Review Article Recent advances in prostate cancer pathology: Gleason grading and beyond

Pathology International 2016; 66: 260 272 doi:10.1111/pin.12398 Review Article Recent advances in prostate cancer pathology: Gleason grading and beyond Rajal B Shah 1 and Ming Zhou 2 1 Division of Urologic Pathology, Miraca Research Institute, Miraca Life Sciences, Irving, Texas and 2 Department of Pathology, New York University Langone Medical Center, New York, New York Since its initial description in 1966 by Dr Donald Gleason, Gleason grading has become the cornerstone in the management of prostate cancer (PCa). With widespread use of Prostate Specific Antigen (PSA) screening and needle core biopsy, the diagnosis and management of PCa have dramatically evolved. In addition, better understanding of the morphological spectrum of prostate cancer and its clinical significance have prompted the refinement of the grading criteria and reporting guidelines commensurate to contemporary practice. The modification of the Gleason grading system implemented by the International Society of Urological Pathology in 2005 and subsequent revision in 2014 has profoundly impacted how PCa is graded and managed. This review aims to provide a concise update on the refinement of the histological criteria for various Gleason patterns and problem areas of Gleason grading, and provide recommendations on how to improve the grading reproducibility. The new proposal to group Gleason scores into clinically meaningful grade groups will also be discussed. Finally, we will discuss how magnetic resonance imaging (MRI)-targeted biopsy and emerging genetic markers may help improve the Gleason grading accuracy and risk stratification currently based on clinicopathological parameters. Key words: Gleason patterns, Gleason score, Grade groups, Intraductal carcinoma of the prostate, ISUP Modified Gleason grading system, multi-gene assays, Magnetic Resonance Imaging (MRI)- targeted prostate biopsy, Prostate cancer, PTEN BACKGROUND The Gleason grading system developed by Dr. Donald Gleason in 1966, remains the cornerstone for the management of prostate cancer (PCa). 1,3 The system is relatively simple and reasonably reproducible to apply. It is one of the key parameters Correspondence: Ming Zhou, MD, PhD, Department of Pathology, New York University Langone Medical Center, 560 First Avenue, TCH-461, New York, NY 10016, USA. E-mail: Ming.zhou@nyumc.org Received 23 November 2015. Accepted for publication 3 February 2016. 2016 Japanese Society of Pathology and John Wiley & Sons Australia, Ltd for therapy-planning (active surveillance vs definitive therapy), and remains as the most important prognostic factor in predicting pathological findings in radical prostatectomy, biochemical failure, local and distant metastasis after therapy and PCa specific mortality. 4 8 Modifications implemented by International Society of Urological Pathology (ISUP) Grading Consensus The Gleason grading system has undergone continuous modification and changes in response to changes in the clinical practice of diagnosis and treatment of prostate cancer since its inception. The most significant changes were introduced in 2005 at the auspices of ISUP and further modification also ensued in 2014 (Fig. 1). The resulting grading system is referred to as the ISUP modified Gleason grading system. 9,10 However, it is important to stress that the changes put forth by ISUP simply codified what have already been used in practice by many pathologists. 9 It is important for surgical pathologists to be acquainted with and apply the modified grading criteria in their practice. Some of the changes are definitional, including precise definition of each Gleason pattern and grading criteria for PCa morphological variants. Others are operational, i.e., how to report Gleason grade in special circumstances, including reporting secondary patterns of lower or higher grade when present to a limited extent, tertiary patterns in both biopsy and prostatectomy specimen, and so forth. It is important to keep in mind that one should rely on the architectural patterns observed at the low to medium power to determine the grade of a tumor. In particular, for a diagnosis of Gleason pattern 4, it needs to be visualized at 10X lens magnification. 9 Definitional changes Gleason patterns 1 and 2 The most important change is perhaps the strict definition of each pattern. Due to its misleading clinical implications, a

Advances in prostate cancer pathology 261 is discussed in detail in the discussion of GP 4 and Table 1. Prostate cancer may demonstrate branching such as U or Y shaped glands, which is not considered as fusion and should be regarded as GP 3 9 (Fig. 2d). Finally, crowded microacinar GP 3 may mimic cribriform or fused GP 4 (Fig. 2e). Gleason pattern 4 Figure 1 2015 modified ISUP Gleason grading system (Reproduced with permission from Indiana University 9 ). Gleason score (GS) of 1 + 1 = 2 should not be rendered, regardless of the specimen type. 10 GS 2 4 should rarely be rendered in needle biopsies, if ever. 10,11 The major limitation of diagnosing GS 4 on needle biopsy is that one cannot see the entire edge of the lesion to determine if it is circumscribed. They rarely may be used in transurethral resection (TURP) and radical prostatectomy (RP) specimens. Therefore, from practical standpoint Gleason pattern (GP) in contemporary practice starts at 3 and GS starts at 6 in prostate biopsy specimens and most TURP and RP specimens. Gleason pattern 3 Pattern 3 represents the most common GP encountered in prostate needle biopsy (NBX). It is strictly defined as discrete, well-formed cancer glands. Variably sized individual glands infiltrating between benign glands represent the classical example of GP 3 (Fig. 2a). In contrast to the original description by Gleason, individual cells are no longer regarded as GP 3. 10 Minute foci of individual, infiltrating tumor glands seen frequently in NBXs represent GS 3 + 3 = 6 (Fig. 2b). Tangentially sectioned GP 3 glands may appear poorly formed and mimic pattern 4 (Fig. 2c). How to distinguish tangentially sectioned GP 3 glands from true poorly formed GP 4 The definition of GP 4 has evolved significantly since the inception of the Gleason grading system. It was first described as large clear cells growing in a diffuse pattern resembling hypernephroma 1. GP 4 was later expanded to include raggedly infiltrating, fused-glandular tumor and glands that are not single and separate, but coalesce and branch. 