Sinonasal squamous cell carcinoma (SNSCC) is the most common malignant tumor of the sinonasal region. Much progress has been made in the treatment of SNSCC including endoscopic endonasal resection, conformal radiation therapy such as intensity modulated radiation therapy and proton therapy, as well as the use of neoadjuvant and adjuvant chemotherapy. These advancements have improved the outcome of patients with SNSCC. Understanding of the molecular biology of SNSCC will further enhance the treatment of patients with SNSCC including developments of effective targeted therapy and immunotherapy.
29 Squamous Cell Carcinoma of the Nasal Cavity and Paranasal Sinuses
Sinonasal squamous cell carcinoma (SNSCC) is the most common histologic subtype of sinonasal cancers and accounts for almost half of all cancer in the sinonasal region. Of the 2,698 patients who had sinonasal cancers treated at MD Anderson Cancer Center, 45% had SNSCC (Fig. 29.1). A recent comprehensive analysis using the U.S. National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) database registry reported the trends in the epidemiology of SNSCC.1 A total of 4,994 cases of SNSCC were identified, composed of 65% males and 35% females, for a 1.81:1 male:female prevalence ratio. The majority of SNSCC tends to occur in people 55 years old or older, with 3,954 (79%) cases reported in patients within this age group. Dividing the data by race showed that 4,120 (82.50%) patients were white, 438 (8.77%) were black, and 436 (8.73%) were “others.” The majority of cases reported the paranasal sinuses (2,693 cases, 53.92%) as the primary sites, with the remainder of cases being in the nasal cavity (2,301 cases, 46.08%). Incidence trend analysis revealed a significant decrease in yearly rates from 1973 to 2009 for the overall population, females, whites, blacks, and “others” (P < .05). This decrease may be partially attributable to decreased exposure to textile dust and heightened awareness and better regulation of the exposure to the carcinogenic effect of industrial substances. Another variable that may have contributed to the decrease in overall incidence of SNSCC is the decline in tobacco smoking.1
The majority of squamous cell carcinoma (SCC) of the paranasal sinuses is keratinizing but tends to be only moderately differentiated. Nonkeratinizing and poorly differentiated carcinomas are less common, and the latter show a more rapid course of growth. Variants of SCC make up 15% of all cases of SCC of the upper aerodigestive tract. There are five main histologic variants of SCC in the head and neck region: verrucous (VSCC), papillary (PSCC), spindle cell (sarcomatoid; SCSC), basaloid (BSCC), and adenosquamous (ASC). Conventional sinonasal SCC has been studied extensively, but far less is known about its major variants. In a recent SEER database analysis, a total of 4,382 cases of conventional sinonasal SCC and 328 cases of its major variants were found.2 Sinonasal BSCC was diagnosed at a significantly lower mean age than sinonasal SCC. Sinonasal SCSC significantly affected the maxillary sinus more commonly than SCC. In the setting of advanced stage disease, sinonasal VSCC, PSCC, and BSCC appear to be associated with a better prognosis than conventional sinonasal SCC, whereas the impact of histologic subtype on prognosis in early stage disease appears to be more limited. Survival for SCSC and ASC, both regarded as more lethal variants, was statistically similar to conventional SCC. This study highlights the importance of distinguishing between conventional sinonasal SCC and its major histologic variants, because histologic subtype appears to carry important prognostic implications.2
Another distinct entity is Schneiderian carcinoma, which commonly represents malignant transformation in a preexisting Schneiderian papilloma (SP). Carcinomas arising from SP are rare, with more carcinomas identified in the inverted type and oncocytic types but only isolated reports describing carcinomas arising from the exophytic type. Carcinoma ex-SP ranges from 2 to 27% in the literature, but based on a recent systematic review conducted in 2014, without referral or academic institution bias, the 1.9% rate might be more accurate.3 In general, the male:female ratio of patients who have Schneiderian carcinoma ranges from 1.2:1 to 6.7:1, for an overall average of about 3.4:1. Patients range in age from 32 to 86 years, with overall mean around 61 years. Most patients experienced a mixed anatomical site presentation: nasal cavity combined with maxillary, ethmoid, sphenoid, and/or frontal sinus, with possible involvement of the nasopharynx and ear. The role of human papillomavirus (HPV) in the oncogenesis of SP and malignant transformation remains to be defined, but some studies have demonstrated that high-risk HPV is more prevalent in dysplastic SP and in those who have malignant transformation.4 , 5 The majority of Schneiderian carcinomas are synchronous (carcinoma present at primary presentation of SP), with 36% metachronous (carcinoma developing after initial treatment of SP). This highlights the importance of complete excision and careful histopathologic assessment of all SP, which minimizes the risk of recurrence and allows comprehensive evaluation of the specimen for the presence of any coexisting malignancy.6 The recurrence rates of SP quoted in the literature vary from less than 5% to as high as 75% and may depend on the surgical approach and the completeness of the surgical excision. Although multicentricity of the tumor has been suggested to be responsible for the high rate of recurrence, inadequate removal of the tumor during the initial resection seems to be the most important predictive factor of local recurrence. This was well demonstrated by Myers et al, who reported less than 5% recurrence rate of adequately resected SP.7
