How does immune checkpoint blockade remodel the metastatic lymph node microenvironment?

Why the metastatic lymph node matters

Immune checkpoint inhibitors that block PD-1 or its ligand PD-L1 are used against advanced melanoma, but only some patients benefit, and the paper argues that events inside metastatic lymph nodes are poorly understood. The lymph node is often the first site melanoma reaches, and the study asks how effective checkpoint blockade changes that node's immune environment rather than only the primary tumor.

The authors built a mouse system in which B16F10 melanoma cells are injected into the hind footpad and spread to the draining popliteal lymph node. After three rounds of in vivo selection, the adapted cells produced more melanin pigmentation in draining nodes than parental cells, giving a reproducible model of lymph node metastasis. The authors state that this model does not fully reproduce the natural progression of spontaneous human melanoma metastasis, a boundary that applies to every result below.

Blockade shrinks nodal metastases, not the footpad tumor

Mice received three intraperitoneal doses of anti-PD-1 and anti-PD-L1 antibodies, or matched isotype control, starting 15 days after tumor inoculation and repeating every 5 days. The combined therapy produced only a modest, statistically non-significant reduction in primary footpad tumor size at the endpoint, and draining lymph node sizes were similar between treated and control mice.

Inside the node, the picture differed. Metastasis area, measured by quantifying melanin-pigmented regions, was significantly reduced in treated mice (P = 0.0014), as was metastatic area as a percentage of total lymph node area (P < 0.0001), with 16 nodes per group. Total lymph node cell counts were also significantly higher in treated mice (P = 0.0246, 7 nodes per group), so the nodes stayed similar in gross size while containing more cells and less pigmented tumor. The authors read this as combined blockade reducing metastatic burden in an immunocompetent model, and they note that the dissociation between nodal and primary tumor response suggests targeting of the metastatic microenvironment.

B cells and specific myeloid subsets expand

Flow cytometry at day 30 showed comparable numbers of T cells, including CD4+ helper and CD8+ cytotoxic subsets, and NK and NKT cells made up about 1% of the CD45+ compartment with no difference between groups. B220+ B cell numbers, by contrast, were significantly increased in combination-treated mice; the authors caution that B220 is also expressed by some plasmacytoid dendritic cells, so this population may include a minor pDC contribution.

Within the CD11b+ myeloid compartment, combination therapy increased overall myeloid numbers, driven by higher proportions of antigen-presenting cells (CD11c+MHCII-high), subcapsular macrophages (F4/80-CD169+), medullary macrophages (F4/80+CD169-), and infiltrating monocytes (CD11b+Ly6C-low). Monotherapy groups showed intermediate numbers of these subsets. On CD11b+F4/80+ macrophages, treated mice had elevated MHC-I, iNOS, and arginase and reduced CD206, a marker pattern the authors describe as a shift toward a pro-inflammatory phenotype.

T and B cell zones are rebuilt in treated nodes

Multiplex immunofluorescence covering more than 35 markers showed that metastatic nodes in control mice had disrupted architecture, with irregular acellular foci and diffuse T and B cell compartmentalization compared with contralateral nodes. In combination-treated mice, the boundary between the T cell paracortex and B cell follicles was restored, and the relative areas of T and B cell compartments were significantly larger, with a slight expansion of stromal areas and compression of myeloid areas.

The authors flag an important measurement limit: the panel did not include tumor-specific markers, so metastatic foci could not be identified directly in these sections. The spatial results therefore describe organization of immune and stromal compartments within the analyzed node, not direct measurement of tumor-associated spatial remodeling. Marker-level analysis showed increased lymphoid markers including CCR6, B220, CD3e, and CD8b, decreased myeloid markers CD301a and GR1, decreased stromal ICAM-1 and CD146, and increased alpha-smooth muscle actin, alongside higher CD38 and Ki67 and lower GITR, consistent with activation and proliferation.

