MDM2 309 polymorphism predicts outcome in platinum-treated locally advanced head and neck cancer

Oral Oncology 48 (2012) 602–607

Contents lists available at SciVerse ScienceDirect

Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology

MDM2 309 polymorphism predicts outcome in platinum-treated locally advanced head and neck cancer Daniela Vivenza a, Milena Gasco b, Martino Monteverde a, Laura Lattanzio a, Nelofer Syed c, Ida Colantonio b, Nerina Denaro d, Guido Natoli b, Alberto Comino e, Elvio Russi f, Marco Merlano b,⇑, Tim Crook c,g, Cristiana Lo Nigro a,g a

Laboratory of Cancer Genetics and Translational Oncology, Oncology Department, S. Croce General Hospital, via Carle 25, 12100 Cuneo, Italy Medical Oncology, Oncology Department, S. Croce General Hospital, via Carle 25, 12100 Cuneo, Italy Faculty of Medicine, Imperial College, St. Dunstan’s Road, London W6 8RP, UK d Oncology Department, University G. Martino, via Consolare Valeria, 98125 Messina, Italy e Pathology Department, S. Croce General Hospital, via Coppino 26, 12100 Cuneo, Italy f Radiation Therapy, S. Croce General Hospital, via Coppino 26, 12100 Cuneo, Italy b c

a r t i c l e

i n f o

Article history: Received 24 August 2011 Received in revised form 14 January 2012 Accepted 26 January 2012 Available online 21 February 2012 Keywords: Head and neck cancer SNP p53 MDM2 Cisplatin

s u m m a r y Chemo-radiotherapy (CRT) with cisplatin-based regimens is curative in a subset of patients with locally advanced (stage III and IV) squamous carcinomas of the head and neck (LAHNSCC), but causes considerable toxicity. To seek predictive biomarkers, we analysed single nucleotide polymorphisms (SNPs) in the p53 and MDM2 genes in LAHNSCC patients treated with cisplatin-based CRT. We analysed germ-line p53 72 Arg/Pro (R/P) and MDM2 309 SNPs and somatic p53 mutational status in 140 LAHNSCC and determined their utility as predictive biomarkers. In cases with wild-type p53, overall survival (OS) was longest in 72RR (median OS = 60.8 months) and less favourable in 72PP (median OS = 6.7 months, p < 0.0001). OS in individuals with 72RP was intermediate between 72RR and 72PP, while in patients with missense p53 mutations, median OS did not reach statistical significance. Median OS was significantly shorter in patients with MDM2 309 SNP genotypes GG or GT, compared to TT (15 vs. 86 months; p < 0.0001). The predictive effect of the G allele was maintained independent of age, gender, stage, primary site, nodal status, performance status, EGFR grade, HPV status, p53 mutation and p53 SNP (HR for death 3.241; 95% CI 1.90–5.52, p < 0.001). The predictive utility of the MDM2 germ-line 309 SNP, which can be easily determined from peripheral blood, implies that it may be of value in the objective selection of patients for radical CRT. In contrast, the predictive utility of the 72 Arg/Pro SNP in p53 requires mutational analysis of p53, limiting its routine clinical use. Ó 2012 Elsevier Ltd. All rights reserved.

Introduction Management of head and neck squamous cell carcinoma (HNSCC) remains a major clinical challenge in oncology. It 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 (stage III–IV) disease and, for such patients, the treatment failure rate after surgery and/or radiotherapy is high and survival poor. Introduction of multi-modal treatment strategies, combining radiation and chemotherapy, has substantially improved both local control and overall survival of advanced HNSCC patients, compared to locore⇑ Corresponding author. Tel.: +39 0171616739; fax: +39 0171616737. g

E-mail address: [email protected] (M. Merlano). These authors contributed equally to this work.

