Salivary gland location.
Pediatric salivary mass. MRI indicates magnetic resonance imaging; CT, computed tomographic.
Bentz BG, Hughes CA, Lüdemann JP, Maddalozzo J. Masses of the Salivary Gland Region in Children. Arch Otolaryngol Head Neck Surg. 2000;126(12):1435-1439. doi:10.1001/archotol.126.12.1435
Noninflammatory masses of the salivary gland region in children are extremely rare. Therefore, very few published individual and institution-based experiences exist.
Retrospective chart review from 1990 through 1997.
University-based children's hospital.
Patients 18 years of age or younger with a tumor in the salivary gland region. Masses of infectious origin were excluded. Hemangiomas and lymphangiomas were tallied for relative incidences only.
Three hundred twenty-four consecutive cases of salivary gland masses were found: 192 hemangiomas (59.2%), 89 lymphangiomas (27.5%), and 43 (13.3%) solid masses. No significant difference was found between the age at presentation of the patients with benign solid tumors and the patients with malignant solid tumors (mean + SEM age, 7.2 + 0.7 years). Sixty-one percent of the masses were found in the parotid region; 18% were localized to the submandibular gland region; and the remaining 21% were located in a minor salivary gland site. The most common benign perisalivary masses were pilomatrixomas (20.9%), followed by pleomorphic adenomas (11.6%). The most common malignant masses were mucoepidermoid carcinomas (9.3%), followed by rhabdomyosarcomas (7.0%). Treatment was individualized to the disease. Twenty-two patients had adequate data for follow-up analysis (mean + SEM follow-up, 30.0 + 8.4 months). Four patients (18.2%) experienced recurrent or residual disease and were alive with disease at last follow-up, and 100% of our population demonstrated disease-specific survival at last follow-up.
Vascular lesions outnumber solid tumors of the salivary gland region. The most common salivary tumors were pleomorphic adenomas, followed by mucoepidermoid carcinomas. Although certain solid salivary masses may demonstrate locally aggressive behavior, the overall prognosis is favorable.
SALIVARY GLAND tumors are a comparatively rare disorder, accounting for only 1% of all neoplasms of the head and neck region.1 Further complicating the analysis of these tumors in the pediatric population is that fewer than 5% of all salivary tumors occur in patients younger than 16 years.2 Not all masses arising in a salivary gland region represent tumors of glandular origin. Also, the wide variation in the claimed incidence of malignancy3- 5 found in pediatric salivary gland tumors prompts physicians who encounter these rare neoplasms to continue to critically review their true nature. Furthermore, the low incidence of these tumors makes individual and even institution-based experiences with the evaluation and treatment of masses of the perisalivary gland region limited. Therefore, the aim of this investigation was to explore our experience with pediatric salivary gland masses in a tertiary care children's hospital in a major metropolitan area. We hope that our findings will add to the cumulative knowledge about these tumors. Finally, we propose an algorithm with which to evaluate and treat masses of the salivary gland region in the pediatric patient.
Medical records, radiographic and pathologic results, surgical records, and outcome data for all patients undergoing diagnostic and/or therapeutic procedures for masses of the salivary gland over a 7-year period (1990-1997) at Children's Memorial Hospital, Chicago, Ill, were retrospectively reviewed. Abscesses and masses of infectious origin were excluded from analysis. A total of 324 cases fit the general descriptors of salivary gland, parotid gland, submandibular gland, or minor salivary gland neoplasm. The total numbers of patients who were found to have lesions consistent with a hemangioma or lymphangioma were tallied for relative incidence, but further clinical data were not recorded. The remaining patients were determined to have a solid tumor of the intrasalivary or extrasalivary gland region. Data were recorded on patient age and sex; location of the mass; occurrence of facial palsy, pain, or skin involvement; size of the lesion; form of treatment; histopathologic diagnosis; stage and grade of tumor; incidence and location of recurrence; and disease-free and total survival. The results are reported as mean + SEM unless otherwise specified.
