Received: 27 December 2022 Revised: 7 August 2023 Accepted: 10 August 2023 DOI: 10.1111/vco.12932 ORIGINAL ARTICLE Standardization of canine meningioma grading: Validation of new guidelines for reproducible histopathologic criteria Sara Belluco 1 | Giuseppe Marano 2 | Thibaut Lurier 3,4 | Giancarlo Avallone 5 | 6 7 8 9 Chiara Brachelente | Stefano Di Palma | Roberta Rasotto | Kerstin Baiker | Andreas Beineke 10 | Anna Oevermann 11 | Frauke Seehusen 12 | Fabiano José Ferreira de Sant'Ana 13 | Patrizia Boracchi 2 | Martí Pumarola 14 | Maria Teresa Mandara 15 1 Université de Lyon, VetAgro Sup, ICE UPSP 2016.A104, Axe Cancérologie, Marcy l'Etoile, Lyon, France 2 Department of Biomedical and Clinical Sciences "L. Sacco", University of Milan, Milan, Italy 3 INRAE, VetAgro Sup, UMR EPIA, Université Clermont Auvergne, Saint-Genès-Champanelle, France 4 INRAE, VetAgro Sup, UMR EPIA, Université de Lyon, Lyon, France 5 Dipartimento di Scienze Mediche Veterinarie, Università di Bologna, Bologna, Italy 6 Dipartimento di Medicina Veterinaria, Università degli Studi di Perugia, Perugia, Italy 7 IDEXX Laboratories, Wetherby, UK 8 Diagnostic Pathology, Dick White Referrals, Station Farm, Cambridgeshire, UK 9 Department of Veterinary Clinical Sciences, Jockey Club College of Veterinary Medicine and Science, City University of Hong Kong, Kowloon, Hong Kong 10 Stiftung Tierärztliche Hochschule Hannover, Institut für Pathologie, Hannover, Germany 11 Division of Neurological Sciences, Vetsuisse Faculty, University of Bern, Bern, Switzerland 12 Institute of Veterinary Pathology, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland 13 rio de Diagno Laborato stico Patolo gico Veterinário, Universidade de Brasília, Brasília, Brazil 14 Dept Medicina i Cirurgia, Animals, Facultat de Veterinària, Barcelona, Spain 15 Laboratorio di Neuropatologia, Dip. di Medicina Veterinaria, Università degli Studi di Perugia, Perugia, Italy Correspondence Sara Belluco, Laboratoire d'histopathologie Abstract vétérinaire, VetAgro Sup, Campus vétérinaire, Canine meningiomas are currently graded using the human grading system. Recently 1, avenue Bourgelat, 69280 Marcy l'étoile, France. published guidelines have adapted the human grading system for use in dogs. The Email: sara.belluco@vetagro-sup.fr goal of this study was to validate the new guidelines for canine meningiomas. To eval- Funding information uate the inter-observer agreement, 5 veterinary surgical pathologists graded VetAgro Sup 158 canine meningiomas following the human grading system alone or with the new guidelines. The inter-observer agreement for histologic grade and each of the grading criteria (mitotic grade, invasion, spontaneous necrosis, macronucleoli, small cells, hypercellularity, pattern loss and anaplasia) was evaluated using the Fleiss kappa index. The diagnostic accuracy (sensitivity and specificity) was assessed by comparing the diagnoses obtained with the 2 grading systems with a consensus grade (consid- ered the reference classification). The consensus histologic grade was obtained by This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2023 The Authors. Veterinary and Comparative Oncology published by John Wiley & Sons Ltd. Vet Comp Oncol. 2023;21:685–699. wileyonlinelibrary.com/journal/vco 685 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 686 BELLUCO ET AL. agreement between 4 experienced veterinary neuropathologists following the guide- lines. Compared with the human grading alone, the canine-specific guidelines increased the inter-observer agreement for: histologic grade (κ = 0.52); invasion (κ = 0.67); necrosis (κ = 0.62); small cells (κ = 0.36); pattern loss (κ = 0.49) and ana- plasia (κ = 0.55). Mitotic grade agreement remained substantial (κ = 0.63). The guide- lines improved the sensitivity in identifying grade 1 (95.6%) and the specificity in identifying grade 2 (96.2%) meningiomas. In conclusion, the new grading guidelines for canine meningiomas are associated with an overall improvement in the inter- observer agreement and higher diagnostic accuracy in diagnosing grade 1 and grade 2 meningiomas. KEYWORDS central nervous system, dog, grading, meningioma, standardization 1 | I N T RO DU CT I O N using the human grading system alone or in combination with the recently published guidelines. Second, we compared the diagnostic Canine meningioma is the most frequent primary central nervous sys- accuracy of the two grading systems, performed by surgical patholo- tem tumour in dogs, accounting for up to 51% of intra-cranial gists, with a consensus grade obtained by agreement between four tumours.1 Intra-cranial meningioma frequently causes seizures, altered experienced veterinary neuropathologists. mentation and vestibular syndrome, while intra-spinal meningioma causes ataxia and paresis. One-year survival after surgery is approxi- mately 50%.2,3 To date, neither clinical factors nor histologic features 2 | METHODS have been correlated with prognosis in canine meningioma patients. Therefore, the biologic behaviour of the tumour and the patient out- The research protocol was reviewed and approved by our institutional comes cannot be predicted in a clinical setting, resulting in a lack of Research Ethics Committee. targeted treatments for affected dogs. In veterinary medicine, classification of nervous system tumours dates back to 1999.4 However, this older classification system has 2.1 | Cases several limitations compared with more recent literature. Therefore, based on clinical, imaging and histologic similarities between canine A total of 158 canine meningiomas, previously used to develop the and human meningiomas, histologic grading of canine meningioma is guidelines, were included in this study.22 The four neuropathologists currently performed by applying the histologic criteria in the 2016 reached a consensus on histologic and mitotic grades for 151 and WHO human grading system.5–8 In human medicine, histologic grade 148 cases, respectively. is a robust predictor of survival and recurrence in meningioma Tumours were surgically resected for the benefit of the animal patients.9–11 However, the human grading system has not been corre- and fixed in 10% buffered formalin. For each case, samples were rou- lated with prognosis in dogs. tinely processed for histopathology, cut into 4-μm sections, stained A reproducible and reliable grading system is necessary to corre- with haematoxylin and eosin and digitized with a NDP scanner (NDP late