1368-8375/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2012.01.022

gional treatment alone, with 5-year survival rates close to 25%.2– 4 Rapidly alternating chemo-radiation is a variation of concurrent chemo-radiation, intended to minimise toxicity5 and has shown a statistically significant improvement in outcomes compared to conventional radiotherapy6 with a more favourable toxicity profile.6–8 For these reasons, rapidly alternating chemo-radiation is currently employed in our Institution as standard treatment for locally advanced, inoperable, HNSCC (LAHNSCC).9 Determinants of response to treatment remain unidentified. Tumour factors such as tumour volume, lymph node invasion and penetration of tumour through the capsule of involved lymph nodes are recognised prognostic variables predictive of poor outcome, but molecular determinants of prognosis and response to treatment are less well defined. Predictive biomarkers to allow rational selection of patients most likely to benefit from treatment and to spare unnecessary toxicity to those with a poor chance of response would be valuable. A common polymorphism exists in

603

D. Vivenza et al. / Oral Oncology 48 (2012) 602–607

exon 4 of the p53 gene, resulting in the expression of either arginine (R) or proline (P) at codon 72, and we have previously showed in a limited series of cases that a functional single nucleotide polymorphism (SNP) in the apoptosis signalling domain of p53 influences the clinical response of HNSCC to chemotherapy in cases with both mutant and wild-type p53.10–12 The stability of p53 is regulated by MDM2, an E3 ubiquitin ligase and p53 target gene. Levels of MDM2 are influenced by a SNP (MDM2 309) in the MDM2 promoter.13 This SNP is located in the first intron of the MDM2 gene which contains AP-1 and p53-responsive elements that increase MDM2 levels after a p53dependent response.14 The SNP MDM2 309 is a T to G change which generates an improved SP1 binding site. Functionally, cells with MDM2 309 have higher level of MDM2 and a lower apoptotic response than cells with homozygous T/T.14 A number of studies both in vivo and in vitro have suggested that MDM2 309 may influence susceptibility to cancer.15 However, very little is known of the effect of this SNP on response to cytotoxic therapies in oncology. Here, we have addressed this question in a well-characterised series of cases of locally advanced, inoperable, stage III–IV HNSCC. This cohort of previously untreated patients represents an ideal group in which to assess the relative effects of the p53 and MDM2 SNPs on response to cytotoxic anti-cancer therapy and to evaluate the candidacy of such SNPs as predictive biomarkers to inform the management of this increasingly common disease. Patients and methods Patients and tumours The patient population comprised 140 cases of locally advanced, inoperable, TNM stage III–IV HNSCC presenting to the Medical Oncology Department of Cuneo General Hospital, between May 1997 and March 2010. The characteristics of the patient population, part of which was previously described11,12 are presented in Table 1. All the patients in this study are heavy smokers (median 45 packs per year). Patients underwent combined treatment consisting of radiotherapy, alternating with one of the following cisplatin-based chemotherapy regimens: (i) Cisplatin 20 mg/m2 day 1 to 5 + 5-fluorouracil 200 mg/m2 day 1 to 5 every 3 weeks for three courses6–8; (ii) Cisplatin 25 mg/m2 day 1 to 3 + 5-fluorouracil 250 mg/m2 day 1 to 3 + paclitaxel 160 mg/m2 day 1 every 3 weeks for three courses16; (iii) Cisplatin 20 mg/m2 day 1 to 5 + gemcitabine 800 mg/m2 day 1 and 12, every 3 weeks17; (iv) Cisplatin 20 mg/m2 day 1 to 5 + 5-fluorouracil 200 mg/m2 day 1 to 5 every 3 weeks up to three courses + Cetuximab 400 mg/m2 at first infusion, then 250 mg/m2 weekly.9 Radiotherapy was administered during the pauses between chemotherapy cycles, alternating with chemotherapy, and consisted of 2 Gy daily fractions up to 70 Gy. Radiotherapy was not given on the days when cytotoxic treatment was delivered. Radiotherapy employed the standard 3 shrinking field technique (6 MV photons) with shielding blocks drawn on simulator films to protect critical organs.2 After a cumulative dose of 40 Gy, medulla was excluded from irradiation and high-energy (8–10 MeV) electron beams were used to treat the posterior regions of the neck. Tissue and blood samples The study was performed with approval from the Ethical Committee of S. Croce General Hospital, Cuneo and informed patient

consent. Tissue was obtained as paraffin sections or fresh-frozen tissue from patients at the time of diagnosis. The diagnosis and presence of an adequate tumour representation was confirmed by histopathological review. For some patients, normal tissue was obtained from blood (available for 21 patients) or uninvolved tissue (confirmed by histopathological analysis to be free from tumour). If blood or uninvolved tissue was not available, normal tissue was obtained by micro-dissection using sections stained with haematoxylin and eosin to identify areas of normal tissue.

