DNA barcoding can reflect the complex cellular makeup of cancerous tumors, according to a study that could help drive the development of precision oncology.
The findings, in Molecular Systems Biology, indicated that, while both solid biopsies and those from blood samples broadly represented composition, results varied between tumors.
This suggests that combining both strategies could provide a more accurate representation of disease.
“DNA barcoding enabled us to investigate entire tumors, solid biopsies and even liquid biopsies. We were then able to accurately quantify how much tumor heterogeneity is captured in biopsies,” explained senior researcher Antonin Serrano, PhD, who is now at the University of Melbourne.
“We found that DNA shedding in the bloodstream varied widely, not only depending on necrosis and tumor burden, but also across preclinical models.
“We also found that barcode diversity in the center of primary tumors was significantly higher than in the periphery, which could have significant implications for the interpretation of solid biopsies.”
Currently, breast cancer diagnosis and prognosis are primarily based on the analysis of solid biopsies and samples collected during surgery.
These identify receptor status and markers of aggressiveness, and are routinely used to identify the subtype and grade of disease, which in turn determines therapeutic decisions.
However, liquid biopsies from blood specimens that analyze cell-free DNA and circulating tumor DNA are emerging as a powerful way to detect and monitor tumor progression and monitor the responses of patients to treatment.
DNA barcoding technology uses lentiviruses to label individual cancer cells with DNA tags, which act as barcodes. This allows the cells to be tracked and identified in tumor cells and matched biopsies.
The researchers used DNA-based cellular barcoding to label six human breast cancer xenograft models and explore the variety of clonal cells captured by solid and liquid biopsies.
Genetic barcoding was used to generate clonally structured tumors in multiple mice to determine how well solid and liquid biopsies captured tumor heterogeneity using two breast cancer cell lines and four patient-derived xenografts (PDXs).
Serrano and co-workers discovered that the superiority of one sampling method over another depended on the model and tumor burden.
The results suggested that solid biopsies could capture a wide range of minor and dominant clones present in primary tumors across models, representing between 60% and 90% of the primary tumor biomass.
However, the clonal variety captured in solid biopsies was strongly biased by spatial distribution, with the researchers noting that, in the clinic, needles are often directed towards the center of the tumor.
As a result, multiple sampling of a given tumor was likely to yield highly variable results in heterogeneous tumors.
Barcodes from cell-free DNA predominantly represented dominant clones from the primary tumors, and the results suggested that underrepresented clones in the primary tumor might not be detected.
Together, the findings suggest that, in the clinic, combining tumor and blood sampling could provide a better assessment of intratumoral heterogeneity.
“While more work will be required to better understand the properties and dynamics of cancer clones across different types of biopsies over time and in response to specific therapies, this study provides new insights into the utility of barcoded models for studying the variability and biases of solid and liquid biopsies,” the researchers reported.
“Linking meaningful clonal information to multi-omics analyses in preclinical models of cancer holds great promise in the development of diagnostic tools and the implementation of personalized medicine.”
