Skip to content

Lung adenocarcinoma — biomarkers

TL;DR — Lung-adenocarcinoma biomarkers perform different jobs: driver variants select targeted therapy, PD-L1 estimates average checkpoint-monotherapy sensitivity, ctDNA can reveal genotype or molecular residual disease, and histology/stage remain independent anatomic and morphologic predictors. A biomarker is not clinically useful merely because it correlates with outcome; it needs an analytically defined assay, cutoff, intended use, comparator, and evidence that acting on the result improves decisions. Broad tissue NGS plus RNA fusion detection and complementary plasma testing is the current diagnostic backbone (Lindeman 2018, PMID 29398453; Rolfo 2021, PMID 34246791). PD-L1 is imperfect but actionable; tumor mutational burden (TMB), STK11/KEAP1 status, and immune-expression signatures remain contextual rather than universal selectors (Reck 2016, PMID 27718847; Skoulidis 2018, PMID 29773717). Post-treatment ctDNA strongly predicts recurrence, but a live PubMed search on 2026-08-29 found intervention protocols rather than mature randomized survival results for MRD-guided action (cohort PMIDs: 28899864, 28445469; protocol PMIDs: 39659920, 41628935).

Biomarker taxonomy

Type Question answered Adenocarcinoma example Common misuse
Diagnostic What is the lesion? TTF-1/Napsin A-supported pulmonary adenocarcinoma Treating marker positivity as proof of lung origin
Predictive Does treatment effect differ by marker? EGFR mutation for EGFR TKI Calling any response correlation predictive
Prognostic What outcome is likely irrespective of treatment? Stage, ctDNA positivity after surgery Assuming it identifies an effective intervention
Pharmacodynamic Is the drug affecting tumor burden/pathway? Early ctDNA fall Switching therapy without validated threshold
Resistance Why did benefit end? EGFR C797S, MET amplification Assuming one biopsy captures every resistant clone
Surveillance Is occult disease returning? Tumor-informed postoperative ctDNA Treating a negative test as proof of cure

Actionable genomic drivers

Biomarker Required resolution Matched-therapy consequence
EGFR Exact exon/variant: exon 19 deletion, L858R, exon 20 insertion, uncommon mutation Different inhibitor evidence and dose sensitivity
ALK/ROS1/RET/NTRK Functional fusion, partner and reading frame where available Fusion-directed CNS-active TKI
KRAS Exact codon/allele; co-mutations G12C-specific inhibitors; non-G12C trial routing
BRAF V600E versus class II/III BRAF–MEK doublet only established for V600E
MET Exon-14 skipping versus amplification; copy number/focality Selective MET TKI for METex14
ERBB2/HER2 Activating mutation versus amplification/IHC Mutation-directed ADC/TKI evidence

TCGA defined recurrently altered pathways but was not a prospective treatment-selection trial (TCGA 2014, PMID 25079552). Clinical actionability requires trial evidence for the exact alteration state.

Tissue and plasma complementarity

Tissue provides morphology, tumor content, immune context, copy-number structure, and RNA. Plasma samples the whole-body shedding compartment and can return results quickly, but sensitivity falls with low burden, thorax-confined disease, indolent biology, and CNS-only disease (Leighl 2019, PMID 30988079; Rolfo 2021, PMID 34246791).

Result pattern Interpretation Next step
Tissue positive, plasma negative Driver present; low shedding likely Treat based on tissue if analytically valid
Plasma positive, tissue insufficient Actionable if variant is tumor-derived and assay validated Consider tissue for morphology and transformation
Both negative with adequate tissue Lower probability of known driver Confirm RNA fusion coverage and panel breadth
Discordant variants Heterogeneity, clonal hematopoiesis, assay artifact, or evolution Review VAF, genes, specimen timing, and orthogonal evidence
Post-treatment new TP53/DNMT3A/TET2 Could be clonal hematopoiesis Matched leukocyte sequencing helps avoid false tumor assignment

Plasma NGS prospectively increased detection of actionable biomarkers when added to tissue and returned results within clinically useful timelines (Leighl 2019, PMID 30988079; Pritchett 2019, PMID 32914040).