2 The 2005 ISUP consensus further expanded this GP to include ill-defined glands with poorly formed glandular lumens in addition to three previously recognized patterns: fused microacinar glands, large cribriform glands and cribriform glands with an irregular border, and hypernephromatoid cells. 10 The 2014 ISUP consensus recognizes cribriform, fused and poorly formed glands. 9 The term hypernephromatoid should not be used. 9 A major departure in the 2014 ISUP modified Gleason grading system from the original Gleason system is that all cribriform carcinomas without comedonecrosis are considered GP 4 912 The 2005 grading system allowed rounded, well circumscribed glands of the same size as normal glands as cribriform GP 3. 10 Subsequently, Latour et al surveyed 10 urological pathology experts with 30 NBX cases that possibly represented cribriform GP 3. 13 Most experts interpreted the cribriform patterns as GP 4. Since the separation of GP 3 and 4 represents the most important division point in Gleason grading, an upward change from pattern 3 to pattern 4 should be based on distinct architectural difference, i.e. individual well-formed glands vs cribriform glands with multiple lumens, rather than subjective evaluation of differences in size, shape and contour of cribriform glands. This approach is supported by several recent studies that link cribriform pattern to adverse pathological and clinical outcomes. By digitally quantifying five high-grade PCa patterns, Iczkowski et al. showed in multivariable analysis that the cribriform pattern had the highest odds ratio for PSA failure. 14 Both large and small cribriform patterns were significantly linked to failure and the cribriform pattern had particularly adverse implications for outcome. A recent study has further validated that cribriform growth was highly predictive for postoperative metastasis and disease-specific death in GS 7 PCa. 15 Based on these additional data, it was accepted by the 2014 ISUP grading consensus that all cribriform PCa, regardless of its shape, contour and size, should be interpreted as GP 4 and not pattern 3 (Fig. 3a). By 2005 ISUP consensus, clusters of poorly formed or illdefined glands where tangential sectioning is ruled out warrant a diagnosis of PG 4 10 (Fig. 3b). In contemporary practice, poorly formed glands is the most common GP 4 sub-pattern

262 R. B. Shah and M. Zhou Figure 2 Gleason pattern 3 prostate cancers. (a) Variably sized discrete glands infiltrating between benign glands represent classical Gleason pattern 3. (b) Minute focus of discrete cancer glands should be graded as 3 + 3 = 6. (c) Tangentially cut glands located immediately adjacent to other wellformed glands (arrows) mimic poorly formed glands Gleason pattern 4 prostate cancer. (d) Prostate cancer glands with U or Y shaped glands do not represent fusion and should be graded as pattern 3. (e) Gleason pattern 3 crowded microacinar cancer glands may mimic fused glands. Table 1 Histological features that are diagnosis for and against poorly formed Gleason pattern 4 by urological pathologists18 Only cancer glands with no or rare luminal formation may be considered poorly formed <= 5 poorly formed glands regardless of their location should not be graded as Gleason pattern 4; minimum # of poorly formed glands in a focus to be graded as Gleason pattern 4 is 6 Poorly formed glands immediately adjacent to other well-formed glands regardless of their number should not be grade as Gleason pattern 4 > 10 poorly formed glands that are not immediately adjacent to wellformed glands are considered Gleason pattern 4 * These recommendations reflect consensus opinion on grading poorly formed glands Gleason pattern 4 by 17 expert urological pathologists participating in an interobserver reproducibility study.18 They are not part of the 2014 ISUP Gleason grading consensus. in NBX.16 In a recent study of PCa diagnosed on NBX, poorly formed glands were identified in 57% of cases, followed by fused glands (53%) (Fig. 3c), cribriform glands (25%) and hypernephromatoid pattern (0.3%).16 Poorly formed GP 4 glands, however, suffer definitional ambiguity. The 2005 consensus stated that only a cluster of such glands, where a tangential section of Gleason pattern 3 glands cannot account for the histology, would be acceptable as Gleason pattern 4, but did not elaborate on how the possibility of tangential sectioning could be ruled out with reasonable certainty. Such a definitional ambiguity may account for the low interobserver reproducibility for the histological distinction of tangentially sectioned GP 3 from GP 4.17 We studied interobserver reproducibility of diagnosing poorly formed glands GP 4 on core needle biopsy among urologic pathologists.18 The overall reproducibility for the diagnosis was only fair (κ = 0.34). However, the diagnosis of poorly formed GP 4 was more reproducible when certain restrictive morphological and quantitative criteria were applied (Table 1). Only glands with no or barely visible lumens are considered poorly formed by expert urologic pathologists (Fig. 4a). Whether poorly formed glands should be graded as GP 4 depends on their quantitative and topographic morphological features. Poorly formed glands that are adjacent to other wellformed small glands likely result from tangential sectioning and therefore should not be graded as pattern 4 (Fig. 4b). Five or fewer poorly formed glands in a cluster regardless of their location should not be graded as GP 4 (Fig. 4c). In contrast, more than 10 poorly formed glands that are not immediately adjacent to well-formed glands are considered GP 4. The

