Lung Cancer Blood Tests for Diagnosis & Monitoring

Blood tests for lung cancer have moved from experimental curiosity to practical clinical tool in under a decade. The U.S. FDA has already approved the first liquid biopsy test for identifying driver mutations in non-small cell lung cancer (NSCLC), allowing doctors to match patients with targeted therapies using a simple blood draw instead of an invasive tissue biopsy.¹ But the field extends well beyond mutation profiling. Researchers are now using blood-based signals to catch lung cancer early, track whether treatment is working, spot recurrence months before a scan would, and even predict who will respond to immunotherapy.

How Blood Tests Detect Lung Cancer Signals

Tumors constantly shed material into the bloodstream. Fragments of their DNA, tiny membrane-bound packages called extracellular vesicles, whole tumor cells that have broken free, and small RNA molecules all circulate in plasma. Each of these carries information about the cancer’s genetics and behavior. The challenge is fishing those faint signals out of a blood sample dominated by normal cells and normal DNA. Different tests target different types of shed material, and each has its own strengths depending on whether the goal is early detection, treatment selection, or ongoing monitoring.

Early Detection Through Circulating Tumor DNA

One of the most promising approaches for catching lung cancer early relies on circulating tumor DNA, or ctDNA. Tumors leave a chemical fingerprint on their DNA through a process called methylation, and certain methylation patterns show up reliably in lung cancer. Genes like SHOX2 and RASSF1A, for instance, display specific methylation changes in plasma samples from people with early-stage disease.² By scanning for these patterns in a blood sample, researchers can flag the presence of cancer without a CT scan or biopsy.

The numbers from validation studies are encouraging, though not yet perfect. In one study using targeted methylation sequencing, the test correctly identified about 75% of stage Ia patients and roughly 86% of stage Ib patients, while correctly ruling out cancer in over 93% of healthy individuals.³ A separate approach, using a two-step model called LunaCAM, achieved over 90% sensitivity for distinguishing lung cancer from healthy participants, and about 83% specificity for telling cancer apart from benign lung diseases like infections or inflammatory nodules.⁴ That second distinction matters a lot in practice, because the hardest diagnostic challenge is not telling a cancer patient from a perfectly healthy person but telling cancer apart from something benign that also shows up on imaging.

Circulating Tumor Cells

Instead of looking for fragments of tumor DNA, some tests hunt for intact cancer cells that have entered the bloodstream. A meta-analysis pooling data from multiple studies found that these circulating tumor cells, or CTCs, had good overall diagnostic value for detecting lung cancer, with high sensitivity and specificity.⁵ One platform tested in NSCLC patients achieved roughly 82% sensitivity and 87% specificity for distinguishing cancer from benign lung disease.⁶

CTCs offer something that DNA fragments cannot: you can actually study the whole cell, including its protein expression and whether it has begun transitioning toward a more invasive form. That said, CTCs are rare in the blood, especially in early-stage disease, and capturing them reliably requires specialized equipment. As a practical matter, ctDNA-based assays have gained more traction in clinical use, though CTC testing remains a valuable complementary approach.

Guiding Targeted Therapy With Liquid Biopsy

Where blood tests have already changed day-to-day cancer care is in molecular profiling for treatment decisions. Many NSCLC tumors carry specific genetic mutations that can be targeted with precision drugs, particularly mutations in the EGFR gene. Traditionally, identifying these mutations required a tissue biopsy, which is not always feasible depending on tumor location, a patient’s overall health, or how much tissue is available. Blood-based ctDNA testing offers an alternative. In one validation study, ctDNA analysis showed roughly 97% agreement with tissue biopsy for detecting the two most common EGFR driver mutations, with over 91% sensitivity and 100% specificity.⁷

This high concordance means that when tissue is unavailable, a blood draw can reliably guide therapy. It also opens the door to repeat testing over time without subjecting patients to additional biopsies.

Tracking Drug Resistance in Real Time

Cancer is a moving target. Even when a targeted therapy works initially, tumors often develop resistance mutations. One of the best-studied examples is the T790M mutation in EGFR, which commonly emerges after first-line treatment with EGFR-targeting drugs. Detecting T790M matters because a specific second-line drug, osimertinib, works against it. Blood tests allow doctors to check for this resistance mutation through repeat blood draws rather than additional biopsies.

A real-world multicenter study that followed patients over time found that the probability of detecting T790M through repeated liquid biopsies stayed consistent across each subsequent draw.⁸ In a Hungarian study covering over 400 samples, T790M was found in about 48% of samples that already showed activating EGFR mutations, though detection rates were lower in plasma samples alone compared with other sample types like pleural fluid.⁹ When tissue biopsy is not an option, liquid biopsy remains a valuable fallback for catching resistance as it develops.

