Grading of bladder cancer: updates, controversies and practical solutions.

Introduction
Bladder cancer is one of the most frequently diagnosed solid organ malignancies. It currently ranks as the tenth most common cancer worldwide. There is a male predominance, such that bladder cancer is the sixth most common cancer in males.
1
By convention, bladder cancer is considered under two broad categories: non‐muscle invasive bladder cancer (NMIBC) and muscle invasive bladder cancer (MIBC). MIBC refers to cancer within the muscularis propria/detrusor muscle of the bladder and is designated using pathologic staging as T2.
2
Approximately one quarter of patients will have MIBC at initial diagnosis.
The NMIBC category accounts for three quarters of initial diagnoses
3
and is a pathologically heterogeneous group which encompasses the following entities: flat in situ carcinoma (carcinoma in situ‐Tis) and papillary carcinoma
4
(Figure 1). Papillary NMIBC can be non‐invasive (Ta) or can show invasion within the lamina propria (T1) which may include invasion of the muscularis mucosa. Recognition of the muscularis mucosa is essential to avoid over‐staging a bladder cancer as T2. The MIBC/NMIBC nomenclature is largely clinical in nature and is used by treating clinicians while the T‐staging nomenclature is used for the pathology report. Urothelial cancer is the most commonly encountered histology in the bladder seen in 90% of cases. Squamous cell and adenocarcinomas may also occur along with neuroendocrine carcinomas. Not uncommonly, divergent differentiation can be seen in urothelial carcinomas such that there is a “mixed” carcinoma with components of urothelial plus another histology (squamous being the most frequent).
5
,
6
,
7
Subtype histology (previously referred to as variant histology) can also be encountered with subtypes such as micropapillary, plasmacytoid, nested, sarcomatoid, and lymphoepithelioma‐like to name a few.
4
,
8
Recognition of subtypes has become increasingly relevant given the knowledge that some confer an adverse prognosis and may not respond to systemic therapies in a similar manner to pure urothelial carcinomas.
9
,
10
,
11
,
12
,
13
,
14

The last decade has seen advances in systemic therapy options and molecular diagnostics in the bladder cancer arena.
15
,
16
,
17
Specifically, the advent of systemic immune checkpoint inhibitors
18
,
19
and now drug antibody conjugates
20
,
21
have opened up new therapeutic avenues for advanced and metastatic bladder cancer.
22

In tandem with this, has been the emergence of novel diagnostic tools including molecular assays and biomarkers.
23
Examples include assessment of fibroblast growth factor receptor (FGFR) status,
24
determining programmed death ligand‐1 (PDL‐1) and nectin‐4 expression
25
and assessing immune infiltrates. The diagnosis of bladder cancer still requires a tissue sample (either biopsy or transurethral resection‐TUR) and for patients who are non‐metastatic MIBC, the gold standard is treatment by surgery, which is radical cystectomy and pelvic lymph node dissection +/− neoadjuvant chemotherapy.
26
,
27
An alternative option is trimodal therapy (TUR with combined chemotherapy and radiation).
Routine examination using standard haematoxylin and eosin (H&E) slides remains the diagnostic tool of choice as this allows examination of tumour type and architecture, presence of divergent differentiation and subtype histology, tumour grade, tumour stage and assessment of additional features such as lymphovascular space invasion, nodal status and associated precursor/preneoplastic lesions. Accurate assessment and reporting of all these features are required to enable the treating clinician to make informed management decisions. The information contained within the pathology report is also used to determine patient prognosis, which further facilitates patient discussion and counselling.
The tumour grade and stage are integral components of pathology reports across all solid organ cancers. The stage at presentation impacts patient survival, with higher stage and more advanced disease correlating with worse outcomes.
28
In addition to stage, the tumour grade at presentation is recognized as an important prognostic variable. This pertains more to NMIBC, as almost all MIBC cases are high grade
12
,
29
,
30
with the current recommendation for the World Health Organization (WHO) to grade all invasive carcinomas as high grade.
4
Within the NMIBC category, T1 tumours (invasion of lamina propria only) are predominantly high grade, with low grade cases being a rare occurrence. There is debate as to how to grade these T1 carcinomas, as substratification within the high grade category has been shown to carry prognostic significance.
31
,
32
,
33
,
34
In contrast, Ta papillary NMIBC shows a significant difference in grade compared with T1 cases, as most are low grade.
35
,
36
,
37
In the NMIBC setting, grade is an important predictor of progression (defined as developing MIBC, ≥T2) but is not as relevant for determining tumour recurrence.
36
,
38
In this review article, we will cover the historical aspects of bladder grading, review grading schemes in current use and their diagnostic criteria, assess clinical guideline recommendations and available molecular data. Additionally, we will examine the pathologist perspective on bladder cancer grading, review grading in the MIBC setting and for non‐urothelial histology, and assess the potential of hybrid grading in the papillary NMIBC setting.