Advances in prostate cancer pathology Figure 3 Gleason pattern 4 prostate cancers. (a) All cribriform carcinomas regardless of size and shape are considered to be pattern 4. (b) Poorly formed glands represent Gleason pattern 4. (c) Focus of raggedly infiltrating fused glands. recommendations to improve the diagnostic reproducibility of poorly formed GP 4 from this study are provided in Table 1.18 When there is uncertainty whether a focus of poorly formed glands represent GP 3 or 4, a conservative approach is recommended with default to GP 3, an approach also endorsed by the 2014 ISUP consensus.9 Gleason pattern 5 The morphological spectrum of GP 5 is wide and includes PCa exhibiting essentially no glandular differentiation, solid tumor sheets, cords, or single cells and comedocarcinoma with central necrosis surrounded by papillary, cribriform, or solid 263 Figure 4 Poorly formed glands Gleason pattern 4 prostate cancer. (a) Only glands with no or barely visible lumens are considered poorly formed by expert urologic pathologists. Glands with discrete lumens are considered well-formed (arrows). (b) Poorly formed glands that are adjacent to other well-formed small glands (arrows) likely result from tangential sectioning and therefore should not be graded as pattern 4. (c) <= 5 poorly formed glands (arrows) in a cluster regardless of their location should not be graded as Gleason pattern 4. masses.10 Several additional patterns not formally recognized by the 2005 ISUP consensus include sheets with vague luminal formation, single file, solid nested and solid cylinder patterns, have also been proposed to represent GP 5.16,19 Gleason pattern 5 PCa has the worst biological behavior and is not uncommon in the contemporary NBX specimens. In a prospective NBX series using modified Gleason grading system, we reported that GP 5 was present in 6% of all biopsies.19 Infiltrating cords was the most common subpattern, seen in

264 R. B. Shah and M. Zhou 96% of pattern 5 cases, followed by single cells (76%), solid sheets (29%), and comedocarcinoma (2%). In particular, the infiltrating tumor cords and single cells subpatterns frequently coexisted. In addition, the GP 5 most commonly presented as the tertiary pattern (the third most common pattern). Fajardo et al. examined the frequency of undergrading in a consultation practice.20 They reported that the pattern 5 was more likely to be missed when it was a secondary or tertiary component. These studies emphasized the importance of recognizing the various morphologic variations of and to establish a minimum quantitative threshold for pattern 5. In a recent study, we studied interobserver reproducibility of diagnosing GP 5 on NBX among urologic pathologists.21 The overall reproducibility for the diagnosis of Gleason pattern 5 PCa for certain morphologies was only fair (κ = 0.34). Among subpatterns, comedocarcinoma had the highest diagnostic reproducibility (κ = 0.499) followed by variant morphology (such as signet ring cells) (κ = 0.443), single cells/cords (κ = 0.369), and nests (κ = 0.347). The diagnosis of Gleason pattern 5 could be improved when certain restrictive morphological and quantitative criteria were applied. Histological features that are diagnostic of Gleason pattern 5 prostate cancer include large tumor nests (>20 cells within the nest) (Fig. 5a), single cells/cords 6 10 in a cluster (Fig. 5b), single cells/cords >10 in a cluster or intermixed with adjacent glands (Fig. 5c) and signet ring-like cells in single cells or within nests. In contrast, the histological features that are against GP 5 include single cells/cords <5 (Fig. 5d), medium-sized tumor nests (<20 cells in the tumor nest) (Fig. 5e), intraluminal amorphous material only, and Paneth cell-like change in tumor nests (Fig. 5f). The recommendations to improve reproducibility of the diagnosis of GP 5 from this study are provided in Table 2.21 Grading prostate cancer with unusual morphologies and patterns Grading of PCa with variant histology and patterns such as glomeruloid structures and mucinous fibroplasia (Table 3) is based on the underlying glandular architecture and the peculiar namesake histological pattern or variation should be ignored. On the other hand, non-acinar carcinomas, such as urothelial and neuroendocrine carcinomas, should not be graded. Foamy gland carcinomas Even though they appear morphologically bland, foamy gland carcinomas usually are intermediate to aggressive tumors Figure 5 Gleason pattern 5 prostate cancers. (a) Large tumor nests (> 20 cells within the nest) with or without vague attempt for glandular differentiation represent pattern 5. (b) Single cells and cords 10 in a cluster or intermixed with adjacent glands represent pattern 5. (c) Single cells/cords 6 10 in a cluster arrangement represents minimum quantitative threshold for pattern 5. (d) Single cells/cords ( 5) should not be classified as pattern 5 as tangential sectioning of poorly formed Gleason pattern 4 cannot be reliably excluded. (e) Small to medium-sizes nests should not be classified as pattern 5 due to poor reproducibility. (f) Solid tumor nests with Paneth cell-like change should not be classified as pattern 5.

Advances in prostate cancer pathology 265 Table 2 Histological Features that are reproducible for the diagnosis of Gleason pattern 5 prostate adenocarcinoma by Urological Pathologists 21* Of solid nests pattern, only large nests (with >20 cells within the nest) should be considered pattern 5; medium or small tumor nests (<20 cells within the nest) are not pattern 5 Single cells/cords > 10 in a cluster or intermixed with adjacent glands is pattern 5; Singe cells/cords < 5 should not be graded as pattern 5; minimum # of single cells/cores to make a diagnosis of pattern5is6 Intraluminal coagulative necrosis is required to make the diagnosis of comedocarcinoma pattern 5 Signet ring cells in single cells or within nests is pattern 5 Paneth cell change in solid nests is not pattern 5 *These recommendations reflect consensus opinion on grading Gleason pattern 5 by 16 expert urological pathologists participating in an interobserver reproducibility study. 21 They are not part of the 2014 ISUP Gleason grading consensus. Table 3 Gleason grading of prostate carcinoma with unusual variant histology and patterns Histologic variant or pattern Foamy gland pattern Gleason grade Graded based on underlying architecture Pseudohyperplastic Pattern 3 Atrophic Pattern 3 PIN-like adenocarcinoma Pattern 3 Ductal adenocarcinoma Pattern 4, 5 with necrosis Glomeruloid structure Pattern 4 Collagenous micronodules Usually 3, graded based on underlying architecture Mucinous (colloid) carcinoma 3 or 4, graded based on underlying architecture Signet-ring cell-like carcinoma Pattern 5 Small cell carcinoma Not graded Adenosquamous and Not graded squamous carcinoma Sarcomatoid carcinoma Not graded biologically. Many tumors in fact represent high grade cancers. They should be graded based on overall architecture rather than foamy appearance. Prostate carcinoma with mucinous fibroplasia Mucinous fibroplasia or collagenous micronodules, representing hypo- or acellular collagenous substance found within or immediately adjacent to cancer glands, is also considered a feature specific to prostate cancer. Most of such cases represent GP 3 based on the underlying glandular architecture (Fig. 6a). Glomeruloid structures are intraglandular proliferation of cancer cells that superficially resemble renal glomeruli (Fig. 6b). There was no consensus on how to grade glomeruloid structures at the 2005 consensus meeting. Since then, studies have found that glomeruloid structures are often associated with GP 4 or higher cancer and transition between small and large glomeruloid and cribriform glands were common 22. Glomeruloid structures are considered a morphological continuum to cribriform glands and therefore are regarded as GP 4. 