Introduction
Mucoepidermoid carcinoma (MEC) is the most common salivary gland malignancy, characterized by a mixed histological pattern of epidermoid, mucous-producing and intermediate cells.Reference Stewart, Foote and Becker1–Reference White, Hyrcza, Lennerz, Thierauf, Lokuhetty and Cree3 While often found in the parotid glands, primary tumours can arise in any major or minor salivary glands.Reference Brenner, Gillis, Demers, Ellison, Finley and Fitzgerald4, Reference Young and Okuyemi5 Prognosis in MEC is variable, with studies showing five-year survival rates between 87 and 98 per cent and 22 and 67 per cent in low-grade and high-grade tumours, respectively.Reference Boahene, Olsen, Lewis, Pinheiro, Pankratz and Bagniewski6–Reference Sama, Komiya and Guddati9
Cancers of unknown primary (CUP) present as metastatic lesions without any known primary site despite investigative work-up.Reference Kolling, Ventre, Geuna, Milan, Pisacane and Boccaccio10 CUPs represent 3–5 per cent of all malignant epithelial tumours, carry a worse prognosis than their counterparts with known primary and can be aggressive with a poor response to empiric treatment.Reference Pavlidis and Pentheroudakis11–Reference Kim, Hannouf, Sarma, Rodrigues, Rogan and Mahmud14 They are estimated to comprise around 5 per cent of head and neck cancers, with five-year overall survival rates reported around 30–40 per cent.Reference Grau, Johansen, Jakobsen, Geertsen, Andersen and Jensen15 Current literature detailing management of head and neck CUP is mostly in the context of squamous cell carcinoma, with little information on the management of MEC-UP available.Reference Ota and Kitahara16, Reference Mackenzie, Watson, Jankowska, Bhide and Simo17 Hence, MEC-UP presents a unique clinical challenge, with only a handful of case reports detailing management and outcomes.
To guide clinical decision making in this rare presentation of head and neck cancer, we embarked on the first systematic review of the MEC-UP literature for additional cases to present alongside our two (Appendix 1). This resulted in nine total cases of MEC-UP, the largest number of MEC-UP cases presented in a single article to date.
Methods
We performed a database search of Medline (Ovid) and Embase from database inception to April 2024 (Appendix 1). This revealed 1560 unique citations that underwent screening and full-text review in duplicate (K.C. and P.S.). We identified five case reports that investigated MEC-UP of the head and neck (Table 1).
Table 1. Summary of published cases of MEC-UP in the head and neck

CRT = chemoradiotherapy; EBV = Epstein–Barr virus; ENE = extranodal extensions; FNAB = fine needle aspiration biopsy; GERD = gastroesophageal reflux disease; HSV = herpes simplex virus; LN = lymph node; LVI = lymphovascular invasion; MEC = mucoepidermoid carcinoma; MEC-UP = mucoepidermoid carcinoma of unknown primary; MRI = magnetic resonance imaging; MRND = modified radical neck dissection; PE = panendoscopy; PET = positron emission tomography; PY = pack-year; RND = radical neck dissection; RT = radiotherapy; SND = selective neck dissection; US = ultrasound.
Case report
Case one
A 41-year-old female presented with an asymptomatic left-sided lateral neck mass and an otherwise unremarkable physical exam. She is a never-smoker with a history of Lyme disease, Epstein–Barr virus and herpes simplex virus. Her fine needle aspiration biopsy (FNAB) demonstrated cellular atypia but was negative for carcinoma and so she underwent an excisional biopsy that demonstrated low-grade MEC. She underwent panendoscopy with base of tongue biopsies and ipsilateral diagnostic tonsillectomy that was negative for any malignancy. She was staged as TXN1M0. Her imaging post-excisional biopsy did not demonstrate any additional lesions of concern. Her case was presented at multidisciplinary tumour boards who recommended ongoing oncological surveillance with consideration for neck dissection if any changes were to arise. The patient decided against an ipsilateral neck dissection and instead chose regular surveillance. She has had subsequent ultrasound, MRI and PET-CT that did not demonstrate any pathological features consistent with a primary site. Presently, she has not developed any recurrent disease or lesions consistent with a primary tumour within 18 months of her initial presentation.
