new 45 glomus tumors 03

Nasopharyngeal Angiofibroma

Epidemiology

Juvenile nasopharyngeal angiofibroma (JNPA) is found more frequently in young pubertal boys (288); it has been shown to contain androgen receptors (4,81) and occasionally to regress with estrogen therapy. Hwang et al. (125), in 24 nasopharyngeal angiofibromas, detected androgen receptors in 18/24 (75%) cases, whereas only two (8.3%) were positive to progesterone. None of the 24 cases was positive for antibodies to estrogen.

JNPA comprises <0.05% of head and neck tumors (168). Patient age at presentation ranges from 9 to 30 years (52,272,288), with a median of 15 years. Females comprise <4% of the total cases (52). Some investigators have suggested chromosomal studies in affected women because this is mainly a male disease (39).

The tumor is believed to originate from the posterolateral wall of the nasal cavity where the sphenoidal process of the palatine bone meets the horizontal ala of the vomer and the roof of the pterygoid process (4) because it is always involved (39). Other investigators agree, because involution of tumor after irradiation usually occurs in this direction (235).

Clinical Presentation and Pathology

Symptoms usually occur 2 to 48 months before diagnosis (52). The most common complaints are nasal obstruction or epistaxis, followed by nasal voice or discharge, cheek swelling, proptosis, diplopia, hearing loss, and headaches (52). In a series byWitt et al. (288), 7/31 patients presented with anomalous sexual development.

Nasopharyngeal angiofibroma may initially extend into the nasal fossae and maxillary antrum and push the soft palate downward, then through the pterygopalatine fossa and superoanteriorly through the inferior orbital fissure or laterally through the pterygomaxillary fissure to the cheek and temporal regions (58).

Beham et al. (15) reported on 32 cases of nasopharyngeal angiofibroma. Most of the tumor vessels, which lacked elastic laminae, were characterized by vascular walls of irregular thickness and variable muscle content. In places endothelial cells were separated from the stroma by only a single attenuated layer of contractile cells, in some more fibrotic hyaline areas, the stromal cells displayed reactivity for smooth muscle actin. The irregularity of the vascular walls, together with the lack of elastic laminae and stromal fibers, explains the pronounced tendency for hemorrhage in these lesions.

Differential diagnosis includes fibrosarcoma, rhabdomyosarcoma, chronic sinusitis, arteriovenous malformation, lymphangioma, neurofibroma, pleomorphic adenoma, lymphoma, pyogenic granuloma, polyps, and hemangioma.

Diagnostic Work-Up

After the history and physical examination, CT scans with and without contrast should be obtained. The pattern of enhancement in this highly vascular tumor is diagnostic (164,192), and many investigators believe carotid angiograms are unnecessary (30) after CT diagnosis of the lesion, unless embolization, which is also controversial, is contemplated.

CT scans are especially helpful in regions involving thin bony structures (paranasal sinuses, orbits), where CT performs better than MRI. In the nasopharynx and parapharyngeal space MRI is superior to CT. Obtaining tumor volumetric data with spiral CT or MRI facilitates 3D treatment planning (102).

Seventy-two patients with JNPA were evaluated with CT and/or MRI (164). Origin of the tumor was in the pterygopalatine fossa at the aperture of the pterygoid (vidian) canal. The tumor extended posteriorly along the pterygoid canal with invasion of the cancellous bone of the pterygoid base and greater wing of the sphenoid in 60% of the patients. The inability to remove the tumor in toto was principally due to deep invasion of the sphenoid; 93% of recurrences occurred with this type of tumor extension.

If intracranial extension is noted and radiation therapy is contemplated, no further studies are indicated. If the lesion is extracranial and surgery is indicated, bilateral carotid angiograms will identify the feeding vessels and delineate the boundaries of the tumor (Fig. 45.13).

Biopsies are not indicated in all patients because of the potential for severe hemorrhage. It is important to perform a biopsy of the lesion when the clinical picture (sex, age, location, and behavior of the lesion) is not consistent with JNPA (250) because some lesions have proven to be sarcomas or chronic sinusitis (39). Two cases of fibrosarcoma have been reported in patients in their 40s (66).

Staging and Prognostic Factors

Two staging schemes have been proposed: (a) the system of Chandler et al. (39) (Table 45.14), and (b) a radiographic staging system by Sessions et al. (235). Stage Ia is limited to the nasopharynx and posterior nares; stage Ib extends to the paranasal sinuses; stages IIa, IIb, and IIc extend to other extracranial locations; and stage III is intracranial.

Table 45.14. Staging of Nasopharyngeal Angiofibromas

In a retrospective review of 44 cases of juvenile nasopharyngeal angiofibroma, invasion of the skull affected two thirds of the patients, and the rate of recurrence was 27.5% (116). Extensions to the intratemporal fossa, sphenoid sinus, base of pterygoids and clivus, the cavernous sinus (medial), foramen lacerum, and anterior fossa were correlated with more frequent recurrence. Long-term radiographic follow-up showed residual disease in nine asymptomatic patients: these remnants gradually involuted.

General Management

The decision of whether surgery or radiation therapy should be used depends in part on the initial extent of the disease. In patients with extracranial tumors (116,288), surgery is the treatment of choice and yields near-zero mortality or any longterm morbidity.

Tumor remnants in symptomfree patients should be kept under surveillance by repeated CT scanning, since involution may occur. Recurrent symptoms may be treated by radiation therapy rather than by extended surgery or combined procedures (39,61,116).

When there is intracranial tumor extension (seen in about 20% of patients) (130,288), the risk of surgically related death increases. In a literature review by Jones et al. (133), it was found to be 14% to 84%. Most of these patients are best treated with irradiation (54).

Some investigators recommend preoperative intraarterial tumor vessel embolization at the time of diagnostic bilateral carotid angiography, claiming a decrease in operative bleeding (154). Salvage with embolization of polyvinyl alcohol has been described (129). Others have reported anecdotal evidence of partial regression with the use of estrogens, believed to be the result of feedback inhibition of the pituitary's production of gonadotropin-releasing hormone.

