The morphology of the current case was reminiscent of pancreatic IPMN with associated colloid carcinoma. Meanwhile, a GNAS R201C mutation, which is a characteristic molecular alteration in pancreatic IPMNs and colloid carcinomas arising from IPMNs [7], was detected in our case. However, the clinical history and the subsequent positron emission tomography result excluded the possibility of a metastatic tumor from the pancreas.

Salivary gland IPMN was an emerging tumor entity, characterized by a papillary-cystic proliferation of mucin-producing cells, reminiscent of pancreatic IPMN. NKX3.1 expression was observed in the neoplastic cells of salivary gland IPMN, as well as in normal mucinous acinar cells [8]. Thus, it is suspected that this tumor may be derived from the mucinous acinar cells of the minor salivary gland. The lack of P63 and P40 positive myoepithelial/basal cells in most cases suggests that the tumor might not be intraductal and has malignant potential. The classification of salivary IPMN remains controversial. In our case, the tumor was adjacent to the proliferating submucosal salivary gland of the bronchus and showed NKX3.1 expression. NKX3.1 is an immunohistochemical marker for mucinous acinar cells of branchial salivary glands but is not expressed in pulmonary adenocarcinomas [9]. Thus, we believe that the tumor may originate from the mucinous bronchial salivary gland. The negative expression of CK5/6, P63, and P40 suggests that the tumor might not be an intraductal lesion, but rather an invasive mucinous adenocarcinoma. However, the AKT1 p.E17K variant, which is frequently observed in salivary gland IPMN, was not detected in our case. PTEN and HRAS gene alterations, previously reported in a small number of salivary gland IPMN cases, were not detected in our case [2, 3, 8].

Bronchiolar adenoma, mucous gland adenoma (MGA), mucinous cystadenoma, glandular papilloma, sialadenoma papilliferum, mucoepidermoid carcinoma, metastatic IPMN, and pure colloid adenocarcinoma are on the list of differential diagnoses. Based on anatomical location, morphology, immunohistochemistry, and genetic analysis, the above tumors can be excluded. Among these, MGA requires special attention. MGAs are rare benign tumors with varied histology, suggested to be a salivary gland-type tumor due to their preferred location in the proximal airways with submucosal glands, histological similarity to bronchial glands, and NKX3.1 immunoreactivity [10]. The tumor typically lacks cellular atypia, mitotic activity, necrosis and invasion. In contrast to MGAs, our case exhibited obvious cellular atypia, invasion, and a high Ki-67 index, indicating a malignant tumor. However, given the similar location, overlapping histological features, and identical NKX3.1 expression to MGA, we cannot exclude the possibility that our case may represent a malignant counterpart of MGA. The reported cases of these tumors are limited, and further research is needed to clarify their relationship.

In addition, a recently reported ETV6::NTRK3 translocation-associated low-grade mucinous bronchial adenocarcinoma should be noted [11]. The tumor cells were located in the peri-bronchial tissue and had some overlapping morphology with our case (cysts, papillary, micropapillary, and cribriform structures). However, the presence of more atypical tumor cells, a higher Ki67 index (40% vs. < 1%), completely negative P63 and Pan-TRK staining, and the absence of NTRK3 fusions in our case indicate different features compared to that case. Furthermore, Miura et al. [12]. reported a case of primary tracheal adenocarcinoma with NKX3.1 expression. The tumor exhibited somewhat similar histological morphology to our case (dilated cysts, papillary structure, and cartilage invasion) and also showed positive NKX3.1 immunoreactivity. However, no genetic mutations or gene fusions were observed in that case. Additionally, in a previous study by Ritterhouse et al. [13], a small number of lung adenocarcinomas with GNAS mutations at codons 201 and 227 were identified (19/2352, 0.8%). Ten of these GNAS-mutated lung carcinomas were invasive mucinous adenocarcinomas, all with GNAS codon 201 mutations and concurrent KRAS or BRAF mutations, and all were located within the lung lobes. In contrast, our case, while also a mucinous adenocarcinoma with a GNAS mutation, presented as a lesion in the bronchus rather than within the lung parenchyma, with no concurrent KRAS or BRAF mutations detected.

In conclusion, we report a case of salivary gland-type mucinous adenocarcinoma harboring GNAS mutation. This case shared overlapped morphological, immunohistochemical, and genetic features with both pancreatic IPMN and salivary IPMN, and may represent a novel lung cancer entity originating from the mucinous bronchial salivary gland. However, a larger cohort of cases is required to elucidate the nature of this tumor further.