Siblini H, Al-Hendy A, Segars J, et al. Assessing the hepatic safety of epigallocatechin gallate (EGCG) in reproductive-aged women. Nutrients. 2023;15(2):320.
To see if giving women 720 mg of epigallocatechin gallate (EGCG) daily for the duration of their menstrual cycle (at least 30–35 days) would have any effect on their liver enzymes or total bilirubin.
There was no evidence that EGCG had any effect on serum test results used to identify an acute hepatic toxicity reaction (alanine transaminase [ALT], aspartate transaminase [AST], or total bilirubin).
Interventional, randomized, controlled trial
The trial consisted of 39 reproductive-aged women (18–40 years), either with or without uterine fibroids. Ethnicities were 59% White, 12.8% Black, 20.5% Asian, and 2.6% Native Islander with 2 participants (or 5.1%) declining to answer. Thirty-six women finished the study, with 2 voluntarily stopping and 1 lost to follow-up.
The participants were divided into 3 arms: Arm 1 was EGCG only (n=16), arm 2 was EGCG + clomiphene citrate (n=11), and arm 3 was EGCG + letrozole (n=12).
Exclusion criteria included hormonal birth control, alcohol use (more than 14 drinks per week), and liver disease (historical or current). Twelve of the subjects had a diagnosis of uterine fibroids; 2 had a diagnosis of endometriosis.
After initial intake, all subjects started the intervention. The intervention consisted of 4 capsules of 400 mg total green tea extract (GTE) with 45% (180 mg) EGCG, taken once at breakfast every day throughout the study. Between cycle days 2 to 5, arms 2 and 3 received 5 days of clomiphene citrate (100 mg/d) or letrozole (2.5 mg/d), respectively.
Study Parameters Assessed
Each subject in this study had 5 total visits, including the intake visit. Liver function (serum ALT, AST, direct/total bilirubin) and folate were measured at screening as well as at visits 1, 2, and 4. The study looked at adverse events and endometrial thickness using pelvic ultrasound at midcycle.
Primary Outcome the Study Was Designed to Assess
Change in AST/ALT/total and direct bilirubin
There was no evidence of elevation of liver labs that qualified as significant (based on ALT or AST ≥3×upper limit of normal (ULN) or bilirubin ≥2×ULN). There were no differences within any arm or between arms.
This study was funded by several grants, listed as “Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), R01 HD100365 (Johns Hopkins University), R01 HD100367 (University of Chicago and University of Illinois Chicago), R01 HD100369 (Yale University).” None of the authors disclosed any conflicts of interest.
The first obvious limitation of this study is the length of time these women were followed. This isn’t clear in the study, but it appears to be as few as 30 days and as many as 35 to 40 days. The study was designed to capture an acute liver toxicity event, as the researchers were identifying only elevations of ALT and AST 3 times the upper limit of normal and total bilirubin 2 times the upper limit of normal. Many other studies with significantly better trial design and more participants have shown that longer-term follow-up is necessary for evaluation of liver-related data.
The Minnesota Green Tea Trial is one such study. This was a controlled, randomized trial with 1,021 women taking 843 mg of EGCG daily or placebo for 12 months.1
All these individuals had normal baseline ALT/AST levels. After 12 months, in the GTE arm of the study, serum ALT increased by 5.4 U/L and AST increased by 3.8 U/L, significantly higher than in the placebo arm (both P<0.001). In total, 26 (5.1%) of the treated subjects showed moderate or severe abnormalities in liver function tests. Seven of 12 (58%) experienced serious adverse events related to ALT elevations. The odds ratio for developing liver function abnormalities was 7.0 (P=0.0002; 95% CI: 2.4–20.3) compared to placebo. The rise-fall pattern of liver enzyme levels followed a challenge-dechallenge-rechallenge cycle of GTE consumption, with evidence not available in the first 30 days of the study. However, in all of these cases, GTE could not be identified as the sole cause of liver enzyme elevation, as all of them were associated with a simultaneous infection, use of new medication, alcohol consumption, or self-reported past medical history of liver-enzyme elevations.