Although radiation therapy is equally effective in extracranial tumors, the low but existing risk of secondary malignancies should limit its use to the most advanced tumors only (10,154,284). In the experience of Cummings et al. (54) covering 20 years, only two radiation-related malignancies were noted (one skin, one thyroid).

Radiation Therapy Techniques

Photon irradiation should be used for these patients, and fields must be individualized to cover the tumor completely with a margin (1 to 2 cm). Treatment portals are similar to those used in carcinoma of the nasopharynx (without irradiating the cervical lymph nodes) or carcinoma of the paranasal sinuses when these structures or the nasal cavity is involved. Opposing lateral portals are suitable in most patients, with larger fields and compensators used for tumors extending into the nose (Fig. 45.14). More extensive disease requires three-field or wedge-pair arrangements of 3DCRT or IMRT that can yield excellent dose distributions, particularly when there is nasopharyngeal or intracranial tumor extension. In all cases the eyes are protected asmuch as possible. The recommended tumor dose ranges from 30 Gy in 15 fractions in 3 weeks to 50 Gy in 24 to 28 fractions in 5 weeks (54). A conventional setup uses 6- to 18-MV photons to treat the lesion with parallel-opposed fields to 50 Gy (2-Gy fractions) (Fig. 45.15).

Figure 45.14. Diagrams of the most commonly used conventional field arrangements for treatment of nasopharynx angiofibroma: lateral opposed field pair and three-field technique.

Figure 45.15. Example of conventional lateral portal used at the Mallinckrodt Institute of Radiology for nasopharyngeal angiofibroma.

The advantages of IMRT for the treatment of extensive and/or recurrent JNPA has been described in three patients on whom the tumor affected the base of skull, pterygopalatine, and intratemporal fossae, posterior orbit, and nasopharynx (151). Tumor dose varied from 34 to 45 Gy. The tumor shrunk radiographically in all three cases, and there was no endoscopic evidence of disease in two cases at 15 months and 40 months. Late toxicity was limited to one episode of epistaxis and persistent rhinitis in one patient.

Results of Therapy

Jones et al. (133) reported the results of 40 patients with JNPA treated with surgery alone. With a mean follow-up of 17 months (6 to 36 months), the control rates according to the Sessions staging system were as follows: 100% (stages I and IIa), 83% (stage IIb), 80% (stage IIc), and 50% (stage III). All failures were controlled with irradiation (n = 18 patients), embolization (n = 8), or surgical resection (n = 8), and the other four were observed only, demonstrating the extremely high salvage rate in this disease (126). These findings are consistent with other series reporting initial surgical control of 86% with an ultimate control rate of 96% (288).

In another report, 18 patients were treated with gross tumor excision; two cases with intracranial involvement required a combined neurosurgical-otolaryngologic approach (61). Recurrent intracranial disease was detected by MRI in three patients, who were treated with 35-, 36-, and 45-Gy external-beam irradiation. Extracranial tumor recurrences were re-excised in seven patients. All patients (followed up with serial MRI) are living without evidence of active disease.

Cummings et al. (54) treated 42 patients primarily with irradiation and 13 for postsurgical failures; all except six had biopsies. Nine had stage IV disease according to Chandler's staging system. Dose was 30 to 35 Gy in 14 to 16 fractions over a 3-week period. Follow-up ranged from 3 to 26 years. The control rate was 80% and was equivalent for all dose ranges. Local control was 89% and 74%, respectively, when three fields versus two fields were used. When the field size was more than 6 by 6 cm, the control rate was 83% versus 55% for smaller portals, indicating the importance of accurately determining the target volume, including any potential tumor extension. Of 11 recurrences, eight were controlled by a second course of irradiation and three by surgery. These tumors regress slowly, with 50% still present at 12 months. At 24 months, 23% of tumors were still present, and half of those recurred. Of the complete responders, only 1/33 had a recurrence. Robinson et al. (224) also found that objective responses after irradiation were noted within 6 months in 60% of patients and within 6 to 20 months in the other 40%. Symptoms, however, resolved in all patients within 6 months of treatment.

At the Mallinckrodt Institute of Radiology, Fields (written communication, 1989) reviewed our experience with 13 patients: 11 surgical failures and two primarily treated with irradiation. Intracranial extension was noted in 38% of patients. Follow-up ranged from 40 to 173 months. Doses ranged from 36 to 52 Gy, with a median of 48 Gy (1.8 to 2 Gy per fraction, 5 days a week). The control rate was 85%; patients failing irradiation were salvaged with embolization. Late morbidity was mostly xerostomia and dental decay.

Ungkanont et al. (272) described results in 20 patients treated before 1974 and 23 treated between 1975 and 1993: 31 had surgery (18 with preoperative embolization), three had irradiation, seven received chemotherapy (four combined with surgery), and two were observed. Diseasefree survival was 67%; 28% of patients survived with residual tumor, and 4.6% died of surgical complications.

Tumor regression usually occurs slowly after either irradiation (54,172) or chemotherapy (100); therefore, the presence of tumor up to 2 years after treatment is not an invariable sign of failure unless it is symptomatic or progressing.

The management of large JNPA with intracranial extension is difficult. In 18 patients with JNPA, preoperative MRI, embolization of feeding branches from the external carotid artery, and attempted complete resection were used in seven patients with intracranial disease (61); serial MRI scans were used for follow-up. Intracranial disease that was persistent or recurrent and demonstrated subsequent growth was irradiated (35 to 45 Gy) or re-excised.

Wiatrak et al. (284) reported on three patients with extensive intracranial extension treated primarily with radiation therapy doses of 36.6, 40, and 50.4 Gy, respectively, without surgical tumor resection. Although there was no complete resolution of the tumors, significant improvement of symptoms was obtained without serious sequelae. Malignant degeneration in JNPA undergoing radiation therapy has been occasionally reported (16,54,224).

