High-resolution ultrasound is an imaging modality of potential applicability for the evaluation of a parotid or submandibular region mass in a child. Ultrasound can be particularly helpful in distinguishing solid from cystic masses. Doppler ultrasound can evaluate the vascularity of the lesion. Ultrasound has the additional advantages of being noninvasive, not involving radiation, and typically being able to be performed with or without the need for sedation [49]. Any solid mass of the parotid or submandibular region, particularly one without associated inflammatory manifestations, should signal the need for more detailed imaging via computed tomography (CT) or magnetic resonance imaging (MRI) [23]. The age of the child with respect to the need for sedation or general anesthesia may dictate which procedure is initially performed.

Even younger children may tolerate the performance of a relatively short duration CT without sedation. CT is useful in evaluating lymph node metastasis (LNM) and is particularly beneficial if there is a concern of osseous involvement or tumor invasion of neural foramina [50]. Glandular signal intensity and enhancement pattern can, however, parallel that of adjacent musculature on CT. This is particularly true in children, because of the relative lack of fat in the parotid space, soft tissue differentiation in this region may be obscured [41].

MRI, on the other hand, clearly demarcates tumor from glandular tissue in multiplanar fashion. MRI also has the advantage of being free of radiation. T1-weighted images of normal parotid have an image signal intermediate between fat and muscle, whereas submandibular tissue is closer to muscle in intensity [51]. Parotid tumors are well visualized on T1-weighted MR images because of the hyperintense background of the gland. The T1-weighted image gives an excellent assessment of the margin of the tumor, its deep extent, and its pattern of infiltration [52]. Coupled with fat-saturation, contrast enhanced T1-weighted imaging can be used to assess for perineural spread [53]. MR features strongly suggestive of parotid malignancy are poorly defined margins with infiltration into the parapharyngeal space and surrounding musculature [54]. The conventional wisdom is that a hyperintense mass on T2-weighted images is benign and a mass of low to intermediate signal intensity is malignant [52]. A number of studies have, however, shown this not to be a reliable discriminative factor [54]. Innovative techniques, such as diffusion weighted MRI and intravoxel incoherent motion MRI [55] have shown some early promise in distinguishing benign from malignant tumors.

Fine needle aspiration (FNA) cytology in adults is a diagnostic tool that has a reliable sensitivity and specificity for the assessment of salivary gland pathology [56]. Although, its use in children has some advocates [57], the role of FNA is controversial because of the limited established specificity in pediatric salivary gland tumors [58] and because of the need for sedation or even general anesthesia for its performance in younger children [22]. Retrospective study review reveals FNA to be rarely undertaken and the accuracy to be as low as 33 % when it was performed [16].

Incisional biopsy is a consideration in masses involving the tail of the parotid region, and is particularly indicated in clinically unresectable lesions for which a diagnostic biopsy alone is required. Under most circumstances, total excision of the submandibular gland or superficial parotidectomy are the preferred techniques, providing both an excisional biopsy specimen and potential definitive therapy [59]. The marginal mandibular and facial nerves, respectively, need be identified and preserved during these procedures.

There is no specific staging system for pediatric salivary gland tumors. For parotid, submandibular, and sublingual neoplasms, the adult tumor node metastases (TNM) clinicopathologic staging system unified between the American Joint Committee on Cancer (AJCC) and the International Union Against Cancer (UICC) is often applied (Table 39.3; [60]). For minor salivary gland malignancies, the staging system for squamous cell carcinomas, depending on tumor site (oral cavity/oropharynx/nasal cavity/nasopharynx) is alternatively used. Children are often diagnosed at an earlier stage than adults [11].

Owing to the wide range of neoplasms and the diversity of their biological behavior, there is no single grading system for salivary gland cancers. Certain cancers are acknowledged as low grade (for example some adenocarcinomas, basal cell adenocarcinoma, and acinic cell adenocarcinoma) and others as high grade (for example some adenocarcinomas, squamous cell carcinoma, and undifferentiated carcinoma) [60]. Alternatively, salivary gland malignancies have been stratified into low-risk and high-risk tumors, low-risk tumors being those that do not require treatment beyond excision, whereas high-risk being those that do (Table 39.4) [61]. The caveat is that high grade versions of ‘intrinsically’ low grade tumors exist as do low grade versions of typically high grade tumors [61].

For some of the more common adult salivary gland malignancies, specific grading systems have been utilized. For example, three grades of AdCC have been described based on their growth pattern. The tubular and cribriform patterns are considered low grade, with increasing solid components contributing to higher grade, grade 3 being almost entirely solid [60, 61]. This system is controversial as some authors grade any tumor over 30 % solid as high grade [62], while others deem that up to 50 % of the tumor can be solid and still be termed intermediate [63].

