Secondary Logo

Journal Logo


Green Tea Extract to Prevent Colorectal Adenomas, Results of a Randomized, Placebo-Controlled Clinical Trial

Seufferlein, Thomas MD1,*; Ettrich, Thomas J. MD1,*; Menzler, Stefan MD2; Messmann, Helmut MD3; Kleber, Gerhard MD4; Zipprich, Alexander MD5; Frank-Gleich, Stefanie MD6; Algül, Hana MD7; Metter, Klaus MD8; Odemar, Frank MD9; Heuer, Theodor MD10; Hügle, Ulrich MD11; Behrens, Rüdiger MD12; Berger, Andreas W. MD1; Scholl, Catharina PhD13; Schneider, Katharina L. MD13; Perkhofer, Lukas MD1; Rohlmann, Friederike BSc14; Muche, Rainer PhD14; Stingl, Julia C. MD15

Author Information
The American Journal of Gastroenterology: June 2022 - Volume 117 - Issue 6 - p 884-894
doi: 10.14309/ajg.0000000000001706
  • Free
  • SDC
  • CME Test
  • CME



Adenomas are precursor lesions of colorectal cancers (CRC). Removal of adenomas during colonoscopy reduces CRC incidence and CRC-related death (1–3). Nonsteroidal anti-inflammatory drugs (NSAID), such as aspirin or COX2 inhibitors (4–6), have a moderate but significant effect on the formation of new polyps. However, they carry the risk of gastrointestinal and genitourinary bleeding (7) or cardiovascular events, which is unfavorable in a preventive setting (8,9). Metformin is chemopreventive in nondiabetic adults but can also have considerable side effects (10). Weighing up the advantages and disadvantages, there is currently no drug recommended for chemoprevention of CRC in an average-risk population (11). Nutritional compounds that are efficacious in preventing colorectal adenomas but have at worst minimal side effects would be desirable for chemoprevention.

Green tea from leaves of Camellia sinensis (L.) Kuntze, the Chinese variant of the tea plant, is a worldwide consumed beverage. It contains 3%–6% caffeine and up to 30% catechins, mainly (−)-epigallocatechin-3-gallate (EGCG), (−)-epigallocatechin (EGC), (−)-epicatechin-3-gallate, and (−)-epicatechin (12). In particular, EGCG has preventive and possibly therapeutic effects against various cancers (13–16). It affects numerous intracellular signaling pathways (17) and acts on various hallmarks of cancer by inducing apoptosis and cell cycle arrest, modulating DNA methylation and promoting DNA repair (17). Both green tea and EGCG substantially reduce the incidence of polyps in adenomatous polyposis colimin mice by targeting the initial stages of carcinogenesis (18). In man, meta-analyses of epidemiological studies are conflicting regarding adenoma and CRC prevention by green tea (19,20). A more recent meta-analysis concludes that tea consumption has an inverse impact on CRC risk, which may have significant public health implications for the prevention of CRC (21). Two small prospective Asian trials showed a significant and remarkable 16% (22) and 18.7% (23) reduction of adenoma frequency at 12 months, respectively. Regarding green tea, the Shanghai Men's Health study reported that regular green tea consumption (defined as “ever drank green tea at least 3 times per week for more than 6 consecutive months”) was associated with a reduced risk of CRC in nonsmokers (multivariable-adjusted hazard ratio 0.54, 95% confidence interval [CI] [0.34–0.86]) (24). However, larger prospective randomized trials are so far not available.

This study, the MIRACLE (Minimizing the Risk of Metachronous Adenomas of the Colorectum with Green Tea Extract, NCT01360320) trial examined the effect of 150 mg of green tea extract (GTE) b.i.d. compared with that of placebo taken over 3 years on the incidence of metachronous colorectal adenomas in a large White CRC screening population.


Ethics statement

The trial was conducted in compliance with the Declaration of Helsinki. The protocol was approved by the responsible ethics committees of the participating centers. All patients provided written informed consent before trial entry. The trial is registered with (NCT01360320) and was funded by a grant from the German Cancer Aid.