Detecting Recurrence After Surgery

For patients who have had lung cancer surgically removed with the intent to cure, the lingering question is whether tiny amounts of cancer remain. This concept, called minimal residual disease, or MRD, is where ctDNA monitoring may offer its most dramatic advantage over traditional follow-up imaging.

Across multiple studies, ctDNA detected after surgery has been a strong signal that cancer would return. One study found that positive ctDNA detection as early as seven days after surgery identified patients at significantly higher risk of recurrence, and that longitudinal monitoring at two or more post-surgical time points flagged relapse an average of about 145 days before imaging did.¹⁰ Another study reported that post-treatment ctDNA detection preceded radiographic progression in about 72% of patients by a median of roughly five months, and that over half of those patients harbored ctDNA mutation profiles associated with favorable responses to available therapies.¹¹

A review of these studies noted that when looking only at the first post-treatment time point, sensitivity for predicting relapse ranged widely, from 36% to 100%. But when surveillance included multiple follow-up draws, sensitivity improved to between 82% and 100%.¹² The message for patients is that a single negative blood draw after surgery is reassuring but not definitive. Serial monitoring over months paints a much more reliable picture.

Reading Immunotherapy Response From Blood

Immunotherapy has transformed outcomes for many lung cancer patients, but it does not work for everyone, and standard imaging struggles to judge early response. A tumor that looks like it is growing on a scan might actually be swollen with immune cells attacking it, a phenomenon called pseudoprogression. Blood-based ctDNA tracking can help distinguish pseudoprogression from true disease worsening.¹³

In a phase 2 trial of pembrolizumab, patients whose ctDNA cleared by the third treatment cycle had dramatically better survival outcomes: median overall survival was not yet reached for responders versus about seven months for those whose ctDNA rose.¹⁴ Other research has confirmed that rapid ctDNA decreases before the first imaging assessment correlate with clinical benefit, suggesting ctDNA dynamics could serve as an early pharmacodynamic readout of immunotherapy response.¹⁵

Beyond monitoring, blood tests can also help predict who might benefit from immunotherapy in the first place. Tumor mutational burden, a rough measure of how many mutations a tumor carries, can be estimated from a blood sample. In one study, patients with higher blood-based tumor mutational burden had roughly four times higher response rates compared to those with low values, and their disease took longer to progress.¹⁶

MicroRNAs, Exosomes, and Other Non-DNA Signals

Not all blood-based lung cancer tests rely on tumor DNA. MicroRNAs, small regulatory RNA molecules, circulate freely in plasma and shift in characteristic patterns in cancer patients. A study using a panel of 15 microRNAs distinguished lung cancer patients from others with over 91% accuracy, and a separate 14-microRNA panel identified early-stage lung cancer with about 76% sensitivity and over 97% specificity.¹⁷ Earlier work identified a three-microRNA combination yielding about 81% sensitivity and 87% specificity even among stage I patients.¹⁸

Exosomes and other extracellular vesicles shed by tumors also carry diagnostic cargo. Profiling proteins on these vesicles has shown promise for identifying lung cancer regardless of stage or subtype.¹⁹ One experimental platform analyzing vesicle membrane proteins achieved 100% sensitivity in a cohort of 80 clinical samples, outperforming traditional protein-detection methods.²⁰ Meanwhile, long non-coding RNAs carried inside exosomes have shown the ability to separate early-stage NSCLC from healthy controls.²¹ These approaches are still largely experimental, but they illustrate how many different molecular channels tumors broadcast on.

Fragmentomics and the Shape of Cell-Free DNA

A newer twist on blood-based detection does not look at what the DNA says but at how it breaks apart. Cancer-derived cell-free DNA tends to fragment differently from normal DNA, producing shorter pieces with distinctive patterns tied to how tumor cells package their chromosomes. Researchers have built machine learning models that read these fragmentation patterns across the entire genome to flag cancer.

In a prospective study of people at risk for lung cancer, a model combining fragmentation features with clinical risk factors and a common protein marker detected 94% of patients with cancer, including 91% of those with stage I or II disease, at 80% specificity.²² Other groups have confirmed that cancer samples show a measurable leftward shift in fragment size distributions compared with healthy or nodule samples.²³ Because fragmentomics does not require deep sequencing of specific gene targets, it could potentially be cheaper and more scalable than mutation or methylation analysis.

Where Blood Tests Can Go Wrong

No test is immune to false positives or negatives, and blood-based cancer tests have unique pitfalls. The most important one to understand is clonal hematopoiesis, sometimes shortened to CHIP. As people age, blood-forming stem cells can accumulate cancer-associated mutations even in the absence of any cancer. Mutations in genes like KRAS and TP53, both common in lung cancer, can show up in blood DNA purely because of these age-related changes in blood cells. A KRAS mutation from CHIP could lead to inappropriate therapy, and rising TP53 levels might falsely suggest cancer recurrence during surveillance.²⁴ Specialized techniques that sequence white blood cells alongside plasma can filter out these false signals, but not all labs routinely perform that step.²⁵

False negatives are also a real concern. Not all tumors shed equal amounts of DNA into the bloodstream. Studies have found that patients with better overall health status, tumors confined to one side of the chest, and lower overall disease burden are less likely to have detectable ctDNA in plasma.²⁶ In plain terms, blood tests are more likely to miss exactly the patients who most need early detection: those with small, localized tumors. This is a fundamental tension the field is still working to resolve.