histologic findings with prognosis in both human medicine12–17 scan 2.5.90, Nanozoomer HT, Hamamatsu) at 20 magnification and veterinary medicine.18–21 The reproducibility of the human (454 nm/pixel). Slides were visualized with the freely available NDP.2 meningioma grading system applied to canine meningioma was previ- viewer (NDP.view2 Viewing software U12388-01, Hamamatsu ously investigated to lay the groundwork for studies on the prognostic Photonics). significance of histologic grade in canine meningiomas. Use of the human grading system to grade canine meningiomas resulted in low reproducibility, possibly due to unclear descriptions of some of the 2.2 | Histologic evaluation of the tumours diagnostic criteria in the veterinary literature.5,6 Therefore new guide- lines were recently published to increase the reproducibility of histo- Each of the 158 slides was analysed twice by 5 board-certified veteri- logic grading applied to canine meningioma.22 nary surgical pathologists to define the mitotic grade and the histo- The goal of this study was to validate the recent guidelines for logic grade. The first reading was performed following the human the assessment of histologic grade in canine meningioma. First, we grading system applied to canine meningioma, as previously published evaluated the inter-observer agreement among surgical pathologists in the veterinary literature (Table 1).5,6 The second reading was done 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 687 TABLE 1 Human histologic grading for canine meningioma.5,6 T A B L E 2 Guidelines for reproducible criteria on canine meningioma grading.22 Grade Criteria Histologic Grade 1 Tumours lacking criteria for grades 2 and 3 criteria Guidelines Grade 2 Tumours with mitotic grade of 2 or tumours with central nervous tissue invasion or tumours displaying at least 3 Mitotic count Evaluated: of the following criteria: • in the most mitotic area • Sheeting architecture • in consecutive high power fields to cover • Small cells 2.37 mm2 area • Hypercellularity When mitotic count is closely lower than the cut- • Macronucleoli off (1–2 mitoses of difference), a supplemental • Spontaneous necrosis count should be performed in another highly mitotic area Grade 3 Tumours with mitotic grade of 3 or tumours with extreme anaplasia Invasion Presence of tumour cells into the brain or the spinal cord, not surrounded by pial layer cells Spontaneous Focal or multifocal presence of spontaneous necrosis necrosis blindly between 18 and 21 months after the first reading; tumours Small and large foci are equally considered were graded following the recently published guidelines (Table 2).22 When located at sample margins or in Since canine meningioma subtypes have not been definitively hemorrhagic areas, artefactual necrosis should be ruled out correlated with tumour behaviour and prognosis, tumour subtype was In abscesses, reported if tumour cell necrosis is not considered for the histologic grade. evident Macronucleoli Focal or multifocal presence of nucleoli visible at 100X 2.3 | Mitotic grade Small cells Focal or multifocal presence of cells with a high nucleus/cytoplasmic ratio or with a lymphocytic The mitotic grade was calculated by counting the number of detected appearance Evaluated at low magnification mitoses according to the human grading system (first reading) or fol- lowing the new guidelines (second reading). Since the cut-offs for Hypercellularity Focal or multifocal Evaluated at low magnification mitotic grade are different in the human grading system and in the Evaluated separately from small cells guidelines, the mitotic grade was calculated based on the two differ- Pattern loss Replace the “sheeting architecture” ent cut-offs, as reported in Table 3, and the histologic grade was sub- Defined as not identifiable architectural pattern in sequently calculated. more than 50% of the tumour surface Inter-observer agreement was calculated for all four mitotic Evaluated at low magnification grades obtained. Anaplasia Replace “extreme anaplasia” Defined as the multifocal or diffuse presence of anaplastic cells, whose meningeal origin is not evident 2.4 | Macronucleoli In the article outlining the new meningioma guidelines, concern was T A B L E 3 Mitotic grades, following what reported in the human raised that the 100 magnification might be too low for assessing literature as published in veterinary medicine (old cut-off)5,6 and macronucleoli in canine meningiomas (comment in the discussion22). following the guidelines (new cut-off).22 Therefore, in order to evaluate how magnification could influence Grade Old cut-off New cut-off inter-observer agreement on this criterion and, as a consequence, on 1 <4 mitoses in 2.37 mm2 <8 mitoses in 2.37 mm2 the histologic grade, macronucleoli were evaluated at 100 and 200 (adjusted to screen size22). Inter-observer agreement was statistically 2 Between 4 and 19 in Between 8 and 40 in 2.37 mm2 2.37 mm2 analysed for macronucleoli and histologic grade at 100 and 200. 3 ≥20 mitoses in 2.37 mm2 ≥41 mitoses in 2.37 mm2 2.5 | Statistical analysis of the 1580 evaluations, we recorded the histologic grade (grade 1, 2 Data consisted of 1580 plus 299 records. The 1580 records included and 3) and the following histologic criteria: mitotic grade (grade results of histopathologic evaluation of the 158 slides performed by 1, 2 and 3) evaluated using the old and new cut-off values; spontane- 5 surgical pathologists in 2 readings. The 299 records comprised the ous necrosis (yes/no); macronucleoli (yes/no); small cells (yes/no); consensus evaluations of histologic grade and mitotic grade (151 and increased cellularity (yes/no); pattern loss (yes/no); invasion (yes/no) 148 records, respectively) performed by neuropathologists. For each and anaplasia (yes/no). 