Study design

MIRACLE is a prospective, randomized, double-blind, placebo-controlled, multicenter trial conducted to prevent colorectal adenomas by comparing the effect of orally taken GTE with that of placebo. The trial involved 40 German centers participating in the quality-assured National German CRC colonoscopy screening program (see Supplementary Table 1, Supplementary Digital Content 1, Recruitment was from November 2011 until June 2015. The trial raised initially great interest and many centers applied to participate. However, quite a few centers did not meet their promised recruitment targets and had to be replaced. As a consequence, the statistical analysis plan also had to be modified and the number of participants had to be reduced over the course of the trial.

Recruitment, randomization, and treatment

Eligible participants were aged between 50 and 80 years and had ≥1 histologically confirmed colorectal adenoma(s) removed within 6 months before recruitment during a complete colonoscopy with no remaining colorectal adenomas. Exclusion criteria were history of familial CRC syndromes, CRC, and other concomitant cancers unless curatively treated, intestinal malabsorption, inflammatory bowel disease, compromised liver function, regular intake of NSAID >3 months per year except low-dose aspirin (LDA, ≤100 mg/d), or intake of GTE >100 mg/d (approximately 500 mL of green tea per day) over at least 6 months in the past 2 years.

Study capsules with decaffeinated GTE (containing all catechins from tea leaves but standardized to contain 150 mg of EGCG) or placebo were manufactured by Dr. LOGES + Co. GmbH/Winsen, Germany, a pharmaceutical company experienced in the production of GTE. The tea leaves (Camellia sinensis (L.) Kuntze) used for extraction in the MIRACLE trial originated from Japan. After informed consent, all participants entered a 4-week run-in period with GTE to assess its tolerance and potential side effects. Liver enzymes were assessed at the beginning and at the end of the run-in phase and subsequently at months 4, 12, 20, and 36, respectively. Participants with no major treatment-related side effects, no elevated liver enzymes (aspartate aminotransferase, alanine aminotransferase, and bilirubin) >2.5 times upper limit of normal, and intake of at least 65% of the capsules were randomized by a central permutated double-blind block randomization stratified according to intake of LDA and study center.

Participants recorded intake of capsules, concomitant medication, and adverse events (AE) in a diary over 36 months. At the 4-monthly visit, this information was recorded, followed by a brief clinical examination and a pill count. Participants underwent a surveillance colonoscopy approximately 36 months after the qualifying examination. Each lesion was recorded, removed, and subsequently histologically examined by the local pathologist. Advanced adenoma was defined as at least 1 adenoma >1 cm, villous, tubulovillous, high-grade dysplasia, and/or invasive; nonadvanced adenoma was defined as tubular adenoma ≤1 cm. Sessile serrated adenomas were classified by pathologists as follows: lesions showing horizontal growth along the muscularis mucosae and dilation of crypt base (basal third of the crypt) and/or serrations extending into the crypt base. The assessment of BRAF mutations and KRAS mutations was not part of the analysis. Endoscopists fulfilled the regularly audited criteria for performing screening colonoscopies of the quality assured National German CRC screening program. These include at least 200 “total colonoscopies” per endoscopist with a color image proving that the cecum and/or the ileocecal valve have been reached and 50 polypectomies within 2 years before applying for participation in the program.

End points

The primary end point of the study was the proportion of participants with 1 or more colorectal adenomas detected at the follow-up colonoscopy within the 3-year observation period. Preplanned exploratory analyses of the primary outcome included sex and LDA intake.

Secondary endpoints were number of carcinomas, serrated adenomas, and advanced lesions. Separate explorative analyses were conducted for localization of lesions.

Statistical analysis

The sample size estimation was derived from the existing chemoprevention studies reporting an incidence of metachronous adenomas of approximately 45% at the follow-up colonoscopy. Using an absolute risk reduction of 8.8% for the incidence of metachronous adenomas by GTE as described by Baron et al. (7) for the treatment with aspirin (47.1% in the placebo group vs 38.3% in the treatment group) and taking the statistical error rates (significance level 5%, power 80%) in a 1-sided χ2 test, a total of 780 (2 × 390) participants should build the modified intention-to-treat (mITT) sample defined as all randomized participants with informed consent (ITT population) and with a follow-up colonoscopy 26–44 months after randomization. Assuming a dropout rate of 11% in the run-in phase and 15% missing follow-up colonoscopies, 1,032 participants had to be recruited into the run-in phase and 918 (2 × 459) participants should enter the randomized study phase. Recruitment had to be finished slightly prematurely, i.e., after recruitment of 1,001 participants into the run-in phase due to slow accrual. The sample size calculation of the original study protocol (25) was amended in 2015 while the study was still in the process of recruiting.

