Abstract 228P
Background
Clinical profiling studies have shed light on molecular features and mechanisms that modulate response or resistance to immunotherapy but their predictive value remains largely unclear. We (Bareche et al., Annals of Oncology 2022 ) and others (Litchfield et al., Cell 2021 ) have recently curated a compendium of public datasets of DNA, RNA and clinical profiles of patients treated with immunotherapy.
Methods
Leveraging our compendium of immunotherapy clinical datasets, we developed, PredictIO, an open-source meta-analysis pipeline to assess the predictive value of molecular predictors. We first used PredictIO to compute the association between immunotherapy response and established biomarkers, such as tumor mutation burden (TNB) or CD8 gene expression, and a collection of 91 molecular signatures curated from the literature. Second, we used PredictIO for de novo RNA signature discovery pipeline to build a new predictor of immunotherapy response.
Results
Using molecular and clinical profiles of ∼3600 patients across 12 tumor types, our meta-analysis pipeline revealed thatTMB and ∼50% of the gene signatures were significantly predictive of immunotherapy response across tumor types, although their predictive value were strongly dependent on specific tumour types. We next developed a de novo gene expression signature from our pan-cancer analysis and demonstrated its superior predictive value over other biomarkers. To identify novel targets, we computed the T-cell dysfunction score for each gene within PredictIO and their ability to predict dual PD-1/CTLA-4 blockade in mice. Two genes, F2RL1 and RBFOX2, were concurrently associated with worse ICB clinical outcomes, T-cell dysfunction in ICB-naive patients and resistance to dual PD-1/CTLA-4 blockade in preclinical models.
Conclusions
Our study highlights the potential of large-scale meta-analyses in identifying novel biomarkers and potential therapeutic targets for cancer immunotherapy. These initial results, while promising, suffer from severe limitations in terms of data availability for specific cancer types and the lack of frameworks to develop and validate multi-omics predictors of immunotherapy response in a collaborative and scalable way.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
University Health Network.
Disclosure
B. Haibe-Kains: Financial Interests, Personal, Advisory Board: BreakThorugh Cancer, IONIQ Sciences, CQDM; Financial Interests, Personal, Speaker, Consultant, Advisor: Code Ocean.
Resources from the same session
175P - Radiomic biomarker of vessel tortuosity for monitoring treatment change: Preliminary findings in prospective evaluation of ECOG-ACRIN EA5163
Presenter: Pushkar Mutha
Session: Poster session 01
176P - Enhancing immunotherapy response prediction via multimodal integration of radiology and pathology deep learning models
Presenter: Marta Ligero
Session: Poster session 01
177P - Revealing differences in radiosensitivity of advanced non-small cell lung cancer (NSCLC)through single-cell sequencing data
Presenter: Peimeng You
Session: Poster session 01
178P - Explainable radiomics, machine and deep learning models to predict immune-checkpoint inhibitor treatment efficacy in advanced non-small cell lung cancer patients
Presenter: Leonardo Provenzano
Session: Poster session 01
179P - Molecular tumor board directed treatment for patients with advanced stage solid tumors: A case-control study
Presenter: Dhruv Bansal
Session: Poster session 01
180P - An HLA-diet-oriented system unveiling organ-specific occurrence of multiple primary cancers (MPC) with prevention strategy: A large cohort study of 47,550 cancer patients
Presenter: Zixuan Rong
Session: Poster session 01
181P - GeNeo: Agnostic comprehensive genomic profiling versus limited panel organ-directed next-generation sequencing within the Belgian PRECISION initiative
Presenter: Philippe Aftimos
Session: Poster session 01
182P - ALK fusion detection by RNA next-generation sequencing (NGS) compared to DNA in a large, real-world non-small cell lung cancer (NSCLC) dataset
Presenter: Wade Iams
Session: Poster session 01
183P - Frequency of actionable fusions in 7,735 patients with solid tumors
Presenter: Kevin McDonnell
Session: Poster session 01
184P - Patient-specific HLA-I genotypes predict response to immune checkpoint blockade
Presenter: Kyrillus Shohdy
Session: Poster session 01