Your Liquid Biopsy Came Back Clear. But Was It Really Looking in the Right Place? 

In pancreatic cancer, a negative ctDNA result might not mean what you think it means. Here’s the data – and what it changes. 

There is a moment in oncology that happens more often than most people realise. 

The liquid biopsy result arrives. ctDNA: not detected.  

The oncologist reviews it, the patient breathes a little easier, and the surveillance plan continues unchanged. 

But what if the cancer was there; and the test simply wasn’t designed to find it in the way it needed to? 

This is not a hypothetical.  

It is what our research, presented at ASCO 2026 in collaboration with the Medical College of Wisconsin Cancer Center, found in metastatic pancreatic cancer patients. And the numbers are significant enough that every oncologist managing advanced solid tumours should know them. 

The Study – What We Did 

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive and genomically complex cancers in oncology. It is also one where liquid biopsy has shown the most promise, and the most frustrating limitations. 

We enrolled 16 advanced metastatic PDAC patients under an IRB-approved protocol at the Medical College of Wisconsin (PREDICT-MCW | NCT05802069). Each patient underwent serial blood collection at baseline and every three months for up to 12 months, giving us up to five timepoints of longitudinal data per patient. 

At each timepoint, we ran two analyses in parallel: 

1. ctDNA profiling: the standard approach, analysing cell-free DNA from plasma at 8,000x sequencing depth using our OncoIndx® 1,080-gene panel
2. Single-cell CTC profiling: using our OncoIncytes™ Cell Biopsy® platform to isolate live single Circulating Tumour Cells from blood and profile their DNA at true single-cell resolution

Then we compared what each found. 

The Finding That Changes the Conversation 

Across all samples, single-cell CTC-DNA profiling identified 32 unique genomic alterations in these patients. 

Of those 32 alterations: 

• 65% were found exclusively in CTC-DNA; completely absent from matched ctDNA 
• Only 29% were shared between the two modalities 
• The remaining alterations were unique to one platform or the other 

Read that first number again. Let that set in.  

65% of the genomic alterations we found would have been invisible to a ctDNA-only liquid biopsy. 

This is not a rounding error. This is not a technical artefact. This is a reflection of a fundamental biological reality: ctDNA and CTCs are capturing different information from different tumour populations, and relying on only one of them means accepting that you will miss a significant portion of what is actually happening in the tumour. 

But There’s More – The KRAS Finding 

Here is where the data gets particularly clinically significant. 

KRAS mutations were detected in ctDNA in multiple patients; as expected, given that KRAS drives the vast majority of pancreatic cancers. But when we profiled the CTCs from those same patients, the KRAS mutations were absent in matched CTC-DNA. 

What does this mean?  

It means the CTCs in those patients were KRAS-independent; driven by entirely different oncogenic pathways than the dominant KRAS-mutant clone detected by ctDNA. 

This is not a minor detail. It is a finding that aligns with a landmark spatial profiling study published in Nature Genetics in 2026 – confirming that KRAS-independent subclones genuinely exist in PDAC, represent biologically distinct tumour populations, and may be specifically associated with therapeutic resistance and alternative pathway dependence. 

In those KRAS wild-type CTCs, we found co-occurring TP53 alterations alongside activation of the MAPK/RTK, PI3K-AKT-mTOR, and TGFβ/SMAD pathways- alternative oncogenic drivers that ctDNA alone would never have identified in those patients. 

What This Means at the Bedside 

Let’s translate this into clinical terms, because this is ultimately what matters. 

A patient with metastatic pancreatic cancer receives a ctDNA liquid biopsy. The result shows a KRAS G12D mutation – consistent with the diagnosis, not unexpected, limited therapeutic options available for this specific variant. 

The oncologist reviews the result. Considers the available KRAS-targeted therapies. Makes a treatment plan. 

But what the ctDNA result did not show, because it could not:- is that this same patient has a population of KRAS-independent CTCs, harbouring PI3K-AKT-mTOR pathway activation, that represent a distinct and potentially therapy-resistant subpopulation of tumour cells. A subpopulation that may be the very cells driving disease progression. A subpopulation that has a completely different therapeutic vulnerability from the KRAS-mutant clone. 

Without CTC profiling, that information does not exist. The treatment plan is built on an incomplete map. 

The Longitudinal Picture: Cancer Caught in Motion 

Beyond the cross-sectional findings, our serial sampling approach revealed something equally important: dynamic clonal evolution over time. 

Across up to five timepoints per patient, single-cell CTC genomics revealed evolving molecular profiles, the emergence of new mutations, the disappearance of others, the gradual enrichment of specific pathway alterations under treatment pressure. 

Post-treatment, CTC genomics showed profiles enriched for activating variants in the PIK3CA-mTOR-AKT pathway– compared to the baseline tissue and ctDNA profiles that were dominated by KRAS and TP53. The tumour had changed. The treatment had driven it to change. And only longitudinal CTC monitoring captured that evolution as it happened. 

This is what minimal residual disease testing looks like at its most granular, not just detecting whether disease is present, but tracking how it is changing in real time. 

Why Pancreatic Cancer Specifically? 

Pancreatic cancer is a particularly important test case for this kind of analysis for several reasons. 

• First, tissue biopsy in PDAC is notoriously difficult- the pancreas is deep, often surrounded by vital structures, and biopsy carries real procedural risk. Liquid biopsy was supposed to be the answer. 
• Second, ctDNA sensitivity in PDAC is known to be lower than in some other cancers – early-stage disease in particular often yields very low or undetectable ctDNA levels. This means the false-negative rate for ctDNA in pancreatic cancer may be higher than in, say, colorectal or lung cancer, making the additional sensitivity of CTC profiling particularly valuable. 
• Third – and perhaps most importantly, PDAC is one of the cancers where treatment options are most limited and where finding any additional actionable target can make a meaningful difference to a patient’s options. In that context, missing 65% of the genomic alterations is not a statistical problem. It is a clinical one. 

What This Means for the Liquid Biopsy Field 

We want to be clear about something: this research is not an argument against ctDNA liquid biopsy. ctDNA testing is a genuine advance in cancer diagnostics and has transformed care for millions of patients. 

This is an argument for completeness. 

ctDNA captures signals from dead and dying tumour cells shedding DNA into the bloodstream. CTC profiling captures signals from live tumour cells actively circulating — the cells with the highest biological relevance to metastasis, resistance, and progression. 

These are not the same signal. They are complementary and together, they deliver a molecular portrait of the tumour that neither provides alone. 

As our translational impact statement from ASCO 2026 put it: “Single-cell liquid biopsy may enable identification of treatment-resistant tumour subpopulations not detectable through conventional ctDNA analysis alone.” 

That is not a research aspiration. That is what the data shows. 

The Practical Upshot 

If you are an oncologist managing advanced PDAC – or any solid tumour type where ctDNA sensitivity may be limited, here are the questions this research raises: 

1. “When my patient’s ctDNA comes back negative or low, am I confident the test has captured the full picture?” 
2. “Are there treatment-resistant subclones in this patient that ctDNA cannot see?” 
3. “Would CTC-based profiling change my understanding of this tumour and this patient’s options?” 

And if you are a patient with pancreatic cancer reading this; know that the science is advancing faster than most people realise. The tools to see what was previously invisible exist. The question is whether they are being used. 

At 1Cell.Ai, our OncoIncytes™ platform and Cell Biopsy® technology were built to make sure they are. 

The full ASCO 2026 abstract- Poster #4216, Medical College of Wisconsin & 1Cell.Ai- is available on our website. To learn more about our Cell Biopsy® platform and OncoIncytes™