RISK FACTORS — A number of factors are linked epidemiologically and etiologically to paranasal sinus cancer.

Occupational or medical exposures — Medical and occupational exposure to thorotrast, nickel, chromium, isopropyl oils, volatile hydrocarbons, organic fibers found in the wood, shoe, and textile industries, and possibly formaldehyde have all been associated with paranasal sinus cancer . Intestinal-type adenocarcinoma of the ethmoid sinus is particularly associated with exposure to wood, leather, textile, cereal, or cement dust and is the most common histologic type of ethmoid tumor in several European series.

Air pollution — Increasing outdoor air pollution may be associated with a higher incidence of nasal and paranasal sinus cancer. In one series of 256 patients living in two large Mexican urban, highly polluted cities, the incidence of nasal and paranasal sinus cancer increased twofold between 1993 and 1998 in one hospital, and almost fivefold during the years 1995 to 1997 as compared with 1976 and 1986 in a second hospital . These increases in incidence of nasal and paranasal sinus tumors were otherwise unexplained.

Tobacco smoke — Exposure to tobacco smoke has been associated with an increased risk of nasal and sinus cancer. In one case-control study, heavy or long-term smokers had double the risk of developing nasal cancer as compared with nonsmokers. Spouses of smokers also had a significantly elevated risk of developing nasal cavity and paranasal sinus cancer.

Viral exposure — Human papilloma viruses (HPV) 6,11,16, and 18 have been detected in squamous carcinomas of the paranasal sinuses .Although the etiologic role of HPV infection in the development of invasive paranasal sinus cancer is unclear, it may be involved in the malignant degeneration of inverted papilloma of the paranasal sinus, a rare, usually benign condition. There is little evidence of a role for Epstein-Barr virus (EBV) in sinonasal carcinomas in contrast to nasopharyngeal cancer, where EBV appears to be an important etiologic agent.

Chronic sinusitis — At least one study suggests that chronic sinusitis is a risk factor for paranasal sinus cancer with a 2.3-fold increased risk compared with the general population.

PATHOLOGY — Squamous cell carcinomas (SCCs) constitute the majority of paranasal malignancies in most American series (45 to 80 percent of cases). This is followed by adenocarcinoma and mucoepidermoid carcinomas. Rarer tumors include adenocystic carcinoma, angiosarcomas, rhabdomyosarcomas, lymphomas, esthesioneuroblastomas (olfactory neuroblastomas), melanomas, and meningiomas .Multiple synchronous cancers involving the sinuses and nasal cavity may be seen.

CLINICAL PRESENTATION — Paranasal sinus cancers are usually well-differentiated and slow-growing. Early neoplasms tend to be asymptomatic until they invade adjacent structures; diagnosis is often delayed. The limited anatomic access of the paranasal sinuses makes early diagnosis difficult, and most patients have advanced disease by the time of diagnosis .

One exception is tumors involving the infrastructure of the maxillary sinus. The maxillary sinus is divided into a suprastructure and an infrastructure by Ohngren's line. This line extends from the inner canthus to the angle of the mandible. Infrastructure tumors tend to present earlier and are easier to resect. This is probably the result of earlier obstruction of sinus drainage.

The most common symptoms in patients paranasal sinus cancer include facial or dental pain, nasal obstruction, and epistaxis. Although earlier symptoms of nasal discharge and epistaxis are typically disregarded by the patient and physician., when they occur in a patients over 40, physical and radiographic investigation should be pursued.

Oral symptoms (eg, ill-fitting dentures) occur in 25 to 30 percent of patients. Pain with unilateral nasal obstruction, or ocular symptoms can be seen in 50 and 25 percent of patients with antral-ethmoidal disease, respectively. A classic triad of facial asymmetry, palpable/visible tumor in the oral cavity, and visible intranasal tumor occurs in 40 to 60 percent of patients with advanced disease. At least one of these signs is present in 90 percent of cases.

