IL-18 Blocks Thymus Recovery After Chemotherapy by Activating NK Cells

The thymus is the body's central factory for producing a diverse, self-tolerant T cell repertoire, but it is acutely vulnerable to damage from chemotherapy, radiation, infection, and stress-induced corticosteroids. After such insults, thymic recovery is slow, leaving patients — especially those undergoing hematopoietic cell transplantation (HCT) — exposed to prolonged T cell lymphopenia, opportunistic infections, and cancer relapse. While some pro-regenerative signals have been identified (IL-22, BMP4, keratinocyte growth factor), no clinically approved therapy exists to accelerate thymic recovery. This study set out to understand what puts the brakes on that recovery.

The research team, based at Fred Hutchinson Cancer Center, used multiple murine models of acute thymic damage: sublethal total body irradiation (SL-TBI, 550 cGy), dexamethasone injection (stress), cyclophosphamide (chemotherapy), and lipopolysaccharide (infection). All four models produced rapid thymic involution and — critically — cleavage of caspase-1 (cl-Cas-1), the enzyme responsible for converting inactive pro-IL-18 and pro-IL-1β into their active, inflammatory forms. ELISA measurements confirmed that active IL-18 levels surged in thymic tissue within 12–72 hours of each injury type. Importantly, IL-18 binding protein (IL-18BP) also rose, but the ratio of free active IL-18 to IL-18BP increased significantly by day 3 post-irradiation, indicating a net excess of bioavailable IL-18 during the early regenerative window.

To establish causality, the team used germline knockout mice. While IL-1R1-deficient mice (Il1r1−/−) showed no improvement in thymic recovery, mice lacking IL-18 (Il18−/−) or its primary receptor (Il18r1−/−) demonstrated significantly greater thymus cellularity 7 days after SL-TBI compared to wild-type controls. Mice lacking the catalytic domain of caspase-1 (Cas1Δ10) showed similar benefit. Conversely, administration of recombinant IL-18 at day 3 post-irradiation — the nadir of thymic cellularity when regeneration normally begins — significantly delayed reconstitution. IL-18 deficiency did not alter cortisol levels, ruling out glucocorticoid-mediated indirect effects.

The mechanistic core of the paper centers on thymic NK cells. Single-cell RNA sequencing and flow cytometry identified a population of mature, cytotoxic NK cells resident in the thymus that highly express the IL-18 receptor (IL-18R1). Following SL-TBI, these NK cells upregulated granzyme B and perforin in an IL-18-dependent manner. Crucially, these activated NK cells were shown to kill thymic epithelial cells (TECs) — both cortical and medullary populations — which are the master regulators of T cell development and thymic regeneration. Depletion of NK cells or ablation of IL-18 signaling protected TECs and restored their numbers during recovery.

Therapeutically, mice receiving an anti-IL-18 monoclonal antibody for 2–3 weeks after allogeneic HCT showed significantly greater thymus cellularity at day 50 post-transplant compared to PBS-treated controls (approximately 1.5–2-fold increase). The study also flags an important clinical consideration: IL-18 is currently being developed as a cancer immunotherapy agent due to its capacity to drive type 1 immune responses and NK/T cell cytotoxicity. These findings raise the possibility that systemic IL-18 administration in oncology settings could inadvertently impair immune reconstitution by suppressing thymic recovery — a meaningful off-target risk to weigh in ongoing clinical trials.