Association of Machine Learning–Based Assessment of Tumor-Infiltrating Lymphocytes on Standard Histologic Images With Outcomes of Immunotherapy in Patients With NSCLC

Abstract

ImportanceCurrently, predictive biomarkers for response to immune checkpoint inhibitor (ICI) therapy in lung cancer are limited. Identifying such biomarkers would be useful to refine patient selection and guide precision therapy.ObjectiveTo develop a machine-learning (ML)-based tumor-infiltrating lymphocytes (TILs) scoring approach, and to evaluate TIL association with clinical outcomes in patients with advanced non–small cell lung cancer (NSCLC).Design, Setting, and ParticipantsThis multicenter retrospective discovery-validation cohort study included 685 ICI-treated patients with NSCLC with median follow-up of 38.1 and 43.3 months for the discovery (n = 446) and validation (n = 239) cohorts, respectively. Patients were treated between February 2014 and September 2021. We developed an ML automated method to count tumor, stroma, and TIL cells in whole-slide hematoxylin-eosin–stained images of NSCLC tumors. Tumor mutational burden (TMB) and programmed death ligand-1 (PD-L1) expression were assessed separately, and clinical response to ICI therapy was determined by medical record review. Data analysis was performed from June 2021 to April 2022.ExposuresAll patients received anti–PD-(L)1 monotherapy.Main Outcomes and MeasuresObjective response rate (ORR), progression-free survival (PFS), and overall survival (OS) were determined by blinded medical record review. The area under curve (AUC) of TIL levels, TMB, and PD-L1 in predicting ICI response were calculated using ORR.ResultsOverall, there were 248 (56%) women in the discovery cohort and 97 (41%) in the validation cohort. In a multivariable analysis, high TIL level (≥250 cells/mm2) was independently associated with ICI response in both the discovery (PFS: HR, 0.71; P = .006; OS: HR, 0.74; P = .03) and validation (PFS: HR = 0.80; P = .01; OS: HR = 0.75; P = .001) cohorts. Survival benefit was seen in both first- and subsequent-line ICI treatments in patients with NSCLC. In the discovery cohort, the combined models of TILs/PD-L1 or TMB/PD-L1 had additional specificity in differentiating ICI responders compared with PD-L1 alone. In the PD-L1 negative (<1%) subgroup, TIL levels had superior classification accuracy for ICI response (AUC = 0.77) compared with TMB (AUC = 0.65).Conclusions and RelevanceIn these cohorts, TIL levels were robustly and independently associated with response to ICI treatment. Patient TIL assessment is relatively easily incorporated into the workflow of pathology laboratories at minimal additional cost, and may enhance precision therapy.

Prognostic and immunological characteristics of CDK1 in lung adenocarcinoma: A systematic analysis

“…The GSVA results revealed that CDK1 also plays a role in immune regulation, especially in T cells. Tumor-infiltrating lymphocytes have been reported as a predictor of disease progression and to be related to the efficacy of immunotherapy in non-small cell lung cancer (NSCLC) patients ( 17 , 18 ). Here, we next performed a comprehensive analysis to investigate the relationship between CDK1 expression and tumor immune microenvironment factors, including lymphocytes, checkpoints, MHC molecules, chemokines and receptors.…”

Tumor Infiltrating Lymphocytes across Breast Cancer Subtypes: Current Issues for Biomarker Assessment

“…Furthermore, the analytical validity may be implemented through web-based platforms aiming at enhancing the interaction among pathologist and increasing the concordance of morphological assessment, as demonstrated in TONIC trial (NCT02499367), in which concordance values between four pathologists were >90%; and through the development of operator-independent approaches (e.g., machine-learningbased) [71][72][73].…”

“…For example, a meta-analysis of 22 studies, including 1569 patients with early BC who received TIL dynamic assessment during NACT, demonstrated that an increase in TILs in TNBC was associated with better DFS in univariate analysis (HR: 0.59; 95% CI: 0.37-0.95; p = 0.03) [74]. Furthermore, TILs levels were assessed in patients with TNBC who had Furthermore, the analytical validity may be implemented through web-based platforms aiming at enhancing the interaction among pathologist and increasing the concordance of morphological assessment, as demonstrated in TONIC trial (NCT02499367), in which concordance values between four pathologists were >90%; and through the development of operator-independent approaches (e.g., machine-learning-based) [71][72][73].…”

An Algorithm as a Biomarker for Response to Immune Checkpoint Inhibitor Therapy

“…A negotiating strategy that leaders use is to ask, “What would it take?” In light of the weak predictive biomarkers used for immune checkpoint inhibitor (ICI) therapy, it may be time to ask the pathologist “What would it take to provide us with a quantitative tumor infiltrating lymphocyte (TIL) score for each patient with non–small-cell lung cancer?” In this issue of JAMA Oncology , Rakaee et al describe a new biomarker associated with response to ICI based on an objective, open-source, machine learning (ML) method using QuPath software to count TILs in lung cancer. They note that TIL has potential to improve selection of responders to ICI and that ML-TIL is easily obtained, accurate, and reproducible.…”

“…Clearly 1 or 2 were not enough, or we (pathology laboratories) would all be digital. Perhaps the sorts of assays like that described by Rakaee and colleagues, combined with similar either open-source or proprietary assays, will eventually provide sufficient improvement in patient care or reimbursement to trigger broad adoption.…”