Anatomic Pathology / PHH3 AND MITOTIC INDEX IN MENINGIOMAS
Prognostic Significance of the Mitotic Index Using the
Mitosis Marker Anti–Phosphohistone H3 in Meningiomas
Yoo-Jin Kim, MD,1 Ralf Ketter, MD,2 Wolf-Ingo Steudel, MD,2 and Wolfgang Feiden, MD1
Key Words: Meningioma grading; Recurrence; Phosphohistone H3; PHH3; Mitosis; Ki-67
DOI: 10.1309/HXUNAG34B3CEFDU8
Abstract
Mitotic activity is one of the most reliable
prognostic factors in meningiomas. The identification of
mitotic figures (MFs) and the areas of highest mitotic
activity in H&E-stained slides is a tedious and
subjective task. Therefore, we compared the results from
immunostaining for the mitosis-specific antibody
anti–phosphohistone H3 (PHH3 mitotic index [MI])
with standard MF counts (H&E MI) and the Ki-67
labeling index (LI). The relationship between these
proliferation indices and prognosis was investigated in
a retrospective series of 265 meningiomas. The PHH3
staining method yielded greater sensitivity in the
detection of MFs and facilitated MF counting. Mitotic
thresholds of H&E MI of 4 or more per 10 high-power
fields (HPF) and PHH3 MI of 6 or more per 10 HPF
were found as the most appropriate prognostic cutoff
values for the prediction of recurrence-free survival. All
3 proliferation indices were univariately associated
with recurrences and deaths. In contrast with the Ki-67
LI, H&E MI and PHH3 MI also remained as
independent predictors in the multivariate Cox hazards
modeling (P = .0007 and P = .0004, respectively).
118
118
Am J Clin Pathol 2007;128:118-125
DOI: 10.1309/HXUNAG34B3CEFDU8
The most important prognostic question regarding
meningiomas is prediction of recurrence and, for malignant
variants, prediction of survival.1 Among negative prognostic
factors, extent of resection, brain invasion, and increased
mitotic activity have been found to be the most predictive.2,3
The mitotic index (MI), which is defined as the sum of
mitotic figures (MFs) per 10 consecutive high-power fields
(HPF) in the area of the highest mitotic activity, has been
shown to be one of the most reliable predictors of the likelihood of meningioma recurrence.3 This is considered in the
current World Health Organization (WHO) classification of
meningiomas,2 which distinguishes 3 grades with increasing
risk of local recurrence by means of histologic features of
prognostic significance. The mitotic thresholds, as suggested
by Perry and coworkers,3 were adopted by the WHO as an
objective grading criterion: benign (WHO grade I) meningiomas do not exceed 4 mitoses per 10 HPF, atypical (WHO
grade II) meningiomas exhibit an MI of 4 or more and fewer
than 20, and anaplastic (WHO grade III) meningiomas show
high mitotic activity with 20 or more mitoses per 10 HPF.
However, reliability and reproducibility of meningioma
classification based on the standard MI in H&E-stained slides
(H&E MI) are limited owing to several factors. These factors
include selection bias of HPF owing to subjective determination of the areas of highest mitotic activity and heterogeneity
of mitotic activity in different areas of the tumor. Further factors are the variation in sample size of tumor biopsy and resection samples and cellularity, both of which influence the number of evaluable cells.4 Distinguishing MFs in H&E-stained
slides from similar chromatin changes, ie, in apoptotic cells or
secondary to crush, distortion, or karyorrhectic debris, pyknosis, or necrosis, is a subjective task.
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
The identification of MFs could be facilitated by the use
of mitosis-specific staining or labeling techniques. The
immunohistochemical labeling of MFs with the mitosis-specific antibody anti–phosphohistone H3 (PHH3) has been suggested as a promising method for identifying MFs.5 AntiPHH3 antibodies specifically detect the core protein histone
H3 only when phosphorylated at serine 10 (Ser10) or serine
28 (Ser28). The phosphorylation of histone H3 is a rare event
in interphase cells but a process almost exclusively occurring
during mitosis.6-8
An antibody highly specific for the phosphorylated form
of the amino terminus of histone H3 (Ser10) was introduced
in 1997 by Hendzel et al,8 and immunohistochemical studies
have documented tight correlation between H3 phosphorylation and mitotic chromosome condensation initiating during
early prophase, whereas no phosphorylation on histone H3
was documented during apoptosis.9 These findings suggested
PHH3 as a mitosis-specific marker. The PHH3 mitosis counting method has been proved as a reliable and facile method for
mitosis counting in different types of tumors,4,10,11 including
meningiomas.12
Until now, no study has shown the correlation between
PHH3 MI and recurrence-free survival in a large cohort of
meningiomas. We evaluated the results of PHH3 staining for
determination of the PHH3 MI in a large retrospective series
of 265 meningiomas and compared the results with those for
H&E MI and the Ki-67 labeling index (LI). Each of the 3 indicators of proliferation activity was also evaluated for prognostic significance. Based on these findings, we recommend
PHH3-specific mitotic thresholds with prognostic and grading
implications.
