The exploration of biomarkers in immunotherapy is constantly evolving, yet only a few have proven to effectively guide clinical decision-making. At the 2024 International Lung Cancer Conference (CLC 2024), held from August 9-11, Dr. Panwen Tian from West China Hospital of Sichuan University delivered a keynote speech titled "Thoughts and Exploration of Biomarkers in Lung Cancer Immunotherapy." This article provides a brief overview of the report.Predictive Models for the Efficacy of Lung Cancer Immunotherapy Combined with Chemotherapy
The success of individualized immunotherapy is influenced by numerous factors, including the host’s immune status, circulating immune cell characteristics, chronic viral infections, circulating cytokines, gut microbiota and its changes, and the use of immunosuppressive drugs. Positive predictive factors for the efficacy of immune checkpoint inhibitors (ICIs) include high tumor mutational burden (TMB), high PD-L1 expression, absence of p53 mutations, no use of steroids or antibiotics, no liver or pleural metastases, a neutrophil-to-lymphocyte ratio (NLR) <5, and an ECOG performance status (PS) <2.
Lung Cancer Immunotherapy Prognostic Index (LIPI): LIPI is a simple and low-cost biomarker for predicting the efficacy of immunotherapy. However, this model is only applicable to the efficacy of monotherapy (PD-1/PD-L1 inhibitors) in advanced non-small cell lung cancer (NSCLC).
Building a Clinicopathological Model: In terms of predicting the efficacy of ICI+chemotherapy, an analysis published by Dr. Panwen Tian and colleagues in Frontiers in Oncology in 2021 showed that the absence of bone metastasis, low dNLR, smoking history, and PD-L1 ≥50% were significantly associated with longer progression-free survival (PFS) and were independent factors influencing efficacy. The study compared a nomogram model based on clinicopathological factors with the LIPI model for predicting the efficacy of ICI+chemotherapy and found the former to be more effective.
Application of Deep Learning Combined with Radiomics in Lung Cancer Immunotherapy
PD-L1 expression is associated with the efficacy of immune checkpoint inhibitors. Since the 2019 V1 edition of the NCCN guidelines, PD-L1 testing has been upgraded from a 2A to a 1-level recommendation, holding equal status with EGFR, ALK, ROS1 testing. Results from several advanced NSCLC studies (KEYNOTE-024/042, KEYNOTE-189, CheckMate-057, IMpower 132, and IMpower 110) indicate that PD-L1 expression is correlated with the efficacy of immune checkpoint inhibitors. Compared to patients with low or negative PD-L1 expression, those with high PD-L1 expression are more likely to benefit from immunotherapy in advanced NSCLC, exhibiting higher objective response rates (ORR).
Limitations of PD-L1 as a Biomarker:
- Patient Selection: Some patients with positive PD-L1 expression do not respond to anti-PD-1 therapy, while some patients with negative PD-L1 expression show an objective response.
- Spatial and Temporal Heterogeneity: Multiple fresh biopsy samples from both primary and metastatic tumors are recommended to reduce the risk of errors; chemotherapy may affect PD-L1 expression.
- Variability in Testing Methods, Platforms, and Thresholds: There is some subjectivity in manual interpretation using IHC.
- Invasive Specimen Collection: Methods such as biopsy/surgery or EBUS-TBNA are invasive.
Optimizing PD-L1 as a Biomarker:
- Reducing the Difficulty of Specimen Collection: Non-invasive testing.
- Improving Interpretation Accuracy: AI-assisted PD-L1 expression diagnosis is more accurate and faster; deep learning combined with radiomics can be used to predict PD-L1 expression in NSCLC and the efficacy of immunotherapy (both monotherapy and immunotherapy combined with chemotherapy). It also allows for the visualization of PD-L1 expression.
Application of Multi-Omics Biomarkers in Lung Cancer Immunotherapy
TMB as a Predictor in Various Solid Tumors: A 2017 study published in the New England Journal of Medicine revealed a correlation between TMB and the response rate to PD-1 inhibitors, suggesting that TMB may serve as a predictor for various solid tumors. A study presented at the 2021 ASCO (#9018) found that the higher the TMB, the higher the ORR, PFS, and overall survival (OS) in patients receiving immunotherapy. A study published by Regan M Memmot and colleagues in Journal of Thoracic Oncology in 2021 showed that patients with high TMB had better outcomes with immunotherapy monotherapy, but the predictive value of TMB was limited in ICI+chemotherapy.
T-Cell Receptor (TCR) as a Predictor of Neoadjuvant ICI±Chemotherapy Efficacy: A Chinese study published in Clinical Cancer Research in 2020 found a positive correlation between intratumoral TCR clonality and pathological tumor regression after neoadjuvant immunotherapy in patients with stage I-IIIA NSCLC. Another study published in Clinical Cancer Research in 2021 showed that in stage IIIA NSCLC patients treated with neoadjuvant immunotherapy combined with chemotherapy, TCR was associated with complete pathological response post-surgery.
T-Cell Inflammation Gene Expression Profile (GEP) and ICI Efficacy: A study by Ott P and colleagues published in Journal of Clinical Oncology in 2018 analyzed tumor samples (N=313) and found that GEP scores were higher in patients who responded to treatment (Responders) and those with longer PFS. GEP scores were significantly correlated with ORR and PFS.
Multi-Omics as the Future of Biomarker Development: Numerous independent factors influence the efficacy of immunotherapy, and the exploration of combined multi-biomarkers is the way forward. Predicting the efficacy of immunotherapy using a single biomarker is akin to the story of “blind men touching an elephant”; the future of biomarkers lies in “high-throughput,” “modeling,” and “algorithmic” approaches.
Exploration of Biomarkers: High TMR (Tissue TMB + Tissue TCR) Predicts Immunotherapy Efficacy: A study conducted by Dr. Panwen Tian and colleagues (MedComm, 2024) found that compared to TMB, PD-L1, and integrated biomarkers such as TIGS or other gene expression biomarkers (including TIDE and IFNG), the TMR score demonstrated superior predictive performance. Patients with high TMR (TMB+TCR) who received immunotherapy had more significant tumor shrinkage; the higher the TMR score, the longer the mPFS; patients with high TMR scores had a higher CDR3 clone frequency, and CDR3 clones were significantly enriched in the high TMR group.
Conclusion: Models built on clinical and pathological features can predict the clinical efficacy of immunotherapy combined with chemotherapy in advanced NSCLC patients. AI-assisted diagnosis of PD-L1 expression is more accurate, faster, and visualized, and can predict the efficacy of ICI±chemotherapy. High-throughput, modeling, and multi-omics biomarkers are the future trend, with TMB+TCR showing potential for efficacy prediction.
Dr. Panwen Tian
- MD, Chief Physician, Doctoral Supervisor
- Deputy Director of the Lung Cancer Center, West China Hospital of Sichuan University
- Medical Team Leader, Department of Respiratory and Critical Care Medicine, West China Hospital of Sichuan University
- Member, Integrative Lung Cancer Committee, Chinese Anti-Cancer Association
- Member, Case Management Committee, Chinese Anti-Cancer Association
- Vice Chairman, Tumor Respiratory Diseases Committee, Sichuan Anti-Cancer Association
- Deputy Leader, Lung Cancer Group, Respiratory Committee, Sichuan Medical Association
- Participated in projects that won the Second Prize of the National Science and Technology Progress Award and the First Prize of the Sichuan Science and Technology Progress Award
- Principal Investigator of projects funded by the National Natural Science Foundation of China and key projects from the Sichuan Provincial Department of Science and Technology
