Targeted detection of nucleic acid mutations for improved lung cancer diagnostics in liquid biopsies

In this project we aimed to improve the current molecular diagnostics approaches for lung cancer- related mutation detection in liquid biopsies. Liquid biopsy refers to sampling and analysis of biological fluids—typically blood—and has potential for screening, diagnosis, predicting and monitoring of treatment response. However, current assays based on circulating tumor DNA (DNA released by tumor cells in the blood) suffer from sensitivity issues due to the low percentage of tumor-derived DNA. Moreover, they rely on techniques that are costly, slow and require highly trained staff for data analysis. Two approaches were explored.

First, the effect of depletion of wild-type sequences on the sensitivity of mutation detection using hybridization-based methods was tested. Here, the limit of mutation detection could be lowered by approximately an order of magnitude. This was verified using lung cancer patient samples and benchmarked. A feasibility test showed transferability to an electrochemical sensor.

Second, it was investigated how extracellular vesicles—small cell-derived structures also present in the blood—can contribute to an increased sensitivity. Using a lung cancer cell line and blood plasma background, the enrichment of vesicles was compared to the standard workflow for mutation detection. Vesicle enrichment, prior to cell-free DNA extraction, led to significantly higher mutant allele frequencies.

Hybridization-based methods do not rely on labels and can be optimized to work at room temperature. Therefore, we believe that the combination of our depletion-approach and an electrochemical read-out can compose a senor that is cheaper and simpler than the current liquid biopsy analysis techniques. By adding a vesicle enrichment step to molecular diagnostic workflows, this can contribute to the sensitivity of current analyses. Since extracellular vesicles appear in very early stages of cancer, their use could be interesting for screening processes.