Intra-Bone Marrow Transplantation of CD34+ Enriched Grafts Following a Non-Myeloablative Conditioning Allows Successful Engraftment In DLA-Identical Canine Litters.

Successful engraftment following transplantation of hematopoietic stem cells (HSCT) depends mainly on pre- and posttransplant immunosuppression, graft type and composition as well as on the HSC numbers infused. Whereas some of the aforementioned parameters can be influenced in the clinical setting, the latter one is more difficult to address. HSCTs of grafts with limited HSC numbers are accompanied by increased graft failure rates, longer cytopenias and increased morbidity. Current concepts to overcome low HSC numbers include the combination of two unrelated grafts, expansion techniques, modification of the graft composition or the site of graft infusion. In preliminary rodent studies intra-bone marrow (IBM) compared to intravenous (IV) HSCT led to faster engraftment which might be explained by closer location of the HSC to the stem cell niches. Aims: To investigate the feasibility and efficiency of IBM-HSCT following a non-myeloablative conditioning regimen in a dog-leukocyte antigen (DLA) identical canine HSCT model.

DLA-identical siblings were used as donor/recipient pairs for HSCTs. Recipients received a single dose of 2 Gy total body irradiation before HSCT (day 0). Pre- and postgrafting immunosuppression consisted of CSA (d-1 to d+35) and MMF (d0 to d+27). Two IBM-HSCT cohorts were investigated and data compared to IV controls (CON). BM-grafts of the respective donors were infused unmodified IV (CON, n=9) or IBM after HSC enrichment using a buffy coat followed by ficoll density centrifugation (IBM-I, n=6; 5ml total volume) or IBM after HSC enrichment using buffy coat centrifugation only (IBM-II, n=6; 25 ml total volume). In the CON group the graft was infused in the cubital vein. In the IBM-groups the grafts were infused through a BM aspiration needle into the BM of the left humerus and femur over a period of 5 minutes. In 4 IBM animals graft migration analyses were performed using technecium99 marking. Chimerism and BM cellularity were determined at injection and opposite sides. Analyses of chimerism were performed via polymorphic nucleotide repeat analyses weekly. BM cellularity was determined biweekly. Complete blood count was performed daily.

Infusion of grafts directly into the BM was feasible: both volumes (5ml, 25ml) could be infused without any leakage at the injection sites. Tc99-marked BM cells stayed predominately at the injection site for the first 24 hours. All animals engrafted. Mean TNC numbers infused were 2.6 ×10⁸/kg (range: 1.6–11.4; CON), 1.6 ×10⁸/kg (range: 1–2.4; IBM-I), 3.7 ×10⁸/kg (range: 2.1–5.8; IBM-II) (IBM-I vs CON: p=0.08, IBM-II vs CON: p=0.9, IBM-I vs II: p<0.02,). Mean CD34+ numbers infused were 0.6 ×10⁶/kg (range: 0.3–2.2; CON), 2.5 ×10⁶/kg (range: 0.3–6.4; IBM-I), 4.3 × 10⁶/kg (range: 1.3–6.5; IBM-II) (IBM-I vs CON: p=0.06, IBM-II vs CON: p=0.01, IBM-I vs II: p=0.4). PBMC chimerisms at d+14, d+28 and d+56 were 25% (range: 8–46), 36% (range: 5–63), 28% (range: 4–54) (CON), 12% (range: 6–18), 23% (range: 14–39), 18% (range: 6–64) (IBM-I) and 40% (range: 23–60), 61% (range: 31–84), 45% (range: 25–70) (IBM-II) (IBM-I vs CON: p=0.03, p=0.05, p=0.1, IBM-II vs CON: p=n.s. (all), IBM-I vs II: p=0.06, p=0.04, p=0.1). Recovery of hematopoiesis occurred significantly slower in both IBM-BMT groups compared to CON (p<0.002): mean leukopenia (<1.0 gpt/l) durations were 4 days (range: 0–15; IBM-I), 0.5 days (range: 0–2, IBM-II) versus 0 days (CON); mean thrombocytopenia durations (<50Gpt/l) were 19 days (range: 2–49, IBM-I), 10 days (range: 8–16, IBM-II) versus 4 days (range: 0–6, CON). However, if only grafts with <2.0 ×10⁶ CD34+/kg were analysed differences concerning leukocytes recovery times diminished between IBM and the IV groups, whereas they persisted in regards to thrombocytopenia. Analyses of cellularity and chimerism within a HSCT recipient (injection vs opposite site) revealed higher initial cellularities and significantly higher BM donor chimerisms up to day +56, both in favour of the injection site (p<0.03).

Infusion of HSC grafts up to volumes of 25ml directly into the BM is feasible and allows successful donor engraftment following non-myeloablative conditioning. Duration of cytopenias following IBM-HSCT is still significant, perhaps due to the loss of precursor cells during graft preparation. Further studies are warranted to determine optimal graft preparation and IBM application techniques.

No relevant conflicts of interest to declare.