29.3 Disease Spread
29.3.1 Local Spread
The most common route of local spread of cancer of the sinonasal tract is through direct extension. Because most sinonasal cancers are relatively asymptomatic when small, local spread often prompts patients to seek medical attention. In the maxillary sinus, direct extension may occur anteriorly into the soft tissues of the cheek, superiorly into the orbit with resultant proptosis and diplopia, inferiorly into the oral cavity, or posteriorly into the pterygomaxillary space, where it may spread along the branches of the maxillary division of the trigeminal nerve (V2). Cancer of the frontal sinus is quite rare, but the most significant direct extension is posteriorly to the frontal lobes. Cancer of the ethmoid sinus often presents with medial extension to the orbit, superior extension to the cribriform plate, and posterior extension into the sphenoid sinus and nasopharynx. Cancers involving the sphenoid sinus may quickly become problematic because of proximity to the optic nerves, the cavernous sinus, and the pituitary fossa.
In addition to direct local extension, cancer of the paranasal sinuses can spread to nearby structures via the many fissures and foramina located in this region. Cancer of the maxillary sinus frequently erodes posteriorly into the pterygopalatine fossa (PPF). Once in the PPF, the tumor may extend laterally through the pterygomaxillary fissure into the infratemporal fossa; superiorly into the orbit via the inferior orbital fissure or into the middle cranial fossa through the foramen rotundum; posteriorly into the vidian canal, with extension to the petrous portion of the temporal bone; or inferiorly into the oral cavity by way of the palatine canal or the sphenopalatine foramen.
From the frontal sinus, cancer may extend into the nasal cavity through the nasofrontal duct. Cancer of the ethmoidal sinuses may also extend into the nasal cavity through the middle meatus and the sphenoethmoidal recess, posteriorly into the nasopharynx and along the Eustachian tube, or inferiorly along the nasolacrimal duct.
29.3.2 Perineural Spread
The dissemination of cancer cells along nerves is a frequent pathologic finding among a variety of cancers, including head and neck, upper gastrointestinal, pancreatic, and prostate carcinomas. Tumors that have a considerable propensity to disseminate along nerves are known as neurotropic cancers. In the head and neck, the most common tumors having a predilection to invade nerves are adenoid cystic carcinomas (ACCs), followed by SCCs.8 , 9 Tumors of the paranasal sinus that exhibit perineural invasion may use this route to spread in a retrograde fashion to the skull base and even progress intracranially. Alternatively, they may spread in an antegrade fashion and along the involved nerve and its terminal branches. In either case, this neural spread makes surgical resection more complicated and makes achieving negative surgical margins less certain. Imaging, particularly MRI, is critical in determining the extent of neural spread of sinonasal cancers, as is discussed later in this chapter, under the section on imaging.10
Ziv et al reported the incidence and pattern of neural invasion (NI) in 208 patients who had cancers of the paranasal sinuses and anterior skull base.11 Forty-one specimens (20%) had evidence of NI. Sinonasal undifferentiated, adenoid cystic, and SCC had a high propensity for NI, whereas melanoma and sarcoma rarely invaded nerves. Intraneural invasion was found in 32% of these cases, and 34% invaded more than 1 cm distal to the tumor. NI was associated with a high rate of positive margins, maxillary origin, and previous surgical treatment (P < 0.04) but not with stage, orbital invasion, or dural invasion. Patients who had NI were more likely to undergo adjuvant radiotherapy (P = 0.003), which significantly improved survival in patients who had minor salivary gland carcinomas (P = 0.04).
29.3.3 Regional Metastases
The lymphatic drainage of the posterior nasal cavities and paranasal sinuses is primarily to the retropharyngeal and lateral pharyngeal nodes at the base of the skull, then to the upper jugular lymph nodes. Cancer of the anterior nasal cavity and those that erode through the maxilla into the soft tissues of the face spread to the submandibular and upper jugular lymph nodes.
