Common Polymorphisms in the MDM2 and TP53 Genes and the Relationship between TP53 Mutations and Patient Outcomes in Glioblastomas

Abstract

MDM2 SNP309 is associated with younger age of tumor onset in patients with Li‐Fraumeni syndrome, and TP53 codon 72 polymorphism decreases its apoptotic potential. Glioblastomas frequently show genetic alterations in the TP53 pathway. In the present study, we assessed MDM2 SNP309 in 360 glioblastomas, and correlated these with patient age and survival, as well as other alterations in the TP53 pathway. Frequencies of the MDM2 SNP309 T/T, T/G and G/G genotypes in glioblastomas were 40%, 46% and 14%, respectively. Multivariate analysis showed that MDM2 SNP309 G/G allele was significantly associated with favorable outcome in female glioblastoma patients (hazard ratio 0.54; 95% CI = 0.32–0.92). There was a significant association between MDM2 SNP309 G alleles and TP53 codon 72 Pro/Pro in glioblastomas. Glioblastoma patients with TP53 codon 72 Pro/Pro genotype were significantly younger than Arg/Arg carriers (mean 50.2 vs. 56.1 years; P = 0.018). Multivariate analysis showed that those with TP53 codon 72 Arg/Pro allele had significantly shorter survival than those with Arg/Arg allele (hazard ratio 1.35; 95% CI = 1.07–1.71). Detailed analyses revealed that TP53 codon 72 Pro allele was significantly associated with shorter survival among patients with glioblastomas carrying a TP53 mutation, and among those treated with surgery plus radiotherapy.

INTRODUCTION

Glioblastoma is the most frequent and malignant human brain tumor (19). Despite progress in surgery, radio‐ and chemotherapy, survival of patients with glioblastoma remains very poor (43), in particular at the population level (29). The majority of glioblastomas are primary (de novo) glioblastomas that develop in older patients, with a short clinical history and without evidence of less malignant precursor lesions 27, 29). Primary glioblastomas show frequent LOH (loss of heterozygosity) 10q (70%), with other common genetic alterations such as EGFR (epidermal growth factor receptor) amplification, TP53 mutations, PTEN (phosphatase and tensin homolog) mutations, and p16^INK4a deletion (25%–36%) (29). MDM2 amplification is infrequent (5%–10%) but specific for primary glioblastomas 2, 28, 38, 39). Other glioblastomas occur in younger patients and develop slowly through progression from low‐grade or anaplastic astrocytomas (secondary glioblastomas). They show frequent LOH 10q (63%) and TP53 mutations (65%) 27, 29).

TP53 protein functions as a transcription factor, and is involved in cell‐cycle control, DNA repair, apoptosis, cell differentiation, cellular senescence and angiogenesis 50, 51). MDM2 is a key negative regulator of TP53 function (4). The TP53‐MDM2 auto‐regulatory feedback loop regulates activity of the TP53 protein and expression of MDM2; wild‐type TP53 protein induces transcription of the MDM2 gene and MDM2 protein inhibits the ability of TP53 to activate transcription from minimal promoter sequence 1, 25, 36). p14^ARF participates in the regulatory feedback loop with TP53 and MDM2, by binding directly to MDM2 and stabilizing both TP53 and MDM2. TP53 function may therefore be impaired not only by TP53 mutations but also by MDM2 amplification or over expression or by loss of p14^ARF due to homozygous deletion or promoter methylation (45).

There is increasing evidence that common polymorphisms in the transformation‐associated genes may significantly affect the development and prognosis of human neoplasms. A common polymorphism at codon 72 (Arg‐ > Pro) of the TP53 gene is located in a proline‐rich region of the p53 protein that is required for the growth suppression and apoptosis mediated by p53 41, 52). It decreases its apoptotic potential (7), influences the behavior of mutant p53 (22), and decreases DNA repair capacity (54). TP53 codon 72 Arg enhances the ability of mutant TP53 to bind its homolog p73 and neutralized p73‐induced apoptosis (22). TP53 codon 72 Pro/Pro induces apoptosis with lower kinetics and suppresses cellular transformation less efficiently than the Arg/Arg genotype 7, 46). TP53 codon 72 Pro allele may be linked with an increased risk of several human neoplasms 16, 54, 55).