Myeloid cells shift toward an immunostimulatory program

CD11b+ myeloid cells sorted from draining nodes were profiled with the Nanostring Myeloid Gene Expression Panel. The paper reports differentially expressed genes between treated and control groups, a heatmap of the top 125 most differentially expressed genes, and gene ontology and pathway enrichment for the upregulated signature, including STAT1/STAT3 signaling as a proposed contributor to macrophage reprogramming. The authors present this transcriptional shift as support for myeloid reprogramming rather than as proof of a molecular mechanism, and they call for further work on how macrophage reprogramming is controlled during immunotherapy.

Imaging added a spatial dimension. Within metastatic areas, macrophages clustered on T cells in control nodes, and treatment increased the distance between T cells and macrophages, quantified as T cell-macrophage pairs within 30 micrometers as a percent area (n = 3 per group). The authors interpret this as possibly reduced suppressive interaction, but they explicitly note that confirmation is needed that T cells are actively migrating away from macrophages, and that future work should profile T cells after disengagement to see whether they recover function or move toward exhaustion.

Cell counts and pixel areas disagree for a reason

Two measures of the myeloid compartment point in opposite directions: flow cytometry shows more myeloid cells after treatment, while immunofluorescence shows a smaller myeloid pixel area. The authors attribute the apparent myeloid area decrease largely to loss of putative GR1+ suppressor cells, with the compartment instead enriched in resident subcapsular and medullary sinus macrophages and CD11c+MHCII+ antigen-presenting cells. They also list technical explanations, including changes in the segmented size of activated or recruited myeloid cells and biases introduced during tissue dissociation.

A similar discrepancy appears for B cells: flow cytometry shows an increased B cell proportion in combination-treated nodes, while B cells occupied a reduced immunofluorescence pixel area. The authors suggest this may reflect changes in cell size and compartmentalization, since follicles were disrupted and diffuse in control metastatic nodes. These mismatches are a reminder that a cell count, a cell proportion, and a stained area are different measurements and should not be substituted for one another.

Stromal cells are active participants, not scaffolding

Stromal populations also changed with treatment. Fibroblastic reticular cells (CD45-GP38+) increased in treated nodes, with trending increases in blood endothelial cells (CD45-GP38-CD31+) and lymphatic endothelial cells (CD45-GP38+CD31+). The authors argue that stromal cells are not passive structural components but agents that can promote anti-tumor T and B cell responses, and that increased alpha-smooth muscle actin, a common marker of activated fibroblasts, indicates immunostimulatory stromal reprogramming. They also note that this activation marker can be upregulated in other contexts, so it is not a clean readout on its own.

For B cells, the paper places the observed expansion alongside their roles in sampling lymphatics and facilitating antigen presentation, and notes that B cells can also take on regulatory, suppressive phenotypes depending on cues from the metastasis. Whether B cell behavior in metastatic nodes resembles that in tertiary lymphoid structures remains an open question in the authors' framing.

A model of released macrophage-mediated suppression

The authors summarize their findings as a model in which macrophage-T cell interactions in metastatic lymph nodes are associated with T cell suppression and tumor growth, and combined PD-1/PD-L1 blockade alters those interactions while remodeling lymphoid, myeloid, and stromal compartments in ways consistent with anti-tumor immunity. The word associated matters: the study reports correlations between treatment, macrophage phenotype, spatial separation, and reduced metastatic area, not a demonstrated causal chain from macrophage reprogramming to tumor control.

The most decision-relevant split in this paper is between the metastatic lymph node and the primary footpad tumor. Combination blockade significantly reduced melanin-pigmented metastatic area and percentage of node area and increased node cellularity, while the primary tumor reduction was modest and non-significant. The authors note that similar footpad studies have found therapeutic tumor reduction within the lymph node but not at the primary site, and they propose this points to unique targeting of the metastatic tumor microenvironment. Testing whether that nodal effect generalizes will require the spontaneous or genetically engineered melanoma models and patient specimens the authors call for, since all data here come from the B16F10 footpad system.

Sources used in this answer