Background The development of immunotherapy, particularly immune checkpoint blockade, represents a significant recent advancement in clinical oncology. However, its effectiveness is restricted to a subset of patients, with many eventually developing acquired resistance. Consequently, identifying biomarkers of therapeutic response is essential for devising strategies to enhance overall response rates.

Methods We retrospectively analyzed PBMCs and plasma from stage III and IV melanoma patients at baseline and after the first anti-PD-1 treatment cycle. T-cell functional state was profiled using a single-cell polyfunctional assay, gene expression differences were determined by single-cell sequencing, and immune cell populations were validated by flow cytometry. Metabolic endpoints were assessed by cellular respiration in PBMCs and plasma metabolomics.

Results Our data shows that patient T-cell polyfunctional subsets cluster differently by response upon stimulation, suggesting distinct combinations of cytokine secretions. Responder patients had a 6-fold higher polyfunctional strength index (% polyfunctional single cells x the average signal intensity of secreted proteins) when compared to non-responders. This included higher expression of effector and stimulatory markers, including granzyme B, IFNγ, TNFα, and IL-2, among other cytokines. To determine potential mechanisms driving endpoints of T-cell polyfunctionality, we performed single-cell RNA sequencing of patient PBMCs. For 12 patients, transcriptomic clustering reproducibly identified 7 significant immune cell populations expected to be present in PBMCs. Since anti-PD1 therapy mainly targets the T-cell population, we focused on genetic changes in this compartment. Gene enrichment analysis of the T-cell population between responder vs. non-responder patients indicated that genes associated with oxidative phosphorylation are the highest modulated pathway. We measured the oxygen consumption rate (OCR) as a surrogate of mitochondrial metabolism by seahorse® analysis to validate this observation. Our data showed that responder patients had an over 50% (*p<0.03, n=30) increase in spare respiratory capacity after the first treatment cycle compared to non-responder patients. This suggests that the increased polyfunctional index in responders may be partly due to better mitochondrial fitness. Metabolomic analysis revealed clustering of patients by response. Furthermore, enrichment analysis indicated that pathways associated with lipid metabolism are associated with response. Furthermore, several carnitine metabolites, including lignoceroylcarnitine and hydroxylated carnitines (3-hydroxydecanoylcarnitine & 3-hydroxyoctanoylcarnitine), were inversely correlated with Overall Survival (N=40, *p<0.05) thus suggesting a potential critical role of metabolism in immune checkpoint blockade treatment outcomes.

Conclusions These studies may identify metabolic biomarkers to predict patient response, guide treatment decisions, and develop new strategies to improve patient outcomes and quality of life.

Acknowledgements This work is supported by the Melanoma Research Foundation Established Investigator Award to DRS-P. We acknowledge the support of the Wake Forest Baptist Comprehensive Cancer Center‘s Shared Resources: Cancer Genomics (CGSR), Bioinformatics (BISR), and Flow Cytometry (FCSR), supported by the National Cancer Institute’s Cancer Center Support Grant, award number P30CA012197.

Ethics Approval All patients provided written informed consent for the research approved by the Wake Forest University Health Sciences Institutional Review Board (IRB), according to the ethical standards put forward by the Belmont Report, federal, state, and local regulations, and policies governing human research.