Host Vs. Graft Mismatches May Impact on Rejection of Haploidentical Bone Marrow Transplants in Thalassemia Patients Using Post-Transplant Cyclophosphamide

Abstract

Bone marrow transplantation (BMT) may cure and normalize health-related quality of life in over 80% of low-risk children with severe thalassemias (ST). Partially matched-related (haploidentical) transplants (Haplo) with the use of post-BMT cyclophosphamide as GVHD/rejection prophylaxis (PTCy) is an established safe and simple approach allowing BMT to be accessible to virtually all candidates even in settings with limited resources where most children with ST live. However, the experience with Haplo-PTCy in ST is still limited, particularly in terms of factors predictive of rejection, a complication relevant to ST patients because of multiple transfusions and immunocompetence. This study assessed a group of 22 patients with ST with a median age of 4.7 years (range 1.5 to 13.5), all with low-risk features, defined as liver size ≤ 2 cm from costal margin, who underwent Haplo-PTCy BMT between March 2017 and August 2018 in 2 centers in India, the South East Asia Institute for Thalassemia in Jaipur (16 cases) and the People Tree Hospital in Bangalore (6 patients). All received partially matched grafts, 5 from the father and 17 from the mother with a uniform regimen modified from Anurathapan et al. BMT 2016 (outlined in Fig. 1). G-CSF-primed bone marrow was used for all patients with a median cell dose of 16.2 nucleated cells/Kg (range 12.1 to 59.2). At a median follow up of 6.4 months (range 1.8 to 19.1) actuarial overall survival is 86%, thalassemia-free survival 72%, transplant-related mortality 9% and rejection 26%. One death was unrelated to BMT. Donor-Specific Antibodies (DSA) status was known in 21 patients, 4 patients (19%) were DSA+\xa0 (MFI > 2.000) and all but one rejected, interestingly, the DSA+ patient who engrafted had a GVH set up (donor HLA-A homozygous). Of 17 DSA− patients 3 rejected and all had a HVG set up, i.e. the recipient was homozygous for one or more HLA specificities while the donor was not: One was DRB1, one A & DPB1 (permissive), and one DRB1 & DQB1 homozygous. No rejections were observed in the absence of DSA-positivity or a HVG set up, so that actuarial rejection proportion went from 63% in DSA+ HVG+ patients to 0 in DSA− HVG− ones (log rank P=0.0023) (see Fig 2). All 15 patients who engrafted had complete donor chimerism and 9 (60%) developed acute GVHD, 8 grade I-II and one grade IV, this patent died of GVHD, interestingly he had a strong GVH set up (father donor A & C homozygous). CMV reactivation occurred in 3 engrafted patients (20%), other relevant complications included one macrophage activation syndrome who responded to steroids, one BK virus infection with severe late hemorrhagic cystitis, renal failure and death. In conclusion, in addition to DSA status, also HVG vs. GVH HLA set up might be relevant in the context of Haplo-PTCy for non-malignant conditions.