Analysis sets.

The efficacy analysis was based on the mITT data set that comprised all randomized participants with informed consent (ITT population) and with a follow-up colonoscopy 26–44 months after randomization because it was awaited that the follow-up colonoscopies were performed according to local practice and not always performed precisely as advised by the protocol. If several follow-up colonoscopies were performed in 1 patient, always the result of the first colonoscopy with the finding of an adenoma and always the last colonoscopy report if there was no adenoma reported was used. In addition, we included participants undergoing colonoscopy within 26 months after randomization and a finding of an adenoma or a colonoscopy beyond 44 months after randomization if there was no finding of an adenoma. Follow-up colonoscopies performed within the first year after the initial colonoscopy were not included because we considered this time frame as too short. The per-protocol (PP) population comprised all mITT participants without major protocol violations, with ≥65% intake of the trial capsules and at least 28 months of treatment. Major protocol violations were as follows: (i) no adenomas in the initial colonoscopy (n = 11); (ii) regular intake of concomitant medication, such as NSAID, COX2 inhibitors, or aspirin >100 md/d, not allowed according to the study protocol (n = 15); (iii) nonadherence (i.e., less than 65% intake of study medication in a 4-month period; n = 37); (iv) follow-up colonoscopy with an adenoma within 1 year after the initial colonoscopy and less than 26 months after randomization (n = 21); (v) follow-up colonoscopy with an adenoma more than 6 months after regular end of study medication (n = 3). See Figure 1 for the CONSORT flowchart.

Figure 1.:
CONSORT flowchart. ITT, intention to treat; PP, per protocol.

Analysis methods.

The confirmatory analysis for the primary outcome of metachronous adenomas was performed by a 1-sided χ2 test at a significance level of 5% by comparing the risks in both groups. Furthermore, the risk difference and the relative risk (RR) and their respective 1-sided 95% CIs are given. A stratified Cochran-Mantel-Haenszel test adjusting for LDA yielded a similar result and is not shown. In addition, a multiple log-binomial mixed model (26) was performed adjusting for LDA, age, sex, and center (random factor) as cofactors. The resulting adjusted RR (aRR) (incl. a 1-sided 95% CI) and the respective 1-sided P value were presented and interpreted in an exploratory fashion. Both analyses were also performed for the PP data set. To check for selection bias, the mITT and PP set baseline characteristics were compared between run-in, randomization, mITT and PP collectives in total, for mITT and PP set by treatment, and for the subgroup analyses. There were no relevant differences in their distribution. Other clinical end points and subgroups were compared in an explorative manner using the methods described earlier and a 2-sided 5% significance level considering the multiple testing problem. Safety analyses were presented by absolute and relative frequencies of AE. All authors had access to the study data and reviewed and approved the final manuscript.


Eight hundred seventy-nine of 1,001 participants in the run-in phase underwent randomization (447 placebo and 432 GTE). Of the 122 participants not undergoing randomization, 40 reported AE potentially linked to the study drug. Another 247 participants had to be excluded mainly because of a missing follow-up colonoscopy. The final mITT population comprised 632 participants (GTE: 309, Placebo: 323, see CONSORT diagram, Figure 1). Baseline characteristics were well balanced between the groups (Table 1). There were slightly more past smokers, less physically active participants, and more participants with high-grade dysplasia adenomas (6.1% vs 3.1%), villous adenomas (2.8% vs 0.2%), and sessile serrated adenomas (16.1% vs 15.0%) in the GTE group. In the placebo group, more participants exhibited tubulovillous adenomas (21.7% vs 19.1%) in their index colonoscopy.

Table 1.:
Baseline characteristics of the participants

Reported overall adherence to the study protocol was very good (see Supplementary Table 1, Supplementary Digital Content 1, Approximately 97.9%, 94.9%, and 92.6% of the mITT population took ≥65% of the study drugs in the first, second, and third years, respectively. Approximately 97.6% of participants in the mITT population avoided comedication that was not permitted by the protocol.

