Abstract 1422P
Background
Cancer immunotherapy induces durable responses in advanced esophageal squamous-cell carcinoma (ESCC). To understand the molecular mechanisms of its antitumor response and identify ideal candidates for tailoring effective immunotherapy, we proposed a robust molecular classification of ESCC.
Methods
A training cohort of 575 ESCC samples with RNA sequencing data from the Gene Expression Omnibus Database (GEO) was analyzed. Non-negative matrix factorization (NMF) algorithm was employed to separated gene expression patterns from tumor, stromal, and immune cell gene. The expression patterns were annotated by using a set of immune- and stromal-related gene signatures, potential immune biomarkers, and clinicopathological features. Total 71 samples from TCGA database were used for validation of immune genomic classification. All analyses were performed between subtypes using software package R (version 4.0.3).
Results
Approximately 42% of ESCCs in the cohort (243/575) were identified to show enriched inflammatory response, enhanced cytolytic activity, and active interferon-γ signaling (P < 0.001). We named this new molecular class of tumors as Immune Class. Further, we found Immune Class contained two distinct microenvironment-based subtypes, characterized by markers of active or exhausted immune response. The Exhausted Immune Class was characterized by enrichment of activated stroma and, WNT/TGF-β activation and poor survival (P < 0.05). In contrast, a low stromal enrichment score, high human papillomavirus (HPV) infection, and favorable prognosis were associated with Active Immune Class (P < 0.001). The robustness of these immune molecular subgroups was verified in the validation cohorts. Active Immune Class showed potential response to programmed cell death-1 blockade (P < 0.01).
Conclusions
Our findings revealed a novel immune classification in ESCC, and identified two subtypes characterized by active or exhausted immune responses, which may provide new insights for patient selection of suitable immunotherapeutic treatments.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.