Exploration of biomarker candidates in epithelial ovarian cancer with special reference to the prognostic role of LMTK3

Abstract

The growing integration of molecular profiling and data-driven analysis has opened new opportunities for precision oncology. In epithelial ovarian cancer (EOC), however, translating these molecular insights into clinically relevant biomarkers for diagnosis and prognosis remains a major challenge. Addressing this gap requires a comprehensive exploration of histotype-specific molecular signatures that reflect the biological diversity and clinical behavior of EOC. This thesis investigates diagnostic and prognostic biomarker discovery in EOC with a focus on histotype stratification. In Study I, an omics-based approach was applied, performing explorative transcriptomic profiling of advanced-stage EOC. We identified differentially expressed genes defining diagnostic and histotype-specific gene panels that include both coding and non-coding molecular signatures and potential biomarker candidates. Functional enrichment analysis revealed histotype-specific regulatory networks and shared molecular pathways driving advanced-stage disease. Prognostic candidates were further evaluated using up to 15 years of follow-up data covering overall and disease-specific survival. In Studies II and III, we used a classical approach, assessing the expression, cellular distribution, and prognostic impact of Lemur tyrosine kinase 3 (LMTK3) in EOC by immunohistochemistry. LMTK3 is a functionally understudied kinase previously reported to have diagnostic and prognostic potential in several cancers. We demonstrated that high nuclear LMTK3 expression indicates a favorable prognosis in early-stage EOC, while broader analyses confirmed its universal expression across EOC and potential involvement in both tumor development and progression. Prognostic significance appeared confined to early-stage disease, with results suggesting a shift toward unfavorable outcomes in advanced stages, indicating that LMTK3 acts as a spatially regulated, stage-dependent biomarker candidate. In conclusion, this work identifies novel diagnostic and prognostic biomarker candidates for EOC and highlights the biological distinctness of its histotype subgroups. These findings advance our understanding of EOC tumor regulation and progression and emphasize the need for tailored biomarker development and therapeutic strategies across histotypes.