9 Mucinous (colloid) carcinoma consists of tumor glands within abundant extravasated mucin occupying >25% of the tumor (Fig. 6c). Such a diagnosis should only be rendered in radical prostatectomy and transurethral resection specimens. Biopsies with mucinous carcinoma should instead be diagnosed as prostate cancer with mucinous component. Previously such tumors were typically considered as a high grade GS 4 + 4 tumor. In 2005 modified Gleason grading system there was no consensus whether all colloid carcinomas should be assigned a GS 8. New data suggest that mucinous carcinomas treated by RP are no more aggressive than and possibly even have better prognosis than the usual acinar carcinomas. 23,24 By 2014 ISUP grading consensus, mucinous adenocarcinomas should be graded based on their underlying glandular growth pattern rather than grading them all as GP 4. 9 Large duct differentiation in PCa is characterized by large glandular contour demonstrating papillary and cribriform morphology with slit like lumens and columnar cells lining (Fig. 6d). It is graded as pattern 4 and pattern 5 if it contains comedonecrosis. Small cell carcinoma should not be graded due to its unique tumor biology and therapeutic significance. Any amount of such differentiation is significant and should be reported. Intraductal prostatic carcinoma (IDC-P) It has been recognized that some invasive prostatic carcinomas had basal cell layers on H&E staining, or more commonly, on basal cell immunostains. These lesions were variably labeled as high grade prostatic intraepithelial neoplasia (HGPIN) and PCa, ductal type, in the past. McNeal et al first demonstrated that they represented an aggressive form of PCa, as they were almost never seen in the absence of an invasive component. 25 If present, the invasive component is almost always of high grade and large volume. The prognosis of cancer harboring these basal cell-positive cancer glands was also significantly worse. The term intraductal prostatic carcinoma (IDC-P) was coined. Since then, the concept of IDC-P has evolved significantly. Several diagnostic criteria have been proposed and refined using both histological and molecular markers, 26 30 but the one proposed by Guo and Epstein has been widely utilized and is reproducible. 27 This diagnostic approach is summarized in Fig. 7a. 31 In addition to the presence of malignant epithelial cells expanding large acini and prostatic ducts with preservation of basal cells, the diagnosis of IDC-P required the presence of solid or dense cribriform pattern (Fig. 7 b,c). If these features are not present, a diagnosis of IDC-P can be made if there is: (i) non-focal comedonecrosis involving >= 2 glands (Fig. 7d); or (ii) marked nuclear atypia, where the nuclei are at least six times larger than adjacent benign nuclei (Fig. 7e). It must be emphasized that these histological features are highly specific but not sensitive for the diagnosis of IDC-P. In our experience a significant proportion of IDC-P may present

266 R. B. Shah and M. Zhou Figure 6 Prostate cancer with variant morphology. (a) Cancer glands with mucinous fibroplasia. Grading is based on the underlying glandular architecture (pattern 3 for this image). (b) Glomeruloid structures representing intraluminal cribriform proliferation is pattern 4. (c) Discrete cancer glands floating within abundant extracellular mucin pool. These tumors should be graded based on its underlying architecture (3 + 4 = 7 in this case). (d) Confluent papillary cancer glands lined by columnar epithelium represent large duct morphology and should be graded as Gleason pattern 4. with low-grade features that fall short of Guo and Epstein s criteria (Fig. 8a) and may mimic HGPIN.30,32 Such lesions have also been referred to as atypical intraductal proliferation. Overall, isolated atypical cribriform lesions not associated with cancer (cribriform HGPIN) are uncommon and the majority of atypical cribriform lesions containing basal cells, even when their constituent cells are of low-grade cytology, are associated with cancer and represents a low grade part of the IDC-P morphological spectrum. Therefore, one must use caution when diagnosing an isolated atypical cribriform lesion as HGPIN in NBX. Repeat biopsy is warranted in this setting to rule out unsampled invasive cancer. Recent studies have shown that ERG protein expression (Fig. 8b) and/or PTEN loss in atypical cribriform lesion (Fig. 8c) would support the diagnosis of IDC-P over HGPIN.28,29,33 IDC may be a precursor lesion in such cases with increased risk of more aggressive cancer.35 In other organ systems, intraductal lesions are not graded using the same grading system for the invasive component. In another rare setting of IDC and GS 3 + 3 cancer in biopsy, 21% of corresponding RP had GS 3 + 3 = 6 PCa along with IDC,36 arguing against diagnosing the preoperative biopsies as GS 3 + 4. The 2014 ISUP Gleason grading consensus recommends that IDC not be graded but its association with aggressive behavior be noted.12 The diagnosis of IDC should be confirmed using basal cells when there is concern whether the biopsy contains IDC-P only or invasive cancer, or when the Gleason grade could change with the diagnosis of IDC-P. Operational changes Grading Intraductal prostatic carcinoma Reporting of IDC-P in NBX deserves special mention. If a high grade (GP 4 and 5) invasive component is present with IDC-P, diagnosis of IDC-P may matter little for patient management purposes but is still worth reporting as it may provide additional independent prognostic information.32 When associated with a GP 3 component, IDC-P should, however, be documented along with its poor prognostic significance.32 One solution could be, and had been suggested, to grade the IDC-P component as pattern 4 or 5 invasive PCa. However, this approach can be problematic. Rarely patients had IDC-P only on biopsy. At RP, 10% of these patients had no invasive carcinoma34 If the biopsies were to be diagnosed as GS >8 PCa, patients would have been labeled as having high grade disease with poor prognosis when in fact they could be cured with IDC only. Miyai et al. suggested that Grading a biopsy specimen with multiple cores positive for cancer of different Gleason grades An important issue in grading a biopsy with PCa of different GS involving multiple cores is whether to render a global GS or worst score. The 2005 ISUP consensus recommended assigning GS to individual cores as long as they were submitted in separate containers or their locations were specified by urologists. It was optional to provide an overall GS10 In one study of NBX with one core showing GS 4 + 4 = 8 PCa and other cores having GS < 7, the pathological stage at radical prostatectomy was comparable to cases with all needle cores showing GS 4 + 4=8.37 In addition, urological surgeons typically use the highest GS of all cores as the case GS for patient management.38 Arias-Stella et al. recently proposed a composite GS which integrates grade patterns from

