Injections of ginkgo in the treatment of cerebral infarction: a systematic review and network Meta-analysis

To assess the clinical effectiveness and safety of injections of ginkgo (GI) combined with Western Medicine (WM) for cerebral infarction (CI). Randomized controlled trials (RCTs) of CI treated by GI were searched in China National Knowledge Infrastructure Database, Wanfang, China Science and Technology Journal Database, Web of Science, Cochrane library, Embase, PubMed and Chinese Biomedical Literature Database, with the publication data no later than April, 2016. The Cochrane risk of bias method was used to evaluate the methodological quality of the RCTs. The data were analyzed by Review Manager 5.3, Stata 13.0, and WinBUGS 14 software. Totally 37 RCTs involving 4330 patients were included. By direct comparison, the results of GI group were significantly superior to the routine WM group in the total effective rates [OR = 3.61, 95% CI (2.93, 4.44), P < 0.0001], the neural function defect score (NFDS) [MD = − 4.39, 95% CI (− 5.47, − 3.32), P < 0.0001]. Network Meta-analysis (NMA) results showed that, between 5 GIs in efficacy, the difference comparing ginaton injections (GbE) to ginkgo-dipyidamolum injections (GD) [OR = 1.74, 95% CI (0.73, 3.65)], shuxuening injections (SXN) [OR = 1.06, 95% CI (0.609, 1.697)] or ginkgolides injections (GK) [OR = 4.711, 95% CI (1.178, 13.21)] reach statistical significance; the difference comparing GD to GK reach statistical significance [OR = 2.791, 95% CI (0.866, 6.908)]; the difference comparing SXN to GK reach statistical significance [OR = 4.537, 95% CI (1.203, 12.41)]. Besides, there was no difference between 4 GIs in NFDS. Probability ranking result showed a great possibility for GK [Surface under the Cumulative Ranking curve (SUCRA) = 80.3%] in improving the total effective rates, which were followed by GD (SUCRA = 73.34%), SXN (SUCRA = 46.59%), GbE (SUCRA = 45.46%), floium ginkgo extract and tertram ethypyrazine sodium chloride injections (FT) (SUCRA = 35.64%). However, GK (SUCRA = 80.3%) or GbE (SUCRA = 69.4%) was better than other GIs in reducing NFDS.GK + WM is the best treatment measures to reduce NFDS in cerebral infarction, which were followed by SXN + WM (SUCRA = 51.6%), GD + WM (SUCRA = 48.1 %). GIs was more effectiveness on CI than the routine Western Medicine. But based on the limitations of the study, more high-quality randomized controlled trials will be necessary.