Cystic change complicates the cytological interpretation of head and neck masses in two settings above all: the salivary glands and the cervical lymph nodes. In the salivary glands, non-neoplastic cysts such as the benign lymphoepithelial cyst coexist with neoplasms prone to cystic change, and low-grade mucoepidermoid carcinoma in particular may consist largely of cyst fluid containing scattered bland cells, making it a leading source of false negative salivary cytology. In cervical lymph nodes, metastatic squamous cell carcinoma of the aerodigestive tract frequently presents as a cystic mass whose aspirate of debris, inflammatory cells and squames can closely mimic an infected keratinous cyst or branchial cleft cyst.

The mechanism of cystic change in nodal metastasis is incompletely understood; proposed explanations include true cyst formation lined by malignant cells of salivary duct or transitional keratinocyte origin, and central keratin breakdown — spontaneous or after radiotherapy — producing a pseudocystic appearance. Whatever the mechanism, the practical rule is that, in the absence of trauma, necrosis or cystic change in a lymph node should raise the possibility of metastatic carcinoma or lymphoma. The present article describes the cytomorphology and pitfalls of cystic salivary gland and lymph node lesions in a prospective series of cystic head and neck masses.

This analysis derives from a prospective study of 90 palpable cystic head and neck lesions aspirated in the Department of Pathology, J.J.M. Medical College, Davangere, between June 2018 and May 2020. Lesions were included when fluid was aspirated and cyst macrophages were present; frankly bloody aspirates, vascular malformations, bleeding diathesis and non-cooperative patients were excluded. Aspiration was performed with a 23-gauge needle on a 5 ml syringe under aseptic precautions after informed consent, with re-aspiration of residual masses. Smears from centrifuged sediment were fixed in 95% ethanol and stained with H&E and Papanicolaou stains; air-dried smears were stained with Giemsa. Salivary gland aspirates were categorised according to the Milan System for Reporting Salivary Gland Cytopathology. Surgical specimens, where received, were processed routinely for histopathological correlation.

Salivary gland cysts

Eleven aspirates (12.2% of the series) were from salivary gland lesions: six benign lymphoepithelial cysts, three pleomorphic adenomas with cystic change and two low-grade mucoepidermoid carcinomas (Table 1). Cystic malignant change was therefore present in two of eleven salivary lesions (18%).

Table 1. Cystic salivary gland and lymph node lesions (n = 14)

All six lymphoepithelial cysts occurred in males; four lay in the right submandibular region and two in the left parotid, with durations of 30 to 90 days and swellings of 2–4 cm. Aspirates were blood-tinged to serous. Smears showed degenerated red cells and epithelial cells in four cases, with lymphoid cells, haemosiderin-laden macrophages and neutrophils in two. None of the patients had HIV infection. Histopathology was not available for correlation.

The three cystic pleomorphic adenomas occurred at 19–35 years, all in the right parotid, with durations of one to two months; aspirates (0.5 ml) were blood-tinged. Smears showed scattered and loosely cohesive round-to-oval epithelial cells with eccentric nuclei, fine chromatin and scant cytoplasm, against myxoid stromal fragments; the characteristic metachromatic fibrillar chondromyxoid stroma secured the diagnosis. Histopathological correlation in two cases confirmed pleomorphic adenoma with epithelial, myoepithelial and stromal components.

Both low-grade mucoepidermoid carcinomas occurred in males with parotid swellings of 4–5 cm, of two months’ and one year’s duration. Aspirates were haemorrhagic and smears showed cuboidal and intermediate cells with muciphages in a background of mucinous material and red cells. Histopathology, available in one case, showed cords of mucous and intermediate cells forming occasional glandular spaces without marked atypia or mitoses, confirming low-grade mucoepidermoid carcinoma.

Cystic lymph node lesions

Two cases of metastatic squamous cell carcinoma with cystic change occurred in male chronic smokers aged 45 and 72 years, one along the upper sternocleidomastoid and one below the parotid at the angle of the mandible; both swellings were firm, 3–5 cm, of two to three months’ duration, and yielded blood-tinged aspirates. Smears showed abnormal squamous cells with dark nuclei, coarse chromatin and pale scant cytoplasm amid neutrophils and macrophages; the primary was clinically considered aerodigestive. Histopathology in one case confirmed metastatic keratinising squamous cell carcinoma. One case of non-Hodgkin lymphoma with cystic change was diagnosed in a 25-year-old man with a right suprahyoid swelling of two months, fever and cough; imaging showed mediastinal widening and a left pleural effusion. Smears showed scattered abnormal lymphoid cells with large nuclei and prominent nucleoli among macrophages and neutrophils.

Benign lymphoepithelial cyst, defined by Bernier and Bhaskar as solitary or multiple cysts within lymph nodes associated with salivary gland and attributed to cystic degeneration of salivary inclusions within nodes (or, by other authors, to branchial arch remnants), was the commonest salivary cyst in this series. Although the lesion has acquired importance through its association with HIV infection, none of our patients was HIV-positive, in keeping with other reports from comparable populations. Cystic change in pleomorphic adenoma is uncommon, and our frequency of 3.3% sits between the 0.1% of Sahni et al. and the 4.5% of Shekhar et al. The recognised pitfalls are threefold: a predominantly epithelial pattern invites confusion with monomorphic adenoma and adenoid cystic carcinoma; abundant metaplastic squamous cells with scant mucoid material may be misread as mucoepidermoid carcinoma; and abundant myxoid material with sparse epithelium may suggest a retention cyst. The metachromatic fibrillar chondromyxoid stroma remains the most reliable discriminator. Low-grade mucoepidermoid carcinoma accounted for 2.2% of cystic lesions, comparable with 3.17% in Firat et al., 1.8% in Sahni et al. and 0.5% in Shekhar et al. It is the classic false negative of salivary cytology: cyst contents may contain only bland epithelial cells and macrophages, and large cystic spaces may be lined by a single row of columnar mucin-secreting epithelium. A stringy mucinous aspirate containing mucin-secreting and intermediate cells should therefore always carry low-grade mucoepidermoid carcinoma in the differential, and apparently benign mucinous salivary lesions should be reported with this caveat. Cystic nodal metastasis of squamous cell carcinoma presents the mirror-image pitfall: reactive atypia in an infected keratinous cyst, squamous metaplasia in Warthin tumour and branchial cleft cyst may all mimic malignancy, while genuinely malignant cystic nodes may yield deceptively bland material. Attention to nuclear features — hyperchromasia and irregular nuclear membranes — is decisive. Published experience indicates that FNAC detects metastasis in essentially all solid nodes but only 50–73% of cystic nodes, Sheahan et al. reporting diagnostic aspirates in 73% of cystic cervical metastases; our concordant histological correlation and the observation of Firat et al. that head and neck metastases are frequently squamous and frequently cystic reinforce the need for vigilance, re-aspiration of any residual solid area, and early recourse to biopsy where cytology is negative but suspicion persists.

Cystic salivary gland and lymph node lesions concentrate the diagnostic hazards of head and neck cytology. FNAC nevertheless achieved accurate categorisation in this series, including both low-grade mucoepidermoid carcinomas and both cystic nodal metastases. Three working rules emerge: treat every mucinous cystic salivary aspirate as potential low-grade mucoepidermoid carcinoma until excluded; anchor the diagnosis of pleomorphic adenoma on its chondromyxoid stroma rather than its epithelial pattern; and regard unexplained cystic or necrotic lymphadenopathy as metastasis or lymphoma until proven otherwise.