The goal of this meta-analysis was to identify the temozolomide (TMZ) regimen with optimal efficacy and tolerance for treatment of recurrent high-grade glioma (HGG). The PubMed and EMBASE databases were searched from the earliest records to February 2015, which identified 33 studies with 1760 participants that met the inclusion criteria. The standard schedule and three most common dose-dense regimens of TMZ therapy for recurrent HGG were included in this meta-analysis. The schedule of 7 days on/7 days off for the treatment of grade IV gliomas was significantly superior to the standard regimen with respect to progression-free survival at 6 months (34.8 %; 95 % confidence interval (CI) 27.0-43.4 %) and 12 months (15.5 %; 95 % CI 10.7-21.8 %). For grade III gliomas, this regimen conveyed a significantly greater overall survival (OS) rate at 12 months (79.0 %; 95 % CI 56.2-91.7 %), as compared to the standard schedule. Also, the 21 days on/7 days off regimen had significantly longer OS rates at 6 months (73.6 %; 95 % CI 63.4-81.8 %) and 12 months (40.6 %; 95 % CI 32.6-48.6 %) than the standard regimen for grade IV gliomas. In addition, the standard schedule showed a significantly higher clinical benefit rate than the 7 days on/7 days off and 21 days on/7 days off regimens. However, the grade 3-4 toxicity rate of lymphopenia of the standard schedule was 76.5 % (95 % CI 45.5-92.7 %), which was the highest among the four regimens. Recurrent HGG patients receiving personalized treatment should be closely followed up, especially those with concurrent hematological diseases.
Keywords: Glioblastoma; Glioma; Meta-analysis; Recurrent; Temozolomide.