- Email: [email protected]
Tp53 status as guide for the management of ethmoid sinus intestinal-type adenocarcinoma
Oral Oncology 49 (2013) 413–419
Contents lists available at SciVerse ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Tp53 status as guide for the management of ethmoid sinus intestinal-type adenocarcinoma Paolo Bossi a,g, Federica Perrone b,⇑,g, Rosalba Miceli c, Giulio Cantù d, Luigi Mariani c, Ester Orlandi e, Carlo Fallai e, Laura D. Locati a, Barbara Cortelazzi b, Pasquale Quattrone b, Paolo Potepan f, Lisa Licitra a,h, Silvana Pilotti b,h a
Head and Neck Medical Oncology Unit, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy Laboratory of Experimental Molecular Pathology, Department of Pathology, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy Statistic Unit, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy d Department of Otolaryngology and Maxillo-facial surgery, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy e Department of Radiotherapy, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy f Department of Radiology, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy b c
a r t i c l e
i n f o
Article history: Received 7 November 2012 Accepted 24 December 2012 Available online 29 January 2013 Keywords: Intestinal-type adenocarcinoma (ITAC) TP53 Overal survival
s u m m a r y Objective: Intestinal-type adenocarcinoma (ITAC) of the ethmoid sinus is a rare, occupational-related tumor. Optimal treatment consists of surgery and radiotherapy, while chemotherapy is still investigational. The molecular profile of ITAC is characterized by the occurrence of TP53 mutations associated with genotoxic agents such as wood dust. We investigated the role of p53 functionality in relation to the primary treatment. Materials and methods: We retrospectively reviewed 100 medical charts of consecutive patients with a first diagnosis of ITAC treated at our Institute; 74 patients were evaluable for TP53 analysis. Thirty (41%) were treated from 1991 to 2006 with craniofacial resection followed by radiotherapy (Group A), compared with 44 patients (59%) treated from 1996 to 2006 with cisplatin-based induction chemotherapy (PFL) followed by standard treatment (Group B). Results: Five-year OS in Group A was 42%, while in Group B it was 70% (p = 0.041); 5-year DFS in Group A was 40%, while in Group B it was 66%, (p = 0.009) (p = 0.061 and 0.003 at Cox multivariable OS and DFS analyses). Analyzing each group according to p53 functional status, only for Group B patients (who received preoperative chemotherapy) both OS and DFS were in favor of functional p53 (p = 0.023 and p = 0.010, respectively). No impact of p53 functional status as a biomarker was observed in Group A. Conclusions: Functional p53 may predict PFL-chemotherapy efficacy, offering a possible increase in survival when induction chemotherapy is given to a selected population. On the other hand, upcoming innovative approaches should be explored in the presence of non-functional p53. Ó 2013 Elsevier Ltd. All rights reserved.
Introduction Carcinomas of the paranasal sinuses are rare, accounting for less than 1% of all cancers and for 3% of those in the upper aerodigestive tract.1–3 Among these, ethmoid sinus carcinomas represent approximately 20% of all malignant tumors.4 Adenocarcinoma is the most frequent histological type in all the European series of ethmoid sinus cancers,5,6 5 year overall survival for all stages is ⇑ Corresponding author. Address: Department of Pathology, Fondazione IRCCS Istituto Nazionale Tumori, Via Venezian, 1-20133 Milan, Italy. Tel.: +39 0223902280; fax: +39 0223902877. E-mail address: [email protected] (F. Perrone). g These authors have contributed equally to this work. h Senior coauthors. 1368-8375/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.oraloncology.2012.12.011
46%, according to a European population-based cancer registries analysis, with no improvement over time.7 Intestinal-type adenocarcinoma (ITAC) has an indisputable occupational etiology related to exposure to organic dusts, mainly wood and leather.8 Consistently, ITAC is characterized by a high occurrence of molecular alterations reported to be associated with genotoxic agents, such as TP53 mutations represented by CG:C > A:T transitions and involving the CpG dinucleotides, tumor suppressor gene promoter methylation and LOH at the 9p21 locus.9–11 There is a strong histopathologic resemblance between ITAC and colon cancer, also in the molecular profile.12 Ethmoid sinus ITACs are rarely considered in single series. In this context, natural history and prognosis of these neoplasms are difficult to determine.13,14
414
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Moreover only a few biological tumor markers with prognostic value have been studied.15 Standard treatment consists of optimal surgical resection followed by radiotherapy.16,17 Multimodality therapies comprising chemotherapy in a variety of schemes have been employed with promising results.18–22 Apparently, despite the improvement in techniques, radiotherapy as an exclusive treatment has not translated into the achievement of better survival over the last 50 years.23 The role of the TP53 mutation in head and neck squamous cell carcinoma (HNSCC) is conflicting.24 TP53 mutations have been associated with reduced survival after radiotherapy25,26 or cisplatin chemotherapy,27,28 whereas contrasting data are reported with surgery as primary treatment.26,29 Few data are reported on ITAC where, similarly to SCC of the oral cavity,30 TP53 wild type seems to be a promising predictive biomarker for pathologic complete response (pCR) to primary cisplatin-based chemotherapy; on the other hand, TP53 mutations have been reported to predict a low response rate to systemic treatment.31 All these findings suggest a predictive value of TP53 status after DNA damaging therapies. Based on these observations, we were interested in investigating the role of TP53 status in ethmoid sinus ITACs, by comparing a homogeneous series of patients who primarily underwent surgery, with a series who in addition received an induction PFLregimen. Patients and methods Patient selection, treatment and follow up We retrospectively reviewed 100 medical charts of consecutive patients with first diagnosis of ethmoid sinus ITAC all treated at our Institution from 1991 to 2006. Among this group, 26 patients were excluded from analysis: 4 because of postoperative death; 13 because the initial diagnostic biopsy was performed elsewhere and was not available for TP53 analysis; and 9 due to the unfeasibility of TP53 gene sequencing (Bouin fixed material). The patients were divided into two groups according to the treatment they received. Group A Thirty treatment-naive patients from February 1991 to December 2006 were examined. Tumor staging included physical examination, chest X-rays, head and neck computerized tomography (CT) and/or magnetic resonance imaging (MRI). Tumor staging was reviewed according to AJCC–UICC 2002 7 h classification criteria32 and INT classification.33 Treatment consisted of anterior craniofacial resection followed by postoperative radiotherapy. Up to the beginning of 2003 radiation therapy was given with 2D conventional treatment; thereafter it was performed with three-dimensional conformal RT based on fully 3D treatment planning. Field set-up to irradiate the target volume (primary tumor and operative bed), usually consisted of one anterior and two lateral or oblique 6 MV photon beams. Radiation was delivered in doses of 2 Gy/fraction at five fractions weekly, for a total dose of 50–66 Gy, according to the absence or presence of close/positive margins. No elective nodal irradiation was performed. Follow-up was conducted according to institutional policies, comprising physical examination every 3 months for the first year, every 6 months for the next 2 years and then annually; head and neck radiological exams (CT or MRI) were requested every 6 months in the first 2 years, then annually for another 3 years, while chest X-rays were performed annually.