A number of quantitative point scoring grading schemes also exist for mucoepidermoid carcinoma [64]. These assess characteristics, such as intracystic component, presence of neural invasion or necrosis, mitotic rate, and cellular anaplasia to stratify tumors into low, intermediate, or high grade [60]. Most mucoepidemoid carcinomas in children have been found to be low grade, with relative frequencies of low (76 %), intermediate (14 %), and high (10 %) grade (Table 39.5).

Cellular differentiation has also been used to grade salivary gland malignancies into well, moderately, poorly, undifferentiated, or anaplastic categories. Using this system, a significantly higher percentage of tumors in children (88 %) are reported to be well differentiated or moderately differentiated in comparisom to adults (49 %) [11].

The treatment of choice of most salivary gland neoplasms is complete removal with adequate margins [14]. For submandibular tumors, complete excision of the gland is recommended [59]. In the case of parotid tumors, depending on the lesion, superficial or total parotidectomy is indicated. Parotidectomy is considered a safer and more definitive procedure than tumor enucleation as the latter results in higher rates of recurrence and facial nerve dysfunction. Superficial parotidectomy is the treatment of choice when the tumor is lateral to the facial nerve and subsequent histology reveals a low-grade malignancy. Total parotidectomy is recommended when there is deep lobe involvement, suspected or confirmed high-grade tumors, or tumors with aggressive malignant potential, such as those with facial nerve involvement, multiple intraparotid masses, or cervical metastasis [14, 59]. Definitive surgery allows for very good local control rates approximating 97 % in some series [16, 42].

Resection of the facial nerve is controversial. Factors that historically have influenced facial nerve resection include a tumor completely encapsulating the nerve, large tumors ( > 3 cm), and undifferentiated tumors [18, 31]. Resection is currently recommended only when there is gross anatomic or histopathologic evidence of neural invasion at the time of the surgery [59]. When there is no direct involvement, sparing the facial nerve does not appear to worsen the prognosis [65]. When resection of the nerve is necessary, immediate reanimation by means of primary anastomosis or free nerve graft is advocated [66]. Timely management in this respect is critical for achieving optimal facial function results.

The role of neck dissection in salivary gland cancer in children is also debatable. The overall incidence of LN metastases at presentation for primary parotid carcinomas in all age groups ranges between 18 and 28 % [67, 68]. In the adult literature, recommendations range from routine elective neck dissection in all patients with primary carcinoma of the parotid gland [67] to neck dissection under specific circumstances including patients with tumors larger than 3 cm, high-grade tumors, facial paralysis, extra-glandular extension, and perilymphatic invasion [69]. Others argue in support of a supra-omohyoid neck dissection in patients with high-grade tumors for staging purposes, reserving an elective full neck dissection for patients with undifferentiated and squamous cell carcinomas [70]. In children, simultaneous neck dissection at the time of primary tumor excision is typically recommended only when cervical nodal metastases are clinically detected [15, 20, 23, 26]. Some advocate concurrent neck dissection when there is clinical evidence of high TNM stage or known high histological grade [14]. The clinical picture in children is further complicated by the fact that the presence of lymph node hyperplasia is common, particularly in younger children [59]. Under such circumstances, nodal biopsy is indicated, and a formal modified neck dissection undertaken if there is intraoperative frozen section confirmation of nodal metastases [26, 30, 69].

Permanent facial nerve paresis or paralysis after surgery for benign parotid tumors in adults ranges from 3 to 5 % [71, 72], while transient facial nerve dysfunction ranges from 46 to 65 % [71, 72]. These figures are notably higher when tumor enucleation rather than formal parotidectomy is undertaken [73]. In children, rates of facial nerve palsy following parotidectomy range from 5 to 33 % [11, 15, 18, 21, 30]. This risk of facial weakness may be greater in children compared with adults, with one study reporting 10 of 21 cases had one or more branches of the facial nerve sectioned intraoperatively [26]. This risk is also higher in infants than older children [21]. Preoperative planning for facial nerve grafting should be considered if facial nerve resection is required [14]. As would be expected, complete parotidectomy has higher rates of transient facial nerve dysfunction and secondary Frey syndrome than superficial parotidectomy [74].

Freyʼs Syndrome, also known as gustatory sweating, is characterized by hyperhidrosis and flushing of the lateral cervicofacial skin following parotid surgery. These symptoms are due to abnormal skin innervation by auriculotemporal nerve parasympathetic fibers that communicate with the sympathetic nervous system. Subjective symptoms may be mentioned by 5–50 % of patients, but formal evaluation via the starch iodine test can reveal the presence of this syndrome in the majority of postparotidectomy patients [75]. In children, reports of Frey’s syndrome ranges from 0 to 47 % [15, 18, 21, 22, 76]. Children are typically managed expectantly [76]. In adults botulinum toxin type A injection is the treatment of choice [75].