Lead-in therapy targeting PD1 and/or LAG3 imposes distinct immune phenotypes in first-line treatment of metastatic melanoma

Simultaneous blockade of lymphocyte activation gene-3 (LAG-3) and programmed-death-1 (PD-1) pathways enhance anti-tumor activity in patients with melanoma. It is not known if this immunotherapy induces unique immune modulation or how this may relate to clinical outcomes, compared to the single modalities. We conducted a randomized three-arm phase 2 trial ( NCT03743766 ) in advanced melanoma comparing lead-in treatment with one cycle of relatlimab-(n=14), nivolumab-(n=15), and nivolumab-relatlimab (n=14) as first-line therapy, followed by combination nivolumab-relatlimab, to assess the impact of the lead-in therapy on immune populations, related to objective response rate, progression-free survival (PFS) and major pathologic response on biopsy (MPRbx). Unexpectedly, diminished efficacy was observed when nivolumab or relatlimab was given as a lead-in monotherapy, despite receipt of the combination subsequent to week 4. MPRbx at week 4 was significantly higher with nivolumab- and combination vs. relatlimab-lead-in, mechanistically correlated with higher IFN-g and TCR signaling in CD8 ⁺ T cells, and was associated with superior PFS. Relatlimab lead-in was characterized by clustering of CD8 ⁺ T and FOXP3 ⁺ T regulatory cells in the tumor microenvironment, enrichment in inhibitory receptor pathways and associated with worse PFS. Combination therapy increased the fraction of CD8 ⁺ T _EM cells in association with response to therapy while patients with progressive disease demonstrated decrease in CD14 ⁺ CD16 ^- HLA-DR ^low CD33 ^dim monocytic populations. The analyses of component agents afforded by this novel lead-in trial has identified differential clinical and immunological modulation of anti-PD1 and anti-LAG3 in comparison to the combination therapy.