Speaker's Highlight

• Don Sin University of British Columbia, St. Paul Hospital (Canada)
  Kenneth R. Chapman Toronto General Hospital Research Institute (Canada)
• Parameswaran Nair McMaster University (Canada)
  Carolyn Calfee UCSF (U.S.A.)
• Gregory P. Downey University of Colorado School of Medicine (U.S.A.)
  David A. Schwartz University of Colorado School of Medicine (U.S.A.)
• Neil Schluger Tuberculosis Control Branch, California Department of Public Health (U.S.A.)
  Nick Kim Critical Care & Sleep Medicine, University of California San Diego (U.S.A.)
• Nicola Hananiah Baylor College of Medicine (U.S.A.)
  Jae-Joon Yim Seoul National University College of Medicine (Republic of Korea)
• Koichiro Asano Tokai University School of Medicine (Japan)
  Diahn-Warng Perng Taipei Veterans General Hospital (Taiwan)
• Konstantinos Kostikas University of Ioannina (Greece)
  Karin Klooster University Medical Center Groningen (Kingdom of the Netherlands)

• Lung cancer immunotherapy: predictive biomarkers

  Immunotherapy has changed the treatment landscape for advanced non-small cell lung cancer (NSCLC), particularly through immune checkpoint inhibitors that target PD-1, PD-L1, or CTLA-4 pathways. Yet response varies widely. Some patients experience durable disease control, while others show primary resistance or relapse after an initial benefit.

  Predictive biomarkers help clinicians estimate whether a treatment is likely to work for an individual patient. They differ from prognostic markers, which describe the expected course of a disease regardless of therapy. In lung cancer, biomarker interpretation increasingly combines tissue pathology, genomic profiling, clinical features, and changes observed during treatment.

  The subject is especially relevant at respiratory medicine congresses, where oncologists, pulmonologists, pathologists, and researchers examine how laboratory findings can guide precision oncology. A useful framework must address both established tests and promising markers that still require prospective validation.

  Why biomarker selection matters

  The tumor immune microenvironment is complex. Cancer cells may express immune checkpoint proteins, conceal themselves through genetic changes, or create a locally immunosuppressive environment. Consequently, a single laboratory result rarely captures every factor influencing treatment response.

  Biomarker-guided decisions can support the choice between immunotherapy alone, chemotherapy-immunotherapy combinations, targeted treatment, or a clinical trial. They can also help identify patients at risk of immune-related adverse events, although toxicity prediction remains less mature than response prediction.

  Testing is most valuable when it is performed before treatment begins and interpreted alongside histology, disease stage, smoking history, performance status, comorbidities, and previous therapies. A result should inform a clinical decision rather than replace multidisciplinary judgment.

  PD-L1 expression and its limitations

  PD-L1 immunohistochemistry is currently the most widely used predictive test for checkpoint inhibitor therapy in NSCLC. The tumor proportion score (TPS) estimates the percentage of viable tumor cells showing membranous PD-L1 staining. Higher expression, particularly a TPS of 50% or more, is associated with a greater probability of response to some first-line monotherapy regimens.

  However, PD-L1 is an imperfect marker. Expression can vary between primary and metastatic sites, change over time, and be affected by prior treatment. Small biopsies may fail to represent the full tumor, while differences among antibody clones, platforms, and laboratory procedures can influence results.

  A low or absent PD-L1 score does not necessarily exclude benefit from immunotherapy, especially when checkpoint inhibition is combined with chemotherapy. Conversely, high expression does not guarantee response. Pathologists should report the assay, specimen quality, scoring method, and percentage clearly so the oncology team can interpret the finding appropriately.

  Genomic features that shape response

  Tumor mutational burden (TMB) measures the number of somatic mutations within a defined amount of tumor DNA. A higher mutational load may generate more neoantigens, potentially making malignant cells easier for T cells to recognize. Although TMB has biological appeal, differences in sequencing panels, cutoffs, sample quality, and analytic methods have limited its role as a universal standalone marker in lung cancer.

  Driver alterations are equally important because they influence treatment sequencing. EGFR mutations and ALK rearrangements are generally associated with lower response rates to immune checkpoint inhibitor monotherapy than smoking-related, mutation-rich tumors. Patients with these alterations typically require an approved targeted therapy strategy, with immunotherapy considered carefully in later treatment settings.

  Other genomic changes, including STK11/LKB1 and KEAP1 alterations, have been associated with an immunologically “cold” tumor environment and reduced benefit in some studies. These signals are clinically interesting but should not automatically determine treatment outside validated decision frameworks. Broad next-generation sequencing can reveal actionable alterations and prevent premature use of immunotherapy.

