The maxilla may be affected by squamous cell carcinoma (SCC) from both oral and sinus sites. We sought to determine whether the site of origin of the maxillary tumour, oral as compared to sinus, influences survival. Univariate Kaplan–Meier and multivariate Cox proportional hazard models analysis of 58 patients with SCC involving the maxilla, treated with curative intent, was conducted. The overall 5-year disease-free survival for the group was 41.7%. Five-year disease-free survival for oral subsite SCC was 56.8%, while for sinus subsite was only 21.6%. Univariate analysis found SCC of sinus origin to be associated with a poorer prognosis, however this was not confirmed on multivariate analysis. T-stage and positive margins were found to be the only independent risk factors. For SCC of the maxilla, sinus origin of the tumour per se does not confer a poorer prognosis; however, as a result of the complex anatomy of the midface, these tumours can present at an advanced stage, while surgical control of the disease can be more difficult, especially posteriorly. Tumour size and positive margins were the determinants of a poor prognosis in this group of patients with maxillary SCC.
The maxilla can be affected by malignant tumours of both oral and sinus origin. Squamous cell carcinoma (SCC) involving the maxilla is relatively uncommon compared with other head and neck subsites, with an incidence of oral SCC involving the palate or alveolar ridges found to be between 0.5% and 3%. Paranasal sinus or sino-nasal carcinomas account for approximately 3% of head and neck malignancies, with the maxillary sinus being the most frequently involved site.
The paranasal sinuses are anatomically complex, consisting of a number of interconnecting spaces, the largest of which is the maxillary sinus. The average volume of the maxillary sinus in young adults is approximately 25 ml for males and 15 ml for females, and the volume gradually decreases with age. This volume may potentially allow a tumour to grow to a significant size before symptoms develop.
Patients with tumours involving the maxilla may present with a variety of symptoms; these include intraoral or facial swelling, facial pain, ulceration or a non-healing wound (such as an extraction socket), epistaxis, nasal obstruction, nasal discharge, and cheek paraesthesia. The maxilla is a key component of the bony midface, which also includes the zygoma, inferior conchae, nasal bones, and vomer. The midface lies adjacent to the oral cavity and orbits and communicates with the base of the skull, oropharynx, and nasopharynx. These anatomical relationships affect the presentation, management, and outcomes of tumours involving the maxilla.
Intraoral SCC involving the maxilla may spread directly from the oral mucosa into either underlying bone or into surrounding tissue of the oral cavity, soft palate, and oropharynx. In contrast, tumours arising within the maxillary air sinus may spread via foramina or directly invade several anatomical sites, including the nasal cavity, orbits, base of the skull, and infratemporal fossa ( Fig. 1 ).
Sino-nasal SCC has previously been reported as having a poor prognosis, while oral cavity SCC affecting the maxilla has a much better prognosis. However, little direct comparison has been made in the past between the two sites of origin for maxillary SCC, oral and sinus, in the same patient population. Thus, the aims of this study were to compare the prognosis of patients with SCC involving the maxilla from oral sites with those of maxillary sinus origin, treated surgically with curative intent at a single institution, and to examine factors that may affect the prognosis of these patients.
A retrospective review of patients treated for SCC involving the maxilla between 1990 and 2005 was conducted. Sixty-three patients with SCC were identified for inclusion in the study. However, complete records were available for only 58 of these patients for inclusion in the analysis. Prior to inclusion in the final analysis of data for this study, the histopathology was reviewed and confirmed as SCC. Any patient found to have another pathology on final histological examination, or for whom histology could not be confirmed as SCC, was excluded. This project was carried out with human research ethics committee approval.
Data collected included: age at time of resection, gender, TNM stage, smoking and alcohol history, site of involvement, surgical margin status, the presence of perineural/lymphovascular invasion, and adjuvant therapy. For the purposes of this study, TNM staging was based on the American Joint Committee on Cancer AJCC Cancer Staging Manual 6th edition classification for oral SCC and maxillary sinus SCC.
A standard treatment protocol for all patients with biopsy proven SCC of either oral or sinus subsites included clinical patient assessment as well as computed tomography (CT) of the head, neck, and chest. When indicated, magnetic resonance imaging (MRI) of the head and neck was also performed as part of the initial assessment in selected patients. Also for those patients with clinically or radiographically suspicious nodes that did not meet formal radiographic diagnostic criteria, fine needle aspiration was also performed to determine or confirm the presence of nodal disease.
Patients deemed to have resectable disease and who were fit for surgery were offered surgical resection as the primary treatment modality for both sinus and oral cavity cancers, in the first instance. Patients who declined surgery, were not medically fit for surgery, or had unresectable disease were then assessed for chemo/radiotherapy as the primary treatment, with either palliative or curative intent. Following surgery, those patients with advanced disease and those who had histologically proven cervical metastasis, perineural/lymphovascular invasion, extracapsular nodal spread, or positive/close margins, underwent adjuvant radiotherapy to the primary tumour site and neck. Close margins were considered to be those that display tumour within 5 mm from the resection margin. Only patients treated surgically with curative intent were included in the current study.
Data were collected using a computer spreadsheet. The statistical analysis was performed using Statview 5.0 (Statview, Cary, NC, USA). Univariate and multivariate analysis was performed to assess disease-free survival between oral origin and sinus origin maxillary SCC and factors affecting survival. The univariate assessment was accomplished for all factors using Kaplan–Meier analysis with log rank tests to assess differences between factors. The Cox proportional hazards model was used to assess independent risk factors on a multivariate level for those univariate factors that were found to be significant. A P -value of less than 0.05 was considered to be significant.