Maintenance of the Synergistic Effects of Cord Blood Cells and Erythropoietin Combination Therapy After Additional Cord Blood Infusion in Children With Cerebral Palsy: 1-Year Open-Label Extension Study of Randomized Placebo-Controlled Trial

Major Points and Findings:

Families often ask whether a second infusion adds to the first. This Korean study is one of very few with data on that question. Children who had finished a randomized, placebo-controlled trial of allogeneic umbilical cord blood (UCB), with or without erythropoietin (EPO), were all offered one further open-label cord blood infusion 15 months after the first treatment and followed for another year. Children who had cord blood twice can therefore be compared with children who had it once.

The authors call their own result “disappointing”. Gains after the second infusion were much smaller than after the first (1.55 versus 5.75 GMFM-66 points over a year), and a second dose did not significantly outperform a single dose. Because every child was treated in the extension, there was no untreated comparison group. The data do suggest that age and cell dose influence the response, and that cord blood combined with EPO may have a lasting advantage in severely affected children.

Aim:

To find out whether the combined effect of cord blood and EPO seen in the original trial persists for two years, and whether an additional infusion adds benefit.

Methods:

The study was a 1-year open-label extension of a randomized, double-blind, placebo-controlled trial at CHA Bundang Medical Center, South Korea (NCT03130816).

The original groups were (A) UCB plus EPO, (B) UCB plus placebo EPO, (C) placebo UCB plus EPO and (D) double placebo. These labels were kept for analysis, and allocation stayed concealed until the last child had received the second infusion.

All children from the first trial were invited. A blood-group-compatible allogeneic unit had to be available with at least 2 × 10⁷ total nucleated cells (TNC) per kg and a match at 3 or more of 6 HLA loci. Exclusions included aspiration pneumonia, genetic disease, coagulopathy, intractable epilepsy, hypertension, liver or kidney impairment and malignancy.

Treatment: At month 15 (T15) every child received one intravenous infusion of allogeneic cord blood from the CHA cord blood bank, washed just before use to remove the cryoprotectant, with chlorpheniramine one hour beforehand. Oral cyclosporine, 7 mg/kg twice daily, was given for 10 days, from 2 days before to 7 days after the infusion. No EPO was given this time. All children continued rehabilitation.

GMFM-66, the Gross Motor Performance Measure (GMPM) and the Bayley Scales of Infant Development II (BSID-II) were scored at T15 and 3, 6 and 12 months later (T18, T21, T27). Subgroups were defined by severity (GMFCS I to III versus IV to V) and by dose above or below the median.

Results:

Of 88 children who completed the first trial, 76 joined the extension and 69 were analysed (16, 19, 16 and 18 in groups A to D). Seven stopped attending. Mean age was 4.29 years (range 1.9 to 7.5), and 41 of the 69 were at GMFCS level IV or V. Group D contained significantly more severely affected children (16 of 18, p = 0.045). The mean dose was 4.4 × 10⁷ TNC per kg, lower than in the first trial (4.8 to 5.0 × 10⁷).

Safety:

• Four adverse events in two children were judged possibly related. One child had cough and nausea during the infusion, which settled within minutes after an antihistamine. Another, who had just had a respiratory infection, developed fever and bronchopneumonia the next day and recovered with four days of antibiotics.
• Eleven serious adverse events occurred in five children, all a month or more after treatment and all judged unlikely to be related. Seizures were the most frequent (six events in four children, three of them febrile). One child had a first-ever seizure, and his EEG before the infusion had already shown a partial seizure pattern.

Motor scores: The children as a whole improved on all measures over the extension year. However, the GMFM-66 gain from T15 to T27 was 1.55 points, against 5.75 points over the first trial year (p below 0.05). Group B received cord blood alone both times and gained 5.69 points the first time and 1.97 the second. Groups C and D, receiving cord blood for the first time but now over 4 years old, gained only 1.22 and 1.05 points.

Children who had cord blood twice (groups A and B) tended towards larger GMFM-66 gains than those who had it once (C and D), but the difference was not significant, and no differences between the four groups were found at 3, 6 or 12 months.

Among the 41 children at GMFCS IV or V, group A (UCB plus EPO, then UCB again) improved significantly more on the GMPM than group D, measured from the first baseline out to T27 (p below 0.05). Group A had only 8 children in this subgroup.

Dose: Children who received at least 4.4 × 10⁷ TNC per kg (n = 35) gained more than those who received less (n = 34) on the GMFM-66 and BSID-II motor scale at 6 and 12 months, and on the BSID-II mental scale throughout (p below 0.05). Dose correlated with the change in mental score over the year (r = 0.536).

Conclusions:

The authors conclude that combining cord blood with EPO may be more effective, that the advantage may last two years in severely impaired children, and that higher cell numbers lead to better outcomes. Contrary to their hypothesis, repeat dosing did not produce a significant cumulative benefit. They think the 15-month gap was too long, noting that other repeat-dose studies used intervals of days to weeks.

They explain the weak second-year response mainly by age. Children at GMFCS level V reach 90% of their expected motor ability at about 2 years 5 months and those at level IV at about 3 years 2 months. This cohort averaged 4.3 years.

The authors state several limitations. The extension had no untreated control group, so improvement cannot be clearly separated from normal development. The groups were uneven in baseline function. The BSID-II had a ceiling (four children reached the maximum mental score). The abstract’s claim that combination therapy is “more effective” rests on the severe-subgroup GMPM comparison. The whole-group analysis found no differences between groups.

Background Information:

Erythropoietin has neuroprotective effects in laboratory models. In newborn mice with brain injury from oxygen deprivation, cord blood cells and EPO together reduced cell death more than either alone. The original trial gave 500 IU/kg intravenously two hours before the cord blood, then five doses under the skin at three-day intervals.

In this protocol, partially HLA-matched cord blood from an unrelated donor is given together with a 10-day course of cyclosporine, an immunosuppressant. The authors judge the regimen safe overall, but they link the one pneumonia case to cyclosporine lowering immune function in a child who had only just recovered from a cold.