Advances in prostate cancer pathology 267 Figure 7 Intraductal carcinoma of the prostate. (a) A diagnostic approach to IDCP (reproduced from Robinson et al). The diagnosis of IDC-P requires the presence of solid (b) or dense cribriform pattern (c) or if there is non-focal comedonecrosis involving >= 2 glands (d), or marked nuclear atypia, where the nuclei are at least 6 times larger than adjacent benign nuclei (e). contiguous positive biopsy sites. Composite GS correlated better with RP GS than the highest biopsy GS.39 Specifically, the former was less likely to be downgraded compared with the latter in RP specimens. In practice, it is fairly common to see that multiple cores are placed in the same container without any identifiers or they are designated only as left or right. Currently there are no clear guidelines to address this issue. We simulated this scenario of multiple individual intact cores with different GS in the same container by analyzing as if these cores were submitted in containers labeled as right and left then determined the global score (all positive cores averaged as one positive core), worst GS, and largest tumor volume GS. Grade correlation with RP was best and the likelihood of significant upgrading at RP was least when the biopsy worst GS was used.40 These results suggest that when multiple cores are submitted in a container without site specification, GS of individual cores or worst GS should be reported. In cases where a container contains multiple pieces of tissue and one cannot be sure if one is looking at an intact core, the consensus is that an overall GS for that container should be provided. Any amount of high grade in the needle biopsy should be reported High-grade (GP 4 and 5) tumor of any quantity in needle biopsy, as long as it can be identified at low to medium magnification, should be included in the final GS.10 This is based on the

268 R. B. Shah and M. Zhou Reporting of tertiary pattern 5 in the biopsy and prostatectomy specimens Prostate cancer often contains >2 Gleason patterns within the same tumor. It is important to report the tertiary Gleason pattern when it is of higher grade than the secondary pattern. The typical scenario is that a biopsy specimen includes tumors with patterns 3, 4, and 5 in various proportions. In such a tumor, both the primary and the highest pattern should be recorded.9,25 If the pattern 3 is the most common pattern, the final GS should be 3 + 5 = 8. This recommendation is also important for patient management. Risk stratification tools, such as US National Comprehensive Cancer Center Network (NCCN) guidelines and D Amico risk stratification, and outcome prediction models to predict the pathological stage and prognosis following radical prostatectomy, such as Partin Tables and Kattan Nomogram, use only primary and secondary patterns reported in the NBX and therefore tertiary pattern of higher grade would be dropped unless reported as the secondary pattern. The reporting of the tertiary pattern in RP is different from biopsy specimens.10 When three patterns are present and the highest pattern is the least common pattern, the highest pattern should be reported as the tertiary pattern. However, the definition of tertiary pattern is variable. Some experts consider the highest pattern as tertiary pattern only when it is <5% of the tumor volume while some would assign it as tertiary pattern even when it is >5% as long as the highest pattern is tertiary in quantity. Moose et al demonstrated that tumors with GS 4 + 3 = 7 and a tertiary pattern 5 at RP behave worse than those with GS 4 + 3 = 7 without a tertiary pattern 5, but have a lower incidence of seminal vesicle invasion and lymph node metastases compared with tumors of GS 4 + 5=9.41 Reporting Percentage of Gleason Pattern 4 on Needle Biopsy and Radical Prostatectomy Figure 8 Intraductal carcinoma of the prostate with low grade morphology. A significant proportion of IDC-P may present with low-grade features that overlap with HGPIN. (a) This atypical cribriform lesion has morphological features worse than HGPIN but falls short of IDC-P. (b) A multiplex stain for basal cell markers (p63 and CK903), AMACR and ERG demonstrates retention of basal cells (brown chromogen) and ERG overexpression (red chromogen), and (c) PTEN protein loss supports the diagnosis of IDC-P. assumption that any amount of high-grade tumor sampled in needle biopsy most likely indicates a significant tumor at RP. In RP, the majority agrees that even a limited amount of highgrade tumor should be reported as either secondary (when it is >5% of the tumor) or tertiary pattern (when it is <5% of the tumor). Relative abundance of pattern 4 PCa significantly impacts the prognosis. GS 3 + 4 and 4 + 3 constitute two prognostically distinct disease groups.42 Stamey et al. and Reese et al. further demonstrated that incorporation of the percentage of pattern 4 cancer improves risk stratification by the Gleason grading system.43,44 Our own study found that the pathological features of RP in patients with minimum (<5%) pattern 4 cancer in biopsy showing 3 + 4 = 7 were similar to patients with biopsy showing 3 + 3=6.45 For patient management, GS 3 + 3 is an accepted criterion for active surveillance. However, some men, depending on other clinicopathological parameters (age, co-morbidity, extent of PCa, MRI findings, patient desire, etc.), could be a candidate for active surveillance with GS 3 + 4 PCa if the pattern 4 is limited46 The consensus recommendation is to report percent pattern 4 in Gleason score 7 biopsy and radical prostatectomy specimens.

