Potentiation of cord blood cell therapy with erythropoietin for children with CP: a 2 × 2 factorial randomized placebo-controlled trial

Abstract

Background: Concomitant administration of allogeneic umbilical cord blood (UCB) infusion and erythropoietin (EPO) showed therapeutic efficacy in children with cerebral palsy (CP). However, no clinical studies have investigated the effects of UCB and EPO combination therapy using a 2 × 2 four-arm factorial blinded design with four arms. This randomized placebo-controlled trial aimed to identify the synergistic and individual efficacies of UCB cell and EPO for the treatment of CP.

Methods: Children diagnosed with CP were randomly segregated into four groups: (A) UCB+EPO, (B) UCB+placebo EPO, (C) placebo UCB+EPO, and (D) placebo UCB+placebo EPO. Based on the UCB unit selection criteria of matching for ≥ 4/6 of human leukocyte antigen (HLA)-A, -B, and DRB1 and total nucleated cell (TNC) number of ≥ 3 × 107/kg, allogeneic UCB was intravenously infused and 500 IU/kg human recombinant EPO was administered six times. Functional measurements, brain imaging studies, and electroencephalography were performed from baseline until 12 months post-treatment. Furthermore, adverse events were closely monitored.

Results: Eighty-eight of 92 children enrolled (3.05 ± 1.22 years) completed the study. Change in gross motor performance measure (GMPM) was greater in group A than in group D at 1 month (△2.30 vs. △0.71, P = 0.025) and 12 months (△6.85 vs. △2.34, P = 0.018) post-treatment. GMPM change ratios were calculated to adjust motor function at the baseline. Group A showed a larger improvement in the GMPM change ratio at 1 month and 12 months post-treatment than group D. At 12 months post-treatment, the GMPM change ratios were in the order of groups A, B, C, and D. These results indicate synergistic effect of UCB and EPO combination better than each single therapy. In diffusion tensor imaging, the change ratio of fractional anisotropy at spinothalamic radiation was higher in group A than group D in subgroup of age ≥ 3 years. Additionally, higher TNC and more HLA-matched UCB units led to better gross motor outcomes in group A. Adverse events remained unchanged upon UCB or EPO administration.

Conclusions: These results indicate that the efficacy of allogeneic UCB cell could be potentiated by EPO for neurological recovery in children with CP without harmful effects.

Trial registration: ClinicalTrials.gov, NCT01991145 , registered 25 November 2013.

Keywords: Cerebral palsy; Clinical trial; Erythropoietin; Functional performance; Umbilical cord blood.

Conflict of interest statement

The authors declare no potential conflicts of interest.

Figures

Fig. 1

Screening, randomization, and follow-up. a The timeline of the study, b the cooperation of investigators to maintain double-blindness, and c the study flow. CP, cerebral palsy; DTI, diffusion tensor image; EEG, electroencephalogram; EPO, erythropoietin; FA, fractional anisotropy; GMFCS, gross motor function classification system; HLA, human leukocyte antigen; MRI, magnetic resonance imaging; PET, positron emission tomography; UCB, umbilical cord blood

Fig. 2

Changes in gross motor outcome. A Changes in (a) GMPM, (b) GMFM change ratio, and (c) GMPM change ratio from baseline to 1, 3, 6, and 12 months post-treatment among group A, B, C, and D. GMPM and GMFM change ratios were calculated as GMPMatthe time point−GMPMatbaselineGMPMatbaseline and GMFMatthe time point−GMFMatbaselineGMFMatbaseline, respectively. Group A (n = 22) received umbilical cord blood (UCB) with erythropoietin (EPO), group B (n = 24) received UCB with placebo EPO (P-EPO), group C (n = 20) received placebo UCB (P-UCB) and EPO, and group D (n = 22) received P-UCB and P-EPO. Data are shown in violin plots where dots represent each value, bold dotted lines represent the median and fine dotted lines represent lower and upper quartiles. Asterisk indicates significant difference in outcome scores between two groups based on post hoc analyses (P < 0.05) (Dunn’s multiple comparison test) following Kruskal-Wallis test. B Changes in GMPM change ratio according to (a) cell dose and (b) HLA disparity in group A. Subgroups with lower and higher TNC were categorized according to the median value of TNC in groups A and B. Subgroup from group A with higher TNC showed significant improvement in GMPM change ratio compared to group D after 12 months post-intervention. Data are also shown in violin plots where dots represent each value, bold dotted lines represent the median and fine dotted lines represent lower and upper quartiles. Asterisk indicates significant difference in outcome scores between two groups based on post hoc analyses (P < 0.05) (Dunn’s multiple comparison test) following Kruskal-Wallis test. The impact of HLA incompatibility was analyzed between HLA full-matched or 1 mis-matched and HLA 2 mis-matched cases in group A and B. In group A, variances of GMFM during baseline to 1 month (P = 0.036) and to 3 months (P = 0.05) were larger among the subjects who received more HLA-compatible UCB (n = 10) than those treated with HLA 2-mismatched UCB (n = 12). Asterisk indicates significant difference in outcome scores between two groups based on Mann-Whitney U test. EPO, erythropoietin; GMFM, gross motor function measure; GMPM, gross motor performance measure; HLA, human leukocyte antigen; TNC, total nucleated cell; UCB, umbilical cord blood

Fig. 3

Electroencephalogram mapping before and after UCB injection. Average delta/alpha band power ratio (DAR) from electroencephalogram (EEG) is depicted on the 2D brain topomap. DAR from EEG taken before treatment, 12 months after treatment, and their difference (post-treatment–pre-treatment) are shown from left towards right. a Taken from group A (n = 20, mean age of pre-treatment EEG was 2.95 ± 1.20 years), b from group B (n = 20, mean age of pre-treatment EEG was 2.71 ± 1.27 years), c from group C (n = 20, mean age of pre-treatment EEG was 3.28 ± 1.27 years), and d from group D (n = 19, mean age of pre-treatment EEG was 4.17 ± 1.41 years). Among the total 88 participants, only 79 EEG data at baseline and 12 months post-treatment were able to be appropriately processed. Six participants lacked follow-up study, and 3 files were invalid on the analyzing program. *P < 0.05 by Mann-Whitney U test comparing the difference between pre- and post-treatment DAR. DAR, delta/alpha ratio; EEG, electroencephalogram; UCB, umbilical cord blood