Goepfert et al. (100) reported on five patients with aggressive nasopharyngeal angiofibromas recurrent after extracranial resection and irradiation who were treated with chemotherapy. Doxorubicin (60 mg/m2 intravenous [IV] push for 1 day) and dacarbazine (250 mg/m2 IV drip for 5 days) were given, with courses being repeated every 3 to 4 weeks. In a second regimen, vincristine, dactinomycin, and cyclophosphamide were administered at usual doses. Excellent tumor regression was noted in all patients. Patients were disease free at 2, 3, 6, and 10 years.

Ochoa-Carrillo et al. (192) reported on 31 patients treated with surgery and/or radiation therapy. Surgery was the treatment chosen in patients with stage II and III disease, while radiation therapy was the treatment in stage IV, but it had low effectiveness, indicating the need to carefully investigate the value of craniofacial approaches in these tumors. Radiation therapy (30 to 55 Gy) was administered to 16 patients; seven with stage III persistent or recurrent tumor, and eight patients as initial treatment for stage IV disease. The diseasefree interval of patients with stage III and IV disease was 80.3% and 19%, respectively, after 36 months of follow-up.

Tranbahuy et al. (267) reported on seven patients with juvenile angiofibroma who underwent direct tumoral embolization. This technique induced marked devascularization and necrosis of the tumor. No neurologic sequelae were encountered.

Sequelae of Therapy

Most investigators agree that surgical mortality increases with intracranial extension of the tumor. The most common radiation therapy sequelae include delayed growth secondary to hypopituitarism and decreased bone maturation (39). There are several well-documented cases of radiation-induced sarcomas in these patients (54,245) with doses ranging from 66 Gy to more than 90 Gy. Spagnolo et al. (245) reported on four patients treated with irradiation who later developed sarcoma. Chen and Bauer (44) reported the same findings in a patient receiving 66 Gy and followed up for 18 years. This patient was 48 years of age, so the original diagnosis of JNPA does not coincide clinically with typical JNPA. Cummings et al. (54) reported two neoplasms developing 13 and 14 years after irradiation; one was a basal cell carcinoma and one metastatic thyroid carcinoma. Both patients are alive without disease. Two patients developed cataracts.

Extracranial Meningiomas

Extracranial meningiomas of the sinonasal tract are rare tumors (264). The overall prognosis is good, without a difference in outcome between benign and atypical meningiomas. Thirty cases of sinonasal tract meningiomas were retrieved from the files of the Otorhinolaryngic Registry of the Armed Forces Institute of Pathology (15 females and 15 males, aged 13 to 88 years [mean 47.5 years]) (264). Patients presented clinically with a mass, epistaxis, sinusitis, pain, visual changes, or nasal obstruction, dependent on the anatomic site of involvement. Symptoms were present for an average of 31.3 months. The tumors affected the nasal cavity (n = 14), nasopharynx (n = 3), frontal sinus (n = 2), sphenoid sinus (n = 2), or a combination of the nasal cavity and ethmoid, frontal, sphenoid, and/or maxillary sinuses (n = 9). Tumors size ranged from 1 to 8 cm (mean, 3.5 cm). Radiographic studies demonstrated a central nervous system connection in six cases; the tumors often eroded the bones of the sinuses (n = 18) and involved the surrounding soft tissues, the orbit, and occasionally the base of the skull. Histologically, the tumors demonstrated features similar to intracranial meningiomas; the majority of the meningothelial type (n = 23), although there were three atypical meningiomas. Immunohistochemical studies confirmed the diagnosis with reactions for epithelial membrane antigen and vimentin (all tested). The differential diagnosis included paraganglioma, carcinoma, melanoma, psammomatoid ossifying fibroma, and angiofibroma.

Surgical excision was used in all patients. Three patients died with recurrent disease (mean, 1.2 years), one was alive with recurrent disease (25.6 years), and the remaining 24 patients were alive or had died of unrelated causes (mean, 13.9 years) at the time of last follow-up (two patients were lost to follow-up).

Nonlentiginous Melanoma

Malignant melanoma accounts for 11% of primary head and neck malignancies (8). Of all malignant melanomas, 20% to 35% are located in the head and neck area (108).

Cutaneous Melanomas

In a review of the literature, Batsakis (11) found that, of all head and neck malignant melanomas, 64% to 78% were cutaneous, 6% to 8% were mucosal, and 14% to 30% were ocular. The superficial spreading and nodular types of malignant melanoma have a metastatic potential of 10% to 30% and 50%, respectively (108). Neurotropic melanoma is an uncommon variant of cutaneous melanoma, with a higher propensity to invade peripheral nerves. A thorough evaluation with CT scans should determine if there is intracranial or base of the skull involvement; Beenken et al. (14) reported on 13 such patients, two of whom failed perineurally.

Treatment of cutaneous melanomas has typically been wide excision of the lesion with a minimum 3-cm margin (8). More recently, margins of at least 2 cmhave been used in the head and neck area compared with wider margins for stage Imelanomas, with equivalent success as noted by the local failure rate of 3% to 6% (155).

The Princess Margaret Hospital treated 16 patients with nodular melanomas with local excision and postoperative radiation therapy (50 Gy in 10 fractions over 2 weeks); 14 exhibited local tumor control, and six were alive and well 2 to 14 years after treatment. These results were comparable with those with wide local excision alone but with less morbidity and fewer cosmetic alterations. Later, at the same institution, Harwood and Cummings (108) treated five patients with definitive radiation therapy for superficial spreading melanoma of the head and neck area. All five lesions were locally controlled; one patient had a lymph node metastasis that was later controlled, and one died of distant metastases. They recommend treating these patients with 45 Gy in 10 fractions over 2 weeks to 50 Gy in 15 fractions in 3 weeks (107,111).

