Fixing bones before birth

Prenatal mesenchymal stem cell (MSC) transplantation for treatment of osteogenesis imperfecta (OI) has been reported in a human fetus.¹  In this issue of Blood, Guillot and colleagues present experimental validation of this approach by demonstrating improvement in the phenotype of the oim mouse after prenatal transplantation of human fetal MSCs.

OI consists of a group of connective tissue disorders usually caused by structural mutations in the COL1A1 and COL1A2 genes that encode type1 collagen. The clinical phenotype is highly variable, ranging from prenatal fractures and perinatal lethality to mild forms without fractures. Current treatments for OI include biphosphonate drugs, physiotherapy, and orthopedic surgery. There is no curative treatment, and therefore, new therapies are desperately needed.

Cellular therapy has been attempted for OI, but has not been curative. Postnatal allogeneic bone marrow transplantation (BMT) resulted in an increase in growth and other parameters that was sustained for only 6 months.²  Infusion of donor MSCs 18 to 34 months after the same BMT protocol resulted in a similar duration of benefit.³  In both reports, low-frequency engraftment of donor-derived osteoblasts was documented, but long-term persistence was not determined. Le Blanc et al reported treatment of a fetus with an expected severe nonlethal OI phenotype at 32 weeks gestation with allogeneic fetal-liver–derived MSCs.¹  The patient had a surprisingly benign clinical course, and low-frequency osteoblast engraftment was documented; however, interpretation was complicated by the variable relation between genotype and phenotype in OI, by the fact that biphosphonate was instituted at 4 months, and by limited duration of follow-up. These reports are promising because apparent benefit was observed despite minimal engraftment of donor osteoblasts. However, they raise important questions regarding the requirements for osteoblast reconstitution and, for that matter, the identity of the stem cell that maintains the osteoblast compartment.⁴ 

Here, Guillot and colleagues clearly demonstrate phenotypic improvement, but not cure, of the oim mouse after in utero transplantation of human, fetal-blood–derived MSCs—a unique and controversial source of donor cells. They convincingly demonstrate low-frequency engraftment, osteoblast differentiation, and function of the donor cells with apparent donor-cell participation in fracture healing. However, engraftment diminished over the course of the study (12 weeks) and no conclusions can be drawn regarding long-term osteoblast reconstitution.

Prenatal cellular therapy requires a compelling rationale, such as the preemption of clinical manifestations of a disease or the presence of unique developmental events that favor engraftment of stem cells. Both apply to the severe forms of OI.⁵  Conceptually, it is attractive to envision transplanting the appropriate osteogenic stem cell with perfect developmental timing to populate the nascently abnormal osteoblast compartment. The reality, however, is that we have not yet identified the appropriate stem cell, and the optimal timing for prenatal transplantation has not been defined. Guillot et al's study is a laudable first step toward gaining experimental insight for the informed design of optimal trials for prenatal cellular treatment of OI. Many more such studies are needed in relevant animal models to bring this promising and rational therapeutic strategy to successful clinical application.