Meningiomas: Objective assessment of proliferative indices by immunohistochemistry and automated counting method
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.217934
Source of Support: None, Conflict of Interest: None
Background: The most reliable histological correlate of recurrence risk in meningiomas is increased mitotic activity. Proliferative index with Ki-67 immunostaining is a helpful adjunct to manual counting. However, both show considerable inter-observer variability. A new immunohistochemical method for counting mitotic figures, using antibody against the phosphohistone H3 (PHH3) protein was introduced. Similarly, a computer based automated counting for Ki-67 labelling index (LI) is available.
Keywords: Immuno ratio, Ki-67 labeling index, meningioma, phosphohistone H3, proliferative indices
The World Health Organization (WHO) classification of meningiomas is primarily based on morphological parameters. The most reliable histological correlate of recurrence risk is increased mitotic activity of four or more mitoses per 10 high-power fields (HPFs). Proliferative index measurement with Ki-67 immunostaining is a very helpful adjunct that is widely used in routine practice. The combination of mitotic index and Ki-67 labeling index (LI) is a useful method for predicting the short-term recurrence of meningiomas., However, both the mitotic count and the Ki-67 LI show considerable interobserver variability. Mitotic count is defined as the number of mitotic figures per 10 consecutive HPFs in the area of highest mitotic activity. However, the mitotic count is subject to variations due to the selection of areas, sample size variations, and interobserver subjective errors of mistaking apoptotic nuclei, karyorrhectic debris, pyknosis, crush, and distortion as mitotic figures. Manual counting of 1000 nuclei for Ki-67 LI is tedious and limited by sample size and heterogeneity of proliferative activity. Although significant correlation between the tumor grade and the Ki-67 LI was found in meningiomas, the cutoff values between the grades were varied, necessitating methods to increase objectivity.
A new immunohistochemical method for counting mitotic figures and differentiating them from apoptosis, using antibody against the phosphorylated form of the amino terminus of histone H3, phosphohistone H3 (PHH3) protein, was introduced.,, Similarly, a computer-based automated counting of positively labeled nuclei for Ki-67/1000 nuclei is available (freely downloadable software from the immuno ratio [IR] website).
This study was designed to observe any significant difference between the mitotic count on light microscopy and PHH3 index using immunohistochemistry (IHC), and also between Ki-67 LI by the manual method and by utilizing IR, in meningiomas grade-wise. This was an attempt to study whether or not the use of these new techniques increases objectivity in the assessment of proliferation indices in meningiomas when compared to the existing methods.
This was a retrospective study of patients with intracranial meningiomas diagnosed at the department of pathology of our institute during the year 2013. The details of age, gender, clinical presentation, location, and recurrence, if any, were noted from medical records. The hematoxylin and eosin (H and E)-stained histological sections were reviewed by two pathologists. The parameters noted were number of mitoses per 10 consecutive HPFs (×40 objective), presence or absence of sheeting, hypercellularity, nuclear pleomorphism, macro-nucleoli, small cell formation, necrosis, and brain invasion. In discrepant cases, slides were viewed together and consensus was obtained. All tumors were graded according to the World Health Organisation (WHO) criteria 2007. The slides of IHC with Ki-67 were reviewed.
Sections of 5-micron thickness collected on poly-L-lysine coated slides were subjected to IHC by indirect immunoperoxidase technique. Briefly, the sections were dewaxed in two changes each in xylene and alcohol and stabilized in 0.1 M phosphate buffer saline solution. Antigen retrieval was performed by high-temperature unmasking using pressure cooker (125°C for 30 s at 25 lb pressure). Sections were then cooled to room temperature and immersed in 3% methanol hydrogen peroxide for blocking the endogenous peroxidase activity. Following blockage of nonspecific binding sites using 3% casein, immunohistochemical staining was performed using the BioGenex Autostainer. Sections were incubated in the primary antibody (PHH3 antibody 1:50, BioGenex, USA, and Ki-67, monoclonal antibody, 1:50, DAKO, USA) for 1 h, followed by incubation in the secondary antibody for 30 min (polymer detection kit, BioGenex, USA). Antigen-antibody complex was visualized using diaminobenzidine as chromogen. Sections were counterstained with Harris hematoxylin, dehydrated and mounted in DPX (a mixture of distyrene, a plasticizer, and xylene used as a synthetic resin mounting media).