Looking Back—the History of Bladder Cancer Grading
While many readers will be familiar with the WHO grading schemes for bladder cancer (both 1973 and 2004),
39
,
40
there have been attempts to design grading schemes for papillary bladder cancers for decades prior to the introduction of the WHO 1973 system (Table 1).
In the 1920s Dr Broders from the Mayo Clinic proposed a grading scheme for epitheliomas of the genitourinary organs.
41
He proposed a 4‐tier scheme for papillary epitheliomas of the bladder based on the amount of undifferentiated cells present. Grade 1 when 25% of the lesion was undifferentiated, Grade 2 if 50% undifferentiated, Grade 3 if 75% and Grade 4 if no tendency to cellular differentiation was observed. The number of mitotic figures and the number of single, hyperchromatic nucleoli were also referenced as important considerations in grading. Bladder epitheliomas comprised one quarter of the study cohort with 80% of bladder cases being male. The breakdown of grades were as follows: 1%–17.5%, 2%–36.7%, 3%–26.7% and 4%–19.2%. A progressive increase in death rate was noted with increasing grade.
In the 1950s a Scandinavian study examined 434 cases
42
and devised a grading scheme that included infiltration beyond the basement membrane (invasion) in addition to the presence of atypia. This scheme had 7 grades with Grade 1 and 2 being non‐invasive, Grade 3 suspicious for invasion and Grade 4–7 reserved for cases with invasion and increasing amounts of nuclear atypia.
In the 1960s Bergkvist et al.
43
proposed a system with 5 grades, again based on cellular differentiation. In this work they studied 300 cases with a minimum of 8 years of follow up and found higher grades correlated with more aggressive behaviour.
In 1973 the three tier WHO 1973 grading scheme was proposed,
39
comprising grades 1–3, again dependent on the degree of cellular anaplasia. Grade 1 was defined as having the least degree of cellular anaplasia compatible with a diagnosis of malignancy, grade 3 with the most severe degree of cellular anaplasia and grade 2 having histologic features between 1 and 3. This grading scheme was incorporated into the 1973 WHO book (Figure 2).
This was followed by a proposed update to the Bergkvist system in 1986 with a 4‐tier classification by Malmstrom et al.,
44
wherein the grade 2 group was divided into 2A and 2B, with 2B cases showing a greater degree of architectural disarray than 2A tumours. Grade 3–4 was one category. Similar to previous grading schemes, assignment of grade was largely dependent on the degree of nuclear atypia. This study included 232 bladder tumours with follow up and showed reduced 5‐year survival rates with increasing grade.
In 1998, the WHO/ISUP proposed a new classification scheme for papillary urothelial neoplasms.
45
This arose from consensus work at several meetings in both the USA and Europe. The entity of papillary urothelial neoplasm of low malignant potential (PUNLMP) was introduced into the grading scheme in addition to categorising carcinomas as low and high grade in a binary fashion. In this system, histologic criteria for assigning cases to each respective category were provided and included descriptions of architectural and cytologic features.
This system was subsequently modified for the WHO 1999 publication,
46
such that a 4‐tier system was proposed consisting of PUNLMP and grade 1–3 papillary urothelial carcinoma; grade 1 representing low grade carcinoma and grade 2 and 3 the high grade carcinomas.
Five years later, the WHO published the 2004 edition
40
and in this book, the 1998 WHO/ISUP system was endorsed such that grading was a 2‐tier system of carcinoma with PUNLMP, low‐ grade, and high‐grade carcinomas. The subsequent iterations in 2016
47
and 2022
4
retained this schema of NMIBC grading.
In the subsequent years, proposals have been made to further adapt the existing WHO 2004 scheme. In 2012, Cheng et al.
48
proposed a 4‐tier grading scheme with PUNLMP reclassified as grade 1, grade 2 representing low‐grade carcinoma and separation of the high‐grade carcinoma category into grade 3 and 4 (similar to the WHO 1999 scheme) with grade 4 representing cases with complete architectural disorder and extreme nuclear anaplasia.
In 2021, an expert opinion paper from the ISUP was published which presented the possibility of a 3‐tier grading scheme representing an amalgamation of both WHO 1973 and WHO 2004.
49
In such a system, the category WHO 2004 low grade would be retained (and would include PUNLMP) while the high grade would be separated into high grade‐grade 2 and high grade‐grade 3, in line with the WHO 1973 nomenclature. The authors also proposed a detailed description for the features of WHO 1973 grade 1 and 3 carcinomas.
Following this paper, an ISUP consensus meeting was held in Basel, Switzerland in September 2022 examining various bladder cancer topics. Working group 1 focused on bladder cancer grading
50
and based on pre‐meeting survey responses proposed 3‐tier hybrid grading as a potential option for papillary bladder cancer grading. Based on the discussion and voting, there was a strong preference to refine the current grading into a 3‐tier scheme with separation of the high‐grade cases into clinically relevant subgroups.
Bladder cancer grading has undergone many iterations through the decades. Nevertheless, all proposed systems have recognized that a progressive change in both architecture and cellular features contributes to a change in grade, and that increasing grade is associated with worse patient outcomes. The contemporary systems have focused on providing detailed histologic descriptions for each grade with the aim of improving inter‐observer reproducibility. These later systems have also been evaluated more thoroughly in large studies to determine their prognostic performance.

Grading Criteria and Grade Heterogeneity
The grade assigned to a papillary NMIBC takes into account multiple histologic parameters including both architectural and cytologic features. Grade exists on a continuum with cumulative changes in the architectural component and cytonuclear features accounting for a progressive increase in the grade (see Figure 3). This means that cut points between each grade must be assigned based on morphologic interpretation of such features in the absence of a clear landmark feature. Table 2 summarizes the diagnostic criteria used to assign grade in the various proposed grading schemes.
The major criticism of the WHO 1973 system was the lack of detailed criteria which resulted in many cases being assigned to the Grade 2 category. From Figure 2, it can be seen that WHO 1973 Grade 2 encompasses cases that are both low‐ and high‐grade in other grading schemes. WHO 2004 was envisaged as a simpler scheme with a binary low/high‐grade approach, removing a middle (grade 2 like) category. The criteria for assigning cases to the low‐ or high‐grade category are well described, but despite this, pathologists self‐report difficulty in distinguishing low grade from high grade
50
and the published inter‐observer data has not demonstrated a significant improvement compared with WHO 1973 (see next section).
An updated description of WHO 1973 criteria was recently published
49
and similar criteria were used in the assessment of a 3‐tier hybrid grading approach.
51
While a combination of architectural changes and cytonuclear features is used in these recent criteria, there is a strong focus on assessing the nuclear features. The Paris System of Urine Cytology (TPS) assesses nuclear size (nuclear cytoplasmic ratio), hyperchromasia, and chromatin pattern to categorize urine cytology cases
52
and these nuclear features can also be applied to histology sections. Specifically, variability in nuclear size and shape and nuclear hyperchromasia are important features for distinguishing high grade, grade 2 from grade 3 cases. For grade 3 cases, these features are often readily appreciated at low magnification. Low grade cases show oval nuclei with some inconspicuous nucleoli but lack the coarse chromatin and pleomorphism seen in high grade cases. From an architectural perspective, loss of normal polarity and absent umbrella cells are seen in high grade cases while low grade ones show reasonable cellular polarity with cells perpendicularly aligned to the basement membrane with umbrella cells usually visible. These features may be difficult to assess when dealing with tangential sectioning and pathologists must try and assess tissue that is well oriented. Mitotic counts are not an established component of any grading scheme and there is no set number/high power field or mm2 that can be used to designate a case into a specific grade category. Mitoses can be appreciated in low grade papillary neoplasms, although they are usually infrequent, do not show atypical forms, and are largely suprabasal in location. In high grade cases, there is often clustering of mitotic figures which can extend throughout the full thickness of the urothelium. Atypical mitoses can frequently be appreciated in high grade, grade 3 carcinomas. Necrosis can also be seen in higher grade carcinomas and is often focal.
Approximately one third of papillary urothelial carcinomas show mixed grades or grade heterogeneity.
53
The WHO 2004
40
recommended that cases be assigned to the high grade category if any high grade component was present, while in 2016,
47
this was updated such that the amount of high grade should be mentioned in the report. In the latest WHO (2022), a 5% cut off was suggested such that if ≥5% high grade component was noted, the case should be assigned to the high grade category
4
and if 5% the case should be called low grade with <5% high grade component. The literature shows that cases with mixed grade tend to have worse outcome than pure low grade carcinomas even with a minor (<5%) high grade focus
54
,
55
,
56
,
57
,
58
however, the percentage cut offs have varied in studies (5%, 10% cut off). At the ISUP Basel Consensus meeting, proposed cut offs (5%, 10%, 20%, none) to designate a carcinoma as high grade were discussed with no consensus at the in‐meeting polling.
12
The majority of participants (55.9%) favoured the 5% cut off suggested by WHO 2022. It is unclear how reproducible this 5% cut off is between pathologists. A small study presented at the 2024 European Congress of Pathology showed the lowest concordance between manual assessors and automated image analysis in assessing percentage high grade in a mixed grade papillary urothelial carcinoma between 2% and 10% threshold.
59