Regional metastases from paranasal sinus cancer are relatively uncommon and have been characterized to a greater extent for maxillary sinus cancer than for other paranasal sites.12 The reported incidence of lymph node metastasis at presentation varies from 10 to 15%, and nodal recurrence may occur in as many as 30% of patients.13 The most common sites of involvement are the retropharyngeal and level II nodes. In patients who have SCC of the maxillary sinus, the risk of having lymph node metastasis on presentation correlates with extension of the primary tumor to the nasopharynx or oral cavity. The risk of developing regional metastasis after treatment correlates with local tumor recurrence.
Lymph node metastases signify more advanced disease and carry worse prognosis. When the primary disease can be addressed surgically and there is clinical evidence of nodal metastasis, a therapeutic nodal dissection should be performed. Management of the clinically N0 neck remains controversial, but elective nodal irradiation may be warranted in patients who have locally advanced disease.12 , 14 , 15 , s. Literatur
29.3.4 Distant Metastases
Although distant metastasis from cancer of the paranasal sinus does occur, failure to control the disease secondary to local recurrence is far more common. For SCC of the maxillary sinus, the rate of distant metastasis is approximately 10%, and it rarely occurs in the absence of local recurrence. Cancer of the ethmoid has a similar rate of distant metastasis, with adenocarcinoma having a slightly higher rate than squamous cell cancer (15–20% vs. 10%). In general, the most common sites for metastasis are the lung and bone.12 , 16 , 17
The most widely used staging system for sinonasal cancers is the American Joint Committee on Cancer (AJCC) tumor–node–metastasis system. There is a different staging system for cancer of the maxillary sinus than is used for ethmoid sinus and nasal cavity cancers. The nodal staging system for maxillary, ethmoid, and nasal cavity cancers is the same as for other sites in the head and neck and depends on the number, size, and laterality of involved lymph nodes. The classification from the most recent version, the 8th edition,18 is shown in Table 29.1.
The majority of patients who have SNSCC (85%) present with advanced stage (T3–T4) cancer. Although the reported incidence of clinically evident lymph node metastasis presentation is around 10 to 15%, the overall risk of nodal involvement from SCC of the paranasal sinuses is closer to 30%.15 , 16 Regional spread to the lymph nodes is uncommon in cancer confined within the sinus walls. After invasion into the overlying soft tissue and adjacent structures (e.g., the oral cavity), nodal involvement and even dissemination to distant sites are noted more frequently. Distant metastasis can be present in up to 10% of patients on initial diagnosis.19
Treatment of SNSCC depends on the stage and extent of disease. Early stage (T1–T2) tumors can be treated by single-modality therapy, more commonly surgery but in some selected cases radiation therapy. Patients who have localized disease showed 5-year survival rates of 86%, 80%, and 78% when receiving surgery, radiation and surgery, and radiation alone, respectively.1 Endoscopic sinus surgery may be applied in selected cases, as discussed earlier, with relatively good outcomes comparable to those associated with open surgery.20 , 21 , 22 The majority of patients who have more advanced and resectable disease are treated using surgery and postoperative radiation. Extension to the skull base is common, and craniofacial resection has enhanced our ability to resect locally advanced tumors successfully.23 , 24 , 25
Surgery followed by postoperative radiation therapy has been the accepted gold standard for most tumors of the sinonasal cavity. Hoppe et al published an 18-year experience at Memorial Sloane-Kettering Cancer Center in which 85 patients were treated for sinonasal cancers using surgical resection and postoperative radiation.26 Most patients had SCC, T4 tumors, and tumors involving the maxillary sinus. Their 5-year estimates of local progression-free, disease-free, and overall survival rates were 62%, 55%, and 67%, respectively. The authors noted that squamous cell histology and cribriform plate involvement were independent predictors of local recurrence.