Our survey found a total of 324 cases of solid or cystic tumors of the perisalivary gland tissues. Two hundred eighty-one (86.7%) of the 324 tumors were vascular proliferations, of which 59.2% were hemangiomas and 27.5% were lymphangiomas.
The solid salivary tumor group consisted of 43 patients (13.3% of all tumors), with a mean age at presentation of 7.2 + 0.7 years (age range, <1-17 years). An equal number of male and female children (22 each, 50%) had solid perisalivary tumors. A majority of these solid tumors were located in a major salivary gland site (79.5%); 20.5% were located in a minor salivary gland region. An overwhelming majority of these cases were located in the parotid region (61%). Dramatically fewer (18%) were located in the submandibular gland region, while 7% were found in the palate, 5% each in the cheek and tongue, and 2% each in the lip and gingiva (Figure 1).
All patients presented with a mass in the area of a salivary gland. In the vast majority of patients (88.6%), this was the sole presenting symptom. Three patients presented with a facial deformity: 1 with a diagnosis of sinus histiocytosis and 2 with neurofibromas. One patient presented with facial nerve paralysis and was diagnosed as having aggressive fibromatosis. No patients presented with facial pain (a known poor clinical prognostic indicator in adult patients with salivary neoplasms). Pilomatrixomas by nature demonstrated skin involvement, but malignant involvement was not found. One patient had concomitant maxillary sinusitis due to the mass effect against the sinus ostia.
The pathologic distribution of solid masses of a glandular region in our pediatric population is seen in Table 1. As in the adult population, the majority of tumors were found to be benign lesions. Hemangiomas outnumbered the rest of the group (59.0% vs 41.0%). Our assessment of the incidence of these lesions probably underestimated the true incidence, since there is a strong bias at our institution to observe these lesions unless they become problematic. The next most common benign lesion found in our series was lymphangioma (27.5%). A full evaluation of lymphangiomas is beyond the scope of this study.
True solid tumors of a salivary gland site represented 13.3% of all tumors found. No significant difference was found in the mean + SEM age of patients with benign (8.9 + 1.1 years) vs malignant (6.9 + 0.9 years) perisalivary tumors; 76.7% of solid tumors were found to be benign. Of all solid masses in and around a salivary gland, pilomatrixomas were found to be the most common benign lesions (20.9%). The only benign lesion of salivary gland tissue origin was pleomorphic adenoma, which composed 11.6% of all solid tumors and 15.2% of all benign lesions. Reparative granulomas were the next most frequently encountered solid tumors, representing 9.3% of all tumors. One granuloma each was localized to the parotid duct; the parenchyma of the parotid gland, with a small focus within a periparotid lymph node, overlying the mandibular ramus; and finally, the hard palate. Reactive lymph nodes and granular cell tumors accounted for 7.0% of the solid tumors, with 2 granular cell tumors located on the tongue and 1 on the posterior oropharynx. The remainder of the benign tumors were sinus histiocytosis neurofibromas (n = 2) and 1 case each of schwannoma, congenital myofibromatosis, sialoblastoma, benign simple cyst, and benign lymph node hyperplasia.
Treatment of these benign lesions was individualized to the pathologic diagnosis. Four patients with pleomorphic adenomas underwent a superficial parotidectomy, and 1 underwent a submandibular gland excision. One patient with sinus histiocytosis underwent a superficial parotidectomy, and 1 underwent a submandibular gland excision. The 2 patients with neurofibromas underwent debulking procedures, with both requiring reexcision for residual disease. The patient with a sialoblastoma had 2 local recurrences and underwent a wide local excision, followed by a superficial parotidectomy with facial nerve grafting, and finally, a total parotidectomy. The particulars of this case have been described elsewhere.6 All other patients with benign lesions underwent either a submandibular gland excision or a wide local excision.
