Silymarin (Silybum marianum (L.) Gaertn. extract) is a widely used botanical for liver disease, yet clinical results remain inconsistent. Most mechanistic work uses supraphysiological aglycones, whereas humans are exposed predominantly to phase II conjugates that are strongly protein-bound and routed by transporters toward bile and the intestinal mucosa. We reframe silymarin activity through a spatial pharmacology lens, proposing three post-intake windows: early (0–2 h) conjugate-dominant exposure with localised β-glucuronidase-mediated reactivation; intermediate (2–8 h) enterohepatic recirculation pulses; and late (8–48 h) microbial catabolite contributions. Each window engages distinct signalling modules—Keap1/NRF2, NF-κB, and AMPK-mTOR-TFEB—via transient redox events (quinone cycling, micro-H2O2 relays) and proteostatic remodelling (autophagy/mitophagy). We synthesise human pharmacokinetic and clinical evidence—with emphasis on MASLD and alcohol-associated liver disease—and show how formulation, meal timing, and microbiome metabotype determine which windows are engaged. Finally, we propose minimum reporting standards and falsifiable hypotheses to reduce between-study heterogeneity and enable precision use of silymarin.
Lastra et al. (2026) studied this question.