Complete surgical extirpation of the tumor (gross total resection [GTR]) was defined as Simpson grade I or II, corresponding to macroscopically determined complete tumor resection
with bipolar coagulation of the dural insertion. Anything less
than GTR was considered subtotal resection (STR).
Materials and Methods
MF Counting on Anti-PHH3–Immunostained Slides
In each case, a serial section from the paraffin block containing the area of highest mitotic activity on H&E-stained
sections was immunostained with a rabbit monoclonal antiPHH3 antibody (Ser10, dilution 1:500; Epitomics,
Burlingame, CA). The standard procedure included
microwave antigen retrieval (10 minutes at 700 W and 30 minutes at 350 W), incubation for 1 hour in a humid chamber at
37°C with the anti-PHH3 antibody, detection of biotinylated
secondary antibodies using the streptavidin-horseradish peroxidase method with diaminobenzidine as the chromogen,
followed by hematoxylin counterstain. PHH3-labeled MFs
were counted in 10 consecutive HPF, in the same manner as in
H&E-stained sections. Because histone H3 phosphorylation
begins just before prophase,5 prophase nuclei are also PHH3labeled. These PHH3+ nuclei were not considered MFs in
mitosis counting. PHH3-labeled MFs were also enumerated at
low power (LP; objective 10×), and the number of MFs per
Cases and Clinical Data
The retrospective cohort study is based on the data for
265 patients with an initial diagnosis of meningioma, operated on from January 1996 to September 2002 at the
Department of Neurosurgery, Saarland University, Medical
School, Homburg, Germany. The mean ± SD age was 58.7 ±
13 years (range, 4-88 years). The cohort consisted of 186
females and 79 males (sex distribution, 7:3). For statistical
analyses of recurrence-free survival, only patients followed up
until recurrence or death or for at least 5 years were considered. Clinical data recorded included dates of birth and death,
sex, sites of disease, date of surgery for primary and recurrent
tumor resections, extent of resection, and date of recurrence or
regrowth as defined radiologically (computed tomography or
magnetic resonance imaging). The extent of resection was
documented along the guidelines suggested by Simpson.13
Histomorphologic Examination and MF Counting on
H&E-Stained Slides
All available slides from primary meningiomas were
reviewed by 2 independent pathologists (Y.-J.K. and W.F.),
who were unaware of the patients’ outcomes. According to the
predominant feature or growth pattern, each tumor was
assigned to a histologic pattern and grade consistent with the
current WHO criteria. Brain invasion was defined as irregular
projections of tumor or tumor cells into adjacent central nervous system parenchyma without an intervening layer of leptomeninges. The MI was obtained by summing the number of
MFs in 10 consecutive HPF in the area of highest mitotic
activity on H&E-stained slides (H&E MI). When there were
discrepancies between the investigators, the case was examined simultaneously by both observers and results recorded
after agreement was reached.
We evaluated 215 benign meningiomas (WHO grade I;
H&E MI, <4 per 10 HPF), 45 WHO grade II meningiomas
(atypical, chordoid, clear cell, or brain-invasive meningiomas;
or H&E MI, 4-19 per 10 HPF; or at least 3 of the following
criteria: hypercellularity, architectural sheeting, macronucleoli, or small cell formations; meningiomas), and 5 WHO
grade III meningiomas (anaplastic, rhabdoid or papillary
meningiomas or H&E MI, ≥20 per 10 HPF) were evaluated.
Among the rare variants, 2 chordoid types and 1 clear cell
meningioma were diagnosed. There were neither papillary nor
rhabdoid meningiomas in this series.
Am J Clin Pathol 2007;128:118-125
© American Society for Clinical Pathology
119
DOI: 10.1309/HXUNAG34B3CEFDU8
119
119
Kim et al / PHH3 AND MITOTIC INDEX IN MENINGIOMAS
visual field was recorded as the PHH3 MI LP. Labeled
objects, which were not clearly identifiable as MFs at low
magnification, were afterwards evaluated and specified at
higher magnification.
Ki-67 Immunohistochemical Analysis
Serial sections from the same paraffin blocks used for
PHH3 immunohistochemical analysis were immunostained
with a monoclonal anti–Ki-67 antibody (clone KiS5, dilution
1:50; DakoCytomation, Glostrup, Denmark) following the
standard procedure as described in the preceding sections. The
percentages of positively stained tumor cell nuclei were determined in digital images of 5 HPF of the highest labeling region
using an automated image analysis device as described by
Kim et al.14 The computed method is described and provided
free at the morphometry section of http://www.uniklinikumsaarland.de/neuropathologie.
Statistical Analyses
Correlations in PHH3 MI, H&E MI, and Ki-67 LI values
and between PHH3 MI and PHH3 MI LP were determined by
linear regression analyses. The interrater agreement in tumor
grading based on PHH3 MIs whether obtained at high or low
magnification was tested by using κ statistics. Receiver operator characteristics (ROC) were used for determination of the
appropriate cutoff levels for each proliferation index regarding
discrimination between recurrent and nonrecurrent meningiomas. The discriminatory power was tested by using the
Mann-Whitney U test. Survival times and recurrence-free survival times were computed and distributions estimated by
using Kaplan-Meier plots. The log-rank test was used for univariate survival analyses. Adjusted for patients’ age and extent
of tumor resection, Cox proportional hazards modeling techniques were used for multivariate analyses. Statistical analyses were performed using MedCalc software, version 8.1 for
Windows (MedCalc Software, Mariakerke, Belgium).
Results
Clinical Data and Follow-up
A GTR was achieved in 238 cases, whereas 27 patients
underwent only STR. The median follow-up for all patients
studied was 42 months (range, 0-120 months). Of the patients,
99 were followed up until recurrence or death or a minimum
of 5 years and were included in survival analyses.
Thirteen patients (mean ± SD age at initial diagnosis, 61
± 9.6 years; range, 51-78 years) died as a result of disease or
postoperative complications after a mean of 11.4 months following primary surgery. Initial diagnoses were WHO grade I
meningioma in 9 of these patients, WHO grade II meningioma
in 3 patients, and WHO grade III meningioma in 1 patient.
The 9 patients with WHO grade I meningiomas and a fatal
course were older than the average (median, 68.9 vs 58.7
years), and the tumors were preferentially located at the skull
base (7/9 [78%]).
The overall recurrence rate was 12.5% (n = 33), and the
relative risk (RR) of recurrence was significantly higher for
meningiomas treated by STR (10/27; RR = 37%) than for
those treated by GTR (23/238, RR = 9.7%). Recurrences
occurred after a mean of 45 months (48 months after GTR and
36 months after STR). The recurrence rate in WHO grade I
meningiomas was 6.0% in the GTR subset (11/184) and 9.8%
(21/215) in the entire cohort (GTR and STR subset).
Recurrences were recorded in 18% (8/45) of WHO grade II
meningiomas and 80% (4/5) of anaplastic meningiomas.
Brain invasions were recorded in 14 tumors (12 grade II and 2
grade III meningiomas) within 42 cases in which the central
nervous system parenchyma was assessable. There was evidence of an increase of median proliferation activity within
the subset of brain-invasive meningiomas compared with
WHO grade I meningiomas ❚Table 1❚. Four brain-invasive
tumors recurred (RR = 29%).
❚Table 1❚
H&E MI, PHH3 MI, and Ki-67 LI Results in 265 Meningiomas*
WHO grade
Benign (grade I)
Atypical (grade II)
Anaplastic (grade III)
Brain invasion
Recurrent vs nonrecurrent
Recurrent
Nonrecurrent
No. of Cases
H&E MI
PHH3 MI
Ki-67 LI
215
45
5
14
0 (0-3)
4 (0-15)
21 (20-38)
1.5 (0-38)
1 (0-14)
5 (0-31)
34 (30-50)
3 (0-50)
2 (0-13.6)
6.3 (0.6-20.6)
6 (3-34.3)
3.5 (0.8-34.3)
33
232
2 (0-31)
0 (0-38)
4 (0-41)
1 (0-50)
3.2 (0.3-25.5)
2.5 (0-34.3)
H&E MI, mitotic figure counts in 10 high-power fields in the area of highest mitotic activity, assessed in H&E-stained slides; Ki-67 LI, Ki-67 labeling index, which is the
percentage of immunolabeled nuclei determined in 5 high-power fields in the areas of highest labeling density; PHH3 MI, mitotic figure counts in phosphohistone H3 in the
same way as in H&E-stained sections; WHO, World Health Organization.
* Data are given as median (range).
120
120
Am J Clin Pathol 2007;128:118-125
DOI: 10.1309/HXUNAG34B3CEFDU8
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
PHH3 MI: Correlation With H&E MI and Ki-67 LI
Examples of PHH3 staining are shown in ❚Image 1❚.
Immunolabeled MFs are easily seen at low or medium (Image
1A) and high magnification (Images 1B-1D). The correlation
between PHH3 MI and PHH3 MI LP was highly significant (correlation coefficient, r = 0.98; regression coefficient, R2 = 0.95; P
< .00001). Corresponding results of the 2 counting methods are
plotted ❚Figure 1❚. Results of tumor grading based on these 2 MF
counting techniques revealed no differences in tumor grade; thus,
the κ statistics revealed a perfect fit of these methods (κ = 1).
Results of mitosis counting on H&E-stained slides and on
serial anti-PHH3–immunostained sections are summarized in
Table 1. The correlation between H&E MI and PHH3 MI was
also highly significant (r = 0.9; R2 = 0.77; P < .00001).
Corresponding results of H&E and PHH3 counting methods
are plotted in a scattergram (Figure 1). The linear regression
analyses indicated a lower but significant correlation between
the Ki-67 LI and the H&E MI (R2 = 0.5; P < .001) or the
PHH3 MI (R2 = 0.6; P < .001).
The PHH3 counting method was more sensitive in the
detection of MFs compared with the traditional H&E-stained
MF counts, with mean MF counts of 3.2 vs 1.6 in H&E-stained
sections and with increased median MF counts in all 3 grades
(Table 1). If the mitotic thresholds suggested by the WHO were
applied to the PHH3 MI results, in 28 cases there would be
changes in grade: 27 of WHO grade I meningiomas would have
A
B
C
D
❚Image 1❚ A, Phosphohistone H3 (PHH3)-labeled mitotic figures (arrows) are easily detectable at medium power (objective 20×).
B, C, and D, Examples of PHH3-labeled mitotic figures shown at higher magnification. B, Clearly stained telophase mitotic
figure seen in the center of the image and a metaphase mitosis in the upper right quarter (objective 40×). Two metaphase plates
(C, objective 100×) and anaphase mitosis (D, objective 100×) shown at high power.
Am J Clin Pathol 2007;128:118-125
© American Society for Clinical Pathology
121
DOI: 10.1309/HXUNAG34B3CEFDU8
121
121
Kim et al / PHH3 AND MITOTIC INDEX IN MENINGIOMAS
increased in grade to atypical meningiomas (WHO grade II),
and 1 atypical meningioma would have increased to an anaplastic meningioma (WHO grade III). Of these, 3 of the WHO
grade I meningiomas and the atypical meningioma recurred.
Recurrence-Free Survival Analyses
Results of ROC analysis for H&E MI, PHH3 MI, and Ki67 LI are specified in ❚Table 2❚. With “recurrences and/or death”
as the classification variable, ROC analysis revealed the following cutoff points for each MI as the most appropriate: H&E MI,
4 or more per 10 HPF; and PHH3 MI, 6 or more per 10 HPF.
The Ki-67 LI did not reach significance in the ROC analysis.
However, at the expense of a low sensitivity, a Ki-67 LI exceeding 8% indicated recurrence or death with high specificity.
With only “recurrence” as the classification variable, a
PHH3 MI of 4 or more per 10 HPF or an H&E MI of 4 or
A
more per 10 HPF significantly indicated cases at risk of recurrence. The PHH3 MI was more sensitive in the prediction of
recurrence probability than the traditional H&E MI. Three
recurrent cases, which otherwise would have been ignored by
the traditional mitosis counting method, could be therefore
identified by regarding the PHH3 MI. However, both MIs
were significantly superior to the Ki-67 LI in the discrimination between recurrent and nonrecurrent cases, as revealed by
comparative ROC analyses (H&E MI vs Ki-67 LI, P = .022;
PHH3 MI vs Ki-67 LI, P = .015). The highest discriminatory
power among the proliferation indices is suggested for the
PHH3 MI, which exhibits the greatest areas under the curve
and the smallest P values by the Mann-Whitney U test.
The results of the log-rank test for H&E MI, PHH3 MI,
and Ki-67 LI are summarized in ❚Table 3❚, and Kaplan-Meier
curves are displayed in ❚Figure 2❚. Univariate analyses
40
40
35
35
30
30
PHH3 MI LP
H&E MI
B
25
20
15
25
20
15
10
10
5
5
0
0
10
20
30
PHH3 MI
40
50
0
0
10
20
30
PHH3 MI
40
50
❚Figure 1❚ Relationship between mitosis counts (H&E) and phosphohistone H3 (PHH3) mitotic index (MI) and between PHH3 MI
assessed at high or low power (PHH3 MI LP). The scatter diagrams plot the H&E MIs and corresponding PHH3 MIs (A) or the
PHH3 MIs at high power and low power (B) for each meningioma. The correlation between H&E MI and PHH3 MI is high, and
most of the data points lie near the regression line. The majority of data points also lie beneath the 45° diagonal line, indicating
higher values for the PHH3 counting method. The linear trend is even more evident in the comparison between PHH3 MIs
assessed at high and low magnification.
❚Table 2❚
Results of Receiver Operator Characteristics in 99 Meningioma Cases
Classification Variable
Recurrence and/or death
H&E MI
PHH3 MI
Ki-67 LI
Recurrence only
H&E MI
PHH3 MI
Ki-67 LI
Cutoff*
Sensitivity (%)
Specificity (%)
Area Under the Curve
P†
≥4 mitoses
≥6 mitoses
>8%‡
33
30
6
96
94
94
0.645
0.658
0.565
.0252
.014
.3273
≥4 mitoses
≥4 mitoses
>8%‡
36
56
6
96
94
94
0.686
0.819
0.571
.0088
<.0001
.4751
H&E MI, mitotic index assessed on H&E-stained slides; LI, labeling index; PHH3-MI, mitotic index assessed in anti–phosphohistone H3–immunolabeled specimens.
* Most appropriate cutoff point with the highest discriminatory power.
† By the Mann-Whitney U test.
‡ Cutoff point for the Ki-67 LI with the most appropriate specificity.
122
122
Am J Clin Pathol 2007;128:118-125
DOI: 10.1309/HXUNAG34B3CEFDU8
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
revealed a significantly higher risk of recurrence and/or death
for higher proliferation, as measured by all 3 methods. The
hazard ratios were 2.1 for an H&E MI of 4 or more per 10
HPF, 2.3 for a PHH3 MI of 6 or more per 10 HPF, and 7.6 for
a Ki-67 LI of more than 8%.
Results of multivariate Cox hazards modeling are summarized in ❚Table 4❚. H&E MI and PHH3 MI could be identified as independent predictors of worse outcome (recurrence
and/or death), adjusted for age and extent of tumor resection
(with STR as the independent variable). The Ki-67 LI did not
reach significance in the multivariate setting.
tumor cells is known to be the most reliable and objective
histopathologic parameter.3 By following the instructions for
the assessment of the MI as suggested by the WHO, we found
the same mitotic threshold of an MI of 4 or more per 10 HPF
as the most appropriate cutoff point for defining meningiomas
with higher risk of recurrence or death, as proposed by the
WHO. Cox regression analysis also revealed the H&E MI as
an independent predictor of worse outcome, as suggested by
Perry et al.3 However, in practice MF counting in H&Estained slides turned out to be a time-consuming and tedious
task, in particular because in many cases the MFs had to be
carefully searched at high power over the whole slide.
Another quantitative indicator of proliferation activity is
the Ki-67 LI, which is determined as the percentage of Ki-67
positively stained tumor cell nuclei. The Ki-67 antigen is a
nonhistone nuclear protein that is expressed in cycling cells.
The determination of the Ki-67 LI has the advantage over the
assessment of the H&E MI that the immunolabeled nuclei can
be identified much easier than MFs in H&E-stained slides.
Discussion
The 3-step grading system of meningiomas, distinguishing well-differentiated, atypical, and malignant meningiomas
by the guidelines suggested by the WHO admits risk estimation based on histomorphologic criteria. Among the proposed
negative prognostic factors, the degree of proliferation of
❚Table 3❚
Univariate Recurrence-Free Survival Analysis in 99 Meningioma Cases
H&E MI, ≥4/10 HPF
PHH3 MI
≥4/10 HPF
≥6/10 HPF
Ki-67 LI >8%
Relative Risk of
Recurrence (%)
Hazard Ratio (95%
Confidence Interval)
Median Recurrence-Free
Survival Time (mo)
P*
20
2.1 (1.1-7.7)
70
.0327
19
27
29
1.3 (0.7-2.9)
2.3 (1.4-11.7)
7.6 (3.6-7,155.3)
81
69
25
.3128
.0077
<.0001
H&E MI, mitotic index assessed in H&E-stained slides; HPF, high-power fields; LI, labeling index; PHH3 MI, mitotic index assessed in anti–phosphohistone H3–immunolabeled
specimens.
* Log-rank test.
A
B
100
Survival Probability (%)
Survival Probability (%)
100
80
60
40
20
80
60
40
20
0
0
0
20
40
60
Months
80
100
120
0
20
40
60
Months
80
100
120
❚Figure 2❚ Relationship of mitotic indices to recurrence-free survival in 99 patients followed up until recurrence and/or death or
for at least 5 years. The Kaplan-Meier curves for the H&E mitotic index (MI; solid line, <4; dashed line, ≥4) (A) and for the
phosphohistone H3 (PHH3) MI (solid line, <6; dashed line, ≥6) (B) illustrate the significantly worse outcome for meningiomas
with higher mitotic activity.
Am J Clin Pathol 2007;128:118-125
© American Society for Clinical Pathology
123
DOI: 10.1309/HXUNAG34B3CEFDU8
123
123
Kim et al / PHH3 AND MITOTIC INDEX IN MENINGIOMAS
The prognostic significance of the Ki-67 LI and other proliferation indices in meningiomas is well known.15-20
According to the current WHO grading system, meningiomas
with high proliferation indices, ie, Ki-67 LI more than 5% to
10%, should be classified as meningiomas with a greater likelihood of recurrence and/or aggressive behavior, whereas no
universal values have been specified for determining recurrence risk. This is due to a remarkably wide range of Ki-67 LI
values suggested as prognostic cutoff levels among studies,
owing to different staining techniques and counting methods.15
We found a highly significant association of recurrent
meningiomas and Ki-67 LI levels exceeding 8%. This threshold value is comparable with the mean Ki-67 LI value of 7.2%
for atypical meningiomas mentioned by Maier et al16 and
cited by the WHO and also with the threshold value of a Ki67 LI more than 10%, proposed by Ho et al17 and Torp et al,18
for distinguishing recurrent from nonrecurrent meningiomas.
However, as mentioned by others,19,20 there is remarkable
overlap of Ki-67 LI values among meningioma grades and
recurrent/nonrecurrent meningiomas. Accordingly, despite the
high specificity of 94% in the detection of recurrent meningiomas, our results suggest insufficient sensitivity because
only 6% of recurrent cases exceeded the threshold level of a
Ki-67 LI of more than 8%. Comparative ROC and multivariate analyses also revealed a significantly higher discriminatory power and predictive value for both MIs compared with the
Ki-67 LI. However, as suggested in previous reports,15,19 the
Ki-67 LI could be applied as a useful ancillary study for
meningiomas with “borderline atypia.”
To overcome the mentioned difficulties and limitations of
yet existing quantitation methods of proliferation activity, in
the present study, the applicability and prognostic significance
of the PHH3 MI was evaluated. At present, to our knowledge
(based on PubMed query), this is the largest series of tumor
samples immunohistochemically studied for PHH3.
In our experience, the anti-PHH3–labeled MF counting
method could be proved a rapid, facile, and sensitive method
for the assessment of mitotic activity. The identification of the
hot spot areas was much easier and more reproducible because
the MFs were clearly visible at lower magnification. We systematically investigated the evaluation of the PHH3 MI at low
power and compared the results with those obtained by summing the MFs at high magnification. The assessment of the
PHH3 MI at low magnification was much faster and as reliable as the traditional summing in 10 consecutive areas at high
magnification. Labeled objects that are not clearly identifiable
as MFs at low magnification had to be specified at high power.
However, this procedure was less tedious and time-consuming
than the summing of MFs at high power. There were neither
significant deviations in the mitotic counts nor changes in
grade when both PHH3+ MF counting methods were compared with each other. The assessment of the PHH3 MI at low
124
124
Am J Clin Pathol 2007;128:118-125
DOI: 10.1309/HXUNAG34B3CEFDU8
❚Table 4❚
Multivariate Recurrence-Free Survival Analysis by Cox
Hazard Regression in 99 Meningioma Cases
Age
Extent of resection
(subtotal resection)
H&E MI
PHH3 MI
Ki-67 LI
Hazard Ratio (95%
Confidence Interval)
P*
1.03 (1.00-1.06)†
2.19 (0.97-4.91)
.0415
.0589
1.08 (1.03-1.14) ‡
1.07 (1.06-1.10) ‡
1.03 (0.96-1.10)§
.0007
.0004
.3645
H&E MI, mitotic index assessed in H&E-stained slides; LI, labeling index; PHH3 MI,
mitotic index assessed in anti–phosphohistone H3–immunolabeled specimens.
* By the log-rank test.
† Hazard ratio per year of patient age.
‡ Hazard ratio per mitosis.
§ Hazard ratio per percentage in the Ki-67 LI.
magnification can be suggested as the fastest way to reliably
assess the MI in meningiomas.
The major advantage of PHH3 labeling is the clear and
unambiguous distinction of MFs from other chromatin
changes, which permits objective, reliable, and reproducible
identification of MFs. We, therefore, confirm the results of
previous reports,4,10-12 in which the advantages of MF counting in anti–PHH3-immunostained tumor specimens over the
conventional MF counting method are described. Our results
also confirm the higher sensitivity in the identification of MFs
by using PHH3 immunoreaction as reported by Ribalta et al.12
They reported a higher sensitivity of the PHH3 mitosis counting method, which resulted in an increase in tumor grade in
17% of cases studied.
Likewise, PHH3+ MF counting provided higher MIs in all
3 meningioma grades and an increase in tumor grade in 10.6%
(28/265) of cases in our series. Allowing for this higher sensitivity of the PHH3 mitosis counting method, the mitotic thresholds proposed by the WHO have to be reconsidered if PHH3
MIs are applied to tumor grading. Accordingly, the threshold
level at a PHH3 MI of 4 or more per 10 HPF did not reach significance in the univariate recurrence-free survival analysis.
Perry and coworkers3 defined the threshold value of 4 or
more mitoses per 10 consecutive HPF based on the results of
multivariate survival analyses on H&E-stained tumor specimens. In the present study, multivariate Cox hazards modeling
technique identified the PHH3 MI as an independent predictor
of recurrence-free survival with the highest sensitivity in the
detection of recurrent cases among proliferation indices. By
using discriminant analyses, prognostic PHH3-specific
threshold values could be defined, which permits a distinction
between each risk group, or rather, tumor grade. However, as
it holds for Ki-67 and other immunohistochemical markers of
proliferation, there is variability between laboratories in terms
of staining results and interpretation of anti-PHH3 immunoreactions. Thus, it is difficult to define universal cutoffs for
© American Society for Clinical Pathology
Anatomic Pathology / ORIGINAL ARTICLE
PHH3 MIs and the thresholds suggested herein may not be
extrapolatable to other laboratories.
In our series, 6 or more PHH3+ mitoses per 10 HPF indicate meningiomas with significantly higher risk of recurrence
or death within 5 years after primary surgery, independent of
patient age or extent of tumor resection. Based on the traditional H&E MI, 36% of patients at risk for recurrence could
be identified, whereas stratification based on the PHH3 MI
indicated recurrent cases with a sensitivity of 56%. Thus, the
PHH3 mitosis counting method also increases the sensitivity
in the identification of cases at risk of recurrence. In our series,
3 recurrent cases could be correctly identified as high-risk
cases based on the PHH3 MI, whereas they would not have
been so identified by the traditional H&E-stained MF counting method. These findings emphasize the reliability and
advantages of the PHH3 mitosis counting method in the
appropriate risk stratification of patients with meningioma.
Owing to the small sample of anaplastic meningiomas in our
series, it is difficult to define a distinct cutoff point. However,
our results suggest a threshold level at 30 mitoses or more for
defining WHO grade III meningiomas because this cutoff
exhibits the least overlap and the best discrimination between
atypical and anaplastic meningiomas.
Conclusions
The assessment of MI based on PHH3 immunolabeling of
MFs has been proved as a reliable method for the quantitation of
the mitotic activity. The PHH3 MI was identified as the most
sensitive and reliable measure of proliferation activity.
Multivariate analyses identified the PHH3 MI as an independent
predictor of recurrence or death in patients with meningioma. In
our series, we found a threshold level of PHH3 MI of 6 or more
mitoses per 10 HPF for defining meningiomas with significantly higher risk for worse outcome and PHH3 MI of 30 or more
mitoses per 10 HPF for defining anaplastic meningiomas.