Harwood and Cummings (108) also reported results in 74 patients treated with three 8-Gy fractions given on days 0, 7, and 21 with shielding of the spinal cord, brain, and eye. Thirty patients were treated postoperatively after neck dissections if they had extracapsular extension, multiple nodal involvement, a node >3 cm, or residual disease. Tumor control in the neck was achieved in 26/30 patients (86.6%) with follow-up of 1 to 4 years. In four patients with microscopic residual disease at the primary site, this postoperative regimen controlled three of four lesions with follow-up of 1 to 3.5 years. The other 40 patients were treated either for gross (13 patients) or recurrent (27 patients) cutaneous melanoma. Complete response was observed in 15/40 lesions (37.5%) and partial response in 12. An update of Harwood's data (personal communication, 1989) showed a neck tumor control rate of 94% in 41 adjuvantly treated patients versus 57% in 48 patients with gross residual or recurrent tumors. He concluded that irradiation alone should be considered for treatment of superficial spreading melanomas when surgery is contraindicated or after a simple excision in all cases of nodular melanoma in which a wide excision may be contraindicated because of age, location, or medical condition. For nodal disease, patients with poor prognostic pathologic factors should receive postoperative irradiation. Recurrent or unresectable tumors also should be irradiated. He recommended high-dose fractions because the local control rate was 71% when the dose per fraction was >4 Gy and 25% with lower fractions (111).

Another approach to the treatment of recurrent or unresectable cutaneous melanomas is combined hyperthermia and high-fraction radiation therapy as reported by Emami et al. (76). The data support the use of high fractions for melanoma because Overgaard's complete response rate was 59% when fractions of more than 4 Gy were used and 33% for lower dose per fraction sizes. However, a randomized study by the Radiation Therapy Oncology Group comparing four fractions of 8 Gy given on days 0, 7, 14, and 21 and 20 fractions of 2.5 Gy in 5 weekly fractions showed no significant difference in tumor response (24.2% and 23.4% complete response and 35% partial response) (218).

Mucosal Melanomas

Primary mucosal melanomas of the head and neck area comprise 2% to 8% of the cases seen each year in the United States (13). They occur more commonly in countries such as Japan, where mucosal melanoma is found in 22% to 32% of patients with malignant melanoma (271). Most occur in the fifth to seventh decades of life; they are extremely rare in the first two decades (0.6% of mucosal melanomas) (91,269,270). The maleto-female ratio approaches 1:1 (269). A review by Batsakis et al. (13) of 204 mucosal melanomas showed 56.4% to be from the upper respiratory tract and 44% from the oral cavity and pharynx. Nasal cavity/paranasal tumors comprise <1% of malignant melanomas and 2% to 9% of head and neck melanomas (91,120). Pigmentation may precede the lesion in up to 28% of patients for more than 1 year (32). In the oral cavity, the most common location is the hard palate (up to 80%), followed in order of decreasing frequency by the upper gingiva and lower gingiva.

Diagnostic Work-Up

An excisional biopsy should be performed when feasible because some reports have suggested possible local or metastatic spread secondary to a punch or incisional biopsy (219), although this has not been noted in cutaneous melanomas (103). Batsakis et al. (13) found that one third of these lesions were amelanotic, and Hoki et al. (120) noted that 25% were amelanotic.

Metastatic melanoma to the mucosa of the head and neck area is less common (115). It can be differentiated from primary tumors by the presence of normal tissue between subepidermal tumor and the basal layer of melanocytes (13). In a review of the literature. The larynx, tongue, and tonsils are the most common locations for metastases (115).

Prognostic Factors

Batsakis et al. (13) found >0.5 mm invasion to be a poor prognostic factor. Trapp et al. (268) noted this to be true only in patients with more than 0.7 mm invasion. Lymph node involvement is not a prognostic factor. Mucosal melanomas fare worse than their cutaneous counterparts (137), suggesting a lack of immunologic competence (136,222).

Management and Results of Therapy

In a review of the Japanese literature, Umeda et al. (271) found a local tumor control rate for stage I and II disease of 58% (7/12) in surgically treated patients with oral melanomas and a minimum follow-up of 3 years. Similar rates of failure have been reported, even with radical en bloc excisions (20% to 42%). Because the main cause of treatment failure is distant metastases and because almost no patient has clinically evident nodal metastases at presentation, an elective neck node dissection is not recommended by some investigators (27). This subject is still controversial because 30% to 60% of patients may later develop nodal disease (23).

Ohya et al. (195) treated six patients with oral cavity melanomas with irradiation as a component of therapy (dose range 2,000 R to 8,900 R, five treated preoperatively, one postoperatively). All lesions were locally controlled with follow-up of 25 to 109 months; 3/7 patients were alive with 4 to 5 years of follow-up. Harwood and Cummings (108) treated 12 cases and added 12 cases from the literature for a total of 24 patients and 25 lesions. Local tumor control was achieved in 11/24 (9 to 54 months' follow-up). Six of seven tumors treated with 4-Gy fractions or larger were controlled, versus 5/18 treated with smaller fractions.

Postoperative radiation therapy was reported by Panje and Moran (203). Five patients were treated with 60Co or 4-MV photons, 50 to 60 Gy, 1.5 to 2 Gy per fraction, over a period of 5 to 6 weeks. Only one patient showed local control; the other patients failed 6 to 12 months after irradiation.

Kingdom and Kaplan (140) described results in 13 patients with mucosal melanoma of the nasal cavity and paranasal sinuses treated with surgical resection. Eight had microscopically negative margins. Seven patients received postoperative irradiation (30 to 62 Gy). The neck was treated in three patients with doses of 30 to 50 Gy. The local tumor recurrence rate was 85% (11/13), with a mean interval from primary tumor treatment to recurrence of 16 months. Metastatic neck disease developed in two patients and distant metastases in four. Patients receiving postoperative irradiation had increased diseasefree interval and prolonged survival. Negative surgical margins were not predictive of a more favorable outcome. The investigators recommend resection of tumor with negative margins and postoperative irradiation for the treatment of all patients with mucosal malignant melanoma.

