Mitosin and phh3 predict poorer survival in astrocytomas WHO grades II and III

JCP Online First, published on July 17, 2015 as 10.1136/jclinpath-2015-202983 Original article Mitosin and phh3 predict poorer survival in astrocytomas WHO grades II and III R K Varughese, 1 T Lind-Landström, 1 A H Habberstad, 1 Ø Salvesen, 2 C S Haug, 1 S Sundstrøm, 3 S H Torp 1,4 Additional material is published online. To view please visit the journal (http:// dx.doi.org/10.1136/jclinpath- 2015-202983) 1 Faculty of Medicine, Department of Laboratory Medicine, Children s and Women s Health, Norwegian University of Science and Technology (NTNU), Trondheim, Norway 2 Faculty of Medicine, Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology (NTNU), Trondheim, Norway 3 Department of Oncology, St. Olav s Hospital, Trondheim, Norway 4 Department of Pathology and Medical Genetics, St Olav s Hospital, Trondheim, Norway Correspondence to R K Varughese, Department of Pathology and Medical Genetics, c/o Dr. Sverre H Torp, St. Olav s Hospital, Trondheim NO 7006, Norway; rosilinv@gmail.com Received 28 February 2015 Revised 24 June 2015 Accepted 28 June 2015 To cite: Varughese RK, Lind-Landström T, Habberstad AH, et al. J Clin Pathol Published Online First: [ please include Day Month Year] doi:10.1136/ jclinpath-2015-202983 ABSTRACT Aims: The limitations of the current WHO classification of astrocytomas call for a sustained effort to improve diagnostic and prognostic accuracy. The relationship between tumour growth and clinical outcome suggests that proliferative activity should be examined. The objective of this study was to evaluate the diagnostic and prognostic value of the proliferation markers mitosin and phosphohistone H3 ( phh3) in infiltrative astrocytomas WHO grades II and III and compare the findings with mitotic count and Ki-67/MiB-1 immunostaining. Methods: Fifty-nine and thirty-three infiltrative astrocytomas WHO grades II and III, respectively, were immunostained with the proliferation markers mitosin and phh3 using standard immunohistochemical procedures. The expression was quantified as a proliferative index (PI) and statistically evaluated with Spearman s rank correlation test, Wilcoxon-Mann- Whitney U test, and univariable and multivariable COX regression survival analyses. Results: Significant positive correlations were found between these proliferation markers. The number of mitoses, phh3 mitotic figures (MFs), the Ki-67/MiB-1 PI and the mitosin PI were greater in WHOgrade III anaplastic astrocytomas compared to WHO grade II diffuse astrocytomas, while phh3 PI only showed a trend. All proliferation markers were associated with poorer prognosis, but mitotic count was not. Ki-67/MiB- 1, mitosin and phh3 MF achieved statistical significance in the univariable analyses of both time to relapse (TTR) and overall survival (OS). Only mitosin remained significant in both multivariable analyses. phh3 was significant in the multivariable analysis of OS but not of TTR. Clinical factors including age, extent of surgical resection and WHO performance status were also significantly correlated with survival. Conclusions: In conclusion, mitosin and phh3 immunostaining have prognostic and diagnostic value in the clinical assessment of patients with infiltrative astrocytomas. The inclusion of proliferation markers in a layered diagnosis should be considered in the upcoming revision of the WHO classification system. INTRODUCTION Infiltrative astrocytomas WHO grades II IV are the most common primary brain tumour in humans, accounting for about 80% of reported cases in adults. 1 Proper classification is important for optimal prognostic information, treatment options and follow-up. 1 2 The established WHO classification of these tumours is based on histopathological features, but distinguishing between grades is still difficult. The current scheme is not optimal and, as it relies on the subjective assessment of histological findings, there is considerable interobserver variability. 3 4 Thus, new biomarkers are being sought to better distinguish between grades, particularly between WHO grades II and III, as well as to improve prognosis prediction for patients with infiltrative astrocytomas. Proliferation plays an important role in tumourigenesis and prognostication. 5 6 Consequently, mitotic counts are a central part of histopathological malignancy grading. However, it can be challenging to detect mitotic figures (MFs) in routine H&E stained sections due to small and squeezed biopsies, and to distinguish them from pycnotic or apoptotic tumour cells. 5 For these reasons proliferation markers have attracted much interest and become important in the histopathological diagnosis of several human malignancies. Ki-67/MiB-1 is one such marker whose proliferative indices (PIs) in astrocytomas significantly correlate with both increasing tumour grade and poorer outcome. 78 However, due to the considerable overlap of PIs between astrocytomas of grades II IV, Ki-67/MiB-1 has not been incorporated into the WHO classification scheme. 1 4 7 Consequently, there is a need for other proliferation markers, and mitosin and phosphohistone H3 (phh3) are promising candidates. 9 12 It is likely that standard HES sections as well as immunohistochemistry will play an important role in layered diagnosis, as has been suggested for the upcoming revision of the WHO classification system. 13 Mitosin, also known as centromere protein-f, is a nuclear phosphoprotein that can be used to measure proliferation due to its expression in all phases of the cell cycle except G0- and early G1-phase. 10 11 The marker has been demonstrated to correlate well with Ki-67/MiB-1, but only a few studies have focussed on astrocytomas in particular. 14 16 Histone H3 is one of the core histone proteins and a major component of chromatin. 12 17 As the phosphorylated state of histone H3, phh3, is only found during mitosis and not interphase or apoptosis, studies suggest that it may facilitate the 12 17 19 process of counting mitoses. Knowledge of mitosin and phh3 in infiltrative astrocytomas is generally lacking, so the aim of this study was to assess the diagnostic and prognostic value of these two proliferation markers in a series of well-described infiltrative astrocytomas WHO grade II and III with thorough follow-up data. Further, we wanted to evaluate whether these markers have any advantage over mitotic activity and Ki-67/MIB-1 immunostaining and to Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983 1 Copyright Article author (or their employer) 2015. Produced by BMJ Publishing Group Ltd under licence.

Original article Downloaded from http://jcp.bmj.com/ on July 23, 2015 - Published by group.bmj.com determine whether they might help improve and refine the current WHO classification. The study is based on robust welldescribed homogeneous patient material with thorough follow-up. MATERIAL AND METHODS This study was approved by the Regional Committee for Medical and Health Research Ethics Central Norway. Waiver of consent was given by the Regional Ethics Committee because patients had either died or were severely disabled. This study is an extension of earlier investigations on grade II and III astrocytomas where mitotic counts and Ki-67/MiB-1 PI immunostaining were described. 14 20 Inclusion criteria for patients in the study were: age >18 years, from regional hospitals in central Norway, and operated on for supratentorial astrocytomas WHO grade II or III. All included patients were operated on and diagnosed at St. Olav s Hospital, Trondheim, Norway and enrolled following a search of the electronic patient files in the pathology department. All patients with grade II astrocytomas operated on between 1987 and 2007 were identified, resulting in 109 cases. 21 In addition, 31 patients with WHO grade III anaplastic astrocytomas were operated on between 1998 and 2006. 14 For the purpose of this study, an extended search was also conducted for anaplastic astrocytomas from 1987 1997 and 2007. This provided an additional 13 cases, for a total of 44 cases of anaplastic astrocytomas. A thorough revision of the collected cases resulted in the exclusion of some due to insufficient material, failure to meet the inclusion criteria, revised diagnosis, or because they proved to be recurrent tumours. This process resulted in the retention of 59 cases of WHO grade II astrocytomas and 33 cases of WHO grade III anaplastic astrocytomas. Due to lack of material, only 57 (96.6%) and 50 (84.7%) of the 59 cases of diffuse astrocytomas WHO grade II were available for mitosin and phh3 immunostaining, respectively. Of the 33 cases of anaplastic astrocytomas WHO grade III, 32 (97.0%) and 31 (93.9%) were eligible for mitosin and phh3 immunostaining, respectively. Clinical data on gender, performance score, tumour location, treatment and follow-up/survival until 31 December 2013 were obtained from the electronic medical records of St. Olav s Hospital, Trondheim or of local hospitals. All histological sections of these tumours were revised and classified by a senior neuropathologist (SHT) according to the most recent 2007 WHO classification scheme for tumours of the central nervous system. 2 Mitoses were counted in 10 high power fields (HPFs) ( 40 objective) in areas with the highest proliferative activity and expressed as the number of mitoses in 10 HPFs. Only distinct MFs were included. The immunohistochemical analyses have been described previously. 14 20 22 Briefly, immunostaining was performed on 4 - mm-thick paraffin sections. After endogenous peroxidase activity was eliminated and antigen retrieval was carried out using a pressure cooker, the sections were incubated at room temperature with primary antibodies (table 1) using an automated immunohistostainer (Dako Techmate 500). Human tonsils served as positive controls, and the primary antibody was omitted in the negative controls. Microscopic analyses were conducted using a Nikon Eclipse 80i. Sections were scanned and mitosin and phh3 immunostaining was scored in hot-spot areas with the highest labelling intensity. 23 Only distinct nuclear staining was considered positive for mitosin. For sections stained with phh3, only positive cell nuclei undergoing mitosis were counted as positive. PIs for mitosin and phh3 were determined as the percentage of positive nuclei in at least a 1000 tumour cell nuclei or in three HPFs in areas with the highest activity using an ocular grid (0.058 mm 2 ). phh3 MFs were determined in the same manner as mitotic counts. 12 19 The nuclei of non-tumour cells, such as vascular endothelial cells and haematopoietic cells, were not counted. phh3 and mitosin immunostaining in grade II astrocytomas was quantified by CSH and RKV, respectively. The measurements of CSH and RKV were coordinated with those of SHT. κ statistics were applied to 20 randomly selected cases. Statistical analyses were carried out using SPSS V.20.0. Correlations between variables were analysed by Spearman s rank correlation test. The closer Spearman s r is to 1, the stronger the correlation between the variables. A Kaplan Meier curve was used to find the estimated median survival time for both time to relapse (TTR) and overall survival (OS). The nonparametric Wilcoxon Mann Whitney U test was conducted to compare the expression of the markers in WHO grade II versus WHO grade III astrocytomas. The Cox proportional hazard regression model was used to analyse the effect of clinical and immunohistochemical variables on both TTR and OS. The assumption of proportional hazard was verified for all variables using Schoenfeld s residuals test. TTR was defined as the time from surgery to either relapse or death related to the same cancer; death by any other cause was censored. 24 OS was defined as the number of months from surgery to death, irrespective of cause. 24 Only variables which showed significance in univariable analysis were used in multivariable analysis. To avoid problems of multicollinearity, the significant histological variables were considered in separate multivariable analyses. p Values <0.05 were considered statistically significant. RESULTS Clinical characteristics are shown in table 2. Patients operated on for grade II astrocytomas (n=59) and grade III astrocytomas (n=33) were a median age of 40 (range: 19 74 years) and 51 (range: 27 78 years), respectively. The median age at surgery for all 94 patients was 46 (range: 19 78 years) with a male/ female ratio of 1.6. For most patients (82.6%), the WHO performance status at diagnosis was good ( 1). All patients underwent surgery; subtotal resection was performed in 55.4%, while only radical resection was possible in 21.7%. Regarding mitoses, the few grade II astrocytomas with observable MFs occurred in cases with large tumour specimens Table 1 Antibodies and immunohistochemical procedures Antibody Source Type Clone Dilution Incubation time (min) Ki-67/MiB-1 Dako, Glostrup, Denmark Monoclonal Ki-67 antigen 1:100 40 phh3 Upstate Biotechnology, Millipore, Billerica, USA Polyclonal H3 (Ser10) 1:2000 60 Mitosin Novus Biologicals, Cambridge, UK Monoclonal 14C10/1D8 1:500 40 2 Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983

Original article Table 2 Variable Clinical data Parameter Diffuse astrocytomas (n=59) Anaplastic astrocytomas (n=33) Total (n=92) Frequency Per cent Frequency Per cent Frequency Per cent Age at surgery (years) 18 39 28 47.5 10 30.3 38 41.3 40 60 22 37.3 15 45.5 37 40.2 >60 9 15.3 8 15.3 17 18.5 Gender Male 34 57.6 23 69.7 57 62.0 Female 25 42.4 10 30.3 35 38.0 WHO performance status (preoperative) 0 46 78.0 3 9.1 49 53.3 1 11 18.6 11 33.3 22 23.9 2 0 0.0 9 27.3 9 9.8 >2 2 3.4 4 12.1 6 6.5 Data missing 0 0.0 6 18.2 6 6.5 Tumour location Frontal lobe 12 20.3 6 18.2 18 19.6 Temporal lobe 10 16.9 7 21.2 17 18.5 Parietal lobe 4 6.8 2 6.1 6 6.5 Deep structures 2 3.4 2 6.1 4 4.3 Combination 31 52.5 16 48.5 47 51.1 Treatment Surgery only 17 28.8 20 60.6 27 29.3 Surgery+adjuvant therapy 42 71.2 10 30.3 62 67.4 Data missing 0 0.0 3 9.1 3 3.3 Extent of surgical resection Biopsy 11 18.6 9 27.3 20 21.7 Subtotal 33 55.9 18 54.5 51 55.4 Radical 15 25.4 5 15.2 20 21.7 Data missing 0 0.0 1 3.0 1 1.1 Relapse treatment Yes 37 62.7 10 30.3 47 51.1 No 9 15.3 5 15.2 14 15.2 No relapse 13 30.3 10 30.3 23 25.0 Data missing 0 0.0 8 24.2 8 8.7 available. Otherwise, they had a benign histology and were clearly a grade II astrocytoma. The difference in mitotic counts between grade II and III astrocytomas was statistically significant ( p=0.001). In several tumour samples mitotic counting was difficult or impossible due to small and squeezed biopsies. There was one case of grade II diffuse astrocytoma with an extremely high phh3 PI. As no mitoses were observed on the regular H&E section of this case and no increased proliferative activity was indicated by either mitosin or Ki-67/MiB-1, this case was excluded from the statistical analysis of phh3. Mitosin revealed distinct immunoreactive tumour cell nuclei with little background staining (figure 1A, B). The phh3 antibody made it easy to identify MFs in neoplastic astrocytes (figure 1C, D). The results of the correlation analyses are presented in table 3. All proliferation markers were significantly mutually correlated Figure 1 (A) Mitosin immunostaining in WHO grade II diffuse astrocytomas; original magnification 400. (B) Mitosin immunostaining in WHO grade III anaplastic astrocytomas; original magnification 400. (C) phh3 immunostaining in WHO grade II diffuse astrocytomas; original magnification 400. (D) phh3 immunostaining in WHO grade III anaplastic astrocytomas; original magnification 400. Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983 3

Original article Downloaded from http://jcp.bmj.com/ on July 23, 2015 - Published by group.bmj.com Table 3 Correlation between proliferation markers Proliferation marker Correlation to p Value Spearman s r Mitosin PI Mitosis count <0.001 0.435 Ki-67/MiB-1 <0.001 0.766 phh3 PI <0.001 0.467 phh3 mitotic figures <0.001 0.657 phh3 PI Mitosis count 0.140 0.173 Ki-67/MiB-1 <0.001 0.397 phh3 mitotic figures <0.001 0.648 phh3 MFs Mitosis count <0.001 0.399 Ki-67/MiB-1 <0.001 0.615 Ki-67/MiB-1 PI Mitosis count 0.001 0.350 except phh3 PI and mitosis counts. The strongest correlation was between mitosin and Ki-67/MiB-1 (Spearman s r=0.766). The median and range of expression for mitosin and phh3, as well as the comparison of expression between malignancy grades, are shown in table 4. Error bars representing the 95% CIs for the medians are shown in figures 2 6. More MFs were identified using phh3 immunostaining than by regular mitotic count. In general there was trend for greater expression of all markers in grade III anaplastic astrocytomas. The Wilcoxon Mann Whitney U test revealed that mitotic count, Ki-67/MiB-1 PI, mitosin PI and phh3 MFs were significantly increased in grade III astrocytomas compared to grade II astrocytomas. Although phh3 PI tended towards significance ( p=0.056), it did not attain statistical significance. The estimated median TTR for grade II diffuse astrocytomas and grade III anaplastic astrocytomas was 48.0 months (95% CI 22.3 to 73.6) and 9.9 months (95% CI 7.5 to 12.3), respectively (figure 7). The estimated median OS for grade II diffuse astrocytomas and grade III anaplastic astrocytomas was 96.6 months (95% CI 72.5 to 120.6) and 17.6 months (95% CI 4.2 to 31.1), respectively (figure 8). The results from the univariable analysis of TTR and OS are summarised in tables 5 and 6. In the univariable analyses of TTR, the following proved to be significant predictors: age, extent of surgical resection, Ki-67/MiB-1 PI, mitosin PI and phh3 MFs. The same factors were significantly associated with poorer survival in the univariable analysis of OS. Only extent of surgical resection and mitosin PI remained significant in the multivariable analyses of TTR with Ki-67/MiB-1 (table 7A), mitosin (table 7B) and phh3 MFs (table 7C). In the multivariable analyses of OS with Ki-67/MiB-1 (table 8A), mitosin (table 8B) and phh3 MFs (table 8C), extent of surgical resection, age, mitosin PI and phh3 MFs remained Table 4 Proliferation marker Median expression and comparison of grades Median value of expression (range) Grade II Grade III p Value Mitotic count 0.0 (0.0 3.0) 1.0 (0.0 15.0) 0.001 Ki-67/MiB-1 PI 4.5 (0.7 18.6) 9.9 (1.0 39.1) <0.001 Mitosin PI 1.7 (0.0 4.4) 3.4 (0.0 14.8) <0.001 phh3 PI 0.8 (0.0 3.8) 1.0 (0.0 2.7) 0.056 phh3 MFs 3.0 (0.0 17.0) 7.0 (0.0 41.0) 0.001 Figure 2 Error bars showing the 95% CI for the median mitotic count (Wilcoxon Mann Whitney U test, p=0.001). significant, and proved to be independent predictors of poorer survival. The κ statistics for mitosin PI, phh3 PI and phh3 MFs were 0.61 ( p=0.003), 0.60 ( p=0.007) and 0.34 ( p=0.085), respectively. DISCUSSION The current WHO classification system for astrocytomas is not optimal and improvements are constantly sought. Consequently, immunohistochemical determination of proliferative activity is interesting because proliferation is closely linked to increased malignancy. The purpose of this study was to determine if the proliferation markers mitosin and phh3 had any prognostic and diagnostic benefits over mitotic counts and Ki-67/MiB-1 immunostaining in infiltrative astrocytomas WHO grade II and grade III. In general, mitoses, Ki-67/MiB-1, phh3 and mitosin were mutually correlated and exhibited a trend to or significantly higher indices in anaplastic astrocytomas compared with grade II tumours. Mitosin was a significant predictor of worse outcome in the univariable analyses of TTR and remained an independent predictor of survival in the multivariable analyses. Both mitosin and phh3 were significantly associated with poorer OS in both univariable and multivariable analyses. Ki-67/MiB1 was significant in the univariable analyses but not the multivariable analyses of either OS or TTR. Mitotic counts did not attain statistical significance in the survival analyses. Mitosis Mitotic activity was positively associated with Ki-67/MiB-1 PI, mitosin PI and phh3 MFs but not with phh3 PI. When the number of mitoses were compared between grade II and III 4 Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983

Original article Figure 3 Error bars showing the 95% CI for the median Ki-67/MiB-1 proliferative index (PI) (Wilcoxon Mann Whitney U test, p<0.001). astrocytomas, our results showed that the number of mitoses was greater in the latter, and that there was a statistically significant difference, as is known from previous literature. 625 Figure 5 Error bars showing the 95% CI for the median phh3 proliferative index (PI) (Wilcoxon Mann Whitney U test, p=0.056). Concerning mitosis, a solitary mitosis did not allow a diagnosis of grade III astrocytoma, particularly if the biopsy was large and the histology benign. 2 5 6 25 In this regard, our grade II Figure 4 Error bars showing the 95% CI for the median mitosin proliferative index (PI) (Wilcoxon Mann Whitney U test, p<0.001). Figure 6 Error bars showing the 95% CI for the median phh3 mitotic figures (MFs) (Wilcoxon Mann Whitney U test, p=0.001). Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983 5