Papillary Urothelial Neoplasm of Low Malignant Potential
The introduction of the WHO 2004 grading scheme was a move to more defined and standardized histologic grade categories; however, the inclusion of the PUNLMP category was seen as controversial. PUNLMP encompasses the extreme end of the low grade category and, in the preceding WHO 1973 system, would have been designated as Grade 1 carcinoma and histologically (see Table 2) would be a very well differentiated low grade papillary urothelial carcinoma with increased urothelial thickness, minimal atypia, and rare basal mitotic figures.
60
Pathologists self‐report difficulty in distinguishing PUNLMP from low grade papillary urothelial carcinoma
50
and inter‐observer studies confirm this finding, with PUNLMP showing the lowest agreement amongst papillary neoplasms between multiple observers.
61
,
62
,
63
Exclusion of PUNLMP from WHO 2004 increases the inter‐observer variability for this scheme (see Table 3). The clinical significance of PUNLMP in terms of patient outcome (recurrence, progression) has also been debated. The European Association of Urology undertook a retrospective review
63
of non‐invasive papillary neoplasms over a period spanning close to three decades to determine the frequency of a PUNLMP diagnosis and assess its prognostic value. Over the study period, the rate of PUNLMP diagnoses fell from 31.3% (1990–2000) to 1.1% (2010–2018) suggesting that many pathologists were reporting PUNLMP under the low grade papillary carcinoma category. There was no difference in time to recurrence between PUNLMP, Ta low, and Ta high grade cases. Similarly, there was no difference in time to progression between PUNLMP and Ta low grade cases. From the molecular perspective, PUNLMP shows remarkably similar features to low grade papillary urothelial carcinoma, including rates of FGFR3 mutations (60% PUNLMP) and TERT mutations (63%)
64
and TERT mutations in 33% PUNLMP and 50% of low grade papillary urothelial carcinomas.
65
From the clinician's viewpoint, PUNLMP and low grade papillary urothelial carcinoma are treated and followed up in a similar manner.
26
,
66
Proponents of PUNLMP terminology cite the psychological impact of a “cancer” diagnosis on patients, along with potential insurance penalties, as reasons to maintain the nomenclature. The Basel Consensus meeting survey
50
showed that one third of pathologists used the term to “avoid” a label of cancer, and 23% used it at first diagnosis in younger patients. Overall, 39.7% of respondents reported using PUNLMP terminology once or twice a year, and 22.9% never used the term. PUNLMP terminology was used more frequently amongst North American pathologists: 54.7% self‐reported never/using 1–2 per year vs 73% of European pathologists. Given the close overlap between PUNLMP and low grade papillary urothelial carcinoma in terms of biology and outcome, it has been suggested to rename both under the category of non‐invasive papillary urothelial neoplasm (NIPUN)
67
however, only 15% of ISUP Basel pre‐meeting survey participants were in favour of removing the “carcinoma” label from low grade non‐invasive papillary neoplasms, and at the in‐meeting voting, 42.9% voted to use the term papillary urothelial neoplasm (PUN) to encompass PUNLMP and low grade papillary urothelial carcinoma. 39.7% wished to continue using PUNLMP as is, and 17.5% wanted to remove the term and report under low grade papillary urothelial carcinoma. In the two decades since its formal introduction under the WHO 2004 grading scheme, PUNLMP remained a conflicting entity with variable opinions on its use. However, available data from a molecular and clinical perspective argue against retaining it as a standalone category. Under hybrid grading proposals, PUNLMP would be considered as a low grade papillary urothelial carcinoma.

Comparison of Contemporary Grading Systems Performance
Almost three quarters of patients first diagnosed with bladder cancer fall into the NMIBC category. Within this group, recurrences are common and can be seen in ~70% to 75% of patients and ~15% will progress at some time point to MIBC, defined as ≥T2.
68
Determining who will progress is a challenge and impacts decisions around follow up and treatment. The probability of tumour recurrence is anywhere from 15% to 70% at 1 year post initial diagnosis
69
,
70
and progression rates at 1 year range from 1% to 17%.
70
,
71
The rates at 5 years post diagnosis are higher.
69
,
70
,
71

Various clinical and histopathologic parameters have been studied to determine their use as prognostic factors and to develop scoring systems that can help predict an individual's likelihood of recurrence and progression. These scoring systems form the basis for NMIBC prognostic risk groups; “low, intermediate, high and very high risk” in the EAU guidelines
72
and “low, intermediate, high risk” in the AUA guidelines.
73

The two most widely used bladder grading schemes are the WHO 2004 and WHO 1973. In the EAU guidelines, dual grading with both WHO 2004 and WHO 1973 is recommended, while in the AUA guidelines, grading with WHO 2004 is recommended. Geographic location of practice and urologist preference will therefore influence how pathologists report the grade of NMIBC. As part of the ISUP Basel consensus meeting survey, respondents were queried as to the location of practice and the grading system used in daily sign‐out for papillary NMIBC. Amongst North American pathologists, 92% reported using WHO 2004, and amongst European pathologists, 60% used both WHO 2004 and WHO 1973 (dual grading). A single respondent reported using WHO 1973 alone. Amongst non‐European and non‐North American pathologists, 68% reported using WHO 2004, and 24% used dual grading.
50