PHH3 positive nuclei were counted in 10 consecutive high power fields (HPFs) and the mean values grade-wise were calculated. For Ki-67, at least, 1000 nuclei were counted at high magnification (×40 objective) without recounting same areas and the average was expressed as percentage. Foci of necrosis and hemorrhage were excluded. These areas were marked and photomicrographs were taken. The images were subjected to Ki-67 analysis using the freely available online IR software. IR gives a quantitative analysis by calculating the percentage of positively stained nuclear area (labeling index) using a color deconvolution algorithm for separating the staining components (diaminobenzidine and hematoxylin) and adaptive thresholding for nuclear area segmentation. It is used for the analysis of estrogen receptor, progesterone receptor, and Ki-67. The mean Ki-67 LI, both by the manual and by the IR method, was calculated grade-wise.
The data obtained was analyzed using the Statistical Package for the Social Sciences SPSS Software, [manufacturer: IBM, NY]. The mean values of mitotic count on H and E and PHH3 values were correlated. Ki-67 LI, both obtained manually (G-Ki-67) as well as using the IR software (G-IR), was compared and correlated. Ki-67 LI, both obtained manually and using the IR software, was correlated with the grade as well as the PHH3 values. Pearson's correlation and Spearman rho were calculated. A P value < 0.05 was considered statistically significant.
A total of 64 intracranial meningiomas were diagnosed during the year 2013. Evaluation on H and E, PHH3, Ki-67 LI, both manual and IR, was done in 32 cases. These were included for further analysis. There were 13 male and 19 female patients with the male:female ratio being 1:1.46. The age of the patients ranged from 10 to 75 years (mean age - 48.03 years). There were 22 Grade I and 10 Grade II meningiomas.
The mitotic count in Grade I was 0.1–3 (mean = 1.38); in Grade II, it was 2–11 (mean = 4.20). The PHH3 values were 0.1–5 in Grade I (mean = 0.74) and 0.1–6 in Grade II (mean = 1.98). The PHH3 labeling was distinct even at low-power magnification [Figure 1]. However, some cells with clear mitoses failed to stain with PHH3 [Figure 1]. There was statistically significant correlation between the grade and PHH3 values (P = 0.02); however, there was no change in the grade of the meningioma based on the PHH3 index. There was statistically significant correlation between the mitotic count in each grade with the PHH3 values (P = 0.02) [Table 1].
Ki-67 labeling index
The Ki-67 LI (manual) count ranged from 1 to 10 in Grade I (mean = 4.09) and 2–27 in Grade II (mean = 8.60). The Ki-67 LI (IR) was 0–10.5 in Grade I (mean = 3.36) and 0.7–15.3 in Grade II (mean = 6.71) tumors [Figure 2] and [Table 2]. Statistically significant correlation was seen between the grade of the tumor and the values of Ki-67 (both manual [P = 0.03] and IR [P = 0.02]) and PHH3 (P = 0.01) [Table 2]. Ki-67 LI was much easier and quicker to obtain using the computerized method compared to the manual counting method though the time for assessment of each procedure was not calculated. There was a significant correlation between Ki-67 manual counting and Ki-67 IR values (P = 0.01) [Table 2]. The mitotic count, PHH3 values, Ki-67 LI (manual), and Ki-67 IR grade-wise are given in [Table 1] and [Table 2].