Original article Downloaded from http://jcp.bmj.com/ on July 23, 2015 - Published by group.bmj.com Table 5 Univariable analyses of time to relapse Variable HR (95% CI) p Value Clinicopathological variables Age 1.03 (1.01 to 1.05) 0.001 Extent of surgical resection 0.48 (0.33 to 0.69) <0.001 WHO performance status 1.46 (1.11 to 1.93) 0.007 Immunohistochemical variables Mitotic count 1.00 (0.89 to 1.13) 0.985 Ki-67/MiB-1 1.08 (1.04 to 1.12) <0.001 Mitosin PI 1.15 (1.06 to 1.24) 0.001 phh3 PI 1.13 (0.79 to 1.62) 0.500 phh3 MFs 1.03 (1.00 to 1.07) 0.036 Figure 7 Kaplan Meier curve of time to relapse (log-rank test, p=0.095). Cum, cumulative. tumours with a few mitoses were found in large tumour samples. Some of the biopsies of anaplastic astrocytomas were small and squeezed, which made identification of MFs somewhat challenging but any finding all the more relevant. 14 In some cases, evaluation of mitotic activity was deemed too difficult but the morphology was consistent with anaplastic astrocytomas (according to the WHO 2007 classification). In these cases we feel that immunohistochemistry was useful to assess proliferative activity. In the survival analyses, mitotic activity was not significant in the univariable analysis of either OS or TTR, and was therefore not included in the multivariable analysis. Our findings can be related to a study by Giannini et al 6 which found that there was no difference in survival between grade II astrocytomas and grade III astrocytomas that exhibited a solitary mitosis. Out of the 92 tumours in this study, 84 were available for the evaluation of mitotic activity. Of these, 60.7% had no MFs, and 29.7% had 2 MFs. Mitotic count is affected by numerous factors, such as tissue preparation, section thickness and biopsy size. 5 These factors may have influenced the identification of MFs and thus the survival analyses. On the other hand, the lack of statistical significance of mitoses in the survival analyses demonstrates the need for better prognostic markers. Ki-67/MiB-1 Ki-67/MiB-1 was significantly correlated with all investigated proliferation markers. It was also weakly correlated with mitosis counts (Spearman s r=0.350). This is in agreement with previous studies. 61415 In addition, we found a significant difference in the expression of Ki-67/MiB-1 between grade II and III astrocytomas, and as one can see in figure 3, there is no overlap in the error bars showing the 95% CIs for the median. This parallels the findings of other studies on astrocytomas. 615 In the survival analyses, Ki-67/MiB-1 was shown to be a reliable predictor of poor prognosis in the univariable analyses of both OS and TTR. However, it did not remain significant in the Table 6 Univariable analyses of overall survival Variable HR (95% CI) p Value Figure 8 Kaplan Meier curve of overall survival (log-rank test, p=0.006). Cum, cumulative. Clinicopathological variables Age 1.04 (1.02 to 1.06) <0.001 Extent of surgical resection 0.34 (0.23 to 0.51) <0.001 WHO performance status 1.81 (1.36 to 2.40) <0.001 Immunohistochemical variables Mitotic count 1.04 (0.92 to 1.18) 0.535 Ki-67/MiB-1 1.10 (1.06 to 1.14) <0.001 Mitosin PI 1.20 (1.10 to 1.30) <0.001 phh3 PI 1.30 (0.90 to 1.88) 0.169 phh3 MFs 1.06 (1.02 to 1.09) 0.001 6 Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983

Original article Table 7 Multivariable analyses of time to relapse Variable HR (95% CI) p Value (A) Significant clinical variables and Ki-67/MiB PI Age 1.01 (0.99 to 1.03) 0.604 Extent of surgical resection 0.55 (0.36 to 0.85) 0.007 WHO performance status 1.17 (0.82 to 1.68) 0.377 Ki-67/MiB-1 1.05 (1.00 to 1.10) 0.061 (B) Significant clinical variables and mitosin PI Age 1.01 (0.99 to 1.03) 0.281 Extent of surgical resection 0.57 (0.37 to 0.88) 0.011 WHO performance status 1.17 (0.86 to 1.59) 0.328 Mitosin PI 1.14 (1.02 to 1.27) 0.019 (C) Significant clinical variables and phh3 MFs Age 1.00 (0.98 to 1.02) 0.963 Extent of surgical resection 0.46 (0.28 to 0.76) 0.002 WHO performance status 1.10 (0.80 to 1.15) 0.545 phh3 MFs 1.03 (0.99 to 1.06) 0.163 multivariable analyses of either. These results agree with our previous findings as well as those of other studies, but the literature can be conflicting, so further investigation of the utility of Ki-67/MiB-1 immunostaining in infiltrative astrocytomas is 6 8 2226 needed. Inclusion of Ki-67/MiB-1 in the WHO classification system would likely help pathologists conducting diagnostic and prognostic procedures. However, one of many reasons this has not yet been done, is the variability between laboratories and the overlap of indices between malignancy grades. 7 Therefore this marker should be used with caution and together with histopathological features associated with malignancy. Mitosin Mitosin is maximally expressed during the G2/M-phase of the cell cycle, and while this nuclear protein not fully understood, its role is gradually being elucidated. Auto-antibodies against mitosin have been found in some cancer patients but are not increased in normal tissues or autoimmune diseases. 