Updates in Histologic Grading of Urologic Neoplasms

The first grading schemes for RCC were described by Hand and Broders¹  in 1932, followed by Griffiths and Thackray²  in 1949. A few decades later, in 1971, Skinner and colleagues³  pioneered the utilization of nuclear features alone to determine histologic grade and they demonstrated significant survival differences for their 4-tier system. Numerous subsequent grading strategies have been proposed, but nuclear grading schemes have maintained superiority over grading systems that lack nuclear information.⁴ 

The Fuhrman grading system, proposed in 1982, has been the most widely accepted and used grading scheme in the United States for the past few decades.⁵  The Fuhrman system was adapted from the grading method put forth by Skinner et al, and defined grades 1 to 3 on the basis of nuclear features, and grade 4, on the presence of nuclear pleomorphism and/or tumor giant cells, determined by the area of highest grade.⁵  Despite its wide acceptance, numerous limitations have been associated with the Fuhrman grading system such as the need to simultaneously evaluate 3 parameters (nuclear size, shape, and nucleolar prominence) as well as account for nuclear pleomorphism. Furthermore, the original study that formed the basis of the Fuhrman grading system had significant limitations, including having grouped all RCCs into a single category despite inclusion of different morphologic types, as well as limited follow-up information, varying treatment modalities, and identification of only 3 prognostically significant groups.⁵  Patients with grade 1, grade 2-3, and grade 4 RCCs correlated with differences in outcome, with most falling in the intermediate grade 2-3 category.⁵  Although nuclear parameters can be objectively measured, the system does not make clear which of the 3 nuclear features is most important when there is a discrepancy amongst them, and studies have shown a wide variation in Fuhrman grade allocation and reproducibility.^6–9  To simplify matters, some pathologists have relied strictly upon nucleolar prominence to designate a grade, rather than apply the full Fuhrman grading criteria.^10–12 

Over the years, many prognostic factors have been identified for RCC, although only a few histologic features have reliably correlated with outcome. The Rochester Renal Cell Carcinoma Consensus Conference in 1997 evaluated a diverse set of parameters and found that histologic type along with histologic grade, sarcomatoid differentiation, gross margin involvement, TNM classification, and metastases have prognostic significance.^13–15  With the Heidelberg classification and identification of RCC subtypes, which correlate well with patient outcome, the applicability of the Fuhrman grading system was further questioned.^13–17  Of the different RCC subtypes, the Fuhrman system correlated best with outcome for clear cell RCC (CCRCC).^6,18–22  The data are limited regarding the prognostic significance of Fuhrman grading in papillary RCC (PRCC), and some studies have shown that it does not significantly predict outcome.^22–25  The 3 nuclear components of the Fuhrman system were assessed independently for prognostic significance and only nucleolar prominence was retained on univariate analysis in PRCCs, further supporting the use of nucleolar grading instead of Fuhrman grading.¹¹  None of the 3 grading components correlated with outcome for chromophobe RCC (ChRCC), and multiple additional studies support that Fuhrman grading should not be applied to ChRCC.^26–30 

Given the intrinsic difficulties associated with the Fuhrman system and studies showing that nucleolar prominence alone has better correlation with outcomes, particularly for CCRCC and PRCC, a new grading system was proposed.^11,12,31,32  In 2012, at the Vancouver ISUP Consensus Conference, a grading system based solely on nucleolar prominence was accepted, and approved for use in both CCRCC and PRCC.^{11,12,25,31,33,34}  Subsequently, this grading scheme was endorsed by the WHO and renamed as the WHO/ISUP grading system.⁹  More recently, the 8th edition of the American Joint Committee Cancer Staging Manual adopted this grading scheme.³⁵ 

The WHO/ISUP grading system slightly modified the original ISUP classification in that it also emphasizes eosinophilia of the nucleoli.³⁶  Grade 1 carcinomas contain nucleoli that are absent or inconspicuous and basophilic at ×400, grade 2 nucleoli are conspicuous and eosinophilic at ×400 and visible but not prominent at ×100, grade 3 nucleoli are conspicuous and eosinophilic at ×100, and grade 4 is defined by severe nuclear pleomorphism, multinucleated giant cells, and rhabdoid and/or sarcomatoid differentiation (Figure; Table).³⁶  The grade should be allocated on the basis of the maximum nuclear pleomorphism that can be found in a single high-power field.³⁶  These slight alterations have resulted in the downgrading of tumors, particularly from grade 2 to 1 and grade 3 to 2, in large part due to the requirement of eosinophilia.^31,34,37 