As disease progresses, symptoms and signs depend upon the involved site. The bony structures between the nasal cavity, sinuses, orbits and cranial vaults are thin and offer little resistance to cancer spread.

   In the ethmoid sinus, locally advanced lesions may extend into the anterior cranial fossa via the cribriform plate or into the orbit through the lamina papyracea. This may result in anosmia or displacement (typically upward and/or outward) of the globe.

   In the sphenoid sinus, disease may directly extend through the lateral bony wall into the cavernous sinus. It may also invade the middle cranial fossa directly or via the infraorbital nerve. Affected patients may complain of diplopia, blurred vision, proptosis, paresthesias in the distribution of the trigeminal nerve or trismus if the pterygoid musculature is invaded. Inferior extension into the oral cavity may also cause painful loose teeth.

Regional nodal metastases are uncommon, occurring in only 10 to 20 percent of patients. The incidence of lymph node involvement increases as tumors extend locally to adjacent sites, especially with extension into the oral cavity. The retropharyngeal nodes comprise the first echelon lymphatic drainage for sinus malignancies. Other regional nodes that may be involved with lymphatic spread are the periparotid and level 1b nodes.

DIAGNOSIS AND STAGING — A detailed history and physical examination should include close attention to the signs and symptoms of orbital extension and/or cranial nerve involvement. Endoscopic examination (rhinoscopy and nasopharyngoscopy) is necessary to assess the local extent of disease spread and to provide biopsy material. Biopsy of lesions involving the maxillary sinus can be obtained intraoperatively via a Caldwell-Luc approach (through the anterior maxilla wall). In the outpatient clinic, biopsy can be performed intranasally (under the inferior turbinates) or through the gingivo-buccal sulcus. Ethmoid sinus lesions are biopsied through an endoscopic or transnasal approach in the operating room as are frontal sinus malignancies via the frontal recess. Frontal sinus trephination is rarely necessary.

CT and magnetic resonance imaging (MRI) are complementary in evaluating disease extent.CT with thin section cuts (3 mm or less), obtained in both axial and coronal projections, is most useful for evaluating tumor invasion into bony structures such as the floor of the orbit and the infratemporal fossa . CT scans are also useful for delineating extent into nasal cavity, medial canthus, oral cavity, and nasopharynx. MRI offers a better assessment of tumor infiltration into adjacent soft tissues such as dura and cavernous sinus. Sagittal MRI views permit optimal assessment of the cribriform plate, base of the skull, and floor of the middle cranial fossa.

The most recent 2002 edition of the TNM staging system of the American Joint Committee on Cancer (AJCC) and the Union Internationale Contre Cancer (UICC) defines separate T-staging for maxillary and ethmoid sinus tumors .The regional node (N) and distant metastasis (M) staging are identical to other sites within the upper aerodigestive tract.No staging system is in place for frontal or sphenoid sinus tumors.

An alternative staging system for primary tumors of the nasal cavity and paranasal sinuses has been proposed.Compared with the AJCC/UICC staging classification, it is applicable to all cancers of the paranasal sinuses as well as the nasal mucosa. Prognosis appears to be reliably estimated based upon tumor stage, and the prediction of mortality risk for T4 lesions may be better than with the AJCC/UICC staging classification. However, most physicians in the United States continue to use the AJCC system.

TREATMENT — The optimal management for carcinoma of the paranasal sinuses is not known. Because of the rarity of these tumors, randomized treatment trials have been impossible to perform. Virtually all of the literature to date consists of retrospective reports from single institutions that employ varying treatment regimens.

Surgery with and without adjuvant RT — Traditionally, surgical resection has been the primary treatment modality for the management of paranasal sinus cancer involving the maxillary or ethmoid sinuses. However, the limitations of surgery alone are obvious given the frequent presentation of advanced disease.

Surgical technique and the overall approach to management of sinus cancers has evolved to incorporate into decision making both histology, and tumor size as well as and location in relation to the adjacent critical structures. In many cases of maxillary and ethmoid sinus SCC, for example, aggressive local therapy includes en bloc resection with or without orbital exenteration, followed by reconstruction and postoperative (adjuvant) radiotherapy (RT).