Advances in prostate cancer pathology 269 Gleason grading in radical prostatectomy with multifocal cancer Prostate cancer frequently presents as multifocal disease with heterogeneity in Gleason score (GS) and genetic alterations. 47,48 The concept of dominant, or index, tumor nodules is adopted for the convenience of reporting Gleason score of the entire case and procurement of tissue for research. Most often, the dominant nodule is the largest tumor, and has the highest stage and grade. However, in a significant number of cases the largest tumor volume, highest GS and staging parameters (such as extraprostatic extension) do not always concur in the same tumor nodule. 49 In these cases, pathologists should de-emphasize the concept of dominant tumor nodules. Instead, they should place the emphasis on the multifocal nature of the disease and document the pathological features of all independent tumor foci that have largest tumor size, highest GS and staging parameters. PROGNOSTIC GLEASON GRADE GROUPING The Gleason grading system has several deficiencies that may have misleading clinical implications. The modified Gleason grading system has practically eliminated GS 2-5 in biopsy and in majority of radical prostatectomy specimens. Therefore the lowest GS in both biopsy and radical prostatectomy is 6. Since 6 is in the middle of the 2 10 numerical scale, patients may erroneously assume that they have a moderately aggressive cancer despite that GS 6 PCa is the least aggressive tumor. GS is often grouped into three tiers (6, 7, 8 10) for prognostic and therapeutic purposes, despite that GS 3 + 4 vs 4+3and8vs 9 10 have significantly different prognosis. To address these issues, Pierorazio et al. recently proposed a novel way of grouping Gleason grades. Nine potential Gleason scores (2 10) were grouped into five groups: Grade group 1 (Gleason score 2 6); Grade group 2 (Gleason score 3 + 4 = 7); Grade group 3 (Gleason score 4 + 3 = 7); Grade group4(gleasonscore8);andgradegroup5(gleasonscore 9 10) to accurately reflect prognosis of prostate cancer. 50 In a subsequent meta-analysis of over 19 000 men treated by radical prostatectomy from five institutions worldwide, grade groups strongly correlated with the risk of biochemical recurrence after surgery. 12 Compared with the grade group 1, the relative risk of progression for grade groups 2 5 were 2.6, 8.5, 16.8, and 29.3. The 5-year biochemical risk free survival was 97.5%, 93.1%, 78.1%, 63.6%, and 48.9% for Grade Groups 1 5, respectively. The grade groups were also validated on biopsy correlating with risk of progression after radical prostatectomy and following radiation therapy. A new grading system was proposed based on these findings (Table 4) and was endorsed by ISUP and accepted by the World Health Organization for the 2016 edition of the genitourinary pathology blue book. It is suggested this grade group be used in conjunction with the Gleason grading Table 4 Histological Definition of Grade Group of Prostate Cancer Grade Group 1 (Gleason score < 6) Grade Group 2 (Gleason score 3 + 4 = 7) Grade Group 3 (Gleason score 4 + 3 = 7) Grade Group 4 (Gleason score4+4=8;3+5=8; 5+3=8) Grade Group 5 (Gleason scores 9 10) Only individual discrete well-formed glands Predominantly well-formed glands with lesser component of poorlyformed/fused/cribriform glands Predominantly poorly formed/ fused/cribriform glands with lesser component of well-formed glands - Only poorly-formed/fused/ cribriform glands or - Predominantly well-formed glands and lesser component lacking glands or - Predominantly lacking glands and lesser component of well-formed glands Lacks gland formation (or with necrosis) with or w/o poorly formed/ fused/cribriform glands For cases with >95% poorly-formed/fused/cribriform glands or lack of glands on a core or at RP, the component of <5% well-formed glands is not factored into the grade. Poorly-formed/fused/cribriform glands can be a more minor component. system. An example could be prostate carcinoma, Gleason score 3+4=7 (Grade group 2). MAGNETIC RESONANCE IMAGING (MRI)-TARGETED PROSTATE BIOPSIES Multiparametric magnetic resonance imaging (mpmri), combining T2 weighted sequence, diffusion weighted imaging (DWI) and dynamic contrast enhancement (DCE) MRI, is now an established imaging modality to detect and characterize PCa 51,52 MRI-targeted biopsy, compared with the standard extended biopsy, has a higher rate of detection of clinically significant PCa with fewer biopsy cores. PCa missed by MRItargeted biopsy is more likely to be clinically insignificant 51,52 MRI-targeted biopsy also provides better representation of the tumor characteristics such as tumor volume and Gleason grade. Lanz et al. found that MRI-targeted biopsy allowed for an accurate estimation of GS in more than 2/3 of patients, while GS misclassification was mostly due to a lack of accuracy in the determination of the secondary Gleason grade. 53 Arsov et al. demonstrated that combining MRI-targeted and standard biopsy significantly reduced the risk of Gleason upgrading in RP compared to standard and targeted biopsy used alone (28.8%, 40.4%, and 50.0%, respectively) 54 Siddiqui et al. further showed that MRI-targeted biopsy upgraded and detected PCa of higher GS in 32% of patients compared with standard 12-core biopsy alone. Targeted biopsy preferentially detects higher-grade PCa while missing lower-grade tumors. 55 These findings suggest that MRI-targeted biopsy can improve the biopsy and RP GS concordance and therefore enhance the prognostic significance of GS derived from the biopsy specimens.

