Milk thistle failed its best liver trials, yet doctors use it against death caps

Key takeaways

• A 2007 Cochrane review of 13 trials with 915 people found milk thistle did not lower overall deaths from alcohol-related or viral liver disease, and its apparent benefit faded in the better-run trials⁷.
• In an NIH-funded trial of 154 people with hepatitis C, high doses of a standardized extract for 24 weeks worked no better than a placebo³⁸.
• A purified, injectable form of milk thistle’s main compound is approved in Europe for death cap poisoning, but in the U.S. it has been given as an investigational drug, and a poison expert said in 2025 that it can be hard to get fast²⁴¹².
• Small trials suggest milk thistle may lower blood sugar in type 2 diabetes, but the evidence is low quality and mostly from one region³¹³.
• Milk thistle rarely causes side effects, but skip it if you’re allergic to ragweed or daisies, and never use it to put off seeing a doctor about liver symptoms³⁵.

On February 17, 2026, San Francisco’s health department sent local doctors a health advisory. Since mid-November, California had logged 40 cases of poisoning from death cap mushrooms or close relatives. Four people had died, and three needed new livers¹.

One step on the advisory’s treatment list traces back to an old herbal remedy. It told doctors to consider giving silibinin by vein within 24 hours¹. Silibinin is the main active compound in milk thistle.

The milk thistle capsules sold as a “liver tonic” have a much weaker record. In careful U.S. trials, high doses of a standardized extract did no better than a placebo, or dummy pill, for hepatitis C or fatty liver disease⁴⁸¹⁰.

Both facts are true at once. Milk thistle shows how a remedy can be real in one setting and oversold in another.

A thistle with a long liver reputation

Milk thistle (Silybum marianum) grows tall and bears big purple flower heads. It is native to Europe, Asia Minor and North Africa, and early colonists brought it to North America³⁴. Other names include Mary thistle and holy thistle³.

People have eaten it as a food and used it for liver complaints for centuries⁴. A Cochrane team traces milk thistle remedies back to ancient Greece⁷. It was also used to increase breast milk³.

The medicine comes from the “seeds.” Botanically, these are small dry fruits, which is why European regulators list the herb as milk thistle fruit, Silybi mariani fructus⁵.

Extracts of the fruit hold a family of plant compounds known together as silymarin. The main ones are silibinin (also spelled silybin), silychristin and silydianin⁴.

Think of silymarin as the whole team and silibinin as its lead player. That difference matters later, because the hospital drug is purified silibinin, not the capsule in your cabinet.

In stores, you’ll find capsules and tablets, teas and extracts⁴⁵. Capsules usually hold 250 to 750 milligrams (mg) of silymarin, taken two or three times a day⁴.

Europe treats milk thistle as a traditional herbal medicine. In 2018, the European Medicines Agency’s herbal committee accepted it for indigestion, a feeling of fullness and support of liver function, based on long, safe use rather than proof from trials⁵. It added a condition: a doctor should first rule out serious disease⁵.

A 2024 review of 427 newer studies and reports found no reason to change that ruling⁶.

The best-run liver trials came up short

A Cochrane review is an independent summary of all the good trials on one question. In 2007, a Cochrane team pooled 13 randomized trials, where people were assigned by chance to milk thistle or a comparison. The trials covered 915 people with liver disease from alcohol or hepatitis B or C⁷.

Milk thistle did not lower the overall death rate⁷. When every trial was counted, it seemed to cut deaths from liver disease by half. In the better-run trials, that benefit shrank and was no longer clear⁷.

Most of the trials were weak. Fewer than half were properly blinded, meaning patients and doctors may have known who got the herb⁷.

That pattern, where loose studies look good and careful ones don’t, is one of the oldest warning signs in research.

Then came a trial called SyNCH, paid for by the National Institutes of Health (NIH)⁹. At four U.S. medical centers, 154 adults with hepatitis C took part. Earlier interferon-based treatment had failed them⁸.

They took a standardized extract called Legalon at 420 or 700 mg, or a placebo, three times a day for 24 weeks⁸⁹. The extract was made and supplied by the Italian-German drug firm Rottapharm Madaus⁹.

Only two people in each group reached the main goal: a normal level of ALT, a liver enzyme that rises when the liver is inflamed, or a drop of at least half⁸. Virus levels didn’t change either⁴. Blood tests confirmed that people in the herb groups were absorbing the compound⁹.

Fatty liver disease came next. Nonalcoholic steatohepatitis, or NASH, is a fatty liver with inflammation that can lead to scarring.

In a U.S. trial published in 2019 and funded by the federal National Center for Complementary and Integrative Health (NCCIH), 78 adults with NASH took Legalon at 420 or 700 mg, or a placebo, three times a day for 48 weeks³¹⁰.

About 15% to 19% of people on the herb improved by the trial’s main measure, compared with 12% on placebo. The gap was too small to count¹⁰.

A central review later found that 63% of the people enrolled didn’t meet the trial’s own biopsy rules, which muddied the result¹⁰. Scarring improved most in the placebo group, though not by a clear margin¹⁰.

The extract came from Rottapharm Madaus, and one author was affiliated with Rottapharm Biotech¹⁰.

A 2017 Malaysian trial gave 99 adults with NASH 700 mg or a placebo three times a day for 48 weeks¹¹. It also missed its main goal: 32.7% improved on the herb vs. 26.0% on placebo, a difference that could be chance¹¹.

But scarring eased by at least one stage in 22.4% of the herb group, compared with 6.0% on placebo¹¹. The authors called for a larger trial to confirm it, and the trial record we opened lists no funder¹¹.

Two careful trials, then, disagree on the one hopeful sign. That is a reason for a bigger study, not a reason to buy a bottle.

In the hospital, purified silibinin is a different story

Death caps (Amanita phalloides) carry amatoxins, poisons that survive cooking, boiling, drying and freezing¹.

The illness often follows a cruel pattern. Vomiting and watery diarrhea start 6 to 24 hours after the meal. Then the person seems to get better while liver damage builds. Within two to four days, the liver can fail¹.

Amatoxins get into liver cells through a doorway protein called OATP1B3. Silibinin appears to block that doorway. That may stop the toxin from cycling back into the liver again and again¹².

In Europe, purified silibinin given by vein is approved for death cap poisoning⁴. A U.S. study plan says the product is registered for this use in more than a dozen European countries¹².

In the U.S., it has been given as an investigational drug. Starting in 2009, doctors could get it through an open study, run under an FDA investigational new drug application, in which every patient got the drug and no one got a placebo¹².

The 2018 version of that study plan, sponsored by the drugmaker Mylan Specialty, laid out the steps. A doctor called Poison Control or a study hotline, the hospital’s review board approved, and then the drug shipped¹².

The dose depended on body weight: 5 mg per kilogram over an hour, then 20 mg per kilogram a day by steady drip, for up to four days or longer¹². For a 154-pound adult, that works out to about 1,400 mg of silibinin a day.

That study ended in April 2020, and its results were posted in 2021, according to the FDAAA TrialsTracker¹⁵. We could not confirm the drug’s current U.S. status or how hospitals obtain it in 2026.

How well does it work? NIH’s LiverTox database notes that no controlled trials of the IV drug have been published, though open studies reported better survival than expected⁴.

An interim report on the first 50 U.S. patients counted 42 full recoveries, five deaths and three transplants¹². Without a comparison group, it’s impossible to know how many would have lived anyway.

Speed is the other problem. During the 2025 outbreak, Craig Smollin, a University of California, San Francisco emergency doctor who directs the city’s branch of the state poison control system, said silibinin is often tried but can be delayed because it’s hard to get².

If someone may have eaten a wild mushroom, don’t reach for capsules. Call Poison Control at 1-800-222-1222 right away, before symptoms start; they can take 6 to 24 hours to show up¹. Our article on respecting nature covers foraging safety.

Blood sugar is the most interesting lead

A 2016 review in the Journal of Diabetes Research pooled five randomized trials with 270 adults who had type 2 diabetes¹³. They took 200 to 600 mg of silymarin a day for 45 days to six months¹³.

Fasting blood sugar fell about 27 mg/dL more than in comparison groups. HbA1c, a measure of average blood sugar over about three months, fell by about 1 point¹³.

Changes that large from a supplement call for caution. The authors rated the evidence low quality, found that results varied widely from trial to trial, and made no recommendation¹³.

Two of the five trials were funded by drug companies¹³. NCCIH adds that most of this research came from Middle Eastern countries, so it’s unclear whether the results apply elsewhere³.

If you take diabetes medicine, talk with your clinician before adding milk thistle, as NCCIH advises for anyone on medicines³. If you do start it, check your blood sugar more often.

Gentle for most people, with a few firm cautions

Milk thistle itself has a clean liver record. LiverTox gives it an E score, meaning it is an unlikely cause of liver injury⁴. In the Cochrane trials, side effects were no more common than with placebo⁷.

• Common side effects. Bloating, nausea and gas top the list³. European regulators add headache, upset stomach and diarrhea⁵.
• Allergy. Reactions are possible, especially if you’re allergic to ragweed, chrysanthemums, marigolds or daisies³. Europe tells people allergic to the daisy family not to use it, and lists allergic reactions that can include asthma and anaphylaxis⁵.
• Pregnancy and breastfeeding. Little is known about safety³. European regulators advise against it⁵.
• Children and teens. Europe doesn’t recommend it under age 18, because safety hasn’t been shown⁵.
• Drug interactions. In lab dishes, silymarin affects enzymes that process many drugs. LiverTox found little evidence that this matters in people, and Europe lists no known interactions⁴⁵. Tell your clinician and pharmacist anyway³.

The biggest risk is false comfort. A “liver support” capsule can tempt someone to skip the tests that find hepatitis, fatty liver or alcohol damage early. European rules tell milk thistle users to get medical advice right away for yellow skin or eyes or a change in urine or stool color⁵.

What a “liver detox” label can’t tell you

In the U.S., milk thistle is sold as a dietary supplement. The FDA does not approve supplements before they go on sale, and makers are responsible for their own safety and labeling³.

Quality varies. NCCIH reports that some products hold far more or less silymarin than the label says. Some have been contaminated with pesticides, microbes or mycotoxins, which are poisons made by molds³.

“Supports liver health” is a marketing claim, not a test result. Our article on how your body already detoxes itself explains what the liver does on its own every day.

Here is where natural-health values and the evidence line up. Tradition pointed to something real: in lab and animal studies, silymarin protected the liver from toxins, including death cap poison⁴.

Respecting that history means being honest about the rest. A plant can earn trust for one job and still fail at another. The best way to protect your liver is still to know what’s harming it. As the Cochrane team notes, most liver disease comes from alcohol and liver-damaging viruses⁷.

Milk thistle by use: what the evidence shows

Use (form and dose studied)	Evidence rating	What research shows
Hepatitis C (Legalon, 420 or 700 mg three times a day, 24 weeks)	Evidence says no, or unsafe	No better than placebo on liver enzymes or virus48
Alcohol-related or hepatitis B or C liver disease (various oral extracts)	Not enough evidence	No effect on deaths overall; better trials showed no clear benefit7
Fatty liver disease, NASH (420 to 700 mg three times a day, 48 weeks)	Promising but early	Both trials missed their main goal; one saw less scarring1011
Type 2 diabetes (200 to 600 mg a day, up to six months)	Promising but early	Lower blood sugar in five small, low-quality trials313
Death cap poisoning (purified silibinin by IV, hospital only)	Promising but early	Approved in Europe; no controlled trials412
More breast milk	Not enough evidence	Unclear whether it helps3
Indigestion and fullness (teas and extracts)	Not enough evidence	Accepted in Europe on long use, not trials5

Your milk thistle profile card and buying checklist

Profile card	Details
Plant and part	Silybum marianum; the fruit, sold as “seed”45
Key compounds	Silymarin, a mix led by silibinin4
Forms	Capsules, tablets, teas and extracts; IV silibinin is hospital-only45
Doses studied	200 to 600 mg a day for blood sugar; 420 to 700 mg three times a day in liver trials81013
Traditional tea	3 to 5 grams of fruit in 100 milliliters of boiling water, two or three times a day5
Skip it if	You’re allergic to the daisy family, pregnant, breastfeeding or under 1835
Check first if	You take diabetes medicine or any prescription drug3

Before you buy, run through this checklist:

1. Find the Latin name and part. The label should say Silybum marianum and fruit or seed.
2. Look for silymarin in milligrams. Trials measured silymarin, so a plain “milk thistle 1,000 mg” tells you little.
3. Look for an independent quality seal. It checks what’s in the bottle, not whether it works.
4. Walk away from cure talk. Skip any product that promises to detox, reverse or cure liver disease.
5. Get tested first. If you’re worried about your liver, ask for blood tests before you self-treat.
6. Bring the bottle. Show it to your clinician and pharmacist with your other medicines.

Questions readers ask

The bottom line

Milk thistle has earned its place in herbal tradition, and one of its compounds has earned a place in the hospital. Purified silibinin, given by vein, is a treatment health officials tell doctors to consider for death cap poisoning, even though it still lacks controlled trials.

The capsules sold for liver health are another matter. Careful trials in hepatitis C, alcohol-related and viral liver disease, and fatty liver found little or no benefit. The herb is gentle for most people, so the main danger is false comfort. Use it, if you like, as a well-tolerated tradition, and let real liver tests guide your care.

Related reading

• Your body already detoxes itself, and a few everyday habits help it work
• Nature makes both medicine and deadly poison, and respecting it means telling them apart
• What gets into supplements that isn’t on the label
• Why supplement labels say ‘supports’ and never say ‘cures’
• Herbal supplements are a $13 billion business, and herbalists need no license

Sources

1. San Francisco Department of Public Health. “Health Advisory: Mushroom Poisoning.” SFDPH, February 17, 2026. media.api.sf.gov
2. UCSF School of Pharmacy, Department of Clinical Pharmacy. “California Poison Control System Responds to Largest Poisonous Mushroom Cluster in Decades.” UCSF, December 12, 2025. clinicalpharmacy.ucsf.edu
3. National Center for Complementary and Integrative Health. “Milk Thistle.” NIH, updated February 2025. nccih.nih.gov
4. LiverTox: Clinical and Research Information on Drug-Induced Liver Injury. “Milk Thistle.” National Institute of Diabetes and Digestive and Kidney Diseases, updated January 21, 2020. ncbi.nlm.nih.gov
5. European Medicines Agency, Committee on Herbal Medicinal Products. “European Union Herbal Monograph on Silybum marianum (L.) Gaertn., fructus.” EMA/HMPC/294187/2013, adopted June 5, 2018. ema.europa.eu
6. European Medicines Agency, Committee on Herbal Medicinal Products. “Addendum to Assessment Report on Silybum marianum (L.) Gaertn., fructus.” EMA/HMPC/509451/2023, adopted January 31, 2024. ema.europa.eu
7. Rambaldi A, Jacobs BP, Gluud C. “Milk Thistle for Alcoholic and/or Hepatitis B or C Virus Liver Diseases.” Cochrane Database of Systematic Reviews, 2007, Issue 4, CD003620. cochrane.org
8. Healio. “Milk Thistle Showed No Benefit in HCV” (on Fried MW, et al., SyNCH trial, JAMA, 2012;308:274-282). Healio, July 17, 2012. cm.healio.com
9. London S. “Silymarin Flops for Treatment of Chronic Hepatitis C” (SyNCH trial report from the AASLD meeting). MDedge, January 3, 2012. mdedge.com
10. Navarro VJ, Belle SH, D’Amato M, et al. “Silymarin in Non-Cirrhotics With Non-Alcoholic Steatohepatitis: A Randomized, Double-Blind, Placebo Controlled Trial.” PLoS ONE, 2019;14(9):e0221683 (Thomas Jefferson University record). jdc.jefferson.edu
11. Chan WK (Wah Kheong Chan), Nik Mustapha NR, Mahadeva S. “A Randomized Trial of Silymarin for the Treatment of Nonalcoholic Steatohepatitis.” Clinical Gastroenterology and Hepatology, 2017;15(12):1940-1949 (University of Malaya record). eprints.um.edu.my
12. Mylan Specialty LP. “Intravenous Milk Thistle (Silibinin-Legalon) for Hepatic Failure Induced by Amatoxin/Amanita Mushroom Poisoning,” study protocol version 4.0, IND 105960, September 7, 2018. ClinicalTrials.gov NCT00915681. cdn.clinicaltrials.gov
13. Voroneanu L, Nistor I, Dumea R, Apetrii M, Covic A. “Silymarin in Type 2 Diabetes Mellitus: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.” Journal of Diabetes Research, 2016. pmc.ncbi.nlm.nih.gov
14. MedlinePlus. “Liver Diseases.” National Library of Medicine, updated July 27, 2026. medlineplus.gov
15. FDAAA TrialsTracker. “NCT00915681: A Trial That Was Reported Late by Mylan Specialty, LP” (completion April 10, 2020; results submitted August 31, 2021). Data updated October 16, 2026. fdaaa.trialstracker.net