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 688 BELLUCO ET AL. F I G U R E 1 Inter-observer agreement among surgical pathologists, following the human grading system alone (first reading) and following the guidelines (second reading). With the guidelines, the inter-observer agreement was increased for histologic grade and five criteria (invasion, spontaneous necrosis, small cells, pattern loss and anaplasia). Agreement for mitotic grade was similar regardless of whether the new guidelines were applied. Agreement decreased when applying the guidelines for two criteria (prominent nucleoli and increased cellularity). *Mitotic grade calculated with the cut-off recommended in the guidelines (new cut-off). Inter-observer agreement among surgical pathologists using The analyses were performed using the R software version 4.2.4 the human grading system (first reading) and the human grading (R Core Team,25 with the packages irrCAC26 and geepack27 added system in combination with the guidelines (second reading) was and Knime Analytics Platform release 4.6.0. assessed by estimates of the Fleiss kappa index. The level of agree- ment was interpreted using the Landis and Koch interpretation23: slight agreement for kappa values between 0.00 and 0.20; fair 3 | RE SU LT S agreement for values between 0.21 and 0.40; moderate agreement for values between 0.41 and 0.60; substantial agreement for values 3.1 | Inter-observer agreement between 0.61 and 0.80; and almost perfect agreement for values between 0.81 and 1.00. Along with the kappa index, the “uncor- The inter-observer agreement among surgical pathologists was rected” (not corrected for chance-expected agreement) percentage greater for histologic grade and 5 out of 8 histologic criteria, when the of agreement was reported, in order to provide additional insights. guidelines were applied than when using the human grading system This index is defined as the percentage of concordant pairs over alone (Figure 1 and Table 4). At the first reading (human the total number of paired results for every pair of surgical system alone), three of the nine criteria (histologic grade, invasion and pathologists. spontaneous necrosis) had moderate agreement, 1 (mitotic grade) had The methods described above were also used to evaluate the substantial agreement and none had almost perfect agreement. At the inter-observer agreement for macronucleoli determined at 100 and second reading (using the guidelines), three of nine criteria had mod- 200 magnification. erate agreement (histologic grade, pattern loss and anaplasia) with The agreement with the consensus histologic grade (considered increased kappa index and three criteria had substantial agreement as the reference classification) was evaluated by estimates of classifi- (mitotic grade, invasion and spontaneous necrosis). Small cells still had cation accuracy. In order to account for the association among the fair agreement, although the kappa index was increased at the second classifications performed on each single slide (“within-slide associa- reading, compared with the first reading. tion”), estimates of sensitivity, specificity and respective 95% confi- In contrast, macronucleoli and hypercellularity had decreased dence intervals were obtained by generalized estimating equation agreement with the guidelines, compared with the human grading (GEE) methods, as described by Genders et al.24 alone, since the agreement went from fair, at the first reading, to Furthermore, differences in sensitivity and specificity between slight, at the second reading. histologic grades evaluated by the human grading alone and using the guidelines were obtained as follows: for each histologic grade (1–3, and), a test of hypothesis on the coefficients of the pertinent 3.2 | Mitotic grade inter-observer agreement GEE model was performed. The null hypothesis of the test was that both sensitivity and specificity are equal between the two methods. For each reading (first and second), the mitotic grade was evaluated In cases of rejection of the null hypothesis ( p < .05), further results using 2 different cut-offs: the old cut-off, based on the human WHO were reported, that is, 95% confidence intervals for the differences system and as previously published in the veterinary literature, and of sensitivity and specificity. These confidence intervals were the new cut-off suggested by the guidelines. Although differences obtained by the non-parametric bootstrap method, with 3000 were not statistically significant (with overlaps between 95% confi- bootstrap samples. dence intervals), all inter-observer agreements for mitotic grade were 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 689 (0.172, 0.432) (0.393, 0.705) Anaplasia 0.302 0.549 92.7% 91.5% Inter observer agreement among surgical pathologists using the human grading as reported in veterinary literature (first reading) and using the guidelines (second reading). (0.258, 0.420) (0.359, 0.614) Pattern loss 0.339 0.486 71.4% 86.6% (0.271, 0.473) (0.068, 0.214) cellularity Increased 0.372 0.141 77.0% 63.2% F I G U R E 2 Inter-observer mitotic grade agreement among surgical pathologists, following the human grading system alone (first reading) and following the guidelines (second reading), when (0.143, 0.374) (0.265, 0.452) 2 different mitotic cut-offs were applied. The cut-off proposed in the Small cells guidelines (new) increased the inter-observer agreement, independent 0.258 0.359 of whether or not the guidelines were followed. Nevertheless, there 79.0% 78.0% was a slightly decreased agreement among pathologists when the guidelines were followed (second reading), compared to that obtained using the human system alone (first reading), independent of which (0.307, 0.479) (0.017, 0.184) Macronucleoli cut-off was used. 0.393 0.101 72.8% 97.1% Note: %: percentage of “concordant” evaluations by two pathologists; kappa = agreement index (Fleiss Kappa). greater and passed from moderate to substantial using the new cut- off, compared with the old one, at both the first and second readings (0.457, 0.619) (0.540, 0.701) (Figure 2 and Table 5). Spontaneous Counting mitoses following the guidelines (second reading), necrosis 0.538 slightly lowered the agreement for mitotic grade calculated with both 77.7% 81.5% 0.62 cut-offs, when compared with the first reading. Nevertheless, the overall agreement for histologic grade increased when the guidelines (0.382, 0.558) (0.584, 0.745) were used (Figure 3 and Table 5). Mitotic grade evaluated following the new cut-offs proposed in the guidelines. Invasion 0.470 0.665 76.9% 85.2% 3.3 | Macronucleoli inter-observer agreement (0.550–0.798) (0.519–0.742) At 100X magnification, the kappa index was slight, while at 200 Mitotic gradea magnification, the kappa index was fair. Despite this difference, agree- ment on the histologic grade only marginally increased when nucleoli 0.674 0.630 94.1% 94.1% were evaluated at 200, compared with 100 magnification (Figure 4 and Table 6). Histologic grade (0.340, 0.484) (0.436, 0.606) Abbreviation: C.I., confidence interval. 