Group B Forty-four treatment-naive patients treated from January 1996 to March 2006 with primary chemotherapy (all with PFL-regimen: cisplatin, fluorouracil and leucovorin) followed by anterior craniofacial resection and radiotherapy were considered; treatment and clinical outcome of 30 of these patients has already been reported in our previous study.31 Staging, postoperative radiation treatment and follow-up procedures were similar to those in Group A. However, in this group, the extent of the initial disease before chemotherapy was considered to define the radiation therapy target volume. Tumor specimen and molecular analysis of TP53 gene Formalin-fixed, paraffin-embedded tissue obtained before any treatment from surgical specimens or diagnostic biopsies were analyzed in Groups A and B, respectively. Pathologic exams were reviewed by the same pathologist (PQ). Genomic DNA was extracted from microdissected 7-lm methylene-blue-stained sections as previously described.9 The samples were analyzed for TP53 mutations in exons 5–8 by PCR and automated DNA sequencing (ABIprism 377, Applied Biosystem), as previously described.34 The detected mutations were confirmed at least twice, by independent amplifications and sequence reactions. All the missense mutations were functionally classified according to transactivation activities (TA) as ‘‘fully functional’’ (median TA >75% and 6140%), ‘‘partially functional’’ (median TA >20% and 675%) or ‘‘non-functional’’ (median TA 620%).35 A p53 functional protein was considered when TP53 was wildtype or carried fully or partially functional missense mutation. Statistical methods The K statistic was calculated to evaluate the agreement between TP53 status and p53 functionality. The association between continuous variables and treatment, TP53 gene status or p53 functionality was assessed using the Mann–Whitney–Wilcoxon test; the chi-square or the Fisher exact test with mid-p correction, as appropriate, were used when categorical variables were involved in testing association. The study end-points were overall survival (OS) and diseasefree survival (DFS). In all the survival analyses the initial date was taken as the date of surgery for Group A and that of starting chemotherapy for Group B. For event-free patients follow-up time was censored at 96 months (approximately the median follow-up time of the Group B patients with the shortest follow-up). OS time was calculated from initial date to the date of death due to any cause, or it was equal to the censoring time for living patients. DFS time was calculated from initial date to the date of disease relapse, second primary tumor or death without evidence of disease, whichever occurred first; DFS time was equal to censoring time for disease-free patients. OS and DFS curves were estimated by the Kaplan Meier method and compared with the log-rank test. Multivariable OS and DFS analyses were performed using multivariable Cox models including as covariates the following: treatment (no induction CT, induction CT); p53 functionality (non-functional, functional); postoperative radiotherapy (no, yes); and INT stage (as a linear trend). INT stage was chosen instead of the AJCC–UICC staging system because of its better discriminative performance.36 To explore whether the treatment prognostic effect was different for patients with non-functional p53 as compared with patients with functional p53, additional multivariable Cox models included the interaction term between treatment and p53 functionality; with such models we could also explore whether p53
415
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
functionality provided different survival according to having received or not having received induction CT. The analyses were carried out using the SASÒ and R software (http://www.r-project.org/, last access November 2, 2012) and the results were considered statistically significant whenever a two-sided P value below 0.05 was achieved. Results Classifications according to TP53 status and p53 functionality were practically overlapping (k index of agreement = 0.84, p < 0.0001); 6 cases with TP53 mutation retained an intact p53 functionality. Thus, in the statistical analyses we chose to classify tumors according to their functional status of p53. Table 1 shows patient and disease characteristics, both overall and according to p53 functional status. No significant association was detected between p53 functionality and the given characteristics. The Groups A (no CT) and B (induction CT) were similar in terms of patient age (p = 0.467), stage (p = 0.278), and delivered radiotherapy dose (p = 0.815), overall and according to p53 functional status. Group A Most patients (73%) presented with advanced disease (AJCC– UICC stage III and IV), consisting of T3 or T4 tumors, while according to INT classification 50% of the patients had a T3–T4 disease; typically no patient showed nodal disease. All pathologic reports showed free margins. Four patients (13%) did not receive postoperative radiation therapy either because of surgical complications or refusal. Fifty percent of patients carried a TP53 mutation, while 47% had p53 non-functional tumors (only one mutation retained an intact p53 functionality). Nineteen patients experienced a local relapse of the primary tumor, which occurred in 79% of cases within the first 2 years. Nine recurrences were detected in patients with p53 non-functional tumors and in 10 cases in p53 functional neoplasms. Two patients presented with a second primary tumor from the gastroenteric tract, 26 and 40 months after initial treatment. Median follow-up
was 120 months (interquartile range: 94–144 months); 7 patients were alive and free of disease; 2 patients had died because of a second primary tumor and 2 of unrelated causes. Group B A high percentage of patients (82%) had advanced disease, with AJCC–UICC T3 and T4 stage, while according to INT stage this percentage fell to 45%; no patient showed nodal disease. No positive surgical pathologic margins were recorded. Eight patients (18%) were not treated with postoperative radiation therapy (5 for achievement of pathologic complete response (pCR), 2 due to clinical contraindication, one because of refusal). Fifty percent of patients carried a TP53 mutation, while 39% had p53 non-functional tumors (5 cases with TP53 mutation retained an intact p53 functionality). Fourteen patients experienced a relapse (10 local, 1 local and nodal, 3 distant), occurring within 2 years in 86% of cases; 10 relapses occurred in p53 non-functional cases. Four patients presented with a second primary tumor, arising between 67 and 118 months after the initial treatment: 2 from the gastroenteric tract, 1 renal and 1 intestinal type adenocarcinoma of ethmoid sinus. This latter case was considered a second primary, because of the elapsed time (118 months) since first therapy and the different TP53 mutational status (wild-type at first diagnosis, mutated in the second tumor); it could be speculated that it was a radio-induced tumor, within the mucosa of a patient exposed to leather dust. Median follow up was 92 months (interquartile range: 76– 113 months); 24 patients were alive and free of disease, while 2 patients had died of unrelated causes. Overall survival (OS) and disease-free survival (DFS) Fig. 1 shows the OS and DFS curves in the two treatment groups. Five-year OS in Group A was 41.8% (95% confidence interval: 27.0– 64.6%), while in Group B it was 70.4% (58.1–85.3%), p = 0.041; 5year DFS in Group A was 40.0% (25.8–62.0%), while in Group B it was 65.9% (53.3–81.5%), p = 0.009. Both OS and DFS according to p53 functional status were in favor of patients carrying p53 functional tumors (p = 0.015 and 0.028, Fig. 2).
Table 1 Patient and disease characteristics overall and according to p53 functional status. Whole series
p53 Functional
p53 Non-functional
P value
74
43 (58.1%)
31 (41.9%)
–
61 (53.3–68.0)
63(55.5–69.5)
58 (51.5–65.5)
0.145
71 (95.9%) 3 (4.1%)
41 (95.3%) 2 (4.7%)
30 (96.8%) 1 (3.2%)
0.784
AJCC–UICC stage 1 2 3 4a 4b
9 (12.2%) 7 (9.5%) 28 (37.8%) 17 (23.0%) 13 (17.6%)
5 (11.6%) 5 (11.6%) 16 (37.2%) 12 (27.9%) 5 (11.6%)
4 (12.9%) 2 (6.5%) 12 (38.7%) 5 (16.1%) 8 (25.8%)
0.465
INT stage 2 3 4
39 (52.7%) 23 (31.1%) 12 (16.2%)
24 (55.8%) 15 (34.9%) 4 (9.3%)
15 (48.4%) 8 (25.8%) 8 (25.8%)
0.183
Treatment Surgery Surgery + RT Induction CT + surgery Induction CT + surgery + RT
4 (5.4%) 26 (35.1%) 8 (10.8%) 36 (48.7%)
2 (4.7%) 14 (32.6%) 7 (16.3%) 20 (46.5%)
2 (6.5%) 12 (38.7%) 1 (3.2%) 16 (51.6%)
0.374
Number of patients Age, yrs Median (IQR) Genotoxic exposition Yes No
416
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Figure 1 Kaplan–Meier overall survival (OS) curves (left panel) and disease-free survival (DFS) curves (right panel) according to treatment group (P = p value at log-rank test).
Figure 2 Kaplan–Meier overall survival (OS) curves (left panel) and disease-free survival (DFS) curves (right panel) according to p53 functionality (P = p value at log-rank test).
Table 2 shows the results of multivariable Cox model analyses of OS and DFS. The treatment prognostic effect observed at univariable analysis was confirmed at multivariable analysis. In particular, Group A prognosis was significantly worse in terms of DFS (p = 0.003), with a threefold risk of developing an unfavorable event as compared with Group B (HR = 2.92, together with a confidence interval not including one). Regarding a molecular biomarker, non-functional p53 was significantly associated with worse OS (p = 0.024), with a HR of 2.32. INT stage was significantly associated with both the end-points (p = 0.001 and < 0.001, for OS and DFS, respectively). When including AJCC–UICC stage instead of INT stage in the Cox model analyses, the results were practically
unchanged in terms of treatment effect; the treatment HR (95% confidence interval) was 1.86 (0.91, 3.80), with p = 0.091 for OS and 2.25 (1.15, 4.42), with p = 0.018 for DFS. p53 Functionality achieved statistical significance for both the end-points (p = 0.017 and 0.041, respectively) and, as well, AJCC–UICC stage was an independent prognostic factor (p = 0.028 and 0.040, respectively). Analyzing each treatment group according to functional status, it was only in Group B, where both OS and DFS were in favor of functional p53, whereas no significant differences were seen in Group A, where the patients were not treated with preoperative chemotherapy (Figs. 3 and 4). Such results were confirmed at multivariable analysis with Cox models including the interaction between treat-
417
P. Bossi et al. / Oral Oncology 49 (2013) 413–419 Table 2 Results of overall and disease-free survival analysis by multivariable Cox regression models in the whole series. OS
DFS
HR
CI
P
HR
CI
Treatment group No CT vs. induction CT
P
1.99
0.97–4.07
0.061
2.92
1.45–5.88
0.003
p53 Functionality Non-functional vs. functional
2.32
1.11–4.81
0.024
1.80
0.91–3.55
0.092
Postoperative radiotherapy No vs. yes
1.68
0.55–5.08
0.359
1.98
0.72–5.45
0.187
INT stage 4 vs. 2
5.68
2.12–15.16
0.001
7.67
2.81–20.89
<0.001
OS: overall survival; DFS: disease-free survival; HR: hazard ratio; CI: 95% HR confidence interval; P: p value at two-sided Wald test. HR is an estimate of the risk increase (if > 1) or decrease (if < 1) associated with a covariate category (or value with continuous covariates) vs. the reference category (value), which is assumed to have a risk of one. CI not including the value of one indicates a significant difference toward the reference category (value).
Figure 3 Kaplan–Meier overall survival (OS) curves according to p53 functionality in treatment Group A (left panel) and B (right panel) (P = p value at log-rank test).
ment and p53 functionality, as shown in Fig. 5. In particular, when considering p53 functional group, the absence of induction CT carried a higher risk both in terms of OS (HR = 3.33) and DFS (HR = 6.32, with the HR confidence interval not including one). On the other hand, in the same Cox model with interaction between treatment and p53 functionality, non-functional p53 was associated with a significantly worse outcome compared with functional p53 only in the induction CT group (about fourfold HRs for both OS and DFS, with confidence intervals not including the value of one). On the contrary, in the group not receiving chemotherapy no significant prognostic role for p53 functionality was highlighted (data not shown). These observations point to a predictive rather than a prognostic role of p53 functionality. Discussion In the present study we compared OS and DFS of two consecutive series of ethmoid sinus ITACs profiled for TP53, both treated with surgery followed by postoperative radiotherapy in a single Institution. Notably, in the first series (Group A) the patients were surgically treated upfront, whereas in the second series (Group
B) surgical resection was performed after primary chemotherapy (PFL regimen). The two series showed a similar distribution of staging (AJCC–UICC stage III–IV in 73% and 82% of Group A and B respectively) and a very similar rate of non-functional TP53 mutations (47% vs. 39%). The results highlighted that the series receiving PFL primary chemotherapy showed a significantly better OS (p = 0.041) and DFS (p = 0.009) compared with the untreated one, and considering the 2 series in aggregate the significance was confirmed at multivariable analysis of DFS (p = 0.003 for DFS and p = 0.061 for OS). Furthermore, the results highlighted that non-functional p53 carried a significantly worse prognosis in terms of both OS and DFS compared with functional p53, with a significance maintained at multivariable analysis of OS (p = 0.024). Considering the aim of our investigation, both analyzing the single series and the two series in aggregate in the Cox multivariable model, a survival benefit was evident for patients with p53 functional cancer only in the group receiving PFL chemotherapy, while no prognostic role could be identified when chemotherapy was not administered. The better outcome obtained with a functional p53 irrespective of the treatment given reinforces the prognostic role of p53. However, the significant correlation between functional p53 with DFS
418
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Figure 4 Kaplan–Meier disease-free survival (DFS) curves according to p53 functionality in treatment Group A (left panel) and B (right panel) (P = p value at log-rank test).
Figure 5 Treatment group hazard ratios (HR, squares) and 95% confidence intervals (continuous lines) according to p53 functional status. The estimates were obtained from multivariable Cox models for overall survival (OS) and disease-free survival (DFS) with interaction treatment and p53.
restricted to the CT induction group strongly supports a predictive rather than a prognostic role of p53 functional status. This observation could guide the management of ITAC in the choice of whether or not to perform induction PFL chemotherapy. We are aware of biases possibly introduced by comparing our two series in terms of outcome mainly because of the lack of randomization and possible confounding factors occurring within the time frame, such as improvements in surgical and radiotherapy techniques. Nonetheless, the induction CT prognostic effect was maintained at multivariable analysis, taking into account possible inhomogeneity between the groups in postoperative radiotherapy and tumor stage. Thus, we believe that the magnitude of benefit on survival rates with induction therapy mainly related to p53 functionality should not be underestimated considering the homogeneity of the analysis, performed on a single histotype and on a mono-institutional basis, in a relatively short period of time for a rare disease. Moreover, a homogeneous molecular approach and the choice to focus on protein functionality rather than simply on TP53 gene status could strengthen the results of the study. In our previous report we underlined among ITAC the predictive value of wild-type TP53 in achieving a pathologic complete remission to induction chemotherapy.31 Here, expanding the number of cases and comparing two suitable series, we reconfirmed the predictive significance of p53 as a biomarker, showing that the gained benefit is potentially ascribable to chemotherapy in patients
carrying functional p53 tumors. These findings are in line with the increasingly widespread belief in the concept that the central role p53 plays in tumor suppression is obtained by mediating the DNA damage response. Further support to the strong dependence of ITAC tumor cells on p53 function is given by the close correlation between nonfunctional p53 and poor sensitivity to chemotherapy. This phenotype, also referred to as ‘‘hypersensivity’’ for tumor suppressor genes (and as ’’oncogene addiction’’ for oncogenes) does not necessarily correlate with the prognosis and in keeping with its role may serve as a drug target mediating drug sensitivity and tumor response.37 Consistently, paclitaxel, which has been proved to be able to mediate the sensitivity of cells devoid of p53 through microtubule-associated protein 4 (MAP4)38 has been identified through a ‘‘synthetic lethal-like’’ screening searching for ‘‘anticancer drugs that are not dependent on intact p53 suppressor gene function for their activity’’.39 Moreover, metformin which by LKB1/AMPK/mTOR axis is known to induce autophagy and therefore to favor cell survival/resistance in p53 proficient cells, has been proved to induce the death of cells through metformininduced hypoxic/glucose withdrawal in p53 disabled cells.40 Very recently, metformin has been demonstrated to serve as a radiosensitizer for HNSCC with TP53 disruptive mutations by inhibiting senescence,41 known to share and closely overlap signs of autophagy.42 Notably, this drug is currently being evaluated in ongoing trials and may be exploited in this clinical setting. Cumulatively, the present preliminary data need to be confirmed on larger series, but they underline that the assessment of the functional status of p53, in addition to defining prognostic patient categories, may also be used to identify the subgroup of chemoresponsive tumors. Furthermore, our results may help in selecting patients with non-functional p53 tumors who are expected to have a worse prognosis but who might actually benefit from innovative approaches, such as the ability to lead p53 mutated cells to kill themselves through their own disability. Conflicts of interest statement None declared.