  Management of the eye — Management of the eye has been controversial in the past. Some surgeons advocate orbital exenteration when the tumor transgresses the orbital floor and/or the periorbita (dense periosteum around the globe that acts as a barrier to spread of cancer). Orbital exenteration is not pursued for patients with advanced sinus malignancy that does not transgress the dense periorbita. Studies have shown that in these cases, orbital exenteration does not improve survival.

However, orbital preservation is oncologically safe and clearly worthwhile to maintain a functional eye, even in the presence of orbital invasion .One study analyzed 256 patients with maxillary sinus tumors that invaded the orbit. All patients had conservative surgery followed by RT and chemotherapy. Five-year survival was 68 percent, but was significantly lower in those with nonsquamous histology. The authors advocate multimodality therapy and orbital preservation to improve function and survival whenever possible This approach may have less survival impact in those with nonsquamous histologies.

  Reconstruction — Improved quality of life (QOL) through functional reconstruction has gained exposure and acceptance as the standard of care over the past five years. Initially, surgeons sought to clearly delineate and classify unique defects resulting from sinus cancer surgery.Defect classification then set the stage for the methodical development of functional and aesthetic reconstructions.

QOL parameters have been studied in patients undergoing different methods of orodental rehabilitation for defects of the complex palatomaxillary area. QOL outcomes are superior for those who undergo free-tissue reconstructions (eg, nonvascularized calvarial bone grafts with a soft tissue free or pedicled temporalis flap .as compared to defect-matched patients whose rehabilitation includes only a prosthetic obturator. Others have shown that adaptability to a denture, thus preserving mastication, is a critical part of maintaining QOL in patients with maxillary defects as a result of treatment for sinus cancer.Prosthetic and prosthodontic considerations are discussed in more detail elsewhere. .

  Adjuvant RT — Regardless of the surgical margin status, adjuvant postoperative RT optimizes local control. However, even with aggressive surgery and adjuvant RT, the results of treatment for most paranasal sinus cancers is poor with local control rates from 50 to 60 percent, and five-year survival rates ranging from 30 to 60 percent.

Preoperative RT has been explored as a means of making these lesions more amenable to surgical resection. However, given the inherent bias in these nonrandomized studies, it is unclear whether preoperative is superior to postoperative RT in enhancing local control and improving outcome.

  Ethmoid sinus tumors — For cancers involving the ethmoid sinus, the surgical technique depends upon the location and size of the lesion. For smaller lesions, a medial maxillectomy is generally performed with an en bloc ethmoidectomy. However, more extensive surgery (eg, a craniofacial resection via craniotomy, or a transglabellar/subcranial approach is needed if tumor involves the fovea ethmoidalis or the cribriform plate). This latter technique involves resection of the entire ethmoid block with an osteotomy that encompasses the roofs of both ethmoid and cribriform plate regions; it is coupled with advanced reconstructive techniques. Postoperative RT is generally recommended.

The range of treatment results is broad, and depends upon the stage and severity of disease. Overall survival is approximately 52 and 74 percent, and 49 and 70 percent at three and five years, respectively for all stages of surgically treated patients:

   A study of 91 patients with T2-4 ethmoid sinus cancer who underwent anterior craniofacial resection included 42 previously untreated, and 49 locally recurrent tumors . In the previously untreated group, three year disease-free (DFS) and overall survival rates for T2, 3 and 4 tumors were 55, 38 and 18 percent, and 75, 52, and 45 percent, respectively. For patients with locally recurrent disease, the corresponding DFS and overall survival rates were 30, 18 and 5 percent and 55, 30 and 22 percent, respectively.

   Outcome may be closely related to the extent of cribriform plate invasion . This was illustrated in a report of 41 patients with ethmoid sinus tumors involving the cribriform plate or anterior fossa, or that were located near the upper roof of the ethmoid sinus. Patients were treated with a combined transfacial and neurosurgical approach followed by postoperative RT; 21 also received preoperative (induction) chemotherapy with fluorouracil (5-FU), cisplatin and epirubicin.

