Published: Jan. 12, 2025.
Cell-free DNA (cfDNA) has emerged as a promising non-invasive diagnostic tool for early diagnosis and prognosis in human oncology. This approach allows cancer detection and monitoring through simple blood samples, circumventing the discomfort and risk associated with surgical biopsies. This study investigates the utility of cfDNA as a potential tumor marker for canine cancer diagnostics by analyzing the size distribution of cfDNA fragments through Whole Genome Sequencing (WGS) data.
The researchers conducted a pattern analysis focused on quantifying fragment sizes within cfDNA WGS data to discern the variances in the distribution of fragment sizes between healthy canine samples and those afflicted with hemangiosarcoma. Leveraging machine learning techniques, specifically Support Vector Machine (SVM), they classified hemangiosarcoma using modified versions of the ichorCNA and WisecondorX algorithms tailored to their specific research needs.
The findings in canine subjects revealed that healthy dogs exhibited a peak at 165 bp, whereas dogs with hemangiosarcoma showed a reduced peak length at 160 bp. Leveraging machine learning for data analysis, particularly the Support Vector Machine (SVM) model, yielded notable diagnostic accuracy. However, including post-operative samples in the machine learning analysis diminished performance metrics, potentially reflecting a post-surgical reduction in disease burden rather than an improvement in the dogs’ prognoses.
The successful application of fragment size distribution and machine learning in the screening of hemangiosarcoma in dogs opens the door to the potential use of this methodology in other canine cancers. The findings advocate for further exploration into the application of this non-invasive diagnostic tool across various cancer types in dogs, highlighting the value of cross-species insights in advancing cancer detection and treatment strategies.
The study reveals that cfDNA patterns closely resemble those seen in humans, indicating the potential of cfDNA fragment size distribution as a tumor marker for cancer screening and monitoring in canines. However, the study had limitations, including a small sample size and a lack of comprehensive assessment across different cancer stages. The study also highlighted the need for further data or experiments to validate the criteria for normal samples.
This study paves the way for integrating cfDNA-based liquid biopsy techniques into veterinary oncology, offering a minimally invasive and effective diagnostic tool for early cancer detection and personalized treatment strategies in dogs.