| Objective: To evaluate the efficacy and safety of high-dose dexamethasone(HDD) versus Conventional-dose prednisone(CDP) for primary immune thrombocytopenia( ITP).Methods: We included randomized controlled trials(RCT) comparing HDD with CDP for primary ITP. The participants were patients with primary ITP, the age ?15 and without limitation of gender, race and prior treatment. The primary outcome were complete response(CR), overall response(OR), platelet transfusion frequency. The secondary outcomes were incidence of severe bleeding, time-to-response, duration of response and relapse. We searched the Ovid MEDLINE, EMBASE, CINAHL, CENTRAL, CBM and Clinical Trials Registry Platform databases(search date January 31, 2016), ect. Two reviewers independently identified the included trials, evaluated their methodological quality and extracted data. The Cocharane collaboration's Review manager 5.3 software was used for data analysis. Continuous data were pooled together, using weighted mean difference(MD) or standard mean difference(SMD)(according to data comparability) and 95% confidence interval(CI) and dichotomous data using the relative risk(RR) and 95% CI.Results: The search retrieved 1226 titles, of which 7 original RCTs were included in this systematic review. There were 737 patients, 231 males and 506 females, including newly diagnosed, acute and chronic cases. The administration of HDD was different. Thus, we divided these studies into 3 subgroup, 1 course HDD(4 studies), 1-2 courses HDD(2 studies) and 1 course HDD then prednisone tapering(1 study). We pooled the subgroup results first and then summarized these 3 subgroups. The results were showed as follow:(1) CR: All of these 7 studies(n=737) reported CR. Compared with CDP, RRs(95% CI)of 1 course HDD, 1-2 courses HDD and 1 course HDD then prednisone tapering group were 1.32(1.01, 1.72), 1.85(1.41, 2.42)and 11.00(2.83, 42.7), respectively. NNT were 8.62, 4.25 and 1.50, respectively. Pooled RR of all 7 RCTs was 1.75(1.45, 2.10)and NNT was 4.52.(2)OR: All 7 studies(n=737) reported OR. RRs(95% CI)of 1 course HDD vs CDP, 1-2 courses HDD vs CDP and 1 course HDD then prednisone tapering vs CDP were 1.23(1.09, 1.39), 1.16(1.04, 1.30)and 1.93(1.29, 2.88), respectively. NNT were 6.52, 8.50 and 2.31, respectively. Pooled RR of these 7 RCTs was 1.93(1.29, 2.88)and NNT was 6.22.(3)Time to response: 1 study(n=182) reported time to response and it compared 1-2 courses HDD with CDP. MD was-2.76(-3.69,-1.83)days.(4) The incidence of splenectomy: 1 study(n=60) reported the incidence of splenectomy and it compared 1 course HDD then prednisone tapering with CDP. RR(95% CI)was 0.25(0.08, 0.80)and NNT was 3.33.(5)Serious adverse reactions(except bleeding): There were 2 RCTs, 1(n=192) comparing 1-2 courses HDD with CDP and the other(n=60) comparing 1 course HDD then prednisone tapering with CDP, RRs of serious adverse reactions were 0.20(0.01, 4.20) and 0.50(0.05, 5.22). The pooled RR(95% CI)was 0.34(0.05, 2.09). No statistically significant difference was found between HDD and CDP.(6)Infections: 4 studies reported this outcome, 2(n=103) compared 1course HDD with CDP and 2 compared 1-2 courses HDD with CDP. RRs(95% CI)were 0.09(0.01, 0.67)and 0.02(0.01, 1.38). The pooled RR(95% CI) was 0.12(0.03,0.49) and NNH was15.9.(7) Withdrawal because of side effect: 1 study(n=192) reported this result and it compared 1-2 courses HDD with CDP showing that RR(95% CI)was 0.20(0.01, 4.20) and the difference was not statistically significant.Conclusions:(1) HDD had a significant advantage over CDP in response to treatment(CR and OR). Among the different administrations of HDD, 1 course HDD then prednisone tapering was superior to either 1course HDD or 1-2 course HDD and might reduce splenectomy.(2) Compared with CDP, 1 course HDD and 1-2courses HDD may reduce infections but not serious adverse reactions(except bleeding) and withdrawal because of side effect significantly. |
PDF Full Text Request