Mitotic count was shown to be the most reliable marker for predicting the likelihood of recurrence in meningiomas. The anti-PHH3 antibodies used in the study specifically detect the phosphorylated core protein histone H3, making it a mitosis-specific marker that also helps to differentiate mitoses from apoptotic nuclei and nuclei in prophase. The PHH3 staining was easy to identify in low-power magnification also and was clear from the surrounding cells. The mean PHH3 values were higher in immunohistochemistry samples of Grade II compared to Grade I tumors and it was statistically significant. The mean PHH3 index correlated with mitotic count on H and E sections, and the Ki-67 LI by both manual and IR methods. These values were statistically significant. Kim et al., observed a highly significant correlation between the mitotic count assessed on H and E section and the PHH3 index and concluded that the PHH3 counting method was more sensitive in the detection of mitotic figures compared with the traditional light microscopic evaluation on H and E stained sections. Similar observations were made earlier.,, Kim et al., also suggested a threshold value of 6 per 10 HPFs associated with a significantly higher risk of recurrence or death within 5 years of primary surgery, independent of the patient's age or the extent of tumor resection. Ribalta et al., observed an increase in the tumor grade in 17% of cases using the PHH3 counting method. No threshold value could be suggested from the present study as the sample size was small and the values were not correlated with recurrence. There was no change in the tumor grade in the present study. A few cells with identifiable mitoses did not stain with PHH3 staining in the present study. This lack of staining may be related to the phosphorylation status of the histone protein. Similar observations were recorded earlier.
Cell proliferation increases as one moves across the spectrum from benign to atypical to anaplastic meningiomas, and actively proliferating meningiomas have a shorter recurrence time. A number of proliferation markers are used to identify rapidly proliferating meningioma subsets, which correlate with aggressive behavior.,,,,,,, Ki-67 is the most commonly used proliferation marker in meningiomas.,,, [Table 3] summarizes the proliferation markers used in meningiomas and their clinical application.
The Ki-67 antigen is a nonhistone nuclear protein that is expressed in the cell nucleus during all the phases of cell cycle except G0. The mean Ki-67 LI were higher in Grade II meningiomas compared to Grade I meningiomas and these were statistically significant. This is in agreement with other studies.,,,, There was significant gradewise correlation between the Ki-67 LI and mitotic count on H and E as well as PHH3 index. There was also correlation of the Ki-67 LI between the manual and automated method by IR.
The determination of the Ki-67 LI has the advantage over the assessment of the mitotic count on H and E sections in that the immunolabeled nuclei can be identified much easier than mitotic figures in the H and E stained sections. However, the different thresholds in different laboratories pose problems in determining a universal cutoff value for the Ki-67 LI. Several automated methods for counting of the Ki-67 labeled nuclei are designed.,,, IR software was used in the present study. Kim et al., used a plugin for the public domain image analysis software and did not observe any significant differences in the manual or automated mean labeling indices. They concluded that the automated method had the advantages of being faster, objective, accurate, reproducible, and easy to use.
Swiderska et al., compared the results of scoring the Ki-67 LI using optical scoring under a microscope software for Ki-67 LI quantification based on hot spots selected by pathologists manually and by automation and concluded that there was a good intra- and inter-observer agreement. The agreement between Ki-67 manual and automatic hotspot selection is also high. The authors observed that automation of the area selection is an effective tool for increasing the reliability of Ki-67 scoring in meningiomas. In this study, the area of high immunoreactivity for Ki-67 was marked manually by the pathologist and automated count was obtained on the computer. There was a statistically significant correlation between the grade of the tumor and the manual and automated Ki-67 LI in the present study.
Both the techniques (IHC with PHH3 and IR for automated counting of Ki-67 labeled nuclei) used in the study had advantage over the existing methods in terms of rapidity of analysis, ease of use, objectivity, and clarity. The techniques correlated well with the existing techniques and can be applied to routine use for predicting the likelihood of recurrence in meningiomas.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]