In an attempt to understand which of these two systems has better prognostic performance, the EAU NMIBC guideline committee performed a systematic literature review of >3500 articles.
38
Despite reviewing 3593 abstracts, there were only 20 original studies identified in the literature that could be used for the systematic review. Data from 931 patients, representing 5 pooled studies, was used to compare the performance of each system in identifying recurrence. Most patients in these studies were Ta (non‐invasive) so that a separate analysis of T1 cases was not possible. Using WHO 1973, the recurrence rates were 33% (G1), 42% (G2) and 63% (G3) while using WHO 2004, these were 20% (PUNLMP), 38% (low grade) and 41% (high grade). Data from 1263 patients, representing 13 studies, was used to assess rates of progression. Notably, there was variability in how progression was defined within these studies, with only seven studies using the definition of progression ≥T2. The other studies also included Ta to T1 disease as progression. The pooled progression rates using WHO 1973 were: 3% (G1), 9% (G2) and 32% (G3). Using WHO 2004, these were: 1% (PUNLMP), 4% (low grade) and 25% (high grade). The authors concluded that both systems were sub‐optimal, that WHO 1973 identified more aggressive cases than WHO 2004, and that using WHO 2004 did not provide any prognostic information in T1 cases, and finally, that larger studies were required to determine which system was a better predictor of outcome.
As part of the follow up from the EAU committee review paper, a study was undertaken to look at the prognostic value of PUNLMP compared to low and high grade papillary NMIBC.
63
This study examined cases from 17 centres over a time period of 1990 to 2018 (n = 3311 Ta papillary urothelial neoplasms). Overall, the diagnosis of PUNLMP was used in 127/3311 tumours (3.8%) and the frequency of the diagnosis declined significantly from 31.3% (1990–2000) to 1.1% (2010–2018). The recurrence rates at 5 years were: 51.2% (PUNLMP), 48.3% (low grade) and 45.4% (high grade). The progression rates at 5 years were: 2.6% (PUNLMP), 2.1% (low grade) and 7.5% (high grade). There was no difference in time to progression for PUNLMP and low grade papillary urothelial carcinoma. Overall, the prognosis of PUNLMP was similar to the Ta low grade cases.
An individual patient study consisting of 5145 primary NMIBC cases (Ta and T1) was subsequently undertaken to perform a direct comparison of WHO 1973 and WHO 2004 systems.
36
In this cohort, 64% of cases were Ta, 51.3% were WHO 2004 low grade, and for WHO 1973, 23.5% were G1, 49.3% G2, and 27.2% G3. The median follow‐up was 3.9 years. Any case diagnosed as PUNLMP was included in the low grade category. On multivariable analysis, neither system was prognostic for tumour recurrence; however, both were prognostic for tumour progression, with WHO 1973 being a stronger predictor (P < 0.00) than WHO 2004 (P = 0.067). The 5‐year progression rates were: 2.6% (WHO low grade), 13.7% (WHO high grade), 1.4% (1973 G1), 5.6% (1973 G2) and 18.8% (1973 G3). Both grading systems were combined to give a 4‐tier approach such that WHO 2004 was divided into low grade G1 and low grade G2, with WHO 2004 high grade divided into high grade G2 and high grade G3. The 5‐year progression rates were as follows: low grade G1 (1.4%), low grade G2 (3.8%), high grade G2 (7.7%) and high grade G3 (18.8%). This hybrid 4‐tier approach gave a higher Harrell's concordance index (0.73) than either system alone (WHO 2004–0.67, WHO 1973–0.71). Substratification of the WHO 2004 high grade group into G2 and G3 highlighted the significantly different progression rates for these tumours.
To date, only a single North American study has systematically examined the impact of grading scheme on papillary NMIBC progression.
35
In this work, 609 cases from a time period from 2000 to 2019 were dual graded using both WHO 2004 and WHO 1973 with a median follow up of 6.5 years. Ta cases comprised 74% of the cohort, 56% were WHO 2004 low grade and for WHO 1973, 29% (G1), 49% (G2) and 22% (G3). Similar to the EAU paper, progression in WHO 2004 low grade was 7% and 40% in high grade. For WHO 1973, it was 5% (G1), 22% (G2) and 50% (G3). Progression rates were higher in T1 than Ta cases. Using a 4‐tier hybrid grading, the progression rates for Ta cases were: low grade G1 (2.9%), low grade G2 (3.3%), high grade G2 (14.3%) and high grade G3 (37%). The area under the curve (AUC) was higher for hybrid grading than both WHO 1973 and WHO 2004 as were the c‐indices. Given the comparable progression rates in the low grade G1 and G2 cases, the authors suggested retaining the WHO 2004 low grade category with substratification of the high grade cases only.
The most recent study
37
to address grading system performance was a cohort of 9712 primary Ta and T1 cases, including the previously reported 5145 cases
36
in which the authors utilized a 3‐tier grading scheme, retaining WHO 2004 low grade category and separating the high grade cases into G2 and G3. Across the 3 cohorts included, all showed time to progression was shorter for high grade G3 cases compared with high grade G2 cases, with 5 year progression rates of 7%–9% for high grade G2 and 18 to 20% for high grade G3. The authors concluded that the prognostic accuracy of hybrid 3‐tier grading was better than either WHO 1973 or WHO 2004 and that using WHO 2004 categorization would result in all high grade cases being treated similarly, which could represent under‐or overtreatment of any individual patient.
Further studies have looked at individual categories such as primary Ta grade 3 cases
74
and primary T1 grade 1 cases.
75
For TaG3 cases, the 5‐year progression rates (13%) are between those of TaG2 (3.6%) and T1G3 cases (20%). Notably, concomitant CIS changes the prognosis such that TaG3 with CIS behave comparably to T1G3 cases. The study that specifically assessed T1G1 cases found differences in 5‐year progression rates between T1G1 (6.3%), TaG1 (0.95%) and T1G2 (9.1%). Both studies show the value in subcategorizing papillary NMIBC beyond the WHO 2004 system.
Tumour grade is acknowledged as an important prognostic factor in papillary NMIBC, and grade is incorporated into various risk stratification models. There is variability in which grading system is used in daily sign out, driven by the clinical guideline recommendation used in the jurisdiction of pathologist practice. The available literature comparing prognostic performance of the two most widely used grading schemes (WHO 2004 and WHO 1973) shows that a hybrid approach utilizing both, with substratification of the clinically heterogeneous high‐grade cases, provides improved patient prognostication.