PD-L1

PD-L1 TPS is the proportion of viable tumor cells with membranous staining. KEYNOTE-024 established predictive clinical utility at TPS ≥50% in EGFR/ALK-negative NSCLC (Reck 2016, PMID 27718847). KEYNOTE-189 showed that chemo-immunotherapy benefits extend across PD-L1 strata (Gandhi 2018, PMID 29658856).

Source of variation Effect
Assay clone/platform Cutoffs are not automatically interchangeable
Small biopsy/cytology Sampling error when expression is heterogeneous
Primary versus metastasis Spatial discordance
Prior radiation/systemic therapy Dynamic expression
Necrosis/low viable tumor Unreliable denominator
Driver-positive biology High TPS may not predict the same benefit as driver-negative disease

PD-L1 is a continuous, imperfect probability marker. It should not be described as “positive/negative” without assay, cutoff, specimen, and clinical context.

Tumor mutational burden

TMB counts somatic mutations per sequenced megabase after filtering. Whole-exome and panel TMB are not numerically interchangeable; panel size, gene content, germline filtering, tumor purity, and sequencing depth change values.

CheckMate 227 made TMB a major research biomarker, but long-term nivolumab–ipilimumab benefit was not confined to a universal TMB-defined population and TMB did not become a mandatory NSCLC selector (Paz-Ares 2022, PMID 34648948).

KEAP1-mutant adenocarcinoma can have high TMB and poor immune response, showing that mutation quantity does not encode clonality, antigen presentation, T-cell access, or myeloid suppression (Marinelli 2020, PMID 32866624).

STK11, KEAP1, and TP53

Marker Evidence Current use boundary
STK11/LKB1 Associated with PD-1 resistance in KRAS-mutant adenocarcinoma (PMID 29773717) Randomized POSEIDON subgroup data support a CTLA-4-containing hypothesis, not a stand-alone licensed selector (PMID 39385035)
KEAP1/NFE2L2 Redox/immune-excluded phenotype and poor outcome (PMIDs: 32866624, 36526124) Prospective biomarker-specific validation remains incomplete despite POSEIDON subgroup evidence (PMID 39385035)
TP53 Often inflamed with KRAS; also chromosomal instability Contextual, not a regimen selector
SMARCA4 Aggressive biology in selected subsets Histologic and genomic context required

Retrospective treatment cohorts are vulnerable to immortal-time bias, treatment-selection bias, correlated genotypes, and incomplete adjustment. A marker becomes predictive through a treatment-by-marker interaction, ideally prospectively tested.

ctDNA for response and resistance

Serial ctDNA can show molecular response before imaging, identify emergent resistance, and reveal multiple progressing clones. Its absence may mean response or merely low shedding.

Setting Potential use Evidence gap
Metastatic baseline Faster driver detection Negative result needs tissue fallback
Early on-treatment Molecular response No universal time point or clearance threshold
Radiographic progression Resistance genotype Tissue still needed for transformation
Oligoprogression Distinguish limited versus systemic molecular escape Low-volume disease may not shed
CNS-only progression Detect resistance Plasma sensitivity is particularly limited

MRD after curative-intent therapy

Chaudhuri and colleagues detected post-treatment molecular residual disease before radiographic recurrence in localized lung cancer using CAPP-Seq (PMID 28899864). TRACERx used phylogenetic, tumor-informed ctDNA to track subclonal relapse (Abbosh 2017, PMID 28445469; TRACERx dissemination analysis, PMID 37055640).

Prospective cohorts link postoperative ctDNA positivity with recurrence and suggest dynamic risk classification (LUNGCA-1, PMID 34844976; Qiu 2021, PMID 34799585).

Requirement for clinical utility Why it matters
Defined sampling window Surgery releases DNA; timing changes sensitivity/specificity
Tumor-informed versus fixed panel Trade-off between personalization, speed, and coverage
Matched leukocyte control Prevents clonal-hematopoiesis false positives
Prospective action rule A prognostic signal alone does not improve outcome
Effective intervention Earlier detection helps only if earlier treatment helps
False-negative management Low shedding cannot justify unsafe de-escalation

Histologic and imaging biomarkers

IASLC grade combines predominant pattern with the presence of ≥20% high-grade patterns; external studies confirm prognostic separation, but interobserver reproducibility and treatment interaction remain limitations (Rokutan-Kurata 2021, PMID 33905897; meta-analysis, PMID 38485464).

