleason score determination. Materials and Methods: We selected 125 consecutive patients treated with radical prostatectomy for a clinically localized prostate cancer diagnosed on magnetic resonance imaging-transrectal ultrasound targeted biopsy and/or systematic biopsy. On multiparametric magnetic resonance imaging each suspicious area was graded according to PI-RADSÔ score. A correlation analysis between multiparametric magnetic resonance imaging and pathological findings was performed. Factors associated with determining the accuracy of Gleason score on targeted biopsy were statistically assessed. Results: Pathological analysis of radical prostatectomy specimens detected 230 tumor foci. Multiparametric magnetic resonance imaging detected 151 suspicious areas. Of these areas targeted biopsy showed 126 cancer foci in 115 patients, and detected the index lesion in all of them. The primary Gleason grade, secondary Gleason grade and Gleason score of the 126 individual tumors were determined accurately in 114 (90%), 75 (59%) and 85 (67%) cases, respectively. Maximal Gleason score was determined accurately in 80 (70%) patients. Gleason score determination accuracy on targeted biopsy was significantly higher for low Gleason and high PI-RADS score tumors. Conclusions: Magnetic resonance imaging-transrectal ultrasound targeted biopsy allowed for an accurate estimation of Gleason score in more than two-thirds of patients. Gleason score misclassification was mostly due to a lack of accuracy in the determination of the secondary Gleason grade. Key Words: prostatic neoplasms, biopsy, magnetic resonance imaging, neoplasm grading
GLEASON score is one of the most relevant pretreatment criteria for patient counseling and decision making. Its reliability is crucial when proposing less aggressive treatment alternatives such as active surveillance, focal treatment or brachytherapy. Multiple reports have
highlighted that its determination by systematic biopsies was achieved with a limited accuracy, with underestimation reported in approximately a third of patients.1e4 These studies also suggested great variability in Gleason score underestimation, especially in cases of Gleason score
0022-5347/16/1951-0001/0 THE JOURNAL OF UROLOGY® Ó 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
Dochead: Adult Urology
http://dx.doi.org/10.1016/j.juro.2015.07.021 Vol. 195, 1-6, January 2016 Printed in U.S.A.
Abbreviations and Acronyms DCE ¼ dynamic contrast enhanced DW ¼ diffusion weighted mp ¼ multiparametric MRI ¼ magnetic resonance imaging RP ¼ radical prostatectomy T2W ¼ T2-weighted TB ¼ targeted biopsy TRUS ¼ transrectal ultrasound Accepted for publication July 2, 2015. The corresponding author certifies that, when applicable, a statement(s) has been included in the manuscript documenting institutional review board, ethics committee or ethical review board study approval; principles of Helsinki Declaration were followed in lieu of formal ethics committee approval; institutional animal care and use committee approval; all human subjects provided written informed consent with guarantees of confidentiality; IRB approved protocol number; animal approved project number. * No direct or indirect commercial incentive associated with publishing this article. † Correspondence: Service d’urologie, h^opital Cochin, 27, rue du Faubourg Saint Jacques, 75014 Paris (telephone: 01 58 41 27 64; FAX: 01 58 41 27 85). ‡ Financial interest and/or other relationship with Koelis.
www.jurology.com
FLA 5.4.0 DTD JURO12755_proof 29 September 2015 1:32 pm EO: JU-15-831
j
1
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114
ACCURACY OF GLEASON SCORE DETERMINATION
2
115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171
3þ3 cancer on biopsy, which can be upgraded to Gleason score 3þ4 on RP specimen in 20% to 66% of cases.1 In cases of Gleason score 3þ4 cancer on biopsy, upgrading to 4þ3 cancer on RP specimen was reported in approximately 25% of cases.1 Several authors attempted to address this issue by proposing transrectal5 or transperineal6 saturation biopsies, considering that increased tissue sampling and more exhaustive mapping of the prostate would improve the accuracy of Gleason score determination. However, even the most invasive saturation biopsy protocols using more than 50 template guided (5 mm grid) transperineal cores still misclassify Gleason score in almost a third of the cases.7 The precise knowledge of tumor location using imaging represents, at least theoretically, the best way to accurately target suspicious foci and, thus, determine Gleason score with optimal accuracy. Multiparametric MRI, combining T2W, DW and DCE MRI, is now an established imaging modality to localize prostatic tumor foci larger than 0.2 cm3.8e11 PI-RADS was developed in 2012 for the detection of clinically significant tumor foci. When MRI is performed before biopsy, suspicious areas can be targeted to take additional cores during standard biopsy. Also, software based MRI-TRUS image registration systems now enable the accurate performance of these magnetic resonance targeted biopsies.12e14 Another potential advantage of MRI-TRUS fusion TB suggested in the recent literature is the improved characterization of the index tumor.15,16 Detecting appropriately the more aggressive foci in the prostate should hypothetically improve the accuracy in Gleason score determination. In this setting Gleason score 7 may receive more attention. Studies comparing biopsy and true (on RP specimen) Gleason score use a binary stratification of Gleason score 7, 3þ4 vs 4þ3. Although the poorer prognosis of Gleason score 4þ3 vs 3þ4 has been demonstrated in several studies,17,18 it may be more relevant to consider that the percentage of grade 4 is a more useful discriminant tool to evaluate tumor aggressiveness. In this study we evaluated the accuracy of MRI-TRUS fusion TB for Gleason score determination, incorporating the percentage of grade 4 as a continuous pattern in Gleason score 7 tumors.