Patients with nasal cavity/paranasal mucosal melanoma have a median survival of 24 months. Five-year diseasefree survival rates of 25% have been reported (269). Patients with laryngeal melanoma had a 13% 5-year diseasefree survival rate (137). Because of the poor results obtained and because 37% of patients had associated adjacent pigmentation (222), some investigators recommend prophylactic excision of all melanocytic nevi. Because the results with irradiation are comparable with those of surgical series and because of the poor survival of these patients due to distant metastases and not locoregional failure, irradiation alone, with surgery for salvage, should be seriously considered as the primary treatment for mucosal melanomas of the head and neck (111)

Lentigo Maligna Melanoma

Natural History

Lentigo maligna (Hutchinson's melanotic freckle [124] or circumscribed precancerous melanosis of Dubreuilh) and its invasive counterpart, lentigo maligna melanoma (LMM), are well-recognized clinicopathologic entities (46). LMM comprises about 10% of all melanomas in the head and neck, occurs predominantly on the face and ears of elderly persons, and generally has a very long natural history, frequently reaching a large size before diagnosis. Approximately one third of lentigo maligna lesions, if left untreated, will eventually transform into invasive LMM.

Tannous et al. (260) hypothesized that lentigo maligna can be divided into two categories: one represents a pigmented lesion that is a precursor to melanoma, and the other melanoma in situ. Also, they hypothesized that in some patients there is a progression to malignant melanoma.

Clinical Presentation and Diagnostic Work-Up

These lesions appear as circumscribed and later as more diffuse areas of hyperpigmentation of the skin. They may develop some superficial nodularity and eventual ulceration as they become more invasive. In 10% of the latter patients, regional and distant metastases eventually develop. The 10% metastatic spread in LMM contrasts with the 25% metastatic tendency in nodular melanomas arising in superficial spreading melanomas and a 50% metastatic spread in nodular melanomas arising de novo.

The diagnostic work-up of these patients is similar to that of patients suspected of having malignant melanoma. Biopsies of the lesion are required to obtain histopathologic confirmation of the diagnosis. Careful physical examination must rule out any areas of extension or regional or distant spread.

General Management

The usual treatment of lentigo maligna and LMM has been surgery, with approximately 1- to 2-cm margin of normal skin (8).

Hill and Gramp (118) reported on 66 cases of lentigo malignant melanoma; 38% of which required two excisions or more to clear the tumor and 32% of cases showed evidence of invasive melanoma. Only one case has recurred thus far, and none have developed metastatic disease.

Because of the low incidence of regional lymph node metastases, elective lymph node dissection is not indicated. For larger lesions, wider surgical excision with skin grafting has been reported to give poor cosmetic results.

Cohen et al. (47) reported their experience with Mohs microsurgery, which was performed in 26 patients with lentigo maligna and 19 patients with lentigo maligna melanoma. After a median follow-up of 58 months (214.3 patient years), there was one recurrence, in a patient with five prior recurrences before Mohs micrographic surgery.

Kuflik and Gage (149) treated 30 patients with cryosurgery. Lesions ranged from 1.3 to 4.5 cm in diameter. Lesions recurred in two patients (recurrence rate of 6.6%) who were successfully retreated with cryosurgery. Eleven patients observed for more than 5 years showed no recurrences.

Radiation therapy with various techniques has been frequently used in the treatment of these patients, particularly those with larger lesions, because of minimal morbidity and generally excellent cosmetic results (Fig. 45.16).

Figure 45.16.Lentigo maligna melanoma of face before (A) and 6 years after (B) 50 Gy in 25 fractions delivered with 9-MeV electrons and bolus.

Radiation Therapy Techniques

As in other skin lesions, the portals should be carefully designed to include the entire tumor with adequate margin (1 cm for lesions <2 cm and 2 cm for larger tumors). Because Miescher's irradiation technique used very superficial x-rays, with 50% depth dose being at approximately 1 mm, there is the possibility of local recurrence if dermal extension is unrecognized. Therefore, Harwood and Lawson (111) recommend using minimum x-ray energies of 100 keVp and preferably 140 to 175 keVp to treat these patients. Superficial x-rays (100 to 200 keVp) with adequate filtration or electrons (6 to 9 MeV) with appropriate thickness of bolus (about 1.5 cm) are adequate for most patients. Doses of 45 to 50 Gy in 15 to 25 fractions delivered over 3 to 5 weeks will control the disease in most patients. We recommend delivering 3 to 3.5 Gy, three times weekly, every other day, to a total of 50 Gy, depending on the size and thickness of the lesion. Elective irradiation of the regional lymphatics is not necessary.

Careful follow-up with clinical examinations and photographs of the lesion is essential to ascertain the continuing regression of the tumor.

In patients on whom surgical excision is performed, postoperative irradiation is recommended if positive margins are found (111). Doses are similar to those stated earlier.

Results of Therapy

Kopf et al. (146) reported six recurrences in 16 patients treated with 50 keV x-rays (500 cGy doses).

Harwood and Lawson (111) described 13 patients with lentigo maligna treated with radiation therapy: 11 had local tumor control, one had an edge recurrence salvaged by irradiation, and one had residual tumor (alive and well 11 years after treatment for the recurrence). One patient alive at 2 years refused further treatment. Of 19 patients irradiated for LMM, 17 had tumor control with radiation therapy alone for periods ranging from 6 months to 6 years. One patient had a central recurrence that was salvaged by surgery (alive and well 5 years after treatment of recurrence). One patient died of intercurrent disease <3 months after irradiation. No patient has developed lymph node or distant metastases in either group.

Tsang et al. (269) described results in 54 patients treated with radiation therapy or surgery. Younger patients with smaller lesions were treated with surgical excision (18 patients) and achieved actuarial tumor control of 94% at 3 years. Older patients with larger lesions located in the head and neck area were treated by radiation therapy (36 patients), with an actuarial tumor control rate of 86% at 5 years. No patient developed metastatic melanoma. The late cosmetic appearance was acceptable in the majority of irradiated patients, with 11% showing poor cosmesis because of progressive skin pallor, atrophy, and telangiectasia in the treated area.

