TERT Promoter Mutation in Adult Glioblastomas: It's Correlation with Other Relevant Molecular Markers
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.310096
Source of Support: None, Conflict of Interest: None
Keywords: EGFR, Glioblastoma, IDH1/2, MGMT promoter methylation, TERT promoter mutations
Glioblastomas (GBM) is the commonest and unfortunately, the most malignant primary brain tumor in adults. They are very heterogenous both clinically and biologically. It is now shown to comprise six distinct molecular subgroups - Isocitrate dehydrogenase (IDH)-mutant, Histone3K27-mutant, Histone3G34-mutant, Receptor Tyrosine Kinase I (RTKI), Mesenchymal and RTKII. RTKI and RTKII subgroups are characterized by platelet-derived growth factor receptor A (PDGFRA) and epidermal growth factor receptor (EGFR) gene amplifications respectively; mesenchymal subgroup is characterized by multiple copy number variations (predominantly chromosome 7 gains and 10 loss). K27, G34, and RTKI groups are predominantly of pediatric age-group; while the other three are predominantly of adults. Currently, post-2016 WHO classification, mutational evaluation for IDH1/2, H3.3 and H3.1, is an integral component for routine classification. Since 2005, evaluation for O6-methylguanine-DNA methyltransferase promoter (pMGMT) methylation as a predictive marker, has become an integral component of routine neuro-oncology practice for adult GBM. Additionally, Telomerase reverse transcriptase (pTERT) promoter mutations have shown to be a distinct, independent, and superior prognostic marker in adult GBMs.,,
pTERT-mutation (C228T and C250T mutations at -124 bp and -146 bp upstream of ATG site) is one of the two deranged telomere maintenance mechanisms in cancers, seen in approximately 80-86% of primary GBMs and in 6-28% in secondary GBMs.,,,,
However, there is a paucity of the data of pTERT mutation, especially its relation with other currently practiced molecular parameters. In this study, thus adult GBMs were evaluated for pTERT-mutations and also correlated with clinicohistological findings, IDH1/2-mutations, EGFR amplification and pMGMT-methylation, to address the lacunae in the literature pertaining to our subcontinent.
Sample: Institutional review board-approved study, included histologically diagnosed adult (>18 years) GBM (which were non-1p/19q co-deleted by fluorescence in-situ hybridization). The relevant clinical details were noted from the medical records.
Histopathology: Independently classified into histological subtypes on Haematoxylin and eosin-stained slides. Immunohistochemistry (IHC) for p53 (1:400 dilution; DO7; DAKO; Denmark), IDH1R132H (1:100; H09; Dianova; Germany) and ATRX protein (1:750; Polyclonal; Sigma; USA) was done in all cases by on Ventana Benchmark XT autoimmunostainer. The findings were recorded independently of the histological subtypes. Expression of p53 protein was stratified (as described in earlier study) into - positive, negative, and focal positive. ATRX protein was interpreted as retained (tumor cells showing nuclear staining), loss (no tumor cells show nuclear staining with reactivity of native cells-like endothelial and/or neurons), and non-contributory (no positivity with the internal control). IDH1R132H was expressed as positive and negative.
Fluorescence In-Situ Hybridisation (FISH): EGFR gene-amplification was evaluated using locus-specific identifier dual-color-probe kit, as described in our earlier study.
Sequencing for IDH1R132, IDH2R172 and pTERT mutations: Genomic DNA was isolated from the representative paraffin block of each sample using the QIAamp DNA FFPE Tissue kit (Qiagen, Germany), quantified using the Nano Drop spectrophotometer (Thermo Scientific, USA), amplified by polymerase chain reaction (PCR) for the TERT promoter, exon 4 of IDH1R132 and IDH2R172 using appropriate primers (see appendix for detailed procedure) and analyzed via Sanger sequencing using Chromas 4.3.4 software.[Additional file 1]
Methylation-specific PCR (MSP) for pMGMT methylation: Done as elaborated in our earlier study. (See also appendix for the detailed procedure).
Statistical analysis: The Chi-square was performed for correlation between pTERT and other parameters using IBM-SPSS-21.
Clinico-histological findings: One hundred fifty-five cases [Table 1], age-ranged from 20-80 years (interquartile range [IQR]: 37-60 years; median: 51 years and mean: 49 years) formed the study-cohort. Frontal and temporal (n: 47 [30.4%] each) were commoner locations.
Histological subtypes were – conventional GBM (cGBM): 134 (86.5%), GBM with atleast one low power field of oligodendroglial-like morphological component (GBM-O): 16 (10.3%), gliosarcoma (GS): 1 (0.6%), giant cell GBM (GBM-G):2 (1.3%) and epithelioid GBM (eGBM): 2 (1.3%).
p53 (n: 152) and ATRX protein (n: 145) expression: Seventy-nine (52%) were p53 protein positive [Figure 1]a, 52 (34.2%) were p53 focal positive and 21 (13.8%) were p53 negative. No significant statistical correlation was identified between p53 protein expression and age-group and histological patterns.
Twenty-nine (20%) showed loss [Figure 1]b and 116 (80%) showed retained ATRX protein expression; while 10 were non-contributory. No case of >60 years age-group showed ATRX-loss. Of the twenty-seven p53 interpretable cases with ATRX-loss - 21 (77.8%) were p53 positive, 5 (18.5%) were p53 focal positive and 1 (3.7%) was p53 negative.
IDH mutation: The total cohort of samples was evaluated for IDH1/2 mutation, p53 protein, and ATRX protein expression [Figure 1]a,[Figure 1]b,[Figure 1]c,[Figure 1]d,[Figure 1]e and [Table 1]. Fifteen were immuno-positive for IDH1R132H; the rest 140 negative cases were sequenced for IDH1R132 and IDH2R172 mutations. Seven showed IDH1R132 heterozygous point mutation [IDH1R132H (c.395G>A): 1; IDH1R132C (c.394 C>T), IDH1R132G (c.394C>G) and IDH1R132S (c.394C>A): 2 each], one hundred nineteen had no mutations and 14 were uninterpretable. No case showed IDH2R172 mutations. Overall, twenty-two were IDH-mutant (mIDH) with age-range: 21-58 years (IQR: 30-51 years; median: 33.5 years). No significant association was identified for mIDH with histological subtypes (P = 0.9). Nineteen of 73 (26%) p53 protein-positive, 1/49 (2%) p53 focal positive and no (0/16) p53 negative cases were mIDH. 64.3% (18/28) of ATRX-loss and 2% (2/104) of ATRX-retained cases were mIDH.
EGFR gene amplification: Thirty-seven were amplified [Figure 1]f and [Table 1], 101 were non-amplified [Table 1] and 17 uninterpretable. EGFR gene amplification was seen across all age-groups (=30 years: 1, 31–40 years: 5, 41-50 years: 6, 51-60 years: 13, >60 years: 12). Twelve (32.5%) were p53 positive, 16 (43.2%) were p53 focal positive and 9 (24.3%) were p53 negative. Of the 98 EGFR non-amplified cases, 60 (61.2%) were p53 positive, 29 (29.6%) focal positive and 9 (9.2%) negative. The association of EGFR non-amplification and p53 protein positivity was statistically significant (P = 0.01). 28/95 (29.5%) EGFR non-amplified and none (0/36) of EGFR amplified case showed ATRX-loss. All the 33 EGFR amplified cases were wtIDH; while 22/93 EGFR non-amplified cases were mIDH. Of 104 wtIDH cases, 33 (31.7%) were EGFR-amplified and 71 (68.3%) were EGFR non-amplified. 68.3% of adult GBMs in this study cohort were non-EGFR amplified/non-IDH mutated.
