Blocking CD38 Revives Exhausted T Cells and Breaks Immunotherapy Resistance

Immune checkpoint blockade (ICB) has transformed cancer treatment, yet a large proportion of melanoma patients develop resistance, often fatally. Understanding why some patients fail to respond—and how to rescue those responses—is one of the most pressing challenges in oncology. This study identifies CD38, an NAD+-consuming ecto-enzyme, as a key driver of T cell dysfunction and ICB resistance in human melanoma.

Using published single-cell RNA sequencing data from melanoma patients, the researchers showed that CD38 expression in CD8+ tumor-infiltrating lymphocytes (TILs) closely mirrors PD-1 expression and marks exhausted T cell states. Exhausted CD8+ T cell clusters (defined by high co-inhibitory receptor expression and low TCF7) showed the highest CD38 levels, while effector memory and early-activation clusters showed low CD38 and high TCF7. CD38 expression in CD8+ T cells was a strong predictor of ICB resistance, achieving an area under the curve (AUC) of 0.87 across melanoma patient cohorts, and was elevated both before and after treatment in non-responders.

Mechanistically, CD38 upregulation was driven by two distinct signals: chronic T cell receptor (TCR) stimulation (mimicking persistent tumor antigen exposure) and type I interferon signaling—both hallmarks of the tumor microenvironment. Disrupting CD38 via pharmacological inhibitors or genetic knockout restored intracellular NAD+ pools, improved mitochondrial bioenergetics (including oxidative phosphorylation and spare respiratory capacity), increased TCF7 expression, and enhanced cytokine production and cytotoxic effector functions in CD8+ T cells.

Critically, the team validated these findings in patient-derived organotypic tumor spheroids (PDOTS)—3D microfluidic ex vivo cultures that preserve autologous tumor, immune, and stromal cells from fresh melanoma surgical specimens. In a cohort of ICB-resistant patient tumors, dual PD-1/CD38 blockade significantly restored anti-tumor immune responses compared to either agent alone, overcoming resistance that was not overcome by PD-1 blockade alone.

These results establish the CD38-NAD+ axis as a targetable metabolic vulnerability in exhausted T cells and provide preclinical human tumor evidence supporting clinical evaluation of CD38-directed therapies—alone or combined with PD-1 inhibitors—for patients with ICB-resistant melanoma. The findings also highlight NAD+ metabolism as a critical determinant of T cell fitness in the tumor microenvironment.