270 R. B. Shah and M. Zhou GENETIC MARKERS Advanced molecular techniques, such as gene expression profiling, comparative genomic hybridization (CGH) and next generation sequencing, provide an unprecedented opportunity for us not only to study the molecular mechanisms of prostate carcinogenesis, but also to develop gene-based assays for diagnosis and prediction of prognosis and treatment response. A potential utility of genetic and molecular markers of PCa is to triage patients for appropriate management. In current practice, the management decision is based primarily on clinical and biopsy pathology findings, including clinical stage, serum PSA level and density, digital rectal examination and biopsy Gleason score and tumor extent. Patients are stratified into different risk groups for which different management regimens are offered. However, such a stratification scheme is far from perfect and sometimes results in unnecessary treatment in patients with low risk disease and delayed treatment in patients with high risk disease. Molecular and genetic markers, used singularly or in combination, can potentially separate indolent PCa from aggressive ones and help identify patients with indolent disease who can therefore be safely followed and patients with aggressive disease who need definitive treatment 56 Detection of loss of phosphatase and tensin homolog (PTEN) tumor suppressor gene represents one such important example. PTEN on chromosome 10q23 is a key tumor suppressor gene that is often deleted or inactivated in prostate cancer. 57 Several studies have demonstrated that PTEN inactivation is consistently associated with a variety of different adverse pathologic outcomes including metastasis and disease specific death 58 60 PTEN loss detected by immunohistochemistry in NBX was predictive of poor clinical outcome 61 and was also associated with upgrading at radical prostatectomy 62 Therefore, immunohistochemical detection of PTEN loss on NBX is potentially a useful tool, along with other clinicopathological features, to refine the risk stratification prior to definitive therapy. PTEN measurement by immunohistochemistry, however, is subjected to both high inter- and intra-tumoral staining heterogeneity and is affected by the fraction of positive cores and pathological features of PCa that is interrogated. 63 Recommendations for optimal yet cost effective strategies to identify PTEN loss in multiple core biopsy setting have been recently described. 63 There are several commercially available multi-gene assays marketed for this purpose. The two most frequently requested ones are Prolaris by Myriad Genetics (Salt Lake City, UT, USA) and Oncotype Dx Prostate marketed by Genomic Health (Redwood City, CA, USA). Both tests use quantitative reversetranscriptase PCR to measure the expression levels of a panel of genes involved in biological pathways important in prostate cancer and generate an assay score based on the gene expression levels (cell-cycle progression [CCP] score in Prolaris and Prostate Cancer Score (GPS) in Oncotype Dx Prostate). In a recent meta-analysis of six published studies of the prognostic value of CCP score, it was found that the CCP scores provided additional prognostic values not captured by other clinicopathological parameters and the test could impact the treatment decision and potentially lead to a decrease in surgical intervention for low-risk patients. 64 Similarly, GPS generated on prostate biopsies has been prospectively validated as a predictor of PCa aggressiveness in biopsy tissue despite of tumor heterogeneity, multifocality, and limited sampling at time of biopsy. The biopsy-based GPS improves prediction of the presence of adverse pathology and may help patients make more informed decisions between active surveillance and immediate treatment. 65 Even though these tests have shown promising results in early studies and started to be requested by physicians and patients, large prospective validation studies are needed on prostate biopsies before they can be recommended in routine clinical practice for risk stratification and patient management. SUMMARY The ISUP 2005 and 2014 modifications of Gleason grading system have profoundly changed the way PCa is graded and managed. The grading criteria for each Gleason pattern as well as histological variants of PCa have been better defined. We also provided recommendations on how to improve the grading reproducibility. New prognostic grade grouping was proposed to group Gleason scores into clinically and therapeutically meaningful grade groups. Finally, MRI-targeted biopsy and emerging genetic markers hold great promise to improve the Gleason grading accuracy. It is very hopeful that conventional clinicopathological parameters such as GS, imaging findings, and molecular markers are integrated together to provide better and more accurate prediction of tumor behavior and ultimately to help patients make the right treatment decision. ACKNOWLEDGMENT MZ was invited to the meeting of 104th The Japanese Society of Pathology in Nagoya, Japan, in 2015 and this manuscript is partly based upon the presentation given at the meeting. None declared. DISCLOSURE STATEMENT REFERENCES 1 Gleason DF. Classification of prostatic carcinomas. Cancer Chemother Rep 1966; 50: 125 8. 2 Gleason DF. Histologic grading of prostate cancer: a perspective. Hum Pathol 1992; 23: 273 9.

Advances in prostate cancer pathology 271 3 Gleason DF, Mellinger GT. Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J Urol 1974; 111: 58 64. 4 Amin MB, Lin DW, Gore JL, et al. The critical role of the pathologist in determining eligibility for active surveillance as a management option in patients with prostate cancer: consensus statement with recommendations supported by the College of American Pathologists, International Society of Urological Pathology, Association of Directors of Anatomic and Surgical Pathology, the New Zealand Society of Pathologists, and the Prostate Cancer Foundation. Arch Pathol Lab Med 2014; 138: 1387 405. 5 Bastian PJ, Gonzalgo ML, Aronson WJ, et al. Clinical and pathologic outcome after radical prostatectomy for prostate cancer patients with a preoperative Gleason sum of 8 to 10. Cancer 2006; 107: 1265 72. 6 Bentley G, Dey J, Sakr WA, Wood DP Jr, Pontes JE, Grignon DJ. Significance of the Gleason scoring system after neoadjuvant hormonal therapy. Mol Urol 2000; 4: 125 29. 