11 It is likely Table 8 Multivariable analyses of overall survival Variable HR (95% CI) p Value (A) Significant clinical variables and Ki-67/MiB PI Age 1.02 (1.00 to 1.04) 0.079 Extent of surgical resection 0.37 (0.23 to 0.58) <0.001 WHO performance status 1.57 (1.08 to 2.28) 0.019 Ki-67/MiB-1 1.04 (0.99 to 1.10) 0.115 (B) Significant clinical variables and mitosin PI Age 1.02 (1.00 to 1.04) 0.020 Extent of surgical resection 0.39 (0.24 to 0.63) <0.001 WHO performance status 1.34 (1.00 to 1.80) 0.051 Mitosin PI 1.19 (1.06 to 1.33) 0.002 (C) Significant clinical variables and phh3 MFs Age 1.02 (1.00 to 1.04) 0.083 Extent of surgical resection 0.32 (0.19 to 0.54) <0.001 WHO performance status 1.22 (0.89 to 1.66) 0.223 phh3 MFs 1.06 (1.02 to 1.10) 0.003 that overexpression of mitosin indicates increased proliferation and is not a trigger for malignant progression, although it is possible that the increased levels of mitosin may affect the early stages of tumourigenesis. 11 However, this does not detract from the value of mitosin as a proliferation marker but rather increases its potential as a prognostic factor. 11 Mitosin was positively associated with both mitosis counts and Ki-67/MiB-1; Spearman s r suggests that it is more strongly correlated with the latter (0.435 and 0.766, respectively). 15 16 Similar results have been found in other reports. Interestingly, our previous study on grade III anaplastic astrocytomas was unable to find a correlation between mitosin and mitotic counts, although there was a positive correlation between mitosin and Ki-67/MiB-1. 14 This is likely related to the fact that there was a larger sample size in this study, and that many of the anaplastic astrocytomas biopsies were small and squeezed. Our results suggest that there is a significant difference in mitosin expression between WHO grade II and grade III astrocytomas. Figure 4 demonstrates that there is no overlap between the error bars showing the 95% CIs for the medians of both grades, and the Wilcoxon Mann Whitney U test had a p value of <0.001. This contrasts with the results of the study by Korkolopoulou et al, which demonstrated that mitosin was unable to discriminate between grade II and III astrocytomas. 15 However, the authors noted that this might be due to their relatively small sample size (n=14). A study on meningiomas was also unable to find a difference in mitosin expression levels between malignancy grades. 16 Regarding survival, mitosin PI was the only marker in this study to remain significant in the univariable and multivariable analyses of both OS and TTR. Similarly, Korkolopoulou et al 15 found mitosin PI to be a significant predictor of OS but not of disease-free survival. phh3 phh3 reaches its maximum expression during mitosis and is negligible during the other phases. 17 27 As the phosphorylation of the histone H3 is a mitosis-specific event and does not occur during apoptosis, phh3 may be particularly useful as a mitotic marker. 919 There were more phh3 immunoreactive MFs than mitoses in routine sections. This indicates that phh3 is more sensitive in detecting MFs that might otherwise be overlooked. Similar observations have been made in other studies. 12 17 19 28 Thus, our results support the view that phh3 can be used as a mitosis-specific marker. Our study was unable to correlate phh3 PI with the number of mitoses. Furthermore, we found no difference in this index between tumour grades. This is clearly illustrated in figure 5, where the error bars representing the 95% CIs for the medians of both grades overlap (Wilcoxon Mann Whitney U test, p=0.056). Additionally, phh3 PI had no significant prognostic value in the survival analyses. However, these results differ from the findings of Colman et al 12 who reported that phh3 PI did have prognostic value in WHO grade II and III infiltrative astrocytomas. Our previous study on WHO grade III anaplastic astrocytomas found higher phh3 PIs were associated with poorer survival in the univariable analysis of OS, but not in the multivariable Cox analysis. 14 Since phh3 is a mitosis-specific marker, most studies focus on mitotic count rather than PI. 9 17 19 28 phh3 MFs, where the number of cells stained by phh3 was counted in 10 HPFs, Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983 7

Original article Downloaded from http://jcp.bmj.com/ on July 23, 2015 - Published by group.bmj.com correlated well with both mitotic count, Ki-67/MiB-1 and mitosin PI in accordance with other studies. 9141719 We found that there was a statistically significant difference in the number of phh3 MFs between the two malignancy grades ( p=0.001), with more in grade III anaplastic astrocytomas. Liu et al 9 also found a significant difference in phh3 expression between grade II and III infiltrative astrocytomas. Although phh3 MFs was significant in the univariable analyses of TTR, it did not remain significant in the multivariable analysis. On the other hand, it was significant in both the univariable and multivariable analyses of OS. The literature is conflicting and scarce concerning phh3 MFs and survival in astrocytomas. In meningiomas, phh3 MFs was found to be significant in both univariable and multivariable analyses of recurrence-free survival. 