This is the first study to report the feasibility and general tolerance of one repeated systemic UC-MSC transfusion in both juveniles and adults with T1D. After 1-year follow-up, 40.7% participants remained clinical remission in MSC-treated group, nearly 2.5 times higher than those achieved in the control group, indicating the efficacy of UC-MSCs in preservation of β-cell function.
Natural history studies in T1D have shown a progressive decline in insulin secretion ever since the pre-clinical stage and continuing after diagnosis [10]. During the 1st year of diagnosis, only about 20% of individuals with T1D maintained C-peptide, and mean levels of C-peptide slide down across all age groups, with 17.6% decrease in adults of 21–46 years and almost 40% decrease in youth of 7–12 years old [11]. Retention of endogenous β-cell function is associated with beneficial outcomes but still challengeable in clinic. Autologous nonmyeloablative hematopoietic stem cell transplantation (AHSCT) is the first approach ever proved to successfully halt disease process. In a multicenter analysis, 59% of individuals with T1D achieved insulin free within 6 months of transplantation. The median duration of insulin independence was 12 months [12], and the longest lasting time was 5 years in our center [13]. Nevertheless, 52% of treated individuals experienced adverse effects with one serious case who died of Pseudomonas aeruginosa sepsis [12]. Therefore, AHSCT is not widely applied and safer cell-based therapeutic options are in demand.
MSCs also have unique immunomodulatory properties and represent the most clinically studied experimental cell therapy in many immune-mediated disorders, such as graft-versus-host disease, Crohn’s disease and systemic lupus erythematosus [2, 14]. Based on encouraging pre-clinical animal data in T1D models, MSCs also have become an attractive potential therapy for T1D. Up to now, 15 trials have been registered on the website of ClinicalTrial.gov. Regimens of MSC treatment including cell origin, dose, frequency and route of transplantation are varied among registered clinical centers.
In our previous pilot trial, four individuals with T1D received a single transfusion of allogenic bone marrow MSCs [7]. No side effect was observed during 4-year follow-up. One responder showed increased postprandial C-peptide during the first year and the other achieved insulin independence for 3 months. MSC treatment appeared to be safe and beneficial for β-cell preservation, which was contemporaneously confirmed by an open-label RCT in Sweden [8]. Twenty newly diagnosed individuals with T1D were randomly assigned to either autologous bone marrow MSCs or insulin-only treatment. After 1 year, a mean 5% increase in the peak C-peptide response and 10% increase in the area under the curve of C-peptide was observed in MSC-treated subjects, significantly higher than the percent changes of C-peptide in the control group. But no case of insulin free was reported. In an earlier double blind RCT, 15 of 29 newly diagnosed T1D received a single transfusion of allogenic UC-MSCs and displayed an obviously increased and constantly higher fasting C-peptide than the control group through 2-year follow-up. Three recipients discontinued insulin 1 year after transplantation, but little information was available for the lasting period [9]. Disparities in therapeutic efficacy among studies probably due to different sources of MSCs or characteristics of enrolled subjects.
In the above-mentioned reports, preserved C-peptide appeared to be a consistent efficacy, whereas insulin independence seemed less achievable, therefore we defined clinical remission in this study to be over 10% increase in either fasting or postprandial C-peptide. Based on an expended sample size and a parallel controlled design, our study demonstrated a significantly higher achievement of clinical remission after one repeated UC-MSC transfusion. Moreover, insulin independence otherwise not the primary endpoint was observed in three recipients in contrast to none of the control group, thus further implying the effect of our MSC therapeutic strategy in improving endogenous insulin secretion.
Among secondary endpoints, changes of C-peptide showed distinct tendencies between the groups. Levels of C-peptide decreased in most participants in the control group, with a mean 28% and 32% decrease in the fasting and postprandial state, respectively. On the other hand, a convergence for improved C-peptide in MSC-treated group was observed, with a mean 9% decrease in fasting C-peptide and 16% increase in postprandial C-peptide, although no statistical significance was achieved. Glycemic control maintained around the targeted level without obviously increased insulin dose in participants received MSCs. Nevertheless, such a benefit in glycemic control was not restricted to MSC intervention, because no difference was observed in levels of HbA1c or dose of insulin when compared to the control group. This founding was consistence with that in the RCT from Sweden [8], probably due to comparable basic diabetic care provided to all participants.
Intravenous administration of MSCs is generally considered to be clinical safe, without side effects recorded in previous trials in T1D [15]. In this study, we increased the treatment course to repeated transfusion of MSCs. Three recipients experienced mild fever but recovered spontaneously within 24 hours. Such a post-infusion febrile reaction was once reported in clinical trials of other diseases [16]. Notably, we did not observe any severe adverse events, and both juvenile and adult recipients tolerated well.
There was a heterogenicity in the therapeutic efficacy of UC-MSCs, which addressed the potential factors associated with therapeutic effects of UC-MSCs for T1D. According to previous experiences in AHSCT, good responders were older at onset, were treated soon after T1D diagnosis, had lower instances of DKA, and had lower HbA1c levels at the time of transplantation [17]. However, there was no difference in age of onset, gender, occurrence of DKA, or duration of T1D between good and poor responders to UC-MSC treatment in this trial. In contrast, we found that subjects who had lower levels of fasting C-peptide at the time of enrollment and relatively higher HbA1c levels were more likely to achieve clinical remission, indicating that UC-MSCs are beneficial for T1D with less residual β-cell function and poor glycemic control. As a technical element, the number of transfused UC-MSCs also influenced the therapeutic outcomes. In our model for predicting clinical remission, the number of UC-MSCs exerted larger contribution than other parameters, reinforcing the evidence of dose-dependent efficacy of MSC treatment. The optimal doses as well as courses of MSC transplantation should be further investigated in future studies.