Tsao 2009.
| Study characteristics | ||
| Methods | RCT, parallel, double‐blind in Japan | |
| Participants | Participants: 41 (male/female: 19/22) participants aged 18‐75 years, with ≥ 1 histologically confirmed, bidimensionally measurable oral premalignant lesions, with Zubrod performance status < 2, adequate hematologic, liver, renal and cardiac function, with one of the following: harbouring at least mild dysplasia, located in a high‐risk area (i.e. floor of mouth, ventrolateral tongue and soft palate), significant extent of lesion tissue involvement and presence of symptoms. 11, 9 and 10 participants were randomised in intervention group A, B and C respectively and 11 participants in control group Recruitment: from August 2002‐March 2008 |
|
| Interventions | Treatment groups: GTE contains high amounts of polyphenols, including EGCG: Group A: 500 mg/m2 GTE, N = 11 (male/female: 5/6) Group B: 750 mg/m2 GTE, N = 9 (male/female: 4/5) Group C: 1,000 mg/m2 GTE, N = 10 (male/female: 4/6) Control group: placebo, N = 11 (male/female: 6/5) Duration: 12 weeks |
|
| Outcomes | Primary outcome Clinical and histologic response of high‐risk oral lesions Secondary outcome Safety data: qualitative and quantitative toxicities of GTE |
|
| Green tea in exposure categories | NA | |
| Notes | Funding: support to En Ltd., including YM Sagesaka as employee of Ito En Ltd | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "The randomization was done with the Pocock‐Simon dynamic allocation scheme" Comment: done |
| Allocation concealment (selection bias) | Low risk | Quote: "The randomization was done with the Pocock‐Simon dynamic allocation scheme" Comment: probably concealed |
| Blinding of participants and personnel (performance bias) Oral premalignant‐lesions | Low risk | Quote: "GTE and placebo capsules were supplied to the pharmacy in blister packs containing 10 capsules each" Comment: participants and personnel probably blinded |
| Blinding of outcome assessment (detection bias) Oral premalignant‐lesions | Unclear risk | No explicit statement on blinded outcome assessment, only for immunohistochemical staining |
| Incomplete outcome data (attrition bias) Oral premalignant‐lesions | Low risk | Data reported for all participants who completed the study. ITT analysis implemented |
| Selective reporting (reporting bias) | Unclear risk | The study protocol is not available and it is not clear if the published reports include all expected outcomes |
| Other bias | Low risk | No other bias |