Cross-sectional analysis of circulating tumor DNA in primary colorectal cancer at surgery and during post-surgery follow-up by liquid biopsy | Journal of Experimental & Clinical Cancer Research | Full Text

Cross-sectional analysis of circulating tumor DNA in primary colorectal cancer at surgery and during post-surgery follow-up by liquid biopsy | Journal of Experimental & Clinical Cancer Research | Full Text

Cross-sectional analysis of circulating tumor DNA in primary colorectal cancer at surgery and during post-surgery follow-up by liquid biopsy

• Matteo Allegretti, 
• Giuliano Cottone, 
• Fabio Carboni, 
• Ettore Cotroneo, 
• Beatrice Casini, 
• Elena Giordani, 
• Carla Azzurra Amoreo, 
• Simonetta Buglioni, 
• Maria Diodoro, 
• Edoardo Pescarmona, 
• Settimio Zazza, 
• Orietta Federici, 
• Massimo Zeuli, 
• Laura Conti, 
• Giovanni Cigliana, 
• Francesco Fiorentino, 
• Mario Valle, 
• Patrizio Giacomini & 
• Francesca Spinella 

Journal of Experimental & Clinical Cancer Research volume 39, Article number: 69 (2020) Cite this article

• 74 Accesses
• 2 Altmetric
• Metricsdetails

Abstract

Background

Liquid biopsy (LB) in early-stage, non-metastatic colorectal cancer (CRC) must be sensitive enough to detect extremely low circulating tumor DNA (ctDNA) levels. This challenge has been seldom and non-systematically investigated.

Methods

Next generation sequencing (NGS) and digital PCR (dPCR) were combined to test tumor DNAs (tDNAs) and paired ctDNAs collected at surgery from 39 patients, 12 of whom were also monitored during the immediate post-surgery follow up. Patients treated for metastatic disease (n = 14) were included as controls.

Results

NGS and dPCR concordantly (100% agreement) called at least one single nucleotide variant (SNV) in 34 tDNAs, estimated differences in allelic frequencies being negligible (±1.4%). However, despite dPCR testing, SNVs were only detectable in 15/34 (44.1%) ctDNAs from patients at surgery, as opposed to 14/14 (100%) metastatic patients. This was likely due to striking differences (average 10 times, up to 500) in ctDNA levels between groups. NGS revealed blood-only SNVs, suggesting spatial heterogeneity since pre-surgery disease stages, and raising the combined NGS/dPCR sensitivity to 58.8%. ctDNA levels at surgery correlated with neither tumor size, stage, grade, or nodal status, nor with variant abundance in paired tDNA. LB sensitivity reached 63.6% when ctDNA was combined with CEA. Finally, persistence and absence of ctDNA on the first conventional (month 3) post-surgery follow-up were associated with fast relapse and a disease-free status in 3 and 7 patients, respectively.

Conclusions

A simple clinical NGS/dPCR/CEA combination effectively addresses the LB challenge in a fraction of non-metastatic CRC patients.