Interobserver Variability in Grading
While much has been written about bladder grading, there are few original studies on the topic, including the reproducibility or inter‐observer variability in grade assignment. As grade is an important prognostic parameter, ensuring that pathologists can reliably and reproducibly grade papillary NMIBC is crucial. As part of the Basel bladder consensus meeting, working group 1 looked at the available data around inter‐ and intra‐observer variability in grade assignment.
50
The pre‐meeting survey queried the ISUP membership on the topics, and the results showed pathologists struggle with distinguishing low and high‐grade carcinoma (54%) and distinguishing PUNLMP from low‐grade carcinoma (38%). Overall, 75% of respondents felt improvements could be made to the histologic criteria for grading (for both WHO 2004 and WHO 1973) and the majority (76%) would like to improve their grading capability using online, interactive learning modules.
Review of the published literature (see Table 3) again highlights the paucity of original work in this area with only 15 original, fully accessible publications since 2000.
51
,
61
,
62
,
63
,
76
,
77
,
78
,
79
,
80
,
81
,
82
,
83
,
84
,
85
,
86
There is wide variability in terms of the number of cases included in these studies, the number of readers/pathologists involved, the case inclusion criteria, and the methodology used. Together, this makes it challenging to directly compare all studies. Broadly, it is noted that in studies that assess intra‐observer variability, the kappa values (k) are similar irrespective of whether WHO 2004 or WHO 1973 is used. The kappa values for WHO 2004 are slightly higher than for WHO 1973 but are still moderate, and the differences are not statistically significant. The diagnosis of PUNLMP was not reproducible amongst pathologists in these studies. In one study, the reproducibility of PUNLMP diagnosis was 48% compared with 72.7% for low grade and 92% for high grade cases.
61
It was noted in several studies that the rate of PUNLMP diagnoses was low, essentially making the WHO 2004 classification into a 2‐tier one which helped increase the inter‐observer kappa values
80
,
81
and that even with teaching sets, the concordance for the diagnosis of PUNLMP did not significantly improve.
85
In one study
63
a single pathologist re‐reviewed papillary NMIBC (n = 314) 14 years apart and found that the intra‐observer agreement rate was low for PUNLMP (24.7%) vs 53.5% for low grade carcinoma and 88.5% for high grade papillary carcinomas. Distinguishing PUNLMP from low grade papillary urothelial carcinoma was noted as challenging in all studies that included PUNLMP in the cohort. Pathologists also had difficulty in separating low from high grade cases using WHO 2004, as evidenced by kappa values that were mostly in the moderate category (k 0.41–0.60). The literature results bear out the self‐reported pathologist issues noted in the Basel survey.
Given the recent observations and literature supporting the merits of the hybrid 3‐tier grading,
33
,
35
,
50
a study was conducted
51
to assess inter‐observer reproducibility amongst a group of 18 international uropathologists. A total of 30 individual cases were graded at 10× and 20× magnification to give 1076 gradings in total. Each pathologist recorded their diagnosis as low grade or, if high grade, grade 2 or 3. Despite half of the participants using WHO 2004 in daily sign‐out who were not used to substratification of high grade cases, the kappa values were substantial (k 0.61–0.80) with European pathologists only slightly higher than North American pathologists (0.663–0.682 vs. 0.623–0.647). These values are higher than reported in previous papers assessing WHO 2004 versus WHO 1973 and suggest that pathologists could readily adapt to using 3‐tier hybrid bladder grading. A limitation of this study was the inclusion of only subspecialty experts, and further work is required to determine how non‐specialists would perform.

Importance of Bladder Cancer Grading in Clinical Guidelines
Disease recurrence and progression to MIBC represent the main clinical outcomes of patients diagnosed with NMIBC. Multiple factors determine both the risk of recurrence and progression after transurethral resection, including patient age, biological sex, grade, stage (Ta or T1), presence of carcinoma in situ, tumour diameter, multifocality, and history of previous recurrence.
70
It is increasingly recognized that grade has no clinically meaningful impact on the recurrence rate, whereas it is one of the strongest prognosticators for the risk of progression to MIBC.
36
,
87
Based on a scoring of the above‐mentioned most significant clinico‐pathological parameters, the clinical guidelines of the two main urological societies, the European Association of Urology (EAU) and the American Urological Association (AUA) as well as several national guidelines, categorize patients with NMIBC into three to four risk categories. Treatment decisions are mainly based on the risk category of the patient.
73
,
88
,
89
,
90
,
91
Although the EAU risk score was originally based on the WHO 1973 grading system, both the WHO 1973 and WHO 2004/2022 grading can be used for the updated 2021 EAU NMIBC scoring model. The latter model assesses the risk of progression of NMIBC patients, but is restricted to those not treated by BCG. Notably, the EAU recommends using the risk group calculation based on the WHO 1973 when available, citing its better prognostic value. In addition, the EAU guideline incorporates PUNLMP within the WHO 2004/2022 LG tumour category. For other NMIBC patients, e.g., those on BCG maintenance, EAU risk scores are entirely based on the WHO 1973 grading system. As a consequence, the EAU guideline gives a recommendation (weak) to use the ISUP hybrid grading system to allow for its use in their various risk score systems, since it combines both grading systems. The Spanish Club Urologico Español de Tratamiento Oncologico (CUETO) scoring model for risk stratification of BCG‐treated patients is also entirely based on the WHO 1973 grading,
89
whereas the risk stratification of Swedish guidelines of NMIBC uses the 4‐tier WHO 1999 classification, distinguishing PUN‐LMP and low‐grade papillary urothelial carcinoma on the one hand and splitting the WHO 2004/2022 into WHO 1973 grade 2 and grade 3, similar to the ISUP hybrid grading. The NICE (National Institute for Care and Health Excellence) risk grouping is also mainly based on the WHO 1973 grading,
91
whereas the AUA risk score system is entirely based on the WHO 2004/2022. Both NICE and Swedish guidelines aim to improve the prognostic utility of high‐grade tumours, particularly by identifying the worse grade 3 subgroup.
The 2021 updated EAU risk score model for progression of patients with NMIBC who had no previous BCG treatment distinguishes four risk categories, that is low, intermediate, high and very high, while the more generally applicable AUA risk categorization distinguishes three risk categories, that is low, intermediate and high. It should be noted that the NCCN guidelines which use the AUA risk categories have a subcategory of “very high risk features” which includes cetain subtype histology and other features (lympovascular space invasion, prostatic urethral involvement). As an example how the EAU and AUA guidelines differ: the 2021 updated EAU low risk category comprises a primary, single, papillary Ta/T1 LG/G1 tumour < 3 cm in diameter without CIS in patients ≤70 years as well as Ta LG/G1 tumours (no CIS) with at most one additional risk factor (that is either age >70, multifocality or tumour diameter >3 cm). The AUA low risk category includes papillary TaLG urothelial carcinomas with a diameter of <3 cm and all PUNLMPS (irrespective of diameter). Since TaLG >3 cm are intermediate risk the only difference in risk category and management between PUNLMP and TaLG would be for the very rare PUNLMP >3 cm. A recent abstract presentation validated the application of risk stratification tools for NMIBC and found that the EAU crteria were more accurate than the AUA in predicting cancer specific survival.
92