The GTE safety profile was overall favorable (Table 2). Most of the AE were gastrointestinal disorders. Elevations of transaminases were detectable in 0.5% of the participants in the run-in phase with 0.1% grade 3 elevations. During the randomized phase, there were no major differences between both groups except for laboratory investigations with more grade 1/2 elevations of alanine aminotransferase (2.9% vs 0.9%) and aspartate aminotransferase (2.7% vs 0.2%) in the GTE group.

Table 2.:
Common Adverse Events, all grades according to CTC-AE, Version 4.0

Colorectal adenomas in the follow-up colonoscopy

The primary end point of the study was the proportion of participants in the mITT population with 1 or more colorectal adenoma(s) detectable at the follow-up colonoscopy. The time intervals between start of application of GTE or placebo to the final follow-up colonoscopy were not different in both groups (GTE: mean 34.35 (SD 8.0) months, placebo: mean 34.66 (SD 7.7) months). Adenomas were detected in 55.7% of the placebo (180/323) and 51.1% of the GTE groups (158/309), respectively, corresponding to a RR reduction of 8.25% and an absolute difference of 4.6% in favor of GTE. This difference was not statistically significant (aRR 0.905, 95% CI [0.788–1.041], P = 0.161; 1-sided 95% CI [- to 1.018], 1-sided P = 0.081). The respective figures for the PP population were 54.3% (151/278) in the placebo and 48.3% (129/267) in the GTE groups, with a relative reduction of 11% and an absolute difference of 6%. This difference was also not statistically significant (aRR 0.883; 95% CI [0.755–1.032], P = 0.117; 1-sided 95% CI [- to 1.006], P = 0.058; Table 3). The mean number of adenomas in the initial colonoscopy was 2.04 in the GTE and 2.13 in the placebo group. In the follow-up colonoscopy, the mean number of adenomas was 1.12 in the GTE and 1.19 in the placebo group. Additional data regarding adenomas detected according to trial center and sex at baseline are summarized in the Supplementary Tables 2 and 3 (see Supplementary Digital Content 1,

Table 3.:
Metachronous adenomas in the follow-up colonoscopy

Adenoma recurrence according to sex and LDA use

The adenoma rate in female participants was 47.9% (58/121) in the placebo and 47.6% (50/105) the GTE groups (aRR 1.019; 95% CI [0.771–1.347]; P = 0.894; Table 3 and Figure 2). The corresponding figures for the PP population were 45.4% (44/97) in the placebo and 46.9% (45/96) in the GTE groups, respectively (aRR 1.014; 95% CI [0.748–1.373]; P = 0.930). Approximately 60.4% (122/202) of male participants had at least 1 adenoma in the follow-up colonoscopy in the placebo and 52.9% (108/204) in the GTE groups (aRR 0.846; 95% CI [0.717–0.999]; P = 0.048; Table 3 and Figure 2). Thus, GTE intake was associated with a significant, 12.4% relative, and 7.5% absolute reduction of metachronous adenomas in male participants. The figures for the male PP population were 59.1% (107/181) in the placebo and 49.1% (84/171) in the GTE groups, i.e., GTE intake correlated with a 10% absolute reduction in adenoma formation in the male participants (aRR 0.803, 95% CI [0.666–0.969]; P = 0.022) (Table 3 and Figure 2).

Figure 2.:
Adenomas in the follow-up-colonoscopy. (a) Metachronous adenomas in the follow-up colonoscopy. Adenomas were detected in 55.7% of the placebo and 51.1% of the GTE groups with a RR reduction of 8.25% and an absolute difference of 4.6% in favor of GTE (aRR 0.905, 95% CI [0.788–1.041], P = 0.161). Adenomas were detected in 60.4% of male participants in the placebo and 52.9% of male participants in the GTE groups (aRR 0.846; 95% CI [0.717, 0.999]; P = 0.048) with a statistically significant risk reduction (12.4% relative and 7.5% absolute). (b) Adenoma recurrence rates according to histology. aRR, adjusted relative risk; CI, confidence interval; GTE, green tea extract; mITT, modified intention-to-treat.