Ten malignant tumors were encountered in the series and accounted for 23.3% of all solid tumors of a salivary gland region. Low-grade mucoepidermoid carcinomas were the most common malignant tumors encountered (9.3% of all solid tumors [n = 4]): 3 presented as stage I disease and 1 as stage II disease. None of the 4 patients was found to have neck metastases on clinical or radiologic evaluation.
We found that the next most frequent malignant tumor was rhabdomyosarcoma (n = 3 [7.0%]). Subclassification of these 3 sarcomas revealed 1 case each of alveolar, embryonal, and undifferentiated subtypes. Also, we found 1 case each of aggressive fibromatosis, anaplastic large cell lymphoma, and liposarcoma.
Surgical therapy for these malignant lesions was again tailored to the disease. A superficial parotidectomy was performed for the anaplastic large cell lymphoma and for 2 of the mucoepidermoid carcinomas. Two smaller mucoepidermoid carcinomas required a wide local excision. One case required a second wide local excision plus a supraomohyoid neck dissection. Two of the 3 rhabdomyosarcomas required a submandibular gland excision with a supraomohyoid neck dissection, and 1 of the 2 also required a partial mandibulectomy. The patient with the other rhabdomyosarcoma underwent an incisional biopsy prior to radiation therapy and chemotherapy. The patient with mucoepidermoid carcinoma who underwent reexcision and neck dissection was unavailable for follow-up. The remaining 3 patients with mucoepidermoid carcinomas were alive without evidence of disease with a mean follow-up period of 29.5 + 8.7 months.
The patient with liposarcoma underwent a total parotidectomy without receiving postoperative chemotherapy or radiation therapy and remained free of disease at the last follow-up visit 18 months after resection. The patient with aggressive fibromatosis was treated with a composite resection and a functional neck dissection. She received postoperative chemotherapy and radiation therapy but experienced a local recurrence that required a second round of chemotherapy and radiation therapy. She remained alive with disease at 61 months.
The mean length of follow-up for the 22 patients for whom follow-up data were available was 30.0 + 8.4 months. The 16 patients with benign disease and follow-up information had a mean follow-up period of 29.1 + 10.7 months, whereas the 6 patients with malignant disease had a mean follow-up period of 31.8 + 13.4 months. Two patients had residual disease at the time of their primary surgical intervention. The 1 patient with neurofibroma with positive margins subsequently underwent reexcision, whereas the patient with alveolar rhabdomyosarcoma underwent a subsequent course of chemotherapy and radiation therapy. At 16 months of follow-up, the patient with rhabdomyosarcoma was without evidence of disease.
Four patients experienced recurrence of their disease at a mean time of 5.0 + 2.0 months (range, 2-12 months). Within this group, there was a patient with neurofibroma who underwent reexcision, a patient with aggressive fibromatosis who underwent chemotherapy and radiation therapy, and a patient with sinus histiocytosis (Rosai-Dorfman syndrome) who had regional recurrence and underwent a superficial parotidectomy with a supraomohyoid neck dissection; these 3 patients were alive with disease at 38, 61, and 156 months. The last case of recurrence occurred in the patient with a sialoblastoma, who had recurrences 6, 12, and 58 months after the initial recurrence. This patient had no evidence of disease 98 months after the first diagnosis. The patient with an anaplastic large cell lymphoma who demonstrated regional involvement at the time of surgical intervention for the salivary gland and underwent subsequent chemotherapy was without evidence of disease at 2 months.
Salivary gland neoplasms represent rare lesions, accounting for fewer than 1% of lesions of the head and neck.1 Furthermore, tumors of the major salivary glands have a wide variety of histopathologic diagnoses and biologic behavior, mandating a broad range of treatment strategies. Clinical examination may not easily distinguish salivary tissue from extrasalivary tissue origin. Therefore, a diagnostic and therapeutic guide for a mass in the salivary region may assist the clinician in accurate diagnosis and treatment in such cases.
