Mesenchymal stromal cells as treatment or prophylaxis for acute or chronic graft-versus-host disease in haematopoietic stem cell transplant (HSCT) recipients with a haematological condition.Cochrane Database Syst Rev. 2019 Jan 30; 1:CD009768.CD
Recipients of allogeneic haematopoietic stem cell transplants (HSCT) can develop acute or chronic, or both forms of graft-versus-host disease (a/cGvHD), whereby immune cells of the donor attack host tissues. Steroids are the primary treatment, but patients with severe, refractory disease have limited options and a poor prognosis. Mesenchymal stromal cells (MSCs) exhibit immunosuppressive properties and are being tested in clinical trials for their safety and efficacy in treating many immune-mediated disorders. GvHD is one of the first areas in which MSCs were clinically applied, and it is important that the accumulating evidence is systematically reviewed to assess whether their use is favoured.
To determine the evidence for the safety and efficacy of MSCs for treating immune-mediated inflammation post-transplantation of haematopoietic stem cells.
We searched for randomised controlled trials (RCTs) in the Cochrane Central Register of Controlled Trials (CENTRAL, the Cochrane Library 2018, Issue 12), MEDLINE (from 1946), Embase (from 1974), CINAHL (from 1937), Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (from 1990) and ongoing trial databases to 6 December 2018. No constraints were placed on language or publication status.
We included RCTs of participants with a haematological condition who have undergone an HSCT as treatment for their condition and were randomised to MSCs (intervention arm) or no MSCs (comparator arm), to prevent or treat GvHD. We also included RCTs which compared different doses of MSCs or MSCs of different sources (e.g. bone marrow versus cord). We included MSCs co-transplanted with haematopoietic stem cells as well as MSCs administered post-transplantation of haematopoietic stem cells.
DATA COLLECTION AND ANALYSIS
We used standard methodological procedures expected by Cochrane.We employed a random-effects model for all analyses due to expected clinical heterogeneity arising from differences in participant characteristics and interventions.
We identified 12 completed RCTs (879 participants), and 13 ongoing trials (1532 enrolled participants planned). Of 12 completed trials, 10 compared MSCs versus no MSCs and two compared different doses of MSCs. One trial was in people with thalassaemia major, the remaining trials were for haematological malignancies. Seven trials administered MSCs to prevent GvHD, whereas five trials gave MSCs to treat GvHD.In the comparison of MSCs with no MSCs, cells were administered at a dose of between 105 and 107 cells/kg in either a single dose (six trials) or in multiple doses (four trials) over a period of three days to four months. The dose-comparison trials compared 2 x 106 cells/kg with 8 x 106 cells/kg in two infusions, or 1 x 106 cells/kg with 3 x 106 cells/kg in a single infusion.The median duration of follow-up in seven trials which administered MSCs prophylactically ranged from 10 to 60 months. In three trials of MSCs as treatment for aGvHD, participants were followed up for 90 or 100 days. In two trials of MSCs as treatment for cGvHD, the mean duration of follow-up was 13.4 months (MSC group) and 23.6 months (control group) in one trial, and 56 weeks in the second trial. Five trials included adults only, six trials included adults and children, and one trial included children only. In eight trials which reported the gender distribution, the percentage of females ranged from 20% to 59% (median 35.8%).The overall quality of the included studies was low: randomisation methods were poorly reported and several of the included studies were subject to a high risk of performance bias and reporting bias. One trial which started in 2008 has not been published and the progress of this trial in unknown, leading to potential publication bias. The quality of evidence was therefore low or very low for all outcomes due to a high risk of bias as well as imprecision due to the low number of overall participants, and in some cases evidence based on a single study. We found that MSCs may make little or no difference in the risk of all-cause mortality in either prophylactic trials (HR 0.85, 95% CI 0.50 to 1.42; participants = 301; studies = 5; I2 = 34% ; low-quality evidence) or therapeutic trials (HR 1.12, 95% CI 0.80 to 1.56; participants = 244; studies = 1; very low-quality evidence), and no difference in the risk of relapse of malignant disease (prophylactic trials: RR 1.08, 95% CI 0.73 to 1.59; participants = 323; studies = 6; I2 = 0%; low-quality evidence) compared with no MSCs. MSCs were well-tolerated, no infusion-related toxicity or ectopic tissue formation was reported. No study reported health-related quality of life. In prophylactic trials, MSCs may reduce the risk of chronic GvHD (RR 0.66, 95% CI 0.49 to 0.89; participants = 283; studies = 6; I2 = 0%; low-quality evidence). This means that only 310 (95% CI 230 to 418) in every 1000 patients in the MSC arm are expected to develop chronic GvHD compared to 469 in the control arm. However, MSCs may make little or no difference to the risk of aGvHD (RR 0.86, 95% CI 0.63 to 1.17; participants = 247; studies = 6; I2 = 0%; low-quality evidence). In GvHD therapeutic trials, we are very uncertain whether MSCs improve complete response of either aGvHD (RR 1.16, 95% CI 0.79 to 1.70, participants = 260, studies = 1; very low-quality evidence) or cGvHD (RR 5.00, 95%CI 0.75 to 33.21, participants = 40, studies = 1; very low-quality evidence).In two trials which compared different doses of MSCs, we found no evidence of any differences in outcomes.