A common polymorphism in the MDM2 promoter region, a T→G change at nucleotide 309 in the first intron (SNP309), increases the affinity of the promoter for the transcriptional activator Sp1, resulting in higher levels of MDM2 mRNA and MDM2 protein and subsequent down‐regulation of the p53 pathway (3). MDM2 SNP309 G alleles were significantly more frequent than in healthy individuals or in patients with endometrial cancer (53), nasopharyngeal carcinoma (59), lung cancer 20, 58), gastric cancer (30) or hepatocellular carcinomas (6), suggesting that SNP309 is associated with susceptibility to develop certain human neoplasms. MDM2 SNP309 G allele is also associated with younger age of tumor onset in Li‐Fraumeni syndrome patients carrying a TP53 germline mutation (40) as well as sporadic tumors, including colorectal cancer (24) and ovarian and peritoneal carcinomas (11). MDM2 SNP309 G allele is also associated with poorer outcome of patients with renal cell carcinoma (12) and advanced gastric cancer (30).

Little is known as to whether MDM2 SNP309 affects susceptibility to development of brain tumors or if it has prognostic value. Tsuiki et al (48) reported a lack of significantly increased G/G genotype in glioma patients, and lack of association between the SNP309 genotype and the histological grade of glioma, or age at disease onset. El Hallani et al (8) showed a lack of predictive value of MDM2 SNP309 in the outcome of glioblastoma patients. The objectives of the present study were to assess MDM2 SNP309 genotype in glioblastomas at a population level and to correlate it with patient outcome and alterations in other components of the TP53 pathway. We assessed MDM2 SNP309 in 360 glioblastomas, and correlated the results with patient age and survival, as well as other alterations in the TP53 pathway (TP53 mutations, TP53 codon 72 polymorphism, MDM2 amplification, p14^ARF homozygous deletion/promoter methylation).

MATERIALS AND METHODS

Glioblastoma samples

We examined 360 glioblastomas diagnosed in the Canton of Zurich, Switzerland between 1980 and 1994 (29). The mean age of patients was 56.4 ± 12.9 years, and the male : female ratio was 1.4. Tumors were considered primary (de novo) glioblastomas (339 cases) if a glioblastoma diagnosis was made at the first biopsy, without clinical or histologic evidence of presence of a less malignant precursor lesion. A diagnosis of secondary glioblastoma (21 cases) was made only in cases with histopathological evidence of preceding low‐grade or anaplastic glioma 27, 29). DNA was extracted from formalin‐fixed, paraffin‐embedded tissue sections as previously described (29). Results on TP53 mutations and distribution of TP53 codon 72 polymorphism have been reported previously (29).

MDM2 SNP309 genotyping

Allele‐specific PCR was carried out for MDM2 SNP309 genotyping. Primer sequences of MDM2 wild‐type and mutant alleles were previously reported (24). Two independent PCR reactions were carried out in a 10 µL volume, containing PCR buffer (20 mM Tris pH 8.4, 50 mM KCl), 1.5 mM MgCl₂, dNTPs (200 µM each), primers (0.4 µM each) and 0.5 U of platinum Taq DNA polymerase. PCR conditions were as follows: initial denaturing at 95°C for 5 minutes, followed by 35 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, extension at 72°C for 30 s and then a final extension at 72°C for 10 minutes. The PCR products were electrophoresed on 3% agarose gels containing ethidium bromide. To confirm the MDM2 SNP309 results obtained by allele‐specific PCR, we also performed direct sequencing of the MDM2 fragments for several samples, and showed that the results of allele‐specific PCR and direct sequencing were consistent.