3.4 | Concordance with the consensus diagnosis 0.412 0.521 70.6% 77.5% The 4 veterinary neuropathologists reached a consensus diagnosis for mitotic grade in 148/158 (93.6%) cases by applying the guidelines and C.I.) C.I.) kappa kappa (95% (95% the new mitotic grade cut-offs. Among the 148 tumours, there were % % 134 (90.6%) grade 1 tumours, 11 (7.4%) grade 2 tumours and 3 (2%) TABLE 4 Reading grade 3 tumours. Since there were few grade 2 and 3 tumours, diag- 2nd nostic accuracy for mitotic grade with the guidelines could not be 1st calculated. a 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 690 BELLUCO ET AL. T A B L E 5 Inter-observer agreement Readings Cut-offs Agreement Mitotic grade Histologic grade among surgical pathologists for mitotic 1st reading Old % 81.5% 71.8% and histologic grades, evaluated with the Kappa 0.507 0.458 human grading system (first reading), or 95% C.I. 0.407, 0.606 0.379, 0.537 with the grading system implemented with the guidelines (second reading), 1st reading New % 94.1% 70.6% applying the two different two cut-offs: Kappa 0.674 0.412 the old cut-off, as published in the 95% C.I. 0.550, 0.798 0.340, 0.484 previous veterinary literature, and the new cut-off, as proposed in guidelines. 2nd reading Old % 78.2% 71.9% Kappa 0.443 0.485 95% C.I. 0.350, 0.535 0.406, 0.564 2nd reading New % 94.1% 77.5% Kappa 0.630 0.521 95% C.I. 0.519, 0.742 0.436, 0.606 Note: %: percentage of “concordant” evaluations by two surgical pathologists. Abbreviations: C.I., confidence interval; kappa, Fleiss Kappa index. F I G U R E 4 Inter-observer agreement among surgical pathologists, for histologic grade and macronucleoli, when macronucleoli were evaluated at 100 and 200 magnification. At 200 magnification, the inter-observer agreement for macronucleoli was increased. F I G U R E 3 Inter-observer histologic grade agreement among However, regardless of which magnification was used, the inter- surgical pathologists, following the human grading system alone (first observer agreement for histologic grade did not change. reading) and following the guidelines (second reading), when 2 different mitotic cut-offs were applied. With the guidelines (second T A B L E 6 Inter- observer agreement among surgical pathologists reading), the inter-observer agreement was increased, independently for macronucleoli, evaluated at 100 and 200 magnification. of which cut-off was used. Magnification Agreement Macronucleoli Histologic grade The 4 veterinary neuropathologists reached a consensus diagnosis 100 % 97.1% 77.5% for histologic grade in 151/158 (95.6%) cases by applying the guidelines Kappa 0.101 0.521 and the new mitotic grade cut-offs. Among the 151 tumours, there were (95% C.I.) 0.017, 0.184 0.436, 0.606 78 (51.6%) grade 1 tumours, 62 (41.1%) grade 2 tumours and 11 (7.3%) 200 % 77.5% 77.3% grade 3 tumours. Kappa 0.374 0.530 The estimates of sensitivities and specificities obtained by the (95% C.I.) 0.281, 0.467 0.446, 0.614 GEE method are reported in Table 7. We found a statistically signif- Note: % = percentage of concordant evaluations by two surgical icant difference between the old grading system and the new pathologists. guidelines for grade 1 ( p = 0.014) and grade 2 ( p = 0.016) tumours. Abbreviation: C.I., confidence interval; kappa, agreement index (Fleiss For grade 1 tumours, the sensitivity was increased at 95.6% Kappa). (+ 3.9%, CI: 0.9%–7.1%) when using the guidelines and this differ- ence was statistically significant. The specificity was decreased For grade 2 tumours, the sensitivity was decreased at 42.6% ( 3.6%, CI 9.1% to 2.1%), but this difference was not statistically ( 4.2%, CI: 10.6% to 2.6%), but the difference was not statisti- significant. cally significant. The specificity was increased at 96.2% (+3.9%, 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 691 TABLE 7 Diagnostic accuracy at the first reading (human system alone) and second reading (applying the new guidelines) for the histologic grade. 2nd reading 2nd reading vs. 1st reading Difference of diagnostic Grade Sensitivity Specificity accuracy: χ2, df, p-value Sensitivity Specificity 1 95.6% (93.0%, 97.3%) 59.2% (50.7%, 67.2%) 8.5 2, 0.014 +3.9% (0.9%, 7.1%) 3.6% ( 9.1%, 2.1%) 2 42.6% (34.7%, 50.8%) 96.2% (93.5%, 97.8%) 8.8, 2, 0.012 4.2% ( 10.6%, 2.6%) +3.9% (1.1%, 6.8%) 3 90.9% (70.9%, 97.6%) 95.1% (92.7%, 96.8%) 2.65, 2, 0.270 +10.9% (0.0%, 20.0%) +0.6% ( 1.3%, 2.3%) F I G U R E 5 Invasion. (A) Presence of invasion. Complete agreement among surgical pathologists. Tumoral cells are not lined by normal meningeal cells and infiltrate the brain parenchyma. (B) Partial agreement among surgical pathologists. Clear meningeal delimitation of the tumour is masked by cell degeneration and multilayer fibrous reaction. (C) Absence of invasion. Complete agreement among surgical pathologists. The meningioma is well defined and almost detached by normal parenchyma. No tumour infiltration into the nervous system is observed. HE stain. 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 692 BELLUCO ET AL. F I G U R E 6 Anaplasia. (A). Presence of anaplasia. Complete agreement among surgical pathologists. In an abundant collagenous stroma, there are small cells with condensed nucleus, consistent with lymphocytes or degenerating tumoral cells, and polygonal viable tumoral cells with clear cytoplasm and finely dispersed chromatin. Tumoral cells are arranged in nests but no typical meningioma pattern is present, making meningioma diagnosis difficult. (B) Partial agreement among surgical pathologists. The tumour presents anaplastic areas, characterized by cells with severe anisocytosis and anisokaryosis (on the left), along with areas with more differentiated cells arranged in a meningothelial pattern (on the right). (C) Absence of anaplasia. Complete agreement among surgical pathologists. Tumour cells are well differentiated and arranged in whorls and bundles, typical of transitional subtype. Neither anisocytosis, nor anisokaryosis are present. HE stain. CI: 1.1%–6.8%) and this difference was statistically significant. 