The IIP, defined by enriched intratumoral tumor-infiltrating lymphocytes (TIL), is a potential tumor-agnostic biomarker of responsiveness to ICI therapy. Here, we validate the IIP, as assessed by Lunit SCOPE IO, an AI-powered spatial TIL analyzer that runs on routine H&E-stained whole-slide images (WSI), for clinical outcome prediction in a large, multi-center international cohort of ICI-treated patients, demonstrating its utility as a practical biomarker to guide ICI treatment planning.
Lunit SCOPE IO was developed using 17,849 H&E WSI of multiple cancer types, annotated by 104 board-certified pathologists (13.5 x 109 µm2 area and 6.2 x 106 TIL). IIP+ tumors were defined as those with ≥ 20% of all 1 mm2 tumor tiles in a WSI classified as having a high intratumoral TIL density. We evaluated the correlation between IIP and ICI treatment outcomes (overall response rate (ORR) and progression-free survival (PFS), assessed by RECIST v1.1) in a real-world dataset of 1,806 patients (>16 primary tumor types) retrospectively collected from Stanford University Medical Center, Samsung Medical Center, Chonnam National University Hospital, Seoul National University Bundang Hospital, and Northwestern University. IIP status was sub-analyzed by PD-L1 22C3 tumor proportion score (TPS, n = 798), microsatellite status, and tumor mutational burden (TMB, n = 130).
The IIP+ phenotype (35.2%, 636 of 1,806) was highly enriched in nasopharyngeal carcinoma (68.0%), melanoma (56.3%), renal cell carcinoma (52.9%), and non-small cell lung cancer (NSCLC, 33.7%). The IIP+ proportion by PD-L1 TPS (< 1% / ≥ 1%) was 21.6% and 40.7%, respectively. While 33.3% of microsatellite unstable (MSI-H) or TMB-high (≥ 10/Mb) tumors were IIP+, a substantial proportion (26.1%) of microsatellite stable (MSS), TMB-low tumors were IIP+. The ORR in IIP+ patients was significantly higher (26.0% vs. 15.8% in IIP-, p < 0.001). Median PFS for IIP+ was 5.3 months (95% CI 4.6-6.9 m), significantly longer than that for IIP- (3.1 m, 95% CI 2.8-3.6 m), with a hazard ratio (HR) of 0.68 (95% CI 0.61-0.76, p < 0.001). The association held after excluding NSCLC patients (n = 909) (HR 0.69, 95% CI 0.59-0.81, p < 0.001). On subgroup analysis, IIP+ correlated significantly with prolonged PFS, regardless of ICI regimen (mono / combo therapy) or PD-L1 TPS (< 1% / ≥ 1%). Of note, IIP+ was predictive of favorable PFS only in the MSS/TMB-low group (n = 88, HR 0.56, 95% CI 0.33-0.96), but not in the MSI-H/TMB-high groups.
The IIP, as evaluated by Lunit SCOPE IO, may represent a practical, clinically-actionable biomarker predictive of favorable ICI treatment outcomes across diverse cancer patient populations, including those with PD-L1 negative, MSS/TMB-low tumors, in whom predictive biomarkers are urgently needed.
Jeanne Shen1,2, Yoon-La Choi3, Taebum Lee4, Hyojin Kim5, Young Kwang Chae6, Benjamin Dulken1, Stephanie Bogdan2, Maggie Huang7, George A. Fisher, Jr.8, Sehhoon Park9, Se-Hoon Lee9, Jun-Eul Hwang10, Jin-Haeng Chung5, Leeseul Kim11, Seunghwan Shin12, Yoojoo Lim12, Heon Song12, Sergio Pereira12, Chan-Young Ock12.
1Department of Pathology, Stanford University School of Medicine. 2Center for Artificial Intelligence in Medicine & Imaging, Stanford University. 3Department of Pathology and Translational Genomics, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. 4Department of Pathology, Chonnam National University Medical School, Republic of Korea. 5Department of Pathology, Seoul National University Bundang Hospital, Seongnam, Republic of Korea. 6Northewestern University. 7UC Davis Health
8Department of Medicine, Stanford University School of Medicine. 9Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. 10Division of Hematology-Oncology, Department of Internal Medicine, Chonnam National University Medical School, Republic of Korea. 11AMITA health Saint Francis Hospital Evanston, Evanston IL, United States. 12Lunit Inc., Seoul, Republic of Korea.
Abstract: Deep learning-based predictive biomarker for adjuvant chemotherapy in early-stage hormone receptor-positive breast cancer
Abstract: Pan-cancer analysis of tumor microenvironment using deep learning-based cancer stroma and immune profiling in H&E images
Abstract: Deep learning-based predictive biomarker for immune checkpoint inhibitor response in metastatic non-small cell lung cancer
Abstract: Comprehensive deep learning analysis of H&E tissue phenomics reveals distinct immune landscape and transcriptomic enrichment profile among immune inflamed, excluded and desert subtypes...