There is strong evidence that the use of combined surgery and adjuvant radiation therapy results in better tumor control and survival than radiation alone in patients who have cancer of the paranasal sinuses.27 In 2009, Mendenhall et al reported the results of 109 patients who had sinonasal cancer treated between 1964 and 2005.28 Within this group, 56 patients were treated using definitive radiation therapy, whereas 53 patients received surgery and postoperative radiation. Although the 5-year local control rate was 82% in patients who had T1 to T3 lesions, those patients who had T4 disease had a lower local control rate of 50%. Local control at 5 years was 43% after definitive radiation therapy versus 84% with primary surgery and adjuvant radiation therapy (p < 0.0001). Cause-specific survival rates were 81% and 54% for stage I to III and stage IV disease, respectively. This group concluded that the probability of local control and cause-specific survival is better after surgery and radiation therapy than after definitive radiation therapy. However, selection bias may have influenced the poor results of radiation therapy, considering that surgery was likely performed only for patients who had resectable disease. Similarly, in 2009, Snyers et al reported a series of 168 patients treated between 1986 and 2006.29 In all, 130 patients were treated with curative intent using surgery followed by postoperative radiotherapy, and 38 were considered inoperable and received radiotherapy alone (n = 21) or radiotherapy and chemotherapy (n = 17). For the entire population, the 5-year local control rate was 62% and regional control was 79%. Distant metastasis-free survival was 79%. Of the cases involving SCC or adenocarcinoma, patients who had stage I to III versus stage IV disease had local control rates at 5 years of 79% and 54%, respectively, comparable with the series reported by Mendenhall and colleagues.28 Locally advanced disease that is not surgically resectable can be managed using radiation alone or using concurrent chemoradiation therapy. Radiation therapy alone in this setting yielded poor results, and more promising outcomes have been reported by the use of intensive regimens of chemotherapy followed by concurrent chemoradiation.30
The details of surgical management of sinonasal cancers are discussed in Chapter 16 of this book.
29.4.2 Management of the Neck
The incidence of lymph nodal metastasis at presentation ranges from 3 to 26%, and several authors have reported high rates of neck recurrences in untreated necks that can reach up to 30%.13 , 14 Although the most frequently reported sites of lymphatic metastasis are level I and II, a significant part of the lymphatic drainage of the paranasal sinuses and the nasal cavity is directed to the retropharyngeal lymph nodes, which are inaccessible for palpation and are frequently overlooked. It is possible, then, that the true incidence of lymphatic spread of sinonasal malignancy is underestimated. The retropharyngeal lymph nodes are best evaluated using high-resolution imaging (CT or MRI) or PET-CT.
The overall risk of nodal metastasis either at diagnosis or as regional recurrence may depend on the histology and the stage and extent of the primary tumor. High-grade tumors such as SCC and sinonasal undifferentiated carcinoma have a relatively higher rate of nodal involvement, do advanced-stage (T3–T4) tumors.16 Tumor extension into the oral cavity and nasopharynx is also associated with increased risk of nodal metastasis.13 , 14
Lymphatic metastasis to the cervical lymph nodes carries with it a poor prognosis in patients who have cancer of the sinonasal tract. In a study of 146 patients who had maxillary sinus cancer treated at MD Anderson Cancer Center,16 patients presenting with node-negative versus node-positive disease had an estimated 5-year OS rate of 56% versus 44%, respectively (p = 0.06; Fig. 29.2).
Patients who have clinically positive lymph nodes require treatment of the neck, but prophylactic treatment of N0 patients remains controversial.13 Because the risk of nodal metastasis is higher in patients who have high-grade and advanced-stage disease, it is generally recommended that the neck be electively treated in such patients. This policy of elective neck irradiation of patients who have advanced-stage and high-grade tumors was adopted at MD Anderson Cancer Center in 1991.16 Fig. 29.3a shows the effect of elective neck radiation therapy (RT) on nodal control for node-negative (N0) patients who have squamous cell or undifferentiated carcinoma. Of the 36 patients in whom the ipsilateral neck was left untreated, 13 (36%) developed nodal recurrence versus only 3 (7%) of the 45 patients in whom elective neck irradiation was administered (p < 0.001). The use of elective neck treatment in these patients translated into a significant reduction in distant metastases (3% in treated vs. 20% in untreated at 5 years; p = 0.045) and an increase in recurrence-free survival (RFS) (67% in treated vs. 45% in untreated at 5 years; p = 0.025; Fig. 29.3b,c).
29.4.3 Radiation Therapy
Radiation therapy is frequently incorporated in the overall management of patients who have cancer of the nasal cavity and paranasal sinuses. Radiation therapy may be given with curative intent or as an adjuvant therapy before or after surgery. Radiation therapy may also be combined with chemotherapy, either as definitive treatment or as an adjunct to planned surgical resection. Radiation therapy may also be used in the palliation of recurrent or unresectable tumors. Regardless of the treatment strategy, there is almost universal agreement that patients who have advanced-stage tumors are best treated using multimodal therapy, including surgery, radiation, and in some cases chemotherapy.28 , 31
External beam radiation or brachytherapy or both may be used as definitive local therapy in selected patients who have early-stage cancers of the nasal cavity.32 Primary radiation therapy, however, has not been a well-accepted approach for definitive treatment of more advanced sinonasal cancers. This conclusion was partly based on the poor outcomes for patients who have advanced lesions and are concerned that radiation therapy does not adequately treat bony invasion, which is a frequent finding in patients who have sinonasal malignancies. In additions, several publications have reported increased incidence of radiation associated optic nerve injury and osteoradionecrosis when radiation therapy is administered as primary treatment.28 , 29 Although some authors propose primary concurrent chemoradiation therapy for this site,17 , 33 the majority of the published data on primary radiation therapy have been for tumors deemed not surgically resectable and have had a selection bias toward advanced disease.28 , 29 , 34 In a study from Memorial Sloane-Kettering Cancer Center, the 5-year disease-free survival (DFS) and overall survival (OS) were 14%, and 15%, respectively, for 39 patients who had unresectable stage IVB paranasal sinus carcinomas treated with RT, with or without chemotherapy.35 The majority of the recurrence (64%) was within the irradiated field. The investigators reported that the only significant factor for improved local progression-free survival and overall survival was a biologically equivalent dose of radiation of > 65 Gy and that treatment outcomes for patients who had unresectable sinonasal malignancies remained poor.