Analysis of p53 and MDM2 Genomic DNA was purified by proteinase K digestion of 10 lm sticks cut from paraffin sections and from frozen normal tissue using a standard xylene–phenol protocol. DNA was extracted from leucocytes of peripheral blood using the EZ1 DNA blood kit (Qiagen, Venlo, Netherlands). p53 mutations were sought in tumour DNA by direct sequencing in the most frequently affected exons 4–10, which were individually amplified. Primer sets, annealing temperatures used for PCR reactions and length of amplicons are shown in Supplementary material (Table 1 online only). Table 1 Patient characteristics (N = 140). N/A: not available. Gender Median age (range)

Male = 117 Female = 23 57.5 (21–77)

Performance status 0 1 2

73 (52.2%) 58 (41.4%) 9 (6.4%)

Primary site Hypopharynx Oropharynx Oral cavity Larynx Rhinopharynx Other sites

40 (28.6%) 36 (25.7%) 26 (18.6%) 30 (21.4%) 6 (4.3%) 2 (1.4%)

Stage III IV

17 (12%) 123 (88%)

Node N0 N1 N2 N3 N/A

26 (18.6%) 13 (9.3%) 79 (56.4%) 21 (15.0%) 1 (0.7%)

Tumour T0 T1 T2 T3 T4 X

1 (0.7%) 2 (1.4%) 25 (17.9%) 30 (21.4%) 75 (53.6%) 7 (5.0%)

Grade 1 2 3 4 X or N/A

10 (7%) 77 (55%) 47 (34%) 2 (1%) 4 (3%)

EGFR Grade 0–1 Grade 2–3 N/A

35 (25.0%) 99 (70.7%) 6 (4.3%)

HPV16 Negative Positive N/A

122 (87,1%) 16 (11.5%) 2 (1.4%)

604

D. Vivenza et al. / Oral Oncology 48 (2012) 602–607

PCR was carried out in a total volume of 10 ll consisting of 1.5 ll of a 5 lM mix solution of both primers, 1.5 mmol/L MgCl2, 200 lM deoxynucleotide triphosphate, 0.25 U of AmpliTaq GoldÒ 360 DNA Polymerase (Applied Biosystems, Foster City, CA), 1 ll of 360 GC Enhancer, 1.5 ll of genomic DNA (25 ng/ll) and adjusted H2O, using a GeneAmp PCR System 9700 Thermal Cycler (Applied Biosystems, Foster City, CA). The PCR program had an initial denaturation step of 10 min at 95 °C followed by 40 cycles of 30 s at 95 °C, 45 s at adequate annealing temperature for each fragment and 45 s at 72 °C, at last a final step at 72 °C for 7 min. Proposed mutations were confirmed at least twice by repeat DNA isolation, independent amplifications and sequence reactions. PCR products of p53 exons (4–10) were directly sequenced in both directions on an ABI 3130 automated sequencer, using the Big Dye terminator cycle sequencing reaction kit following manufactures’ instructions (Applied Biosystems). Analysis was performed using Variant Reporter Software v1.0 (Applied Biosystems). Single nucleotide polymorphisms (SNPs) in p53 at codon 72 (rs1042522) and in MDM2, 309 SNP (rs2279744), were analysed by direct sequencing after PCR amplifications of genomic DNA from blood or normal tissue, using forward primer 50 -AGAAT GCCAGAGGCTGCTCC-30 and reverse primer 50 -GCAACTGACCGTGCAAGTCA-30 and forward primer 50 -CGGGAGTTCAGGGTAAAGGT30 and reverse primer 50 -AATCCCGCCCAGACTACGC-30 , respectively. Genotypes were distributed typically for a Northern Italian population. PCR was performed with an initial denaturation step of 10 min at 95 °C followed by 40 cycles of 30 s at 95 °C, 45 s of annealing at 58 or 60 °C for p53 and MDM2 SNPs, respectively, and 40 s at 72 °C with a final step at 72 °C for 7 min.

