Sage Journals: Discover world-class research

Abstract

Purpose

Adequate biomarkers are still required to optimize therapy in patients with locally advanced head and neck squamous carcinomas (HNSCC) treated with chemoradiotherapy (CRT).

Methods

We updated the follow-up of 66 HNSCC patients treated with CRT we described more than 10 years ago, focusing on SNP Arg/Pro (R/P) at codon 72 and somatic mutations in TP53 and on SNP309 in the MDM2 gene.

Results

In wild-type TP53 cases, overall survival (OS) was longer in 72RR and less favorable in 72PP (p = 0.005); when TP53 was mutated, OS was longest in 72PP and less favorable in 72RR and 72RP (p = 0.058). Median OS was significantly shorter in patients with MDM2 SNP309 GG or GT genotypes compared with the TT genotype (p = 0.002).

Conclusions

TP53 SNP72 may be useful in selecting patients for CRT, but has to be related to somatic TP53 mutations. The MDM2 SNP309, easily determined in peripheral blood, might be more convenient as a predictive biomarker.

Keywords

Head and neck cancer TP53 MDM2 Chemoradiation Predictive markers

Introduction

The management of head and neck squamous cell carcinoma (HNSCC) remains an important clinical challenge. HNSCC is the sixth most common cancer in the world and accounts for 4% of all malignant neoplasms in Italy (1). Although early-stage disease is curable with surgery or radiation therapy, more than 60% of patients present with locally or regionally advanced disease (stage III-IV) (1). For such patients the treatment failure rate after surgery and/or radiotherapy is high and the survival poor.

Despite improved local control and overall survival (OS), optimal protocols for the administration of combined-modality treatments (combining radiation and chemotherapy) have not been definitively identified in HNSCC, nor has it been established which subgroup of patients with advanced HNSCC will benefit from integrated treatments (2-3-4). Rapidly alternating chemoradiation is a variation of concurrent chemoradiation that is intended to minimize toxicity (5). It has led to a statistically significant improvement in outcome compared to conventional radiotherapy (6), with a more favorable toxicity profile (7, 8).

Because of the significant treatment-related toxicity associated with combined-modality regimens, it is critical to link treatments to molecular predictors that maximize the therapeutic antitumor effect and minimize toxicity.

Tobacco use and alcohol abuse remain the best established risk factors, but accumulating evidence indicates that viral infections, such as HPV and EBV infections, are associated with squamous cell lesions of the oropharynx and nasopharynx, respectively (9, 10). In particular, it is accepted worldwide that HPV-related oropharynx tumors depicts a distinct molecular, clinical and pathologic disease entity (11).

In earlier work we have shown that a functional single-nucleotide polymorphism (SNP) in the apoptosis-signaling domain of TP53 (R/P at codon 72) influences the clinical response of HNSCC to chemotherapy in cases with mutant and wild-type (wt) TP53 (12, 13).

In squamous cell carcinomas as well as other solid tumors, mutations occurring in the arginine allele (72R) are detected more commonly than those in the proline allele (72P) (12, 14-15-16-17). Many studies have also shown that 72R is more active than 72P in inducing apoptosis (18). Due to its effects as a modulator of TP53 mutant function, it is an interesting possibility that treatment outcome in human cancer is influenced by this polymorphism (16). In particular, we have already demonstrated that clinical response following cisplatin-based chemoradiotherapy (CRT) for advanced HNSCC is influenced by this polymorphism: cancers expressing 72R mutants have lower response rates than those expressing 72P mutants (13).

Furthermore, the stability of the TP53 protein is regulated by the E3 ubiquitin ligase MDM2 protein (mouse double minute 2 homolog). The SNP309 T>G in the promoter region of this gene influences the protein levels (19).

In a more recent study, we showed that the MDM2 SNP309 is indeed highly predictive of outcome, irrespective of mutant or wt TP53 status, and retains independent predictive power compared with recognized clinicopathological variables (20).

In the present study, we investigated the correlation of somatic mutations and polymorphisms in the TP53 and MDM2 genes with clinical outcome in platinum-treated locally advanced HNSCC patients with more than 10 years’ follow-up, partially described in previous articles (12, 13).

Material and methods

Patients and Tumors

We analyzed 66 cases of patients with locally advanced, inoperable, stage III/IV HNSCC who had been referred to the Medical Oncology Department of Cuneo Teaching Hospital between May 1997 and May 2002. The clinical-pathological and molecular characteristics of the patient population had been partially described previously (12, 13, 20) and are summarized in Table I. All patients were heavy smokers (median 45 pack/year) and were homogeneously treated with a combined approach consisting of alternating CRT, as detailed by Merlano (5).

Table I

Clinicopathological and molecular characteristics of head and neck cancers

Sex	Age	Primary	Subsite	T	N	Stage	TP53 SNP72	TP53 gene	DSR/NDSR	Mutant allele	MDM2 SNP	EGFR	Response	OS	PFS	P/NP	A/D
Oropharynx
M	68	Oropharynx		2	2c	IV	Arg/Arg	179 Cat>Tat	DSR	Arg	wt	3+	CR	160	130	NP	D
M	36	Oropharynx	Soft palate	3	2c	IV	Arg/Arg	133 Atg>Ctg	NDSR	Arg	wt	0	CR	187	163	NP	A
M	70	Oropharynx	Tonsil	4	2b	IV	Arg/Arg	na	na	na	wt	3+	CR	83	83	NP	D
M	67	Oropharynx	Base of tongue	4	2b	IV	Arg/Pro	wt	-	None	wt	3+	CR	86	57	P	D
F	68	Oropharynx	Tonsil	4	0	IV	Arg/Pro	wt	-	None	wt	2+	CR	165	154	NP	A
M	62	Oropharynx	Base of tongue	3	2c	IV	Arg/Pro	wt	-	None	HET	2+	CR	13	13	P	D
M	59	Oropharynx		4	0	IV	Arg/Pro	248 Cgg>Tgg	DSR	Pro	GG	3+	CR	200	187	P	A
F	76	Oropharynx	Tonsil	4	2c	IV	Arg/Pro	176 tGc>tAc	NDSR	Arg	GG	0	CR	98	79	NP	D
M	59	Oropharynx	Tonsil	4	2b	IV	Arg/Pro	206 tGg>tAg (STOP)	DSR	Arg	GG	2+	CR	73	58	P	D
M	54	Oropharynx	Tonsil	4	1	IV	Arg/Pro	245 Ggc>Agc	NDSR	Arg	GG	2+	CR	14	10	P	D
M	57	Oropharynx	Tonsil	4	2b	IV	Arg/Pro	163 Tac>Cac	NDSR	Pro	GG	3+	CR	19	19	NP	D
M	44	Oropharynx	Base of tongue	4	2a	IV	Arg/Arg	wt	-	None	GG	2+	PR	72	72	NP	D
M	56	Oropharynx	Base of tongue	4	3	IV	Pro/Pro	wt	-	None	GG	2+	CR	7	7	P	D
M	56	Oropharynx	Tonsil	4	1	IV	Arg/Arg	na	na	na	na	3+	CR	154	128	P	A
M	66	Oropharynx	Base of tongue	2	2c	IV	Arg/Arg	wt	-	None	na	2+	CR	20	20	NP	D
M	55	Oropharynx	Base of tongue	4	2c	IV	Arg/Arg	245 gGc>gAc	DSR	Arg	na	0	CR	50	31	P	D
Non-oropharynx
M	55	Hypopharynx	Pyriform sinus	2	3	IV	Pro/Pro	220 tAt>tGt	NDSR	Pro	wt	3+	CR	167	133	NP	A
M	49	Hypopharynx		1	3	IV	Arg/Arg	275 tGt>tAt	NDSR	Arg	wt	0	CR	64	64	NP	D
M	76	Hypopharynx	Pyriform sinus	2	2c	IV	Arg/Arg	wt	-	None	wt	0	CR	76	63	NP	D
M	41	Hypopharynx	Pyriform sinus	3	2b	IV	Arg/Arg	wt	-	None	wt	0	CR	180	180	NP	A
M	67	Hypopharynx	Pyriform sinus	3	2b	IV	Arg/Arg	wt	-	None	wt	0	CR	30	29	P	D
M	47	Hypopharynx	Pyriform sinus	3	1	III	Arg/Arg	wt	-	None	wt	0	PR	155	30	P	A
M	50	Hypopharynx	Pyriform sinus	4	2b	IV	Arg/Pro	220 tAt>tGt	NDSR	Arg	wt	2+	CR	97	69	NP	D
M	55	Hypopharynx	Pyriform sinus	3	0	III	Arg/Arg	175 cGc>cAc	NDSR	Arg	wt	2+	CR	169	137	NP	D
M	40	Hypopharynx	Pyriform sinus	4	0	IV	Arg/Arg	220 tAt>tGt	NDSR	Arg	HET	2+	PR	7	5	P	D
M	49	Hypopharynx		4	2c	IV	Arg/Arg	na	na	na	HET	0	CR	126	52	P	D
M	67	Hypopharynx		3	0	III	Arg/Pro	wt	-	None	HET	3+	CR	65	47	NP	D
M	65	Hypopharynx		3	2c	IV	Arg/Pro	179 cAt>cGt	NDSR	Arg	HET	3+	PR	19	12	P	D
F	72	Hypopharynx	Pyriform sinus	4	0	IV	Arg/Pro	wt	-	None	HET	0	CR	20	20	P	D
M	57	Hypopharynx		4	3	IV	Arg/Arg	179 Cat>Tat	DSR	Arg	HET	0	PR	17	10	P	D
M	59	Hypopharynx	Pyriform sinus	3	1	III	Arg/Pro	158 Cgc>Ggc	NDSR	Arg	HET	2+	CR	15	15	NP	D
M	60	Hypopharynx	Pyriform sinus	4	2b	IV	Arg/Arg	wt	-	None	HET	3+	CR	184	133	NP	D
M	59	Hypopharynx	Pyriform sinus	4	2b	IV	Arg/Arg	Frameshift	DSR	Arg	HET	3+	CR	36	16	P	D
M	50	Hypopharynx		4	3	IV	Arg/Pro	wt	-	None	GG	3+	CR	140	98	NP	D
M	41	Hypopharynx	Pyriform sinus	2	2c	IV	Pro/Pro	175 cGc>cAc	NDSR	Pro	GG	3+	CR	151	121	NP	A
M	58	Hypopharynx	Pyriform sinus	4	2c	IV	Arg/Pro	276 Gcc>Acc	NDSR	None	GG	0	CR	34	34	NP	D
M	56	Hypopharynx		2	3	IV	Arg/Arg	245 Ggc>Agc	NDSR	Arg	GG	3+	CR	27	16	P	D
M	57	Hypopharynx	Pyriform sinus	4	2a	IV	Arg/Pro	280 aGa>aAa	NDSR	Arg	na	3+	PR	15	11	P	D
M	62	Hypopharynx	Pyriform sinus	4	2b	IV	Arg/Arg	175 cGc>cAc	NDSR	Arg	na	2+	CR	8	8	NP	D
M	74	Hypopharynx	Pyriform sinus	3	0	III	Arg/Pro	wt	-	None	na	2+	CR	80	77	P	D
M	59	Larynx		2	2c	IV	Arg/Arg	wt	-	None	wt	3+	CR	80	72	P	D
M	75	Larynx	Epiglottis	2	3	IV	Arg/Arg	249 agG>agT	DSR	Arg	wt	3+	PD	1	1	P	D
M	66	Larynx		3	2b	IV	Arg/Pro	172 gTt>gCt	NDSR	Arg	HET	2+	CR	45	31	P	D
M	53	Larynx		4	2	IV	Arg/Arg	wt	-	None	HET	0	CR	133	101	NP	D
M	54	Larynx		0	3	IV	Arg/Pro	110 cGt>cAt	NDSR	Pro	HET	3+	NE	150	150	NP	A
M	67	Larynx		4	0	IV	Arg/Arg	wt	-	None	HET	0	TD	1	1	NP	D
M	59	Larynx		4	0	IV	Arg/Arg	175 cGc>cAc	NDSR	Arg	GG	3+	NE	14	14	P	D
M	65	Larynx	Vocal cord	3	1	III	Arg/Pro	wt	-	None	GG	3+	CR	18	15	P	D
F	59	Larynx	Vocal cord	4	0	IV	Arg/Arg	wt	-	None	GG	3+	CR	32	31	NP	D
F	66	Larynx	Trachea	4	1	IV	Arg/Pro	wt	-	None	GG	2+	CR	21	14	P	D
M	68	Larynx	Glottis	3	1	III	Arg/Arg	na	na	na	na	3+	CR	28	28	NP	D
M	46	Larynx	Glottis	4	2c	IV	Arg/Pro	245 gGc>gAc	DSR	Pro	na	0	CR	54	54	P	D
M	54	Larynx	Vocal cord	3	0	III	Arg/Arg	176 tGc>tTc	DSR	Arg	na	0	PR	148	15	P	A
M	61	Larynx	Epiglottis	3	0	III	Arg/Pro	266 gGa>gTa	NDSR	Arg	na	2+	CR	10	10	NP	D
M	47	Oral cavity	Oral floor	4	2b	IV	Arg/Pro	241 tCc>tTc	DSR	Pro	wt	2+	CR	156	124	NP	D
M	66	Oral cavity	Soft palate	3	2c	IV	Arg/Pro	Frameshift	DSR	Arg	HET	0	CR	55	51	NP	D
F	55	Oral cavity	Tongue	4	0	IV	Arg/Pro	175 cGc>cAc	NDSR	Arg	HET	2+	PR	8	4	P	D
M	70	Oral cavity	Tongue	2	2c	IV	Arg/Pro	248 cGg>cAg	DSR	Arg	HET	3+	TD	2	2	NP	D
M	65	Oral cavity	Oral floor	4	0	IV	Arg/Arg	wt	-	None	na	0	CR	60	52	P	D
M	60	Oral cavity	Oral floor	4	2a	IV	Arg/Pro	273 Cgt>Tgt	NDSR	Arg	na	3+	CR	36	25	P	D
F	62	Rhinopharynx		4	2a	IV	Arg/Arg	wt	-	None	wt	0	CR	92	87	P	D
M	51	Rhinopharynx		x	2b	IV	Pro/Pro	179 Cat>Tat	DSR	Pro	wt	2+	CR	202	202	NP	A
F	30	Rhinopharynx		4	0	IV	Arg/Arg	273 cGt>cAt	NDSR	Arg	wt	0	NE	39	39	NP	A
M	54	Rhinopharynx		4	3	IV	Arg/Arg	wt	-	None	wt	3+	PR	21	17	P	D
M	37	Rhinopharynx		4	1	IV	Arg/Arg	173 Gtg>Ctg	NDSR	Arg	HET	0	PD	21	12	P	D
M	37	Rhinopharynx		4	2	IV	Arg/Arg	na	na	na	HET	3+	PD	39	20	P	D

T = tumor size; N = nodal status; OS = overall survival; PFS = progression-free survival; P = progression; NP = no progression; A = alive; D = dead; M = male; F = female; DSR = disruptive; NDSR = nondisruptive; CR = complete response; PR = partial response; TD = death due to toxicity; PD = progressive disease; NE = not evaluable; na = not available; wt = wild type; HET = Heterozygous.

This work was carried out in accordance with the code of ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. Informed consent for tissue collection and use for scientific purposes was obtained from each patient enrolled in the study.

Analysis of TP53 and MDM2

TP53 mutations and SNP at codon 72 (rs1042522) in 61/66 tumor DNAs extracted from formalin-fixed paraffin-embedded (FFPE) tissue were sought by direct sequencing in the most frequently affected exons 4-10, which were individually amplified as already described (12, 20). Amplification in 5 patients failed for some exons, while SNP genotyping at codon 72 was obtained for all 66 patients.

In germline 72 RP, we identified the polymorphic allele in which the mutation occurred, in order to divide our population in wt or mutant alleles (mutant-72R or mutant-72P) (13).

The mutations were evaluated for possible functional importance at the protein level using available TP53 databases (http://p53.iarc.fr/ and http://p53.free.fr) and the published literature. In particular missense, frameshift and nonsense mutations, together with alterations affecting the splice-site regions were considered as functional mutations. Silent mutations not affecting the amino-acid sequence or located outside the exons and splice-site regions were considered not functional mutations and were allocated to the wt group.

In MDM2, the SNP at position 309 (rs2279744) was analyzed by direct sequencing after PCR amplification of genomic DNA for the 54 patients in whom blood was available for this purpose, as already described (12, 13, 20).

Egfr Expression

EGFR was detected in all patients by immunohistochemistry (IHC) using the EGFR pharmDx™ kit (Dako) and quantified according to FDA guidelines on a scale from 0 to 3+.

HPV16 and p16 Expression

Due to the lack of adequate tissue slides of the oldest cases, HPV16 and p16 were investigated in half of the patients (33/66: 50%) by PCR and IHC, respectively.

HPV was searched using specific primer pairs for the E1 gene of HPV type 16 (NCBI Ref. Seq: NC_001526.2, clone 114/K): E1 forward 5’-GAGATGCAGTACAGGTTCTAAAACG-3’ and E1 reverse 5’-TGCCATACCCGCTGTCTTC-3.

IHC staining for p16 was performed on 4-micron sections from FFPE tissue blocks mounted on glass slides, using prediluted ready-to-use monoclonal antibodies: clone E6H4 (MTM Laboratories) on a Ventana Benchmark ULTRA automated immunostainer. Positivity for p16 was defined as ≥70% positive cells.

Statistical Analysis

OS and progression-free survival (PFS) curves, according to TP53, MDM2 and EGFR status, were calculated using the Kaplan-Meier life-table method; the log-rank test was used to evaluate the statistical significance. OS was defined as the time from treatment start to death from any cause. Surviving patients were withdrawn at the last follow-up or at the last contact. PFS analyses were based on the time from treatment start to the first event (locoregional recurrence, distant metastasis); patients without an event were censored at their last follow-up.

The comparison was done between 2 groups using the chi-square test, with Yates’ correction or Fisher's exact test when appropriate. Fisher's 2-tailed exact test was used to calculate p values for the 5-year survival rates. For all statistical tests a significance level of 0.05 was used. All of the reported p values are 2-sided.

Multivariate Cox proportional hazards stepwise analysis was carried out using binary variables with 95% confidence intervals (CI) and was performed using SPSS version 13. However, due to the low number of HNSCC patients, stratification by primary site was not possible. Moreover, HPV16 and p16 results were excluded from the analysis and not used for any correlation studies since half of the FFPE samples were no longer available.

All the other variables with clinical significance were included in the model and a backward regression analysis was used to identify those factors that significantly affected the risk of death. Since evaluation of the validity of different prognostic factors for survival was the main objective of the study, we kept these factors in the model and applied the backward regression analysis to the other covariates.

When the multivariate analysis did not obtain an overall p value of <0.05 we excluded from the analysis the parameters without statistical significance in the Kaplan-Meier analysis.

Results

Description of TP53 SNP and Mutations in Patient Population

The clinicopathological and molecular data on SNP72 and the mutational status in TP53 are reported in Table I, where patients are grouped by primary tumor site.

At the time of this second analysis, 82% of patients had died (54/66). The median time to death was 4.5 years (95% CI 2.0-7.1). The distribution of the SNP72 genotypes in the 66 cases was RR = 51.5% (34/66), RP = 42.4% (28/66), and PP = 6.1% (4/66), not different from that in the general population: RR = 53.7%; RP = 38.3%; PP = 8% (p = 0.74) (21).

Mutations in the TP53 gene were detected in 60.7% of patients (37/61). Most of the mutations (92%, 34/37) were missense (MS); 1 patient had a nonsense (NS) mutation and 2 had frameshift mutations. Overall, in line with data in the published literature (22), 21% (13/61) were disruptive and 39% (24/61) were nondisruptive mutations (Tab. I).

Twenty of the 24 patients with wt TP53 (83%) experienced a complete response (CR) to treatment, whereas this happened to only 25 of the 37 patients with TP53 mutations (68%). The difference was not significant (p = 0.14).

The presence of mutations resulting in dysfunctional TP53 was not significantly related to any patient, tumor or biological/molecular parameter except age, with a significant difference between the wt and mutant groups of patients (chi-square test p = 0.02) (Tabs. II and III).

Table II

Prognostic factors in patients with wild-type versus mutant TP53

Parameter	Wt – Number (%)	Mutant – Number (%)	P value
No. of patients	24	37
Male	19 (79.2%)	34 (91.9%)	0.244^a
Female	5 (20.8%)	3 (8.1%)
Age <65 years ≥65 years	12 (50%) 12 (50%)	30 (81.1%) 7 (18.9%)	0.022^a^*
ECOG PS 0-1 PS 2	22 (91.7%) 2 (8.3%)	32 (86.5%) 4 (10.8%)	0.680^b
na	0	1 (2.7%)
Site of primary tumor
Oropharynx	6 (25%)	8 (21.6%)	0.805^a
Hypopharynx	9 (37.5%)	14 (37.8%)
Rhinopharynx (UCNT excluded)	2 (8.3%)	3 (8.1%)
Oral cavity	1 (4.2%)	5 (13.5%)
Larynx	6 (25%)	7 (19%)
Stage
Stage III	4 (16.7%)	4 (10.8%)	0.700^b
Stage IV	20 (83.3%)	33 (89.2%)
Tumor size
T1-2	3 (12.5%)	8 (21.6%)	0.455^a
T3-4	21 (87.5%)	28 (75.7%)
na	0	1 (2.7%)
Nodal status
N0-N1-N2	21 (87.5%)	31 (83.8%)	0.733^b
N3	3 (12.5%)	6 (16.2%)
Grade
Gx	4 (16.7%)	8 (21.6%)	0.844^a
G1-G2	12 (50%)	16 (43.2%)
G3	8 (33.3%)	13 (35.2%)
Final response
NE	0	1 (2.7%)	0.471^a
PD	0	3 (8.1%)
CR	20 (83.3%)	25 (67.6%)
PR	3 (12.5%)	7 (18.9%)
TD	1 (4.2%)	1 (2.7%)

ECOG PS = Eastern Cooperative Oncology Group performance status; UCNT = undifferentiated carcinoma of the nasopharynx; na = not available; CR = complete response; PR = partial response; TD = death due to toxicity; PD = progressive disease; NE = not evaluable.

chi-square test

Fisher's exact probability test

significant result.

Table III

Biological prognostic factors in the wild-type versus mutant patients’ group (n = 61)

Parameter	Wild type	Mutant	P value
No. of patients (%)	24 (39.3%)	37 (60.7%)
Arg/Pro SNP72
Arg homo (n = 29)	14 (58.3%)	15 (40.5%)	0.382^a
Het (n = 28)	9 (37.5%)	19 (51.3%)
Pro homo (n = 4)	1 (4.2%)	3 (8.2%)
EGFR
neg; 1+; 2+ (n = 39)	16 (66.7%)	23 (62.2%)	0.468^a
3+ (n = 22)	8 (33.3%)	14 (37.8%)
MDM2
Wt (n = 19)	9 (37.5%)	10 (27%)	0.63^a
GG/GT (n = 32)	12 (50%)	20 (54%)
na (n = 10)	3 (12.5%)	7 (19%)

Arg homo = Arg homozygote; Het = heterozygote Arg/Pro; Pro homo = Pro homozygote; EGFR = epidermal growth factor receptor; wt = wild type; na = not available.

Pearson's chi-square test.

Role of Codon 72 Polymorphism and TP53 Mutation as Predictive Biomarker

Clinical outcome in head and neck cancer is influenced by TP53 polymorphism

Taking the whole patient population (n = 66), there was no significant difference in OS between germline 72RP genotypes (p = 0.09), although a trend was observed towards more favorable OS in individuals with 72PP (n = 4) (Fig. 1A; median OS not reached). Intermediate OS in 72RR (n = 34) (median OS 5 years) and less favorable OS in 72RP (n = 28) (median OS 3 years) were also observed.

Fig. 1

Effect of TP53 RP SNP72 on overall survival (OS) is dependent on TP53 mutational status. The Figure shows the Kaplan-Meier analysis of OS as a function of TP53 RP SNP72 in the whole patient population (n = 66) (A); in cases with wt TP53 (n = 24) (B); in cases with missense TP53 mutations (n = 34) (C). Three patients were excluded: 1 with a nonsense mutation (SNP72 RP) and 2 with mutations resulting in frameshifts (1 with SNP72 RP and 1 with SNP72 RR).

In the 24 cases with wt TP53 (Fig. 1B), OS was shortest in patients with 72PP (median OS 0.6 versus 6.0 years in 72RR; p = 0.003). In individuals with 72RP, OS was intermediate (median OS 5.4 years) between 72RR (p = 0.57) and 72PP (p = 0.02).

Clinical Outcome and TP53 Missense Mutations

In the 34 patients with MS TP53 mutations, OS was longest in the 3 patients carrying 72PP (no statistics were computed because none of the patients had any events) and less favorable in the 17 patients carrying 72RP (Fig. 1C; median OS 2.8 years; log rank p = 0.058). In 14 individuals with germline 72RR we observed a behavior similar to that of individuals with the 72RP genotype (median OS 3.2 years, RR versus RP p = 0.55). Patients with NS and frameshift mutations were excluded.

Identification of Both 72P and Mutation in the same allele is a Predictive Biomarker of Clinical Outcome

Survival curves split by the presence (n = 37) or absence (n = 24) of mutations did not show any statistical significance in relation to OS or PFS (log-rank p = 0.91 and p = 0.97, respectively; Fig. 2A and 2B).

Fig. 2

Clinical outcome in head and neck cancer is influenced by TP53 mutations in relation to RP SNP72. Survival curves for all patients analyzed only by the presence or absence of mutations did not show any significant difference in overall survival (OS) or progression-free survival (PFS) (log-rank p = 0.91 and p = 0.97, respectively) (A and B). If mutations exist (n = 8 + 29 = 37), those occurring in the 72R allele predict worse OS (log-rank p = 0.009) (C).

In cancers carrying TP53 mutations (n = 37), mutations were detected more commonly in the 72R allele: 29 mutations in 72R and 8 mutations in 72P (Tab. I). Patients with mutations in the 72P allele had a significantly longer OS than patients carrying mutations in the 72R allele. Analysis by the log-rank test of the whole population showed that TP53 mutational status and SNP72 polymorphism correlated significantly with survival (OS p = 0.009) (Fig. 2C).

When analyzing OS, we found that 37.5% (3/8) of patients with mutations in 72P had died versus 90% of patients with mutations in 72R (26/29). The median survival was not reached at 15 years in patients with 72P mutations (95% CI not reached) versus 2.8 years in patients with 72R mutations (95% CI 1.2-4.4, 72P vs. 72R, log-rank test p = 0.004; data not shown). The wt TP53 population showed an intermediate value (87.5% dead; median OS 5.7 years, 95% CI 1.1-9.7; wt vs. 72R, log-rank test p = 0.21; wt vs 72P, log-rank test p = 0.015; data not shown).

When analyzing PFS in patients with mutations in the 72P allele, we found that 37.5% (3/8) progressed compared with those having mutations in the 72R allele, 55% (16/29) of whom relapsed (median PFS 15.6 years, 95% CI 0-31.4 vs. 2.1 years, 95% CI 1.5-3.6, respectively; log-rank test p = 0.030, data not shown). The wt TP53 population showed an intermediate value of progression with 14 of 24 patients developing a relapse (58.3%) (median PFS 6 years; 95% CI 3.5-8.5). Analysis with the log-rank test on the whole population showed no significant correlation between TP53 mutational status and SNP72 polymorphism and PFS (p = 0.08, data not shown).

MDM2 SNP309 is an Independent Predictive Biomarker

Genotyping for MDM2 SNP309 was performed in 54 patients; in 51 of these patients also information on TP53 mutational status was available. The results were 27.8% GG, 35.2% GT, and 37.0% TT. Median OS was significantly shorter in patients carrying GG or GT (n = 34) compared with the germline genotype TT (n = 20) (Fig. 3A; log-rank p = 0.002).

Fig. 3

MDM2 SNP309 is predictive of outcome irrespective of TP53 genotype and mutational status. (A) Kaplan-Meier analysis of overall survival (OS) in all patients analyzed for MDM2 SNP309 (n = 54). (B) Kaplan-Meier analysis of OS in cases with wild-type TP53 (n = 21/51 analyzed). (C) Kaplan-Meier analysis of OS in cases with TP53 missense mutations (n = 30/51 analyzed). (D) Kaplan-Meier analysis of OS in cases with TP53 alleles carrying R [RR and RP; number of patients analyzed = 44/47; 1 nonsense (RP SNP72) and 2 frameshift mutated patients (1 with RP and 1 with RR SNP72) were excluded.]

Interestingly, this effect was maintained regardless of the TP53 status, with a log-rank test that showed a trend in the wt subpopulation (n = 21) (Fig. 3B) and reached statistical significance in the other subgroups (Fig. 3C and 3D).

Even more interestingly, this predictive effect of the G allele was maintained independent of age, TP53 status, mutant 72P allele and EGFR grade [supplementary Table S1 – Main prognostic factors for overall survival (OS) in multivariate analysis (Cox proportional hazards model), available online at http://www.biological-markers.com].

Discussion

We performed this study to verify, with a follow-up exceeding 10 years, the influence of TP53 SNP72 on the outcome of HNSCC patients treated with CRT previously reported with a shorter follow-up (12, 13).

We are aware that in the last decade it was discovered that HNSCC with biologically active HPV is a distinct molecular entity (11, 23) and patients generally have a better prognosis (24), especially those with oropharynx tumors (11). However, in the present study, due to the small cohort of HNSCC patients (n = 66), we were not able to perform any stratification for primary sites nor for HPV infection and we analyzed the cohort as a whole.

In addition, we evaluated MDM2 SNP309 as a predictive factor in the same series of patients because of its close interaction with TP53. The TP53 gene is one of the most important players in cancer biology and it is now widely acknowledged that TP53 mutations are the most common genetic event in human cancer (25). It is less known which are the genes and proteins involved with TP53 in cancer growth, and whether some of them might be used as biomarkers to identify patients who will benefit from CRT.

In previous papers we have described how TP53 SNP72 influences outcome via functional effects on both mutant and wt TP53 (12, 13). In particular, we demonstrated that treatment outcome in patients with advanced HNSCC undergoing treatment with combined-modality CRT regimens is less favorable in cases expressing mutant 72R TP53 (13). Nevertheless, the overall predictive value of analysis of the germline SNP alone (i.e., independent of the presence or absence of TP53 mutations) is not statistically significant (20). This means that the use of TP53 R/P SNP72 as a predictive marker would require sequence analysis of TP53 in tumor tissue, which, apart from being time-consuming and expensive, is not always possible.

A limitation of our study is that we were not able to analyze the TP53 mutations found in our cohort of patients in the manner suggested by Poeta et al (22), i.e., classified as disruptive versus nondisruptive, and to correlate this with clinical outcome. The reason for this is the small number of patients analyzed and disruptive mutations found (we found 13 disruptive and 24 nondisruptive mutations).

Interestingly, this series of patients with advanced HNSCC with a follow-up longer than 10 years suggests that the analysis of both TP53 R/P SNP72 and TP53 mutation might increase the power of long-term outcome prediction. The result may be explained by the strong influence of this SNP on TP53 protein function, although with different effects, in both wt and mutated tumors.

The ring domain E3 ubiquitin ligase MDM2 is the most important negative regulator of TP53. It serves to keep TP53 levels in check under unstressed conditions (26-27-28). Additionally, also MDMX (alias MDM4), a homolog of MDM2, negatively regulates TP53 (29). Not surprisingly, both MDM2 and MDMX are overexpressed in different neoplasms (30). MDM2, which is strongly linked to TP53, has shown a particular polymorphism at position 309 of the promoter region: a T to G change (MDM2 309) that is associated with enhanced MDM2 expression and protein levels, causing attenuation of the TP53 pathway in vitro (19, 31).

We found a significant correlation between the MDM2 SNP309 genotype and clinical outcome; this is confirmed and maintained independently of other factors that might influence outcome. Thus, we demonstrated in this study and in our previous work (20) that MDM2 SNP309 is highly predictive of outcome, irrespective of whether TP53 is wt or mutant, and retains its independent predictive value compared with recognized clinicopathological and molecular variables of HNSCC such as age, stage, performance status, nodal status, EGFR and TP53 mutation.

Wan et al (32) performed a meta-analysis on cancer risk that involved 27,813 cases with various tumor types and 30,295 controls to clarify the potential interaction between MDM2 SNP309 and TP53 mutational status. The review confirmed our finding that the variant homozygote GG and the heterozygote TG were unfavorable variants in most populations (32).

In conclusion, our results strongly support the candidacy of MDM2 SNP309, determined in simple peripheral blood samples, as a robust predictive marker to stratify HNSCC patients treated with CRT.

Footnotes

Financial support: This work has been partially supported by AIRC and Fondazione Veronesi.

Conflict of interest: The authors have no conflict of interest to declare.

References

Argiris

Karamouzis

Raben

Ferris

Head and neck cancer. Lancet 2008; 371 (9625): 1695–1709

Pignon

Bourhis

Domenge

Designé L. Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: three meta-analyses of updated individual data. MACH-NC Collaborative Group. Lancet 2000; 355 (9208): 949–955

Argiris

Brockstein

Haraf

. Competing causes of death and second primary tumors in patients with locoregionally advanced head and neck cancer treated with chemoradiotherapy. Clin Cancer Res 2004; 10 (6): 1956–1962

Pignon

le Maître A, Bourhis J; MACH-NC Collaborative Group. Meta-Analyses of Chemotherapy in Head and Neck Cancer (MACH-NC): an update. Int J Radiat Oncol Biol Phys 2007; 69 2 Suppl S112–S114

Merlano

Alternating chemotherapy and radiotherapy in locally advanced head and neck cancer: an alternative? Oncologist. 2006; 11 (2): 146–151

Merlano

Benasso

Corvò

. Five-year update of a randomized trial of alternating radiotherapy and chemotherapy compared with radiotherapy alone in treatment of unresectable squamous cell carcinoma of the head and neck. J Natl Cancer Inst 1996; 88 (9): 583–589

Corvò

Benasso

Sanguineti

. Alternating chemoradiotherapy versus partly accelerated radiotherapy in locally advanced squamous cell carcinoma of the head and neck: results from a phase III randomized trial. Cancer 2001; 92 (11): 2856–2867

Lefebvre

Rolland

Tesselaar

et al; EORTC Head and Neck Cancer Cooperative Group; EORTC Radiation Oncology Group. Phase 3 randomized trial on larynx preservation comparing sequential vs alternating chemotherapy and radiotherapy. J Natl Cancer Inst 2009; 101 (3): 142–152

Shen

Zhang

Sun

Qian

Understanding the interplay between host immunity and Epstein-Barr virus in NPC patients. Emerg Microbes Infect 2015; 4 (3): e20

10.

Moore

II Mehta

The growing epidemic of HPV-positive oropharyngeal carcinoma: a clinical review for primary care providers. J Am Board Fam Med 2015; 28 (4): 498–503

11.

Gillison

Koch

Capone

. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers Natl Cancer Inst 2000; 92 (9): 709–720

12.

Bergamaschi

Gasco

Hiller

. p53 polymorphism influences response in cancer chemotherapy via modulation of p73-dependent apoptosis. Cancer Cell 2003; 3 (4): 387–402

13.

Sullivan

Syed

Gasco

. Polymorphism in wild-type p53 modulates response to chemotherapy in vitro and in vivo. Oncogene 2004; 23 (19): 3328–3337

14.

Marin

Jost

Brooks

. A common polymorphism acts as an intragenic modifier of mutant p53 behaviour. Nat Genet 2000; 25 (1): 47–54

15.

Tada

Furuuchi

Kaneda

. Inactivate the remaining p53 allele or the alternate p73? Preferential selection of the Arg72 polymorphism in cancers with recessive p53 mutants but not transdominant mutants. Carcinogenesis 2001; 22 (3): 515–517

16.

Furihata

Takeuchi

Matsumoto

. p53 mutation arising in Arg72 allele in the tumorigenesis and development of carcinoma of the urinary tract. Clin Cancer Res 2002; 8 (5): 1192–1195

17.

Soussi

Béroud C. Assessing TP53 status in human tumours to evaluate clinical outcome. Nat Rev Cancer 2001; 1 (3): 233–240

18.

Dumont

Leu

Della Pietra AC III, George DL, Murphy M. The codon 72 polymorphic variants of p53 have markedly different apoptotic potential. Nat Genet 2003; 33 (3): 357–365

19.

Bond

. A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell 2004; 119 (5): 591–602

20.

Vivenza

Gasco

Monteverde

. MDM2 309 polymorphism predicts outcome in platinum-treated locally advanced head and neck cancer. Oral Oncol 2012; 48 (7): 602–607

21.

Bonafè

Olivieri

Mari

. P53 codon 72 polymorphism and longevity: additional data on centenarians from continental Italy and Sardinia. Am J Hum Genet 1999; 65 (6): 1782–1785

22.

Poeta

Manola

Goldwasser

. TP53 mutations and survival in squamous-cell carcinoma of the head and neck. N Engl J Med 2007; 357 (25): 2552–2561

23.

Schlecht

Burk

Adrien

. Gene expression profiles in HPV-infected head and neck cancer. J Pathol 2007; 213 (3): 283–293

24.

Ang

Harris

Wheeler

. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010; 363 (1): 24–35

25.

Levine

Oren

The first 30 years of p53: growing ever more complex. Nat Rev Cancer 2009; 9 (10): 749–758

26.

Poyurovsky

Prives

Unleashing the power of p53: lessons from mice and men. Genes Dev 2006; 20 (2): 125–131

27.

Manfredi

The Mdm2-p53 relationship evolves: Mdm2 swings both ways as an oncogene and a tumor suppressor. Genes Dev 2010; 24 (15): 1580–1589

28.

Marine

Lozano

Mdm2-mediated ubiquitylation: p53 and beyond. Cell Death Differ 2010; 17 (1): 93–102

29.

Marine

Dyer

Jochemsen

MDMX: from bench to bedside. J Cell Sci 2007; 120 Pt 3 371–378

30.

Marine

Francoz

Maetens

Wahl

Toledo

Lozano

Keeping p53 in check: essential and synergistic functions of Mdm2 and Mdm4. Cell Death Differ 2006; 13 (6): 927–934

31.

Alhopuro

Ylisaukko-Oja

Koskinen

. The MDM2 promoter polymorphism SNP309T—>G and the risk of uterine leiomyosarcoma, colorectal cancer, and squamous cell carcinoma of the head and neck. J Med Genet 2005; 42 (9): 694–698

32.

Wan

Yin

Guan

Zhou

MDM2 SNP309, gene-gene interaction, and tumor susceptibility: an updated meta-analysis. BMC Cancer 2011; 11208

Correlation of TP53 and MDM2 Genotypes and Clinical Outcome in Platinum-Treated Head and Neck Cancer Patients with More than 10 Years’ Follow-Up

Abstract

Purpose

Methods

Results

Conclusions

Keywords

Introduction

Material and methods

Patients and Tumors

Analysis of TP53 and MDM2

Egfr Expression

HPV16 and p16 Expression

Statistical Analysis

Results

Description of TP53 SNP and Mutations in Patient Population

Role of Codon 72 Polymorphism and TP53 Mutation as Predictive Biomarker

Clinical Outcome and TP53 Missense Mutations

Identification of Both 72P and Mutation in the same allele is a Predictive Biomarker of Clinical Outcome

MDM2 SNP309 is an Independent Predictive Biomarker

Discussion

Footnotes

References