4 | DI SCU SSION For grade 3 tumours, sensitivity was increased at 90.9% (+10.9%, 0% to 20%) and specificity was increased at 95.1% (+0.6%, CI: In canine patients, the 1-year survival rate following surgical resection of 1.3% to 2.3%), but these differences were not statistically intracranial meningiomas is approximately 50%.2,3 Canine meningiomas, significant. as well as human meningiomas, exhibit a wide range of histologic patterns Overall, the diagnostic accuracy of the grading systems was and clinical behaviour. In human patients, histopathologic findings, and, in statistically greater when performed following the new guidelines particular tumour grade, play a key role in prognostication and therapeutic for grade 1 and 2 tumours, and it seemed to be greater for grade decision-making.10 However, it is currently unknown if and how the 3 tumours (although not statistically significant). human tumour grade could be used in dogs to predict clinical outcome 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 693 F I G U R E 7 Spontaneous necrosis. (A) Presence of necrosis. Complete agreement among surgical pathologists. Large coalescent areas of necrosis, characterized by nuclear loss but preserved cellular outline (coagulative necrosis) separate small areas of viable tumour cells. (B). Partial agreement among surgical pathologists. Small areas of cellular degeneration and necrosis, characterized by loss of cellular details and presence of amorphous eosinophilic material, are scattered throughout the sample. Since the sample is composed of small fragments, interpretation of spontaneous versus artefactual necrosis was not consensual. (C) Absence of necrosis. Complete agreement among surgical pathologists. No necrotic areas are present. HE stain. and response to treatment. Standardization of histologic grading is a pre- 4.1 | Inter-observer agreement requisite for reliable and clinically relevant prognostic studies. Previous data showed that the human histologic grading system has suboptimal Histologic grade. For the histologic grade, application of the guidelines reproducibility when applied to dogs,22 far below that reported in human resulted in an increased inter-observer agreement shown by the medicine.12 New guidelines have been recently proposed to increase the kappa index, which shifted from 0.41 to 0.52. The improved agree- reproducibility of the human grading system, but these guidelines have ment is comparable to that obtained in veterinary medicine with other not been validated.22 Therefore, the goal of this study was to assess the grading systems and tumours (for gliomas κ = 0.519; for canine soft reproducibility and diagnostic accuracy of these guidelines. tissue sarcoma κ = 0.4320). Nevertheless, this result is below the 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 694 BELLUCO ET AL. F I G U R E 8 Small cells. (A) Presence of small cells. Complete agreement among surgical pathologists. The tumour is composed of small cells with increased nucleus-cytoplasmic ratio. Cytoplasm is scant and nucleus is small, often elongated and irregular. (B). Partial agreement among surgical pathologists. Throughout the tumour there are occasional clusters of small cells characterized by increased nucleus- cytoplasmic ratio. The nucleus is small and elongated. These areas were interpreted by some pathologists as artefactual or degenerative, or in a too low number to be representative of the tumour. (C) Absence of small cells. Complete agreement among surgical pathologists. The tumoral cells are moderate to large in size. Two psammoma bodies are also evident on the right (arrows). HE stain. agreement obtained for human meningioma (κ = 0.71–0.84, depend- those from neurologic cases, to their preferred diagnostic laboratory. ing on tumour grade).12 The lower agreement obtained in the dog Thus, surgical pathologists working in these facilities regularly receive might be because the agreement for human meningioma was calcu- nervous system samples. If their diagnostic reproducibility in defining lated among neuropathologists, while for our study it was calculated the tumour type is good,21 they would certainly benefit from clear among surgical pathologists. In veterinary medicine, there are few and precise guidelines to define tumour grade. specialized neuropathology laboratories, and veterinarians, especially Histologic grade of meningiomas is defined by eight criteria, sepa- if not board-certified in neurology, usually send all samples, including rated into main criteria and soft criteria. The three main criteria are 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 695 F I G U R E 9 Pattern loss. (A) Presence of pattern loss. Complete agreement among surgical pathologists. Tumoral cells are arranged in sheets, with any specific cellular pattern. No streams, bundles, or whorls are present. The tumour subtype given for this tumour ranged from rhabdoid, chordoid, meningothelial and anaplastic, suggesting the lack of a clear pattern. (B). Partial agreement among surgical pathologists. Tumoral cells are arranged in sheets, but they lack atypia and other soft criteria of malignancy. At second reading, all surgical pathologists agreed about the meningothelial subtype. The lack of a specific organization in the meningothelial subtype could be interpreted as pattern loss. (C) Absence of pattern loss. Complete agreement among surgical pathologists. The neoplastic cells are clearly arranged in whorls and fascicles, typical of a transitional subtype. HE stain. mitotic grade, invasion and anaplasia. Their presence alone is suffi- while anaplasia reached a moderate agreement with an increase of cient to increase the grade from 1 to 2 or 3. The five soft criteria are more than 0.2 points in the kappa index. necrosis, macronucleoli, pattern loss/sheeting, small cells and hyper- Mitotic count and grade. Mitotic count is often correlated with cellularity. To diagnose a grade 2 tumour in the absence of one of the prognosis and for this reason it is integrated into most grading sys- main criteria, the presence of at least three soft criteria is tems.20,28–32 In human medicine, mitotic grade is the most important 5,6,10 required. criterion for a grade 2 tumour diagnosis and it is directly related to Main criteria. When using the guidelines, two of the three main survival.33–35 Despite the wide use of mitotic grade, disagreements criteria, mitotic grade and invasion, showed a substantial agreement, are frequently reported in veterinary medicine.36,37 In our study, 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 696 BELLUCO ET AL. F I G U R E 1 0 Increased cellularity. (A). Presence of increased cellularity. Complete agreement among surgical pathologists. Most of the tumour presents a high concentration of cells per unit area. Interstitial connective tissue is scant and cells are small, with almost overlapping nuclei. (B) Partial agreement among surgical pathologists. Interstitial connective tissue is scant, but cells are medium in size. (C) Absence of increased cellularity. Complete agreement among surgical pathologists. Microcystic subtype is characterized by large cells containing cytoplasmic vacuoles. The large cell size causes reduced number of cells per unit area. HE stain. inter-observer agreement was substantial, regardless of whether the already applied the recommendations to standardize mitotic count guidelines were applied, similar to that reported in human medicine during the first reading.37 (κ = 0.51–0.67). 