Abstract: Deep-learning analysis of H&E images to define three immune phenotypes to reveal loss-of-target in excluded immune cells as a novel resistance mechanism of immune checkpoint inhibitor...
Abstract: Deep learning-based immune phenotype analysis reveals distinct resistance pattern of immune checkpoint inhibitor in non-small cell lung cancer
Reducing Domain Gap by Reducing Style Bias
Predicting breast tumor proliferation from whole-slide images: The TUPAC16 challenge
Accuracy and efficiency of an artificial intelligence tool when counting breast mitoses
Abstract: Artificial intelligence-powered spatial analysis of tumor infiltrating lymphocytes (TIL) to reflect target gene expressions of novel immuno-oncology agents.
Abstract: Distinct subset of immune cells assessed by multiplex immunohistochemistry correlates with immune phenotype classified by ...
Abstract : Pathologic validation of artificial intelligence-powered prediction of combined positive score of PD-L1 immunohistochemistry in urothelial carcinoma.
Abstract : Clinical performance of artificial intelligence-powered annotation of tumor cell PD-L1 expression for treatment of immune-checkpoint inhibitor (ICI) ...
Abstract : Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes predicts survival after immune checkpoint inhibitor therapy across multiple cancer types.
Abstract : Deep learning based radiomic biomarker for predicting the presence of high-grade histologic patterns in lung adenocarcinoma
Abstract : Artificial intelligence-powered tissue analysis reveals distinct tumor-infiltrating lymphocyte profile as a potential biomarker of molecular subtypes in endometrial cancer
Abstract : Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes reveals distinct genomic profile of immune excluded phenotype in pan-carcinoma
Deep learning from HE slides predicts the clinical benefit from adjuvant chemotherapy in hormone receptor-positive breast cancer patients
Abstract : Interim results of phase I dose escalation study of YBL-006: A novel anti-PD-1 monoclonal antibody in advanced solid tumors
Abstract : Assistance with an artificial intelligence-powered PD-L1 analyzer reduces interobserver variation in pathologic reading of tumor proportion score in non-small cell lung cancer
Abstract : AI-powered whole-slide image analysis of tumor-infiltrating lymphocytes for prediction of prognosis in colorectal cancer
Deep Learning Analysis of CT Images Reveals High-Grade Pathological Features to Predict Survival in Lung Adenocarcinoma
Artificial Intelligence–Powered Spatial Analysis of Tumor-Infiltrating Lymphocytes as Complementary Biomarker for Immune Checkpoint Inhibition in Non–Small-Cell Lung Cancer
Artificial Intelligence–Powered Hematoxylin and Eosin Analyzer Reveals Distinct Immunologic and Mutational Profiles among Immune Phenotypes in Non–Small-Cell Lung Cancer
Abstract : Artificial Intelligence-Powered Analyzer Reduces Inter-observer Variation in PD-L1 Tumor Proportion Score of Non-Small Cell Lung Cancer
Abstract : Artificial Intelligence-Powered Tumor Purity Assessment From H&E Whole Slide Images Correlates...
Abstract : Deep learning-based H&E analyzer reveals distinct immune profiles and clinical outcomes among immune phenotypes in uterine corpus endometrial carcinoma
Abstract : Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes reveals immune-excluded phenotype is correlated with TGF-beta pathway related genomic aberrations
Observer Performance Study to Examine the Feasibility of the AI-powered PD-L1 Analyzer to Assist Pathologists’ Assessment of PD-L1 Expression Using Tumor Proportion Score in Non-Small Cell Lung Cancer
Artificial intelligence-powered human epidermal growth factor receptor 2 (HER2) analyzer in breast cancer as an assistance tool for pathologists to reduce interobserver variation
Artificial intelligence-powered whole-slide image analyzer reveals a distinctive distribution of tumor-infiltrating lymphocytes in neuroendocrine tumors and carcinomas
Artificial Intelligence (AI) - powered spatial analysis of tumor-infiltrating lymphocytes (TIL) for prediction of response to neoadjuvant chemotherapy (NAC) in triple negative breast cancer (TNBC)
Artificial intelligence (AI)-powered pathology image analysis merged with spatial transcriptomics reveals distinct TIGIT expression in the immune-excluded tumor-infiltrating lymphocytes
Trastuzumab plus FOLFOX for Gemcitabine/Cisplatin refractory HER2-positive biliary tract cancer: a multi-institutional phase II trial of the Korean Cancer Study Group (KCSG-HB19-14)
Safety and efficacy of YBL-006, an anti-PD-1 monoclonal antibody in advanced solid tumors: a phase I study
Robust artificial intelligence-powered imaging biomarker based on mammography for risk prediction of breast cancer.