Case two
A 65-year-old female presented with an asymptomatic left-sided neck mass. Three years prior to presentation, she underwent a FNAB of this neck mass outside of Canada that suggested Warthin’s tumour. She is a never-smoker, and her past medical history included hypothyroidism, hypertension and gastroesophageal reflux disease (GERD). Due to continued growth, the mass was investigated by repeat FNAB, which showed atypical cells. CT neck with contrast showed a heterogeneous high-density mass in the left neck (Figure 1). A subsequent core needle biopsy indicated low-grade MEC, and an MRI neck with gadolinium enhancement showed an intermediate-enhancing lesion in the left upper cervical area and a small mixed solid/cystic lymph node (Figure 2). Panendoscopy revealed no obvious mucosal lesions and left tonsillectomy and direct biopsy of the left, central and right tongue base were negative for malignancy. She was staged as TXN1M0.

Figure 1. CT neck with contrast in coronal (A) and transverse (B) demonstrating a 2.2 x 2.3 x 3.2 cm heterogeneous mass between left submandibular gland and left SCM.

Figure 2. MRI neck with gadolinium contrast in coronal (A) and axial (B) demonstrating the same left sided level IIA neck mass measuring alongside a small mixed solid/cystic lymph node.
Primary management was ipsilateral selective neck dissection of levels I–III with partial level IV and resection of peri-facial nodes and the tail of the parotid gland. Surgical pathology showed 1/19 lymph nodes were positive for MEC alongside negative margins and no extranodal extension. The patient was referred for consideration for adjuvant radiotherapy (RT), but due to the lack of high-risk pathological features, the team instead opted for ongoing surveillance. Presently, she does not have any disease recurrence within 7 months of primary treatment.
Discussion
To the best of our knowledge, this report is the largest published series of patients with MEC-UP of the head and neck (Table 1). Across these studies, the median age at presentation was 54 years (range: 43–67 years), and most patients were male, presenting with asymptomatic neck masses. We found that 44 per cent of these patients had a history of smoking, but smoking status remains an unclear risk factor in MEC despite being associated with other salivary gland malignancies.Reference Sawabe, Ito, Takahara, Oze, Kawakita and Yatabe18, Reference Broaddus, Mason, Ernst, King, Lazarus and Murray19 In one case, we found that MEC was diagnosed three years following a FNAB suggestive of Warthin’s tumour. Warthin’s tumour rarely presents outside of the parotid gland, but cases of MEC developing in a Warthin’s tumour have been reported.Reference Yu, Song, Xiao, Lin and Dong20–Reference Williamson, Simmons, el-Naggar and Medeiros22 There are also described cases of Warthin’s tumour-like MEC, which are MEC tumours that possess cellular features similar to Warthin’s tumour.Reference Balasubiramaniyan, Sultania, Sable, Muduly and Kar23 These uncommonly described tumours may represent malignant transformation of Warthin’s tumour, or “tumour-to-tumour” metastasis from a distant primary, which is a phenomenon that has been described in other cancers.Reference Balasubiramaniyan, Sultania, Sable, Muduly and Kar23, Reference Minezaki, Misawa, Tsukayama, Shibuya, Wada and Sano24 Zhang and colleagues suggested that testing for the MAML2 gene via fluorescence in situ hybridization can identify Warthin’s tumour-like MEC and prevent cytological misdiagnosis.Reference Zhang, Liao, Tang, Meyer, van Dyke and Liu25 In summary, head and neck MEC-UP often presents as an asymptomatic neck mass wherein initial diagnosis is often guided by FNAB with consideration for core needle and/or open biopsy if FNAB results are unclear.
Further diagnostic work-up of the primary site in MEC-UP should be modelled after head and neck mucosal CUP. We propose that clinicians consider employing imaging modalities such as CT, MRI and fluorodeoxyglucose (FDG)-PET/CT, in addition to panendoscopy and biopsies of any suspicious mucosal sites. In our review, four studies performed an FDG-PET/CT scan to help localize and stage the primary site following initial work-up with CT and MRI. In oncology, FDG-PET/CT scans are recommended to identify and stage CUPs for any anatomic region.Reference Nissan, Amit, Baron, Zabatani, Urban and Barshack26–Reference Yoo, Henderson and Walker-Dilks29 Moreover, FDG-PET/CT scans are better than CT alone in detecting primary sites, nodal deposits and distant metastases in salivary gland cancer.Reference Nakajima, Patel, Katabi, Flukes, Mauguen and Ganly30, Reference Roh, Hwan, Choi, Seung, Hyun and Cho31 Hence, we recommend that clinicians consider FDG-PET/CT in cases of suspected MEC-UP when CT and MRI do not identify any sites suspicious for a primary cancer. Diagnostic tonsillectomy is often considered for head and neck mucosal CUP.Reference Podeur, Mancini, Delgrande, Santini, Salas and Wong32, Reference Randall, Johnstone, Foss and Martin33 Although rare, authors have reported primary MEC affecting minor salivary glands within subsites of the oropharynx, including the palatine tonsils.Reference Talani, Frånlund and Unguras34–Reference Ramael, Van Steelandt, Puls and Ramael37 Hence, clinicians should consider panendoscopy, targeted biopsies and diagnostic ipsilateral tonsillectomy for MEC-UP of the head and neck, especially if any suspicious areas are noted on diagnostic work-up.
