Abstract
A prognostic model based on selected cell state and cellular community scores in patients with advanced melanoma treated with immune checkpoint inhibitors (Ecotype-ICI score) as a predictor of ICI immunotherapeutic benefits
Journal of clinical oncology, Vol.42(16_suppl), pp.9568-9568
06/01/2024
DOI: 10.1200/JCO.2024.42.16_suppl.9568
Abstract
9568 Background: We conducted an enhanced immune cell state atlas analysis of the data of patients with melanoma treated with ICI to investigate predictors of therapeutic benefits and mechanisms of resistance. Methods: Leveraging RNA-seq from melanoma patients (n=141) treated with ICI within the ORIEN Avatar project (NCT03977402), we constructed a prognostic model based on selected cell state and cellular community scores, also known as ecotype-ICI score ( Li. AACR 2023). We evaluated the model’s predictive value using transcriptomic data from high-risk melanoma patients treated with adjuvant ipilimumab (n=471) or interferon-α (n=248) as part of the E1609 phase 3 trial (1). RNA-seq data were initially deconvoluted for cellular community signatures using EcoTyper. The six prognostic carcinoma ecotype (CE) signatures identified (CE1, CE2, C6, CE7, CD9, CD10) were utilized for calculating a risk score based on a multivariable Cox model, which was trained using Avatar data. For additional external validation, we analyzed data from patients with metastatic melanoma treated with anti-CTLA4 ( Va. n=40), anti-PD1 ( Liu. n=121), and anti-PD1 alone or combined with anti-CTLA4 ( Gide. n=66). A series of ad hoc survival analyses, including Kaplan-Meier analysis, log-rank tests, and Cox regression were further applied. Results: The ecotype-ICI score showed strong prognostic significance in predicting overall survival (OS) in the entire group of E1609 patients (N=719; Cox P < 0.0001, log-rank P < 0.0001), the ipilimumab cohort (Cox P < 0.000143, log-rank P =0.00011), but less so with interferon (Cox P= 0.06). CE9, characterized by its proinflammatory nature and IFN-gamma signaling, and CE10, noted for its canonical T cell state signatures, both demonstrated a leading predictive value according to multivariable Cox regression analysis. Interestingly, CE2, noted for its lymphocyte deficiency signature, also demonstrated complementary predictive value for outcomes following ipilimumab. Within the validation cohorts of public data sets noted above, the ecotype-ICI scores showed similarly strong prognostic significance in predicting OS (Gide. P = 0.045; Liu. P = 0.029; Va. P = 0.011). Ecotype-ICI also demonstrated a distinct distribution in the responder group (complete or partial response) compared to the progressive disease group, underscoring its potential as a discriminative biomarker for ICI response and survival outcomes. Conclusions: Our analysis has successfully established the utility of the immune cell state atlas in predicting therapeutic benefits with ICIs across diverse ICI treatment regimens in melanoma. We will update our results with data related to gene ontology, KEGG pathways and the most complementary pathway signatures at the meeting. 1. Tarhini, 2020.
Details
- Title: Subtitle
- A prognostic model based on selected cell state and cellular community scores in patients with advanced melanoma treated with immune checkpoint inhibitors (Ecotype-ICI score) as a predictor of ICI immunotherapeutic benefits
- Creators
- Ahmad A. Tarhini - Moffitt Cancer CenterTingyi Li - Moffitt Cancer CenterSandra J. Lee - Dana-Farber Cancer InstituteIslam Eljilany - Moffitt Cancer CenterWilliam LaFramboise - Allegheny General HospitalF. Stephen Hodi - Dana-Farber Cancer InstituteMartin McCarter - University of Colorado DenverAbdul Rafeh Naqash - University of OklahomaHoward Colman - Huntsman Cancer InstituteMargaret E Gatti-Mays - The Ohio State University Comprehensive Cancer Center – Arthur G. James Cancer Hospital and Richard J. Solove Research InstituteAakrosh Ratan - University of VirginiaSusanne M. Arnold - University of KentuckyIgor Puzanov - Roswell Park Comprehensive Cancer CenterMichelle L. ChurchmanWilliam S. DaltonGeorge J. Weiner - University of IowaHoward Streicher - National Cancer InstituteWalter J Storkus - UPMC Hillman Cancer CenterJohn M. Kirkwood - UPMC Hillman Cancer CenterXuefeng Wang - Moffitt Cancer Center
- Resource Type
- Abstract
- Publication Details
- Journal of clinical oncology, Vol.42(16_suppl), pp.9568-9568
- DOI
- 10.1200/JCO.2024.42.16_suppl.9568
- ISSN
- 0732-183X
- eISSN
- 1527-7755
- Language
- English
- Date published
- 06/01/2024
- Academic Unit
- Hematology, Oncology, and Blood & Marrow Transplantation; Internal Medicine; Pharmaceutical Sciences and Experimental Therapeutics
- Record Identifier
- 9984656660002771
Metrics
10 Record Views