pMGMT methylation: Fifty-three (43%) were methylated [Figure 1]g, 70 (57%) unmethylated [Figure 1]h and 32 uninterpretable [Table 1]. Age-range for pMGMT-methylated was 21-75 years (IQR: 37-60 years; median: 51 years) and for unmethylated group was 18-80 years (IQR: 32-58 years; median: 50 years). No significant statistical correlation was observed between pMGMT-methylation and age and histological patterns (p value = 0.6 and 0.8 respectively). 26.5% (13/49) and 73.5% (36/49) of pMGMT-methylated cases were mIDH and wtIDH respectively; while 7.4% (5/68) and 92.6% (63/68) of pMGMT-unmethylated cases were mIDH and wtIDH respectively. A significant statistical correlation was observed between mIDH and pMGMT-methylation (p value = 0.00); however, it is to be noted that 73.5% of pMGMT-methylated were wtIDH. Of the 50 pMGMT-methylated cases, 12 (24%) were EGFR-amplified and 38 (76%) were EGFR non-amplified. 28.6% (18/63) and 71.4% (45/63) of pMGMT-unmethylated were EGFR-amplified and non-amplified respectively. No significant statistical correl ation between EGFR and pMGMT was observed (P = 0.6).
pTERT mutation: One hundred fifteen were interpretable [Figure 1]i, [Figure 1]j and [Table 1], [Table 2]; rest 40 yielded noisy sequencing data and rendered uninterpretable. 78 showed genetic polymorphism i.e., -245 T > C (rs2853669; heterozygous: 55 and homozygous: 23) and 50 (43.5%) showed C228T/C250T mutations (C228T: 36[72%; 3 homozygous and 33 heterozygous]; C250T: 14 [28%; 14 heterozygous.]). 23 of 36 C228T mutated and 13 of 14 C250T mutated were associated with polymorphism. 42 showed only polymorphism without C228T/C250T and 23 showed no sequence alterations. C228T and C250T mutations were mutually exclusive and were grouped as pTERT-mutant for further comparative evaluation. Age-range of pTERT-mutated cases (n: 50): 30-75 years (IQR: 47-61 years; median: 55 years) and for pTERT wild-type cases (n: 65): 18-80 years (IQR: 30-60 years; median: 44 years). 70% (35/50) of pTERT-mutant were >50 years but the correlation was statistically insignificant (P = 0.8). 44/99 cGBM, 5/13 GBM-O and 1/1 eGBM showed pTERT mutation.
Of the 49 pTERT-mutant cases, 24 (49%) were p53 protein positive. None of the ATRX-loss cases were pTERT-mutant and all 47 pTERT-mutant cases showed retained ATRX protein, which was statistically significant (P = 0.00). 95.7% (44/46) of pTERT-mutant cases were of wtIDH and 4.3% (2/46) were mIDH (cGBM with C250T mutation and polymorphism; GBM-O with C228T and without polymorphism). 34% (19/56) and 66% (37/56) of pTERT wild-type cases were mIDH (polymorphism: 16 [9 heterozygous; 7 homozyogous]) and wtIDH (polymorphism: 56 [41 heterozygous; 15 homozyogous]) respectively. The association of wtIDH and pTERT-mutation was statistically significant (P = 0.00).
Twenty-one (42%) and 29 (58%) of 50 pTERT-mutant cases were EGFR-amplified (C228T: 15[8 with polymorphism], C250T: 6[5 with polymorphism]) and non-amplified (C228T: 21[15 with polymorphism], C250T: 8[all with polymorphism]) respectively. Nine (15%) and 50 (85%) of 59 pTERT wild-type cases were EGFR-amplified (6 with polymorphism) and non-amplified (33 with polymorphism) respectively. EGFR amplification was more commonly associated with pTERT-mutation (42%) as compared to 15.2% pTERT wild-type (statistically significant; P = 0.00).