Primary salivary gland malignancies are typically treated via primary surgery addressing the primary site and any nodal metastases with indications for adjuvant treatment reserved for high-risk pathological features.Reference Sama, Komiya and Guddati9, Reference Chen, Lau, Farwell, Luu and Donald38 When accounting for high-risk pathological features, survival in salivary gland MEC is associated with histological grade, with five-year survival ranging from 26 per cent to 95 per cent in high and low histological grade disease, respectively.Reference Pires, de Almeida OP, de Araújo and Kowalski39, Reference Peraza, Gómez, Beltran and Amarista40 In MEC-UP, we found neck dissection to extirpate any gross nodal disease was most often performed as primary treatment.Reference Gogna, Kashyap and Gupta41 One study reported primary chemoradiotherapy (CRT) followed by salvage neck dissection for a patient presenting with distant metastases who did succumb to his advanced disease within two months of primary treatment.Reference Ghazali, Flashburg and Ord42 The use of adjuvant RT in MEC varies based on pathological features and grade of disease, with one case series of MEC reported an overall rate of RT of 41 per cent, where most low-grade disease was treated with surgery alone and most high-grade disease was treated with surgery and adjuvant RT.Reference Taylor, Kaya, Bunn, Guss, Mitchell and Fairbanks43 This is supported by another case series, which reported the rate of adjuvant RT in salivary gland MEC as 7 per cent in low-grade disease and 70 per cent in high-grade disease.Reference Nance, Seethala, Wang, Chiosea, Myers and Johnson44 Studies suggest improved locoregional control with adjuvant RT in MEC patients with high histological grade, positive surgical margins, peri-neural invasion, and/or in advanced stage disease, but this remains controversial.Reference Sama, Komiya and Guddati9, Reference Chen, Lau, Farwell, Luu and Donald38, Reference Taylor, Kaya, Bunn, Guss, Mitchell and Fairbanks43, Reference Olsen, Mitchell and Miles45 Most MEC-UP patients in this series received adjuvant CRT, which was typically indicated due to advanced nodal disease.Reference Ghazali, Flashburg and Ord42, Reference Friedrich and Zustin46–Reference Yulian, Hwei, Tambun, Siswoyo, Ham and Suroyo50 While no difference in overall survival has been shown in MEC patients receiving adjuvant RT versus CRT, studies suggest that adjuvant CRT confers a greater locoregional control, especially in those with high-risk pathological features.Reference Sama, Komiya and Guddati9, Reference Peraza, Gómez, Beltran and Amarista40, Reference Onderdonk, Vokes, Gwede, Blair, Agrawal and Haraf51 In cases of MEC-UP, survival may be worsened due to the lack of treatment to the primary site, so, clinicians should have a low threshold for recommending adjuvant RT to affected nodal sites and sites at risk for harbouring the MEC primary site.Reference Pinkiewicz, Dorobisz and Zatoński52
• MEC-UP is a head and neck cancer that is poorly described in the literature
• We detail the largest number of cases and management of MEC-UP through the introduction of two cases and a review of the literature
• Work-up of MEC-UP tends to include PET-CT, panendoscopy, targeted biopsies and consideration for diagnostic tonsillectomy
• Primary management of MEC-UP is complete surgical excision
• There is a lower threshold for consideration of adjuvant RT or CRT than in salivary gland malignancies with known primary
• MEC-UP outcomes may be worse due to incomplete identification and removal of a primary site, if one exists
Conclusion
Herein, we reported the largest review of MEC-UP of the head and neck, which is a rare but challenging surgical diagnosis. From the cases reported, clinicians perform PET-CT in addition to panendoscopy and targeted biopsies with consideration of diagnostic tonsillectomy. As with other salivary gland cancers, primary surgical resection of gross disease is the most common treatment with low threshold for considering adjuvant RT to regions at risk for harbouring the primary malignancy, especially in cases of high-grade histopathology.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022215124002147.
Acknowledgements
None.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
Competing interests
The authors declare no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethical approval
Our institution does not require approval from an ethics board for the reporting of case series or individual cases.
Statements of informed consent
Written consent was obtained from the patients for their anonymised information and photos to be used for educational purposes.