Cancer evolution and multi-omic profile of relapsed colorectal liver metastases after treatment
摘要
Recurrence following resection of colorectal cancer liver metastases remains a major obstacle to prolonged patient survival, often resulting in treatment-refractory disease with limited understanding of the underlying evolutionary drivers. To investigate these mechanisms, we performed an in-depth, patient-specific study of genomic and microenvironmental alterations in relapsed colorectal cancer liver metastases from nine individuals.
MethodsClonal deconvolution and phylogenetic analysis was conducted on DNA sequencing data from multiregion liver metastasis and relapse samples from the same patient. Archived primary tumor specimens were included to trace the clonal lineages responsible for relapse. In parallel, transcriptomic data from the liver metastasis and relapse samples were analyzed to characterize tumor-infiltrating immune cell populations.
ResultsPhylogenetic analyses of relapsed metastases revealed two distinct patterns: (1) relapses that retain a clone already present in ancestral liver metastasis, and (2) relapses with no evident clonal link to the original metastasis. Relapses in the first group carried a chemotherapy-associated mutational signature, which then appeared across all relapse samples. In one patient, the relapsing clone had already diversified within the primary tumor. Tumor microenvironment analyses exposed heterogeneous responses followwing relapse, ranging from decreased to markedly increased infiltration by multiple immune cell types, accompanied by shifts in consensus molecular subtypes and altered neoantigen profiles.
ConclusionsOur study reveals patient-specific evolutionary trajectories underlying relapse in colorectal cancer, highlighting diverse routes to recurrence, including chemotherapy-driven clonal expansion and early divergence from the primary tumor. Our findings reveal a landscape of patient-specific evolutionary trajectories in relapsed colorectal cancer, underscoring the potential value of personalized approaches for understanding and monitoring recurrent disease.