Definition

(ICD-O code 8200/3)

Adenoid cystic carcinoma is an infiltrating malignant tumor, often with a characteristic cribriform appearance, derived from minor seromucinous salivary glands that can be found throughout the mucosa of the nose and paranasal sinuses.

Etiology

Since it was first described by Billroth in 1856, no obvious causative factors have been found.1

Synonyms

Synonyms have been numerous and include cylindroma, adenocystic carcinoma, cribriform adenocarcinoma.

Incidence

A common aphorism is that 70% of minor salivary glands tumors are malignant, although the figures vary from 65 to 88% depending on the series as none are especially numerous.2–5 Nonetheless, adenoid cystic carcinoma (ACC) is rare, constituting <2% of all tumors of the nose and sinuses.6 In a study of 242 malignant tumors in the nose and sinuses in Denmark between 1995 and 2004, 5% were adenoid cystic carcinoma.7

Site

As a consequence of the origin from minor salivary glands, the majority occur in the oral cavity, especially on the palate (~25%) from whence they may involve the nasal cavity and maxillary sinus.6,8 After the palate, the majority affect the maxillary sinus (57%)9 and nasal cavity (14–32%).10–12 From there the tumor invades the skull base and the pterygopalatine and infratemporal fossae. ACC has also been described extending from the sphenoid or nasopharynx into the clivus.13 Very rarely they may be metastases from elsewhere such as the trachea.14 For our own series, the distribution is shown in Table 7.6.

Diagnostic Features

Histological Features and Differential Diagnosis

Adenoid cystic carcinoma is notorious for its low-key presentation. Macroscopically normal mucosa may yield positive biopsies even in the absence of an obvious mass. Histopathologists have classified the tumor into tubular, cribriform (“Swiss cheese”), and solid in order of increasing aggressiveness, although this has not always been supported and mixed forms frequently occur. The appearance is often rather characteristic, but it must be distinguished from other salivary gland tumors, ameloblastoma, and even basal cell or basaloid squamous cell carcinoma. Immunohistochemistry may be helpful showing myoepithelial and epithelial differentiation with reactions to p63, S100, SMA, and pancytokeratin.21–23 The cells are sometimes strongly immunoreactive to c-Kit (CD117) which may help distinguish it from other salivary gland tumors24 and could have therapeutic implications (see later).

Estrogen receptor α was found in 75% of adenoid cystic tumors, while only 17% were positive for estrogen receptor β, the clinical relevance is unknown.25

Perineural spread has been attributed to perineural lymphatic infiltration and occurs in many patients.26,27

Adenoid cystic carcinoma: personal series 1976–2010

	n	Age		M:F	CFR	Maxill	LR	MFD	ESS	Other
		Range (y)	Mean (y)
	54	31–89	52.6	1.7:1	19	11 (2 bilateral)	9	4	2	9
Orbital clearance	19				11	5	3
DXT	24				10	5	1	2	1	5
Overall actuarial survival
5 y					61%
10 y					31%
15 y					31%
Abbreviations: CFR, craniofacial resection; Chemo, chemotherapy; DXT, radiotherapy; ESS, endoscopic sinus surgery; F, female; LR, lateral rhinotomy; M, male; Maxill, maxillectomy; MFD, midfacial degloving; y, year(s).

Furthermore, embolization along nerves has been observed.20 This propensity compromises complete surgical extirpation.

Natural History

Adenoid cystic carcinoma has a unique natural history that almost invariably results in the demise of the patient unless some other fatal event intervenes. However, this can extend over several decades. Indeed, it is an enormous disappointment for both patient and clinician to find true recurrence occurring 20 years after an ostensible cure. The majority of patients present with advanced disease (77%) though few have lymphatic or metastatic disease (2 to 3%).12 However, the disease is characterized by frequent local recurrence and early perineural and hematogenous spread (38%). It has one of the highest rates of local recurrence of all sinonasal malignancy (75 to 90% and possibly 100% with long enough follow-up) (Table 7.7), but it should be noted that not all patients experience an indolent course. It is unknown why this occurs rather than conventional lymphatic spread, although the latter has been reported in 9 to 16% in some series.3,27,34,35 Systemic disease may affect 50% or more, including the lungs, brain, bone, liver and even skin, rarely occurring at presentation, sometimes taking decades to appear. Notwithstanding this, patients may survive for significant periods with systemic disease which produce little clinical symptoms, particularly in the lung.32 This has led to more aggressive management of this problem.36 In ACC the time to metastasis can range from 13 to 77 months37 and the disease-free interval can range from 1 month to 19 years.20,33 Given the often significant time interval, fine needle aspiration of the lung lesion should be undertaken to confirm that it relates to the original ACC.38

Treatment

Given the natural history, any treatment aims to prolong the interval to local recurrence balanced against the morbidity of the treatment. Radical surgery has usually been undertaken in the form of craniofacial resection and/or maxillectomy together with orbital clearance. Even bilateral maxillectomies have been undertaken in some cases using a midfacial degloving approach,39 but there has been an increasing reticence to radically extirpate all cranial nerves in the vicinity to the skull base and beyond as no survival advantage can be shown to justify the resulting complications. Resection of the cavernous sinus contents and internal carotid artery is associated with high morbidity and mortality as might be expected.40

Most recently endoscopic techniques have been tentatively undertaken,41–48 although surgeons should be familiar with the unusual behavior of this tumor and the difficulties in determining extent. There are only a few anecdotal cases in the literature thus far with limited follow-up. Endoscopic techniques may be combined with external approaches in selected cases.