Differential PCR for MDM2 amplification

Differential PCR to detect MDM2 amplification was carried out as described previously with some modifications (2). Briefly, PCR was carried out in a 10 µl volume, containing PCR buffer (20 mM Tris pH 8.4, 50 mM KCl), 1.5 mM MgCl₂, dNTPs (200 µM each), 0.4 µM primers for MDM2, 0.6 µM primers for dopamine receptor (DR) reference sequence, and 0.5 U of platinum Taq DNA polymerase. After denaturing DNA at 95°C for 5 minutes, 30 cycles of PCR (95°C for 30 s, 55°C for 30 s, and 72°C for 45 s) were carried out with a final extension at 72°C for 5 minutes with T3000 thermocycler (Biometra, Archamps, France). The primer sequences for differential PCR were as follows: 5′‐GAG GGC TTT GAT GTT CCT GA‐3′ (forward) and 5′‐GCT ACT AGA AGT TGA TGG C‐3′ (reverse) for MDM2 and 5′‐CCA CTG AAT CTG TCC TGG TAT G‐3′ (forward) and 5′‐GTG TGG CAT AGT AGT TGT AGT GG‐3′ (reverse) for DR. The PCR product was analyzed on an 8% polyacrylamide gel and stained with ethidium bromide. Densitometry of the PCR fragments was performed using ImageQuant Version 5.0 software (Molecular Dynamics, Inc., Sunnyvale, CA). The MDM2/DR ratio from normal blood DNA was 0.9, with a standard variation of 0.23. A value of more than 2.5 (2x mean + 3x SD) for the MDM2/DR ratio was regarded as positive for MDM2 amplification (2).

Differential PCR for p14^ARF homozygous deletion

To assess homozygous deletion of p14^ARF, we performed differential PCR with primers covering exon 1β of the p14^ARF gene, using the GAPDH (glyceraldehyde‐3‐phosphate dehydrogenase) sequence as a reference. PCR conditions and the primer sequence were as previously described with some modifications (26). PCR reactions were carried out in a 10 µl volume, containing PCR buffer (20 mM Tris pH 8.4, 50 mM KCl), 1.5 mM MgCl₂, dNTPs (200 µM each), 0.2 µM primers for p14^ARF, 0.6 µM primers for GAPDH, and 0.5 U of platinum Taq DNA polymerase. PCR condition was as follows: 95°C for 5 minutes as initial denaturation, followed by 30 cycles of denaturation at 95°C for 1 minute, annealing at 60°C for 1 minute, extension at 72°C for 1 minute, with a final extension at 72°C for 5 minutes. The primer sequences were as follows: 5′‐GAG TGA GGG TTT TCG TGG TT‐3′ (forward) and 5′‐GCC TTT CCT ACC TGG TCT TC‐3′ (reverse) for p14^ARF and 5′‐AAC GTG TCA GTG GTG GAC CTG‐3′ (forward) and 5′‐AGT GGG TGT CGC TGT TGA AGT‐3′ (reverse) for GAPDH. The PCR product was analyzed on an 8% polyacrylamide gel and stained with ethidium bromide. Densitometry of the PCR fragments was performed using ImageQuant Version 5.0 software. Samples presenting <20% of the control signals (normal DNA samples extracted from paraffin sections) were considered p14^ARF homozygous deletion, as previously described 32, 56).