Chromophobe RCC is associated with a more favorable prognosis than that of other subtypes, but owing to the inherent nuclear pleomorphism (ie, irregular nuclei and prominent nucleoli) they are paradoxically given a higher Fuhrman grade. In 2010, Paner and colleagues³⁸  proposed a 3-tiered novel chromophobe tumor grading system that discounts the nuclear atypia and is instead based upon nuclear anaplasia and crowding assessed within 2 distinct areas. In this system, ChRCC grade 1 lacks nuclear crowding and anaplasia. Grade 2 contains “geographic nuclear crowding,” which is defined as clustering of cells with high nuclear/cytoplasmic density that is detectable at ×100 and in which some nuclei are directly contacting each other at ×400, as well as the presence of nuclear pleomorphism (≥3-fold variation in size and irregularities of nuclear chromatin). Grade 3 contains frank anaplasia demarcated by tumor giant cells/multilobation or sarcomatoid differentiation.³⁸  Both Fuhrman and chromophobe tumor grades were found to be associated with tumor stage, likely driven by the tumors with sarcomatoid differentiation.³⁸  When ChRCCs with sarcomatoid change were excluded, only the chromophobe tumor grade significantly associated with tumor stage.³⁸  Subsequently, additional studies have compared the chromophobe tumor grade to the Fuhrman grading schemes with conflicting results.^39,40  The current recommendation is that ChRCC should not be graded.^26,40 

Sarcomatoid RCC, composed of malignant spindle cells, is thought to represent epithelial-mesenchymal transformation and is considered to be a radical form of dedifferentiation seen in numerous subtypes of RCC.^6,41  It is associated with aggressive behavior with 77% of cases presenting at advanced stage with metastasis, and has a dismal cancer-specific survival of 15% to 22% at 5 years and a median survival of 4 to 9 months.^42–45  While the presence or absence of sarcomatoid differentiation is clearly significant, the prognostic significance of the amount of sarcomatoid component is inconclusive. Most reports do not find the amount or proportion of sarcomatoid differentiation to be significant, although one study found patients with a smaller sarcomatoid component (<20%) had better outcomes.^{41–44,46,47}  ISUP recommends reporting the presence or absence of sarcomatoid differentiation as well as the underlying RCC subtype (eg, CCRCC with sarcomatoid differentiation).²⁵  In the rare occasion that the neoplasm is composed purely of sarcomatoid carcinoma, it is best classified as RCC unclassified with a sarcomatoid component.²⁵ 

Renal cell carcinomas with rhabdoid differentiation, which resemble rhabdomyoblasts but contain a different immunophenotype, are composed of tumor cells with abundant eosinophilic cytoplasm and a large eccentric nucleus with a prominent nucleolus. Like sarcomatoid morphology, rhabdoid differentiation is thought to represent a form of dedifferentiation and has been associated with numerous subtypes of RCC.^25,48–54  Also similar to sarcomatoid carcinoma, rhabdoid differentiation is associated with advanced stage with metastasis in up to 70% of cases and a generally poor prognosis, as mirrored by its cancer-specific mortality rate of 40% to 50% and median survival of 8 to 31 months.^{25,48–50,55–59}  ISUP also recommends reporting the presence or absence of rhabdoid differentiation and if there is histologic evidence, to also report the underlying RCC subtype.²⁵  In the uncommon occurrence that the neoplasm has purely rhabdoid morphology, it is best classified as RCC unclassified with a rhabdoid component.²⁵ 

Given their poor associated outcomes, RCCs with sarcomatoid and/or rhabdoid differentiation, in addition to tumors displaying anaplasia and/or tumor giant cells, are assigned WHO/ISUP grade 4.²⁵  ISUP does not recommend a minimum amount of either morphologies to make a diagnosis of sarcomatoid or rhabdoid differentiation.²⁵ 