STAS associates with recurrence, especially after limited resection, but remains a histologic descriptor rather than a TNM category (Travis 2024, PMID 38508515; Chen 2019, PMID 30914285).

Radiomics and AI can predict grade, genotype, or outcome retrospectively, but discrimination does not prove calibration or clinical utility. External, prospective, multi-scanner validation and a defined decision threshold are prerequisites (ANORAK, PMID 38200244).

Emerging multi-omic biomarkers

Single-cell and spatial atlases reveal malignant-cell states, fibroblast programs, tertiary lymphoid structures, exhausted T cells, and macrophage niches (Lavin 2017, PMID 28475900; Sinjab 2021, PMID 33972311). These maps generate composite hypotheses but do not yet provide routine assays.

TRACERx shows that subclonal selection and chromosomal instability shape recurrence and treatment response (Frankell 2023, PMID 37046096). A small biopsy may therefore measure only one branch of an evolutionary tree.

Biomarker reporting checklist

  1. Intended use: diagnosis, selection, prognosis, monitoring, or resistance.
  2. Specimen, date, site, processing, tumor fraction, and assay.
  3. Exact analyte and unit; avoid ambiguous “positive.”
  4. Predefined cutoff and evidence source.
  5. Sensitivity limits and non-shedding/low-content interpretation.
  6. Competing explanation such as clonal hematopoiesis.
  7. Treatment action supported by prospective evidence.
  8. Version date because panels, labels, and guidelines change.

Open questions

  • Can MRD-guided escalation or de-escalation improve overall survival rather than only risk classification?
  • Which composite immune biomarker adds reproducible value beyond PD-L1?
  • Are STK11 and KEAP1 treatment-predictive or mainly prognostic across regimens?
  • How should tissue and plasma discordance be adjudicated when both assays are analytically valid?
  • Can spatial/single-cell states be reduced to robust clinical assays without losing biological meaning?
  • Which imaging biomarker improves a real decision rather than only retrospective AUC?

References

  1. Lindeman NI, et al. Updated Molecular Testing Guideline for the Selection of Lung Cancer Patients for Treatment With Targeted Tyrosine Kinase Inhibitors: Guideline From the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. J Mol Diagn. 2018. PMID 29398453
  2. Rolfo C, et al. Liquid Biopsy for Advanced NSCLC: A Consensus Statement From the International Association for the Study of Lung Cancer. J Thorac Oncol. 2021. PMID 34246791
  3. Leighl NB, et al. Clinical Utility of Comprehensive Cell-free DNA Analysis to Identify Genomic Biomarkers in Patients with Newly Diagnosed Metastatic Non-small Cell Lung Cancer. Clin Cancer Res. 2019. PMID 30988079
  4. Pritchett MA, et al. Prospective Clinical Validation of the InVisionFirst-Lung Circulating Tumor DNA Assay for Molecular Profiling of Patients With Advanced Nonsquamous Non-Small-Cell Lung Cancer. JCO Precis Oncol. 2019. PMID 32914040
  5. Reck M, et al. Pembrolizumab versus Chemotherapy for PD-L1-Positive Non-Small-Cell Lung Cancer. N Engl J Med. 2016. PMID 27718847
  6. Gandhi L, et al. Pembrolizumab plus Chemotherapy in Metastatic Non-Small-Cell Lung Cancer. N Engl J Med. 2018. PMID 29658856
  7. Paz-Ares LG, et al. First-Line Nivolumab Plus Ipilimumab in Advanced NSCLC: 4-Year Outcomes From the Randomized, Open-Label, Phase 3 CheckMate 227 Part 1 Trial. J Thorac Oncol. 2022. PMID 34648948
  8. Skoulidis F, et al. STK11/LKB1 Mutations and PD-1 Inhibitor Resistance in KRAS-Mutant Lung Adenocarcinoma. Cancer Discov. 2018. PMID 29773717
  9. Ricciuti B, et al. Diminished Efficacy of Programmed Death-(Ligand)1 Inhibition in STK11- and KEAP1-Mutant Lung Adenocarcinoma Is Affected by KRAS Mutation Status. J Thorac Oncol. 2022. PMID 34740862
  10. Marinelli D, et al. KEAP1-driven co-mutations in lung adenocarcinoma unresponsive to immunotherapy despite high tumor mutational burden. Ann Oncol. 2020. PMID 32866624
  11. Scalera S, et al. Clonal KEAP1 mutations with loss of heterozygosity share reduced immunotherapy efficacy and low immune cell infiltration in lung adenocarcinoma. Ann Oncol. 2023. PMID 36526124
  12. Chaudhuri AA, et al. Early Detection of Molecular Residual Disease in Localized Lung Cancer by Circulating Tumor DNA Profiling. Cancer Discov. 2017. PMID 28899864
  13. Abbosh C, et al. Phylogenetic ctDNA analysis depicts early-stage lung cancer evolution. Nature. 2017. PMID 28445469
  14. Abbosh C, et al. Tracking early lung cancer metastatic dissemination in TRACERx using ctDNA. Nature. 2023. PMID 37055640
  15. Xia L, et al. Perioperative ctDNA-Based Molecular Residual Disease Detection for Non-Small Cell Lung Cancer: A Prospective Multicenter Cohort Study (LUNGCA-1). Clin Cancer Res. 2022. PMID 34844976
  16. Qiu B, et al. Dynamic recurrence risk and adjuvant chemotherapy benefit prediction by ctDNA in resected NSCLC. Nat Commun. 2021. PMID 34799585
  17. Cancer Genome Atlas Research Network, et al. Comprehensive molecular profiling of lung adenocarcinoma. Nature. 2014. PMID 25079552
  18. Frankell AM, et al. The evolution of lung cancer and impact of subclonal selection in TRACERx. Nature. 2023. PMID 37046096
  19. Lavin Y, et al. Innate Immune Landscape in Early Lung Adenocarcinoma by Paired Single-Cell Analyses. Cell. 2017. PMID 28475900
  20. Sinjab A, et al. Resolving the Spatial and Cellular Architecture of Lung Adenocarcinoma by Multiregion Single-Cell Sequencing. Cancer Discov. 2021. PMID 33972311
  21. Rokutan-Kurata M, et al. Validation Study of the International Association for the Study of Lung Cancer Histologic Grading System of Invasive Lung Adenocarcinoma. J Thorac Oncol. 2021. PMID 33905897
  22. Hegedűs F, et al. Prognostic impact of the IASLC grading system of lung adenocarcinoma: a systematic review and meta-analysis. Histopathology. 2024. PMID 38485464
  23. Travis WD, et al. The International Association for the Study of Lung Cancer (IASLC) Staging Project for Lung Cancer: Recommendation to Introduce Spread Through Air Spaces as a Histologic Descriptor in the Ninth Edition of the TNM Classification of Lung Cancer. Analysis of 4061 Pathologic Stage I NSCLC. J Thorac Oncol. 2024. PMID 38508515
  24. Chen D, et al. Tumor Spread Through Air Spaces in Non-Small Cell Lung Cancer: A Systematic Review and Meta-Analysis. Ann Thorac Surg. 2019. PMID 30914285
  25. Pan X, et al. The artificial intelligence-based model ANORAK improves histopathological grading of lung adenocarcinoma. Nat Cancer. 2024. PMID 38200244
  26. Skoulidis F, et al. CTLA4 blockade abrogates KEAP1/STK11-related resistance to PD-(L)1 inhibitors. Nature. 2024. PMID 39385035
  27. Wang K, et al. Adjuvant osimertinib therapy guided by ctDNA-assessed MRD in resected EGFR-mutated stage IA-IIA non-small-cell lung cancer: a randomized clinical trial study protocol. Am J Cancer Res. 2024. PMID 39659920
  28. Andersen ME, et al. ctDNA guided immunotherapy in patients with advanced non-small cell lung cancer: a nationwide Danish, randomised, intervention study (PRELUCA) — study protocol. BMJ Open. 2026. PMID 41628935