Delivering effective doses of radiation (60–70 Gy) for treatment of advanced sinonasal cancer using conventional radiotherapy is associated with serious morbidity, including blindness, brain necrosis, radiation-induced endocrinopathy attributed to hypothalamic–pituitary radiation damage, and osteoradionecrosis.28 , 29 The use of three-dimensional conformal radiotherapy (3D-CRT) and intensity-modulated radiotherapy (IMRT) increases treatment accuracy by delivering tumoricidal doses to the tumor bed while reducing radiation doses to nearby critical structures such as the optic nerves and the brain. Duprez et al examined the ocular complications of IMRT.36 They reported on 130 patients treated using IMRT up to 70 Gy, of whom 101 patients were in the postoperative setting. The 5-year local control and overall survival were 59% and 52%, respectively. There was no radiation-induced blindness in 86 patients available for 6-month follow-up; 10 patients reported grade III tearing, and 1 patient had grade III visual impairment from ipsilateral retinopathy and neovascular glaucoma. Brain necrosis and osteoradionecrosis occurred in 6 patients and 1 patient, respectively. Chen et al analyzed 127 patients treated between 1960 and 2005.37 The cohort was treated using conventional RT in 59 patients, 3D conformation in 45 patients, and IMRT in 23 patients. The 5-year OS, local control (LC), and DFS were not significantly different when analyzed by decade. However, the incidence of grade III to IV toxicity was 53%, 45%, 39%, 28%, and 16% for patients treated in the 1960s, 1970s, 1980s, 1990s, and 2000s. The authors concluded that improvements in therapeutic ratio were responsible for decreasing incidence of complications for patients treated throughout these decades.
More recently, charged particle radiation using beams of protons, carbon ions, helium ions, or other charged particles has held the promise of further enhancing high-dose delivery to tumor targets while limiting toxicity to normal tissue. The unique physical properties of charged particle therapy—with rapid fall-off of dose beyond the Bragg peak (a sharp deposition of dose at a specific depth in tissue)—and its increased biological effectiveness compared with photon therapy might further augment treatment outcomes, not only by reducing the incidence and severity of complications but also by allowing an escalation in radiation dose to improve tumor control and survival, which cannot be achieved using photon therapy. A recent systematic review and meta-analysis compared the clinical outcomes of patients treated using charged particle therapy with those of individuals receiving photon therapy.38 The study included 43 cohorts from 41 noncomparative observational studies, of which 30 cohorts were treated with photon therapy (1,186 patients) whereas 13 received charged particle therapy (286 patients). Median follow-up for the charged particle therapy group was 38 months (range 5–73), and that for the photon therapy group was 40 months (14–97). The pooled event rate of overall survival for charged particle therapy was significantly higher than that for photon therapy at the longest duration of follow-up (relative risk 1·27, 95% CI [confidence interval] 1·01–1·59; p = 0·037) and at 5 years (1·51, 1·14–1·99; p = 0·0038); see Table 29.2.
Locoregional control was also significantly better at the longest duration of follow-up for patients treated with charged particle therapy than for those receiving photon therapy (1·18, 1·01–1·37; p = 0·031), but not at 5 years (1·06, 0·68–1·67; p = 0·79). The pooled 5-year disease-free survival event rate was significantly higher for charged particle therapy than for photon therapy (1·93, 1·36–2·75; p = 0·0003), but not at longest follow-up (1·51, 1·00–2·30; p = 0·052). Table 29.3 shows the comparison of primary outcomes for cohorts receiving proton beam therapy versus those given IMRT. Disease-free survival at 5 years and locoregional control at longest follow-up were significantly higher in the proton beam therapy group. However, no other difference was noted between proton beam therapy and IMRT.