For the roughly 50% of colorectal cancer patients who develop liver metastases, chemotherapy is a lifeline — but the liver damage it causes may simultaneously open the door for tumors to return. A newly identified immune mechanism could explain why some patients relapse despite aggressive treatment, and it points toward a targetable molecular culprit.

Using single-cell transcriptomic profiling of liver immune cells from ten patients with colorectal liver metastases who developed chemotherapy-associated steatohepatitis (CASH) following irinotecan treatment, researchers identified a distinct macrophage population marked by the scavenger receptor MARCO. These MARCO+ macrophages, confirmed across 24 matched human tissue pairs and an irinotecan-induced mouse model, accumulate specifically during the steatohepatitis phase — not during acute liver injury — suggesting their emergence is linked to the inflammatory-metabolic state rather than direct drug toxicity. Their presence correlates with higher recurrence rates in colorectal cancer patients and with a broader immunosuppressive niche: exhausted CD8+ T cells co-expressing the transcription factor TOX, and stressed DNAJB1+ natural killer cells. Crucially, genetic deletion of MARCO in mice substantially suppressed liver metastasis growth and reduced the infiltration of exhausted CD8+ TOX+ T cells.

These findings matter beyond oncology. MARCO is a pattern-recognition receptor previously implicated in dampening innate immune responses to pathogens; its role in shaping a pro-metastatic niche in a chemotherapy-injured liver is a meaningful conceptual extension. The exhausted T-cell signature (TOX is a master regulator of T-cell exhaustion) is already a target of checkpoint immunotherapy, and the co-occurrence of MARCO+ macrophages with TOX+ T cells raises the possibility that macrophage-directed therapies could re-sensitize tumors to existing immune checkpoint blockade. Limitations include the relatively small discovery cohort and the translational gap between genetic MARCO ablation in mice and pharmacological targeting in humans. Still, this is more than incremental: it identifies a mechanistically coherent, stage-specific immune axis that could inform both biomarker development and combination therapy design.