We undertook this retrospective review to compare and contrast the experience of a single university-based children's hospital with the previously published reports. Our review found that an overwhelming majority of cases of perisalivary tumors are of vascular origin, with hemangiomas outnumbering lymphangiomas. This finding does not agree with several published series. The largest series reported was from the Armed Forces Institute of Pathology (AFIP).7 The most common neoplasms found in the AFIP series were benign mixed tumors, followed by hemangiomas. The second largest series, which was also reported from the AFIP,8 likewise found that benign mixed tumors represented the largest percentage of cases, followed by vascular proliferations and mucoepidermoid carcinomas. These reported histologic reviews probably do not reflect the true incidence of these vascular proliferative lesions, since a large percentage of these lesions, most often hemangiomas, would not be studied histologically. Surgical intervention is usually reserved for those lesions that demonstrate cosmetic deformity or a functional compromise. Furthermore, the AFIP is a central pathologic referral institution that tends to study only difficult histologic cases.
A single institution report from the Mayo Clinic9 reviewed a 52-year experience of 74 pediatric salivary gland tumors. The most common lesions of the parotid region were benign mixed tumors, accounting for 29.7% of the cases. The second most common lesions in the Mayo Clinic series were hemangiomas, which represented 21.6% of the series. Again, this figure may underestimate the true incidence of hemangioma cases, since surgery was performed on only those lesions demonstrating an increase in tumor size, rapid tumor growth and disfigurement, failure to involute, or intralesional hemorrhage. Furthermore, the Mayo Clinic approach to hemangiomas was to delay surgical intervention for at least 1 year. This delay may cause patients to seek definitive surgical care in an institution that has a higher likelihood of resection. Within our institution, a wide variety of approaches to vascular lesions are adopted by the services that manage these cases. Since a large number of vascular lesions are managed by the plastic surgery service, cosmetic deformity may play a more prominent role in the decision to resect.
Of the 43 patients with solid salivary region tumors in our series, 33 (76.7%) presented with benign disease, and 10 (23.3%) presented with malignant disease. The most common benign solid tumor of the salivary region was the pilomatrixoma (20.9%). The most common salivary gland tissue tumor was the benign mixed or pleomorphic adenoma (11.6%). We are not aware of any other published series including the pilomatrixoma in the differential diagnosis of a mass in the salivary gland region. Both reports from the AFIP7,8 and the review from the Mayo Clinic9 found that the most common tumor of the pediatric salivary gland region was the pleomorphic adenoma. Our decision to include the pilomatrixoma in our analysis was based on the fact that before radiographic analysis, clinical differentiation of an intraglandular mass and an extraglandular mass can be quite difficult. Our data are in agreement with others that the next most common tumor is the pleomorphic adenoma, followed by a variety of much rarer tumors.
Our series reports a 23% incidence of malignancy in children with solid tumors of the glandular region. This incidence of malignancy falls between the extremes in the literature. Furthermore, we found a wider variety of malignant neoplasms. A series of 22 patients from the Massachusetts Eye and Ear Infirmary5 disagreed with the observation of Schuller and McCabe3,4 that almost 60% of all pediatric parotid masses represent malignancies. In that series, there was only 1 child with a mucoepidermoid carcinoma, while 8 children had pleomorphic adenomas and 13 children demonstrated inflammatory processes.