Sarcomas of the Head and Neck

Natural History

Sarcomas account for <1% of malignant neoplasms in the head and neck. The most frequent histological type is malignant fibrohistocytoma (29%), while the least common is liposarcoma (1%) (35). The histology is complex and requires immunochemical analysis including osteosarcoma, angiosarcoma, chondrosarcoma, hemangiosarcoma, leiomyosarcoma, liposarcoma, malignant fibrous sarcoma, rhabdomyosarcoma, malignant schwannoma, neurofibrosarcoma, and synovial sarcoma. Fibrosarcoma, angiosarcoma, leiomyosarcoma, and rhabdomyosarcoma are the most common types but this varies in published reports. Table 45.15 summarizes the distribution of various histologic types in 1,127 patients. Distribution of these sarcomas was 33% in the scalp or face, 26% in the orbit or paranasal sinuses, 14% arising from upper aerodigestive tract including larynx, and 27% in the neck. Synovial sarcomas are rare soft-tissue malignancies in the head and neck region; they account for 3% to 5% of head and neck tumors. Histologic, immunohistochemical, and characteristic chromosomal translocation findings are necessary for diagnosis. The poor prognosis of this sarcoma justifies radical surgery with postoperative radiation.

Table 45.15. Histologic Distribution of Head and Neck Sarcomas: Review of the Literature

Radiation-induced sarcoma of the head and neck is a rare long-term complication of treatment. The rarity of this tumor is reflected in the very few series reported in the English language medical literature (132,207). When they do occur, most appear at least 10 years following radiation therapy. There is a possibility of a postirradiation sarcoma whenever a suspicious lesion is seen, regardless of the amount of time that has passed since radiation therapy was administered. The original pathology should be re-examined to ensure that the original tumor was diagnosed correctly. Electron microscopy can be useful in differentiating sarcomatous-appearing epithelial lesions from true soft-tissue sarcomas.

The incidence of radiation-induced sarcomas of the head and neck is, however, likely to increase due to progressive aging of the population combined with improved survival in head and neck cancer patients. This problem can be extremely challenging and the overall outlook has been reported to be very bleak. Patel et al. (207) reviewed 69 cases reported in the English medical literature since 1966 and pooled this information with their experience in treatment of 10 patients. This group was compared for survival with 124 patients with a diagnosis of head and neck sarcoma registered on the Head and Neck Sarcoma database at the Royal Marsden Hospital. There was no site prediction for radiation-induced sarcoma of the head and neck, but malignant fibrous histiocytoma was the most common pathologic diagnosis. The period of latency between initial radiation therapy and diagnosis ranged from 9 to 45 years with a median of 17 years. Surgery was the mainstay of treatment, and follow-up ranged from 6 months to 15 years with a median of 48 months. The actuarial 5-year diseasefree survival rate in these patients was 60%.

Clinical Presentation and Diagnostic Work-Up

Clinical presentation varies with the primary site of disease.

Tumors arising from the aerodigestive tract usually present with nasal bleeding, a palpable mass in the neck, or difficulty in swallowing or breathing. Of tumors arising from the base of skull or the nerve sheath, cranial nerve deficit is the most common presentation. Diagnostic work-up follows that of softtissue sarcomas of other sites in the body. With early lesions, radiographs or CT may show only nonspecific opacification, soft-tissue swelling, and occasionally bone destruction. Table 45.1 outlines the suggested diagnostic work-up. MRI, especially with gadolinium contrast, may be used as a supplement or alternative to CT scanning (161). A CT scan of the chest is also mandatory for staging work-up.

The American Joint Committee on Cancer (AJCC) staging system for soft-tissue sarcomas is based on histologic grade, the tumor size and depth, and the presence of distant or nodal metastases. The staging system is the same as for sarcomas of the extremities, although specific staging for head and neck sarcomas is not standardized (49).

Prognostic Factors

Prognostic factors for predicting local recurrence or diseasefree survival include anatomic site, treatment modality, tumor histology and grade, tumor size, extension of disease, and surgical margins (69,73,286).

A report from Royal Marsden Hospital showed anatomic location and treatment modality to be independent prognostic factors for local recurrence; tumors of the head had a better local recurrencefree survival than did those of the neck (73). Patients treated with a combination of surgery and radiation therapy had a better recurrencefree survival than did those treated with surgery or irradiation alone. The only significant independent prognostic factor for overall survival was the implementation of definitive surgery versus biopsy. In the above report, the prognostic impact of tumor stage and grade did not reach statistical significance. In contrast, Tran et al. (266) reported that 90% of patients with low-grade tumors were free of disease versus only 16% with high-grade lesions.

Bentz et al. (19) reviewed 111 head and neck sarcoma patients; median duration of follow-up was 51 months; the actuarial 5-year relapsefree disease-specific, and overall survivals were 55%, 52%, and 44% respectively. By multivariate analysis, size and grade significantly influenced all survivals, whereas margin status additionally influenced relapsefree survival.

In 352 patients treated at the Mayo Clinic for primary soft-tissue sarcoma of the head and neck, the 5-year overall survival for nonmetastasizing tumors such as dermatofibrosarcoma, protuberans, and desmoid tumors was 100%. Of the tumors with metastasizing potential that presented without lymph node metastasis or distant metastasis, the 5-year overall survival rate was 62.9% but decreased to 25% when lymph node or distant metastasis was detected at presentation (87).

In 109 soft-tissue sarcomas of all sites, a French study demonstrated that quality of the surgery was one of the most important variables for predicting local recurrences. Tumor size, surgical margins, presence of tumor necrosis, and adequacy of the excision correlated with metastasisfree survival (169).

In 57 patients with soft-tissue sarcomas of the head and neck treated at Massachusetts General Hospital, angiosarcoma had a considerably poorer prognosis than other histologies (257). In addition to tumor grade and size, direct tumor extension to neurovascular structures, bone, contiguous organs, and skin was associated with a higher incidence of distant metastasis. The actuarial 5-year freedom-from-distant-metastasis rates were 70% and 100% for tumors with or without direct extension, respectively.