Sixteen (36.4%) and 28 (63.6%) of 44 pTERT-mutant cases were pMGMT-methylated (C228T: 11[7 with polymorphism], C250T: 5[all with polymorphism]) and unmethylated (C228T: 20[11 with polymorphism], C250T: 8[7 with polymorphism]) respectively. pTERT-mutation was more commonly observed with pMGMT-unmethylation (63.6%) than methylation (36.4%). However, this correlation was not statistically significant (P = 0.08).
Based on mIDH and EGFR gene amplification, the study cohort was classified into three molecular groups - mIDH, EGFR-amplified and wtIDH/EGFR-nonamplified.
mIDH group (n: 22; 15.6%): Age-range was 21-58 years (IQR: 30-51 years; median: 33.5 years). None were of >60 years age group. 16 were IDH1R132H, two each were IDH1R132C, IDH1R132G and IDH1R132S mutations. No IDH2R172-mutation was observed. None of the mIDH cases were p53 negative. 18 (90%) of mIDH cases showed ATRX-loss and 2 (10%) showed retained ATRX protein. All mIDH cases were EGFR non-amplified. 9.5% (2/21; both of these cases showed retained ATRX) mIDH cases were pTERT-mutant and 72.2% (13/18) of mIDH showed pMGMT-methylation.
EGFR-amplified group (n: 37; 27%): Age-range was 26-74 years (IQR: 46-64 years; median: 55 years). 32.4% (12/37) EGFR amplified cases were p53 positive as compared to 61.2% (60/98) EGFR non-amplified cases, the latter association is statistically significant (P = 0.00). None of the EGFR amplified cases showed ATRX-loss (n: 36) and mIDH (n: 33). 40% (12/30) of EGFR amplified were pMGMT-methylated as compared to 46% (38/83) of EGFR non-amplified. 70% (21/30) of EGFR amplified cases were pTERT-mutant as compared to 36.7% (29/79) in EGFR non-amplification and was statistically significant (P = 0.00).
wtIDH/EGFR-nonamplified group (n: 71): Age-range for this group was 18-75 years (IQR: 39-61 years; median: 51 years). 54.3% (38/70) were p53 positive, 37.1% (26/70) were p53 focal positive and 8.6% (6/70) were p53 negative. 86.8% (59/68) were ATRX-retained and interestingly 13.2% (9/68) cases showed ATRX-loss. 49% (25/51) and 35% (21/60) of this group showed pTERT-mutation and pMGMT-methylation respectively.
Seventy-six cases were interpretable for IDH-mutations, EGFR-amplification, pTERT-mutations and pMGMT-methylation [Table 3] and [Figure 2]. In this cohort, the mIDH group showed a significantly lower median age (and age-range), frequent (13/14; 93%) p53 protein positivity/over-expression, frequent (12/14; 85.7%) loss of ATRX protein expression and infrequent (2/14; 14.3%) pTERT-mutations than the other two molecular groups. pMGMT-methylation showed significant statistical correlation for mIDH (9/14; 64.3%, P = 0.03); while it was not statistically significant for EGFR-amplification (9/22; 41%) whereas it was uncommon (12/40; 30%) in wtIDH/EGFR non-amplified group. pTERT-mutation was rare with mIDH; though, was more frequent in EGFR-amplified (17/22; 77.3%) as compared to that in wtIDH/EGFR non-amplified (19/40; 47.5%) group but it is not statistically significant .