Approximately 18.0% (58/323) in the placebo and 15.5% (48/309) in the GTE groups reported a regular use of LDA (≤100 mg/d) at randomization. At least 1 adenoma was detected in 60.3% (35/58) of participants in the LDA placebo and 52.1% (25/48) in the LDA GTE mITT groups, respectively, in the follow-up colonoscopy. This corresponded with an absolute risk reduction of 8.2% in favor of GTE (aRR 0.813; 95% CI [0.575–1.150]; P = 0.239; Table 3). The respective figures for the PP population were 59.6% (28/47) in the LDA placebo and 47.6% (20/42) in the LDA GTE groups (aRR 0.712; 95% CI [0.462–1.099]; P = 0.124). The percentage of participants with colorectal adenomas in the follow-up colonoscopy was not lower in the LDA placebo group than in the placebo alone group, but absolute numbers of patients were rather low (mITT: 60.3% LDA placebo vs 55.7% placebo alone) (Table 3).

Histology and localization of adenomas

Ideally, chemoprevention targets advanced adenomas bearing the highest risk of progressing to cancerous lesions (27). In our study, the percentage of participants with at least 1 advanced adenoma in the follow-up colonoscopy was 11.6% (36/311) in the placebo and 11.2% (33/294) in the GTE groups (mITT; aRR 0.996; 95% CI [0.643–1.544]; P = 0.986; Table 3). For nonadvanced adenomas/low-grade adenomas (i.e., tubular adenomas ≤1 cm), the corresponding figures were 37.6% (117/311) in the placebo and 32.4% (96/296) in the GTE groups (aRR 0.831; 95% CI [0.672–1.028]; P = 0.088). A similar pattern was observed in the male subgroup: There was no difference between the placebo and the GTE groups regarding advanced adenomas (P: 10.8% [21/194]; GTE: 11.3% [22/194]; aRR 1.061, 95% CI [0.609–1.848]; P = 0.835) or serrated adenomas (P: 7.8% [25/322]; GTE: 6.8% [21/308]; aRR 0.864, 95% CI [0.496–1.503]; P = 0.603). By contrast, nonadvanced adenomas/low-grade colorectal adenomas were significantly reduced in the male GTE subgroup: 34.4% (67/195) vs 43.8% (85/194) in the male placebo group (aRR 0.752; 95% CI [0.587–0.964]; P = 0.025). In the female population, there were no differences between the groups independently of the adenoma histology. Thus, GTE/EGCG intake had no effect on advanced adenomas in the study population including the male subgroup. There were numerically more adenomas with high-grade intraepithelial neoplasia/CRC in the placebo compared with those in the GTE group in the follow-up colonoscopy: 1.9% (6/322) vs 1.0% (3/308; Table 3 and Figure 2). Among these, there were 2 CRC detected in the placebo (2/322; 0.6%) and 1 in the GTE groups (1/308; 0.3%).

Regarding adenoma localization, 34.7% of participants in the GTE and 38.1% in the placebo group had adenomas located on the right side in their follow-up colonoscopy (mITT; P: 123/323; GTE: 107/308; aRR 0.906; 95% CI [0.743–1.106]; P = 0.331). The figures for left-sided adenomas were 30% in the placebo and 27% in the GTE group (P: 97/323; GTE: 83/308; aRR 0.907; 95% CI [0.711–1.157]; P = 0.431). These data do not suggest an effect of GTE on new adenoma formation in the left or right colon, respectively.


GTE and EGCG have been proposed for chemoprevention of metachronous colorectal adenomas (22,23). In this study, the proportion of participants with adenomas after 3 years of exposure to GTE or placebo was 51.1% in the GTE and 55.7% in the placebo groups, respectively. This 4.6% difference in favor of GTE was not statistically significant. The PP analysis did also not show a significant effect of GTE on new adenoma formation in the whole study population. This study is the largest trial so far examining the effect of GTE taken b.i.d. for 3 years compared with that of placebo on new adenoma formation, and it is the first trial to examine the chemopreventive effect of GTE in a White CRC screening population. Smaller Asian studies reported a more than 18% reduction in colorectal adenoma recurrence by GTE/EGCG (22,23), which could not be demonstrated in this study even with a higher dose of EGCG given b.i.d. over a longer period of time than in these studies (22,23). This study is a randomized trial and has a substantially larger sample size than the Asian trials, but there may also be ethnogenetic differences regarding the efficacy of EGCG on adenoma recurrence. The number of participants recruited for this trial compared well with other trials investigating the role of specific compounds on secondary prevention of colorectal polyps. The follow-up interval of 3 years has been used in various chemoprevention trials and was the recommended follow-up interval when the study was designed and was part of the regular surveillance algorithm in Germany. Approximately 25% of the randomized participants did not undergo follow-up colonoscopy in line with published figures (4,6). However, most of the participants received their follow-up colonoscopy between 26 and 44 months after randomization and were exposed to GTE or placebo, respectively, for the scheduled time.

An interesting aspect of this study was the prespecified analysis according to sex showed a 7.5% difference in adenoma recurrence rate in the male GTE group, which was statistically significant (GTE: 52.9% vs 60.4% in the placebo group; 2-sided P = 0.048; aRR 0.846). However, the effect of GTE/EGCG on adenoma recurrence in male participants was limited to tubular adenomas but could not be shown for advanced adenomas. This appears in line with data in adenomatous polyposis colimin mice suggesting that GTE/EGCG act early in colorectal carcinogenesis but are not efficient in preventing adenoma progression (18). In the female study group, there was no difference at all between GTE and the placebo groups (P: 47.9%, GTE: 47.6%; 2-sided P = 0.894). The reason for the sex difference observed in our study is currently unclear. However, assuming that the chemopreventative effect is cumulative over time and assuming a type II error for the female GTE would most likely have to be given over many years for a clinically meaningful decrease of CRC risk.

It has been shown that the local bioavailability of GTE in the intestinal epithelium contributes to its cancer protective effect (28). Local bioavailability of GTE catechins at the intestinal epithelium may be influenced by estrogens that inhibit efflux transporters such as the multidrug resistance–related proteins (MRP transporter) in the intestinal epithelium, which confer resistance to a range of natural chemoactive substrates by higher drug excretion through the intestinal barrier (29,30). In our nutrikinetic study on the MIRACLE-GTE preparation, we found a 55% lower volume of distribution of EGC in female participants with use of oral contraceptives compared with that in male participants, suggesting a higher concentration of EGC in plasma of female participants (31). Lower efflux transport activity at the physiological barriers may lead to lower substrate excretion to the intestine and thus less local bioavailability of green tea catechines in the intestinal lumen (and higher plasma concentrations). This may well explain both the lower effects of green tea intake on adenoma occurrence observed for female participants in the MIRACLE trial and the lower volume of distribution of EGC in female participants observed in our nutrikinetic study. Other data also support the idea of a general sex-specific effect of green tea, e.g., on cognitive function in male individuals, but not in female individuals (32).

Safety is important for a chemopreventive agent taken over a longer period of time. We used a daily dose of GTE standardized to contain 300 mg of EGCG, a dose that was slightly higher than in other chemoprevention trials but corresponded to the upper limit of the mean daily intake of EGCG resulting from the consumption of green tea (33). The European Food Safety Agency EFSA states that the mean daily intake of EGCG resulting from the consumption of green tea beverages ranges from 90 to 300 mg/d (33). There was only 0.1% grade 3 liver enzyme elevation in the nonrandomized run-in phase and no difference in grade 3/4 side effects between placebo and GTE in the randomized phase of the trial. We observed only grade 1/2 elevations in liver enzymes in the GTE group compared with those in the placebo group, demonstrating a good safety profile (10,27). However, due the limited follow-up period of 3 years, long-term safety of GTE cannot be delineated from our data.

It is known that the proportion of participants experiencing AE, particularly hepatotoxicity, increases with higher doses of EGCG such as 800–1,200 mg/d, which is used for the treatment of, e.g., amyloidosis (34). However, in this therapeutic setting, treatment duration is generally shorter.

In conclusion, GTE/EGCG intake over 3 years was safe but had no significant effect on colorectal adenoma recurrence in the whole population of a nationwide CRC screening program. In a subgroup analysis, male but not female participants seemed to benefit from GTE. The reasons for this sex difference are unclear at the moment and should be addressed by future research.


Guarantor of the article: Thomas Seufferlein, MD.

Specific author contributions: T.S., T.J.E., J.C.S., and R.M.: involved in the design and development of the trial. T.S. and T.J.E.: led and coordinated the study. T.S., T.J.E., S.M., H.M., G.K., A.Z., S.F.-G., H.A., K.M., F.O., T.H., U.H., R.B., C.S., K.L.S., A.W.B., and L.P.: recruited patients and/or contributed to data collection. F.R., R.M., T.S., T.J.E., and J.C.S.: analyzed data that were interpreted by all authors. R.M. and F.R.: responsible for the statistical analysis. T.S., T.J.E., R.M., L.P., and J.C.S.: drafted the manuscript with input from all other authors. All authors have seen and approved the final report.

