Aims: Secretory carcinoma (SC) is characterized by ETV6 rearrangements, most often ETV6-NTRK3 fusion. Given its histologic overlap with other salivary gland tumors (SGTs), SCs can be difficult to diagnose without genetic confirmation. A recently developed pan‐TRK (tropomyosin receptor kinase) antibody shows promise for identifying tumors with NTRK (neurotrophic tyrosine kinase receptor 3) fusions. The aim of this study was to evaluate the utility of pan‐TRK immunohistochemistry in distinguishing SCs from mimics and selecting patients eligible for TRK inhibitor clinical trials. We examined whole‐tissue sections from 111 SGTs with molecular characterization, including 26 SCs (23 with ETV6-NTRK3 fusion and 3 with ETV6-RET fusion detected by ligation-dependent reverse transcription-polymerase chain reaction, next-generation sequencing and 85 non-SC SGTs (no ETV6-NTRK3 fusion). Immunohistochemistry was performed with a pan‐TRK rabbit monoclonal antibody. When any pan‐TRK staining (nuclear or cytoplasmic with any staining intensity) was considered to indicate positivity, 22 of 23 SCs with ETV6-NTRK3 fusion (95.7%) and 33 of 85 non‐SC (38.8%) salivary neoplasms were positive, mainly basal cell adenoma, pleomorphic adenomas, adenoid cystic carcinomas, and epithelial-myoepithelial carcinomas. All SCs with ETV6-RET fusion were entirely negative. When only nuclear pan-TRK staining with any staining intensity was considered positive, 18 of 23 SCs with ETV6-NTRK3 fusion (78.3%) were positive, 11 among them with diffuse staining (>30% of cells). All non-SCs and SCs with ETV6-RET fusion were entirely negative. In comparison to molecular analysis (ligation-dependent reverse transcription-polymerase chain reaction, next-generation sequencing), nuclear pan-TRK IHC has a sensitivity of 78.3% and a specificity of 100% for diagnosing SCs with ETV6-NTRK3 fusion, 69% and 100% for SCs (all fusions). Pan-TRK is a reasonable screening test for diagnosing SCs among SGTs when taking only nuclear staining into account. Although pan-TRK expression is not entirely sensitive for SCs, nuclear staining is highly specific for SCs with ETV6-NTRK3 fusion. The lack of pan-TRK immunoreactivity in a subset of SCs is suggestive of atypical exons 4 to 14 or exons 5 to 14 ETV6-NTRK3 fusion or non-NTRK alternative fusion partners such as ETV6-RET. Pan-TRK staining can serve as a strong diagnostic marker to distinguish SC from it mimics and to select patients eligible for TRK inhibitor clinical trials.

Background Pseudomyxoma peritonei (PMP) is a complex and partially understood disease defined by mucin deposits in the peritoneal cavity, mostly of appendiceal origin caused by the rupture of a mucocele often containing Low or High grade Appendiceal Mucinous Neoplasm (LAMN/HAMN). Other origins include primitive ovarian mucinous cystadenoma or cystadenocarcinoma almost always with an associated teratoma, but to our knowledge no case of ovarian teratomatous appendiceal-like mucocele with LAMN has been reported as a cause of PMP. Case presentation A 25-year old female with infertility was diagnosed with an isolated left ovarian tumor in a context of PMP. Histological examination revealed an ovarian teratoma containing an appendiceal-like structure with mucocele and LAMN, without any associated lesion of the appendix on full histological analysis. Molecular characterization of the ovarian lesion showed co-KRAS and GNAS mutations, as described in PMP of appendiceal origin, while only KRAS mutations are reported in primitive ovarian mucinous tumor. Conclusions Detection of co-KRAS and GNAS mutations in our case of ovarian teratomatous appendiceal-like mucocele with LAMN shows that when PMP derives from a mucinous ovarian lesion (with histological proof of none-appendiceal involvement), it is probably of a digestive teratomatous origin, emphasizing the need to actively search for tetatomatous signs in a context of ovarian PMP.

No abstract