Four of 14 patients without radiographic evidence of cribriform plate involvement had invasion of this structure at exploration. Significant operative complications developed in 16 of the 41 patients, including cerebrospinal fluid (CSF) leak, osteomyelitis, or meningitis. For all patients, the three and five-year survival rates were 53 and 36 percent. Survival was significantly better among patients whose tumors were close to the cribriform plate but without radiographic involvement, compared with patients with radiographic evidence of cribriform involvement (three year survival 83 versus 38 percent; five year survival 50 and 32 percent, respectively).

   Adenocarcinoma — Adenocarcinomas of the ethmoid sinus complex have been successfully treated with topical application of 5-FU after surgical debulking. In a series of 62 consecutive patients undergoing debulking and repeated cycles of topical 5-FU and necrotomy, the adjusted DFS rates at two and five years were 96 and 87 percent, respectively.

  Maxillary sinus tumors — T1 or T2 lesions of the maxillary sinus can be successfully excised using true maxillectomy in which the antral walls are removed without sacrificing the orbital periosteum, zygoma or the palatal mucosa. An extended or radical maxillectomy is used for more advanced lesions. The orbit is spared whenever possible.

A large single institution retrospective analysis included 149 patients with carcinoma of the maxillary sinus treated with radical surgery and postoperative RT (55 to 60 Gy in six weeks).The five-year actuarial overall and cancer-specific survival rates were 36 and 42 percent, respectively. For patients with stage II, III, or IV disease, survival rates were 75, 36, and 11 percent, respectively.

Contraindications for surgery of maxillary and ethmoid sinus cancers include tumor extension to the lateral base of the skull and intracranial contents.Most surgeons believe that invasion into the cavernous sinus stands as an additional contraindication. Possible sequelae of surgical treatment include meningitis, hemorrhage, wound infection, CSF leak, trismus, and total blindness. Perioperative mortality rates range from 0 to 13 percent.

Multimodality treatment — As in other head and neck cancers, the incorporation of both adjuvant chemotherapy and RT into treatment has produced encouraging results and has allowed orbital preservation in many cases .

The success of multimodality approaches for paranasal sinus cancers can be illustrated by the following:

   In one report, 12 patients with locally advanced epithelial nonadenocarcinoma cancers of the paranasal sinuses and nasal fossa were treated with three cycles of cisplatin and infusional 5-FU followed by preoperative RT (48 Gy) and limited surgery. The objective response rate from preoperative chemotherapy was approximately 70 percent with no clinical complete responses, although histopathologic analysis of resected specimens showed no viable tumor in eight patients, and minimal microscopic disease in three. Local control was achieved in 11 of 12 patients, and ten were alive and recurrence-free at a mean follow-up of 27 months.

   In a second study induction 5-FU and cisplatin was followed by surgical resection and postoperative RT with or without concomitant chemotherapy in 12 consecutive patients with stage III or IV, newly diagnosed paranasal sinus carcinoma.Eleven were alive and free of disease at a median follow-up of 55 months (range, 13 to 105 months).

   These same authors published a second series of 19 consecutive patients with stage III and IV locoregionally advanced paranasal sinus carcinoma. Sixteen patients received induction chemotherapy consisting of three cycles of cisplatin and 5-FU, followed by traditional resection (14 patients) or debulking (two patients). Surgery was followed by concomitant chemoradiotherapy with hydroxyurea and 5-fluorouracil in 15 patients and standard RT in one. Three other patients underwent surgical resection followed by concomitant chemoradiotherapy. For the entire group undergoing multimodality therapy, the DFS and overall survival rates at five years were 73 and 67 percent; 54 percent were still alive at ten years.