Tumor necrosis is recognized as a notable prognostic indicator associated with poorer outcomes. As such, the 2012 Vancouver ISUP Consensus Conference recommended that the presence or absence as well as the percentage of necrosis should be included in pathology reports for CCRCC.^25,60–62 

In 2013, Delahunt et al⁶³  proposed a novel grading system for CCRCC, in which the ISUP grade was incorporated and substratified on the presence or absence of microscopic tumor necrosis. This new system reportedly outperformed ISUP grade alone.⁶³  The proposed modified grading system stratifies tumors into the following 4-tiered system: modified grade 1 composed of grade 1 or nonnecrotic grade 2, modified grade 2 defined as necrotic grade 2 or nonnecrotic grade 3, modified grade 3 composed of necrotic grade 3 or nonnecrotic and non-sarcomatoid/rhabdoid grade 4, and modified grade 4 consisting of grade 4 tumors with either necrosis and/or sarcomatoid/rhabdoid differentiation.⁶³  This study further noted that the addition of microscopic tumor necrosis did not add statistically significant prognostic information for PRCC and ChRCC, like it did for CCRCC.⁶³ 

Similar findings were reported in a subsequent study by Verine and colleagues⁶⁴  in which integrating tumor necrosis also provided greater prognostic information than the WHO/ISUP grading alone. Another study performed by Khor et al⁶⁵  applied this modified grading scheme and found patients could be significantly tiered into 3 prognostic groups, especially when the amount of necrosis is quantified in tumors of higher grade. Group 1 was composed of nonnecrotic grade 1-3 tumors, group 2 contained necrotic grade 1-3 tumors as well as grade 4 tumors with less than 10% necrosis, and group 3 was composed of grade 4 tumors with more than 10% necrosis.⁶⁵  Recently, Dagher and colleagues⁶⁶  also found that WHO/ISUP grade 2-4 tumors with necrosis had a significantly worse prognosis than those without necrosis and were associated with a higher risk of metastasis.

Although the prognostic significance associated with tumor necrosis is evolving and appears to be building, no further formal recommendations have been made aside from reporting the presence or absence of tumor necrosis and both the macroscopic and microscopic percentage of tumor necrosis.²⁵ 

Grading has not been validated for other subtypes of RCC beyond CCRCC and PRCC, but may be used for illustrative purposes with the understanding that it does not correlate with prognosis.³⁶  For instance, it is well known that certain subtypes of RCC, such as tubulocystic carcinoma, typically have prominent nucleoli consistent with WHO/ISUP grade 3 but are associated with indolent behavior.^36,67 

Prostate cancer (PCa) is extremely common in the developed world and demonstrates a wide range of malignant potential.^68,69  Histologic grading of PCa is an important component of predicting tumor behavior and thereby optimizing therapeutic approach.

The current system of PCa grading is founded on the Gleason system, which was developed by Donald F. Gleason and colleagues in the 1960s and 70s based on prospective, randomized studies comparing treatment regimens in more than a thousand patients across the Veterans Administration Hospital System.^70,71  The original Gleason system defined 5 architectural patterns (1–5) of “increasing apparent histologic malignancy” based on the low magnification (×40–×100) appearance of the tumor without regard to cytology. Histologic variation was accounted for by combining the first and second most predominant patterns (by area) to yield a score (range, 2–10) that correlated to patient mortality, particularly with deaths due to cancer only.⁷⁰ 

Numerous modifications to the Gleason system have been proposed since its inception, usually with the intention of either improving agreement or accounting for the changing landscape of PCa, including the advent of prostate-specific antigen, image-based, and biopsy screening, the routine use of immunohistochemistry, new treatment modalities, rising support for active surveillance, and changing patient populations. After the major revisions to the Gleason system by the 2005 and 2014 ISUP Consensus Conferences, the current state of PCa grading differs considerably from the original description and is reviewed below.

Consistent with Gleason's original approach, contemporary prostate cancer grading is based on assigning architectural patterns (1–5) to a tumor at low magnification (×40–×100). The Gleason score (eg, 3 + 4 = 7) is composed of the 2 most prevalent architectural patterns (1–5) in resection specimens or the most prevalent and worst remaining pattern (3–5) in needle biopsy specimens.^72–74  When only 1 pattern is present it is counted twice (eg, 3 + 3 = 6). While cytologic features may help establish a diagnosis of PCa, they are not considered in grading. Likewise, while high-power examination (×200–×400) is recommended to confirm the low-power impression (eg, the presence of fused glands or single cells), it does not by itself contribute to the assignment of a Gleason pattern. A “Grade Group” is assigned to a tumor on the basis of the composite Gleason score.

Gleason patterns have been modified over time to increase interobserver agreement or to reflect a better understanding of tumor behavior.

Gleason patterns 1 and 2 should not be assigned except in rare circumstances and can be ignored in routine practice.

Gleason's original description of patterns 1 and 2—well-circumscribed nodules with a varying degree of uniformity and glandular shapes/sizes including cribriform patterns—is only applicable to larger resection specimens (where the entire nodule can be evaluated) and likely contained a mixture of entities such as adenosis, intraepithelial neoplasias, as well as low- and high-grade adenocarcinomas. While the 2005 ISUP Consensus Conference included modified versions of patterns 1 and 2 and recognized that they were occasionally diagnosed by expert genitourinary pathologists on resection specimens, they strongly cautioned against assigning patterns 1 and 2 on needle biopsies owing to (1) poor reproducibility, (2) poor correlation with grade at prostatectomy, and (3) risk of misguiding patients and clinicians.^73,75,76  Since that time, and despite their potentially appropriate use in transurethral resection specimens, patterns 1 and 2 have all but disappeared from routine practice.^76–79  Furthermore, the WHO and ISUP adoption of the Grade Group system effectively eliminates the distinction between patterns 1, 2, and 3 by lumping all Gleason scores less than or equal to 6 into Grade Group 1.⁷⁴ 

Gleason pattern 3 is the most commonly encountered pattern and consists of variably sized individual glands that are well formed and discrete at low magnification (×40–×100).⁷³  Small, well-formed glands that are separated by even minimal stroma are considered pattern 3. Individual well-formed glands that demonstrate branching patterns without definitive cribriform formation are considered pattern 3. A predominance of well-formed glands with only a few poorly formed glands, especially when the poorly formed glands are present at the periphery of a nodule or only identified at higher power, is considered consistent with pattern 3 and likely represents tangential sectioning of well-formed glands. Cases with morphology that is borderline between patterns 3 and 4 should be conservatively assigned pattern 3.

As of the 2014 ISUP Consensus Conference, all fused or cribriform glands are considered pattern 4 and should not be assigned pattern 3.⁷⁴  Multiple studies have shown poor outcome associated with cribriform glands.^80–83 

Gleason pattern 4 is the second most commonly encountered pattern and perhaps the most heterogeneous. Complex glands that are either cribriform, fused, or glomeruloid are convincingly pattern 4. Ill-defined glands with poorly formed lumina are also considered pattern 4, although they must be carefully distinguished from small, well-formed glands, tangential cuts, or processing artifacts of pattern 3. A cluster of at least 10 poorly formed glands not adjacent to well-formed glands is recommended to assign pattern 4.⁸⁴ 

All cribriform glands are currently considered Gleason pattern 4, except when associated with necrosis (now considered pattern 5, see below). A few studies have associated cribriform pattern 4 with a worse prognosis than other variants of pattern 4,^83,85  but this distinction has not been incorporated into current reporting.

Gleason pattern 5 represents the least differentiated and consequently most aggressive architectural pattern. Single and cords of infiltrative cells, solid masses with no or at most hardly recognizable glandular differentiation, and carcinomas with well-developed necrosis (intraluminal comedonecrosis and/or karyorrhexis) are convincing examples of pattern 5. Masses or sheets of cells with signet ring formation or rosette-like polarized spaces without recognizable glandular lumina formation are considered pattern 5.^73,74  Small solid cylinders of cells and focal but unequivocal comedonecrosis are consistent with pattern 5.

Pathologists should be conservative in reporting small foci of pattern 5. Tangentially cut poorly formed glands and glands with crush or processing artifact should not be overcalled.

