Abstract 1125P
Background
Long-term clinical outcomes post immune checkpoint blockade (ICB) for metastatic melanoma (MM) are highly variable. Whilst this is in part reflects inter-tumour heterogeneity, the degree patient germline genetic variation influences ICB responses is poorly characterised. Previous pan-tumour patient cohort studies suggested that homozygosity at Class I MHC alleles is associated with reduced survival post ICB, although these observations have not been reproduced in meta-analyses. Here we explore the relationship between homozygosity at one, two and three alleles at Class I MHC in a longitudinal follow-up cohort of patients receiving ICB for MM, integrating observations with immunological and transcriptomic data at single-cell resolution.
Methods
232 MM patients receiving standard of care ICB were HLA typed using genome-wide genotyping and followed for up to 9 years. Pre-treatment and on treatment CD8 T cells RNA and T cell receptor (TCR) sequencing of was performed for all samples (n>500), as well as focused single-cell RNA seq. The relationship between homozygosity at HLA-A, HLA-B and HLA-C and survival parameters were explored with results integrated with transcriptomic and immunophenotyping.
Results
50.9% of patients survived for >5 years. Notably, no effect of homozygosity at either one or two Class I HLA alleles on outcomes was noted (60.5% vs. 49.8% - no homozygosity, P=0.17). However, patients homozygous at three alleles (3.0% cohort) have catastrophic outcomes, independent of tumour features, with 1 year survival 17.8% vs. 75.2%, P<0.0001. Pre-treatment these patients have reduced CD8 TCR diversity, and marked immune dysregulation. Single-cell RNA sequencing demonstrates impaired responses to ICB from treatment initiation.
Conclusions
Understanding ICB resistance is of vital importance in stratifying MM treatments. We find homozygosity at one or two Class I MHC alleles does not impact outcomes. Conversely, we demonstrate a subset of MM patients have homozygosity at all three Class I MHC alleles. They show markedly impaired immunological and transcriptomic responses to ICB, with very poor clinical outcomes. This work marks the first description of germline encoded ICB resistance.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
University of Oxford.
Funding
Wellcome Trust, Cancer Research UK.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
1198P - Phase Ib portion of the ACTION-1 phase Ib/3 trial of RYZ101 in gastroenteropancreatic neuroendocrine tumors (GEP-NET) progressing after 177Lu somatostatin analogue (SSA) therapy: Safety and efficacy findings
Presenter: Jonathan Strosberg
Session: Poster session 13
1199P - MAVERIC: Phase II randomized study of everolimus as maintenance therapy for metastatic neuroendocrine neoplasms (mNEN) with pulmonary or gastroenteropancreatic (GEP) origin. Results on behalf of the GOIRC
Presenter: Lorenzo Antonuzzo
Session: Poster session 13
1200P - A phase II single-arm interventional trial evaluating the activity and safety of CABOzantinib (CBZ) plus TEMozolomide (TMZ) in lung and gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) progressed after SSA therapy, everolimus, sunitinib or PRRT: CABOTEM Trial
Presenter: oOttavia Clemente
Session: Poster session 13
1201P - Evaluation of efficacy TEMCAP regiment as first-line or further line therapy in patients with advanced, unresectable, progressive GEP-NET. Real-world data
Presenter: Agnieszka Kolasinska-Cwikla
Session: Poster session 13
1202P - Final analysis of TENEC trial: A phase II trial of temozolomide (TMZ) as second-line treatment for advanced neuroendocrine carcinomas (NEC) in patients (pts) with a poor performance status (PS)
Presenter: Claudia von Arx
Session: Poster session 13