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Acknowledgments The authors acknowledge the nurses involved in patients’ care at Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy. This work was supported in part by Associazione Italiana per la Ricerca sul Cancro (S.P.) and Ministero della Salute Programma Integrato Oncologia (S.P. RFPS-2006-2-342237.5). References 1. Robin PE, Powell DJ, Stansbie JM. Carcinoma of the nasal cavity and paranasal sinuses: incidence and presentation of different histological types. Clin Otolaryngol Allied Sci 1979;4(6):431–56. 2. Sisson Sr GA, Toriumi DM, Atiyah RA. Paranasal sinus malignancy: a comprehensive update. Laryngoscope 1989;99(2):143–50. 3. Roush GC. Epidemiology of cancer of the nose and paranasal sinuses: current concepts. Head Neck Surg 1979;2(1):3–11. 4. Frazell EL, Lewis JS. Cancer of the nasal cavity and accessory sinuses. A report of the management of 416 patients. Cancer 1963;16:1293–301. 5. Cheesman AD, Lund VJ, Howard DJ. Craniofacial resection for tumors of the nasal cavity and paranasal sinuses. Head Neck Surg 1986;8(6):429–35. 6. Cantù G, Solero CL, Mariani L, et al. Intestinal type adenocarcinoma of the ethmoid sinus in wood and leather workers: a retrospective study of 153 cases. Head Neck 2011;33(4):535–42. 7. Gatta G, Bimbi G, Ciccolallo L, Zigon G, Cantù G. Survival for ethmoid sinus adenocarcinoma in European populations. Acta Oncol 2009;48(7):992–8. 8. Barnes L. Intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Am J Surg Pathol 1986;10(3):192–202. 9. Perrone F, Oggionni M, Birindelli S, et al. TP53, p14ARF, p16INK4a and H-ras gene molecular analysis in intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Int J Cancer 2003;105(2):196–203. 10. Wu TT, Barnes L, Bakker A, Swalsky PA, Finkelstein SD. K-ras-2 and p53 genotyping of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Mod Pathol 1996;9(3):199–204. 11. Holmila R, Bornholdt J, Suitiala T, et al. Profile of TP53 gene mutations in sinonasal cancer. Mutat Res 2010;686(1–2):9–14. 12. Frattini M, Perrone F, Suardi S, et al. Phenotype-genotype correlation: challenge of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Head Neck 2006;28(10):909–15. 13. Wax MK, Yun KJ, Wetmore SJ, Lu X, Kaufman HH. Adenocarcinoma of the ethmoid sinus. Head Neck 1995;17(4):303–11. 14. Klintenberg C, Olofsson J, Hellquist H, Sokjer H. Adenocarcinoma of the ethmoid sinuses. A review of 28 cases with special reference to wood dust exposure. Cancer 1984;54(3):482–8. 15. Gallo O, Franchi A, Fini-Storchi I, et al. Prognostic significance of c-erbB-2 oncoprotein expression in intestinal-type adenocarcinoma of the sinonasal tract. Head Neck 1998;20(3):224–31. 16. Jiang GL, Morrison WH, Garden AS, et al. Ethmoid sinus carcinomas: natural history and treatment results. Radiother Oncol 1998;49(1):21–7. 17. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant paranasal sinus tumors: report of an international collaborative study. Head Neck 2005;27(7):575–84. 18. Rosen A, Vokes EE, Scher N, Haraf D, Weichselbaum RR, Panje WR. Locoregionally advanced paranasal sinus carcinoma. Favorable survival with multimodality therapy. Arch Otolaryngol Head Neck Surg 1993;119(7):743–6. 19. Roux FX, Brasnu D, Devaux B, et al. Ethmoid sinus carcinomas: results and prognosis after neoadjuvant chemiotherapy and combined surgery – a 10-year experience. Surg Neurol 1994;42(2):98–104. 20. Bjork-Eriksson T, Mercke C, Petruson B, Ekholm S. Potential impact on tumor control and organ preservation with cisplatin and 5-fluorouracil for patients with advanced tumors of the paranasal sinuses and nasal fossa. A prospective pilot study. Cancer 1992;70(11):2615–20.
419
21. Knegt PP, Ah-See KW, Vd Velden LA, Kerrebijn J. Adenocarcinoma of the ethmoidal sinus complex: surgical debulking and topical fluorouracil may be the optimal treatment. Arch Otolaryngol Head Neck Surg 2001;127(2):141–6. 22. Licitra L, Locati LD, Cavina R, et al. Primary chemotherapy followed by anterior craniofacial resection and radiotherapy for paranasal cancer. Ann Oncol 2003;14(3):367–72. 23. Chen AM, Daly ME, Bucci MK, et al. Carcinomas of the paranasal sinuses and nasal cavity treated with radiotherapy at a single institution over five decades: are we making improvement? Int J Radiat Oncol Biol Phys 2007;69(1):141–7. 24. Bosch FX, Ritter D, Enders C, et al. Head and neck tumor sites differ in prevalence and spectrum of p53 alterations but these have limited prognostic value. Int J Cancer 2004;111(4):530–8. 25. Gallo O, Chiarelli I, Boddi V, Bocciolini C, Bruschini L, Porfirio B. Cumulative prognostic value of p53 mutations and bcl-2 expression in head-and-neck cancer treated by radiotherapy. Int J Cancer 1999;84(6):573–9. 26. Alsner J, Sorensen SB, Overgaard J. TP53 mutation is related to poor prognosis after radiotherapy, but not surgery, in squamous cell carcinoma of the head and neck. Radiother Oncol 2001;59(2):179–85. 27. Temam S, Flahault A, Périé S, et al. P53 gene status as a predictor of tumor response to induction chemiotherapy of patients with locoregionally advanced squamous cell carcinomas of the head and neck. J Clin Oncol 2000;18(2):385–94. 28. Cabelguenne A, Blons H, de Waziers I, et al. P53 alterations predict tumor response to neoadjuvant chemotherapy in head and neck squamous cell carcinoma: a prospective series. J Clin Oncol 2000;18(7):1465–73. 29. Poeta ML, Manola J, Goldwasser MA, et al. TP53 mutations and survival in squamous-cell carcinoma of the head and neck. N Engl J Med 2007;357(25):2552–61. 30. Perrone F, Bossi P, Cortelazzi B, et al. TP53 mutations and pathologic complete response to neoadjuvant cisplatin and fluorouracil chemotherapy in resected oral cavity squamous cell carcinoma. J Clin Oncol 2010;28(5):761–6. 31. Licitra L, Perrone F, Bossi P, et al. Prediction of TP53 status for primary cisplatin, fluorouracil, and leucovorin chemotherapy in ethmoid sinus intestinal-type adenocarcinoma. J Clin Oncol 2004;22(24):4901–6. 32. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. American joint committee on cancer staging manual. seventh ed. New York: Springer; 2009. 33. Cantù G, Riccio S, Bimbi G, et al. Craniofacial resection for malignant tumours involving the anterior skull base. Eur Arch Otorhinolaryngol 2006;263(7):647–52. 34. Licitra L, Perrone F, Bossi P, et al. High-risk human papillomavirus affects prognosis in patients with surgically treated oropharyngeal squamous cell carcinoma. J Clin Oncol 2006;24(36):5630–6. 35. Petitjean A, Mathe E, Kato S, et al. Impact of mutant p53 functional properties on TP53 mutation patterns and tumor phenotype: lessons from recent development in the IARC TP53 database. Hum Mutat 2007;28(6):622–9. 36. Cantù G, Solero CL, Miceli R, et al. Which classification for ethmoid malignant tumors involving the anterior skull base? Head Neck 2005;27(3):224–31. 37. Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogene and nononcogene addiction. Cell 2009;136(5):823–37. 38. Zhang CC, Yang JM, Bash-Babula J, et al. DNA damage increases sensitivity to vinca alkaloids and decreases sensitivity to taxanes through p53-dependent repression of microtubule-associated protein 4. Cancer Res 1999;59(15):3663–70. 39. Weinstein JN, Myers TG, O’Connor PM, et al. An information-intensive approach to the molecular pharmacology of cancer. Science 1997;275(5298):343–9. 40. Buzzai M, Jones RG, Amaravadi RK, et al. Systemic treatment with the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth. Cancer Res 2007;67(14):6745–52. 41. Skinner HD, Sandulache VC, Ow TJ, et al. TP53 disruptive mutations lead to head and neck cancer treatment failure through inhibition of radiation-induced senescence. Clin Cancer Res 2012;18(1):290–300. 42. Narita M, Young AR, Arakawa S, et al. Spatial coupling of mTOR and autophagy augments secretory phenotypes. Science 2011;332(6032):966–70.
