Abstract 1166P
Background
In-transit metastases (ITM) in melanoma represent a clinical entity where tumor lesions are found between the primary tumor and the draining lymph node. ITM is associated with worse prognosis. However, a significant subset of patients have discordance with high burden of locoregional relapses, but durable distant progression free survival. We studied two such patients to profile the dynamics and evolution of longitudinal tumor samples and identify mechanisms underlying these heterogeneous phenotypes.
Methods
We used whole exome sequencing (WES), bulk RNAseq, snRNAseq, and high-plex Cyclic Immunofluorescence (CyCIF) imaging to profile 28 biopsies over 7 years from pt1 (acral lentiginous melanoma, with progressive disease) and 22 biopsies over 11 years from pt2 (cutaneous melanoma, with durable remission).
Results
Pt1 (progressive disease) tumors had low to absent levels of immune infiltrate, low tumor mutational burden (TMB) and 5-hmC loss. Phylogenetic analysis revealed co-evolution of seven genetic lineages with multiple independent resistance-associated alterations; the majority of distant metastases emerged from a single lineage characterized by high aneuploidy and TMB, and a likely pathogenic missense mutation in TET2. The brain metastasis diverging early in molecular evolution but emerging late in disease had the highest TMB and aneuploidy. In another lineage, we identified WNT-beta catenin transcriptional signature contributing to TIL exclusion. In contrast, phylogenetic analysis of pt2 (durable remission) revealed six co-evolved genetic lineages with acquisition of mutational signature 11 and high TMB (53 mut/MB) after dacarbazine chemotherapy. In line with this, several tumor samples contained abundant Ki67+ CD8+ TILs and regression-like stromal changes, reflective of an autonomous effective host-immune response leading to a complete remission independent of any systemic therapy.
Conclusions
This study identifies genomic and phenotypic features linked to aggressive phenotypes and distant metastasis, including high aneuploidy, TET2 mutation, and 5hmC loss. Additional spatial analysis (GeoMx) is currently ongoing.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Doris Duke foundation.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
1159P - Characterization of melanoma of unknown primary in the era of immunotherapy and targeted therapy in Spain: Results from the prospective real-world study GEM 1801
Presenter: Pablo Cerezuela-Fuentes
Session: Poster session 13
1160P - Methods of nivolumab administration in advanced melanoma: A comparison of patients’ clinical outcomes treated with flat dose or weight-adjusted dose, FLATIMEL study
Presenter: Iona Campo le Brun
Session: Poster session 13
1161P - Therapeutic outcome of molecular profiling of melanoma patients resistant to standard treatment: Real-world data
Presenter: Madona SAKKAL
Session: Poster session 13
1162P - Prolonged exposure to proton pump inhibitors (PPI) at the time of initiation of immune checkpoint blockade (ICB) mediates better clinical outcomes in patients with metastatic melanoma
Presenter: Kyrillus Shohdy
Session: Poster session 13
1163P - CD39 affect the prognostic role of NLR via N2 neutrophils in metastatic melanoma patients treated with immunotherapy
Presenter: Domenico Mallardo
Session: Poster session 13
1164P - Changes of TCR repertoire in metastatic melanoma and renal cell carcinoma patients treated with nivolumab correlate with overall survival
Presenter: Martin Klabusay
Session: Poster session 13
1165P - Single cell spatial features of in-transit melanoma associated with patient outcome to immunotherapy
Presenter: Xinyu Bai
Session: Poster session 13
1168P - Tumor PD-L1 predicts the outcome of PD-1-based immunotherapy in metastatic melanoma depending on the type of tissue examined
Presenter: Jan-Malte Placke
Session: Poster session 13
1169P - Tumour transcriptional and spatial protein profiling in Mexican patients reveals that acral lentiginous melanoma is characterized by an immunosuppressive microenvironment
Presenter: Martha Estefania Vázquez-Cruz
Session: Poster session 13
1170P - Survival outcome prediction of primary melanoma tumours from histology images using deep learning
Presenter: Céline Bossard
Session: Poster session 13