Back when Shaw first entered the field, identifying and decoding these clues was a mammoth task. However, the science has changed dramatically in the past 25 years.

"The big revolution has been the technologies and the types of approaches we can use," Shaw says. High-throughput sequencing, combined with machine learning approaches capable of analysing enormous datasets, has transformed what researchers can detect and how confidently they can interpret it. Just as important, she says, has been the accumulation of clinical evidence demonstrating where these tests genuinely improve patient care.

Those advances have gradually moved liquid biopsy from an academic curiosity toward routine clinical practice. In England, circulating tumour DNA testing is now reimbursed for selected patients with lung and breast cancer, reflecting growing evidence that, in the right settings, blood-based testing can complement existing diagnostic pathways while offering a cost-effective alternative when repeated tissue sampling is not feasible.

For researchers like Shaw, who have dedicated decades to proving the validity of the tool alongside tissue biopsy this is an exciting development. But, she stresses, it is not a case of one tool replacing another. Instead, the strength of liquid biopsy lies in answering questions that tissue simply cannot.

"You will never collect a tissue biopsy from a patient every three months or every two weeks," says Dr Florent Mouliere, group leader at the Cancer Research UK National Biomarker Centre in Manchester. "That's where liquid biopsies have a unique selling point compared to tissue biopsies, that you can do this real-time understanding about what's happening in the patient's body."

Much of the field has focused on circulating tumour DNA, or ctDNA, fragments of DNA shed by cancer cells that can be sequenced to identify tumour-specific mutations. But researchers are increasingly looking beyond mutations alone. Epigenetic markers, such as DNA methylation can provide clues about where a tumour originated, while RNA, proteins, and extracellular vesicles each offer a different window into tumour biology.

"We're not trying to focus solely on the technology," Mouliere explains. "We're really trying to have a composite approach where we improve the biological understanding of cell-free DNA to do other things."

This shift towards multimodal liquid biopsy reflects a growing recognition that no single biomarker tells the whole story. Cell-free DNA, for example, is largely released when cells die through processes such as apoptosis or necrosis. Extracellular vesicles, by contrast, are secreted by living cells, offering complementary information about what a tumour is doing in real time.

"There are no magic bullets," says Mouliere. "There's a lot to be learned by combining the different analytes together because they represent different biological processes."

"It isn't a globally useful tool," says Shaw. "Not every patient at every point in time will have a signal." A negative result, she explains, does not necessarily mean the absence of disease. Some tumours shed very little DNA into the bloodstream, while others release enough material to be detected with relative ease, making repeat sampling and an understanding of tumour biology essential to interpreting results.

Those biological differences help explain why liquid biopsy has progressed unevenly across oncology. In colorectal and lung cancers, where tumours typically shed substantial amounts of DNA into circulation, blood-based testing has demonstrated clear clinical value. Brain tumours remain more challenging because the blood-brain barrier limits the release of tumour DNA into the bloodstream, while researchers are still trying to understand why renal cancers often produce weaker signals.

For Shaw, many of those unanswered questions remain rooted in breast cancer. Her group continues to investigate why circulating tumour DNA performs exceptionally well in some breast cancer subtypes, but less consistently in oestrogen receptor-positive disease, the most common form of breast cancer. They are also exploring whether blood-based biomarkers could help refine breast screening pathways by identifying which women with suspicious mammograms are most likely to require biopsy.

"Could a blood test triage women as high- or low-risk?" she asks. "That may or may not need to have that needle biopsy." Such an approach could reduce unnecessary procedures for patients while easing pressure on healthcare systems, although Shaw is quick to stress that these applications still require robust clinical validation.

Beyond individual biomarkers, researchers are increasingly looking toward multimodal approaches that combine multiple layers of biological information. Rather than relying on circulating tumour DNA alone, future tests may integrate DNA mutations, methylation patterns, RNA, proteins, and extracellular vesicles to build a more complete picture of disease. Moreover, artificial intelligence is expected to play a growing role in making sense of those increasingly complex datasets.

"I think both," says Dr Florent Mouliere, when asked whether the next breakthrough will come from biology or computational advances. "AI will change the way we are analysing our data. It is already changing the way we're analysing our data." But he argues that improved computational methods alone will not be enough. A deeper understanding of the biology of circulating nucleic acids will be equally important in unlocking new applications, both within oncology and beyond.

For now, however, both researchers return to a more measured conclusion. Liquid biopsy is unlikely to replace tissue biopsy, nor will it provide a universal answer for every patient or every tumour. Its impact instead lies in adding another layer of information to cancer care, one that allows clinicians to monitor disease over time, personalise treatment decisions and, increasingly, intervene before conventional tools reveal that anything has changed.

"There is a lot still to do," Shaw says. "There's a good career for the next generation to be had in this field."

After spending more than 25 years watching liquid biopsy evolve from an unconventional research idea into an emerging clinical reality, she still describes each new discovery with the same enthusiasm that first drew her into cancer genomics. The excitement, it seems, comes how much knowledge is still to be uncovered.