   In 22 patients with ethmoid sinus cancer invading the skull base (usually considered inoperable), preoperative cisplatin-based chemotherapy was followed by en bloc resection via a combined neurosurgical and transfacial approach. The three year rates of survival, local control, nodal recurrence, and distant metastasis were 68, 66, 5, and 10 percent, respectively.

   Intraarterial chemotherapy with 5-FU, bleomycin, methotrexate and cisplatin has been used to increase tumor drug exposure while reducing systemic toxicity. Studies suggest that intraarterial therapy with RT may achieve results comparable to those with surgery and RT, particularly with maxillary sinus tumors.These therapies might be particularly useful in patients who would otherwise need orbital exenteration. In one series of 75 patients with maxillary sinus cancer, the use of this combined modality approach was associated with five and ten-year survival rates of 76 and 66 percent, respectively; all 23 patients with orbital involvement retained the orbital contents .Even patients with locally advanced disease may successfully preserve the orbital contents through this approach, although appropriate patient selection is critical.

These results appear promising, however, further investigation is needed. Patients should be encouraged to participate in well-designed experimental trials.

It is possible that a molecular marker exists for chemotherapy responsiveness, at least for adenocarcinomas of the ethmoid sinus. A striking correlation between p53 mutation status and response to induction cisplatin, 5-FU and leucovorin (PFL) chemotherapy was reported in a prospective series of 30 patients with intestinal type ethmoid sinus adenocarcinoma. Chemotherapy was followed by surgery and RT. The pathologic complete response rates for patients with wild type versus mutated p53 were 83 and 11 percent, respectively, while with 55 month median follow-up, the relapse rates were 0 and 44 percent for pathologic complete responders and nonresponders, respectively. A similar relationship (although not nearly so striking) has been seen with SCCs of the head and neck undergoing induction chemotherapy. (See "Neoadjuvant chemotherapy for locoregionally advanced head and neck cancer: General aspects", section on Impact of p53 status on response to neoadjuvant therapy).

Radiation therapy — RT alone has been successful for the treatment of early (stage I) lesions of the paranasal sinuses. In one report of 48 patients with malignant tumors of the nasal cavity, ethmoid sinus, or sphenoid sinus treated with curative intent RT, the continuous DFS rate at ten years was 86 percent for stage I, but dropped to 42 and 22 percent for patients with stage II and III disease.

Slightly better results were reported in a series of 29 patients with ethmoid sinus tumors (66 percent T4 lesions) treated with primary RT .The five year overall survival was 39 percent, however 12 of the 18 deaths were due to cancer, and surgical salvage was only successful in two of six patients. However, these data were not confirmed in a retrospective review of 180 patients with all sites of paranasal sinus cancer who were treated with curative intent RT at the same institution. Ninety-five patients (53 percent) failed initial management, of whom only 34 could undergo surgical salvage (the others had distant metastases, unresectable disease, were medically unfit, or refused surgery). The overall five year survival was only 22 percent for the entire group.

RT is rarely used as the sole modality of treatment except for cancers of the frontal and sphenoid sinuses; the challenging anatomy makes these tumors unsuitable for en bloc surgical resection. Inoperable lesions are also treated with primary RT. In most cases, elective treatment of the neck is unnecessary, although this is a controversial area.

  Dosing and complications — RT should be delivered by megavoltage external beam approaches. Due to the close proximity of critical structures such as the central nervous system and eye, great care should be taken in designing the treatment ports to minimize the risk of serious toxicity. CT based treatment planning and conformal RT should be used for planning to ensure adequate tumor coverage while minimizing the dose to normal critical structures.

The initial treatment port is usually large, including all the adjacent structures at risk. Upon delivery of 45 to 50 Gy, this field is modified to give an additional boost of 15 to 20 Gy to the primary site, thus delivering a total dose of 60 to 70 Gy in 2 Gy daily fractions. While 50 Gy is usually sufficient either preoperatively or postoperatively in the patient with negative margins, doses of more than 66 Gy are required in unresectable cases, or in those with gross residual disease following attempted resection.