PATIENTS AND METHODS Patient Inclusion We recently conducted a correlation analysis between MRI-TRUS TB and RP specimens in 125 consecutive patients, and described the characteristics of tumor foci missed with a TB strategy alone.16 These patients form Dochead: Adult Urology
the basis of the present study. The study was approved by the institutional review board, which issued a waiver of informed consent for the review of clinical, biological, histological and MRI data.
Magnetic Resonance Imaging MRI was performed using a 1.5T scanner with integrated endorectal and pelvic phased array coils. The endorectal coil was inserted and inflated with air to a volume of approximately 80 to 100 ml. The diffusion weighted and DCE images had the same orientation as the transverse T2W images. DCE images were evaluated qualitatively on a cine loop of subtracted gradient echo images, and regions of interest were drawn on foci showing an early and focal enhancement to display the curve type of the kinetics of gadolinium. A typical suspicious lesion was well circumscribed and of low signal intensity on T2W imaging, showing restricted diffusion and early and intense enhancement with rapid washout on DCE imaging. Each suspicious area was further characterized according to the European Society of Urogenital Radiology PI-RADS Likert-like global score, as score 1dclinically significant disease highly unlikely to be present, score 2dclinically significant cancer unlikely to be present, score 3dpresence of clinically significant cancer is equivocal, score 4dclinically significant cancer likely present, and score 5dclinically significant cancer highly likely present.19
Prostate Biopsies All patients underwent 10 to 12-core random systematic biopsies. At least 2 additional TBs were performed in the suspicious areas detected on mp-MRI using an elastic MRI-TRUS image registration system (KoelisÒ).
Histological Evaluation All biopsy cores were individually labeled. The number of cores involved with cancer, the total length of tissue sampled, total length of cancer detected and Gleason score were determined. The overall percentage of Gleason 4 in each target was calculated by averaging the percentage of grade 4 in each targeted core. After radical prostatectomy the glands were cut into 4 mm sections perpendicular to the posterior plane, including apex and base, according to a modified Stanford protocol. Paraffin embedded blocks were cut to produce 5 mm whole mount sections, and were stained with hematoxylin and eosin. All slides were digitized with a high resolution scanner (Hamamatsu, Japan). The total number of tumor foci and their locations were recorded. If the distance between 2 tumor foci was greater than 4.5 mm, they were considered separate.20 Tumor foci were graded according to the modified Gleason grading system and the percentage of grade 4 was recorded.21 Pathological stage was determined according to the 2002 TNM classification.22 Tumor volume was calculated by computerized planimetry. Tumor foci were considered clinically insignificant if they were organ confined with a volume less than 0.5 cc and Gleason score 6 or less.
Measurements and Statistical Analysis Each pathological whole mount section previously digitized was matched to the corresponding MRI on the basis
FLA 5.4.0 DTD JURO12755_proof 29 September 2015 1:32 pm EO: JU-15-831
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228
ACCURACY OF GLEASON SCORE DETERMINATION
of the location of the ejaculatory ducts, diameter of the 229 prostate and approximate distance from the base or apex. 230 The limit between anterior and posterior gland was 231 defined as the transverse section running through the 232 urethra. A correlation analysis between mp-MRI findings 233 and TB results and whole mount pathological analysis 234 was then performed for each individual tumor focus. 235 This analysis was conducted by 2 radiologists and 236 1 pathologist who interpreted images by consensus. 237 Population description was performed using proportions 238 for categorical data, mean and standard error of the mean 239 for normally distributed data, and median and range for other quantitative data. Factors associated with the ac240 curacy of Gleason score determination on TB were 241 assessed with stepwise logistic regression. Analyses were 242 conducted using MedCalcÒ software with p