12 Recommendations have been recently published to In human grading, mitotic cut-offs were reported to be 4 and standardize mitotic count evaluation in veterinary medicine,36,37 and 20 mitoses in 1.60 mm.238,39 Nevertheless, in most papers, and in the they were integrated into the guidelines for canine meningioma grad- application of human grading to dog meningioma, the 1.60 mm2 sur- 22 ing. In our study, the lack of a substantial improvement between the face area was replaced by the more commonly used 10 high power first and second reading may be because the surgical pathologists fields (HPFs, old cut-off),12,33–35,40–42 creating reduced reproducibility 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 697 in human and veterinary medicine.22,43 In order to solve this problem, called “sheeting” in the human grading system) is recognized as a sub- in the most recent human WHO edition, HPF is clearly replaced by jective criterion even in human medicine, causing confusion with the 10 surface area. In the guidelines for canine meningioma grading, the confluent syncytial patterns of the meningothelial subtype when human cut-offs calculated for 1.60 mm2 were adapted to the stan- observed at higher magnification.33 Small cells are difficult to identify, 2 22,36,37 dardized veterinary area of 2.37 mm (new cut-off). It is inter- since the identification of variably sized clusters of smaller tumour esting to note that the new cut-off applied to canine meningiomas is cells is qualitative and often subjective. Therefore, in a prognostic more reproducible than the old cut-off. This is due to the distribution grading system for canine meningioma, these two criteria should be of the mitotic count in the 158 samples: 85% of the tumours had less deleted, interpreted with caution or evaluated differently, to increase than 8 mitoses and almost 30% of the tumours had a mitotic count their reproducibility. around 4 in 2.37 mm2. Thus, until otherwise proven by prognostic Macronucleoli. Reproducibility was low when assessing macronu- studies, the new mitotic cut-off should be applied to canine cleoli, both with and without the guidelines. In human medicine, a meningioma. macronucleolus is defined as visible at 100 magnification. In our Invasion (Figure 5). In human meningioma, invasion is directly cor- study, at 100 magnification, corrected for screen size,22 very few to no related with tumour relapse and poor prognosis.11 Invasion limits the nucleoli were visible. Moreover, the kappa index at 100 was slight, benefits of excisional surgery, since the tumour cannot be completely while at 200 it was fair, supporting the comment of some authors that, resected and thus will likely regrow. In the human grading system if really needed, nucleoli should be evaluated at 200. Nevertheless, applied to canine meningiomas, invasion was not clearly defined. despite the low reproducibility, macronucleoli have a very low impact on There was confusion about dura mater/bone invasion and brain inva- histologic grade assessment, regardless of the magnification. Therefore, sion, the latter considered with or without well-defined pial delimita- we suggest omitting macronucleoli from any future grading system cor- tion of tumour cells. Therefore, the guidelines for the assessment of related with prognosis for canine meningioma. canine meningioma define invasion as when the pial layer does not Hypercellularity (Figure 10). We found hypercellularity to be line tumour cells apparently infiltrating the nervous tissue, as reported another criterion with only fair to slight reproducibility with both in the original human grading.39 Use of this definition increased the the human grading and when applying the guidelines. This criterion reproducibility of invasion assessment. Nevertheless, when using should be interpreted in relation to the mean cellularity for each the guidelines, the agreement and the kappa index are still below that specific subtype of meningioma, which remains subjective. If con- reported in human medicine, where the kappa index is approximately sidered for inclusion in a new grading system for canine meningi- 12 0.76. This discrepancy could be explained by the fact that immuno- oma, then increased cellularity should be evaluated quantitatively, histochemical markers, which are regularly used to assess invasion in not qualitatively. human medicine, are currently lacking in veterinary medicine. Anaplasia (Figure 6). The reproducibility of anaplasia assessment was also increased by applying the guidelines for canine meningiomas 4.2 | Diagnostic accuracy and became comparable to that obtained in human medicine.12 Since anaplasia is a sufficient criterion to upgrade tumours to grade 3, its When applying the guidelines, grade 3 tumours were identified almost detection is pivotal to identify a poorer prognosis. perfectly, with a sensitivity of 91% and a specificity of 95%. The Soft criteria. When the guidelines were applied, agreement increase in sensitivity and specificity when using the guidelines was increased for three soft criteria (spontaneous necrosis, small cells and not statistically significant, but this could be due to the low number of pattern loss). Spontaneous necrosis reached a substantial agreement. grade 3 tumours in the study. Although meningioma grades have not In contrast, when applying the guidelines, agreement decreased for been fully validated and accurately linked to prognosis in dogs, several two criteria, macronucleoli and hypercellularity. publications have associated anaplastic or grade 3 tumours with Spontaneous necrosis (Figure 7). Evaluation of necrosis in meningi- shorter survival times (less than 4 months post-surgery).2,44,45 These oma can be challenging, since it can vary from single cell necrosis to data highlight the importance of an accurate diagnosis. vast areas of karyolytic and karyorrhectic debris. Moreover, iatrogenic The guidelines also allowed better detection of grade 1 tumours, tissue damage due to sample handling may be present, which can with significantly increased sensitivity, and grade 2 tumours, with signifi- mimic intratumoral necrosis. Therefore, surgical pathologists with little cantly increased specificity. However, using the guidelines, the grade experience in neuropathology could over-interpret this criterion. The 1 specificity and grade 2 sensitivity were slightly decreased, remaining guidelines allowed an increase in inter-observer agreement, which between 40% and 60%, supporting the idea that many grade 2 meningio- became a substantial agreement, comparable with that reported in mas can be misdiagnosed as grade 1 tumours. This is a well-known prob- human medicine for meningioma (κ = 0.66) and higher than that lem for three-tier grading systems,46 supporting the proposal of two-tier reported in canine soft tissue sarcomas (κ = 0.46). 