An intra-bone marrow (IBM) hematopoietic stem cell transplantation (HSCT) is assumed to optimize the homing process and therefore to improve engraftment as well as hematopoietic recovery compared with conventional i.v. HSCT. This study investigated the feasibility and efficacy of IBM HSCT after nonmyeloablative conditioning in an allogeneic canine HSCT model. Two study cohorts received IBM HSCT of either density gradient (IBM-I, n = 7) or buffy coat (IBM-II, n = 6) enriched bone marrow cells. An historical i.v. HSCT cohort served as control. Before allogeneic HSCT experiments were performed, we investigated the feasibility of IBM HSCT by using technetium-99m marked autologous grafts. Scintigraphic analyses confirmed that most IBM-injected autologous cells remained at the injection sites, independent of the applied volume. In addition, cell migration to other bones occurred. The enrichment process led to different allogeneic graft volumes (IBM-I, 2 × 5 mL; IBM-II, 2 × 25 mL) and significantly lower counts of total nucleated cells in IBM-I grafts compared with IBM-II grafts (1.6 × 10/kg versus 3.8 × 10/kg). After allogeneic HSCT, dogs of the IBM-I group showed a delayed engraftment with lower levels of donor chimerism when compared with IBM-II or to i.v. HSCT. Dogs of the IBM-II group tended to reveal slightly faster early leukocyte engraftment kinetics than intravenously transplanted animals. However, thrombocytopenia was significantly prolonged in both IBM groups when compared with i.v. HSCT. In conclusion, IBM HSCT is feasible in a nonmyeloablative HSCT setting but failed to significantly improve engraftment kinetics and hematopoietic recovery in comparison with conventional i.v. HSCT.

The canine hematopoietic stem cell transplantation (HSCT) model has become accepted in recent decades as a good preclinical model for the development of new transplantation strategies. Information on factors associated with outcome after allogeneic HSCT are a prerequisite for designing new risk-adapted transplantation protocols. Here we report a retrospective analysis aimed at identifying risk factors for allograft rejection in the canine HSCT model. A total of 75 dog leukocyte antigen-identical sibling HSCTs were performed since 2003 on 10 different protocols. Conditioning consisted of total body irradiation at 1.0 Gy (n = 20), 2.0 Gy (n = 40), or 4.5 Gy (n = 15). Bone marrow was infused either intravenously (n = 54) or intraosseously (n = 21). Cyclosporin A alone or different combinations of cyclosporine A, mycophenolate mofetil, and everolimus were used for immunosuppression. A median cell dose of 3.5 (range, 1.0 to 11.8) total nucleated cells (TNCs)/kg was infused. Cox analyses were used to assess the influence of age, weight, radiation dose, donor/recipient sex, type of immunosuppression, and cell dose (TNCs, CD34(+) cells) on allograft rejection. Initial engraftment occurred in all dogs. Forty-two dogs (56%) experienced graft rejection at median of 11 weeks (range, 6 to 56 weeks) after HSCT. Univariate analyses revealed radiation dose, type of immunosuppression, TNC dose, recipient weight, and recipient age as factors influencing long-term engraftment. In multivariate analysis, low radiation dose (P < .001) and low TNC cell count (P = .044) were identified as significant independent risk factors for graft rejection. Peripheral blood mononuclear cell chimerism ≥30% (P = .008) and granulocyte chimerism ≥70% (P = .023) at 4 weeks after HSCT were independent predictors of stable engraftment. In summary, these data indicate that even in low-dose total body irradiation-based regimens, the irradiation dose is important for engraftment. The level of blood chimerism at 4 weeks post-HSCT was predictive of long-term engraftment in the canine HSCT model.