Outcomes of Autologous Bone Marrow Mononuclear Cell Administration in the Treatment of Neurologic Sequelae in Children With Spina Bifida

Major Points and Findings:

This is the largest published study of cell therapy given after birth to children who already live with the consequences of spina bifida, namely constipation and soiling, a bladder that does not empty or hold properly, and weak legs. Eleven children in Hanoi, Vietnam, received two intrathecal infusions of their own bone marrow mononuclear cells, six months apart. The route is the lumbar injection many cell clinics use, but the cells differ. They were a fresh, unexpanded bone marrow fraction in which fewer than one cell in a thousand was a mesenchymal stem cell.

The results are encouraging, especially for bowel function, and no serious adverse event occurred. However, the study had no control group, no blinding and no statistical testing, and several measurements were available in only six or seven of the eleven children. It is a safety study with promising signals and does not establish that the treatment works.

Aim:

To evaluate the safety of intrathecal autologous bone marrow mononuclear cell (BMMNC) infusion and its effect on bowel, bladder and leg function in children with spina bifida.

Methods:

An open-label, uncontrolled phase I trial at Vinmec Times City International Hospital, run from July 2016 to December 2020 and registered retrospectively in July 2022 (NCT05472428).

Eligible children were of either sex, aged 6 months to 15 years, with lumbar spina bifida that had already been surgically closed. They had to have both a bowel disorder (constipation or faecal incontinence) and urinary dysfunction (retention or leakage). Excluded were children with spinal clefts at chest, neck or other levels, a clotting disorder, acute or chronic infection, kidney or liver failure, or complex heart disease.

Bone marrow was taken from both anterior iliac crests under general anaesthesia. The volume was set by body weight at 8 mL per kg for children under 10 kg, and at 80 mL plus 7 mL for each kg above 10 kg for heavier children, up to 200 mL. Mononuclear cells were separated by density-gradient centrifugation in a cleanroom, washed twice and suspended in 10 mL of saline. Release criteria were endotoxin below 5.0 EU/mL, viability above 80% and a negative mycoplasma test.

For the infusion, an 18-gauge spinal needle was placed under general anaesthesia between the 4th and 5th lumbar vertebrae. Three millilitres of cerebrospinal fluid were allowed to drain, then the 10 mL cell suspension was infused by electric pump over 30 minutes. The whole procedure, including a fresh marrow harvest, was repeated six months later. Children stayed in hospital for 72 hours after each infusion and were followed for 12 months.

Outcomes were adverse events, bowel function (Bristol stool scale, Rome IV criteria, anorectal manometry), bladder function (symptoms, cystometry, voiding cystourethrogram) and leg function (walking ability, manual muscle testing). Only descriptive statistics were used.

Results:

Eleven children were treated (mean age 4.5 years, five boys and six girls). Every child needed help to pass stool at baseline and ten used clean intermittent catheterisation. The mean total dose was 839 million mononuclear cells at the first infusion and 959 million at the second, containing about 66 and 63 million CD34-positive cells, with viability above 95%. Mesenchymal stem cells made up only 0.083% of the infused cells.

Safety: No severe adverse events occurred. Seven adverse events were judged possibly related to the treatment: headache (3), vomiting (1), fever (1), pain in the urethra (1) and swelling at the marrow harvest site (1). All settled without treatment.

Bowel:

• Children needing full assistance to empty the bowel (enema, abdominal pressure or manual removal) numbered 11 of 11 at baseline, 6 after the first infusion and 3 after the second. At the end 8 of 11 could pass stool unaided, and 2 children showed no bowel improvement at all.
• Of the seven children old enough to report it, none could feel the urge to defecate at baseline. After the first infusion 5 could, and after the second 6.
• Hard, pellet-like stool (Bristol type 1) fell from 10 children to 2. Constipation by Rome IV criteria fell from 11 children to 6.

Bladder: Changes in bladder function were more modest. Ten of 11 children used intermittent catheterisation at the start, and the same ten still did at the end. No child began to pass urine voluntarily. Urine leakage fell from 9 children to 3, and the number who could feel the urge to urinate rose from 1 to 4 of the seven who could be assessed. In the seven children who had cystometry, mean bladder capacity rose from 127.7 to 158.3 mL at 12 months and maximum detrusor pressure fell from 32.4 to 21.9 cm H2O. The fall in pressure is relevant because high bladder pressure threatens the kidneys. Vesicoureteral reflux disappeared in one of two affected children. Abnormal bladder wall appearance (6 of 10 children) did not change.

Legs: Children walking independently rose from 6 to 8. One of the two new walkers had been under 12 months old at baseline, too young to walk. Those needing ankle-foot orthoses fell from 3 to 2, and one child’s stiff ankle was unchanged. Children with normal leg muscle strength on manual testing went from 2 to 3 of the 9 who could be tested.

The paper contains an inconsistency. Bladder capacity expressed as a percentage of the capacity expected for the child’s age fell, from a median of 78.4% to 59.8%, which the text acknowledges. The figure caption says it “improved”. The bladders grew, but more slowly than normal growth would predict.

Conclusions:

The authors conclude that intrathecal autologous BMMNC infusion is safe and may improve bowel, bladder and motor function in children with spina bifida, and they call for a larger randomised trial “to obtain an accurate conclusion”.

Several features of the study support that caution. There was no comparison group, and these were young children (mean age 4.5) who were growing, developing and receiving ongoing care over the year, so some change would be expected anyway. The paper describes several changes as “significant” although no statistical tests were performed. Assessors and parents knew the treatment had been given. Anorectal manometry was missing in five children (two were under 2 years old and the parents of three refused it) and cystometry was missing in four. The article also carries a correction notice (May 2023) stating that a reference to a patient was removed for confidentiality.

Background Information:

Treatment costs were covered by a charity, the Kind Heart Foundation, and families paid nothing. All authors are employees of the Vinmec healthcare system or its stem cell research institute, which also paid the registration and publication costs.

On the choice of the intrathecal route, the authors argue that cells placed in the spinal fluid can reach the cord directly, whereas most cells given into a vein are trapped in the lungs or spleen. Lumbar injection has long been used in children for chemotherapy, baclofen and spinal anaesthesia.

The paper cites two earlier human reports. A 2019 study by Boruczkowski and colleagues gave Wharton’s jelly mesenchymal stem cells to children with spina bifida and was reported as safe, with gains in motor, bladder, bowel and cognitive function. A series of 13 patients by Sharma and colleagues received bone marrow mononuclear cells without complications. Neither was a controlled trial. The proposed mechanisms (growth factor release, reduced cell death, new blood vessels) all come from animal work.