General Management

Surgery is the preferred initial treatment modality for sarcomas (50). Unfortunately, it is often difficult to achieve complete resection of the tumor, and a high recurrence rate has been observed with surgery alone (82). Extracapsular enucleation of the tumor results in 90% local recurrence because of the presence of microscopic pseudopodia, which tend to grow through the pseudocapsule into the surrounding tissue and the presence of skipped lesions some distance from the main tumor mass. Pathologic analysis of the surgical bed often discloses microscopic extension of tumor. Farhood et al. (82), in a review of 176 cases of adult head and neck sarcomas, reported that the pathologic margins of surgical specimens obtained by wide local excision were positive in more than 50% of cases. This resulted in inferior overall survival for sarcomas of the head and neck when compared with extremity sarcomas (225). Wide local excision, with a 5-cm margin around the pseudocapsule in extremity sarcomas, is associated with better outcome, although approximately 20% will have local recurrence. The criteria for surgical resection are impractical for head and neck sarcomas because of anatomic limitations (49); wide local excision is rarely possible because the tumors extend beyond the confines of origin and in the proximity of vital neurovascular structures. Some retrospective studies have suggested improved local tumor control when combined surgery and external irradiation are used. In 130 patients with soft-tissue sarcomas of the head and neck treated with surgery alone at Royal Marsden Hospital, the overall 5-year survival was 50%; local tumor control was only 47%, and local recurrence was the cause of death in 63% of cases. Patients treated with combined-modality treatment (surgery and irradiation) had less extensive surgery, yet local recurrencefree survival was longer (73).

Radiation therapy, by external beam or brachytherapy, plays an important adjunctive role in the management, especially for tumors where en-bloc resection with negative margin is not possible (255,286). Chemotherapy regimens are available for soft-tissue neoplasms primarily designed to improve local tumor control (175). Survival is predicted on the incidence of local recurrence and risk of distant metastasis, both of which are influenced by tumor grade.

A systematic review of radiation therapy trials was performed by the Swedish Council of Technology Assessment in Health Care (SBU) (255). This synthesis of the literature on radiation therapy for soft-tissue sarcomas is based on data from five randomized trials. Moreover, data from six prospective studies, 25 retrospective studies, and three other articles were used. In total, 39 scientific articles were included, involving 4,579 patients. The results were compared with those of a similar overview from 1996 which included 3,344 patients. There was evidence that adjuvant radiotherapy improves local tumor control in combination with conservative surgery with negative, marginal, or minimal microscopic positive surgical margins. There are still insufficient data to establish that preoperative radiotherapy is favorable compared to postoperative radiotherapy in patients presenting primarily with large tumors. The preoperative setting results in more wound complications. There is no randomized study comparing external beam radiotherapy and brachytherapy. The data suggest that external beam radiotherapy and low-dose-rate brachytherapy result in comparable local control for high-grade tumors. Some patients with low-grade soft-tissue sarcomas benefit from external beam radiotherapy in terms of local control. Brachytherapy with a low dose rate for low-grade tumors seems to be of no benefit, but data are sparse. In two small studies investigating hyperfractionation schedules there was no indication of improvements compared to daily fractions of 2 Gy.

Mesenchymal chondrosarcoma of the sinonasal tract is a rare, malignant tumor of extraskeletal origin (141). Thirteen patients with sinonasal mesenchymal chondrosarcoma presented with nasal obstruction (n = 8), epistaxis (n = 7), or mass effect (n = 4), or a combination of these. The maxillary sinus was the most common site of involvement (n = 9), followed by the ethmoid sinuses (n = 7) and the nasal cavity (n = 5). All cases were managed by surgery with adjuvant radiation therapy (n = 4) and/or chemotherapy (n = 3). The overall mean survival was 12.1 years, although 5/6 patients who developed local recurrences died of disease (mean survival, 6.5 years). Six patients were alive and disease free (mean survival, 17.3 years), and two patients were lost to follow-up (141).

A multidisciplinary discussion before the initiation of treatment is required to formulate the best approach for radiation delivery, surgical technique, and mode of reconstruction.

Radiation Therapy Techniques

The general principles for radiation therapy of head and neck sarcomas are similar to those of soft-tissue sarcomas. Complete coverage of the surgical bed and scar with adequate margins (3 to 5 cm) is required (210). However, because of the proximity of critical and radiosensitive organs (eyes, spinal cord, brainstem), selecting optimal portal margins without seriously compromising the functioning of these organs is an art. Techniques similar to those used in epithelial tumors of the head and neck can be applied to sarcomas. In general, 55 to 60 Gy is needed for postoperative adjuvant irradiation, and an additional 10- to 15-Gy boost is recommended if the surgical margins are close (≤3 mm) or involved by tumor. Some institutions prefer preoperative irradiation of 45 to 50 Gy. Special attention should be directed to limiting the dose to critical structures. Use of a 3D treatment technique can be considered as demonstrated in Fig. 45.11.

Results of Therapy

Because of the propensity for sarcomas to invade the surrounding tissues, complete surgical clearance may be difficult (1). In a series from UCLA, attempted en-bloc resection left residual tumor at the surgical margins in 52/127 patients (266). The incidence of local recurrence was high (60%) with surgery alone.

In a retrospective report of 73 patients with sarcomas of the head and neck treated at Princess Margaret Hospital, the 5-year cause-specific survival was 62%, with a local recurrence rate of 41% and a distant metastasis rate of 31% (159). Extension to adjacent structures, high-grade tumor, and tumor >10 cm were associated with poor survival. Gross residual tumor after surgery was also associated with a high local recurrence rate (75%) despite the addition of radiation therapy. Patients with clear surgical margins or only microscopic involvement fared much more favorably and had a similar local tumor control rate (74% and 70%, respectively), provided adjuvant irradiation was given. Because of the difficulty in obtaining wide surgical margins, 68% of the patients died as a result of uncontrolled local disease. These data substantiate the importance of surgical margins as well as the contribution of adjuvant irradiation (159).

Colville et al. (49) reported on 41 male and 19 female patients treated with head and neck soft-tissue sarcomas, overall 5-year survival was 60%. Twenty-five patients had surgery alone, 20 surgery and pre- or postoperative radiation therapy and 15 received nonsurgical treatment. With mean follow-up of almost 4 years the 5-year local tumor control was 56% in the surgical and 40% in the nonsurgical group (more advanced and aggressive tumors). The 5-year survival was 70% and 40%, respectively.

