Innovative ctDNA model could improve cancer progression monitoring

In metastatic cancer surveillance, monitoring the actual concentrations of circulating tumor DNA (ctDNA) may be critical. Researchers showed that absolute ctDNA concentration thresholds can be defined to rule out or predict impending cancer progression. They introduce a dual threshold model in a novel study in The Journal of Molecular Diagnostics, published by Elsevier, that may improve cancer surveillance, patient stratification, and risk-informed, personalized treatment by providing more accurate and timely assessment of disease progression.

Lead investigator Geert A. Martens, MD, PhD, Department of Laboratory Medicine, AZ Delta General Hospital, Roeselare; and Department of Biomolecular Medicine, Ghent University (Belgium), explains, "Monitoring cancer progression in metastatic breast cancer currently relies primarily on medical imaging, supplemented by poorly specific, inexpensive biomarkers such as CA15-3. Monitoring tumor-specific mutations in circulating DNA, a concept called 'liquid biopsy' is far superior. However, clinicians currently do not know how to interpret ctDNA concentrations. We came up with a solution."

The researchers involved in the study conducted long-term (two years), frequent (five weekly) measurements of ctDNA levels in patients with advanced breast cancer and investigated whether the ctDNA level could be used to predict or rule out impending disease progression. They measured ctDNA using a variety of techniques, including targeted deep sequencing and digital PCR, which showed perfect correlation. The choice of technique used would be dictated mostly by the pathology laboratory's total cost of ownership and logistic aspects such as turnaround times, they say.

Dr. Martens explains, "We confirmed that ctDNA levels are superior to old school biomarkers such as CA15-3, and that frequent ctDNA measurement results in earlier (three months) recognition of tumor progression. But most importantly, we were able to develop a very simple dual threshold classifier that gives a clear result in 90% of blood draws. At ctDNA levels below 10 mutant copies/mL (0.25% VAF) it can reassure patients that progression is unlikely, while levels above 100 copies/mL (2.5% VAF) are associated with at least a 90% chance of progression. We launched this somewhat provocatively as the '0/10/100 copy model.' We do realize that ctDNA concentrations in blood vary with tumor type and stage, but we are confident that our novel statistical approach to the data can and should be generalized."

The investigators recommend that advanced cancer centers replace conventional protein biomarkers like CA15-3 by patient-personalized, mutation-specific digital PCR tests, and start monitoring frequently for advanced cancer surveillance and early cancer minimal residual disease. Such ctDNA monitoring holds great value: more sensitive and specific monitoring, better use of radiology resources, fewer hospital visits, less anxiety and overall, a positive health-economic impact. By looking at the actual ctDNA concentrations using relatively inexpensive PCR tests, doctors can also select the appropriate time for retesting the tumor or the liquid biopsy using Comprehensive Genomic Profiling.

This research was also able to confirm these same thresholds for surveillance of metastatic non-small-cell lung cancer patients.

Dr. Martens concludes, "In terms of practicality, our concept goes beyond the 'cohort analyses' of Kaplan-Meier survival curves. We provide the statistical framework so our work can be critically reproduced and applied retrospectively to ANY data set with registered progression outcomes. We hope this work can inspire other scientists to apply our concept. The actual concentrations of ctDNA hold strong diagnostic potential for cancer progression. We should prepare for ctDNA concentration-guided scheduling of care in advanced cancers."