Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Multi-Organ Genomic Analysis of Resistant Cutaneous Melanoma Provides Insights for Potential Improvement of Treatment Strategies

Findings from genomic analysis in a rapid autopsy cohort study
16 May 2023
Translational Research

Genomic analysis performed in a rapid autopsy cohort study of patients with melanoma identify the genetic and transcriptomic landscape of melanoma with acquired resistance to MAPK or immune checkpoint inhibitors, providing insights for the potential improvement of treatment strategies. Multiple forms of genomic instability may cause and/or result from resistance evolution, with treatment implications that warrant mechanistic studies.

The study team uncovered a preliminary set of organ-specific metastatic signatures, tumour macroenvironment crosstalks and immune contextures that characterise resistant cutaneous melanoma, justifying expanded rapid autopsy melanoma-based and functional analyses. The findings are published by Dr. Roger S. Lo of the David Geffen School of Medicine, University of California in Los Angeles, CA, US, and colleagues on 27 April 2023 in the Nature Medicine.  

Cutaneous melanoma exhibits UV-related high mutational burden. Mutually exclusive BRAF and NRAS mutations drive MAPK addiction in approximately 70% of metastatic cutaneous melanoma cases. Cutaneous melanoma genomes also harbour a high burden of structural variants and chromothripsis. Current knowledge of this mutational landscape is derived from tumours naive to highly active treatments developed recently and inclusive of earlier-stage disease. How MAPK or immune checkpoint inhibitors alter the mutational landscape and, thereby, cause death remains largely unknown.

Metastasis and treatment failure represent the most common causes of mortality in patients with cutaneous melanoma. Despite experimental metastasis studies, the difficulty of accessing patient-derived metastatic tissues has limited understanding of clinical metastatic and organ-specific evolution. Although the concept of an organ-specific pre-metastatic niche has been demonstrated experimentally, little is known clinically regarding co-adaptations between metastases and their site-specific macroenvironments.

Acquired MAPK inhibitor resistance in cutaneous melanoma has been evaluated at an omics scale in a few cohorts, but knowledge of clinically acquired immune checkpoint inhibitor resistance is limited. Metastases to accessible anatomic sites overrepresent current datasets on acquired resistance. Monitoring and treatment strategies to counter resistance require insights into multi-organ mechanisms and heterogeneity.

It is unknown whether rare signatures of a particular malignancy, such as UV-related cutaneous melanoma, are common with respect to late mutations, potentially due to the influence of a particular treatment. Such signatures that emerge later during tumour evolution might represent targetable pathway defects or synthetic lethalities.

The study team assembled a rapid autopsy melanoma cohort from patients with BRAF- or NRAS-mutated cutaneous melanoma who were treated with and responded initially to MAPK or immune checkpoint inhibitors, but later died because of disease progression. This cohort includes 71 distinct metastatic tumours, 41 tumour-adjacent ‘normal’ tissues representing organ-specific tumour macroenvironments and 38 tumour-non-adjacent normal.

The study team generated and analyzed whole-exome sequencing (WES) from tumours and patient-matched tumour-non-adjacent normal as well as whole-genome sequencing from a subset. To dissect the contribution of distinct treatments, they comparatively analyzed WES data from longitudinal pre- and post- tumours from patients with cutaneous melanoma who had progressed on either MAPK only or immune checkpoint inhibitor only treatment.

Moreover, the study team developed organ-specific metastatic signatures based on tumour cell-enriched transcriptomes and analyzed ligand–receptor signalling between tumour and tumour-adjacent ‘normal’ tissues. Finally, they deconvolved tumour, tumour-adjacent ‘normal’ and tumour-non-adjacent normal transcriptomes to decipher organ-specific immune contextures.

Analyzing data from separate melanoma cohorts including 345 treatment-naive patients and 35 patients with patient-matched pre-treatment and post-acquired resistance tumour samples, the study team performed cross-cohort analyses to identify MAPK or immune checkpoint inhibitors as respective contributors to gene amplifications and deletions enriched in autopsy versus treatment-naive tumours.

In the autopsy cohort, private/late mutations and structural variants display shifted mutational and rearrangement signatures, with MAPK inhibitor specifically selecting for signatures of defective homologous-recombination, mismatch and base-excision repair. Transcriptomic signatures and crosstalks with tumour-adjacent macroenvironments nominated organ-specific adaptive pathways. An immune-desert, CD8-positive-macrophage-biased archetype, T-cell exhaustion, and type-2 immunity characterised the immune contexture.

The authors commented that by comparative analysis of acquired-resistant cutaneous melanoma with patient-matched pre-treatment tumours, they resolved how each treatment distinctly and convergently shapes the high mutational, copy number alteration (CNA) and structural variant burdens of acquired-resistant cutaneous melanoma. Significantly mutated genes and genes altered by CNAs and structural variants enrich in immune-evasive processes that may confer cross-treatment resistance, accelerating lethal disease progression.

Notably, evolution of MAPK versus immune checkpoint inhibitor resistance shifts the mutational signatures, implicating treatment-elicited DNA damage and/or deficiency in repair pathways as culprits. The evolution of late/private structural variants, regardless of treatment history, enriches for RS2. Analysis of breakpoint-junctional sequences of structural variants suggests non-homologous end joining as a MAPK or immune checkpoint inhibitor co-target.

Computationally depleting bulk metastatic tumour transcriptomes of patient-matched and organ-matched normal tissue-derived transcriptomes shed light on organ-specific metastatic signatures. Liver and spleen metastases display neural differentiation, suggesting treatment targets. Melanoma brain metastasis displays signatures of IFN signalling, oxidative phosphorylation and PI3K-AKT signalling. The brain-specific macroenvironment appears to be a predominant source of IFN ligands.

Overall, rapid autopsy melanoma tumours, including melanoma brain metastasis, strongly display an immune-desert but CD8-positive-macrophage-biased archetype with enrichment of T-cell exhaustion. For melanoma brain metastasis, loss of antigen presentation and enrichment of type-2 immunity suggest TGFβ blockade and upregulation of cytotoxic NK cell-mediated or CD4-positive T cell-mediated anti-tumour immunity as potential treatment strategies. The pro-tumourigenic tumour-associated macrophage phenotype of melanoma brain metastasis also suggests treatment co-targets.

The authors concluded that this study begins to build foundational insights into highly evolved and lethal cutaneous melanoma resistant to MAPK or immune checkpoint inhibitors. 


Liu S, Dharanipragada P, Lomeli SH, et al. Multi-organ landscape of therapy-resistant melanoma. Nature Medicine; Published online 27 April 2023. DOI: https://doi.org/10.1038/s41591-023-02304-9

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings
  • Necessary cookies enable core functionality. The website cannot function properly without these cookies, and you can only disable them by changing your browser preferences.