This 2020 USP review found 25 different pesticide residues in green tea extract samples.
In 2018 Hu et al published a review of green tea adverse events from 159 human interventional studies. Based on the 48 clinical studies assessing liver injury specifically, the incidence of hepatotoxicity was approximately 4.9%, calculated from elevated liver-function biomarkers in 111 events out of 2,269 subjects who were consuming green tea preparations (including green tea, GTE, or individual catechins).2 So clearly this is a problem, at least from some preparations being studied.
The trial currently under review stated that this limited safety study was prompted by a 2016 United States Pharmacopeia (USP) review that resulted in a mandated “cautionary labeling statement” for all manufacturers of powdered decaffeinated green tea extract and follow-up in 2020 by a USP appointed “Green Tea Extract Hepatotoxicity Expert Panel.”3 The panel found there is some concern, noting, “Published adverse event case reports associate hepatotoxicity with EGCG intake amounts from 140 mg to ∼1000 mg/day and substantial inter-individual variability in susceptibility, possibly due to genetic factors.”
The committee reviewed 331 relevant articles, which included clinical trials, preclinical animal research, and 75 cases of human hepatologic toxicity from green tea, catechins, polyphenols (ie, EGCG), and green tea extract. The toxicologic studies identified contamination issues (solvent residues, pesticide residues, pyrrolizidine alkaloids, and elemental impurities) or issues with the high concentrations of EGCG found in the finished products.
The reviewers identified several potential sources of hepatotoxicity. First, for extraction of the catechins from green tea, alcohols or solvents (acetone or ethyl acetate) can be used, resulting in residual material in the finished product. Second, the decaffeination process can involve solvents—chloroform or dichloromethane, both of which may be found in the finished product. Third, the process of extraction has the potential to increase the levels of catechins, pesticide residues, toxic metals, or pyrrolizidine alkaloids in the powdered extract.
One green tea extract, Exolise—an 80% ethanolic dry extract standardized at 25% EGCG catechins—reportedly caused liver injury and was subsequently banned in France and Spain in 2003. Both articles referenced are not available in English, and so I could not access them to determine whether the injury may have been due to high catechin amounts or contaminants in the finished product.
Green tea, as well as white and black teas, have been shown to contain significant organophosphate and organochlorine pesticides, including DDE (dichlorodiphenyldichloroethylene), the metabolite and human adipose storage form of DDT (dichlorodiphenyltrichloroethane). This 2020 USP review found 25 different pesticide residues in green tea extract samples. There was no indication if any of them were grown under USDA organic certification restrictions. While pesticide residues are logically concerning, there were no cases of acute hepatotoxicity from GTE that could be traced directly to pesticide contamination. Pesticide contamination of green and black tea is common, and researchers postulate that extracting the tea leaves to create GTE would extract/concentrate those pesticides as well.4
The 2020 review study identified several studies that noted the accumulation of hepatotoxic metals in green tea leaves, including arsenic, cadmium, chromium, copper, lead, mercury, and manganese. Because none of the case studies looked for the presence of these metals in the products consumed, there is no information about metal contamination in reported cases of liver injury.
Pyrrolizidine alkaloids (PAs) are commonly found as contaminants in green tea due to the incidental coharvesting of neighboring plants with high levels of these alkaloids. According to the 2020 review, PAs are unlikely to be the cause of green tea–hepatotoxicity events. PAs are known to cause hepatic sinusoidal obstruction syndrome (aka hepatic veno-occlusive disease), not the hepatocellular damage seen in the cases related to GTE intake.
Regardless of the inconsistencies of the above studies, it does appear that using organic green tea extracts and whole leaves that have been tested for contamination (metals, pesticides) is a prudent decision. And a conscientious clinician may want to ask the manufacturer for the actual data for that product lot. If the manufacturer cannot provide transparency (from the finished product, not the raw material), it may be advisable to choose one that can.