Abstract
Melanoma, a highly aggressive and immunogenic skin cancer, frequently escapes the body’s immune defences. Treatment of melanoma, which is currently based on targeted therapy and immunotherapy, has dramatically advanced over the past decade. Advances in targeted therapy have been based on inhibition of the MAPK pathway, while for immunotherapy, advances have been based on blocking immune checkpoint proteins such as the PD-1/PD-L1 axis. Immune checkpoint inhibitors (ICI), such as anti-PD1 and anti-PDL1 antibodies lead to reactivation of immune pathways, promoting rejection of melanoma. However, the benefits of ICI therapy remain limited to a relatively small proportion of patients. The precise mechanisms underlying innate and acquired ICI resistance remain unclear. Only 30-50% of melanoma patients respond, or they respond poorly to ICI therapy. Yet, all ICI treated the patients are susceptible to significant and potentially serious side-effects or adverse events from ICI treatment. To date, no accurate biomarkers have been reported for prediction of who will respond to ICI treatment.
In this project, I have investigated differences in melanoma tissues in responder and non-responder patients to anti-PD1 therapy in terms of tumour and immune cell gene-associated signatures. I performed multi-omics investigations, including RNA-Seq, NanoString, tumour mutation burden (TMB), and Infinium MethylationEpic genome-wide methylation analysis on melanoma tumour tissues, which were collected from patients before starting treatment with anti-PD1 immune checkpoint inhibitors. Patients were subsequently categorized into responder and non-responder categories to anti-PD1 therapy based on RECIST criteria.
As I have used archival formalin-fixed paraffin-embedded (FFPE) clinical tissue samples, the first aim of this study was to identify good quality samples to generate reproducible and reliable data for performing transcriptomic and methylomic studies. Following on from that, as part of the initial phase of this study, I analysed RNA-Seq data as well as Oncomine targeted DNA sequencing data from the FFPE tissues to understand the biology of the responding and non-responding melanoma tumour bed and its association with the tumour microenvironment. From the RNA-Seq data I carried out HLA phenotyping as well as gene enrichment analysis, and immune cell deconvolution studies. Consistent with previous studies, the data from this study showed that responders to anti-PD1 therapy had higher immune scores compared to the non-responders. Responder melanomas were more highly enriched with a combination of CD8+ T cells, dendritic cells and an M1 subtype of macrophages. In addition, melanomas from responder patients exhibited a more differentiated gene expression pattern, with high proliferative- and low invasive-associated gene expression signatures, whereas tumours from non-responders exhibited high invasive- and frequently neural crest-like cell type gene expression signatures. This study suggests that non-responder melanomas exhibit a de-differentiated gene expression signature, which is associated with poorer immune cell infiltration, and which establishes a gene expression pattern that characterises innate resistance to anti-PD1 therapy.
This study also confirmed from Oncomine targeted DNA sequencing that TMB is not an accurate predictor for melanoma patients’ response to anti-PD1 therapy. However TMB had a strong positive correlation with expressed neoantigens, and a moderate positive relationship with immune scores in melanoma patients. In terms of global methylation landscapes, the responding melanomas exhibited distinct patterns of methylation compared to the non-responding melanomas, which was supported by PCA analysis and differential methylation patterns. A comparative analysis between methylation and transcription data revealed that several genomic loci were commonly altered in both methylation and transcription datasets, which suggests there may be functionally relevant changes to the genomic regulation of responding versus non-responding melanomas.
In conclusion, in this project I successfully profiled transcriptomes and methylomes associated with melanoma tissues from responder and non-responder patients to ICI therapy, and I have identified multiple features and genomic markers that distinguish responder from non-responder patients. Based on the findings I have presented in this thesis, in future work with further validation in an independent melanoma cohort, it will be possible to choose methylation markers that are co-ordinated with gene expression differences, which could be further investigated as potential biomarkers to differentiate between responding and non-responding melanomas to anti-PD1 therapy.