It is estimated that approximately 27%–31% of NSCLC patients are unable to provide a suitable specimen upon diagnosis. In addition, invasive tissue biopsy procedures can pose a health risk to patients. The development of liquid biopsies has therefore allowed mutations to be detected from circulating tumour DNA (ctDNA), DNA fragments tumours release into the bloodstream.

Although liquid biopsies are cheaper and simpler in procedure than tissue biopsies, the method has been plagued by sensitivity issues.

Tumours release limited amounts of ctDNA, making it more difficult to identify them and subsequently determine the best treatment options. In a bid to improve the technique, scientists from the Memorial Sloan Kettering Cancer Center, the MD Anderson Cancer Center, and the Dana-Farber Cancer Institute have collaborated with DNA sequencing giant Illumina and its liquid-biopsy spinoff Grail to develop a new assay, or laboratory test, with a higher degree of accuracy.

The new assay used Illumina’s “ultradeep next-generation sequencing,” whereby the genomic region of interest was sequenced approximately 50,000 times in order to detect targetable oncogenic drivers and resistance mutations.

The sequencing information was then fed to a machine-learning algorithm developed by Grail to determine mutation readouts. The assay was tested on 127 patients with advanced, metastatic NSCLC, 91 of whom had mutations identified from tissue biopsies.

Of those, 68 had the variant detectable in ctDNA, resulting in a clinical sensitivity of 75%. The 75% detection rate demonstrated in the study suggests that a tissue biopsy might still be required.

However, the combination of liquid biopsy with machine-learning generated valuable information about cancer genetic characteristics with quick turnaround time.

The non-invasive nature makes the technology more beneficial, and with refining, the technique has the potential to replace invasive tissue biopsies.