Reading EGFR in Plasma: What ctDNA Testing Can and Can't Tell You in Lung Cancer

The clinical question

When someone is diagnosed with advanced non-small cell lung cancer (NSCLC), one of the first and most consequential questions is whether the tumor carries an activating EGFR mutation. The answer determines eligibility for a well-established class of oral targeted drugs — EGFR tyrosine kinase inhibitors (TKIs), including osimertinib — rather than committing the patient to chemotherapy by default. But the mutation type matters, not just its presence. Classic sensitizing alterations (exon 19 deletions and the L858R point mutation) gate first-line TKI therapy; exon 20 insertions point toward distinct agents; the T790M variant signals acquired resistance; and atypical mutations (S768I, L861Q, G719X) occupy a more nuanced middle ground.

Historically this answer came from tissue. The problem is that tissue isn't always obtainable — some tumors are anatomically awkward to biopsy, some patients are too frail, and some specimens simply don't yield enough viable tumor cells for molecular testing. That's the gap a plasma-based assay is meant to fill: can we read the EGFR genotype from a tube of blood?

What's measured, and how

The analyte is circulating tumor DNA (ctDNA) — short fragments of tumor-derived DNA shed into the bloodstream and recovered from the plasma fraction of a blood draw. The two dominant platform families are PCR-based assays, which interrogate a defined panel of known hotspot mutations, and next-generation sequencing (NGS), which casts a wider net across the gene and can detect a broader range of variants [2].

The critical caveat — and it's one worth repeating to patients and trainees — is sensitivity. ctDNA testing detects mutations when tumor DNA is actually present in the plasma at a detectable level. Tumors shed DNA at highly variable rates depending on burden, site, and biology, so a negative plasma result does not reliably mean the mutation is absent. This is the fundamental asymmetry of liquid biopsy: a positive plasma result is highly informative and actionable, whereas a negative result often needs tissue confirmation before chemotherapy is assumed to be the only path. Peer-reviewed laboratory guidance published in 2025 lays out integration strategies for exactly this reality, describing reflex workflows for both NGS and PCR platforms that cover every actionable result state — exon 19 del, L858R, exon 20 insertion, T790M, and atypical variants — and endorsing plasma-first approaches specifically in tissue-limited settings [2].

The evidence

The regulatory precedent here is older than many clinicians realize. In June 2016, the FDA approved the cobas EGFR Mutation Test v2 for use with plasma — the first liquid-biopsy companion diagnostic in NSCLC, cleared to detect exon 19 deletions and L858R to gate first-line EGFR TKI therapy [1]. This mattered beyond its immediate indication. It established that a plasma ctDNA assay could meet regulatory standards as a companion diagnostic and set the template for essentially every liquid-biopsy CDx approval that followed [1]. The approval was deliberately framed as an alternative to tissue, not a replacement — tissue testing remained the recommended fallback when plasma was negative, precisely because of the limit-of-detection problem described above.

Nearly a decade on, the question has shifted from "can plasma substitute for tissue at baseline?" to "what else can plasma tell us over time?" Here the evidence is genuinely promising but should be read as investigational rather than practice-defining. A 2026 systematic review and meta-analysis examined ctDNA clearance as an early on-treatment signal in unresectable EGFR-mutant NSCLC treated with osimertinib [3]. The evidence suggests that patients whose plasma EGFR ctDNA clears at 2–4 weeks after starting osimertinib have better progression-free and overall survival than those in whom it persists [3]. That reframes ctDNA as a dynamic, pharmacodynamic readout — a way to watch whether the drug is biologically working — rather than solely a one-time, baseline genotyping tool [3]. It's an attractive concept, particularly for adjuvant and locally advanced disease where monitoring questions loom large, but it's exactly the kind of fast-moving area that warrants caution: clearance is a prognostic correlation, not yet a validated trigger for changing therapy.

Research is also pushing ctDNA upstream of diagnosis. A prospective feasibility study registered in December 2024 (NCT06716580) is evaluating plasma ctDNA EGFR screening in high-risk East Asian and Latinx populations — groups disproportionately affected by EGFR-mutant lung cancer yet frequently excluded from standard low-dose CT screening criteria, which were largely built around smoking history [4]. If feasible, this could extend the biomarker's utility from treatment selection into early detection in populations current screening paradigms underserve [4]. For now, it's a feasibility question, not a screening recommendation.

Where it stands

It helps to separate what's validated from what's investigational. The validated core is narrow but solid: plasma ctDNA detection of sensitizing EGFR mutations as a companion diagnostic to gate first-line TKI therapy, backed by the 2016 FDA approval and operationalized through the laboratory integration strategies formalized in 2025 [1,2]. In practice, a plasma-first reflex workflow — test blood when tissue is limited, confirm negatives with tissue where feasible — is well supported [2]. The resistance mutation T790M and the therapeutically distinct exon 20 insertions are part of the actionable landscape these workflows are built to report, since each routes patients toward a different drug class [2].

The investigational frontier is where the excitement — and the appropriate restraint — lives. Using ctDNA clearance to monitor osimertinib response is biologically compelling and backed by meta-analytic correlation with survival, but it is not yet a validated basis for MRD-guided treatment decisions [3]. Similarly, ctDNA-based early detection in screening-ineligible high-risk groups remains a feasibility question under active study [4]. None of these dynamic or screening applications should be read as directing individual treatment choices; they inform eligibility for drug classes and, increasingly, our understanding of who benefits and when.

The throughline is proportionality. Plasma EGFR testing has earned a durable place in the first-line diagnostic workup — a genuine, regulator-endorsed alternative when tissue falls short. What it can't yet do is replace clinical judgment about negatives, or serve as a standalone trigger for stopping or switching therapy. Keeping those boundaries clear is what lets a blood draw do real work without promising more than the biology delivers.

---

References

1. Kwapisz D. The first liquid biopsy test approved. Is it a new era of mutation testing for non-small cell lung cancer? Annals of Translational Medicine. 2017. PMC5326656.

2. Fernández-Galán E, Puig-Butillé JA. Integrating ctDNA testing for EGFR analysis in advanced non-small cell lung cancer: strategies for clinical laboratories. Advances in Laboratory Medicine. 2025;6(3):233–244. DOI:10.1515/almed-2025-0012. PMID:40977812.

3. Li P, et al. ctDNA clearance predicts survival in unresectable EGFR-mutant NSCLC. Frontiers in Oncology. 2026. DOI:10.3389/fonc.2026.1743159.

4. Circulating Tumor DNA (ctDNA) Assay in Lung Cancer and EGFR Gene Mutation. ClinicalTrials.gov NCT06716580. 2024.