3 Protein Patterns Predict Melanoma Immunotherapy Response

Cutaneous melanoma responds strongly to immunotherapy, yet system-level mechanisms linking mutational burden, genomic alterations, and antigen patterns remain incompletely defined, according to findings published on Tuesday.

Two new companion studies developed a molecular atlas of melanoma and a classification method to predict patient response before treatment begins.

Researchers Build Protein Atlas for Melanoma

Immune checkpoint inhibitors have improved melanoma survival over the last 15 years, but a large percentage of patients fail to respond. To uncover why some tumors resist therapy, researchers examined the proteome, which represents the entire set of proteins expressed in a tissue. Proteins act as the molecular machines carrying out cellular work.

In a study published in Cancer Discovery, researchers profiled 183 cutaneous melanomas using 13 omic layers. The multi-omics approach included immunopeptidomics, metabolomics, protein ubiquitylation, and glycosylation to build a comprehensive atlas of tumor immunogenicity.

"To understand the proteomic landscape in melanoma, we needed two different sets of analyses," said Diwakar Davar, co-senior author and associate professor of medicine at the University of Pittsburgh.

The primary tumor analysis linked genomic changes to protein consequences, revealing therapeutic vulnerabilities in untreated tumors.

Identifying Three Distinct Subtypes

A companion study published in Clinical Cancer Research utilized data from 207 or 208 preserved tumor samples obtained prior to immunotherapy from patients whose skin melanoma had spread. The team identified three distinct melanoma subtypes characterized by unique protein patterns.

One group exhibited increased immune activity, another showed elevated cell proliferation features, and a third displayed characteristics associated with tumor growth and metastasis. The immune-rich group demonstrated the highest response rate to combination immunotherapy.

Researchers trained a classifier model to sort tumors into these three categories. The model predicted treatment response and progression-free survival with a high degree of certainty across validated patient cohorts.

Translating Findings Into Clinic Tests

The next step involves translating these biological discoveries into a practical clinical test for hospital laboratories. Amanda Paulovich of Fred Hutch Cancer Center noted the identification of a panel of 127 proteins capable of distinguishing among the three melanoma subtypes.

"Our next step is to translate these findings into a practical clinical test," said Amanda Paulovich, of Fred Hutch Cancer Center. "We’ve identified a panel of 127 proteins that can distinguish among three melanoma subtypes associated with different responses to immunotherapy."

Researchers aim to develop a targeted mass spectrometry assay utilizing routine biopsy samples to classify individual patient tumors.

Questions About Melanoma Proteomics

What data did the multi-omic atlas analyze?

Researchers profiled 183 cutaneous melanomas using 13 omic layers, including the genome, proteome, metabolome, immunopeptidomics, protein ubiquitylation, and glycosylation.

How many patient samples were used in the Clinical Cancer Research study?

The study analyzed 208 tumor samples obtained from patients prior to receiving immunotherapy, drawing on contributions from National Cancer Institute-funded Melanoma SPORE programs and 12 participating institutions.

What is the purpose of the 127-protein panel?

The panel of 127 proteins distinguishes among three distinct melanoma subtypes associated with different responses to combination immunotherapy, forming the basis for a planned targeted mass spectrometry assay.