Significant Incidence of Mixed Donor Chimerism Despite TBI Conditioning in Paediatric B-ALL Patients Bridged to Transplant with Blinatumomab – a Report from the UK and Ireland Paediatric Bone Marrow Transplant/Cellular Therapies Group

Introduction: Blinatumomab is an effective tool to bridge paediatric patients with ALL to HSCT. Blinatumomab containing regimens pre HSCT are associated with lower levels of MRD and less toxicity compared with conventional chemotherapy relapse protocols (Locatelli, F et al. JAMA. 2021;325(9):843-854. doi:10.1001/jama.2021.0987). This allows more children to proceed to HSCT with negative MRD. Historically, post-HSCT mixed chimerism (MC= <95% donor) identifies patients with high RR. The impact of prior blinatumomab on post HSCT engraftment dynamics, chimerism or the impact on RFS/ OS is not well understood.

Methods: Data were collected from 57 consecutive paediatric patients aged 1–16 years who received blinatumomab prior to HSCT in 7 UK and Republic of Ireland transplant centres between 2014-2022. Data were collected from a second study group who received conventional chemotherapy relapse protocols without blinatumomab (chemo-only group) pre HSCT at 2/7 UK centres between 2010-2021. Infants, Primary Refractory, Mixed phenotype acute leukemia and patients with less than 100 days follow-up (FU) post-HSCT were excluded. Peripheral blood chimerism at 1, 3, 6, 9, 12 mo and last FU post HSCT were analysed. Primary and secondary graft failure (GF) were defined as failure to achieve donor engraftment at 1 mo post HSCT or initial engraftment followed by whole blood donor chimerism <50% at any time point, respectively. Relapse was defined as disease >0.01% MRD (flow cytometry or molecular).

Results: The blinatumomab (n=45) and chemo-only (n=40) groups were comparable in terms of demographics, conditioning regimen (93% TBI-based), in-vivo-T cell depletion, donor type and stem cell source. Median FU was 27 mo for blinatumomab and 29 mo for the chemo-only group. In the blinatumomab group the incidence of MC at 12 mo was 53% (24/45) and 57.7% (26/45) with a longer FU (median FU 27 mo) while in the chemo-only group the incidence was 20% (8/40) MC (median FU 29 mo). Increasing number of blinatumomab cycles was associated with a significantly higher incidence of MC (1 cycle 32% vs >2 cycles 74%, p= 0.02). Mixed chimerism developed early post HSCT with 91% MC occurring within 6 mo in both groups. The levels of MC varied in the blinatumomab group: 7/26 patients had 95%-99% donor chimerism, 12/26 had 50%-94%, 5/26 had 20%-49% and 2/26 had <20% MC. There were 2 primary and 4 (8.8%) secondary GF at 12 mo FU and 7 (15.5%) with longer FU (median 29 mo) in the Blinatumomab group. No GF observed in the chemo-only group (p= 0.013). Overall RR was 24.7%: 17.7% (8/45) in the blinatumomab and 32.5% (13/40) in the chemo-only group (p= 0.137). Even with an increased FU period (from 12 mo to 27 mo), OS was significantly higher in the blinatumomab group compared to the chemo-only group 87.4% vs. 63.8%. (p= 0.03). Relapse free survival was also superior in the blinatumomab group (74.7%) compared to the chemo-only group (67.2%) (P= 0.1). Within the blinatumomab group, the OS was 92% in the MC group vs. 82% in the full donor chimerism (FDC) (p= 0.47) group, while the RFS was similar with 75.5% in the MC and 74.9% in the FDC group (p= 0.72). Eight of 45 patients (17.7%) in the blinatumomab group relapsed of whom 4 had MC and 4 FDC. Of the 4 relapses in the MC blinatumomab group, 2/4 had 95%-99% donor chimerism, 1/4 at 50%-95% and 1/4 20-49%. Interestingly there were no relapses in the <20% MC sub-group. Immunosuppression cessation or DLI in response to MC had no significant impact on RFS in the analysed groups.

Conclusion: This is the first report of pre HSCT blinatumomab being associated with significantly increased incidence of MC post HSCT in paediatric ALL patients. We hypothesise that the MC reflects reduced pre transplant treatment intensity as the current UK Relapse Guidelines combines blinatumomab with less intensive reinduction chemotherapy than IntReALL or COG protocols. However, the improved OS and RFS in the blinatumomab group despite MC is difficult to explain. The incidence of GF was increased in the blinatumomab group (8.8% at 12 mo and 15.5% at 27 mo FU), and this remains of some concern, although this had no significant impact on RFS and OS. The RR in the blinatumomab group did not correlate with the degree of MC. This may reflect recovering autologous immunity post blinatumomab leading to allogeneic rejection. These intriguing findings need longer FU, larger patient numbers and prospective trials to further understand these data and draw firm conclusions.