Assignment to risk category depends on the number of additional clinical risk factors of tumour of the same grade and stage. This holds true for both the EAU, CUETA, AUA and NICE risk categories. The EAU categorization is more complex as compared to that of the AUA and NICE because it takes into account more adverse clinical risk factors than the latter guidelines. Thus, according to the 2021 updated EAU risk score, a TaHG tumour without CIS and no other risk factors is intermediate risk, TaHG without CIS but with at least two risk factors or TaHG with CIS and up to two risk factors is high risk, and TaHG with CIS and 3 risk factors is very high risk. According to the AUA guideline, a papillary TaHG carcinoma with a size < 3 cm and no CIS is intermediate risk, but those with a size larger than 3 cm as well as recurrent TaHG are high risk. The EAU guideline further considers all patients with subtype pathology very high risk, while the AUA and NCCN consider under the high risk with a subcategory of very high risk features which includes subtype pathology. Follow‐up and treatment preferences for patients within a specific risk category may differ dependent on the guideline and geographic jurisdiction.

Grading and Molecular‐Genetics
The most common genetic alteration in bladder cancer is activating point mutations of the FGFR3 gene. FGFR3 mutations are the hallmark of the low grade papillary pathway of the dual‐pathway model of bladder carcinogenesis.
93
,
94
They are detected in about 80% of PUN‐LMP and low grade papillary urothelial carcinomas, in 16% of grade 3 carcinomas, while they are essentially absent in primary dysplasia and carcinoma in situ.
94

FGFR3 mutant urothelial carcinoma mostly represents the luminal papillary molecular subtype of urothelial carcinoma. TP53 mutations are characteristic of the carcinoma in situ‐high grade carcinoma pathway of bladder carcinogenesis.
93
,
95
Carcinomas of this pathway are often of the basal squamous and genetically unstable molecular subtype, but their genotype is much more heterogeneous as compared to the luminal papillary subtype.
95
High grade carcinomas are also characterized by alterations in several cell‐cycle regulatory genes and associated with elevated Ki‐67 scores reflecting their increased proliferation. Cell‐cycle markers, however, failed to improve prediction of recurrence and progression of the EAU and CEUTA risk models.
96

Although uncommon, low grade papillary urothelial carcinomas may evolve into high grade urothelial carcinomas, a process manifested morphologically by grade heterogeneity. This progression of papillary low grade urothelial carcinoma is associated with loss of CDKN2A.
97
In a study of papillary urothelial carcinomas with grade heterogeneity, this loss of CDKN2A and increased proliferative activity (Ki‐67 score) were often present both in the low grade and high grade tumour areas.
98
This finding implies that genetic alterations associated with progression precede the histopathologically visible change in grade. In a large long‐term follow‐up study of non‐invasive (Ta) papillary low grade urothelial carcinomas, Villegas et al.
99
also noted that adverse transcriptomic changes of more aggressive carcinoma were already present in four TaLG patients who progressed to metastatic carcinoma, further highlighting the discrepancy between phenotype and genotype for a small subset of TaLG.
Jackson et al. categorized Ta bladder cancers as urothelial‐like, urothelial‐like/KRT5 high, or genomically unstable. The urothelial‐like/KRT5 group had the best outcomes of all groups, suggesting a combination of cell cycle activity and basal‐squamous gene expression may offer better guidance than cell cycle activity alone.
100
Molecular subtyping may thus objectively identify patients with Ta cancer at high risk of recurrence and progression, particularly by grouping tumours based on multiple expression signatures, beyond a single measure of cell cycle activity. This could guide the decision to treat with intravesical BCG and offer early cystectomy in patients with multiple recurrences.
95

Using FGFR3 status and immunohistochemical expression of Ki‐67, P53 and P27, multivariable analysis of clinical data supported a four‐tiered grading system using the 1973 and 2004 WHO classifications, with one low‐grade (PUNLMP/LG/G1) category, two intermediate grade (LG/G2 and HG/G2) categories and a separate high‐grade (HG/G3) category.
101
Two studies in patients with NMIBC reported a higher frequency of both focal genomic amplifications and a total number of chromosomal breaks in grade 3 compared to grade 1 and 2 tumours.
102
,
103
In the latter study, 161 patients with NMIBC who were not treated with BCG, grade 3 (WHO 1999) was associated with progression, whereas WHO 2004 high grade was not. Thus, based on outcome, molecular information from chromosomal and mutational data revealed larger differences between grade 2 and 3 than between grade 1 and 2.
Molecular grading using FGFR3 status and Ki‐67 score was proposed as an alternative for bladder cancer grading.
104
In multivariate analysis with progression and disease‐specific survival as end points, the combination of FGFR3 and MIB‐1 proved independently significant.
104
Molecular grading using gene‐expression data from 144 patients has been proposed as a relevant complement to standard pathological grading
105
based on the finding that the grade signature for grade 3 was associated with increased risk of progression in a heterogeneous cohort receiving different treatments. Furthermore, a previously defined 12‐gene signature for progression in NMIBC, that was validated in a prospective multicenter setting adjusting for EORTC‐risk score, was independently associated with progression.
106

Grading Invasive Bladder Cancers and Non‐urothelial Histology

Grading of Muscle‐Invasive Bladder Cancer
Most muscle invasive bladder cancer (MIBC) is high grade tumours. Invasion is an unexpected finding in low grade tumours, and by convention, it is much more commonly encountered in high grade lesions, reaching 70%–96% in some series.
107
,
108
No standardized criteria have been reported to define low‐grade invasive urothelial carcinoma.
4
The 2022 WHO blue book recommends that invasive urothelial carcinoma should be graded as “high grade”, acknowledging the rare occurrence of invasive low‐grade urothelial carcinoma.
4

There is a paucity of data on the grading of ≥T2 tumours, where grade probably has no value as the higher stage supplants grade as the determinant of prognosis.
12
All MIBC are now considered high‐grade under the current WHO 2022 classification of urinary tract tumours.
4

Grading of morphologic subtypes and cases with divergent differentiation is problematic because of their atypical and diverse histologic features.
12
Some morphologic subtypes such as nested urothelial carcinoma, tubular‐microcystic urothelial carcinoma and large nested urothelial carcinoma have deceptively bland nuclei and have previously been referred to as “bland” subtype. In contrast, others such as sarcomatoid urothelial carcinoma, lymphoepithelioma‐like urothelial carcinoma, and giant cell urothelial carcinoma have innately high‐grade nuclei.
4
The unique morphologic patterns exhibited by certain subtypes, such as micropapillary, plasmacytoid, lipid‐rich urothelial carcinoma, and poorly differentiated urothelial carcinoma, in conjunction with tumours that harbour non‐urothelial carcinoma components such as squamous, glandular, trophoblastic, and sarcomatoid presents challenges to grading using the current WHO system for urothelial carcinoma.
4
No unique histologic grading has been designed specifically for any urothelial carcinoma subtype or divergent differentiation.
12
The 2022 WHO system recommended that all subtypes and divergent differentiation be considered as high‐grade. Sarcomatoid, plasmacytoid and micropapillary urothelial carcinoma show worse (sarcomatoid) and likely inferior (plasmacytoid, micropapillary subtypes) outcomes when compared with stage matched conventional urothelial carcinoma.
12
Thus, this is a pragmatic approach to these invasive tumours which behave at least comparable to or worse than invasive conventional urothelial carcinoma with the possible exception of the rare lymphoepithelioma‐like subtype (Table 4).
4
,
12
At the Basel Consensus meeting, 85% of participants voted to apply “high grade” to all recognized urothelial carcinoma subtypes and divergent differentiation.
12

Grading of T1 (Lamina Propria Invasive) Carcinoma
There is more limited data on impact of grade on T1 carcinomas compared with their Ta counterparts. Historical data has shown that the vast majority of invasive cases are high grade with <10% designated as low grade.
31
,
34
,
109
Indeed, the current WHO recommendation is to report all invasive carcinomas as high grade. Two studies of note specifically looked at invasive, low grade carcinomas. In the first study,
29
4/23 cases showed T2 disease in addition to T1 disease. Some unusual histologic features for low grade carcinomas were described in this cohort including apoptotic debris and brisk mitotic activity, which are more commonly encountered in high grade lesions. The second study
30
examined 41 cases and concluded that the rates of recurrence and progression were in keeping with those seen in Ta carcinomas. The WHO 1973 system allows high grade T1 cases to be categorized as G2 or G3 and using this scheme
32
it was shown that WHO 1973 grading was prognostic for outcome (cancer specific and progression free survival). A further study
110
also observed significant differences in progression and survival between WHO 1973 grade 2 and 3 T1 carcinomas. In analysis of large NMIBC cohorts, T1 cases represent usually one third of any cohort and progression rates always exceed those of Ta cases. When T1 cases are examined separately from Ta cases, applying WHO 1973 grading will stratify for progression with improved Harrell's C‐indices (0.60 vs 0.55 WHO 2004
36
) and shows differences in 5 year survival. Application of a hybrid grade in these cases also show differences in T1 high grade 5 year outcomes (0.41 for high grade, G3 vs 0.52 for high grade, G2).
35
Hybrid grading schemes confer an advantage over WHO 2004 grading for T1 cases by enabling substratification of a cohort that otherwise are mixed together under the term “high grade” in WHO 2004.

Grading of Adenocarcinoma of the Bladder
Adenocarcinoma of the bladder has been traditionally graded based on the degree of differentiation (gland formation and cyto‐nuclear features) into well, moderately, and poorly differentiated like adenocarcinomas (see Figure 4) from other organ sites.
4
However, contemporary studies with a central review on grading bladder adenocarcinoma are also limited.
12
The earlier study by Anderström et al.
111
showed nonsignificant worse survival rates for poorly differentiated bladder adenocarcinomas compared to moderately and well‐differentiated tumours. Zaghloul et al.
112
showed significantly different distant metastasis‐free rates (46% for grade 1 vs. 38% for grade 3) amongst the three adenocarcinoma grades, although it had no significant effect on survival. Several recent studies on bladder adenocarcinomas (including urachal adenocarcinoma) from the SEER database showed grade (with varied cutoffs) to be an independent predictor of outcome.
113
,
114
,
115
Grading invasive pure adenocarcinoma of the bladder according to the degree of differentiation (as defined above) was recently endorsed by the 2023 Dublin International Society of Urological Pathology (ISUP) Consensus Conference.
116
,
117

There are no specific histological grading systems that have been proposed for urachal adenocarcinomas and non‐glandular carcinomas in the literature.
118
Grading approaches to these tumours are usually in similar manners to those used in their histologically analogous tumours (e.g., adenocarcinoma, urothelial carcinoma, and squamous cell carcinoma) from the urinary tract or other organ sites.
117
,
119

Grading Bladder Adenocarcinoma of Mullerian Type
Rare adenocarcinomas of suggested Mullerian derivation may arise in the bladder, including clear cell adenocarcinoma and endometrioid adenocarcinoma. Clear cell adenocarcinoma has been considered an aggressive tumour, with a 5‐year survival rate of 40% for bladder clear cell adenocarcinoma.
12
However, the clinical course is stage‐driven, with patients presenting at an advanced stage more often than those with urothelial carcinoma. No specific grade system applies to clear cell adenocarcinoma of the bladder.
4
Endometrioid carcinoma of the urinary tract shows morphological similarities to its uterine counterpart. It is characterized by a glandular pattern and back‐to‐back glands, with grades based on different degrees of differentiation (low, moderate, and high) and nuclear pleomorphism.
4

Grading of Squamous Cell Carcinoma
Pure squamous cell carcinoma of the bladder, like those from other organ sites, has been traditionally graded based on the degree of differentiation (keratinization and intercellular bridges, and nuclear pleomorphism) into well, moderately, and poorly differentiated.
4
However, contemporary studies with central pathology review on the grading of bladder squamous cell carcinoma are limited.
12
Lagwinski et al.
120
graded 45 bladder squamous cell carcinomas as well, moderately, and poorly differentiated, and did not find the grade to correlate with the outcome. Interestingly, the degree of differentiation tends to be better in bilharzial than non‐bilharzial squamous cell carcinomas.
121
A recent review of 1210 squamous cell carcinoma from the SEER database reported that the grade of such tumours is not a significant predictor of cancer‐specific survival.
122
In contrast to pure squamous cell carcinoma, urothelial carcinoma with divergent squamous differentiation should be considered as high‐grade according to the current WHO grading system.
4
Grading invasive pure squamous cell carcinoma of the bladder according to the degree of differentiation (as defined above) was endorsed by the ISUP Consensus Conference on Urinary Bladder Cancer held in Basel, Switzerland on September 3, 2022.
12

