Dataset Information

Improved prediction of immune checkpoint blockade efficacy across multiple cancer types.

ABSTRACT: Only a fraction of patients with cancer respond to immune checkpoint blockade (ICB) treatment, but current decision-making procedures have limited accuracy. In this study, we developed a machine learning model to predict ICB response by integrating genomic, molecular, demographic and clinical data from a comprehensively curated cohort (MSK-IMPACT) with 1,479 patients treated with ICB across 16 different cancer types. In a retrospective analysis, the model achieved high sensitivity and specificity in predicting clinical response to immunotherapy and predicted both overall survival and progression-free survival in the test data across different cancer types. Our model significantly outperformed predictions based on tumor mutational burden, which was recently approved by the U.S. Food and Drug Administration for this purpose¹. Additionally, the model provides quantitative assessments of the model features that are most salient for the predictions. We anticipate that this approach will substantially improve clinical decision-making in immunotherapy and inform future interventions.

SUBMITTER: Chowell D

PROVIDER: S-EPMC9363980 | biostudies-literature | 2022 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Improved prediction of immune checkpoint blockade efficacy across multiple cancer types.

Chowell Diego D Yoo Seong-Keun SK Valero Cristina C Pastore Alessandro A Krishna Chirag C Lee Mark M Hoen Douglas D Shi Hongyu H Kelly Daniel W DW Patel Neal N Makarov Vladimir V Ma Xiaoxiao X Vuong Lynda L Sabio Erich Y EY Weiss Kate K Kuo Fengshen F Lenz Tobias L TL Samstein Robert M RM Riaz Nadeem N Adusumilli Prasad S PS Balachandran Vinod P VP Plitas George G Ari Hakimi A A Abdel-Wahab Omar O Shoushtari Alexander N AN Postow Michael A MA Motzer Robert J RJ Ladanyi Marc M Zehir Ahmet A Berger Michael F MF Gönen Mithat M Morris Luc G T LGT Weinhold Nils N Chan Timothy A TA

Nature biotechnology 20211101 4

Only a fraction of patients with cancer respond to immune checkpoint blockade (ICB) treatment, but current decision-making procedures have limited accuracy. In this study, we developed a machine learning model to predict ICB response by integrating genomic, molecular, demographic and clinical data from a comprehensively curated cohort (MSK-IMPACT) with 1,479 patients treated with ICB across 16 different cancer types. In a retrospective analysis, the model achieved high sensitivity and specificit ...[more]

PMID: 34725502

Similar Datasets

Project description:BackgroundsThe high morbidity and mortality of lung cancer are serious public health problems. The prognosis of lung cancer and whether to apply immune checkpoint blockade (ICB) are currently urgent problems to be solved.MethodsUsing R software, we performed Kaplan-Meier (K-M) analysis, Cox regression analysis, functional enrichment analysis, Spearman correlation analysis, and the single-sample gene set enrichment analysis.ResultsOn the Tumor IMmune Estimation Resource (TIMER2.0) website, we calculated the abundance of tumor-infiltrating immune cells (TIICs) of lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) patients. B cell and myeloid dendritic cell (DC1) were independent prognostic factors for LUAD and LUSC patients, respectively. Enrichment analysis confirmed that genes highly related to B cell or DC1 were closely related to the immune activation of lung cancer patients. In terms of adaptive immune resistance markers, CD8A, CD8B, immunomodulators (immunostimulants, major histocompatibility complex, receptors, and chemokines), immune-related pathways, tumor microenvironment score, and TIICs, high B cell/DC1 infiltration tissue was inflamed and immune-activated and might benefit more from the ICB. Genes most related to B cell [CD19, toll-like receptor 10 (TLR10), and Fc receptor-like A (FCRLA)] and DC1 (ITGB2, LAPTM5, and SLC7A7) partially clarified the roles of B cell/DC1 in predicting ICB efficacy. Among the 186 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, there were three and four KEGG pathways, which partially explained the molecular mechanisms by which B cell and DC1 simultaneously predicted the prognosis and efficacy of immunotherapy, respectively. Among five immune subtypes, the abundance of B cell/DC1 and expression of six hub genes were higher in immune C2, C3, and C6.ConclusionB cell and DC1 could predict the prognosis and ICB efficacy of LUAD and LUSC patients, respectively. The six hub genes and seven KEGG pathways might be novel immunotherapy targets. Immune C2, C3, and C6 subtypes of lung cancer patients might benefit more from ICB therapy.

Project description:Siglec-9, a cell surface transmembrane receptor mainly expressed on B cells, CD56+ NK cells, and CD4+ and CD8+ T cells, is strongly related to the tumor immune microenvironment. However, the expression pattern of Siglec-9 and its prognostic potential have not been investigated in a pan-cancer perspective. This study aimed to explore the association of Siglec-9 with prognosis, tumor stage, molecular subtype, and the immune microenvironment in pan-cancer. The mRNA expression of Siglec-9 was obtained from The Cancer Genome Atlas (TCGA), the Broad Institute Cancer Cell Line Encyclopedia (CCLE), and Genotype-Tissue Expression (GTEx). The relationship between Siglec-9 mRNA expression and prognosis was evaluated by the Kaplan-Meier analysis. The correlation between Siglec-9 and tumor-infiltrating immune cells, immune subtype, and molecular subtype was evaluated on Tumor Immune Estimation Resource (TIMER) and Integrated Repository Portal for Tumor-Immune System Interactions (TISIDB). The correlation between Siglec-9 expression and immune checkpoint, mismatch repair (MMR), DNA methyltransferase (DNMT), tumor mutation burden (TMB), and microsatellite instability (MSI) was also analyzed. It showed that Siglec-9 expression was significantly altered in most TCGA tumors. Siglec-9 expression was associated with the prognosis of patients with adrenocortical carcinoma (ACC), lung adenocarcinoma (LUSC), thymoma (THYM), colon adenocarcinoma (COAD), glioblastoma multiforme (GBM), prostate adenocarcinoma (PRAD), esophageal carcinoma (ESCA), and brain lower-grade glioma (LGG). Particularly, increased Siglec-9 expression was strongly correlated with poor prognosis in LGG. Correlation between Siglec-9 expression and tumor stage was also observed in various cancers. In addition, Siglec-9 was positively associated with infiltration of immune cells including neutrophils, dendritic cells (DCs), macrophage, and CD4+ and CD8+ T cells. Moreover, a significant correlation between Siglec-9 and MSI, TMB, MMR, DNMT, immune checkpoint, immune subtype, molecular subtype, and immunomodulators was observed in multiple cancers. Specifically, poor prognostic value and strong correlation to immune cell infiltration were verified with the LGG dataset from the Chinese Glioma Genome Atlas (CGGA). These findings indicated that Siglec-9 can be a novel biomarker and a potential target for cancer immunotherapy.

Dataset Information

Improved prediction of immune checkpoint blockade efficacy across multiple cancer types.

Publications

Improved prediction of immune checkpoint blockade efficacy across multiple cancer types.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets