Editorial: Plants against animals

Many plants have the ability to make toxins in order to protect themselves against plant-eating animals. Over time, herbivores, physiologically and anatomically designed for eating plants, will avoid these poisonous plants. Some plants can be toxic except processed in a certain way. Toxic effects of undercooked kidney beans in France were recently reported, due to phytohaemagglutinin, a plant lectin, not destroyed due to undercooking.¹ Choi and co-workers in the Drug-induced liver injury network (DILIN) published an interesting report on a cohort of patients with acute liver injury due to plant-based food supplement (PBS).² The DILIN has contributed immensely on the knowledge on acute liver injury, due to prescription drugs as well as on Herbal and Dietary Supplements since it was initiated in 2004.³ Some landmark papers have increased our understanding on hepatotoxicity of green tea extract,⁴ garcinia cambogia⁵ and turmeric.⁶ Interestingly, the median latency from ingestion of the PBS in the current study² was only 3 days, whereas the median latency of green tea extract containing supplements was 72 days,⁴ garcinia cambogia 51 days⁵ and turmeric supplements 86 days,⁶ indicating an immune-mediated injury. The short latency of the PBS suggests, as the authors point out, a toxic effect characterised by abdominal pain (59%), nausea, fever, itching and rash (12%) but diarrhoea was not reported. Approximately, 50% had jaundice at presentation, 24% hospitalised but no serious consequences were observed. Interestingly, the ALT level at presentation was the peak value in all except one of the 17 patients, suggesting a very rapid resolution of liver injury and liver enzymes normalised in all patients. The quick resolution is very unusual for idiosyncratic DILI, which often takes weeks or months to resolve, but rather resembles liver injury associated with hypoxic hepatitis (HH).⁷ In HH, the recovery of ischaemic/hypoxic nature of the injury, after stabilisation of the patient is related to the half time of ALT which is 2–3 days and AST 1–2 days. The authors speculated that the liver injury was due to an oxidant injury, related to the metabolism of baikiain found in Tara flour, into reactive metabolites, similar to hepatotoxicity associated with acetaminophen. The only liver biopsy seems to point towards another type of injury which revealed an acute hepatitis, mild ductular reaction, mild lymphocytic and eosinophilic portal inflammation.² This is perhaps not highly suggestive of oxidant injury and rather points towards allergic reaction, with eosinophilic portal inflammation as well as itching and rash in symptomatic patients. The presence of eosinophilia has been found to be associated with a favourable prognosis in patients with DILI^8-10 as was the case in the current report.² In summary, the current paper by Choi et al. is well written and the study and its results are very interesting. Chemical analysis of the samples obtained with thorough exclusions of various hepatotoxic agents and finding Tara flour, adds important scientific flavour to the study.

Einar S. Björnsson: Conceptualization; methodology; writing – original draft.

The author has no conflicts of interest.

This article is linked to Choi et al paper. To view this article, visit https://doi.org/10.1111/apt.18116