Pathologist and Clinician Perspective on Grading Schemes
As part of the ISUP Basel consensus meeting 2022, two online surveys with 10 questions were circulated. The first was to the ISUP membership and the second to the EAU membership. There were 214 ISUP respondents and 191 EAU respondents. For both cohorts, most respondents practiced in an academic/university affiliated health centre. Within the EAU group, 175/191 (91.6%) were practicing urologists and from the ISUP group, 176/214 (82.2%) were practicing pathologists with 65% in practice over 11 years.
123

There were differences between the pathology and urology groups in terms of grading systems used in practice (Table 5). Within the pathology community, WHO 2004 was most widely used (predominantly amongst North American pathologists) while dual grading using both WHO 1973 and 2004 was the grading scheme most urologists saw in practice (49%). A small minority of both groups (<1% pathologists and 8.4% urologists) exclusively utilized WHO 1973. There was a marked difference in the number of PUNLMP cases encountered annually. Almost 48% of urologists encountered this diagnosis 1–2 times per year while 40% of pathologists encountered at the same frequency. The highest rates of PUNLMP diagnoses were reported by North American pathologists: 54% stated never/rarely reporting vs 74% European pathologists.
50
The surveyed members from both organizations were aligned on their perspective on management of PUNLMP with 59% of pathologists and 58% of urologists believed PUNLMP and Ta low grade papillary carcinomas should be managed similarly. The EAU respondents reported that WHO 1973 grade 3 should be distinguished from WHO 2004 high grade as it would influence clinical decisions in both Ta and T1 NMIBC (37.4%) versus 21% of pathologists. Almost a quarter of pathology respondents (23.8%) indicated they were uncertain about this question versus 11% of practicing urologists. Interestingly 21% of urologists were in favour of reverting back to WHO 1973 versus <3% of pathologists. The urology respondents favoured a 3‐tier reporting scheme in current practice (42%) compared with 40% of pathologists. The pathology preference was for a 2‐tier scheme (52%) vs 30% of urologists. When asked about the future of papillary NMIBC grading, the preferred option for both ISUP and EAU members was a hybrid 3‐tier approach leaving the WHO 2004 low grade category in place and splitting the high grade category into grade 2 and 3 cases (33.7% EAU and 36% ISUP).
It should be noted that the urology members of the EAU are mostly practicing in Europe and likely adhere to the EAU guideline statements on grading which recommend dual grading of papillary NMIBC using both WHO 2004 and 1973.
72
However, the survey data showed that 38% of European pathologists reported using only the WHO 2004 scheme in their day to day practice which closely aligned with the urologist response; that they saw WHO 2004 used in reporting (41%).
A recent editorial
124
has also highlighted the potential utility of hybrid 3‐tier grading for stratifying patients into different arms of clinical trials as well as ensuring those at the more extreme end of the high grade spectrum (HG‐G3) be considered for earlier aggressive intervention. In addition to the EAU guidelines, the United Kingdom Royal College of Pathologists have included the hybrid grade as part of their urinary system reporting tissue dataset.
89
Overall, the pathology and urology communities are aligned with respect to their opinions regarding PUNLMP and preference for hybrid 3‐tier grading. The survey results show urologists favour “more information” with a lower preference for 2‐tier grading compared with pathologists and a significantly higher preference for 4‐tier grading (5% vs. 18%).

Conclusion
Grading of bladder cancer has undergone multiple iterations over the preceding century. The two most widely adopted systems for papillary neoplasms have been the three tier WHO 1973 and WHO 2004 grading schemes. Both are problematic in their own way. Inter‐observer reproducibility is sub‐optimal (more for WHO 1973) and pathologists self‐report challenges in distinguishing low‐ from high‐grade cases using WHO 2004 and distinguishing PUNLMP from low grade. The evolving role of artificial intelligence will likely impact further studies on grading and grade reproducibility. The readers are directed to the paper on artificial intelligence in bladder cancer in this annual review issue for more information.
The clinical utility in stratifying patients is another relevant issue with use of the WHO 1973 system resulting in many “G2” cases which sometimes did well and other times had adverse outcomes, limiting its ability to discriminate good from bad outcomes. WHO 2004 has not resolved this issue as all high grade cases are categorized together with the potential to either under‐ or overtreat patients and provides no risk stratification for T1 carcinomas. From the limited molecular studies to date, it is known that high grade G3 cases are different to high grade G2 cases, the latter showing more molecular similarities with low grade cases. Clinical outcome studies also support this distinction within high grade cancer with increased, and faster time to progression in G3 vs G2 high grade cases in both non‐invasive and invasive tumours.
The diagnosis of PUNLMP has been decreasing over the last three decades and is largely used to circumvent the label of “cancer”, particularly in young patients. Biologically, PUNLMP is similar to very low grade papillary urothelial carcinoma in terms of molecular profile and clinical course and can be best considered under the category of low grade papillary urothelial carcinoma.
Pathologists and clinicians are aligned on the need to provide optimal patient outcomes through improved risk stratification and appropriate therapy selection.
49
,
50
,
123
,
124
The number of “tiers” to be used in grading has been debated, with some proponents advocating hybrid grading using a 4‐tier system essentially subdividing the WHO 2004 low grade into grade 1 and 2 with grade 1 encompassing PUNLMP.
125
The high grade cases would be subdivided into grade 2 and 3. Advocates of this approach emphasize the importance of identifying extreme low grade cases (e.g. low grade, grade 1 cases) as these exist further from the low grade/high grade, grade 2 cut off and would allow clinicians to tailor follow up to be more conservative in these cases. Additionally, utilizing more grade categories could provide a better description of where a tumour lies in the grade continuum. In contrast, pathologists and urologists have a preference to use hybrid 3‐tier grading, which encompasses the ease of WHO 2004 (in maintaining a single low grade category) with the prognostic performance of WHO 1973 (subdividing high grade category into G2 and G3) as evidenced by urologist responses to the pre‐meeting survey and the pathologist in‐meeting votes at the Basel Consensus meeting. Recent publications have shown the clinical utility of hybrid 3‐tier grading
35
,
36
,
37
and how updated/refined grading criteria can improve inter‐observer agreement in assessing cases.
51
Hybrid grading, either 3‐ or 4‐tier has potential benefits over existing grading schemes. Hybrid 3‐tier grading is preferred as it shows good reproducibility and represents a pragmatic and clinically relevant alternative to the grading schemes in current daily use.

Author contributions
MRD, THvdK, ALB, and LC wrote the review article. All authors provided substantial contributions to this review, drafted and critically revised the article, and approved the final version for publication.

Conflict of interest
The authors have no conflicting interests to disclose.

Funding information
This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors.