Keywords
ATRX - glioblastoma - IDH-1 - immunohistochemistry - p53
Introduction
Glioblastoma (GBM) is recognized as the most aggressive type of diffuse astrocytic
glioma, characterized histologically by features such as nuclear atypia, cellular
pleomorphism, brisk mitotic activity, microvascular proliferation, and/or necrosis.
Brain tumors affect approximately 17,000 individuals annually, with gliomas accounting
for nearly 60% of these cases.[1] Among gliomas, GBM is the most frequently diagnosed, showing a higher incidence
in males.[1] It is associated with a dismal prognosis, and most patients succumb within a year
of diagnosis.[2]
Although GBMs often arise sporadically, familial forms have also been documented.
These tumors predominantly develop in the supratentorial compartment—especially the
frontal lobe—and their highly infiltrative nature makes them difficult to distinguish
from adjacent normal brain parenchyma.[3] In recent years, diagnostic approaches have expanded beyond histology to include
immunohistochemistry (IHC), molecular pathology, and biomarker profiling, all of which
now play a pivotal role.[4]
According to existing data, around 60% of GBMs originate de novo, while the remaining
40% evolve from lower grade gliomas such as diffuse or anaplastic astrocytomas, often
over a period of 4 to 5 years.[5] These tumors commonly show altered expression of key regulatory genes and proteins,
including ATRX (alpha-thalassemia/mental retardation syndrome X-linked), p53, and
IDH-1 (isocitrate dehydrogenase 1).[1]
The significance of integrating molecular data with histopathological findings was
emphasized in the revised 2016 World Health Organization (WHO) classification of central
nervous system (CNS) tumors. Accurate classification now requires both morphological
assessment and molecular profiling, as tumors with similar histology may behave differently
depending on their genetic makeup and response to treatment.[6]
Standard treatment involves surgical resection followed by radiotherapy and chemotherapy.
Additional strategies, including integrin inhibitors and immunotherapy, are under
investigation. However, prognosis continues to depend on multiple variables, such
as anatomical site, histological grade, molecular characteristics, proliferative index,
and patient age. Although there has been a surge in research investigating the biology
and treatment of GBM, comprehensive clinical analysis remains limited.[3]
[7]
[8]
Despite aggressive multimodal therapy, recurrence is common due to the molecular heterogeneity
of GBM. Therefore, profiling protein expression of key molecular markers such as IDH-1,
p53, and ATRX via IHC is crucial for subclassification and understanding tumor biology.
This study aims to classify GBM cases using an IHC-based algorithm, in accordance
with the 2016 WHO classification of CNS tumors, which was the standard during the
study period (March 1, 2016–March 1, 2021).
Objectives of the Study
To subclassify GBM cases based on immunohistochemical expression patterns of IDH-1,
p53, and ATRX.
Materials and Methods
This was a descriptive cross-sectional study conducted on histologically confirmed
GBM cases received in the Department of Pathology over a 5-year period.
Inclusion Criteria
All newly diagnosed GBM cases (WHO Grade IV), classified as per the 2016 WHO classification
of CNS tumors.
Exclusion Criteria
Recurrent GBMs were excluded, as comprehensive data on their primary tumor characteristics
were often unavailable.
Tissue Processing and Immunohistochemistry
Representative tissue sections from tumor specimens were selected, processed, and
stained with hematoxylin and eosin (H&E). Cases diagnosed as GBM on histopathology
were included for further immunohistochemical analysis using antibodies against IDH-1,
p53, and ATRX. IDH-1 showed cytoplasmic staining. It was assessed and categorized
as either positive or negative. Antibody used was IDH-1 R132H (Clone QMOO2), a mouse
monoclonal from Quartett. ATRX which showed nuclear staining was evaluated and classified
as retained (positive in >10% of cells) or lost . Antibody used was Anti-ATRX, a rabbit
polyclonal IgG (HPA001906) from Sigma Life Science-Prestige.
p53 showed nuclear staining which was semi-quantitatively scored based on the proportion
of positive nuclei, using a four-tiered scale ([Table 1]). Antibody used was Anti-p53, a rabbit monoclonal (Clone QR025) from Quartett.
Table 1
Scoring of p53
|
Description
|
Score
|
Result
|
|
<10% of nuclei stained
|
0
|
Negative
|
|
10–30% of nuclei stained
|
1+
|
Positive
|
|
30.1–50% of nuclei stained
|
2+
|
Positive
|
|
>50% of nuclei stained
|
3+
|
Positive
|
Statistical Analysis
Data entry was performed using Microsoft Excel, and statistical analyses were conducted
using SPSS software (version 23). A p-value less than 0.05 was considered statistically significant.
Results
This study aimed to classify GBM cases based on immunohistochemical markers as per
the 2016 WHO classification of CNS tumors.
A total of 95 patients with histologically confirmed GBM were included. The mean age
of the study population was 50.2 years, ranging from 10 to 70 years, with 53 patients
above 50 years. The cohort comprised 56 males and 39 females.
Immunohistochemical Expression of IDH-1, p53, and ATRX
IDH-1 positivity was observed in 21 patients (22.1%). p53 overexpression was noted
in 36 patients (37.9%). ATRX loss of expression was seen in 44 patients (46.3%; [Table 2]).
Table 2
Immunoexpression of IDH-1, p53, and ATRX in enrolled GBMs
|
IHC marker
|
Positive
|
Negative
|
|
IDH-1
|
21 (22.1%)
|
74 (77.9%)
|
|
P53
|
36 (37.9%)
|
59 (62.1%)
|
|
ATRX
|
51 (53.7%)
|
44 (46.3%)
|
Abbreviation: GBM, glioblastoma.
Analysis of Two-Marker Combinations
[Table 3] summarizes the results of dual-marker IHC. The most frequent IDH-1/ATRX combination
was IDH-1–/ATRX+ (44.2%). For IDH-1/p53, IDH-1–/p53–(50.5%) was most common. Least
prevalent patterns included IDH-1 +/ATRX+ (9.5%) and IDH-1 +/p53+ (10.5%). Statistical
analysis showed no significant association between expression patterns in the two-marker
combinations.
Table 3
Immunohistochemical results of various combinations of two protein pairs
|
IHC
|
Two-protein pairs
|
No. of cases (%)
|
p-Value
|
|
IDH-1/ATRX
|
IDH-1 +/ATRX−
|
12 (12.6%)
|
0.509
|
|
IDH-1 +/ATRX+
|
9 (9.5%)
|
|
IDH-1 −/ATRX−
|
32 (33.7%)
|
|
IDH-1 −/ATRX+
|
42 (44.2%)
|
|
IDH-1/p53
|
IDH-1 +/p53+
|
10 (10.5%)
|
0.298
|
|
IDH-1 +/p53−
|
11 (11.6%)
|
|
IDH-1 −/p53+
|
26 (27.4%)
|
|
IDH-1 −/p53−
|
48 (50.5%)
|
|
p53/ATRX
|
p53 +/ATRX−
|
13 (13.6%)
|
0.11
|
|
p53 +/ATRX+
|
23 (24.2%)
|
|
p53 −/ATRX−
|
31 (32%)
|
|
p53 −/ATRX+
|
28 (29.4%)
|
Analysis of Three-Marker Combinations
[Table 4] outlines the results of combined expression for IDH-1, p53, and ATRX. The most common
profile was IDH-1–/p53–/ATRX + , seen in 27.3% of cases ([Fig. 1]). The least common profile, IDH-1 +/p53 +/ATRX–, was found in only 3.1% of cases
([Fig. 2]). A statistically significant correlation was found between p53 and ATRX expression
in IDH-1 mutant tumors (p = 0.01).
Table 4
Immunohistochemical results of various combinations of 3 proteins
|
IDH-1
|
p53
|
ATRX
|
Frequency
|
%
|
p-Value
|
|
IDH-1+
|
p53+
|
ATRX−
|
3
|
3.15
|
0.01
|
|
p53+
|
ATRX+
|
7
|
7.4
|
|
p53−
|
ATRX−
|
9
|
9.5
|
|
p53−
|
ATRX+
|
2
|
2.1
|
|
IDH-1−
|
p53+
|
ATRX−
|
10
|
10.5
|
0.54
|
|
p53+
|
ATRX+
|
16
|
16.8
|
|
p53−
|
ATRX−
|
22
|
23.1
|
|
p53−
|
ATRX+
|
26
|
27.3
|
Fig. 1 The figure shows representative pictures of one of the GBM cases with IDH-1-/p53-/
ATRX + . (A) H&E (20 × ), (B) IDH-1-, (C) p53-, (D) ATRX + , (B–D) 40 × . GBM, glioblastoma.
Fig. 2 The figure shows representative pictures of one of the GBM cases with IDH-1 +/ p53 +/ATRX-.
(A) H&E (20 × ), (B) IDH-1 + , (C) p53 score 2, (D) ATRX-, (B–D) 40 × . GBM, glioblastoma.
Histological Subtypes of Glioblastoma
The distribution of histological subtypes is shown in [Table 5]. Classic GBM was the most common histological subtype ([Fig. 3]).
Table 5
Morphological subtypes of GBM cases
|
Morphological subtypes
|
Frequency
|
Percentage
|
|
Classical GBM
|
71
|
74.7
|
|
Giant cell GBM
|
18
|
18.9
|
|
Gemistocytic GBM
|
2
|
2.1
|
|
Oligodendroglial
|
1
|
1.05
|
|
Primitive neuronal
|
1
|
1.05
|
Abbreviation: GBM, glioblastoma.
Fig. 3 The figure shows (A) microvascular proliferation (B) pseudopalisading necrosis in classical GBM (A, B) 20 × . GBM, glioblastoma.
A total of 18 cases were giant cell GBM ([Fig. 4A]), 2 cases were gemistocytic GBM, 1 case was GBM with primitive neuronal component
([Fig. 4B]), and 1 case was GBM with oligodendroglial component.
Fig. 4 (A) Giant cell glioblastoma (20 × ) and (B) GBM with primitive neuronal component (40 × ). GBM, glioblastoma.
Among giant cell GBM cases, 72.2% were IDH-1 negative, 61.1% showed p53 positivity,
and 55.5% had loss of ATRX expression.
Discussion
GBM is the most malignant and aggressive primary brain tumor, known for its resistance
to standard treatment modalities. This resistance is largely attributed to the tumor's
molecular heterogeneity, involving variations in key regulatory genes. Our study aimed
to assess the immunohistochemical expression of IDH-1, p53, and ATRX in GBM cases,
and to evaluate marker combinations that could effectively classify tumors based on
their molecular profile as per the 2016 WHO classification of CNS tumors.
Out of 95 GBM cases analyzed, 56 (59%) were males and 39 (41%) were females, closely
aligning with the findings of Chaurasia et al, who reported a similar male predominance
(58.2%) in their cohort of 163 cases.[1] Lee et al observed an equal gender distribution (75 males and 75 females out of
150 cases).[2]
Regarding patient age, our cases ranged from 10 to 70 years with a mean of 50.2 years.
This finding corresponds to studies by Chaurasia et al (mean: 49.4 years; range: 21–79)[1] and Lee et al (mean: 58.5 years; range: 19–85).[2]
In terms of IDH-1 expression, we found 22.1% positivity, which was higher than those
of Chaurasia et al (10.4%)[1] and Lee et al (11.1%),[2] but comparable to that of Dahuja et al, who reported 31.1% positivity in a smaller
cohort of Indian patients.[4]
p53 overexpression was identified in 37.9% of cases in our study, aligning closely
with Dahuja et al (40%).[4] Higher rates were reported by Chaurasia et al (48.4%) and Lee et al (49.5%).[1]
[2]
Loss of ATRX expression was seen in 44 cases (46.3%). This was higher than the 15.3%
reported by Chaurasia et al,[1] 4.8% by Gülten et al,[9] and 24.4% by Dahuja et al.[4] Other studies such as Reuss et al (18%)[10] and Liu et al (26%)[11] have shown intermediate values. These findings reinforce the role of IHC in detecting
ATRX mutations, as it correlates well with molecular alterations.[12]
[13]
[14]
In the analysis of dual-marker combinations, IDH-1–/p53–was the most prevalent pattern
(50.5%), similar to Chaurasia et al (46%).[1] This may be due to limitations of IHC in detecting non-R132H mutations of IDH-1,
which require gene sequencing for confirmation. Also, not all TP53 mutations lead
to protein accumulation; nonsense mutations may yield negative IHC results.
The combination IDH-1–/ATRX+ was seen in 44.2% of our cases, whereas Chaurasia et
al found this pattern in 78.5% of cases. For p53 and ATRX, our data showed that 29.4%
had p53–/ATRX + , compared to 42.9% in Chaurasia et al's cohort.[1] These dual-marker combinations did not yield statistically significant associations
in our study.
When evaluating all three markers (IDH-1, p53, ATRX) together, we identified eight
molecular profiles. The most frequent was IDH-1–/p53–/ATRX + , seen in 27.3% of cases,
mirroring Chaurasia et al's result, where 39% had the same profile. Among IDH-mutant
cases, IDH-1 +/p53–/ATRX– was the most common combination in our series (9.5%), higher
than the 2.5% reported by Chaurasia et al.[1] A statistically significant association between p53 and ATRX expression in IDH-1
mutant cases was observed (p = 0.01), in line with other reports.
Regarding histological subtypes, the classical GBM was most common, followed by giant
cell GBM. In giant cell GBMs, most cases were IDH-1-negative, p53-positive, and showed
ATRX loss. These findings are consistent with Cantero et al, who noted IDH-1 negativity
in 97.2%, p53 expression in 89%, and ATRX loss in 24% of giant cell GBM or GBMs with
a giant cell component.[15]
Conclusion
Our study revealed distinct immunohistochemical expression patterns among GBM subtypes.
Among IDH-wild-type GBMs, the predominant phenotype included negative p53 expression
with retained ATRX staining. In contrast, IDH-mutant tumors most frequently exhibited
negative p53 expression along with ATRX loss.
These patterns suggest that analyzing the combined expression of IDH-1, p53, and ATRX
proteins may provide meaningful subclassification of GBM cases. This approach can
aid in diagnosis, influence treatment strategies, and may potentially offer prognostic
value.
Our findings contribute to a deeper understanding of the molecular diversity and biological
aggressiveness of GBM. By identifying molecular profiles through IHC, this study supports
efforts to develop more targeted and individualized treatment options.
Limitations
A key limitation of this study is its relatively small sample size. Larger studies
involving broader populations are essential to validate and generalize these findings.
Additionally, while IHC serves as a practical tool in resource-limited settings, it
cannot detect all forms of IDH and TP53 mutations—particularly non-R132H IDH-1 variants,
IDH-2 mutations, and null mutations of p53. These require molecular confirmation via
gene sequencing, which was not feasible in our setting.
Future research incorporating the 2021 WHO classification and molecular diagnostic
techniques will be crucial for refining GBM subtyping, enhancing prognostication,
and improving therapeutic outcomes.