Pairing Blood Tests With CT Screening

Because of these limitations, blood tests are increasingly seen not as replacements for imaging but as complements. One proposed workflow uses a high-sensitivity blood-based fragmentomic assay as a pre-screen to help identify individuals who would benefit most from a low-dose CT scan, potentially expanding screening to people who currently fall outside guidelines.²⁷ Another study found that combining molecular and immune markers from blood with CT results significantly improved screening accuracy and reduced the number of unnecessary follow-up procedures for benign nodules.²⁸ If a nodule shows up on CT and the blood test also flags a concerning signal, confidence in pursuing biopsy goes up. If the blood test is clean, a wait-and-watch approach feels safer.

Multi-cancer early detection tests, which scan a single blood sample for signals from dozens of cancer types, represent an even more ambitious version of this idea. In the prospective PATHFINDER study, the test did detect a case of lung cancer, though it presented at stage III, highlighting the ongoing challenge of catching lung cancer early even with broad screening tools.²⁹

Sample Handling Matters More Than You Might Think

A detail that rarely makes headlines but directly affects test accuracy is how the blood sample is collected and processed. Cell-free DNA is fragile, and if white blood cells in the sample break open during transport, they flood the plasma with normal genomic DNA that drowns out the faint tumor signal. The type of collection tube used makes a real difference. Standard EDTA tubes, the kind used for routine blood work, allow significant DNA contamination if processing is delayed. In one study, EDTA tubes held for 72 hours before processing saw a sixfold increase in total DNA yield from cell lysis, and the detection rate for KRAS mutations in known-positive patients dropped from 50% (when processed quickly) to just 20%.³⁰ Specialized cell-free DNA collection tubes that contain preservatives to prevent this lysis maintained stable DNA quality even after 72 hours at room temperature.³¹ If you are having a liquid biopsy drawn, confirming that the lab uses the right tube type is a small detail with outsized impact.

Cost and Access Considerations

Blood tests are less invasive and generally less expensive than tissue biopsies, but “less expensive” does not automatically mean affordable. Cost-effectiveness depends heavily on the clinical context. For patients with advanced NSCLC who cannot undergo tissue biopsy, a European modeling study found that comprehensive liquid biopsy testing was cost-effective, particularly for EGFR-mutated patients, where matching the right targeted therapy produced enough benefit to justify the cost.³² A separate analysis from Spain using the FoundationOne Liquid CDx test found an incremental cost of under €10,000 per quality-adjusted life year gained, well within typical willingness-to-pay thresholds.³³

For population-level screening of healthy individuals, the math changes. A Brazilian analysis estimated that liquid biopsy screening in high-risk people would only become cost-effective if lung cancer prevalence in the screened population exceeded 4%.³⁴ In most real-world screening populations, prevalence sits well below that threshold, which means widespread blood-based screening will need either significant cost reductions or better-targeted enrollment to make economic sense.

Equity Gaps in Testing

Molecular testing, whether tissue-based or blood-based, has the potential to either reduce or widen existing health disparities. The frequency of certain targetable mutations varies by ancestry. EGFR mutations, for example, appear in roughly 45% of lung cancer patients of East Asian descent but only about 10% of those with European or African ancestry.³⁵ This biological variability means that the practical benefit of ctDNA testing is not uniform across populations, but the underlying technology performs similarly regardless of race. A real-world analysis found that similar proportions of all racial and ethnic subgroups had actionable mutations detected through ctDNA testing, with comparable variant allele frequencies across groups.³⁶ The barrier, then, is not the technology’s performance but whether patients are offered testing in the first place. Ensuring equitable access to liquid biopsy, particularly in underserved communities, could help narrow some of the gaps in lung cancer outcomes.

Pleural Fluid as an Alternative Source

Blood is the most accessible body fluid for liquid biopsy, but it is not the only one. For patients who develop pleural effusions, the fluid that accumulates around the lungs, that fluid can be richer than plasma in tumor-derived products including ctDNA, microRNAs, and circulating tumor cells.³⁷ Since pleural fluid is often drained for symptomatic relief anyway, analyzing it for molecular markers is a practical way to extract additional diagnostic information from a procedure that was already happening. Studies comparing T790M detection rates found markedly higher detection in non-plasma samples, including pleural fluid, compared with plasma alone.⁹ When available, these organ-specific fluids can complement blood-based testing, especially in cases where plasma ctDNA levels are low.