Liquid Biopsy (Circulating Tumour DNA)

Liquid biopsy detects tumour derived material in blood, most often circulating tumour DNA (ctDNA), which is short fragments of DNA released by dying tumour cells into plasma. It gives a non-invasive way to find tumour mutations, monitor treatment, and detect disease after surgery, and it can be repeated easily.

ctDNA is a small fraction of all cell free DNA, from under 0.01% in early disease to over 10% in advanced cancer. It has a half life of about 15 minutes to 2 hours, so levels reflect the current tumour burden. Methods include digital droplet PCR for known variants, and next generation sequencing with error correction for broad panels. Tumour informed assays follow patient specific variants found in the resected tumour.

Blood (10 to 20 mL) is collected in tubes that stabilise cells, or in EDTA tubes processed within 4 to 6 hours, and plasma is separated by two rounds of centrifugation. Results are ready in 5 to 14 days. The test is used when tissue is not available, when quick results are needed, or to follow disease over time.

For known driver variants in lung cancer, ctDNA genotyping has a specificity of over 95%, and a sensitivity of about 60 to 80% compared with tissue, depending on tumour burden. A positive result is reliable, and a negative result does not exclude a mutation.

Indications.

• Genotyping of advanced cancer (lung, colorectal, breast, prostate and others) when a tissue biopsy is not possible, or the result is awaited and treatment is urgent.
• Detection of resistance mutations at progression (for example, EGFR T790M in lung cancer, ESR1 in breast cancer, RAS in colorectal cancer after anti-EGFR treatment).
• Monitoring of response to treatment, by the fall or rise of ctDNA levels.
• Detection of minimal residual disease after surgery, for example in colorectal and breast cancer, to predict recurrence, and in research to direct adjuvant treatment.
• Early detection of recurrence, before it is seen on imaging.
• Tumour profiling in cancer of unknown primary, and in settings where tissue is exhausted.

Not indicated as a screening test for cancer in healthy people, since multi-cancer methylation tests are not validated for routine use. A negative result is not used alone to exclude disease.

Specimen requirements.

• 10 to 20 mL whole blood in cell stabilising tubes (stable for up to 3 to 7 days at room temperature), or EDTA tubes spun within 4 to 6 hours.
• Plasma separated by a first spin (800 to 1600 g, 10 minutes) and a second high speed spin (about 16000 g, 10 minutes), then stored at -80°C, and shipped frozen.
• Avoidance of haemolysis, and of tubes that release DNA from white cells.
• Blood drawn before a new treatment cycle, and not within a few days of surgery or biopsy, which raise background DNA.
• Clinical details: tumour type, stage, treatment and previous molecular results.

Normal appearances.

Healthy people have low levels of cell free DNA from white cells, with no tumour variants. A negative result means no variant was detected over the detection limit of the assay, and the limit (variant allele frequency, for example 0.1 to 0.5%) is stated in the report.

Interpretation.

• Positive for a targetable variant: usually accepted for treatment selection, particularly in lung cancer, with the variant and its allele frequency reported.
• Negative for a variant: not conclusive, and tissue testing is done if the clinical picture allows, since low shedding tumours (slow growing, brain metastases, small volume disease) give false negatives.
• Resistance mutations: emergence at progression informs the next treatment, and multiple resistance clones can appear.
• Dynamic changes: a fall of ctDNA by 50% or more within weeks of starting treatment predicts response, and a rise precedes progression on imaging by weeks to months.
• Minimal residual disease after surgery: a positive result identifies patients with a high chance of recurrence, and a negative one a lower risk, though not zero.
• Variants of white cell origin (clonal haematopoiesis, in genes such as DNMT3A, TET2, ASXL1, TP53 and PPM1D) can be mistaken for tumour variants, and matched white cell sequencing helps to separate them.

Critical values and findings needing immediate action.

• A targetable variant at high allele frequency in a patient needing urgent treatment: start the matched treatment through the oncology team.
• A rising ctDNA with no radiological change in a patient on treatment: early imaging and clinical review.
• A negative result in a patient with strong suspicion of an actionable alteration: arrange tissue biopsy without delay.
• A germline like variant (allele frequency near 50%) in a hereditary cancer gene: referral for genetic counselling.

Limitations and artefacts.

• Low sensitivity when tumour shedding is low, so a negative result is unreliable, particularly in early stage disease, brain tumours and some histologies.
• Pre-analytical errors: haemolysis, delay in processing, wrong tubes, one step centrifugation and freeze thaw cycles.
• Clonal haematopoiesis and germline variants produce false positive tumour calls.
• Different assays differ in genes covered, detection limits and error correction, and results are not interchangeable.
• Copy number changes and fusions are detected less reliably than point mutations, particularly in DNA-based panels.
• Variants of uncertain significance are reported often, and their interpretation changes over time.
• Costs, turnaround and limited evidence for use of the result in treatment decisions outside defined settings.
• Tissue testing remains the reference, since it also gives histology and protein biomarkers.

Special groups.

• Patients who cannot have a biopsy (poor performance status, unsafe site): plasma testing is the main option.
• Patients with cancer of unknown primary and limited tissue: plasma testing adds information.
• Patients with high tumour burden and liver metastases: higher yield of ctDNA.
• Patients after surgery for colorectal cancer: tumour informed assays for minimal residual disease are used in research and selected practice.
• Pregnant women: foetal DNA can be present in plasma, and a tumour derived signal is considered with specialist input.
• Regions: liquid biopsy needs good sample handling, cold chain shipping and sequencing capacity, which are available in few places. Where tissue testing is unavailable, plasma testing with a small, validated panel for a key target (for example, EGFR) can be a practical alternative, and costs and access to targeted drugs decide its value.

Follow up tests.

• Tissue biopsy for histology, immunohistochemistry and molecular tests.
• Full next generation sequencing on tumour tissue.
• Matched white cell sequencing to exclude clonal haematopoiesis.
• Serial ctDNA at set intervals during treatment and follow up.
• Staging and response imaging.
• Tumour markers (for example, carcinoembryonic antigen, PSA, CA 15-3).
• Germline genetic testing and counselling when indicated.
• Multidisciplinary meeting.