Morphologic variants of PCa are relatively uncommon, usually admixed with classic PCa, and, with the exceptions of small cell carcinoma and intraductal carcinoma, should be graded by their underlying architecture.^73,74  Prostate cancer with treatment-related (radiation, chemotherapy, and/or androgen-deprivation therapy) histologic changes should not be formally graded.⁸⁶ 

Small cell carcinoma of the prostate has distinct clinical, biologic, and therapeutic implications and should not be graded.⁷³ 

Intraductal carcinoma of the prostate, which represents cancerization of preexisting ducts and acini with proliferation and atypia exceeding that seen in prostatic intraepithelial neoplasia, is not currently graded, despite striking correlation with high-grade and stage disease.^87–91  As of the 2014 ISUP Consensus Conference, the current recommendation is to report intraductal carcinoma with a comment denoting its “invariable association with aggressive prostate cancer.”⁷⁴ 

The 2014 ISUP Consensus Conference and the 2016 4th edition of the WHO Bluebook formalized the adoption of a 5-level prognostic “grade grouping” system as proposed by Epstein et al.⁹²  The new Grade Groups (GGs) are based on modified Gleason scoring and offer several advantages, including (1) simplification of the 25 variations of Gleason score into 5 prognostically significant groups^92–95 ; (2) creation of a more intuitive scale where the lowest grade disease is called grade 1 (as compared to low-grade disease being Gleason score 6 of 10); and (3) clearer separation of biologically distinct Gleason score combinations (ie, separating 3 + 4 = 7 from 4 + 3 = 7).

Gleason scores less than or equal to 6 are grouped into GG1 and enjoy the best overall prognosis. Gleason score 7 is separated into GG2 (Gleason 3 + 4) and GG3 (Gleason 4 + 3), a distinction that has long been supported by evidence.^96,97  All combinations of Gleason score 8 (3 + 5 = 8, 5 + 3 = 8, or 4 + 4 = 8) comprise GG4, and Gleason scores 9 and 10 (4 + 5 = 9, 5 + 4 = 9, or 5 + 5 = 10) comprise the worst prognostic GG5.

It is currently recommended that the 2014 WHO/ISUP modified Gleason score be reported in conjunction with the GG in prostate biopsy and resection specimens.

In biopsy specimens, the Gleason score and GG are based on combining the most predominant pattern with the worst remaining pattern. At the 2014 ISUP Consensus Conference, there was agreement that tertiary patterns should not be reported in biopsy specimens and that a minor component (<5%) of a lower pattern should not be reported. In contrast, high-grade patterns should be included in the grade regardless of quantity.

Another consensus agreement was that Gleason score and GG should be reported for each positive core if they are separately designated by the urologist. The quantity of cancer should be reported as a percentage of the core involved with optional inclusion of cancer length. In cases of discontinuous involvement, the percentage should include the full span from the first cancer gland to the last with a description of the intervening tissue.

The percentage of pattern 4 should be reported in Gleason score 7 biopsy specimens as “small amounts of pattern 4 may still warrant active surveillance.”^98–100 

Current recommendations are to assign an individual Gleason score and GG to each dominant tumor nodule within a radical prostatectomy specimen.^73,101  The reasoning is that the highest-grade nodule likely drives prognosis (ie, separate nodules of scores 4 + 4 = 8 and 3 + 3 = 6 likely behave more similarly to a score of 4 + 4 = 8 than an average Gleason score of 7). Dominant nodules are defined as discrete tumor nodules of largest size, highest grade, or highest stage. Usually a single nodule shares all 3 features; however, occasionally smaller nodules are of higher grade or higher stage,¹⁰²  and should also be assigned a size, Gleason score, and GG.⁷³ 

In most cases, the Gleason score in prostatectomy specimens should represent the most prevalent pattern followed by the second most prevalent pattern. In cases where a higher pattern is the third most prevalent pattern and composes more than 5% of the nodule, then that higher pattern should be incorporated into the score.⁷²  A minor (tertiary) pattern is assigned only in cases where a worse pattern represents less than 5% of the nodule. Minor (tertiary) patterns do not change the GG (ie, Gleason score 4 + 3 = 7, GG3, with minor pattern 5).

