Keratocystoma is a salivary gland tumor which was described by Nagao, et al. in a two-case report in 2002 [1], likely the same neoplasm previously described in 1999 as a choristoma by Seifert [2]. The diagnosis remained ambiguous for two decades, as it was a tumor defined purely by morphology which resembled other benign and malignant tumors. In 2024, however, it was found that most keratocystomas harbor IRF2BP2::RUNX2 fusions [3]. With this discovery, pathologists now have a molecular gold standard, allowing them to make this rare benign diagnosis more comfortably [4]. Moreover, awareness of this fusion should allow pathologists to fully realize the morphologic boundaries of keratocystoma, as has been done with other fusion-driven salivary gland tumors [5].

The genetic story of keratocystoma may, however, not be so simple. For example, one case in the 2024 series was positive for RUNX2 rearrangement by break apart fluorescence hybridization (FISH), but negative for IRF2BP2::RUNX2 by reverse transcriptase polymerase chain reaction, suggesting an alternate RUNX2 fusion partner. Moreover, one case in that paper diagnosed as keratocystoma was altogether negative for RUNX2 rearrangements by FISH. In this study, we describe 3 cystic squamous parotid gland tumors harboring RUNX1 or RUNX2 fusions heretofore undescribed in salivary gland neoplasia.