Contents lists available at SciVerse ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Tp53 status as guide for the management of ethmoid sinus intestinal-type adenocarcinoma Paolo Bossi a,g, Federica Perrone b,⇑,g, Rosalba Miceli c, Giulio Cantù d, Luigi Mariani c, Ester Orlandi e, Carlo Fallai e, Laura D. Locati a, Barbara Cortelazzi b, Pasquale Quattrone b, Paolo Potepan f, Lisa Licitra a,h, Silvana Pilotti b,h a
Head and Neck Medical Oncology Unit, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy Laboratory of Experimental Molecular Pathology, Department of Pathology, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy Statistic Unit, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy d Department of Otolaryngology and Maxillo-facial surgery, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy e Department of Radiotherapy, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy f Department of Radiology, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy b c
a r t i c l e
i n f o
Article history: Received 7 November 2012 Accepted 24 December 2012 Available online 29 January 2013 Keywords: Intestinal-type adenocarcinoma (ITAC) TP53 Overal survival
s u m m a r y Objective: Intestinal-type adenocarcinoma (ITAC) of the ethmoid sinus is a rare, occupational-related tumor. Optimal treatment consists of surgery and radiotherapy, while chemotherapy is still investigational. The molecular profile of ITAC is characterized by the occurrence of TP53 mutations associated with genotoxic agents such as wood dust. We investigated the role of p53 functionality in relation to the primary treatment. Materials and methods: We retrospectively reviewed 100 medical charts of consecutive patients with a first diagnosis of ITAC treated at our Institute; 74 patients were evaluable for TP53 analysis. Thirty (41%) were treated from 1991 to 2006 with craniofacial resection followed by radiotherapy (Group A), compared with 44 patients (59%) treated from 1996 to 2006 with cisplatin-based induction chemotherapy (PFL) followed by standard treatment (Group B). Results: Five-year OS in Group A was 42%, while in Group B it was 70% (p = 0.041); 5-year DFS in Group A was 40%, while in Group B it was 66%, (p = 0.009) (p = 0.061 and 0.003 at Cox multivariable OS and DFS analyses). Analyzing each group according to p53 functional status, only for Group B patients (who received preoperative chemotherapy) both OS and DFS were in favor of functional p53 (p = 0.023 and p = 0.010, respectively). No impact of p53 functional status as a biomarker was observed in Group A. Conclusions: Functional p53 may predict PFL-chemotherapy efficacy, offering a possible increase in survival when induction chemotherapy is given to a selected population. On the other hand, upcoming innovative approaches should be explored in the presence of non-functional p53. Ó 2013 Elsevier Ltd. All rights reserved.
Introduction Carcinomas of the paranasal sinuses are rare, accounting for less than 1% of all cancers and for 3% of those in the upper aerodigestive tract.1–3 Among these, ethmoid sinus carcinomas represent approximately 20% of all malignant tumors.4 Adenocarcinoma is the most frequent histological type in all the European series of ethmoid sinus cancers,5,6 5 year overall survival for all stages is ⇑ Corresponding author. Address: Department of Pathology, Fondazione IRCCS Istituto Nazionale Tumori, Via Venezian, 1-20133 Milan, Italy. Tel.: +39 0223902280; fax: +39 0223902877. E-mail address: [email protected] (F. Perrone). g These authors have contributed equally to this work. h Senior coauthors. 1368-8375/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.oraloncology.2012.12.011
46%, according to a European population-based cancer registries analysis, with no improvement over time.7 Intestinal-type adenocarcinoma (ITAC) has an indisputable occupational etiology related to exposure to organic dusts, mainly wood and leather.8 Consistently, ITAC is characterized by a high occurrence of molecular alterations reported to be associated with genotoxic agents, such as TP53 mutations represented by CG:C > A:T transitions and involving the CpG dinucleotides, tumor suppressor gene promoter methylation and LOH at the 9p21 locus.9–11 There is a strong histopathologic resemblance between ITAC and colon cancer, also in the molecular profile.12 Ethmoid sinus ITACs are rarely considered in single series. In this context, natural history and prognosis of these neoplasms are difficult to determine.13,14
414
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Moreover only a few biological tumor markers with prognostic value have been studied.15 Standard treatment consists of optimal surgical resection followed by radiotherapy.16,17 Multimodality therapies comprising chemotherapy in a variety of schemes have been employed with promising results.18–22 Apparently, despite the improvement in techniques, radiotherapy as an exclusive treatment has not translated into the achievement of better survival over the last 50 years.23 The role of the TP53 mutation in head and neck squamous cell carcinoma (HNSCC) is conflicting.24 TP53 mutations have been associated with reduced survival after radiotherapy25,26 or cisplatin chemotherapy,27,28 whereas contrasting data are reported with surgery as primary treatment.26,29 Few data are reported on ITAC where, similarly to SCC of the oral cavity,30 TP53 wild type seems to be a promising predictive biomarker for pathologic complete response (pCR) to primary cisplatin-based chemotherapy; on the other hand, TP53 mutations have been reported to predict a low response rate to systemic treatment.31 All these findings suggest a predictive value of TP53 status after DNA damaging therapies. Based on these observations, we were interested in investigating the role of TP53 status in ethmoid sinus ITACs, by comparing a homogeneous series of patients who primarily underwent surgery, with a series who in addition received an induction PFLregimen. Patients and methods Patient selection, treatment and follow up We retrospectively reviewed 100 medical charts of consecutive patients with first diagnosis of ethmoid sinus ITAC all treated at our Institution from 1991 to 2006. Among this group, 26 patients were excluded from analysis: 4 because of postoperative death; 13 because the initial diagnostic biopsy was performed elsewhere and was not available for TP53 analysis; and 9 due to the unfeasibility of TP53 gene sequencing (Bouin fixed material). The patients were divided into two groups according to the treatment they received. Group A Thirty treatment-naive patients from February 1991 to December 2006 were examined. Tumor staging included physical examination, chest X-rays, head and neck computerized tomography (CT) and/or magnetic resonance imaging (MRI). Tumor staging was reviewed according to AJCC–UICC 2002 7 h classification criteria32 and INT classification.33 Treatment consisted of anterior craniofacial resection followed by postoperative radiotherapy. Up to the beginning of 2003 radiation therapy was given with 2D conventional treatment; thereafter it was performed with three-dimensional conformal RT based on fully 3D treatment planning. Field set-up to irradiate the target volume (primary tumor and operative bed), usually consisted of one anterior and two lateral or oblique 6 MV photon beams. Radiation was delivered in doses of 2 Gy/fraction at five fractions weekly, for a total dose of 50–66 Gy, according to the absence or presence of close/positive margins. No elective nodal irradiation was performed. Follow-up was conducted according to institutional policies, comprising physical examination every 3 months for the first year, every 6 months for the next 2 years and then annually; head and neck radiological exams (CT or MRI) were requested every 6 months in the first 2 years, then annually for another 3 years, while chest X-rays were performed annually.