28 Clinical factors Our data confirm the robust prognostic value of lower age, greater extent of surgical resection and WHO performance status as positive predictors of survival, in full agreement with the literature. 29 Limitations The limitations of our study are its retrospective approach and the moderate size of the series, so the survival analyses should be assessed with caution. We had more diffuse astrocytomas WHO grade II (n=59) than anaplastic astrocytomas WHO grade III (n=33). The reasons for this are not entirely clear, but do demonstrate a weakness of retrospective studies. Another flaw may be that immunostaining was affected by pre-analytical conditions and antigen preservation. As mentioned in the Material and methods section, small and squeezed biopsies were excluded, which may have given rise to a selection bias. The material available for evaluation depends on several factors, including the surgeon, the surgical techniques used, the location of the tumour and lack of tissue due to earlier experiments. These factors are thought to be distributed at random, but there is always a risk of unforeseen bias. Nevertheless, one of the strengths of our study is that it is population based and all tumours were removed at the same hospital according to the same surgical policies. Further, as our follow-up data are thorough and based on official public statistics, the prognostic and diagnostic values of Ki-67/MiB-1, mitosin and phh3 immunostaining are promising. κ-statistics were performed on 20 randomly selected cases. To better manage interobserver variation, this should perhaps have been conducted on all cases; however, as the senior pathologist (SHT) continuously supervised all co-authors, this source of error has been greatly reduced. Take home messages Proliferation markers can be useful for the diagnosis and prognosis of astrocytomas. Mitosin has potential as a proliferation marker in grade II and III astrocytomas. phh3 can be used as a mitosis-specific marker. Mitosin, phh3 and Ki-67/MiB-1 have been shown in this study to have diagnostic and prognostic value. Conclusion In the present study we have shown that the proliferation markers phh3 and mitosin are useful in the prognostic and diagnostic work-up of infiltrative astrocytomas WHO grade II and III. Mitosin seems to have a slight advantage over Ki-67/ MiB-1 and this should be tested in a larger series of astrocytomas. However, none of the markers should be used alone but rather in combination with established criteria for malignancy. We believe that proliferation markers, such as mitosin, phh3 and Ki-67/MiB-1, have a role to play in layered diagnosis, and should be considered in the upcoming revision of the WHO classification system. 13 Nevertheless, further research is required to identify specific markers that could identify patients with aggressive astrocytomas; perhaps a panel of proliferation markers could also be developed. Handling editor Cheok Soon Lee Acknowledgements We are very grateful to Professors Ivar Skjåk Nordrum and Christina Vogt for their critical reading of the manuscript. We would also like to thank the Norwegian Cause of Death Registry for their cooperation. Contributors RKV: immunohistochemical analyses, statistical analyses and writing of the manuscript; TL-L, AHH: collection of data and immunohistochemical analyses; ØS: statistical analyses; CSH: immunohistochemical analyses; SS: assessment of clinical data and critical reading of the manuscript; SHT: study design, collection of data and writing of the manuscript. Competing interests None declared. Ethics approval The Regional Committee for Medical and Health Research Ethics Central Norway approved this study. Provenance and peer review Not commissioned; externally peer reviewed. REFERENCES 1 Louis DN, Ohgaki H, Wiestler OD, et al. WHO classification of tumours of the central nervous system. 4th edn. Lyon: International Agency for Research on Cancer, 2007. 2 Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 2007;114:97 109. 3 van den Bent MJ. Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician s perspective. Acta Neuropathol 2010;120:297 304. 4 Kros JM. Grading of gliomas: the road from eminence to evidence. J. Neuropathol Exp Neurol 2011;70:101 9. 5 Prayson RA. Cell proliferation and tumors of the central nervous system. Part 1: evaluation of mitotic activity. J Neuropathol Exp Neurol 2002;61:501 9. 6 Giannini C, Scheithauer BW, Burger PC, et al. Cellular proliferation in pilocytic and diffuse astrocytomas. J Neuropathol Exp Neurol 1999;58:46 53. 7 Johannessen AL, Torp SH. The clinical value of Ki-67/MIB-1 labeling index in human astrocytomas. Pathol Oncol Res 2006;12:143 7. 8 Hsu DW, Louis DN, Efird JT, et al. Use of MIB-1 (Ki-67) immunoreactivity in differentiating grade II and grade III gliomas. J Neuropathol Exp Neurol 1997;56:857 65. 