  Comparing established and emerging markers

  No biomarker should be viewed in isolation. PD-L1 offers practical clinical value, while genomic and immune-microenvironment measures may add context. Blood-based markers are attractive because they can be repeated, but many remain investigational.

  Biomarker	What it measures	Potential clinical value	Main limitations
  PD-L1 TPS	PD-L1 on tumor cells	Helps select some monotherapy approaches	Spatial variation, assay differences, imperfect sensitivity
  Tumor mutational burden	Number of somatic mutations	May reflect neoantigen potential	Platform-dependent cutoffs and inconsistent validation
  EGFR or ALK status	Oncogenic driver alterations	Guides targeted therapy and treatment sequencing	Does not directly predict every immunotherapy outcome
  STK11 or KEAP1 alterations	Tumor suppressor and stress-response pathways	May identify resistance-associated biology	Not yet a universal treatment-selection test
  Tumor-infiltrating lymphocytes	Immune-cell presence in the tumor	Describes immune engagement	Sampling and scoring are not standardized
  Circulating tumor DNA	Tumor-derived DNA fragments in blood	Supports molecular monitoring and residual disease research	Variable shedding and limited routine validation

  Blood-based and dynamic biomarkers

  Circulating tumor DNA (ctDNA) is being studied as a minimally invasive way to detect molecular response, emerging resistance, and minimal residual disease. A rapid decline in ctDNA after treatment may correlate with tumor shrinkage, while persistent or rising levels can signal residual disease before imaging changes become obvious.

  Other blood-based candidates include circulating immune cells, cytokine patterns, cell-free DNA characteristics, and circulating tumor cells. These approaches could make biomarker assessment more dynamic than a single pretreatment biopsy. They may also be useful when tissue is scarce or when repeated sampling is impractical.

  At present, most liquid biopsy findings should be interpreted as complementary rather than definitive for immunotherapy selection. A negative ctDNA result may reflect low tumor shedding rather than the absence of cancer. Standardized assays, validated thresholds, and prospective trials are needed before routine adoption becomes widespread.

  Integrating biomarkers into clinical decisions

  A reliable workflow begins with adequate tissue acquisition and pathology confirmation. Molecular testing should cover guideline-relevant driver genes, while PD-L1 testing should use an assay appropriate to the intended treatment. When tissue is insufficient, a validated plasma genomic test can sometimes provide important information, although a negative plasma result may need tissue confirmation.

  Clinicians should also consider immune-related contraindications and the possibility of overlapping toxicity with later therapies. For example, the sequence of immunotherapy and certain targeted treatments may affect the risk of adverse events. Biomarkers therefore belong within a broader treatment plan that includes medication history, organ function, patient preferences, and access to specialist monitoring.

  The COVID-19 health notice also illustrates the wider role of respiratory medicine congresses: clinical knowledge develops alongside safe participation, infection-control awareness, and communication among healthcare professionals.

  Practical priorities for multidisciplinary teams

  A balanced biomarker strategy can improve consistency while avoiding overinterpretation of preliminary findings. Teams involved in lung cancer care should:

  • Confirm histology, stage, specimen adequacy, and the timing of each test.
  • Use validated PD-L1 immunohistochemistry and document the assay and scoring method.
  • Complete broad molecular profiling before selecting immunotherapy when an actionable driver is possible.
  • Treat TMB, STK11, KEAP1, immune signatures, and ctDNA as context-dependent unless supported by validated evidence.
  • Reassess disease biology at progression when a new biopsy or liquid biopsy could change management.

  Research is moving toward composite biomarkers that combine PD-L1, genomic alterations, T-cell activity, ctDNA kinetics, and clinical characteristics. Machine-learning models may eventually integrate these data, but transparent validation and equitable access will be essential.

  Attend respiratory oncology discussions, review current evidence, and bring biomarker interpretation into every multidisciplinary lung cancer pathway. Careful testing and disciplined clinical judgment can help ensure that immunotherapy is offered to the patients most likely to benefit while avoiding decisions based on a single incomplete signal.

  Richard Russell Nuffield Department of Clinical Medicine, University of Oxford (United Kingdom)
• Mona Bafadhel King’s College London (United Kingdom)
  David Jackson Guy’s and St Thomas’ Hospital, King’s College London (United Kingdom)
• James Chalmers University of Dundee (United Kingdom)
  David Price University of Aberdeen (United Kingdom)