Group B Forty-four treatment-naive patients treated from January 1996 to March 2006 with primary chemotherapy (all with PFL-regimen: cisplatin, fluorouracil and leucovorin) followed by anterior craniofacial resection and radiotherapy were considered; treatment and clinical outcome of 30 of these patients has already been reported in our previous study.31 Staging, postoperative radiation treatment and follow-up procedures were similar to those in Group A. However, in this group, the extent of the initial disease before chemotherapy was considered to define the radiation therapy target volume. Tumor specimen and molecular analysis of TP53 gene Formalin-fixed, paraffin-embedded tissue obtained before any treatment from surgical specimens or diagnostic biopsies were analyzed in Groups A and B, respectively. Pathologic exams were reviewed by the same pathologist (PQ). Genomic DNA was extracted from microdissected 7-lm methylene-blue-stained sections as previously described.9 The samples were analyzed for TP53 mutations in exons 5–8 by PCR and automated DNA sequencing (ABIprism 377, Applied Biosystem), as previously described.34 The detected mutations were confirmed at least twice, by independent amplifications and sequence reactions. All the missense mutations were functionally classified according to transactivation activities (TA) as ‘‘fully functional’’ (median TA >75% and 6140%), ‘‘partially functional’’ (median TA >20% and 675%) or ‘‘non-functional’’ (median TA 620%).35 A p53 functional protein was considered when TP53 was wildtype or carried fully or partially functional missense mutation. Statistical methods The K statistic was calculated to evaluate the agreement between TP53 status and p53 functionality. The association between continuous variables and treatment, TP53 gene status or p53 functionality was assessed using the Mann–Whitney–Wilcoxon test; the chi-square or the Fisher exact test with mid-p correction, as appropriate, were used when categorical variables were involved in testing association. The study end-points were overall survival (OS) and diseasefree survival (DFS). In all the survival analyses the initial date was taken as the date of surgery for Group A and that of starting chemotherapy for Group B. For event-free patients follow-up time was censored at 96 months (approximately the median follow-up time of the Group B patients with the shortest follow-up). OS time was calculated from initial date to the date of death due to any cause, or it was equal to the censoring time for living patients. DFS time was calculated from initial date to the date of disease relapse, second primary tumor or death without evidence of disease, whichever occurred first; DFS time was equal to censoring time for disease-free patients. OS and DFS curves were estimated by the Kaplan Meier method and compared with the log-rank test. Multivariable OS and DFS analyses were performed using multivariable Cox models including as covariates the following: treatment (no induction CT, induction CT); p53 functionality (non-functional, functional); postoperative radiotherapy (no, yes); and INT stage (as a linear trend). INT stage was chosen instead of the AJCC–UICC staging system because of its better discriminative performance.36 To explore whether the treatment prognostic effect was different for patients with non-functional p53 as compared with patients with functional p53, additional multivariable Cox models included the interaction term between treatment and p53 functionality; with such models we could also explore whether p53
415
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
functionality provided different survival according to having received or not having received induction CT. The analyses were carried out using the SASÒ and R software (http://www.r-project.org/, last access November 2, 2012) and the results were considered statistically significant whenever a two-sided P value below 0.05 was achieved. Results Classifications according to TP53 status and p53 functionality were practically overlapping (k index of agreement = 0.84, p < 0.0001); 6 cases with TP53 mutation retained an intact p53 functionality. Thus, in the statistical analyses we chose to classify tumors according to their functional status of p53. Table 1 shows patient and disease characteristics, both overall and according to p53 functional status. No significant association was detected between p53 functionality and the given characteristics. The Groups A (no CT) and B (induction CT) were similar in terms of patient age (p = 0.467), stage (p = 0.278), and delivered radiotherapy dose (p = 0.815), overall and according to p53 functional status. Group A Most patients (73%) presented with advanced disease (AJCC– UICC stage III and IV), consisting of T3 or T4 tumors, while according to INT classification 50% of the patients had a T3–T4 disease; typically no patient showed nodal disease. All pathologic reports showed free margins. Four patients (13%) did not receive postoperative radiation therapy either because of surgical complications or refusal. Fifty percent of patients carried a TP53 mutation, while 47% had p53 non-functional tumors (only one mutation retained an intact p53 functionality). Nineteen patients experienced a local relapse of the primary tumor, which occurred in 79% of cases within the first 2 years. Nine recurrences were detected in patients with p53 non-functional tumors and in 10 cases in p53 functional neoplasms. Two patients presented with a second primary tumor from the gastroenteric tract, 26 and 40 months after initial treatment. Median follow-up
was 120 months (interquartile range: 94–144 months); 7 patients were alive and free of disease; 2 patients had died because of a second primary tumor and 2 of unrelated causes. Group B A high percentage of patients (82%) had advanced disease, with AJCC–UICC T3 and T4 stage, while according to INT stage this percentage fell to 45%; no patient showed nodal disease. No positive surgical pathologic margins were recorded. Eight patients (18%) were not treated with postoperative radiation therapy (5 for achievement of pathologic complete response (pCR), 2 due to clinical contraindication, one because of refusal). Fifty percent of patients carried a TP53 mutation, while 39% had p53 non-functional tumors (5 cases with TP53 mutation retained an intact p53 functionality). Fourteen patients experienced a relapse (10 local, 1 local and nodal, 3 distant), occurring within 2 years in 86% of cases; 10 relapses occurred in p53 non-functional cases. Four patients presented with a second primary tumor, arising between 67 and 118 months after the initial treatment: 2 from the gastroenteric tract, 1 renal and 1 intestinal type adenocarcinoma of ethmoid sinus. This latter case was considered a second primary, because of the elapsed time (118 months) since first therapy and the different TP53 mutational status (wild-type at first diagnosis, mutated in the second tumor); it could be speculated that it was a radio-induced tumor, within the mucosa of a patient exposed to leather dust. Median follow up was 92 months (interquartile range: 76– 113 months); 24 patients were alive and free of disease, while 2 patients had died of unrelated causes. Overall survival (OS) and disease-free survival (DFS) Fig. 1 shows the OS and DFS curves in the two treatment groups. Five-year OS in Group A was 41.8% (95% confidence interval: 27.0– 64.6%), while in Group B it was 70.4% (58.1–85.3%), p = 0.041; 5year DFS in Group A was 40.0% (25.8–62.0%), while in Group B it was 65.9% (53.3–81.5%), p = 0.009. Both OS and DFS according to p53 functional status were in favor of patients carrying p53 functional tumors (p = 0.015 and 0.028, Fig. 2).
Table 1 Patient and disease characteristics overall and according to p53 functional status. Whole series
p53 Functional
p53 Non-functional
P value
74
43 (58.1%)
31 (41.9%)
–
61 (53.3–68.0)
63(55.5–69.5)
58 (51.5–65.5)
0.145
71 (95.9%) 3 (4.1%)
41 (95.3%) 2 (4.7%)
30 (96.8%) 1 (3.2%)
0.784
AJCC–UICC stage 1 2 3 4a 4b
9 (12.2%) 7 (9.5%) 28 (37.8%) 17 (23.0%) 13 (17.6%)
5 (11.6%) 5 (11.6%) 16 (37.2%) 12 (27.9%) 5 (11.6%)
4 (12.9%) 2 (6.5%) 12 (38.7%) 5 (16.1%) 8 (25.8%)
0.465
INT stage 2 3 4
39 (52.7%) 23 (31.1%) 12 (16.2%)
24 (55.8%) 15 (34.9%) 4 (9.3%)
15 (48.4%) 8 (25.8%) 8 (25.8%)
0.183
Treatment Surgery Surgery + RT Induction CT + surgery Induction CT + surgery + RT
4 (5.4%) 26 (35.1%) 8 (10.8%) 36 (48.7%)
2 (4.7%) 14 (32.6%) 7 (16.3%) 20 (46.5%)
2 (6.5%) 12 (38.7%) 1 (3.2%) 16 (51.6%)
0.374
Number of patients Age, yrs Median (IQR) Genotoxic exposition Yes No
416
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Figure 1 Kaplan–Meier overall survival (OS) curves (left panel) and disease-free survival (DFS) curves (right panel) according to treatment group (P = p value at log-rank test).
Figure 2 Kaplan–Meier overall survival (OS) curves (left panel) and disease-free survival (DFS) curves (right panel) according to p53 functionality (P = p value at log-rank test).
