Elucidation of the Application of Blood Test Biomarkers to Predict Immune-Related Adverse Events in Atezolizumab-Treated NSCLC Patients Using Machine Learning Methods

The Multifaceted Role of Neutrophils in NSCLC in the Era of Immune Checkpoint Inhibitors

Lung cancer is the most common cause of cancer-related death in both males and females in the U.S. and non-small-cell lung cancer (NSCLC) accounts for 85%. Although the use of first- or second-line immune checkpoint inhibitors (ICIs) exhibits remarkable clinical benefits, resistance to ICIs develops over time and dampens the efficacy of ICIs in patients. Tumor-associated neutrophils (TANs) have an important role in modulating the tumor microenvironment (TME) and tumor immune response. The major challenge in the field is to characterize the TANs in NSCLC TME and understand the link between TAN-related immunosuppression with ICI treatment response. In this review, we summarize the current studies of neutrophil interaction with malignant cells, T-cells, and other components in the TME. Ongoing clinical trials are aimed at utilizing reagents that have putative effects on tumor-associated neutrophils, in combination with ICI. Elevated neutrophil populations and neutrophil-associated factors could be potential therapeutic targets to enhance anti-PD1 treatment in NSCLC.

Predictive value of NLR and PLR for immune-related adverse events: a systematic review and meta-analysis

BACKGROUND

Currently, there is a lack of affordable and accessible indicators that can accurately predict immune-related adverse events (irAEs) resulting from the use of immune checkpoint inhibitors (ICIs). In order to address this knowledge gap, our study explore the potential predictive value of two ratios, namely the neutrophil-lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR), for irAEs in cancer patients.

METHODS

A systematic search was performed in PubMed, Embase, and the Cochrane library. Studies involving NLR or PLR with irAEs were included. Quality and risk of bias of the selected studies were assessed. Forest plots were created based on Cox model analysis. Random effects meta-analyses were conducted to estimate odds ratio (OR) and its 95% confidence interval (CI).

RESULTS

After screening 594 studies, a total of 7 eligible studies with 1068 cancer patients were included. Analysis based on Cox regression showed that low neutrophil-lymphocyte ratio (L-NLR) (OR = 3.02, 95% CI 1.51 to 6.05, P = 0.002) and low platelet-lymphocyte ratio (L-PLR) (OR = 1.83, 95% CI 1.21 to 2.76, P = 0.004) were associated with irAEs. In the subgroup analysis of cut-off value, when the NLR cut-off value was 3, irAEs was significantly correlated with NLR (OR = 2.63, 95% CI 1.63 to 4.26, P < 0.001).

CONCLUSIONS

Both L-NLR and L-PLR have been found to be significantly associated with irAEs. Consequently, patients identified as being at a higher risk for irAEs should be subjected to more diligent monitoring and close observation.

Prediction of Effectiveness and Toxicities of Immune Checkpoint Inhibitors Using Real-World Patient Data

PURPOSE

Although immune checkpoint inhibitors (ICIs) have improved outcomes in certain patients with cancer, they can also cause life-threatening immunotoxicities. Predicting immunotoxicity risks alongside response could provide a personalized risk-benefit profile, inform therapeutic decision making, and improve clinical trial cohort selection. We aimed to build a machine learning (ML) framework using routine electronic health record (EHR) data to predict hepatitis, colitis, pneumonitis, and 1-year overall survival.

METHODS

Real-world EHR data of more than 2,200 patients treated with ICI through December 31, 2018, were used to develop predictive models. Using a prediction time point of ICI initiation, a 1-year prediction time window was applied to create binary labels for the four outcomes for each patient. Feature engineering involved aggregating laboratory measurements over appropriate time windows (60-365 days). Patients were randomly partitioned into training (80%) and test (20%) sets. Random forest classifiers were developed using a rigorous model development framework.

RESULTS

The patient cohort had a median age of 63 years and was 61.8% male. Patients predominantly had melanoma (37.8%), lung cancer (27.3%), or genitourinary cancer (16.4%). They were treated with PD-1 (60.4%), PD-L1 (9.0%), and CTLA-4 (19.7%) ICIs. Our models demonstrate reasonably strong performance, with AUCs of 0.739, 0.729, 0.755, and 0.752 for the pneumonitis, hepatitis, colitis, and 1-year overall survival models, respectively. Each model relies on an outcome-specific feature set, though some features are shared among models.

CONCLUSION

To our knowledge, this is the first ML solution that assesses individual ICI risk-benefit profiles based predominantly on routine structured EHR data. As such, use of our ML solution will not require additional data collection or documentation in the clinic.

