Abstract 416P
Background
Locally advanced oesophageal adenocarcinoma (EAC) remains difficult to treat because of common resistance to neoadjuvant therapy and frequent recurrence.
Methods
We performed a multi-omic study with a multi-timepoint strategy to investigate neoadjuvant treatment response at clonal resolution and in the tumor microenvironment. EAC samples from chemotherapy responding (REPs) and chemotherapy non-responding (NRPs) patients were collected at three time points (before, during, and after neoadjuvant therapy) within the DKTK-funded MEMORI trial[1]. Deep whole exome sequencing was performed on 48 samples from 17 REPs and on 22 samples from 10 NRPs. Matched RNA sequencing was performed on 54 samples from 17 REPs and on 26 samples from 10 NRPs. Longitudinal imaging mass cytometry (IMC) to characterize immune response was performed on 27 REP-samples and 16 NRP-samples and T-cell receptor sequencing in 18 REP-samples and 9 NRP-samples.
Results
EAC showed major transcriptomic changes during treatment with upregulation of immune, stromal and oncogenic pathways such as EMT, KRAS, Hedgehog and WNT. However, phylogenetic analysis revealed that clonal sweeps were rare, suggesting that gene expression changes were unlikely to be clonally driven. Immune escape through HLA-LOH was linked to an inactive T-cell phenotype and to poor clinical treatment response. Moreover, IMC analyses showed a lower ratio of cytotoxic GranzymeB+ to PD-1+ CD4 and CD8 T cells in NRPs compared to REPs at all timepoints, indicating a less effective T-cell phenotype in NRPs. TCR analyses revealed a lack of clonal T-cell expansions in patients with poor treatment response.
Conclusions
Based on a multi-timepoint approach, we integrated multi-omic analyses for a comprehensive understanding of treatment resistance. Our study highlights profound transcriptional changes and EMT during treatmentwithout underlying clonal replacement, suggesting phenotypic plasticity as a driver for therapy resistance. The presence of HLA-LOH, a less activated T-cell phenotype and a lack of clonal T-cell expansions in patients with poor clinical treatment response suggest combined immuno-chemotherapy as a potential efficacious approach.
Clinical trial identification
NCT02287129.
Legal entity responsible for the study
The authors.
Funding
German Cancer Aid Society (Deutsche Krebshilfe) (M.B.), Cancer Research UK (M.B, A.B. and T.A.G.), the US NIH via the Cancer Systems Biology Consortium U54 scheme (CA217376) (T.A.G.), German Research Foundation (DFG 3772/1) (M.Q.), AIRC under MFAG 2020 (ID. 24913 project) (G.C.), German Cancer Consortium (DKTK) (J.T.S.), Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) (405344257 (SI 1549/3–2)) (J.T.S.), German Federal Ministry of Education and Research (BMBF; 01KD2206A/SATURN3) (J.T.S.). The MEMORI trial was supported by the German Cancer Consortium (DKTK). The authors acknowledge DKTK Partner site Essen, DKTK Partner site Freiburg, DKTK Partner site München. The authors acknowledge iBioTUM-Tissue and study coordinator Jens-Peter Zimmermann for management of biosamples.
Disclosure
J.T. Siveke: Other, Personal and Institutional, Invited Speaker, Funding: AstraZeneca, Bayer, Boehringer Ingelheim, Bristol Myers Squibb; Other, Personal, Invited Speaker: Immunocore, MSD Sharp & Dohme, Novartis, Servier; Other, Personal, Invited Speaker, Funding: Roche/Genentech; Other, Institutional, Funding: Abalos Therapeutics, Bristol Myers Squibb; Other, Personal and Institutional, Advisory Board: Pharma15. All other authors have declared no conflicts of interest.