Liquid biopsy for early cancer detection: novel approaches show promising early signals

Detecting early-stage cancer continues to present a significant technical challenge in oncology as “There is a substantial signal-to-noise problem: the number of target molecules in stage I disease is extremely low and no liquid biopsy currently available has demonstrated adequate sensitivity and specificity,” explains Prof. Caroline Dive from the Cancer Research UK National Biomarker Centre, University of Manchester, UK. While many pioneering tests such as Galleri have represented meaningful progress toward this goal, the development of highly sensitive and highly specific detection methods are still sought (Ann Oncol. 2021;32:1167–1177; Lancet. 2023;402:1251–1260). Preliminary findings using innovative approaches presented at MAP Congress 2026 (London, 10–11 September) show potential for the detection of early-stage cancers.

A first oral presentation focused on epitranscriptomics, i.e. the study of biochemical RNA modifications. “Over 150 RNA modifications have been identified to date, with each one representing a potential target molecule to detect. Critically for liquid biopsy, some of these modifications influence RNA stability, overcoming the issue of rapid RNA degradation in blood,” highlights Dive. A 5-nucleoside circulating epitranscriptomic signature was identified from 130 patients with stage I–IV colorectal cancer (CRC) and using supervised machine learning, was shown to discriminate tumour samples with high accuracy (cross-validation F1-score ≥0.95, including stage I disease) (Abstract 7O). The assay was subsequently credentialled in a separate cohort of 47 patients with CRC, and further investigations demonstrated that epitranscriptomic profiles could distinguish pre-cancerous lesions from healthy samples. According to Dive, “This is a novel approach and although the study is signal seeking and not powered to be conclusive, the ability to pick up stage I disease and pre-cancerous lesions (polyps) from a small blood volume is potentially very exciting and definitely warrants further study.”

A second study presented at the Congress described an assay for detecting early-stage solid cancer based on the detection of sub-100 base pair CCCTC-binding factor-protected cell-free DNA (cfCTCF-DNA) fragments using chromatin immunoprecipitation sequencing (Abstract 3O). “What is particularly exciting here is that this assay seems to have overcome the obstinate challenge of the signal-to-noise ratio that has held this field back, with very high specificity for tumour-derived material,” says Dive. The training cohort comprised samples from 11 healthy patients, 10 with chronic inflammation and 49 patients with cancer, including colorectal, pancreatic, lung, breast, liver and bile duct cancer across stages I to IV. “Although the training cohort is small, the researchers reported 100% sensitivity for tumour detection, which is a genuinely exciting result,” enthuses Dive. In an independent, blinded validation cohort (31 lung cancer, 28 CRC, 22 healthy controls), respective sensitivity and specificity rates of 93% and 95% were reported.

Despite their promising findings, both studies were relatively small, and confirmation of the results in larger independent cohorts will be required to evaluate the true potential of these methods for early cancer detection.

“The roadmap from biomarker discovery to routine clinical deployment of fit for purpose tests is long and demanding,” concludes Dive. “Beyond analytical performance and clinical utility, cost-effectiveness is also a critical consideration within resource-constrained healthcare systems. Ultimately, the clinical incentive driving researchers’ efforts in this area is clear: if we can detect cancers early, the chances of cure or much longer survival with good quality of life are so much better.”