The percentage of pattern 4 should be reported for Gleason score 7 (GG2 and GG3) in both needle biopsy and radical prostatectomy specimens. The reasoning is that the percentage of pattern 4 is continuously correlated with prognosis,¹⁰³  and the quantity may impact patient management,^98–100  such as qualification for active surveillance or radiation therapy. In cases where the percentage of pattern 4 is near 50%, reporting the percentage helps the urologist understand that the prognosis may be borderline between GG2 and GG3. Several studies have demonstrated that increasing percentage pattern 4 at radical prostatectomy correlates with an increased risk of biochemical recurrence after radical surgery.^104,105  The method for estimating this percentage is not specifically specified, although a reasonable suggestion is as follows: 5% or less, 6% to 10%, 11% to 20%, and so on.¹⁰⁶ 

Urothelial carcinoma is the most common form of bladder cancer in industrialized countries and the second most common malignancy of the genitourinary tract after prostate cancer.^107,108  Approximately two-thirds of newly diagnosed urothelial carcinomas are noninvasive or early invasive papillary carcinomas (Ta and T1).¹⁰⁸  Among all prognostic and predictive factors for Ta and T1 papillary tumors, including multifocality, history of recurrence, concomitant carcinoma in situ, and short disease-free survival, grading seems to be among the most consequential with a well-demonstrated impact on survival and recurrence rates.^72,108–112  Grading is the most important predictive factor for noninvasive (Ta) disease.

Several grading systems have been proposed in the last half-century, and urothelial carcinoma grading has been a topic of controversy and debate among pathologist and working groups.^{108,113–116}  Currently the European Association of Urology guidelines on non–muscle-invasive urothelial carcinoma of the bladder recommends reporting both the 1973 and the 2004/2016 WHO grading systems for stratification of patients.¹¹⁷ 

In the 1973 WHO histologic typing of bladder tumors, papillary neoplasms were classified into “papilloma” and carcinomas grades 1 to 3. Papillomas were described as a papillary neoplasm containing delicate fibrovascular cores, lined by unremarkable urothelium identical to that seen in the normal bladder, and with absent or nearly absent mitotic figures.¹¹⁸  The diagnosis of carcinoma was made on the basis of presence of any degree of anaplasia, which the authors of the 1973 WHO classification defined as increased cellularity, nuclear crowding, disruption of nuclear polarity, dysmaturation of the epithelium, polymorphism, irregular nuclear shape and chromatin pattern, displaced or abnormal mitotic figures, and/or giant cells. The degree of anaplasia was the basis of the 1973 grading system. In this scheme, grade 1 (G1) tumors display the least degree of anaplasia sufficient for a diagnosis of carcinoma, grade 3 (G3) carcinomas display marked or severe cellular anaplasia, and grade 2 (G2) carcinomas comprise all tumors in between.¹¹⁸ 

The 1973 grading system has been heavily criticized for lacking specific diagnostic criteria for the 3 different grade groups, leading to high interobserver variability and poor reproducibility. These problems are reflected in the commonplace usage of ambiguous grades (eg, grade 1-2 of 3 or grade 2-3 of 3), which fail to stratify patients into specific risk groups, and a widely variable assignment of grade 2 ranging between 13% to 69% of total cases.^{109,115,116,119,120} 

The 1998 WHO/ISUP classification and grading system was designed around the objective of unifying terminology that could be widely accepted and to improve interobserver variability and reproducibility.^109,116,120  In 2004, the 1998 WHO/ISUP system was adopted by the WHO classification of tumors and it is the basis of the current 2016 grading and classification system.^{109,116,119,120} 

In this current scheme, papillary neoplasms are divided into 4 categories: papilloma, papillary urothelial neoplasm of low malignant potential, low-grade carcinoma (LGUC), and high-grade carcinoma (HGUC). The 1998 WHO/ISUP system also introduced specific descriptions and histologic criteria to define the various grades and categories.^113,119,120 

The 2016 WHO classification system is a refinement of the 2004 WHO classification and still recommends the use of the 1998 WHO/ISUP classification and grading system. Grading of urothelial neoplasms is of most importance in noninvasive disease, particularly papillary neoplasms. Of invasive tumors, 95% are HGUCs with a small proportion presenting as LGUCs (usually confined to the lamina propria). Therefore, pTa and pT1 tumors are graded into LGUC or HGUC, while pT2 tumors are considered HGUC by default (with some exceptions). The term noninvasive should be used to differentiate noninvasive from invasive disease. All 2004/2016 WHO categories/grades have specific criteria and descriptions obtaining uniformity between groups of tumors.^72,116 