Neck dissection is rarely required and certainly has no prophylactic role. However, resection of pulmonary metastases is increasingly being undertaken either by wedge resection or lobectomy as this may “buy” a reasonable period of survival.36 Patient selection will be determined by local control and whether there is any other disseminated disease as well as the individual’s cardiopulmonary reserve.

Although radiotherapy is often given, its role is not well determined. It is rarely used as a single primary modality as the tumor is not classically radiosensitive. Combined with surgery it may delay recurrence but cannot be shown to significantly improve “cure.” Although there have been recent reports of the use of IMRT,49,50 determining the field might be problematic and long-term follow-up in reasonable numbers of patients is lacking. In a study by Coombs et al advocating IMRT in 46 patients with sinonasal malignancy, 20 of whom had adenoid cystic carcinoma, median follow-up was only 16 months (range 3–40 months).49 Even with this short follow-up, local control dropped from 85% at 1 year to 49% at 3 years using this as the primary modality, and distant control was 83% at 1 year and 0% at 2 years in the group as a whole.

Proton beam therapy is also being used in a few centers, but again numbers and follow-up preclude any conclusions.51,52 Disease-free survival was 56% at 5 years in a series of 23 patients with ACC in the skull base. Stereotactic radiosurgery with the gamma knife has also been reported53 in 29 of 34 patients with salivary gland malignancy involving the skull base after failed neutron beam therapy.

Chemotherapy has been used for some years to treat both systemic and advanced local disease.54–57 Regimes generally used cisplatin and/or doxorubicin, 5-FU, mitomycin-C, or cyclophosphamide. More recently epirubicin, mitoxantrone, and gemcitabine have been used in Phase II trials58–60 and there has also been interest in docetaxel–cisplatin and paclitaxel–carboplatin in advanced disease.61,62 C-Kit expression has led to trials of oral imatinib which has a synergistic interaction with cisplatin63 but without encouraging results in vivo.64,65

The neuropathic pain experienced by patients may be helped by drugs such as carbamazepine and gabapentin.

Outcome

As already stated, this disease can recur throughout a lifetime so 5-year survival figures are particularly deceptive and some patients may surprisingly survive for several years with secondary disease (Table 7.8). Spiro et al found a 10-year survival of 7% in their early series of 242 cases4 but this improved in a later series (1960–1986)32 with reduction in both treatment failure and local recurrence, which they attributed to the regular addition of postoperative radiotherapy in more recent cases. However, there was no significant improvement in the frequency of distant metastases with or without radiotherapy. Lupinetti et al11 also showed that surgery with postoperative radiotherapy provided the best overall and disease-specific survival compared with other treatment modalities (p = 0.018 and p = 0.05, respectively), while surgery alone significantly improved survival (p < 0.001) as compared with radiotherapy or chemotherapy, although this may also reflect patient selection.

In our craniofacial series the patients fared somewhat better, with a 5-year survival that dropped from 61% to 31% at 15 years.66 As might be expected, skull base invasion is a significant factor in survival in this and other series.12 In a series of 59 patients of which 13 (22%) were sinonasal, the overall survival rates were 76% and 40%, and 87% and 65% for disease-free survival at respectively 5 and 10 years after combined surgery and radiotherapy, but figures for the sinonasal lesions alone were not given. However, the authors confirmed that clinical nerve involvement was an important prognostic factor.67 As previously noted, adjunctive radiotherapy may delay recurrence but cannot be shown to improve cure.11,12,32,33

Overall, major salivary gland adenoid cystic carcinomas do better than minor salivary gland lesions, and palatal lesions do better than those closer to the skull base.32 In a large European study of 2,611 cases of adenoid cystic carcinoma at all sites, nasal cavity lesions did worse than oral cavity lesions.68 The large MD Anderson study12 showed that cribriform lesions did best and solid forms worst of the histological types.

The numbers treated endoscopically are too few and with too short a follow-up from which to draw any conclusions, usually in single numbers in otherwise small series of mixed malignant tumors.41–48 However, one could argue that in a disease with a generally poor outcome, wide-field endoscopic resection with its associated low morbidity might offer advantages in a carefully selected group of patients.

Patients may die from local disease, often without secondaries,33 although they can succumb to metastatic disease alone. Continued treatment of local recurrence can therefore be worthwhile, even in the presence of distant disease. Similarly, the resection of pulmonary metastases is increasingly undertaken. In the short term, pulmonary resection may be helpful. Liu reported 84% 5-year survival in a small group of patients with ACC in the head and neck, too short a follow-up to be meaningful, and all were dead at 14 years.69 Interestingly, no difference was shown between patients with a single nodule and those with multiple lesions. A more compelling study by Bobbio et al70 considered 9 patients with pulmonary metastases from ACC without locoregional disease who underwent metastasectomy who had a mean survival of 72 months compared with a mean survival of 62 months in 11 patients with lung secondaries who did not undergo surgery. Unfortunately, the second group may have had more extensive lung involvement, so conclusions are difficult to draw. Overall, distant metastases remain one of the most important factors in determining survival.11,32,68