From the 1Institute of Neuropathology and the 2Department of
Neurosurgery, Saarland University, School of Medicine,
Homburg/Saar, Germany.
Address reprint requests to Dr Kim: Institute of Neuropathology,
Saarland University, School of Medicine, Bldg 90.3, D-66421
Homburg-Saar, Germany.
References
1. Burger PC, Scheithauer BW, Vogel FS, eds. Surgical Pathology
of the Nervous System and Its Coverings. New York, NY:
Churchill Livingstone; 2002.
2. Louis DN, Scheithauer BW, Budka H, et al. Meningiomas.
In: Kleihues P, Cavenee WK, eds. Pathology and Genetics of
Tumours of the Nervous System. Lyon, France: IARC Press;
2000:176-184. World Health Organization Classification of
Tumours.
3. Perry A, Stafford SL, Scheithauer BW, et al. Meningioma
grading: an analysis of histological parameters. Am J Surg
Pathol. 1997;21:1455-1465.
4. 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;30:657-664.
5. Shibata K, Inagaki M, Ajiro K. Mitosis-specific histone H3
phosphorylation in vitro in nucleosome structures. Eur J
Biochem. 1990;192:87-93.
6. Shibata K, Ajiro K. Cell cycle–dependent suppressive effect
of histone H1 on mitosis-specific H3 phosphorylation. J Biol
Chem. 1993;268:18431-18434.
7. Gurley LR, D’anna JA, Barhan SS, et al. Histone
phosphorylation and chromatin structure during mitosis in
Chinese hamster cells. Eur J Biochem. 1978;84:1-15.
8. Hendzel MJ, Wei Y, Mancini MA, et al. Mitosis-specific
phosphorylation of histone H3 initiates primarily within
pericentromeric heterochromatin during G2 and spreads in an
ordered fashion coincident with mitotic chromosome
condensation. Chromosoma. 1997;106:348-360.
9. Hendzel MJ, Nishioka WK, Raymond Y, et al. Chromatin
condensation is not associated with apoptosis. J Biol Chem.
1998;18:24470-24478.
10. Baehner R, Weidner N. Enhanced mitotic figure counting
in breast carcinomas using a mitosis-specific antibody:
anti–phosphohistone-H3 (PHH3) [abstract]. Mod Pathol.
2000;13:17A.
11. 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;30:83-89.
12. Ribalta T, McCutcheon IE, Aldape KD, et al. The mitosisspecific antibody anti–phosphohistone-H3 (PHH3) facilitates
rapid reliable grading of meningiomas according to WHO
2000 criteria. Am J Surg Pathol. 2004;28:1532-1536.
13. Simpson D. The recurrence of intracranial meningiomas after
surgical treatment. J Neurol Neurosurg Psychiatry. 1957;20:2239.
14. Kim YJ, Romeike BFM, Uszkoreit J, et al. Automated nuclear
segmentation in the determination of the Ki-67 labeling-index
in meningiomas. Clin Neuropathol. 2006;25:67-73.
15. Perry A, Stafford SL, Scheithauer BW, et al. The prognostic
significance of MIB-1, p53, and DNA flow cytometry in
completely resected primary meningiomas. Cancer.
1998;82:2262-2269.
16. Maier H, Wanschitz J, Sedivy R, et al. Proliferation and DNA
fragmentation in meningioma subtypes. Neuropathol Appl
Neurobiol. 1997;23:496-506.
17. Ho D, Hsu C, Ting L, et al. Histopathology and MIB-1
labeling index predicted recurrence of meningiomas: a
proposal of diagnostic criteria for patients with atypical
meningiomas. Cancer. 2002;94:1538-1547.
18. Torp SH, Lindboe CF, Gronberg BH, et al. Prognostic
significance of Ki-67/MIB-1 proliferation. Clin Neuropathol.
2005;24:170-174.
19. Abramovich CM, Prayson RA. Histopathologic features
and MIB-1 labeling indices in recurrent and nonrecurrent
meningiomas. Arch Pathol Lab Med. 1999;123:793-800.
20. Kolles H, Niedermayer I, Schmitt C, et al. Triple approach for
diagnosis and grading of meningiomas: histology, morphometry
of Ki-67/Feulgen stainings, and cytogenetics. Acta Neurochir
(Wien). 1995;137:174-181.
Am J Clin Pathol 2007;128:118-125
© American Society for Clinical Pathology
125
DOI: 10.1309/HXUNAG34B3CEFDU8
125
125