Molecular diagnosis of lung tumors with a history of pancreatic cancer: Spatial profiling-based differential markers and functional insights into invasive mucinous adenocarcinoma

Background: Differentiating pancreatic cancer lung metastasis (PCLM) from primary lung cancer, particularly invasive mucinous adenocarcinoma (IMA), sometimes be difficult due to their pathological similarities such as mucin production. This study aimed to identify diagnostic markers for distinguishing IMA from PCLM and to elucidate the molecular pathogenesis of mucinous lung adenocarcinoma using spatial multi-omics analyses. Methods: We performed spatially resolved transcriptomics by GeoMx digital spatial profiler on surgically resected lung tumors, from nine patients with a history of pancreatic cancer and three patients of IMA without such history. Multiple regions of interest (ROIs) were selected in tumor components from each sample. Based on GeoMx results, validation was performed by immunohistochemistry (IHC) on tissue specimens. Subsequently, functional analysis was conducted using the lung adenocarcinoma cell line. Results: By spatial transcriptomics profiling, positive UGT2B15 and negative S100A4 were suggested as candidate markers for distinguishing IMA from PCLM. The expression status of these proteins were confirmed by IHC on tissue specimens. Transcription factor analyses suggested that HNF1A was involved in UGT2B15 expression in the IMA, and functional analyses using lung cancer cell lines also demonstrated these associations. Conclusion: We identified UGT2B15 and S100A4 as potential candidate markers for distinguishing IMA from PCLM. The involvement of the transcription factor HNF1A in the expression of UGT2B15 in the IMA was suggested.