Our search retrieved 7770 references. We first screened the titles and abstracts and excluded 7751 records that were not relevant to RIC for AML in CR and MDS or not RCTs. After further examined fulltexts of the remained 19 records, we excluded 10 references that were not RCT studies, not relevant to RIC or not compared with MAC regimens and the duplicated reports. In the end, we included 6 RCTs reported in 9 references into meta-analyses. All authors agreed to include the six studies (Bornhäuser et al.,9 Kröger et al.,10 Ringdén et al.,11 Scott et al.,12 Beelen et al.8 and MC-FludT·14/L Trial I;7 for flow diagram see Fig. 1). Studies Bornhäuser et al.,9 Kröger et al.,10 Ringdén et al.11 and Scott et al.12 reported the long-term follow up data.11,26−28
The six included studies with 1413 participants (711 in the RIC group and 702 in the MAC group) all focused on the efficacy and safety of RIC compared with MAC followed by allo-HSCT for AML in CR and MDS. Four studies focused on RIC vs. TBI/Bu based MAC while two studies focused on RIC vs. treosulfan based MAC regimens. All the studies used peripheral stem cell and bone marrow as stem cell sources. Donors included matched related, mismatched related and matched unrelated donors. The demographic characteristics of the two treatment arms were similar in the included studies and are shown in Table 1. All included studies displayed low risk of bias. Details of quality assessment of the included studies are shown in Table 2 and Supplement 3. All the studies used the intention-to-treat method to analyze OS, CIR and LFS. There was no selective reporting in all the included studies. Because funnel plots and meta-regression should only be used with more than 10 studies, we did not use them to detect publication bias in our analysis.29
OS was not statistically different between RIC and MAC (HR = 0·95, 95% CI 0·64–1·4, P = 0·80). Heterogeneity of the meta-analysis was significant (P = 0·003, I2 = 72%) (Fig. 2A). The result was also similar in the RIC vs. TBI/Bu based MAC subgroup analysis (HR = 0·84, 95% CI 0·5–1·4, P = 0·50) with significant heterogeneity (P = 0·04, I2 = 65%), but in the RIC vs. treosulfan 30 g/m2 based MAC subgroup analysis, RIC was significantly inferior to the treosulfan based MAC conditioning regimen (HR = 1·63, 95% CI 1·17 − 2·28, P = 0·004). The combined long-term follow-up data also showed there was no difference between RIC and MAC (HR = 0·86, 95% CI 0·53–1·41, P = 0·56) with significant heterogeneity (P = 0·01, I2 = 73%) (Fig. 4).
There was no difference in CIR (HR = 1·18, 95% CI 0·88–1·59, P = 0·28) between RIC and MAC (Fig. 2B). There was also no difference in CIR in the three subgroup analyses. Heterogeneity in the meta-analysis and in the RIC vs. TBI/Bu-based MAC subgroup was significant. Bornhäuser et al.,9 Kröger et al.10 and Scott et al.12 reported LFS, the combined result showed RIC had similar LFS to MAC (HR = 1·09, 95% CI 0·69–1·74, P = 0·71) with significant heterogeneity (P = 0·05, I2 = 66%) (Fig. 2C).
RIC significantly reduced NRM compared to TBI/Bu based MAC (HR = 0·53, 95% CI 0·36–0·8, P = 0·002) with no heterogeneity (P = 0·40, I2 = 0%) (Fig. 3A). However, the treosulfan 30 g/m2 based MAC8 significantly reduced NRM compared to RIC (HR = 1·67, 95% CI 1·02–2·72, P = 0·04). RIC did not show a significant difference compared with treosulfan 42 g/m2 based MAC (MC-FludT·14/L Trial I;7 HR = 0·76, 95% CI 0·45–1·30, P = 0·32).
RIC showed a tendency to reduce aGVHD (Fig. 3B) and III-IV aGVHD (Supplement 4) compared to TBI/Bu based MAC (HR = 0·79, 95% CI 0·60–1·03, P = 0·08) and (RR = 0·61, 95% CI 0·36–1·04, P = 0·07) with no significant heterogeneity (P = 0·15, I2 = 43%) and (P = 0·19, I2 = 39%). In the Beelen et al.8 and MC-FludT·14/L Trial I7 studies, RIC still did not show a significant difference from treosulfan based MAC (either 30 g/m2 or 42 g/m2).
There was no difference between RIC and MAC in cGVHD (Fig. 3C) and extensive cGVHD (Supplement 4) (HR = 1·01, 95% CI 0·79–1·28, P = 0·96 and RR = 1·03, 95% CI 0·77–1·37, P = 0·84, respectively) with significant heterogeneity (P = 0·08, I2 = 49% and P = 0·09, I2 = 51%, respectively). There was also no difference between RIC and MAC in the subgroup analyses.
RIC showed a trend of increasing GF (OR 2·19, 95% CI 0·96–5·03, P = 0·06) without heterogeneity (P = 0·34, I2 = 12%). The incidence of GF in the RIC and MAC arms were both rare, 2.57% (18 events in 701 participants) and 1.16% (8 events in 690 participants), respectively. RIC did not show significant difference from MAC in terms of overall organ toxicity and oral mucositis, with significant heterogeneity. On the other hand, RIC significantly reduced renal and urinary disorders (RR 0.61, 95% CI 0·39–0.97, P = 0·04) and infection (RR 0.87, 95% CI 0·78–0.97, P = 0·01) without heterogeneity (Supplement 4).
We repeated the meta-analyses for the OS, CIR and long-term OS with the fixed-effect model because of their significant heterogeneity and the results did not change the overall conclusions of these endpoints (Supplement 5).
We removed one study at a time and then repeated the meta-analysis in the sensitivity analysis. The pooled HRs ranged from 0.84 to 1.05 for OS and from 1.02 to 1.26 for CIR. The results after removing any study (including Beelen et al.8 and Scott et al.12 studies) were overall stable. After we removed the Scott et al.12 study, the heterogeneity of CIR disappeared (Supplement 6) and the results of CIR did not change. Eight CML patients were included in the Ringdén et al.11 study. When we removed it in the sensitivity analysis there were no significant changes were observed in the results of OS, CIR, and NRM (Supplement 7).
The quality of evidence for the OS, CIR, LFS and cGVHD endpoints was moderate. The quality of the NRM and aGVHD endpoints was high (Supplements 8 and 9).