Characteristics of Mesenchymal Multipotent Stromal Cells Determine Their Effectiveness for Acute Graft Versus Host Disease Prophylaxis after Allogeneic Bone Marrow Transplantation

Introduction

Allogeneic bone marrow transplantation (allo-BMT) is currently the only way to cure many hematoproliferative disorders. However, allo-BMT use is limited by severe complications, among which the most challenging is graft-versus-host disease (GVHD). As the conventional methods of GVHD prophylaxis are often inefficient new method involving the use of donors’ multipotent mesenchymal stromal cells (MSC) was developed. In some cases prophylaxis of acute GVHD (aGVHD) failed. The reasons of the failure could be either the result of particular qualities of donor-recipient interaction, patient status or characteristics of MSC samples. The results of the aGVHD prophylaxis with donors’ MSC injections after allo-BMT in patients with hematological malignancies included in the randomized clinical trial (Clinicaltrials.gov NCT01941394) were analyzed. In order to discriminate between effective and ineffective for aGVHD prophylaxis MSC samples were thoroughly analyzed. The growth and differentiation characteristics, relative expression levels of different genes were investigated in all MSC samples.

Methods

The study included 77 patients who received allo-BMT from related donors after informed consent. The patients were randomized into 2 groups: the first received standard prophylaxis of aGVHD and the second were additionally infused with MSC from the bone marrow of corresponding hematopoietic stem cells donor at day of WBC reconstitution >1*10⁹/l. MSC were cultivated in aMEM with 4% donors’ platelet lysate. MSC were administered intravenously when the blood counts indicated recovery (peripheral blood leukocytes reached 1x10⁹/l). MSCs and colony-forming unit-fibroblasts (CFU-Fs) from the bone marrow of those donors were analyzed. For this purpose MSC were cultivated in standard conditions (aMEM, 10% fetal calf serum) for 5 passages. Relative expression level (REL) of 30 genes involved in proliferation, differentiation and immunomodulation was estimated by RT-qPCR in all MSC samples.

Results

The infusion of MSC reduced the incidence of aGVHD 2 times and increased the 5 years overall survival of patients (p=0.047). Four of 39 MSC samples have been ineffective for preventing aGVHD. Analysis of individual donor characteristics (gender, age, body mass index), the MSC properties of these donors (growth parameters, REL of 30 genes involved in proliferation, differentiation and immunomodulation) found no significant differences between the MSC, effective and ineffective for preventing aGVHD. However the analysis revealed that cumulative MSC production and CFU-F concentrations in bone marrow decreased with donor age. MSC populations revealed the hierarchy that changed during cultivation, resulting in an increase in the impact of mature cells and a decrease in the subpopulation of cells with high proliferation potential.

The combination of predictors that characterize the most suitable for the prevention of aGVHD MSC samples was revealed by multiple logistic regression analysis. A model calculating the probability of the success of MSC samples application was proposed:

logit(P)=0,75+10,897*FGFR1-4,272*PPARG-2,014*IGF1,

where logit(P) = ln[P/(1-P)], P – probability of successful prophylaxis, FGFR1, PPARG and IGF1 – REL of corresponding genes in tested MSC sample. Chi-square goodness of fit test p= 0.0053. The calculated efficiency of this model was 94%.

The following parameters of MSC were essential for the success of aGVHD prophylaxis: increased REL of FGFR1 combined with reduced REL of PPARG and IGF1 genes. Depending on the chosen value for probability of successful application of MSC, this model can correctly predict the outcome of the use of MSC in 82-94% of cases.

Conclusions

These data confirm the presence of hierarchy as well as heterogeneity in MSC population. The high variability of all analyzed characteristics among MSC from different donors was shown. The mathematical model revealed the combination of parameters enabling to distinguish effective and ineffective MSC samples. By means of the proposed model ineffective MSC samples could be discharged and replaced by effective MSC sample from the third part donor. Such strategy hopefully will prevent the development of aGVHD in the maximum number of patients.