12,20 grading systems for some animal tumours.32,47 Small cells and pattern loss (Figures 8 and 9). Reproducibility was In conclusion, the guidelines allowed an overall increased agree- increased when small cells and pattern loss were evaluated with the ment among surgical pathologists as well as better detection of grade guidelines, although the agreement remained fair. These results are 1 and grade 3 tumours, thus they should be integrated into the cur- comparable to those obtained in human medicine.12 Pattern loss (also rent grading system for canine meningioma. Nevertheless, a low 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 698 BELLUCO ET AL. reproducibility remains for some minor criteria and these criteria 7. Sturges BK, Dickinson PJ, Bollen AW, et al. Magnetic resonance imag- appear to have little impact in defining histologic grade. Furthermore, ing and histological classification of intracranial meningiomas in 112 dogs. J Vet Intern Med. 2008;22(3):586-595. doi:10.1111/j.1939- in human medicine, all the soft/minor criteria seem to not be signifi- 1676.2008.00042.x cantly correlated with tumour relapse or patient survival.34,35 There- 8. Miller AD, Miller CR, Rossmeisl JH. Canine primary intracranial can- fore, in the future, a new grading system for canine meningiomas cer: a clinicopathologic and comparative review of glioma, meningi- should be developed, either avoiding or quantitatively evaluating the oma, and choroid plexus tumors. Front Oncol. 2019;9:1151. doi:10. 3389/fonc.2019.01151 less reproducible histological parameters, and correlating the grade to 9. Arie Perry DJB. Neuropathology patterns and Introduction. Practical tumour prognosis. Surgical Neuropathology a Diagnostic Approach. Vol 1. 2nd ed. Elsevier; 2018. ACKNOWLEDGEMEN TS 10. International Agency for Research on Cancer. WHO Classification of Tumours. Central Nervous System Tumours. Vol 6. 5th ed. Lyon, France: The authors thank all the technicians and pathologists involved in International Agency for Research and Cancer; 2021. their daily work and Dr. Leah Cannon for editing support. Without 11. Banan R, Abbetmeier-Basse M, Hong B, et al. The prognostic signifi- their excellent work, studies like this one would never be completed. cance of clinicopathological features in meningiomas: microscopic brain invasion can predict patient outcome in otherwise benign meningiomas. Neuropathol Appl Neurobiol. 2021;47:724-735. doi:10. FUND ING INFORMATION 1111/nan.12700 The authors disclose the following financial support for the research, 12. Rogers CL, Perry A, Pugh S, et al. Pathology concordance levels for authorship, and/or publication of this article: Impulsion Santé Globale meningioma classification and grading in NRG oncology RTOG trial 2019 grant from VetAgro Sup, Lyon. 0539. Neuro-Oncol. 2016;18(4):565-574. 13. Williams JF, Zhao M, Najdawi F, et al. Grading of medullary thyroid carcinoma: an interobserver reproducibility study. Endocr Pathol. CONF LICT OF IN TE RE ST ST AT E MENT 2022;33(3):371-377. doi:10.1007/s12022-022-09718-0 The authors declared no potential conflicts of interest with respect to 14. Taylor C, Puzyrenko A, Iczkowski KA. Trends in disagreement with the research, authorship, and/or publication of this article. outside genitourinary pathology diagnoses at an academic center. Pathol Res Pract. 2022;236:153997. doi:10.1016/j.prp.2022.153997 15. Van Bockstal MR, Berlière M, Duhoux FP, Galant C. Interobserver DATA AVAI LAB ILITY S TATEMENT variability in ductal carcinoma In situ of the breast. Am J Clin Pathol. The data that support the findings of this study are available from the 2020;154(5):596-609. doi:10.1093/ajcp/aqaa077 corresponding author upon reasonable request. 16. Prayson RA, Agamanolis DP, Cohen ML, et al. Interobserver repro- ducibility among neuropathologists and surgical pathologists in fibril- ORCID lary astrocytoma grading. J Neurol Sci. 2000;175(1):33-39. doi:10. 1016/S0022-510X(00)00274-4 Sara Belluco https://orcid.org/0000-0002-7209-6879 17. van den Bent MJ. Interobserver variation of the histopathological Giancarlo Avallone https://orcid.org/0000-0002-5203-7679 diagnosis in clinical trials on glioma: a clinician's perspective. Acta Fabiano José Ferreira de Sant'Ana https://orcid.org/0000-0002- Neuropathol (Berl). 2010;120(3):297-304. 3956-1329 18. Meuten DJ, Moore FM, Donovan TA, et al. International guidelines for veterinary tumor pathology: a call to action. Vet Pathol. 2021; 58(5):766-794. doi:10.1177/03009858211013712 RE FE R ENC E S 19. Krane GA, Shockley KR, Malarkey DE, et al. Inter-pathologist agree- 1. Song RB, Vite CH, Bradley CW, Cross JR. Postmortem evaluation of ment on diagnosis, classification and grading of canine glioma. Vet 435 cases of intracranial neoplasia in dogs and relationship of neo- Comp Oncol. 2022;20:881-889. doi:10.1111/vco.12853 plasm with breed, age, and body weight. J Vet Intern Med. 2013;27(5): 20. Yap FW, Rasotto R, Priestnall SL, Parsons KJ, Stewart J. Intra- and 1143-1152. doi:10.1111/jvim.12136 inter-observer agreement in histological assessment of canine soft 2. Forward AK, Volk HA, Cherubini GB, et al. Clinical presentation, diag- tissue sarcoma. Vet Comp Oncol. 2017;15(4):1553-1557. doi:10. nostic findings and outcome of dogs undergoing surgical resection for 1111/vco.12300 intracranial meningioma: 101 dogs. BMC Vet Res. 2022;18(1):88. doi: 21. Belluco S, Avallone G, Di Palma S, Rasotto R, Oevermann A. Inter- 10.1186/s12917-022-03182-y and Intraobserver agreement of canine and feline nervous system 3. Lacassagne K, Hearon K, Berg J, et al. Canine spinal meningiomas and tumors. Vet Pathol. 