Table 2 shows the results of multivariable Cox model analyses of OS and DFS. The treatment prognostic effect observed at univariable analysis was confirmed at multivariable analysis. In particular, Group A prognosis was significantly worse in terms of DFS (p = 0.003), with a threefold risk of developing an unfavorable event as compared with Group B (HR = 2.92, together with a confidence interval not including one). Regarding a molecular biomarker, non-functional p53 was significantly associated with worse OS (p = 0.024), with a HR of 2.32. INT stage was significantly associated with both the end-points (p = 0.001 and < 0.001, for OS and DFS, respectively). When including AJCC–UICC stage instead of INT stage in the Cox model analyses, the results were practically
unchanged in terms of treatment effect; the treatment HR (95% confidence interval) was 1.86 (0.91, 3.80), with p = 0.091 for OS and 2.25 (1.15, 4.42), with p = 0.018 for DFS. p53 Functionality achieved statistical significance for both the end-points (p = 0.017 and 0.041, respectively) and, as well, AJCC–UICC stage was an independent prognostic factor (p = 0.028 and 0.040, respectively). Analyzing each treatment group according to functional status, it was only in Group B, where both OS and DFS were in favor of functional p53, whereas no significant differences were seen in Group A, where the patients were not treated with preoperative chemotherapy (Figs. 3 and 4). Such results were confirmed at multivariable analysis with Cox models including the interaction between treat-
417
P. Bossi et al. / Oral Oncology 49 (2013) 413–419 Table 2 Results of overall and disease-free survival analysis by multivariable Cox regression models in the whole series. OS
DFS
HR
CI
P
HR
CI
Treatment group No CT vs. induction CT
P
1.99
0.97–4.07
0.061
2.92
1.45–5.88
0.003
p53 Functionality Non-functional vs. functional
2.32
1.11–4.81
0.024
1.80
0.91–3.55
0.092
Postoperative radiotherapy No vs. yes
1.68
0.55–5.08
0.359
1.98
0.72–5.45
0.187
INT stage 4 vs. 2
5.68
2.12–15.16
0.001
7.67
2.81–20.89
<0.001
OS: overall survival; DFS: disease-free survival; HR: hazard ratio; CI: 95% HR confidence interval; P: p value at two-sided Wald test. HR is an estimate of the risk increase (if > 1) or decrease (if < 1) associated with a covariate category (or value with continuous covariates) vs. the reference category (value), which is assumed to have a risk of one. CI not including the value of one indicates a significant difference toward the reference category (value).
Figure 3 Kaplan–Meier overall survival (OS) curves according to p53 functionality in treatment Group A (left panel) and B (right panel) (P = p value at log-rank test).
ment and p53 functionality, as shown in Fig. 5. In particular, when considering p53 functional group, the absence of induction CT carried a higher risk both in terms of OS (HR = 3.33) and DFS (HR = 6.32, with the HR confidence interval not including one). On the other hand, in the same Cox model with interaction between treatment and p53 functionality, non-functional p53 was associated with a significantly worse outcome compared with functional p53 only in the induction CT group (about fourfold HRs for both OS and DFS, with confidence intervals not including the value of one). On the contrary, in the group not receiving chemotherapy no significant prognostic role for p53 functionality was highlighted (data not shown). These observations point to a predictive rather than a prognostic role of p53 functionality. Discussion In the present study we compared OS and DFS of two consecutive series of ethmoid sinus ITACs profiled for TP53, both treated with surgery followed by postoperative radiotherapy in a single Institution. Notably, in the first series (Group A) the patients were surgically treated upfront, whereas in the second series (Group
B) surgical resection was performed after primary chemotherapy (PFL regimen). The two series showed a similar distribution of staging (AJCC–UICC stage III–IV in 73% and 82% of Group A and B respectively) and a very similar rate of non-functional TP53 mutations (47% vs. 39%). The results highlighted that the series receiving PFL primary chemotherapy showed a significantly better OS (p = 0.041) and DFS (p = 0.009) compared with the untreated one, and considering the 2 series in aggregate the significance was confirmed at multivariable analysis of DFS (p = 0.003 for DFS and p = 0.061 for OS). Furthermore, the results highlighted that non-functional p53 carried a significantly worse prognosis in terms of both OS and DFS compared with functional p53, with a significance maintained at multivariable analysis of OS (p = 0.024). Considering the aim of our investigation, both analyzing the single series and the two series in aggregate in the Cox multivariable model, a survival benefit was evident for patients with p53 functional cancer only in the group receiving PFL chemotherapy, while no prognostic role could be identified when chemotherapy was not administered. The better outcome obtained with a functional p53 irrespective of the treatment given reinforces the prognostic role of p53. However, the significant correlation between functional p53 with DFS
418
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Figure 4 Kaplan–Meier disease-free survival (DFS) curves according to p53 functionality in treatment Group A (left panel) and B (right panel) (P = p value at log-rank test).
Figure 5 Treatment group hazard ratios (HR, squares) and 95% confidence intervals (continuous lines) according to p53 functional status. The estimates were obtained from multivariable Cox models for overall survival (OS) and disease-free survival (DFS) with interaction treatment and p53.
restricted to the CT induction group strongly supports a predictive rather than a prognostic role of p53 functional status. This observation could guide the management of ITAC in the choice of whether or not to perform induction PFL chemotherapy. We are aware of biases possibly introduced by comparing our two series in terms of outcome mainly because of the lack of randomization and possible confounding factors occurring within the time frame, such as improvements in surgical and radiotherapy techniques. Nonetheless, the induction CT prognostic effect was maintained at multivariable analysis, taking into account possible inhomogeneity between the groups in postoperative radiotherapy and tumor stage. Thus, we believe that the magnitude of benefit on survival rates with induction therapy mainly related to p53 functionality should not be underestimated considering the homogeneity of the analysis, performed on a single histotype and on a mono-institutional basis, in a relatively short period of time for a rare disease. Moreover, a homogeneous molecular approach and the choice to focus on protein functionality rather than simply on TP53 gene status could strengthen the results of the study. In our previous report we underlined among ITAC the predictive value of wild-type TP53 in achieving a pathologic complete remission to induction chemotherapy.31 Here, expanding the number of cases and comparing two suitable series, we reconfirmed the predictive significance of p53 as a biomarker, showing that the gained benefit is potentially ascribable to chemotherapy in patients
carrying functional p53 tumors. These findings are in line with the increasingly widespread belief in the concept that the central role p53 plays in tumor suppression is obtained by mediating the DNA damage response. Further support to the strong dependence of ITAC tumor cells on p53 function is given by the close correlation between nonfunctional p53 and poor sensitivity to chemotherapy. This phenotype, also referred to as ‘‘hypersensivity’’ for tumor suppressor genes (and as ’’oncogene addiction’’ for oncogenes) does not necessarily correlate with the prognosis and in keeping with its role may serve as a drug target mediating drug sensitivity and tumor response.37 Consistently, paclitaxel, which has been proved to be able to mediate the sensitivity of cells devoid of p53 through microtubule-associated protein 4 (MAP4)38 has been identified through a ‘‘synthetic lethal-like’’ screening searching for ‘‘anticancer drugs that are not dependent on intact p53 suppressor gene function for their activity’’.39 Moreover, metformin which by LKB1/AMPK/mTOR axis is known to induce autophagy and therefore to favor cell survival/resistance in p53 proficient cells, has been proved to induce the death of cells through metformininduced hypoxic/glucose withdrawal in p53 disabled cells.40 Very recently, metformin has been demonstrated to serve as a radiosensitizer for HNSCC with TP53 disruptive mutations by inhibiting senescence,41 known to share and closely overlap signs of autophagy.42 Notably, this drug is currently being evaluated in ongoing trials and may be exploited in this clinical setting. Cumulatively, the present preliminary data need to be confirmed on larger series, but they underline that the assessment of the functional status of p53, in addition to defining prognostic patient categories, may also be used to identify the subgroup of chemoresponsive tumors. Furthermore, our results may help in selecting patients with non-functional p53 tumors who are expected to have a worse prognosis but who might actually benefit from innovative approaches, such as the ability to lead p53 mutated cells to kill themselves through their own disability. Conflicts of interest statement None declared.