Our data also were in general agreement with the published literature that the most common malignant salivary lesion in children is the mucoepidermoid carcinoma. The next most common malignancy found in our pediatric population was the rhabdomyosarcoma. These findings are in direct disagreement with those in almost all other published reports. The AFIP reports,7 as well as the Mayo Clinic review,9 found the next most common malignancy to be the acinic cell carcinoma. We found no examples of this tumor. The only review in which the rhabdomyosarcoma was included in the analysis was a report from the St Jude Medical Center,10 which reviewed 17 cases involving malignancies of the major salivary glands and found that 8 were carcinomas, of which 6 were mucoepidermoid carcinomas, 1 was a malignant mixed tumor, and 1 was an acinic cell carcinoma. The investigators included rhabdomyosarcomas in their series and found that the remaining 9 cases consisted of 6 embryonal rhabdomyosarcomas and 1 each of alveolar, poorly differentiated, and poorly differentiated embryonal rhabdomyosarcomas. They recommended that all carcinomas be surgically removed, while incisional biopsies, followed by definitive chemotherapy and radiation therapy, should be performed on rhabdomyosarcomas.
Our outcome data demonstrated a high likelihood of long-term, disease-free survival in these patients. A report from the Head and Neck Service and the Department of Pathology of the Memorial Sloan-Kettering Cancer Center documented 38 pediatric tumors, which represented 1.7% of a total 2135 patients with tumors of the major salivary glands.11 Their series excluded cysts, lymphangiomas, hemangiomas, granulomas, lipomas, inflammatory lesions, sarcomas, and metastatic lesions. Interestingly, the overall 5- and 10-year determinate survival rates for malignant lesions were reported to be 94% and 93%, respectively. Our data are in agreement with these results; our overall determinate survival rate, with a mean follow-up period of 30.0 + 8.4 months, was 100%.
Based on the collective experience of these series, we propose a diagnostic and treatment strategy (Figure 2). When a clinician encounters a child with a mass in the region of a salivary gland, the first step is to determine whether the mass is inflammatory or noninflammatory. Inflammatory lesions should be treated with a trial of antibiotics. Masses without signs of inflammation are palpated to determine if the mass is cystic or solid. Lesions in which this may be difficult to determine by physical examination may undergo fine-needle aspiration. Cystic lesions include hemangiomas and lymphangiomas.
If the lesion has been determined to be of a solid character, magnetic resonance imaging, computed tomography, and/or physical examination may help determine if the mass is intrasalivary or extrasalivary. If there is still difficulty in distinguishing between the 2 types, fine-needle aspiration can be used again to further evaluate the mass.
Extrasalivary masses, such as pilomatrixomas, may be treated with surgical excision or observed. Intrasalivary solid masses should undergo surgical excision for pathologic determination. A frozen section should be sent intraoperatively for histopathologic diagnosis, and if the lesion represents a benign process, then a wide excision or a formal superficial gland excision is adequate. If the frozen-section diagnosis is malignancy, then the determination of the malignancy's pathologic nature is necessary. A carcinoma is treated with surgical excision, with or without neck dissection, and, rarely, adjuvant chemotherapy or radiation therapy. The facial nerve is sacrificed only if it is obviously involved either clinically or intraoperatively. If the diagnosis is a rhabdomyosarcoma or lymphoma, then radiation therapy, chemotherapy, or both would be the next step, as dictated by the final pathology report.
Tumors of a salivary gland region are extremely rare in the pediatric population. We found that 76.7% of solid perisalivary masses in the pediatric patient are benign and that 23.3% are malignant. The most common benign lesions are pilomatrixomas, followed by pleomorphic adenomas. The most common malignant lesions are mucoepidermoid carcinomas, followed by rhabdomyosarcomas. Based on our experience, we have proposed a management algorithm that can be used to facilitate diagnosis and treatment with the hope that outcomes may be improved.
Accepted for publication June 28, 2000.
Presented in part at the 14th Annual Meeting of the American Society of Pediatric Otolaryngology and the Third Biannual Meeting of the Interamerican Association of Pediatric Otolaryngology, Palm Desert, Calif, April 28, 1999.
Corresponding author: John Maddalozzo, MD, Division of Otolaryngology–Head and Neck Surgery, Department of Surgery, Children's Memorial Hospital, 9300 Children's Plaza, Box 95, Chicago, IL 60614.