The Impact of Umbilical Cord Mesenchymal Stem Cells on Motor Function in Children with Cerebral Palsy: Results of a Real-world, Compassionate use Study

Major Points and Findings:

This retrospective study from Poland reviews the records of 152 children with cerebral palsy who received Wharton’s jelly (umbilical cord) mesenchymal stem cells outside a clinical trial, under the Polish hospital exemption for experimental treatment. Motor function was compared before the first injection and two months after the fourth.

Gross motor scores rose by a median of 1.9 points, and the smaller groups who could walk improved on walking distance and on a timed mobility test. Younger children gained more on the GMFM. There was no control group, so the study cannot separate the effect of the cells from development, physiotherapy or the passage of time.

Two of the three authors are employees of Polski Bank Komórek Macierzystych (PBKM, part of the FamiCord Group), the Warsaw cord blood and tissue bank. Medical writing for the paper was paid for by the same company, and families paid for the treatment themselves, in part through crowdfunding.

Aim:

To analyse the effect of umbilical cord MSCs (hUC-MSCs) on motor function in children with cerebral palsy, using the GMFM, the 6-Minute Walk Test (6-MWT), the Timed Up and Go test, Lovett’s muscle strength test and the Clinical Global Impression (CGI) scale.

Methods:

The children were treated between May 2018 and August 2021 as a medical therapeutic experiment (hospital exemption), approved by the Bioethics Committee at the Regional Chamber of Physicians and Dentists in Lublin (152/2018/KB/VII and 119/2021/KB/VIII). Data were taken from medical records after treatment. Registration on ClinicalTrials.gov is described as retrospective and ongoing.

The cells came from donated umbilical cords processed within 48 hours of delivery. They were grown in serum-free medium for no more than five passages, checked against the International Society for Cellular Therapy marker criteria, frozen, and thawed before use. Viability was at least 90%. The authors state that proliferation rate and markers of cell ageing were not monitored.

Treatment ran in up to two courses of five injections given every 2 months. Injections were intravenous or intrathecal, with doses of 10, 20, 30 or 40 × 10⁶ cells by body weight to give 1 × 10⁶ MSCs/kg per injection. The first injection was always intravenous. Later injections were switched to intrathecal or back depending on factors that included parental preference, and the paper does not say how many children had each route.

Only visit 1 (baseline) and visit 5 (after four injections, assessed before the fifth) were analysed. Changes were tested with the Wilcoxon test, subgroups with Mann-Whitney U and Kruskal-Wallis tests, and correlations with Spearman’s R and Kendall’s Tau.

Results:

The 152 children had a median age of 5.0 years (IQR 3.0 to 8.0) and weighed 7 to 56 kg. Of these, 84 (55.3%) were boys and 75 (49.3%) had epilepsy. Most were severely affected: 74 (49%) were GMFCS level V and 33 (22%) level IV, with 16 at level I, 15 at level II and 14 at level III.

GMFM (all 152 children): The median change from visit 1 to visit 5 was 1.9 points (IQR 0.0 to 8.0, range −9.0 to 34.4, p < 0.000001). The change correlated with age (Spearman’s R = −0.38), with younger children gaining more, and was unrelated to sex (p = 0.40) or epilepsy (p = 0.17). By baseline GMFCS level the median gains were:

• Level I: 2.9 points (p = 0.001872)
• Level II: 1.0 point (p = 0.071190, not significant)
• Level III: 6.4 points (p = 0.018604)
• Level IV: 2.0 points (p = 0.000014)
• Level V: 1.7 points

The difference between GMFCS groups was not significant (p = 0.35). All five GMFM areas changed significantly, although the median change in four of the five areas was 0.0 points and 1.0 point for sitting.

6-Minute Walk Test (26 children, mainly GMFCS I and II): Median distance went from 120 m to 200 m, a median gain of 75 m (IQR 20.0 to 115.0, p = 0.000058). The gain correlated with GMFM change (R = 0.47) but not with age.

Timed Up and Go (26 children): Median time fell from 12 to 10 seconds, a median change of −2 seconds (IQR −3 to −1, p = 0.000027). Older children improved less (R = 0.46).

Lovett’s muscle strength (63 children): Changes were very small. Two children gained one grade in biceps brachii and four gained in quadriceps femoris (p = 0.068). Eight gained one grade in gluteus maximus (p = 0.01). Strength changes did not correlate with GMFM, Up and Go or 6-MWT changes.

CGI (151 children): At baseline 70 (46.4%) were rated seriously ill and 44 (29.1%) significantly ill. At visit 5, 8 (5.3%) were rated very much improved, 28 (18.5%) much improved, 60 (39.7%) minimally improved, 54 (35.8%) unchanged and 1 (0.7%) minimally worse. Greater GMFM gains went with better CGI ratings (R = −0.28).

Safety: The paper does not report adverse events for this cohort.

Conclusions:

The authors conclude that the cells improved functional performance, with varied individual responses. Children with severe disability also benefited, though less than those who were less disabled at the start, and younger children responded better while older children could still benefit.

The authors list these limitations: no control group and no randomisation, with families who could pay for treatment possibly introducing bias. They also list mixed intravenous and intrathecal routes, non-standard dosing and an analysis that used only two time points. They add that CGI ratings are subjective and that results in real-world use are usually weaker than in trials.

Other limitations are apparent from the full text:

• Without a comparison group, a 1.9-point median GMFM gain over about 8 months cannot be attributed to the cells. Young children with cerebral palsy gain GMFM points with growth and therapy, which would also fit the finding that younger children improved more.
• Two authors work for the company that produced the cells, the company funded the medical writing, and the treating physician was also the person collecting the data.
• The walking tests cover 26 children, about one in six of the cohort, mostly at GMFCS I and II.
• No adverse events, serious events or dropouts are reported, and only children with complete data at visits 1 and 5 were analysed.
• The second course and the fifth injection fall outside the analysis, so the paper gives no information on longer follow-up.
• The discussion states that younger children did better “in all GMFCS subgroups” and describes a “moderate correlation” between gross motor gains and muscle strength. Neither statement matches the results as shown, which report no age analysis by subgroup and no correlation between strength and GMFM change.

Background Information:

Cerebral palsy is the most common motor disability of childhood, and standard therapies aim to manage symptoms. Umbilical cord MSCs are described as easy to obtain and of low immunogenicity, and the authors attribute their effects mainly to paracrine signalling (release of anti-inflammatory, angiogenic and neurotrophic factors) rather than replacement of brain cells. This reasoning was also used to justify switching between intravenous and intrathecal delivery.

In Poland, MSC products are given under the hospital exemption for advanced therapy medicinal products, as a last option when other treatments have failed. An earlier paper by two of the same authors (Boruczkowski and Zdolińska-Malinowska, 2019) reported parent-rated quality of life and self-sufficiency in children treated with the same cells.