Penel et al. (211) recorded their experience with 28 adult head and neck soft-tissue sarcomas. The median age was 45.7 years (range: 18 to 86). The male/female ratio was 15:13. The most common subtype was rhabdomyosarcoma (seven cases). Twenty-two patients presented with previous inadequate resection performed elsewhere before admission. Nineteen patients had surgery (complete resection in 13 cases). Associated treatments were neoadjuvant chemotherapy, adjuvant chemotherapy, and postoperative radiotherapy in 4, 3, and 10 cases, respectively. The 2-year overall survival rate was 56%.

Pandey et al. (202) reported on 22 cases of head and neck sarcomas (neck, lower jaw, tongue, cheek, scalp, and maxilla were the commonest sites affected). None of the patients had palpable neck nodes or distant metastasis at presentation. All the patients were treated with primary surgical resection, followed by adjuvant treatment in 14 cases (63.6%). After a median follow-up of 14.5 months, two patients died, six developed local recurrence, four developed metastatic disease, and another patient developed a second primary sarcoma. The overall 5-year survival was 80%, while the 5-year diseasefree survival rate was 24.1%.

Barker et al. (9) published a review of 44 patients diagnosed with nonmetastatic soft-tissue sarcoma in a head and neck. The most common tumor histologies included malignant fibrous histiocytoma (15 patients), angiosarcoma (nine patients), fibrosarcoma (six patients), and leiomyosarcoma (six patients). The median overall survival for all patients was 79 months. The actuarial 5-year local tumor control was 55% and was highly correlated with the extent of surgical excision: 25% for subtotal resection/debulking, 65% for wide local excision, and 100% for radical excision. Local tumor control at 5 years was 60% for patients treated with both surgery and radiotherapy, 54% surgery alone, and 43% for radiation alone. Adjuvant radiation therapy significantly improved the local control rates (from 25% to 54%) for patients with close (<2 mm) or positive surgical margins. Of 14 patients with locoregional failure in whom salvage was attempted, nine (64%) were rendered disease free.

Rapidis et al. (221) reported on 25 patients with head and neck sarcomas with follow-up ranging from 8 to 144 months. Twenty-three patients were treated with surgery as the primary modality; 14 with surgery alone. Clear margins were obtained in all of them and local control was achieved in 12/13. The 5-year survival for the entire group was 40%. Reported results of treatment of soft-tissue sarcomas is summarized in Table 45.16.

Table 45.16. Treatment Results of Adult Soft-Tissue Sarcomas of the Head and Neck

Tumor Characteristics

Several series have shown that tumor grade and size dictate the outcome of patients with head and neck sarcomas such as leiomyosarcoma, rhabdomyosarcoma, and malignant fibrous sarcoma (50). Farhood et al. (82), in a review of 176 adult head and neck sarcomas, found that only 20% of the patients with high-grade tumors were alive 10 years after treatment, as opposed to 88% of patients with low-grade tumors. Greager et al. (101) noted mean survival of 93 months in patients with lowgrade tumors smaller than 5 cm versus 15 months for those with high-grade lesions larger than 5 cm. Weber et al. (282) described a 45% 10-year survival rate for patients with tumors smaller than 5 cm versus 10% for those with tumors 5 cm or larger.

Many series have reported that chondrosarcoma is not a radiosensitive tumor, and radiation therapy has no role in it treatment. However, some reports have demonstrated the contribution of radiation therapy in this histology. Harwood et al. (110) reported that 6/12 patients with chondrosarcoma of the bone receiving irradiation alone were alive with no evidence of disease at periods ranging from 3 to 16 years. McNaney et al. (180) described a 65% survival rate at 2 years in 20 chondrosarcoma patients who received primary radiation therapy. Tumor grade was the most important prognostic factor.

Osteogenic sarcoma of the head and neck has a pattern of recurrence different from similar tumors elsewhere in the body. Head and neck osteosarcomas are usually high grade; they have a very high incidence of local recurrence but a lower risk of distant metastases. Several studies have used adjuvant irradiation and chemotherapy, which commonly results in improved locoregional tumor control and survival. Chambers and Mahoney (38) reported a 73% 5-year survival rate in patients with osteogenic sarcoma of the head and neck treated with highdose preoperative irradiation followed by wide surgical excision. Similar observations were reported by Tran et al. (266).

Chemotherapy

Head and neck soft-tissue sarcomas frequently metastasize; 25% of patients in the UCLA study had distant metastases (266). The role of adjuvant chemotherapy to improve diseasefree survival in sarcoma of the head and neck is controversial. The results were disappointing in a randomized trial conducted by the National Cancer Institute (99). Also, in a review of 11 randomized trials of adjuvant chemotherapy in nonextremity soft-tissue sarcomas, Elias and Antman (74) observed that none showed a statistically significant benefit in survival. Unlike with softtissue sarcomas of the extremities, in which distant metastasis is the most common cause of death, the majority of deaths in sarcomas of the head and neck are associated with local failure. Approximately half of the distant metastases were detected after local recurrence occurred (159). Chemotherapy did not appear to affect local tumor control (74).

McKenna et al. (179) reported on 16 adult patients with highgrade soft-tissue sarcomas of the head and neck treated with surgery, radiation therapy, and chemotherapy. With a median follow-up of 43 months, 12 patients (75%) achieved local tumor control, and 10 patients were disease free. In a series of 94 patients treated at UCLA (266), local control was achieved in 52% of patients treated with surgery alone and 90% of those receiving adjuvant irradiation and/or chemotherapy.

For preoperative neoadjuvant chemotherapy, which supplements radiation therapy to downstage disease before surgery, satisfactory results are available only for sarcomas of extremities (212,227). The Institute Gustave-Roussy reported that two thirds of locally advanced lesions could be rendered operable after neoadjuvant chemotherapy and/or radiation therapy (227).

With the exception of rhabdomyosarcoma, postoperative adjuvant chemotherapy should be given only in a clinical trial setting.