Financial support: The trial was fully funded by a grant from the German Cancer Aid (No. 109122). The funding party had no influence on data collection, analysis, or interpretation; trial design; patient recruitment; or any aspect pertinent to the study.

Potential competing interests: None to report.

Data transparency statement: Additional data are available in the supplementary material. The final study protocol (Vs. 6) and deidentified data sets that underlie the reported results are available for a period of 5 years after the publication date on reasonable request. Proposals for access should be sent to [email protected].

Study Highlights


  • ✓ Adenomas are common colonic lesions that are precursors of colorectal cancers.
  • ✓ Nonsteroidal anti-inflammatory drugs such as aspirin or COX2 inhibitors have a moderate effect on new adenoma formation but carry the risk of gastrointestinal and genitourinary bleeding or serious cardiovascular events.
  • ✓ Epigallocatechingallate (EGCG)/green tea extract (GTE) have multiple biological effects including induction of apoptosis and modulation of DNA methylation.
  • ✓ Small clinical trials have suggested a chemopreventive effect of EGCG.


  • ✓ To the best of our knowledge, this is the largest randomized trial so far examining the effect of GTE on adenoma recurrence in a White colorectal cancer screening population.
  • ✓ GTE 150 mg b.i.d. over 3 years was well tolerated with a virtually absent side effect profile.
  • ✓ There was no statistically significant difference in the chemopreventive effect on colorectal adenoma recurrence between GTE and placebo in the whole study population.
  • ✓ A preplanned subgroup analysis revealed a significant difference in the adenoma recurrence rate in favor of GTE exclusively in the male study population pointing to a potential sex-specific difference in chemoprevention that warrants further investigations.


We thank all participants and their families, the participating institutions, the German Cancer Aid for funding, the KKS of the Martin-Luther-University Halle-Wittenberg (legal sponsor), and the “Stiftung Lebensblicke” for support.