The median age of adult GBMs, in this study was 51 years, which is a decade younger than the west with temporal and frontal as common supratentorial locations, which is concordant with the relevant literature.,,
Parsons et al., first reported somatic heterozygous point-mutation at codon 132 of IDH1 gene in 12% cases of GBM. In this study, IDH1/2 mutation was seen in 15.6% and expectedly IDH1R132H was the commonest.,, No case in this study showed IDH2172 mutations, re-emphasizing of its rarity in GBMs., The median age for mIDH in the present study was 33.5 years with age-range of 21–58 years, which was also concordant with the literature., The positive correlation between mIDH1 and p53 protein overexpression in this study and the literature re-emphasizes the association of TP53 and IDH mutations., 90% (18/20) of mIDH cases showed ATRX-loss; but only 64.3% (18/28) of ATRX-loss cases were mIDH i.e., not all mIDH GBM cases show loss of ATRX protein expresion and vice-versa. No case of >60 years age-group showed ATRX-loss and mIDH, suggesting the virtual non-occurrence of IDH-mutations and ATRX-alterations in >60 years.,
The reported frequency of EGFR-amplification in GBMs is approximately 30-50% with age-range of 25-85 years.,, However, in the current study 27% were EGFR-amplified with age-range of 26-74 years. The uncommoness of association of EGFR amplification with p53 over-expression, observed in the current study is concordant with the literature.,, As expected, mIDH and EGFR-amplification were mutually exclusive.,,
The reported frequencies of pTERT-mutation in de-novo and secondary GBMs was 80-86% and 16-28% respectively; C228T as the commonest type of mutation., In this study 43.5% of enitre study cohort of GBMs showed pTERT-mutation (with 72% of C228T type). 9.5% of mIDH and 54.3% of wtIDH (70% in EGFR-amplified group and 49% of wtIDH/EGFR non-amplified group) were pTERT-mutant respectively. Unlike Killela et al. and Labussiere et al. who reported median age of 59.6 years and 53.6 years for pTERT-mutated and pTERT wild-type groups respectively, the median age in the current study for pTERT-mutant cases was higher than the wild-type (55 and 44 year respectively), but was not statistically significant., However, the frequency of pTERT-mutation is higher (35/63;55.5%) in the >50 years as compared to 29% (15/52) in the =50 years age group; which is inverse to the mIDH. 54.3% (44/81) of wtIDH group showed pTERT-mutation as compared to 9.5% (2/21) mIDH group. Two cases of pTERT-mutant and mIDH were 1p19q intact on FISH, thus excluding the possibility of anaplastic oligodendroglioma. This findings do suggest the rare occurrence of pTERT-mutation in mIDH GBM, which is also reported in earlier studies.,,, On the contrary, 70% (21/30) of EGFR amplified cases showed pTERT-mutations, which is in concordance with the existing literature.,,
54.2% of pTERT wild-type were pMGMT-methylated as compared to 36.4% of pTERT-mutant group, though this association was statistically insignificant but similar pattern is also reported in the literature.
TP53-mutation is more frequent in the pTERT wild-type group (47%) as compared to pTERT-mutant group (30%). Though the present study is limited by lack of TP53 mutational evaluation, however p53 protein overexpression was commoner i.e., (63.5%) in the pTERT wild-type group as compared in pTERT-mutant group (i.e. 49%). pTERT-mutation and ATRX-alterations as expected were mutually exclusive, even in the current study. Interestingly, 49% and 13.2% of the cases in the study showed isolated (without mIDH and EGFR-amplification) pTERT-mutation and ATRX-loss respectively, these subsets probably represent a distinct biological group.
To conclude, this is a correlative study on pTERT-mutation with the currently known significant molecular parameters in adult GBMs. The frequency of pTERT-mutation in the present study was 43.5% (9.5% in mIDH, 70% in EGFR-amplified and 49% in wtIDH/EGFR non-amplified groups) and C228T mutation was commonest type. pTERT-mutation was more common with wtIDH cases, more frequent in EGFR-amplified group than non-amplified. Though, pMGMT methylation was more frequent in pTERT wild-type group but was statistically insignificant. Notedly, pMGMT-methylation was uncommon in wtIDH/EGFR non-amplified group as compared to mIDH and EGFR-amplified groups; while no frequency of pTERT-mutation was more common in EGFR-amplified than EGFR non-amplified/wtIDH group and it is rare in mIDH group; possibly suggesting pMGMT-methylation and pTERT-mutation as unrelated molecular prognostic parameters. This study also identifies a distinct subset of adult GBM with isolated ATRX-loss, which may need to be explored further.
Financial support and sponsorship
The study was funded by Tata Memorial Centre, intramural Institutional grant.
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]