7 Bostwick DG, Grignon DJ, Hammond ME, et al. Prognostic factors in prostate cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000; 124: 995 1000. 8 Shah RB. Current perspectives on the Gleason grading of prostate cancer. Arch Pathol Lab Med 2009; 133: 1810 6. 9 Epstein JI, Egevad L, Amin MB, Delahunt B, Srigley JR, Humphrey PA. The 2014 International Society of Urological Pathology (ISUP) consensus conference on Gleason grading of prostatic carcinoma: Definition of grading patterns and proposal for a new grading system. Am J Surg Pathol 2016; 40: 244 52. 10 Epstein JI, Allsbrook WC Jr, Amin MB, Egevad LL. The 2005 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am J Surg Pathol 2005; 29: 1228 42. 11 Epstein JI. Gleason score 2-4 adenocarcinoma of the prostate on needle biopsy: a diagnosis that should not be made. Am J Surg Pathol 2000; 24: 477 8. 12 Epstein JI, Zelefsky MJ, Sjoberg DD, et al. A Contemporary Prostate Cancer Grading System: A Validated Alternative to the Gleason Score. Eur Urol 2016; 69: 428 35. 13 Latour M, Amin MB, Billis A, et al. Grading of invasive cribriform carcinoma on prostate needle biopsy: an interobserver study among experts in genitourinary pathology. Am J Surg Pathol 2008; 32: 1532 9. 14 Iczkowski KA, Torkko KC, Kotnis GR, et al. Digital quantification of five high-grade prostate cancer patterns, including the cribriform pattern, and their association with adverse outcome. Am J Clin Pathol 2011; 136: 98 107. 15 Kweldam CF, Wildhagen MF, Steyerberg EW, Bangma CH, van der Kwast TH, van Leenders GJ. Cribriform growth is highly predictive for postoperative metastasis and disease-specific death in Gleason score 7 prostate cancer. Mod path: an official journal of the United States and Canadian Academy of Pathology, Inc 2015; 28: 457 64. 16 Gottipati S, Warncke J, Vollmer R, Humphrey PA. Usual and unusual histologic patterns of high Gleason score 8 to 10 adenocarcinoma of the prostate in needle biopsy tissue. Am J Surg Pathol 2012; 36: 900 7. 17 McKenney JK, Simko J, Bonham M, et al. The potential impact of reproducibility of Gleason grading in men with early stage prostate cancer managed by active surveillance: a multi-institutional study. JUrol2011; 186: 465 9. 18 Zhou M, Li J, Cheng L, et al. Diagnosis of Poorly Formed Glands Gleason Pattern 4 Prostatic Adenocarcinoma on Needle Biopsy: An Interobserver Reproducibility Study Among Urologic Pathologists With Recommendations. Am J Surg Pathol 2015; 39: 1331 9. 19 Shah RB, Tadros Y. Adenocarcinoma of the prostate with Gleason pattern 5 on core biopsy: frequency of diagnosis, morphologic subpatterns, and relation to pattern distribution based on the modified Gleason grading system. Hum Pathol 2014; 45: 2263 9. 20 Fajardo DA, Miyamoto H, Miller JS, Lee TK, Epstein JI. Identification of Gleason pattern 5 on prostatic needle core biopsy: frequency of underdiagnosis and relation to morphology. Am J Surg Pathol 2011; 35: 1706 11. 21 Shah RB, Li J, Cheng L, et al. Diagnosis of Gleason Pattern 5 Prostate Adenocarcinoma on Core Needle Biopsy: An Interobserver Reproducibility Study Among Urologic Pathologists. Am J Surg Pathol 2015; 39: 1242 9. 22 Lotan TL, Epstein JI. Gleason grading of prostatic adenocarcinoma with glomeruloid features on needle biopsy. Hum Pathol Apr 2009; 40: 471 7. 23 Lane BR, Magi-Galluzzi C, Reuther AM, Levin HS, Zhou M, Klein EA. Mucinous adenocarcinoma of the prostate does not confer poor prognosis. Urology 2006; 68: 825 30. 24 Osunkoya AO, Nielsen ME, Epstein JI. Prognosis of mucinous adenocarcinoma of the prostate treated by radical prostatectomy: a study of 47 cases. Am J Surg Pathol 2008; 32: 468 72. 25 McNeal JE, Yemoto CE. Spread of adenocarcinoma within prostatic ducts and acini. Morphologic and clinical correlations. Am J Surg Pathol 1996; 20: 802 14. 26 Cohen RJ, Shannon BA, Weinstein SL. Intraductal carcinoma of the prostate gland with transmucosal spread to the seminal vesicle: a lesion distinct from high-grade prostatic intraepithelial neoplasia. Arch Pathol Lab Med 2007; 131: 1122 5. 27 Guo CC, Epstein JI. Intraductal carcinoma of the prostate on needle biopsy: Histologic features and clinical significance. Mod Pathol 2006; 19: 1528 35. 28 Han B, Suleman K, Wang L, et al. ETS gene aberrations in atypical cribriform lesions of the prostate: Implications for the distinction between intraductal carcinoma of the prostate and cribriform highgrade prostatic intraepithelial neoplasia. Am J Surg Pathol 2010; 34: 478 85. 29 Morais CL, Han JS, Gordetsky J, et al. Utility of PTEN and ERG immunostaining for distinguishing high-grade PIN from intraductal carcinoma of the prostate on needle biopsy. Am J Surg Pathol 2015; 39: 169 78. 30 Shah RB, Magi-Galluzzi C, Han B, Zhou M. Atypical cribriform lesions of the prostate: relationship to prostatic carcinoma and implication for diagnosis in prostate biopsies. Am J Surg Pathol 2010; 34: 470 7. 31 Robinson B, Magi-Galluzzi C, Zhou M. Intraductal carcinoma of the prostate. Arch Pathol Lab Med 2012; 136: 418 25. 32 Shah RB, Zhou M. Atypical cribriform lesions of the prostate: clinical significance, differential diagnosis and current concept of intraductal carcinoma of the prostate. Adv Anat Pathol 2012; 19: 270 8. 33 Shah RB. Clinical applications of novel ERG immunohistochemistry in prostate cancer diagnosis and management. Adv Anat Pathol 2013; 20: 117 24. 34 Robinson BD, Epstein JI. Intraductal carcinoma of the prostate without invasive carcinoma on needle biopsy: emphasis on radical prostatectomy findings. J Urol 2010; 184: 1328 33. 35 Miyai K, Divatia MK, Shen SS, Miles BJ, Ayala AG, Ro JY. Heterogeneous clinicopathological features of intraductal carcinoma of the prostate: a comparison between precursor-like and regular type lesions. Int J Clin Exp Pathol 2014; 7: 2518 26. 36 Khani F, Epstein JI. Prostate Biopsy Specimens With Gleason 3 + 3 = 6 and Intraductal Carcinoma: Radical Prostatectomy Findings and Clinical Outcomes. Am J Surg Pathol Oct 2015; 39: 1383 9. 37 Kunz GM Jr, Epstein JI. Should each core with prostate cancer be assigned a separate gleason score? Hum Pathol 2003; 34: 911 4. 38 Rubin MA, Bismar TA, Curtis S, Montie JE. Prostate needle biopsy reporting: how are the surgical members of the Society of Urologic Oncology using pathology reports to guide treatment of prostate cancer patients? Am J Surg Pathol 2004; 28: 946 52.