1. Atkin WS, Morson BC, Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med 1992;326:658–62.
2. Zauber AG, Winawer SJ, O'Brien MJ, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012;366:687–96.
3. Winawer SJ, Zauber AG, O'Brien MJ, et al. Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. The National Polyp Study Workgroup. N Engl J Med 1993;328:901–6.
4. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355:885–95.
5. Baron JA, Sandler RS, Bresalier RS, et al. A randomized trial of rofecoxib for the chemoprevention of colorectal adenomas. Gastroenterology 2006;131:1674–82.
6. Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the prevention of sporadic colorectal adenomas. N Engl J Med 2006;355:873–84.
7. Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med 2003;348:891–9.
8. Drazen JM. COX-2 inhibitors—A lesson in unexpected problems. N Engl J Med 2005;352:1131–2.
9. Zell JA, Pelot D, Chen WP, et al. Risk of cardiovascular events in a randomized placebo-controlled, double-blind trial of difluoromethylornithine plus sulindac for the prevention of sporadic colorectal adenomas. Cancer Prev Res 2009;2:209–12.
10. Higurashi T, Hosono K, Takahashi H, et al. Metformin for chemoprevention of metachronous colorectal adenoma or polyps in post-polypectomy patients without diabetes: A multicentre double-blind, placebo-controlled, randomised phase 3 trial. Lancet Oncol 2016;17:475–83.
11. Whitlock EP, Williams SB, Burda BU, et al. Aspirin use in adults: Cancer, all-cause mortality, and harms. A systematic evidence review for the U.S. Preventive Services Task Force. Evidence Synthesis No. 132. AHRQ Publication No. 13-05193-EF-1 2015.
12. Balentine DA, Wiseman SA, Bouwens LC. The chemistry of tea flavonoids. Crit Rev Food Sci Nutr 1997;37:693–704.
13. Johnson R, Bryant S, Huntley AL. Green tea and green tea catechin extracts: An overview of the clinical evidence. Maturitas 2012;73:280–7.
14. Yuan JM. Cancer prevention by green tea: Evidence from epidemiologic studies. Am J Clin Nutr 2013;98:1676S–1681S.
15. Kanwar J, Taskeen M, Mohammad I, et al. Recent advances on tea polyphenols. Front Biosci (Elite Ed) 2012;4:111–31.
16. Rahmani AH, Al Shabrmi FM, Allemailem KS, et al. Implications of green tea and its constituents in the prevention of cancer via the modulation of cell signalling pathway. Biomed Res Int 2015;2015:925640.
17. Khan N, Afaq F, Saleem M, et al. Targeting multiple signaling pathways by green tea polyphenol (-)-epigallocatechin-3-gallate. Cancer Res 2006;66:2500–5.
18. Ju J, Hong J, Zhou JN, et al. Inhibition of intestinal tumorigenesis in Apcmin/+ mice by (-)-epigallocatechin-3-gallate, the major catechin in green tea. Cancer Res 2005;65:10623–31.
19. Sun CL, Yuan JM, Koh WP, et al. Green tea, black tea and colorectal cancer risk: A meta-analysis of epidemiologic studies. Carcinogenesis 2006;27:1301–9.
20. Wang ZH, Gao QY, Fang JY. Green tea and incidence of colorectal cancer: Evidence from prospective cohort studies. Nutr Cancer 2012;64:1143–52.
21. Chen Y, Wu Y, Du M, et al. An inverse association between tea consumption and colorectal cancer risk. Oncotarget 2017;8:37367–76.
22. Shimizu M, Fukutomi Y, Ninomiya M, et al. Green tea extracts for the prevention of metachronous colorectal adenomas: A pilot study. Cancer Epidemiol Biomarkers Prev 2008;17:3020–5.
23. Shin CM, Lee DH, Seo AY, et al. Green tea extracts for the prevention of metachronous colorectal polyps among patients who underwent endoscopic removal of colorectal adenomas: A randomized clinical trial. Clin Nutr 2018;37:452–8.
24. Yang G, Zheng W, Xiang YB, et al. Green tea consumption and colorectal cancer risk: A report from the Shanghai Men's Health Study. Carcinogenesis 2011;32:1684–8.
25. Stingl JC, Ettrich T, Muche R, et al. Protocol for minimizing the risk of metachronous adenomas of the colorectum with green tea extract (MIRACLE): A randomised controlled trial of green tea extract versus placebo for nutriprevention of metachronous colon adenomas in the elderly population. BMC Cancer 2011;11:360.
26. Greenland S. Model-based estimation of relative risks and other epidemiologic measures in studies of common outcomes and in case-control studies. Am J Epidemiol 2004;160:301–5.
27. Hull MA, Sprange K, Hepburn T, et al. Eicosapentaenoic acid and aspirin, alone and in combination, for the prevention of colorectal adenomas (seAFOod Polyp Prevention trial): A multicentre, randomised, double-blind, placebo-controlled, 2 x 2 factorial trial. Lancet 2018;392:2583–94.
28. Hao X, Sun Y, Yang CS, et al. Inhibition of intestinal tumorigenesis in Apc(min/+) mice by green tea polyphenols (polyphenon E) and individual catechins. Nutr Cancer 2007;59:62–9.
29. Loe DW, Almquist KC, Cole SP, et al. ATP-dependent 17 beta-estradiol 17-(beta-D-glucuronide) transport by multidrug resistance protein (MRP). Inhibition by cholestatic steroids. J Biol Chem 1996;271:9683–9.
30. Vaidyanathan JB, Walle T. Transport and metabolism of the tea flavonoid (-)-epicatechin by the human intestinal cell line Caco-2. Pharm Res 2001;18:1420–5.
31. Scholl C, Lepper A, Lehr T, et al. Population nutrikinetics of green tea extract. PLoS One 2018;13:e0193074.
32. Xu H, Wang Y, Yuan Y, et al. Gender differences in the protective effects of green tea against amnestic mild cognitive impairment in the elderly Han population. Neuropsychiatr Dis Treat 2018;14:1795–801.
33. EFSA Panel on Food Additives and Nutrient Sources added to Food (ANS), Younes M, Aggett P, et al. Scientific opinion on the safety of green tea catechins. EFSA J 2018;16:e05239.
34. Meshitsuka S, Shingaki S, Hotta M, et al. Phase 2 trial of daily, oral epigallocatechin gallate in patients with light-chain amyloidosis. Int J Hematol 2017;105:295–308.

Supplemental Digital Content

© 2022 by The American College of Gastroenterology