Currently, predictive biomarkers for response to immune checkpoint inhibitor therapy in lung cancer are limited. To address this issue, researchers have developed a machine-learning-based tumour-infiltrating lymphocytes (TILs) scoring approach to evaluate TIL association with clinical outcomes in patients with advanced non–small cell lung cancer (NSCLC).1 The findings of this study were published in JAMA Oncology.
In recent years, the introduction of immune checkpoint inhibitors (ICIs) dramatically improved the outcome for patients with NSCLC. Unlike traditional chemotherapeutic agents, ICIs boost the body’s natural tumour-killing response by inhibiting immune checkpoint activation. ICIs have been shown to improve survival in advanced NSCLC, however, the response to immunotherapy varies between patients.2 Unfortunately, there is a lack of predictive biomarkers for a response to immune checkpoint inhibition in NSCLC.
A recent multicentre retrospective discovery-validation cohort study was performed using a cohort of 685 ICI-treated patients with NSCLC.1 Median follow-up was 38.1 months for the discovery (n=446) and 43.3 months for validation (n=239) cohort. A machine-learning automated method was developed to count tumour and stroma cells and tumour-infiltrating lymphocytes (TILs) in whole-slide haematoxylin-eosin-stained images of NSCLC tumours. Tumour mutational burden (TMB) and programmed death ligand-1 (PD-L1) expression were assessed separately. Objective response rate (ORR), progression-free survival (PFS), and overall survival (OS) were determined by blinded medical record review.
In a multivariable analysis, higher TIL levels (≥250 cells/mm2) were found to be associated with an improved PFS (HR: 0.71, p=0.006 and HR: 0.80, p=0.01) and OS (HR: 0.74, p=0.03 and HR: 0.75, p=0.001) in both the discovery as the validation cohort. A higher TIL level was associated with improved PFS and OS in both first- and subsequent-line ICI treatments in patients with NSCLC.
As such, these findings indicate that TIL levels in tumour tissue are robustly and independently associated with a response to ICI treatment in patients with NSCLC. Interestingly, this machine-learning automated TIL assessment is relatively easy to incorporate into the workflow of pathology laboratories at minimal additional cost. If these findings are validated, a TIL assessment could become part of the routine practice to predict ICI treatment effectiveness and enhance precision therapy.