The term papillary urothelial neoplasms of low malignant potential (PUNLMP) was introduced to describe noninvasive urothelial neoplasms with negligible risk of progression but increased risk of recurrence and some potential for adverse clinical outcomes requiring clinical surveillance.^72,109  PUNLMPs consist of papillary tumors with discrete, slender, and nonfused papillae, lined by normal-looking urothelium that exhibits increased cellularity and/or thickness when compared to normal urothelium. Normal architecture should be preserved with no loss of polarity, and umbrella cells should be present. Cellular atypia should be absent or minimal with only slight nuclear enlargement and overall the tumor should appear monotonous. PUNLMP nuclei display evenly distributed chromatin with absent or inconspicuous nucleoli; nuclear grooves may be present but mitoses are absent or very rare and limited to the basal layer. Necrosis should not be seen in PUNLMP.^{108,109,113,116,120}  The main histologic distinction from papilloma is that PUNLMPs show increased urothelial thickness and/or enlarged nuclei. PUNLMP recurrence rates can be significant (0% to 47%); however, progression rates are low (2% to 14%).^108,116,119  Although PUNLMP is still a controversial category, it circumvents the impact of a cancer diagnosis in a group of patients with indolent disease while keeping the need of long-term surveillance.^108,116 

To simplify the 1973 WHO classification system and eliminate the ambiguity created by the G2 intermediate grade group, the WHO/ISUP system stratifies papillary urothelial carcinoma (PUC) into low-grade (LG) and high-grade (HG) only. LGPUCs are composed of delicate papillae exhibiting extensive branching but minimal fusion, lined by thickened/cellular urothelium with an overall orderly appearance but showing some loss of polarity, pleomorphism, and variation in chromatin distribution when examined closely. Mitoses may be present at any level but usually are found in the lower half of the urothelium and limited to typical forms.^108,113,116  It should be recognized that tangential sectioning of LGPUCs at the base of the papillae can result in overrepresentation of the mitotically active immature urothelium and the appearance of broadly fused papillae; these findings should not be overinterpreted as high-grade.^113,116  It is important to keep in mind that the lowest interobserver reproducibility of the current system is in the distinction of PUNLMP and LGPUC, as histologic differences between them may be subtle.^116,121  The difference may be clinically subtle too, and although significant differences in recurrence and progression rates between PUNLMP and LGPUC have been shown by some authors,^120,122  a recent systematic review by Soukup et al¹¹⁹  demonstrated overall similar rates of recurrence and progression between these entities.

HGPUCs show a disorderly appearance due to moderate-to-marked architectural and cytologic irregularities that can be recognized at low magnification. Papillae are frequently fused and loss of cellular polarity is notable. Nuclei are usually pleomorphic with prominent nucleoli, numerous mitoses are readily seen including atypical forms, and some cases show frank anaplasia.^{108,113,116,120}  HGPUCs are aggressive lesions. The rates of progression to lamina propria and muscularis propria invasion are 25% and 5%, respectively, and reported recurrence rates are consistently around 60%.^108,119  The presence of marked anaplasia has been linked with shorter time to recurrence and progression.¹⁰⁸ 

It is important to note that HGPUCs often show histologic intratumoral variability with a mixture of grades reported in up to 45% of cases. When variability is present, the carcinoma should be assigned the highest grade present as long as it accounts for at least 5% to 10% of tumor volume.^{113,116,120,123} 

Although not currently widely used, molecular markers could have a significant role in determining tumor grade accurately. It has been shown that FGFR3 mutation is detected in 84% of LGUCs, while P53 mutation is associated with HGUCs, advanced stage, and frequent recurrence. Molecular tools may be used in the future to aid in the diagnosis of difficult cases.¹¹⁶ 

The histologic grading of genitourinary neoplasms is likely to remain relevant for years to come. Recent updates have made significant strides to improving biologic relevance and interobserver agreement for this important topic.