2019;56:349. doi:10.1177/0300985818824952 nerve sheath tumours in 34 dogs (2008-2016): distribution and long- 22. Belluco S, Marano G, Baiker K, et al. Standardization of canine menin- term outcome based upon histopathology and treatment modality. gioma grading: inter-observer agreement and recommendations for Vet Comp Oncol. 2018;16(3):344-351. doi:10.1111/vco.12385 reproducible histopathologic criteria. Vet Comp Oncol. 2022;20:509. 4. Koestner A. Histological Classification of Tumors of the Nervous System doi:10.1111/vco.12802 of Domestic Animals. Washington DC: Armed Forces Institute of 23. Landis JR, Koch GG. The measurement of observer agreement for Pathology, in cooperation with the American Registry of Pathology categorical data. Biometrics. 1977;33(1):159-174. doi:10.2307/ and The World Health Organization Collaborating Center for World- 2529310 wide reference on Comparative Pathology; 1999. 24. Genders TSS, Spronk S, Stijnen T, Steyerberg EW, Lesaffre E, 5. Higgins RJ, Bollen AW, Dickinsons PJ, Siso-Llonch S. Tumors of the Hunink MGM. Methods for calculating sensitivity and specificity of nervous system. In: Meuten DJ, ed. Tumors in Domestic Animals. Vol 1. clustered data: a tutorial. Radiology. 2012;265(3):910-916. doi:10. John Wiley & Sons, Inc.; 2017:834-891. 1148/radiol.12120509 6. Vandevelde M, Higgins RJ, Oevermann A. Neoplasia. Veterinary Neu- 25. R Foundation for Statistical, Computing, Vienna, Austria. R: the R Pro- ropathology: Essentials of Theory and Practice. John Wiley & Sons, Inc.; ject for Statistical Computing. Accessed June 10, 2021. https://www. 2012. r-project.org/ 14765829, 2023, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vco.12932 by Readcube (Labtiva Inc.), Wiley Online Library on [27/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License BELLUCO ET AL. 699 26. Kilem L. Gwet. irrCAC: Computing Chance-Corrected Agreement Coeffi- 39. Louis DN, Ohgaki H, Wiestler OD, et al. In: Jaffe ES, Lakhani SR, cients (CAC). Accessed June 10, 2021. https://CRAN.R-project.org/ Ohgaky H, eds. WHO Classification of Tumours of the Central Nervous package=irrCAC System. Vol 1. International Agency for Research on Cancer (IARC); 27. Højsgaard S, Halekoh U, Yan J. The R package geepack for general- 2016. ized estimating equations. J Stat Softw. 2006;15:1-11. doi:10.18637/ 40. Goldbrunner R, Minniti G, Preusser M, et al. EANO guidelines for the jss.v015.i02 diagnosis and treatment of meningiomas. Lancet Oncol. 2016;17(9): 28. Avallone G, Rasotto R, Chambers JK, et al. Review of histological e383-e391. doi:10.1016/S1470-2045(16)30321-7 grading Systems in Veterinary Medicine. Vet Pathol. 2021;58(5):809- 41. Saygin I, Cakir E, Ercin ME. Interobserver variability in mitotic count 828. doi:10.1177/0300985821999831 for meningioma grading: how can we reduce it? Turk Neurosurg. 2020; 29. Dagher E, Abadie J, Loussouarn D, Campone M, Nguyen F. Feline inva- 30(5):643-650. doi:10.5137/1019-5149.JTN.26252-19.2 sive mammary carcinomas: prognostic value of histological grading. Vet 42. Salles D, Santino SF, Malinverni ACM, Stávale JN. Meningiomas: a Pathol. 2019;56(5):660-670. doi:10.1177/0300985819846870 review of general, histopathological, clinical and molecular character- 30. Goldschmidt M, Peña L, Rasotto R, Zappulli V. Classification and grad- istics. Pathol Res Pract. 2021;223:153476. doi:10.1016/j.prp.2021. ing of canine mammary tumors. Vet Pathol. 2011;48(1):117-131. doi: 153476 10.1177/0300985810393258 43. Barresi V, Caffo M, Tuccari G. Classification of human meningiomas: 31. Spangler WL, Kass PH. The histologic and epidemiologic bases for lights, shadows, and future perspectives. J Neurosci Res. 2016;94(12): prognostic considerations in canine melanocytic neoplasia. Vet Pathol. 1604-1612. doi:10.1002/jnr.23801 2006;43(2):136-149. doi:10.1354/vp.43-2-136 44. Greco JJ, Aiken SA, Berg JM, Monette S, Bergman PJ. Evaluation of 32. Kiupel M, Webster JD, Bailey KL, et al. Proposal of a 2-tier histologic intracranial meningioma resection with a surgical aspirator in dogs: grading system for canine cutaneous mast cell tumors to more accu- 17 cases (1996–2004). J Am Vet Med Assoc. 2006;229(3):394-400. rately predict biological behavior. Vet Pathol. 2011;48(1):147-155. doi:10.2460/javma.229.3.394 doi:10.1177/0300985810386469 45. Petersen SA, Sturges BK, Dickinson PJ, et al. Canine intraspinal 33. Backer-Grøndahl T, Moen BH, Torp SH. The histopathological spec- meningiomas: imaging features, histopathologic classification, and trum of human meningiomas. Int J Clin Exp Pathol. 2012;5(3): long-term outcome in 34 dogs. J Vet Intern Med. 2008;22(4):946-953. 231-242. doi:10.1111/j.1939-1676.2008.0106.x 34. Karabagli P, Karabagli H, Mavi Z, Demir F, Ozkeles EY. Histopatholog- 46. Dal Col P, Garaix T, Massard A, et al. Meningioma sampling: how much ical and clinical features as prognostic factors of atypical meningio- is enough for the accurate grading of atypical meningiomas? Pathology mas. Turk Neurosurg. 2020;30(5):575-746. doi:10.5137/1019-5149. (Phila). 2021;20:602-607. doi:10.1016/j.pathol.2020.10.024 JTN.31161-20.1 47. Koehler JW, Miller AD, Miller CR, et al. A revised diagnostic classifica- 35. Loewenstern J, Shuman W, Rutland JW, et al. Preoperative and histo- tion of canine glioma: towards Validation of the canine glioma patient logical predictors of recurrence and survival in atypical meningioma as a naturally occurring preclinical model for human glioma. after initial gross Total resection. World Neurosurg. 2019;128:e148- J Neuropathol Exp Neurol. 2018;77(11):1039-1054. doi:10.1093/jnen/ e156. doi:10.1016/j.wneu.2019.04.069 nly085 36. Donovan TA, Moore FM, Bertram CA, et al. Mitotic figures-normal, atypical, and imposters: a guide to identification. Vet Pathol. 2021; 58(2):243-257. doi:10.1177/0300985820980049 37. Meuten DJ, Moore FM, George JW. Mitotic count and the field of How to cite this article: Belluco S, Marano G, Lurier T, et al. view area: time to standardize. Vet Pathol. 2016;53(1):7-9. doi:10. Standardization of canine meningioma grading: Validation of 1177/0300985815593349 new guidelines for reproducible histopathologic criteria. Vet 38. Perry A. 13 – Meningiomas. In: Perry A, Brat DJ, eds. Practical Surgical Neuropathology: A Diagnostic Approach. 2nd ed. Elsevier; 2018:259- Comp Oncol. 2023;21(4):685‐699. doi:10.1111/vco.12932 298. doi:10.1016/B978-0-323-44941-0.00013-8