Methylation‐specific PCR for p14^ARF promoter methylation

Sodium bisulfite modification was performed using EZ DNA Methylation Kit^TM (Zymo Research, Orange, CA). Approximately 200 ng of DNA was denatured with dilution buffer for 15 minutes at 37°C, incubated with CT conversion reagent at 50°C for 16 h and cleaned up and desulfonated by using columns. Methylation‐specific PCR was carried out in a 10 µl volume, containing PCR buffer (20 mM Tris pH 8.4, 50 mM KCl), 2 mM MgCl₂, dNTPs (250 µM each), primers (0.5 µM each), and 0.5 U of platinum Taq DNA polymerase. CpGenome^TM Universal methylated DNA (Chemicon International Inc. Temecula, CA) was used as a methylation‐positive control for the methylated p14^ARF promoter, and DNA sample from peripheral blood leukocytes from a healthy individual was used as a control for unmethylated p14^ARF promoter. Universal methylated DNA and control for unmethylated p14^ARF promoter were sodium‐bisulfite modified as described above. PCR amplification was carried out in the T3000® thermocycler (Biometra, Archamps, France) with initial denaturing at 95°C for 10 minutes followed by 40 cycles of denaturing at 95°C for 45 s, annealing at 62°C (for methylated reaction) or 60°C (for unmethylated reaction) for 45 s and extension at 72°C for 45 s, which was followed by final extension at 72°C for 10 minutes. Template‐free distilled water was included as a negative control for the PCR. Primer sequences for methylated and unmethylated PCR were previously reported by Esteller et al (9). PCR products were electrophoresed on 4% agarose gels containing ethidium bromide.

Statistical analyses

The Mann‐Whitney non‐parametric test was used to assess differences in frequency of genetic alterations between primary and secondary glioblastomas. Student's t‐test was used to analyze differences in mean ages between different polymorphisms and alterations in the TP53 pathway. Kaplan‐Meier's methods and the log‐rank test were performed to compare survival of patients with different genetic alterations and polymorphisms. Fisher's exact test was used to assess the relationship between different genetic alterations. All of the statistical analysis was performed with Stat‐View for Windows 5.01 software® (SAS Institute Inc., Cary, NC). Multivariate Cox regression models were used to assess the relationship between polymorphisms and patients' survival. Adjustment was made for age (<50, 50–59, 60–69, ≥70 years), gender and treatment (surgery alone, surgery plus radiotherapy). Logistic regression models were used to analyze the association between polymorphisms of genes. Statistical significance was set at a P value of <0.05.

RESULTS

MDM2 SNP309 genotype

Frequencies of the T/T, T/G, and G/G genotypes of MDM2 SNP309 in glioblastomas were 40%, 46% and 14%, respectively (Table 1). These frequencies were similar to those previously reported in healthy Caucasian individuals 3, 8, 24). There was no significant difference in these frequencies between primary and secondary glioblastomas. The mean age of glioblastoma diagnosis was 56.8 years for T/T, 55.6 years for T/G and 57.9 years for G/G MDM2 SNP alleles (P = 0.265).

Table 1.

MDM2 SNP309 and TP53 codon 72 polymorphisms in glioblastomas.

Polymorphism	No. of cases	Frequency	Mean age at glioblastoma diagnosis	Mean survival (months)	Hazard Ratio** (95% CI)
MDM2 SNP309
T/T	143	40%	56.8 ± 12.0 years	9.9	1
T/G or G/G	217	60%	56.1 ± 13.5 years	9.3	0.98 (0.78–1.23)
T/G	166	46%	55.6 ± 13.5 years	9.1	1.01 (0.79–1.28)
G/G	51	14%	57.9 ± 13.3 years	9.8	0.91 (0.65–1.26)
TP53 codon 72
Arg/Arg	188	58%	56.1 ± 11.0 years	10.9	1
Arg/Pro or Pro/Pro	136	42%	54.4 ± 14.6 years	8.9	1.27 (1.02–1.58)
Arg/Pro	111	34%	55.4 ± 14.4 years	8.5	1.35 (1.07–1.71)
Pro/Pro	25	8%	50.2 ± 15.3 years*	10.9	1.00 (0.66–1.52)

^* Significantly younger than patients with TP53 codon 72 Arg/Arg (P = 0.018).

^** Adjusted for age, gender and treatment.

Female patients with MDM2 SNP309 T/G or G/G alleles tended to show more favorable outcome (median survival 7.9 months) than those with T/T allele (6.7 months; log‐rank test, P = 0.0610). Multivariate analyses adjusted for age and treatment showed that female patients with MDM2 SNP309 G/G had a significantly longer survival than those with T/T (hazard ratio 0.54; 95% CI = 0.32–0.92; Table 2). There was no such tendency in male patients (Table 2).

Table 2.

Distribution and hazard ratios of glioblastoma patients according to MDM2 SNP309 andTP53 codon 72 polymorphisms.

	Gender				Treatment
	Males		Females		Surgery alone		Surgery plus radiotherapy
	N	Hazard Ratio* (95% CI)	N	Hazard Ratio* (95% CI)	N	Hazard Ratio** (95% CI)	N	Hazard Ratio** (95% CI)
	MDM2 SNP309
TT	96	1	47	1	32	1	91	1
T/G or G/G	117	1.09 (0.82–1.46)	100	0.73 (0.50–1.06)	88	1.09 (0.71–1.66)	119	1.00 (0.74–1.34)
T/G	90	1.05 (0.77–1.44)	76	0.81 (0.55–1.21)	67	1.09 (0.70–1.71)	94	1.06 (0.78–1.45)
G/G	27	1.22 (0.79–1.90)	24	0.54 (0.32–0.92)	21	1.08 (0.60–1.91)	25	0.82 (0.51–1.30)
TP53 codon 72
Arg/Arg	116	1	72	1	69	1	119	1
Arg/Pro or Pro/Pro	78	1.23 (0.92–1.64)	58	1.31 (0.91–1.87)	49	1.23 (0.84–1.80)	86	1.33 (1.01–1.77)
Arg/Pro	64	1.31 (0.96–1.79)	47	1.38 (0.94–2.02)	39	1.27 (0.85–1.90)	71	1.48 (1.09–2.00)
Pro/Pro	14	0.94 (0.53–1.66)	11	1.04 (0.52–2.08)	10	1.07 (0.52–2.20)	15	0.92 (0.53–1.60)

^* Adjusted for age and treatment.

^** Adjusted for age and gender.

N = numbers of cases.

MDM2 amplification

MDM2 amplification was found in 17 of 317 (5%) primary glioblastomas, and these cases lacked TP53 mutations. All cases with MDM2 amplification had T/G (12 cases) or G/G (5 cases) genotypes of MDM2 SNP309. There was a significant positive association between MDM2 amplification and T/G or G/G genotype (P = 0.0008). No MDM2 amplification was found in any of the secondary glioblastomas.

TP53 codon 72 polymorphism

Frequencies of the Arg/Arg, Arg/Pro and Pro/Pro genotypes of TP53 codon 72 in glioblastomas were 58%, 34% and 8%, respectively (Table 1), which were similar to those in healthy Caucasian individuals 42, 47). There was no significant difference in these frequencies between primary and secondary glioblastomas. Patients with the Pro/Pro genotype were significantly younger (mean, 50.2 years) than Arg/Arg carriers (56.1 years; P = 0.018; Table 1). TP53 mutations were equally common for either codon 72 allele (data not shown).

Glioblastoma patients with TP53 codon 72 Arg/Pro alleles had significantly shorter survival (median 7.3 months) than those with Arg/Arg alleles (8.6 month; log‐rank test, P = 0.0111). Multivariate analysis after adjustment for age, gender, and treatment confirmed this finding (hazard ratio 1.35; 95% CI = 1.07–1.71; Table 1).

Among patients treated with surgery plus radiotherapy, TP53 codon 72 Arg/Pro alleles were associated with shorter survival (median, 10 months) than Arg/Arg alleles (11.4 months; log‐rank test, P = 0.0101). Multivariate analyses after adjustment for age and gender showed that TP53 codon 72 Arg/Pro alleles were associated with significantly shorter survival in patients treated with surgery plus radiotherapy, but not in those treated with surgery alone (Table 2).

Univariate and multivariate analyses revealed TP53 codon 72 Arg/Pro or Pro/Pro alleles were significantly associated with shorter survival of patients with glioblastomas carrying a TP53 mutation, but not in glioblastomas with wild‐type TP53 (Figure 1; Table 3).

Figure 1.

TP53 codon 72 Arg/Pro or Pro/Pro alleles were significantly associated with shorter survival of patients with glioblastomas carrying a TP53 mutation (A), but not in glioblastomas with wild‐type TP53 (B).

Table 3.

Multivariate analysis for the effect of TP53 codon 72 polymorphism on survival of glioblastoma patients.

	TP53 Wild‐type		TP53 Mutated
	No. of cases	Hazard ratio* (95% CI)	No. of cases	Hazard ratio* (95% CI)
	TP53 codon 72
Arg/Arg	130	1	58	1
Arg/Pro, Pro/Pro	93	1.06 (0.81–1.39)	43	1.66 (1.07–2.56)

^* This model is adjusted for age, gender and treatment.

MDM2 SNP309 and TP53 codon 72 polymorphism

There was a significant positive association between TP53 codon 72 Pro/Pro genotype and MDM2 SNP309 T/G or G/G genotypes, after adjustment for gender and age (P = 0.015). However, there were no significant synergistic effects of MDM2 SNP G and TP53 codon 72 Pro alleles in survival of patients in multivariate analysis (data not shown).

p14^ARF alterations

Homozygous deletion and promoter methylation of the p14^ARF gene were detected in 22% and 11% (total 33%) of glioblastomas, respectively. There was no significant difference in the frequency of p14^ARF alterations between primary and secondary glioblastomas (35% vs. 14%, P = 0.1171).

Alterations in the TP53/MDM2/P14^ARF pathway

There was a significant inverse correlation between MDM2 amplification and TP53 mutation (P = 0.0056) and between p14^ARF alterations and TP53 mutations (P = 0.012). More than half of the glioblastomas (184/338; 54%) showed at least one alteration in the TP53 pathway (TP53 mutations, p14^ARF homozygous deletion, p14^ARF promoter methylation, MDM2 amplification). There were no significant differences in mean age at diagnosis (54.8 ± 10.9 vs. 57.3 ± 14.3 years; P = 0.084) or survival (median 7.6 vs. 7.9 months; P = 0.477) of glioblastoma patients with and without alterations in at least one of alterations in the TP53 pathway. There was also no significant difference in the frequency of overall alterations in the TP53 pathway between primary and secondary glioblastomas (53% vs. 71%, P = 0.164).

DISCUSSION

MDM2 SNP309 G/G genotype is associated with over expression of MDM2, which is recognized by immunohistochemistry in gliomas (48). In esophageal tissue, MDM2 G/G genotype carriers had significantly higher MDM2 mRNA level than the T/T genotype carriers (14). Changes in the levels of MDM2 have significant effects on cellular biology. Transgenic mice with a hypomorphic allele of MDM2 that results in as little as a 20% reduction in MDM2 protein levels were markedly more sensitive to ionizing radiation (23).

In the present study, we show that MDM2 SNP309 G allele was significantly associated with favorable outcome in female patients, but not in male patients. The molecular basis of this observation remains to be elucidated, but may be related at least in part to the fact that MDM2 expression is regulated by several hormones such as estrogen, through a region of the MDM2 promoter containing SNP309 18, 31, 35, 37). Since the G allele of SNP309 increases the affinity of the MDM2 promoter for Sp1, which is a well‐characterized cotranscriptional activator for multiple hormone receptors, including the estrogen receptor, the SNP309 locus may alter the effects of hormones such as estrogen on tumorigenesis, and therefore contribute to the gender differences in human neoplasms 17, 34, 44).

In the present study, all of the glioblastomas with MDM2 amplification had either T/G or G/G genotype, and there was a significant positive correlation between SNP309 T/G or G/G genotype and MDM2 amplification. The molecular mechanisms of preferential MDM2 amplification in glioblastomas with SNP309 G alleles remain to be elucidated.

A polymorphism at codon 72 (Arg‐ > Pro) of the TP53 gene may be linked with increased risk of several human neoplasms such as cancer of the lung (54), prostate (55) and breast (16). Pro alleles were also associated with shorter survival of patients with ovarian and peritoneal carcinomas (11), breast cancer (47), and malignant peripheral nerve sheath tumors (13). van Heemst et al (49) carried out a meta‐analysis of 65 published studies and reported that carriers of TP53 codon 72 Pro/Pro genotype have increased cancer risk compared to Arg/Arg carriers.

Little is known about the role of the TP53 codon 72 polymorphism on brain tumor risk and prognosis of brain tumor patients. In a study of 135 brain tumors, Parhar et al (33) reported a significant increase in the Arg/Pro alleles among high‐grade astrocytomas compared with non‐astrocytomas. In a population‐based case–control study in Sweden (205 gliomas, 164 meningiomas and 374 controls), Malmer et al (21) showed the absence of overall associations for codon 72 polymorphism in gliomas and meningiomas. In the present study, patients with TP53 codon 72 Pro/Pro genotype were significantly younger (mean, 50.2 years) than Arg/Arg carriers (56.1 years; P = 0.018), suggesting that the Pro/Pro genotype may be associated with early onset of development of glioblastomas.

In the present study, glioblastoma patients with TP53 codon 72 Arg/Pro alleles showed significantly shorter survival than those with Arg/Arg alleles. Detailed analyses revealed that this was true in patients treated with surgery plus radiotherapy, but not in those treated with surgery alone, suggesting that this common TP53 polymorphism may affect the sensitivity to radiotherapy. There was no such association in homozygous Pro/Pro alleles. This is likely explained by the very small number of cases with Pro/Pro allele, which may limit statistical power. Many previous studies therefore also combined Arg/Pro and Pro/Pro alleles for statistical analyses 10, 15, 57).

We showed that TP53 codon 72 Arg/Pro and Pro/Pro alleles were significantly associated with shorter survival in patients with glioblastomas carrying a TP53 mutation, but not in glioblastomas with wild‐type TP53. The molecular basis of this observation remains to be elucidated. However, it may be at least in part due to the fact that presence of TP53 mutations leads to more DNA damage in neoplastic cells (51), and that TP53 codon 72 Pro/Pro genotype induces apoptosis less efficiently than Arg/Arg alleles 7, 46).

It has been reported that gene‐gene interaction of MDM2 SNP309 G/G and TP53 codon 72 Pro/Pro genotypes increases lung cancer risk in a supermultiplicative manner (58). Bougeard et al (5) reported that both MDM2 SNP309 G and TP53 codon 72 Arg alleles were significantly associated with younger age of tumor onset in French carriers of the TP53 germline mutation, and there was a cumulative effect of MDM2 SNP309 and TP53 codon 72 polymorphisms on the age of tumor onset. The present study showed a significant association between MDM2 SNP309 G and TP53 codon 72 Pro alleles in glioblastomas, but there was no significant cumulative effect on survival of glioblastoma patients.

In the present study, 54% of glioblastomas showed alterations in one or more components of the TP53 pathway (TP53 mutation, MDM2 amplification, p14^ARF homozygous deletion/promoter methylation). Overall alterations in the TP53 pathway were more frequent in secondary glioblastomas (71%) than in primary glioblastomas (53%), but the difference was not statistically significant. Thus, although TP53 mutations are typical for secondary glioblastomas, disruption of the TP53 pathway is an important and frequent genetic event in the development of both glioblastoma subtypes. Supporting the view that TP53, MDM2 and p14^ARF genes function in the same signaling pathway (45), the present study noted significant inverse correlations between TP53 mutations and MDM2 amplification, and between TP53 mutations and p14^ARF alterations.