Answer: Adenoid cystic carcinoma

Histology: This is a discrete stroma-rich lesion in the liver in a patient without prior evidence of malignancy, which certainly suggests the preoperative diagnosis of cholangiocarcinoma. However, the lesional cells are small and blue and quite bland relative to what one typically sees in cholangiocarcinoma. The tumor shows areas of tubular and cribriform growth which strongly suggests adenoid cystic carcinoma. Markers for neuroendocrine neoplasms were negative, as was HEPAR-1. A metastatic adenoid cystic carcinoma was suspected, but no clinical history of adenoid cystic carcinoma was provided. Using FISH, the presence of MyB-NFIB gene fusion was demonstrated, a characteristic finding in adenoid cystic carcinoma. Only then, upon further questioning, the patient revealed a long standing history of a swelling in the area of her submandibular gland. This area of swelling was subsequently excised and found to be a widely infiltrative adenoid cystic carcinoma.

Discussion: A carcinoid tumor should label for neuroendocrine markers, and the cribriform pattern seen in the case would be unusual. Cholangiocarcinoma as indicated above would demonstrate greater cytologic atypia. Hepatocellular carcinoma generally lack desmoplastic stroma, and would label for HEPAR-1.

This case highlights the potential for aggressive behavior by submandibular gland salivary gland carcinomas, and the capacity for adenoid cystic carcinoma to present as a metastasis with an occult primary.