ISH iVIEWTM Blue Plus Detection Kit (Ventana); staining of the slides was performed with a BenchMark-XT (Ventana) automated slide preparation system. Some cases were confirmed by qualitative PCR, using forward primer 50 -GAGATGCAGTACAGGTTCTA AAACG-30 and reverse primer 50 -TGCCATACCCGCTGTCTTC-30 . Amplification reaction was carried out as above described. The PCR program had an initial denaturation step of 10 min at 95 °C followed by 40 cycles of 30 s at 95 °C, 30 s of annealing at 54 °C and 40 s at 72 °C, at last a final step at 72 °C for 7 min.

Human papillomavirus (HPV)

Utility of p53 codon 72 SNP as a predictive biomarker requires mutational analysis

Samples were analysed for HPV genomic sequences by immunohistochemistry (IHC), using INFORM HPV III Family 16 Probe B (Ventana Medical Systems, San Mateo, California, USA) and the

A

Wild-type and mutant p53

1.0

Epidermal growth factor receptor (EGFR) EGFR was detected by IHC using the EGFR pharmDx™ kit (Dako, Glostrup, Denmark) and quantified according to FDA guidelines on a scale from 0 to 3+. Statistical analysis Overall survival (OS) curves, according to p53 and MDM2 status, were calculated using the Kaplan–Meier life-table method and the log-rank test was used to evaluate the significance. OS was defined as time from treatment start to death from any cause. Surviving patients were withdrawn at the last follow-up. Multivariate Cox proportional hazards stepwise analysis was carried out using binary variables with 95% confidence interval (95% CI). Fisher’s two-tailed exact test was used to calculate P values for the 5-years survival rates. Results

Mutations in p53 were detected in 73/140 patients (52%). Of the 73 patients with p53 mutations, 15 had a non-sense (NS) and 62 a

B

0.6 PP

0.4 RR

0.2 0

0.6 0.4

RR

50 100 Overall survival (months)

150

Mutant p53 (missense)

0

D

RP

PP

0

P=0.218

1.0

0.8

0.2

RP

0

C

Surviving fraction

Surviving fraction

0.8

Wild-type p53

1.0

1.0

P < 0.0001 3

150

120 90 60 Overall survival (months)

Mutant p53 (all)

0.8

0.6 0.4 RR

0.2

Surviving fraction

Surviving fraction

PP PP

0.8 0.6 0.4

RR

0.2 RP

0

RP

0

P=0.031 0

50 100 Overall survival (months)

150

P=0.062 0

50 100 Overall survival (months)

150

Figure 1 Effect of p53 72 Arg/Pro SNP is dependent on p53 mutational status. The figure shows Kaplan–Meier analysis of overall survival as a function of p53 72 Arg/Pro SNP. (A) In the whole patient population (N = 140). (B) In cases with wild-type p53. (C) In cases with MS p53 mutation. (D) In all cases with mutation (MS and NS).

605

D. Vivenza et al. / Oral Oncology 48 (2012) 602–607

missense (MS) mutation (four patients had both a MS and a NS mutation). p53 mutations are shown in full in Supplementary material (Table 2 online only). Taking the patient population as a whole, there was no significant difference in OS between germ-line 72 Arg/Pro genotypes (p = 0.22; log rank, Mantel Cox), although there was a trend towards more favourable OS in individuals with 72RR (median OS = 32.1 months), intermediate OS in 72RP (median OS = 18.6 months) and least favourable in 72PP (median OS = 7.2 months) (Fig. 1A). In the 67 cases with wild-type p53, overall survival was longest in 72RR (median OS = 60.8 months)

A

1.0

All patients

and less favourable in 72PP (median OS = 6.7 months, p < 0.0001). OS in individuals with 72RP was intermediate between 72RR and 72PP (median OS = 21.1 months; Fig. 1B). In the 62 patients with missense p53 mutations, overall survival was longest in 72PP mean survival time = 102.8 months (median OS = not yet reached) and less favourable in 72RP with a mean survival time = 40.7 months (median OS = 17.4 months), p < 0.05. In individuals with germ-line 72RR, mean survival time = 53.7 months (median OS = 26.9 months), intermediate between 72RP and 72PP (Fig. 1C). Inclusion of nonsense mutations in the p53 mutants did

B

Wild-type p53

0.8

0.6 TT

0.4

Surviving fraction

Surviving fraction

0.8

1.0

0.2

0.6 TT

0.4 0.2

GG/GT

0

50 100 Overall survival (months)

0

C

GG/GT

0

P < 0.0001

1.0

150

60

3

0

90

Mutant p53

D

1 .0

72 Arg/Arg

0 .8

TT

0 .4

Surviving fraction

0.6

0 .6 TT

0 .4 0 .2

0.2 GG/GT

0

150

120

Overall survival (months)

0.8 Surviving fraction

P < 0.0001

GG/GT

0

P < 0.0001 0

50 100 O Overallll survival i l ((months) th )

P < 0.0001 3

0

150

120 90 60 Overall survival (months)

150

Figure 2 MDM2 309 SNP is predictive of outcome irrespective of p53 genotype and mutational status. (A) Kaplan–Meier analysis of overall survival in all patients. (B) Kaplan– Meier analysis of overall survival in cases with wild-type p53. (C) Kaplan–Meier analysis of overall survival in cases with mutant p53. (D) Kaplan–Meier analysis of overall survival in cases with p53 72 Arg/Arg.

HPV16

1.0

1.0

0.8

0.8

0.6

0.4 -

Surviving fraction

Surviving fraction

EGFR

0.6

0.4 HPV16 -

0.2

0.2

P=0.196 0

+

P=0.550 50

100

Overall survival (months)

150

0

50

100

150

Overall survival (months)

Figure 3 Kaplan–Meier analysis of overall survival in all patients as a function of EGFR expression and HPV16 status.

606

D. Vivenza et al. / Oral Oncology 48 (2012) 602–607

not modify the survival effects (Fig. 1D). These results show that the predictive utility of the p53 72 Arg/Pro SNP is dependent on the presence or absence of mutations in p53. MDM2 309 SNP is an independent predictive biomarker Out of the 126 patients we genotyped for the MDM2 309 SNP, 76 were of germ-line genotype GG or GT, whereas 50 were genotype TT. In the whole patient population, median OS was significantly shorter in patients with genotypes GG or GT, compared to germ-line genotype TT (15 vs. 86 months; log rank p < 0.0001; Fig. 2A). Furthermore, 5-year survival was significantly worse in GG or GT genotypes (11/65 vs. 24/26 months; Fisher’s exact test p < 0.0001). The effect of the MDM2 309 SNP on OS was maintained in cases with wild-type p53 (Fig. 2B), mutant p53 (Fig. 2C) and p53 72 SNP (Fig. 2D). The predictive effect of the G alleles was also maintained independently of age, gender, stage, primary site, nodal status, performance status, EGFR grade and HPV positivity (Fig. 3, Table 2). Discussion The ability to identify patients most likely to respond to radical yet highly toxic cancer therapy remains a key goal of translational oncology. This objective is nowhere more relevant than in LAHNSCC, in which combined modality CRT offers a potentially curative therapeutic option to some patients, but at the price of high morbidity, mortality and economic cost. Robust biomarkers to individualise the use of CRT have not, however, been described. In the present study, we demonstrated striking effects of two SNPs in the p53 pathway on clinical outcome in this patient population.

Table 2 Multivariate analyses of MDM2 309 SNP and p53 72 SNP. HR = hazard ratio; CI = confidence interval; PS = performance status; N = nodal status. Predictor

P value

HR

95% CI

Multivariate analysis of MDM2 309 SNP Overall P value = 0.003 MDM2 309 GG/GT <0.001 3.241 1.90–5.52 Age > 65 years 0.372 1.29 0.74–2.24 Male gender 0.982 0.99 0.48–2.04 PS > 1 0.852 1.10 0.42–2.86 Stage > III 0.872 0.93 0.36–2.35 N>2 0.818 1.08 0.55–2.14 Hypopharynx/Larynx 0.351 0.78 0.47–1.31 EGFR 0.166 1.49 0.85–2.62 HPV16 0.705 1.15 0.56–2.33 p53 mutation 0.372 1.24 0.77–2.00 p53 SNP72 PP 0.161 0.50 0.19–1.31 Multivariate analysis of wild-type p53 72 Pro/Pro SNP Overall p53 72 PP 0.005 7.624 Age > 65 years 0.117 1.79 Male gender 0.458 0.64 PS > 1 0.464 0.61 Stage > III 0.403 0.61 N>2 0.265 1.74 Hypopharynx/Larynx 0.471 0.77 HPV16 0.654 1.30 EGFR 0.232 1.68

P value = 0.02 1.85–31.44 0.86–3.70 0.20–2.07 0.17–2.27 0.19–1.95 0.66–4.58 0.37–1.58 0.41–4.07 0.72–3.93

Multivariate analysis of mutant p53 72 Pro/Pro SNP Overall P value = 0.171 p53 72 PP 0.054 0.24 0.57–1.02 Age > 65 years 0.557 1.27 0.57–2.83 Male gender 0.468 1.43 0.55–3.72 PS > 1 0.282 1.98 0.57–6.87 Stage > III 0.842 0.87 0.23–3.30 N>2 0.254 1.71 0.68–4.30 Hypopharynx/Larynx 0.970 1.01 0.48–2.14 HPV16 0.314 1.63 0.63–4.22 EGFR 0.275 1.44 0.75–2.80

In particular, we have shown for the first time that the germ-line MDM2 309 SNP is a highly informative predictive biomarker of outcome and our results imply that determination of the germ-line genotypes at codon 72 of p53 and MDM2 309 may have utility in prediction of clinical outcome in patients treated with cisplatinbased CRT and therefore have important implications in facilitating identification of patients appropriate to receive such therapy. We previously showed that codon 72 SNP in p53 influences outcome via functional effects on both mutant and wild-type p53.11,12 In the present study, we have confirmed in a larger cohort that the p53 72 Arg/Pro SNP strongly influences outcome. Because of the opposing effects of the 72 SNP on wild-type and mutant p53 function, the overall predictive value of analysis of the germ-line SNP alone (i.e. independent of the presence or absence of p53 mutations) is not significant. The implication of this is that use of the p53 72 Arg/Pro SNP to triage patients would require sequence analysis of p53. In the routine clinical setting, this could limit the utility of analysis of this SNP. The MDM2 309 SNP is, however, highly predictive of outcome, irrespective of whether p53 is wild-type or mutant and retains independent predictive power compared to recognised clinical– pathological variables for HNSCC. The lack of significance of HPV and EGFR as prognostic factors in this population of heavy smoking patients, with unresectable LAHNSCC has already been described.18,19 The importance of treatment intensification as a factor able to overcome the role of known prognostic factors related to poor prognosis in patients treated with surgery or radiotherapy has also been reported previously by Benasso et al.20 In that study, multiple logistic regression analysis, considering age at diagnosis, gender, site of the primary tumour, size of the primary, nodal involvement, performance status and the treatment, indicated that only treatment was associated with complete response and overall survival (p = 0.0001). Our results therefore support the hypothesis that treatment intensification in such a patient population, through the concomitant administration of chemotherapy and radiation, overcomes the prognostic disadvantage of an EGFR-positive and HPV16-negative status. Therefore, in these patients, MDM2, which maintains its prognostic value, becomes even more relevant. There are few studies of the effect of the MDM2 309 SNP on outcomes following cancer therapy, although one study suggests that the two SNPs affect outcome in early stage, surgically-treated, HNSCC patients receiving post-operative radiotherapy.21 The priority is now to validate our results in independent clinical series. Conformation of our data would strongly support the candidacy of MDM2 309 SNP (determined from routine peripheral blood sampling) as a viable means to stratify patients for suitability to receive CRT. Conflict of interest statement We state explicitly that there are no potential conflicts of interest. Acknowledgment C.L.N., L.L., M.M. and M.G. were partially supported by the Associazione Italiana Ricerca Cancro. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.oraloncology.2012.01.022.

D. Vivenza et al. / Oral Oncology 48 (2012) 602–607

References 1. Argiris A, Karamouzis MV, Raben D, Ferris RL. Head and neck cancer. Lancet 2008;371(9625):1695–709. 2. Pignon JP, Bourhis J, Domenge C, 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. Meta-analysis of chemotherapy on head and neck cancer. Lancet 2000;355: 949–55. 3. Argiris A, Brockstein BE, Haraf DJ, Stenson KM, Mittal BB, Kies MS, et al. Competing causes of death and second primary tumors in patients with locoregionally advanced head and neck cancer treated with chemoradiotherapy. Clin Cancer Res 2004;10:1956–62. 4. Pignon JP, le Maitre A, Bourhis J. Meta-analyses of chemotherapy in head and neck cancer (MACH-NC): an update. Int J Radiat Oncol Biol Phys 2007;69: S112–4. 5. Merlano M. Alternating chemotherapy and radiotherapy in locally advanced head and neck cancer: an alternative? Oncologist 2006;11:146–51. 6. Merlano M, Benasso M, Corvò R, Rosso R, Vitale V, Blengio F, et al. 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;88:583–9. 7. Corvò R, Benasso M, Sanguineti G, Lionetto R, Bacigalupo A, Margarino G, et al. Alternating chemoradiotherapy versus partly accelerated radiotherapy in locally advanced squamous cell carcinoma of the head and neck. Cancer 2001;92:2856–67. 8. Lefebvre JL, Rolland F, Tesselaar M, Bardet E, Leemans CR, Geoffrois L, et al. Phase III randomized trial on larynx preservation comparing sequential versus alternating chemotherapy and radiotherapy. J Natl Cancer Inst 2009;101: 142–52. 9. Merlano M, Russi E, Benasso M, Corvò R, Colantonio I, Vigna-Taglianti R, et al. Cisplatin-based chemoradiation plus cetuximab in locally advanced head and neck cancer: a phase II clinical study. Ann Oncol 2011;22:712–7. 10. Marin MC, Jost CA, Brooks LA, Irwin MS, O’Nions J, Tidy JA, et al. A common polymorphism acts as an intragenic modifier of mutant p53 behaviour. Nat Genet 2000;25:47–54.

607

11. Bergamaschi D, Gasco M, Hiller L, Sullivan A, Syed N, Trigiante G, et al. p53 polymorphism influences response in cancer chemotherapy via modulation of p73-dependent apoptosis. Cancer Cell 2003;3:387–402. 12. Sullivan A, Syed N, Gasco M, Bergamaschi D, Trigiante G, Attard M, et al. Polymorphism in wild-type p53 modulates response to chemotherapy in vitro and in vivo. Oncogene 2004;23:3328–37. 13. Bond GL, Hu W, Bond EE, Robins H, Lutzker SG, Arva NC, et al. A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell 2004;119:591–602. 14. Bond GL, Hu W, Levine AJ. A single nucleotide polymorphism in theMDM2 gene: from a molecular and cellular explanation to clinical effect. Cancer Res 2005;65:5481–4. 15. Post SM, Quintás-Cardama A, Pant V, Iwakuma T, Hamir A, Jackson JG, et al. A high-frequency regulatory polymorphism in the p53 pathway accelerates tumor development. Cancer Cell 2010;18:220–30. 16. Merlano M, Russi EG, Numico G, Colantonio I, Garrone O, Pelissero A, et al. Paclitaxel, cisplatin, 5-fluorouracil and radiotherapy in the management of advanced squamous cell carcinoma of the head and neck: a phase II trial. Radiother Oncol 2005;75:193–6. 17. Numico G, Russi EG, Vitiello R, Sorrentino R, Colantonio I, Cipolat M, et al. Gemcitabine and cisplatin in a concomitant alternating chemoradiotherapy program for locally advanced head-and-neck cancer: a pharmacology-guided schedule. Int J Radiat Oncol Biol Phys 2006;66:731–7. 18. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân PF, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;363(1):24–35. 19. Numico G, Russi EG, Colantonio I, Lantermo RA, Silvestris N, Vitiello R, et al. EGFR status and prognosis of patients with locally advanced head and neck cancer treated with chemoradiotherapy. Anticancer Res 2010;30(2):671–6. 20. Benasso M, Bonelli L, Numico G, Corvò R, Sanguineti G, Rosso R, et al. Treatment with cisplatin and fluorouracil alternating with radiation favourably affects prognosis of inoperable squamous cell carcinoma of the head and neck: results of a multivariate analysis on 273 patients. Ann Oncol 1997;8(8):773–9. 21. Tu HF, Chen HW, Kao SY, Lin SC, Liu CJ, Chang KW. MDM2 SNP 309 and p53 codon 72 polymorphisms are associated with the outcome of oral carcinoma patients receiving postoperative irradiation. Radiother Oncol 2008;87:243–52.