Validation of a Machine Learning Model to Predict Immunotherapy Response in Head and Neck Squamous Cell Carcinoma

Head and neck squamous-cell carcinoma (HNSCC) is a disease with a generally poor prognosis; half of treated patients eventually develop recurrent and/or metastatic (R/M) disease. Patients with R/M HNSCC generally have incurable disease with a median survival of 10 to 15 months. Although immune-checkpoint blockade (ICB) has improved outcomes in patients with R/M HNSCC, identifying patients who are likely to benefit from ICB remains a challenge. Biomarkers in current clinical use include tumor mutational burden and immunohistochemistry for programmed death-ligand 1, both of which have only modest predictive power. Machine learning (ML) has the potential to aid in clinical decision-making as an approach to estimate a tumor's likelihood of response or a patient's likelihood of experiencing clinical benefit from therapies such as ICB. Previously, we described a random forest ML model that had value in predicting ICB response using 11 or 16 clinical, laboratory, and genomic features in a pan-cancer development cohort. However, its applicability to certain cancer types, such as HNSCC, has been unknown, due to a lack of cancer-type-specific validation. Here, we present the first validation of a random forest ML tool to predict the likelihood of ICB response in patients with R/M HNSCC. The tool had adequate predictive power for tumor response (area under the receiver operating characteristic curve = 0.65) and was able to stratify patients by overall (HR = 0.53 [95% CI 0.29-0.99], p = 0.045) and progression-free (HR = 0.49 [95% CI 0.27-0.87], p = 0.016) survival. The overall accuracy was 0.72. Our study validates an ML predictor in HNSCC, demonstrating promising performance in a novel cohort of patients. Further studies are needed to validate the generalizability of this algorithm in larger patient samples from additional multi-institutional contexts.

Cutaneous immune-related adverse events to immune checkpoint inhibitors: from underlying immunological mechanisms to multi-omics prediction

The development of immune checkpoint inhibitors (ICIs) has dramatically altered the landscape of therapy for multiple malignancies, including urothelial carcinoma, non-small cell lung cancer, melanoma and gastric cancer. As part of their anti-tumor properties, ICIs can enhance susceptibility to inflammatory side effects known as immune-related adverse events (irAEs), in which the skin is one of the most commonly and rapidly affected organs. Although numerous questions still remain unanswered, multi-omics technologies have shed light into immunological mechanisms, as well as the correlation between ICI-induced activation of immune systems and the incidence of cirAE (cutaneous irAEs). Therefore, we reviewed integrated biological layers of omics studies combined with clinical data for the prediction biomarkers of cirAEs based on skin pathogenesis. Here, we provide an overview of a spectrum of dermatological irAEs, discuss the pathogenesis of this "off-tumor toxicity" during ICI treatment, and summarize recently investigated biomarkers that may have predictive value for cirAEs via multi-omics approach. Finally, we demonstrate the prognostic significance of cirAEs for immune checkpoint blockades.

The artificial intelligence and machine learning in lung cancer immunotherapy

Since the past decades, more lung cancer patients have been experiencing lasting benefits from immunotherapy. It is imperative to accurately and intelligently select appropriate patients for immunotherapy or predict the immunotherapy efficacy. In recent years, machine learning (ML)-based artificial intelligence (AI) was developed in the area of medical-industrial convergence. AI can help model and predict medical information. A growing number of studies have combined radiology, pathology, genomics, proteomics data in order to predict the expression levels of programmed death-ligand 1 (PD-L1), tumor mutation burden (TMB) and tumor microenvironment (TME) in cancer patients or predict the likelihood of immunotherapy benefits and side effects. Finally, with the advancement of AI and ML, it is believed that "digital biopsy" can replace the traditional single assessment method to benefit more cancer patients and help clinical decision-making in the future. In this review, the applications of AI in PD-L1/TMB prediction, TME prediction and lung cancer immunotherapy are discussed.

Predictive Biomarkers for Checkpoint Inhibitor Immune-Related Adverse Events

Immune-checkpoint inhibitors (ICIs) are antagonists of inhibitory receptors in the immune system, such as the cytotoxic T-lymphocyte-associated antigen-4, the programmed cell death protein-1 and its ligand PD-L1, and they are increasingly used in cancer treatment. By blocking certain suppressive pathways, ICIs promote T-cell activation and antitumor activity but may induce so-called immune-related adverse events (irAEs), which mimic traditional autoimmune disorders. With the approval of more ICIs, irAE prediction has become a key factor in improving patient survival and quality of life. Several biomarkers have been described as potential irAE predictors, some of them are already available for clinical use and others are under development; examples include circulating blood cell counts and ratios, T-cell expansion and diversification, cytokines, autoantibodies and autoantigens, serum and other biological fluid proteins, human leucocyte antigen genotypes, genetic variations and gene profiles, microRNAs, and the gastrointestinal microbiome. Nevertheless, it is difficult to generalize the application of irAE biomarkers based on the current evidence because most studies have been retrospective, time-limited and restricted to a specific type of cancer, irAE or ICI. Long-term prospective cohorts and real-life studies are needed to assess the predictive capacity of different potential irAE biomarkers, regardless of the ICI type, organ involved or cancer site.