PRIUM-Cell: In Utero Myelomeningocele Repair Programme Using Mesenchymal Stromal Cells

Major Points and Findings:

This paper is a research protocol, not a report of results. It describes PRIUM-Cell, a planned clinical trial at Armand-Trousseau Hospital in Paris that will add a patch of donor umbilical cord mesenchymal stromal cells to standard open fetal surgery for spina bifida (myelomeningocele). The patch is placed directly on the baby’s exposed spinal cord before the defect is closed, between 23 and 26 weeks of pregnancy. The original article is in French, with an English abstract.

No patient had been treated when the paper was written, so it contains no human safety or efficacy data. The case for the trial rests on the team’s own experiments in fetal lambs, where the patch improved movement and continence, and on safety data for mesenchymal cells used in other diseases. The trial is small (13 pregnancies), has no control group, and its main goal is to show that the patch can be applied safely. It is designed to decide whether a larger trial is worthwhile, not to prove that the treatment works.

Aim:

To assess whether a patch of allogeneic (donor) umbilical cord-derived mesenchymal stromal cells (UC-MSCs) can be applied during open fetal surgery for myelomeningocele, and whether it is safe for the newborn. Secondary aims are safety for the mother and for the child up to 36 months, and early signs of benefit for the child up to 36 months.

Methods:

The paper explains the reasoning behind the trial. Fetal surgery for spina bifida has been shown since 2011 to improve outcomes compared with surgery after birth, with a gain of one to two motor levels in 40% of children and half the risk of hydrocephalus needing treatment. The benefit is partial: only 71% of operated children walk independently and 62% still need daily bladder catheterisation. The team therefore developed a cell patch intended to improve spinal cord repair during the operation.

Over eight years the team tested patches made from sheep umbilical cord cells, and later from human umbilical cord cells, in a sheep model of spina bifida. With sheep cells they reported clearly better motor and urinary function, better preservation of spinal cord tissue and nerve cells, and less scarring between the cord and the skin. With human cells in sheep the motor benefit was smaller. No maternal, fetal or newborn side effects were seen, the cells did not travel away from the patch, and no cells were found in the spinal cord 53 days after surgery. The authors take this last point to mean the cells do not turn into nerve cells and probably work by releasing protective and anti-scarring factors. In the laboratory, the cells secreted factors such as BDNF, NGF, Galectin-1, MFG-E8 and HGF.

Results:

As a protocol, the paper reports what is planned rather than results.

Design: A single-centre, open-label, non-comparative, prospective phase I/II trial using a one-stage A’Hern design. There is no control group and no blinding.

Who can take part: Adult pregnant women with health insurance cover in France who speak French, carrying a single fetus with myelomeningocele or myeloschisis (a flat form of open spina bifida) whose upper level lies between the first thoracic (T1) and first sacral (S1) vertebra, with a Chiari type 2 malformation, at less than 26 weeks of pregnancy, and with no other malformation or genetic anomaly. Exclusions include twins or more, a high risk of preterm birth (for example a cervix shorter than 25 mm or a previous preterm birth), a previous uterine scar on the body of the uterus, placenta praevia, a body mass index above 35, a uterine malformation, an abnormal angle of the fetal spine, inability to have general anaesthesia, and HIV, hepatitis B or hepatitis C.

Number of patients: 13 pregnancies.

Cell product and dose: A patch of 17 million allogeneic UC-MSCs on a semi-rigid scaffold (the scaffold material is not described in this paper), produced to Good Manufacturing Practice standards by the Centre MEARY cell and gene therapy facility of the Paris public hospitals (AP-HP). The cell number was chosen from the team’s animal work.

Timing and procedure: The patch is applied to the whole exposed fetal spinal cord during the centre’s standard open fetal surgery at 23 to 26 weeks, and the defect is then closed in two layers over it. Mothers stay in hospital for 7 days of close monitoring, then are seen weekly until delivery.

Primary outcome: A combined measure of success. The whole patch must be applied and the defect closed over it, and the baby must have no complication at birth, meaning a healed scar with no wound infection and no cerebrospinal fluid leak. The trial will be judged worth pursuing if at least 8 of the 13 procedures succeed. The thresholds are a 30% success rate below which the technique is considered not effective enough and 70% as the target.

Secondary outcomes:

• Maternal complications during surgery, pregnancy, delivery and after birth (for example bleeding needing transfusion, early rupture of the membranes before 37 weeks, infection in the womb, pre-eclampsia).
• Newborn safety: birth before 37 weeks and serious complications of prematurity, plus checks for tumours at 8 weeks (clinical examination, head ultrasound and brain MRI) and at 36 months (examination and brain MRI).
• Efficacy at 8 weeks: brain and spinal MRI or ultrasound (ventricle size, Chiari malformation, tethered cord, syrinx), motor level compared with the level seen before birth, urine stream and bowel habits.
• Efficacy at 36 months: MRI, need for a shunt or repeat spinal surgery, walking (independently, with braces or in a wheelchair), urinary tract ultrasound, urodynamics, urinary infections, need for catheterisation or bladder medicine, constipation, laxatives, enemas and nappies, Bayley-4 development testing and whether the child has started nursery school.

Funding and status: Funded by the French national hospital clinical research programme (PHRC national 2023, project PHRC-23-0079). The paper does not give a trial registration number, a start date or the regulatory authorisation status, and no patient results are reported.

Conclusions:

The authors hope the patch will improve spinal cord repair, and with it motor and bladder outcomes, in human babies as it did in sheep. If the results are positive, they plan a Europe-wide phase III trial, a version for keyhole (fetoscopic) surgery, and possibly other spinal cord operations in children and adults.

Some caution is needed. Everything said about benefit comes from animal studies, and with human cells the benefit in sheep was smaller than with sheep cells. The trial will compare its 13 children with no control group, so any later claim of better function will be hard to separate from the known effect of fetal surgery itself. The primary outcome is mainly surgical and safety success, not improved function. The authors write that looking for tumours is a “theoretical” choice because they do not expect harm, but a sample of 13 can only detect common problems. The paper also leaves out some practical details such as the scaffold material and registration information.

Background Information:

The team at Armand-Trousseau Hospital (AP-HP, Sorbonne University) runs the French rare disease reference centre for spina bifida and previously ran the PRIUM (NCT01983345) and PRIUM 2 (NCT04770805) fetal surgery trials. They see one to two referred patients a week. In France more than 90% of cases are diagnosed before birth, and 78 to 90% of these pregnancies end in medical termination depending on the region.

To support safety, the authors cite a meta-analysis of 55 randomised trials (2,696 patients) of mesenchymal cells given into the bloodstream, in which a passing fever was the only side effect linked to the cells, and a 2024 meta-analysis of 9 randomised trials (540 adults) of cells injected into the spinal fluid that found no serious side effects. Neither involved pregnancy or fetal treatment. They note that the only other team doing this work is Diana Farmer’s group at UC Davis, which uses early-gestation placental mesenchymal cells and began the CuRe trial (NCT04652908) in 2021. At the time of writing, its results had not been published.

The authors declare no conflicts of interest. Several of them published the earlier sheep studies on which the trial is based.