9 Liu X, Chen N, Wang X, et al. Apoptosis and proliferation markers in diffusely infiltrating astrocytomas: profiling of 17 molecules. J Neuropathol Exp Neurol 2006;65:905 13. 10 Ma L, Zhao X, Zhu X. Mitosin/CENP-F in mitosis, transcriptional control, and differentiation. J Biomed Sci 2006;13:205 13. 11 Varis A, Salmela AL, Kallio MJ. Cenp-F (mitosin) is more than a mitotic marker. Chromosoma 2006;115:288 95. 12 Colman H, Giannini C, Huang L, et al. Assessment and prognostic significance of mitotic index using the mitosis marker phospho-histone H3 in low and intermediate-grade infiltrating astrocytomas. Am J Surg Pathol 2006;30:657 64. 13 Louis DN, Perry A, Burger P, et al. International Society of Neuropathology Haarlem consensus guidelines for nervous system tumor classification and grading. Brain Pathol 2014;24:429 35. 14 Habberstad AH, Gulati S, Torp SH. Evaluation of the proliferation markers Ki-67/ MIB-1, mitosin, survivin, phh3, and DNA topoisomerase IIalpha in human anaplastic astrocytomas an immunohistochemical study. Diagn Pathol 2011;6:43. 15 Korkolopoulou P, Patsouris E, Konstantinidou AE, et al. Mitosin and DNA topoisomerase IIalpha: two novel proliferation markers in the prognostication of diffuse astrocytoma patient survival. Appl Immunohistochem Mol Morphol 2001;9:207 14. 8 Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983

Original article 16 Konstantinidou AE, Korkolopoulou P, Kavantzas N, et al. Mitosin, a novel marker of cell proliferation and early recurrence in intracranial meningiomas. Histol Histopathol 2003;18:67 74. 17 Fukushima S, Terasaki M, Sakata K, et al. Sensitivity and usefulness of anti-phosphohistone-h3 antibody immunostaining for counting mitotic figures in meningioma cases. Brain Tumor Pathol 2009;26:51 7. 18 Tapia C, Kutzner H, Mentzel T, et al. Two mitosis-specific antibodies, MPM-2 and phospho-histone H3 (Ser28), allow rapid and precise determination of mitotic activity. Am J Surg Pathol 2006;30:83 9. 19 Ribalta T, McCutcheon IE, Aldape KD, et al. The mitosis-specific antibody anti-phosphohistone-h3 (PHH3) facilitates rapid reliable grading of meningiomas according to WHO 2000 criteria. Am J Surg Pathol 2004;28:1532 6. 20 Lind-Landstrom T, Habberstad AH, Sundstrom S, et al. Prognostic value of histological features in diffuse astrocytomas WHO grade II. Int J Clin Exp Pathol 2012;5:152 8. 21 Lind-Landstrom T, Habberstad AH, Torp SH. Proliferative activity and histopathological features in diffuse grade II astrocytomas. APMIS 2012;120: 640 7. 22 Torp SH, Alsaker M. Ki-67 immunoreactivity, basic fibroblastic growth factor (bfgf) expression, and microvessel density as supplementary prognostic tools in low-grade astrocytomas. An immunohistochemical study with special reference to the reliability of different Ki-67 antibodies. Pathol Res Pract 2002;198:261 5. 23 Prayson RA. Cell proliferation and tumors of the central nervous system, part II: radiolabeling, cytometric, and immunohistochemical techniques. J Neuropathol Exp Neurol 2002;61:663 72. 24 Punt CJ, Buyse M, Kohne CH, et al. Endpoints in adjuvant treatment trials: a systematic review of the literature in colon cancer and proposed definitions for future trials. J Natl Cancer Inst 2007;99:998 1003. 25 Coons SW, Pearl DK. Mitosis identification in diffuse gliomas: implications for tumor grading. Cancer 1998;82:1550 5. 26 Hilton DA, Love S, Barber R, et al. Accumulation of p53 and Ki-67 expression do not predict survival in patients with fibrillary astrocytomas or the response of these tumors to radiotherapy. Neurosurgery 1998;42:724 9. 27 Margraf LR, Gargan L, Butt Y, et al. Proliferative and metabolic markers in incompletely excised pediatric pilocytic astrocytomas--an assessment of 3 new variables in predicting clinical outcome. Neuro Oncology 2011;13:767 74. 28 Kim YJ, Ketter R, Steudel WI, et al. Prognostic significance of the mitotic index using the mitosis marker anti-phosphohistone H3 in meningiomas. Am J Clin Pathol 2007;128:118 25. 29 Stieber V. Low-grade gliomas. Curr Treat Options Oncol 2001;2:495 506. Varughese RK, et al. J Clin Pathol 2015;0:1 9. doi:10.1136/jclinpath-2015-202983 9

Mitosin and phh3 predict poorer survival in astrocytomas WHO grades II and III R K Varughese, T Lind-Landström, A H Habberstad, Ø Salvesen, C S Haug, S Sundstrøm and S H Torp J Clin Pathol published online July 17, 2015 Updated information and services can be found at: http://jcp.bmj.com/content/early/2015/07/17/jclinpath-2015-202983 Supplementary Material References Email alerting service These include: Supplementary material can be found at: http://jcp.bmj.com/content/suppl/2015/07/10/jclinpath-2015-202983.d C1.html This article cites 28 articles, 2 of which you can access for free at: http://jcp.bmj.com/content/early/2015/07/17/jclinpath-2015-202983 #BIBL Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article. Notes To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions To order reprints go to: http://journals.bmj.com/cgi/reprintform To subscribe to BMJ go to: http://group.bmj.com/subscribe/