P. Bossi et al. / Oral Oncology 49 (2013) 413–419
Acknowledgments The authors acknowledge the nurses involved in patients’ care at Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy. This work was supported in part by Associazione Italiana per la Ricerca sul Cancro (S.P.) and Ministero della Salute Programma Integrato Oncologia (S.P. RFPS-2006-2-342237.5). References 1. Robin PE, Powell DJ, Stansbie JM. Carcinoma of the nasal cavity and paranasal sinuses: incidence and presentation of different histological types. Clin Otolaryngol Allied Sci 1979;4(6):431–56. 2. Sisson Sr GA, Toriumi DM, Atiyah RA. Paranasal sinus malignancy: a comprehensive update. Laryngoscope 1989;99(2):143–50. 3. Roush GC. Epidemiology of cancer of the nose and paranasal sinuses: current concepts. Head Neck Surg 1979;2(1):3–11. 4. Frazell EL, Lewis JS. Cancer of the nasal cavity and accessory sinuses. A report of the management of 416 patients. Cancer 1963;16:1293–301. 5. Cheesman AD, Lund VJ, Howard DJ. Craniofacial resection for tumors of the nasal cavity and paranasal sinuses. Head Neck Surg 1986;8(6):429–35. 6. Cantù G, Solero CL, Mariani L, et al. Intestinal type adenocarcinoma of the ethmoid sinus in wood and leather workers: a retrospective study of 153 cases. Head Neck 2011;33(4):535–42. 7. Gatta G, Bimbi G, Ciccolallo L, Zigon G, Cantù G. Survival for ethmoid sinus adenocarcinoma in European populations. Acta Oncol 2009;48(7):992–8. 8. Barnes L. Intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Am J Surg Pathol 1986;10(3):192–202. 9. Perrone F, Oggionni M, Birindelli S, et al. TP53, p14ARF, p16INK4a and H-ras gene molecular analysis in intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Int J Cancer 2003;105(2):196–203. 10. Wu TT, Barnes L, Bakker A, Swalsky PA, Finkelstein SD. K-ras-2 and p53 genotyping of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Mod Pathol 1996;9(3):199–204. 11. Holmila R, Bornholdt J, Suitiala T, et al. Profile of TP53 gene mutations in sinonasal cancer. Mutat Res 2010;686(1–2):9–14. 12. Frattini M, Perrone F, Suardi S, et al. Phenotype-genotype correlation: challenge of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Head Neck 2006;28(10):909–15. 13. Wax MK, Yun KJ, Wetmore SJ, Lu X, Kaufman HH. Adenocarcinoma of the ethmoid sinus. Head Neck 1995;17(4):303–11. 14. Klintenberg C, Olofsson J, Hellquist H, Sokjer H. Adenocarcinoma of the ethmoid sinuses. A review of 28 cases with special reference to wood dust exposure. Cancer 1984;54(3):482–8. 15. Gallo O, Franchi A, Fini-Storchi I, et al. Prognostic significance of c-erbB-2 oncoprotein expression in intestinal-type adenocarcinoma of the sinonasal tract. Head Neck 1998;20(3):224–31. 16. Jiang GL, Morrison WH, Garden AS, et al. Ethmoid sinus carcinomas: natural history and treatment results. Radiother Oncol 1998;49(1):21–7. 17. Ganly I, Patel SG, Singh B, et al. Craniofacial resection for malignant paranasal sinus tumors: report of an international collaborative study. Head Neck 2005;27(7):575–84. 18. Rosen A, Vokes EE, Scher N, Haraf D, Weichselbaum RR, Panje WR. Locoregionally advanced paranasal sinus carcinoma. Favorable survival with multimodality therapy. Arch Otolaryngol Head Neck Surg 1993;119(7):743–6. 19. Roux FX, Brasnu D, Devaux B, et al. Ethmoid sinus carcinomas: results and prognosis after neoadjuvant chemiotherapy and combined surgery – a 10-year experience. Surg Neurol 1994;42(2):98–104. 20. Bjork-Eriksson T, Mercke C, Petruson B, Ekholm S. Potential impact on tumor control and organ preservation with cisplatin and 5-fluorouracil for patients with advanced tumors of the paranasal sinuses and nasal fossa. A prospective pilot study. Cancer 1992;70(11):2615–20.
419
21. Knegt PP, Ah-See KW, Vd Velden LA, Kerrebijn J. Adenocarcinoma of the ethmoidal sinus complex: surgical debulking and topical fluorouracil may be the optimal treatment. Arch Otolaryngol Head Neck Surg 2001;127(2):141–6. 22. Licitra L, Locati LD, Cavina R, et al. Primary chemotherapy followed by anterior craniofacial resection and radiotherapy for paranasal cancer. Ann Oncol 2003;14(3):367–72. 23. Chen AM, Daly ME, Bucci MK, et al. Carcinomas of the paranasal sinuses and nasal cavity treated with radiotherapy at a single institution over five decades: are we making improvement? Int J Radiat Oncol Biol Phys 2007;69(1):141–7. 24. Bosch FX, Ritter D, Enders C, et al. Head and neck tumor sites differ in prevalence and spectrum of p53 alterations but these have limited prognostic value. Int J Cancer 2004;111(4):530–8. 25. Gallo O, Chiarelli I, Boddi V, Bocciolini C, Bruschini L, Porfirio B. Cumulative prognostic value of p53 mutations and bcl-2 expression in head-and-neck cancer treated by radiotherapy. Int J Cancer 1999;84(6):573–9. 26. Alsner J, Sorensen SB, Overgaard J. TP53 mutation is related to poor prognosis after radiotherapy, but not surgery, in squamous cell carcinoma of the head and neck. Radiother Oncol 2001;59(2):179–85. 27. Temam S, Flahault A, Périé S, et al. P53 gene status as a predictor of tumor response to induction chemiotherapy of patients with locoregionally advanced squamous cell carcinomas of the head and neck. J Clin Oncol 2000;18(2):385–94. 28. Cabelguenne A, Blons H, de Waziers I, et al. P53 alterations predict tumor response to neoadjuvant chemotherapy in head and neck squamous cell carcinoma: a prospective series. J Clin Oncol 2000;18(7):1465–73. 29. Poeta ML, Manola J, Goldwasser MA, et al. TP53 mutations and survival in squamous-cell carcinoma of the head and neck. N Engl J Med 2007;357(25):2552–61. 30. Perrone F, Bossi P, Cortelazzi B, et al. TP53 mutations and pathologic complete response to neoadjuvant cisplatin and fluorouracil chemotherapy in resected oral cavity squamous cell carcinoma. J Clin Oncol 2010;28(5):761–6. 31. Licitra L, Perrone F, Bossi P, et al. Prediction of TP53 status for primary cisplatin, fluorouracil, and leucovorin chemotherapy in ethmoid sinus intestinal-type adenocarcinoma. J Clin Oncol 2004;22(24):4901–6. 32. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. American joint committee on cancer staging manual. seventh ed. New York: Springer; 2009. 33. Cantù G, Riccio S, Bimbi G, et al. Craniofacial resection for malignant tumours involving the anterior skull base. Eur Arch Otorhinolaryngol 2006;263(7):647–52. 34. Licitra L, Perrone F, Bossi P, et al. High-risk human papillomavirus affects prognosis in patients with surgically treated oropharyngeal squamous cell carcinoma. J Clin Oncol 2006;24(36):5630–6. 35. Petitjean A, Mathe E, Kato S, et al. Impact of mutant p53 functional properties on TP53 mutation patterns and tumor phenotype: lessons from recent development in the IARC TP53 database. Hum Mutat 2007;28(6):622–9. 36. Cantù G, Solero CL, Miceli R, et al. Which classification for ethmoid malignant tumors involving the anterior skull base? Head Neck 2005;27(3):224–31. 37. Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogene and nononcogene addiction. Cell 2009;136(5):823–37. 38. Zhang CC, Yang JM, Bash-Babula J, et al. DNA damage increases sensitivity to vinca alkaloids and decreases sensitivity to taxanes through p53-dependent repression of microtubule-associated protein 4. Cancer Res 1999;59(15):3663–70. 39. Weinstein JN, Myers TG, O’Connor PM, et al. An information-intensive approach to the molecular pharmacology of cancer. Science 1997;275(5298):343–9. 40. Buzzai M, Jones RG, Amaravadi RK, et al. Systemic treatment with the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth. Cancer Res 2007;67(14):6745–52. 41. Skinner HD, Sandulache VC, Ow TJ, et al. TP53 disruptive mutations lead to head and neck cancer treatment failure through inhibition of radiation-induced senescence. Clin Cancer Res 2012;18(1):290–300. 42. Narita M, Young AR, Arakawa S, et al. Spatial coupling of mTOR and autophagy augments secretory phenotypes. Science 2011;332(6032):966–70.