A Liquid Biopsy-Based Approach Decoding Solid Tumor Models in the Era of Precision Oncology

"We used to think our fate was in our stars. Now we know, in large measure, our fate is in our genes." With this statement, James Watson captured the shift that placed the genome at the center of modern biomedical science. In oncology, this shift has enabled increasingly precise characterization of tumor biology and greater integration of genomic data into clinical decision-making. This thesis applies liquid biopsy as a minimally invasive and dynamic tool to study tumor-derived DNA in retinoblastoma, the most common pediatric intraocular malignancy. Although classically initiated by biallelic loss of RB1, retinoblastoma displays a broader molecular spectrum than historically appreciated, including mosaic variants, complex structural alterations, and variable tumor fractions, especially when sampling peripheral blood. These features complicate diagnosis and underscore the need for approaches capable of capturing the full diversity of genomic events contributing to disease onset and progression. In recent years, efforts to refine retinoblastoma genomics have increasingly relied on next-generation sequencing (NGS) of cell-free DNA obtained from plasma and, more effectively, from aqueous humor. In this work, combined cfDNA analysis from both sources is used as a single, integrated NGS approach that can delineate the genetic basis of disease in one step, providing comprehensive insight into both inherited predisposition and tumor-associated alterations. Parallel to these developments, the emergence of Third-Generation Sequencing (TGS) has opened new opportunities for genomic resolution not achievable with short-read technologies. Long-read whole-genome sequencing, though still scarcely explored in retinoblastoma, enables direct interrogation of native DNA molecules and offers the ability to resolve structural configurations and genomic regions that remain poorly accessible to NGS. Building on these complementary strengths, this thesis proposes an integrated diagnostic framework that unites high-sensitivity cfDNA sequencing with long-read whole-genome analysis. The combined approach enhances the detection and interpretation of genomic events across scales, from point mutations to large, previously unresolvable structural rearrangements, allowing a more complete reconstruction of individual genomic profiles. By bridging the depth of liquid biopsy with the breadth of long-read sequencing, this strategy advances the molecular characterization of retinoblastoma and lays the groundwork for next-generation precision diagnostics in pediatric ocular oncology.