The effects of idarubicin versus other anthracyclines for induction therapy of patients with newly diagnosed leukaemia.

BACKGROUND: Anthracycline combined with cytarabine has been the standard for induction therapy of newly diagnosed acute myeloid leukaemia (AML) for several decades. Due to theoretical advantages, idarubicin (IDA) might be the most effective and tolerable anthracycline. However, there is no evidence that would definitively prove the superiority of IDA over other anthracyclines.

OBJECTIVES: To assess the efficacy and safety of IDA versus other anthracyclines in induction therapy of newly diagnosed AML.

SEARCH METHODS: We identified relevant randomised controlled trials (RCTs) by searching the Cochrane Central Register of Controlled Trials (The Cochrane Library 2014, Issue 8), MEDLINE (from 1946 to 3 August 2014), EMBASE (from 1974 to 3 August 2014), Chinese BioMedical Literature Database (1978 to 3 August 2014), relevant conference proceedings and databases of ongoing trials.

SELECTION CRITERIA: RCTs that compared IDA with other anthracyclines in induction therapy of newly diagnosed AML.

DATA COLLECTION AND ANALYSIS: Two review authors independently extracted data and assessed the quality of studies according to methodological standards of the Cochrane Collaboration. We estimated hazard ratios (HRs) for time-to-event data outcomes using the inverse variance method, and risk ratios (RRs) for dichotomous data outcomes using the Mantel-Haenszel method. We adopted a fixed-effect model and repeated the main meta-analysis by a random-effects model in a sensitivity analysis.

MAIN RESULTS: We identified 2017 references. Ultimately, 27 RCTs (including 22 two-armed RCTs and five three-armed RCTs) involving 9549 patients were eligible. The consolidation treatments adopted in the studies were comparable and had no impact on the results. Overall, the risk of bias of the studies was unclear to high.Eighteen RCTs (N = 6755) assessed IDA versus daunorubicin (DNR). The main meta-analyses showed that IDA compared with DNR prolonged overall survival (OS) (12 studies, 5976 patients; HR 0.90, 95% confidence interval (CI) 0.84 to 0.96, P = 0.0008; high quality of evidence) and disease-free survival (DFS) (eight studies, 3070 patients; HR 0.88, 95% CI 0.81 to 0.96, P = 0.004; moderate quality of evidence), increased complete remission (CR) rate (18 studies, 6692 patients; RR 1.04, 95% CI 1.01 to 1.07, P = 0.009; moderate quality of evidence), and reduced relapse rate (four studies, 1091 patients; RR 0.88, 95% CI 0.80 to 0.98, P = 0.02; moderate quality of evidence), although increased the risks of death on induction therapy (14 studies, 6349 patients; RR 1.18, 95% CI 1.01 to 1.36, P = 0.03; moderate quality of evidence) and grade 3/4 mucositis (five studies, 2000 patients; RR 1.22, 95% CI 1.04 to 1.44, P = 0.02; moderate quality of evidence). There was no evidence for difference in the risks of grade 3/4 cardiac toxicity (six studies, 2795 patients; RR 0.98, 95% CI 0.70 to 1.37, P = 0.91; moderate quality of evidence) and other grade 3/4 adverse events (AEs). None of the studies reported on quality of life (QoL).Eight RCTs (N = 2419) evaluated IDA versus mitoxantrone (MIT). The main meta-analyses showed that there was no evidence for difference between arms in OS (six studies, 2171 patients; HR 0.98, 95% CI 0.89 to 1.08, P = 0.69; high quality of evidence), DFS (four studies, 249 patients; HR 0.88, 95% CI 0.70 to 1.10, P = 0.26; low quality of evidence), CR rate (eight studies, 2411 patients; RR 0.97, 95% CI 0.92 to 1.03, P = 0.32;moderate quality of evidence), the risks of death on induction therapy (five studies, 2055 patients; RR 1.10, 95% CI 0.88 to 1.38, P = 0.39; moderate quality of evidence) and relapse (three studies, 328 patients; RR 0.99, 95% CI 0.80 to 1.22, P = 0.89; moderate quality of evidence). There was no evidence for difference in the risks of grade 3/4 cardiac toxicity (one study, 160 patients; RR 0.67, 95% CI 0.11 to 3.88, P = 0.65; low quality of evidence) and other grade 3/4 AEs. None of the studies reported on QoL.Two RCTs (N = 211) compared IDA with doxorubicin (DOX). Neither study assessed OS. One study showed that there was no evidence for difference in DFS (63 patients; HR 0.62, 95% CI 0.34 to 1.14, P = 0.12; low quality of evidence). The main meta-analysis for CR rate showed an improved CR rate with IDA (two studies, 187 patients; RR 1.28, 95% CI 1.03 to 1.59, P = 0.02; low quality of evidence). Neither study provided data for the risks of death on induction therapy and relapse. One trial showed that there was no evidence for difference in the risk of grade 3/4 cardiac toxicity (one study, 100 patients; RR 0.31, 95% CI 0.01 to 7.39, P = 0.47; very low quality of evidence). Neither study reported on QoL.Two RCTs (N = 1037) evaluated IDA versus zorubicin (ZRB). Neither study assessed OS. One trial showed that there was no evidence for difference in DFS (one study, 155 patients; HR 1.25, 95% CI 0.83 to 1.88, P = 0.29; low quality of evidence). The main meta-analyses for CR and death on induction therapy both showed that there was no evidence for difference (CR rate: two studies, 964 patients; RR 1.04, 95% CI 0.96 to 1.13, P = 0.31; low quality of evidence. risk of death on induction therapy: two studies, 964 patients; RR 0.75, 95% CI 0.50 to 1.13, P = 0.17; moderate quality of evidence). Neither study reported the risks of relapse and grade 3/4 cardiotoxicity. One trial showed that IDA reduced the risk of grade 3/4 mucositis. Neither study reported on QoL.

AUTHORS' CONCLUSIONS: Compared with DNR in induction therapy of newly diagnosed AML, IDA prolongs OS and DFS, increases CR rate and reduces relapse rate, although increases the risks of death on induction therapy and grade 3/4 mucositis. The currently available evidence does not show any difference between IDA and MIT used in induction therapy of newly diagnosed AML. There is insufficient evidence regarding IDA versus DOX and IDA versus ZRB to make final conclusions. Additionally, there is no evidence for difference on the effect of IDA compared with DNR, MIT, DOX or ZRB on QoL.