Serious complications of RT occur in 5 to 10 percent of treated patients and include serous otitis media, meningitis, unilateral or bilateral blindness, optic neuritis, cataracts, and central nervous system damage.When doses above 50 Gy are needed, at least one series suggests that the rate of radiation retinopathy is lower with hyperfractionated as compared to conventional fractionation RT (37 versus 13 percent, respectively).

  Hyperfractionated RT — Hyperfractionated RT has been studied in patients with unresectable or locally advanced paranasal sinus cancer .In one study of 21 patients with locally advanced cancer of the nose or paranasal sinus (T3, T4, or recurrent tumors), split course hyperfractionated RT (1.2 Gy per fraction, twice daily for three separate two week sessions, each separated by one to two weeks) was administered with concomitant infusional cisplatin. Sixteen of the 17 patients treated with curative intent achieved a clinical complete response, of whom 50 percent remained alive from 36 to 126 months after therapy. Although these data are encouraging, the superiority of hyperfractionated RT over standard fractionation is unproven.

  Intensity modulated RT — Intensity modulated RT (IMRT) is an advanced form of three-dimensional conformal RT that can create a dose distribution around a complex and irregular target volume. In contrast to conventional or conformal RT, in which a constant dose rate is administered to a defined field, IMRT delivers nonuniform beam intensities to the target volume. The intensity of the beam is changed either by fluctuating the opening of the radiation beam (collimator) with a fixed gantry position or by changing the beam opening during an arc.

While IMRT has the potential to improve the targeting of tumor bearing areas while sparing more normal tissue during the treatment of paranasal sinus cancer, only short-term results are available in a small number of patients.

Advanced disease — Patients with unresectable disease are often offered concomitant chemotherapy and RT, while those with recurrent or metastatic disease are usually managed with chemotherapy alone.

  Concomitant chemoradiotherapy — Concomitant chemoradiotherapy is a feasible option for patients with unresectable squamous cell carcinoma of the head and neck. . The results in patients with paranasal sinus cancer are limited but have been encouraging when compared to the results of standard treatment in historical series .As an example, one series of patients with stage IV unresectable cancers of the head and neck studied split-course hyperfractionated RT with concomitant infusional cisplatin . The three year local control rate for cancers of the paranasal sinuses was 78 percent.

Another technique for patients with advanced cancer of the paranasal sinuses combines rapid high-dose intraarterial delivery of cisplatin, simultaneous intravenous infusion of the cisplatin antagonist thiosulfate, and external beam RT .Encouraging preliminary results using this approach were reported in 19 patients with paranasal sinus cancer, 16 with T4 disease.With a median follow-up of 53 months, the actuarial two and five-year survival rates were 68 and 53 percent, respectively. One patient had a myocardial infarction during treatment, but no other treatment-limiting toxic effects were noted. These impressive results, particularly in this unfavorable cohort, require independent confirmation.

  Chemotherapy — Traditionally, chemotherapy has been reserved for palliative treatment of advanced or recurrent disease. This topic is addressed elsewhere.  High-grade esthesioneuroblastomas (olfactory neuroblastomas) have a response rate of about 50 percent to cisplatin-based regimens.

POSTTREATMENT FOLLOW-UP — The goal of posttreatment surveillance is improved survival through early detection of recurrent disease and identification of second primary cancers. Patients with HNC are more likely to develop second primary cancers than any other group of patients with malignancy. This probably reflects the wide distribution of the toxic effects of tobacco and alcohol, the major risk factors for HNC.

The late development of second primary tumors is the most common cause of posttreatment "failure" after 36 months. The major sites are head and neck, lung, and esophagus. Patients with supraglottic laryngeal cancer are at a particularly high risk of developing metachronous lung cancers.  section on Second and multiple primaries).

Despite the lack of defined survival benefit from any posttreatment surveillance strategy, surveillance protocols are in widespread clinical use after curative-intent therapy for HNC  In general, the intensity of follow-up is greatest in the first two to three years, which is the period of greatest risk for disease recurrence.