Double-Unit Cord Blood (CB) Transplantation (DCBT) Combined with Haplo-Identical CD34+ Selected Peripheral Blood Stem Cells (PBSC) Is Associated with Enhanced Neutrophil Recovery, Universal Haplo Rejection and Frequent Pre-Engraftment Syndrome

Introduction While DCBT provides high rates of sustained donor engraftment, delayed neutrophil recovery is frequent and associated with increased transplant-related mortality.

Methods We are investigating the combined transplantation of a 4-6/6 double-unit CB allograft with CD34+ selected haplo-identical PBSC (haploCD34+) to speed myeloid recovery in patients (pts) with high-risk hematologic malignancies. This analysis evaluates sustained neutrophil recovery, engraftment patterns, and potential factors associated with haplo-rejection in 36 DCBT-haploCD34+ recipients.

Results 36 DCB-haplo-CD34+ pts (median 48 years, range 16-69) were transplanted 9/2012-6/2014. Diagnoses included 28 acute leukemias (24 CR or aplasia, 4 not in CR) and 8 advanced lymphoid malignancies. 2 pts had had a prior allograft. Conditioning was myeloablative (2 high-dose TBI, 34 reduced intensity) with CSA/MMF and no ATG. CB units had a median infused TNC (x 10⁷/kg) of 2.26 (larger unit) & 1.87 (smaller unit), & a median HLA-allele match of 5/8 (range 2-7/8). Haplo-identical donors (median 37 years, range 18-71) had a median HLA-match of 4/8 (range 4-6/8). The median infused haplo-CD34+ dose was 3.1 x 10⁶/kg (range 1.1-7.5). The median infused haplo-CD3+ dose was 1.5 x 10³/kg (range 0.3-8.0). One patient died early post-transplant (multiorgan failure). In 35 evaluable pts, engraftment patterns [groups (Gp) 1-4] based on count recovery & chimerism are shown (Table). Overall, 34/35 (97%) pts recovered neutrophils (median day 13, range 11-38) whereas one 2nd allograft pt with donor-specific HLA-Abs (DSA) to haplo & both units had both primary haplo-rejection and CB graft failure (GF, Gp 4). Additionally, one pt (Gp 4) with transient haplo-mediated neutrophil recovery had subsequent CB GF in the setting of low CD34+ dose in the dominant unit. Both GF pts were salvaged with single-unit CBT. In other pts, haplo-rejection was universal & was associated with dominant CB unit derived T-cell engraftment although rejection speed varied, low haplo dose may have contributed in 2 pts, and 3 of 4 pts overall with haplo-DSA had no haplo-engraftment.

Of 35 evaluable pts, 27/35 (77%) developed pre-engraftment syndrome (PES, fever >38.3°C not due to infection +/- rash, capillary leak) at a median 10 day onset (range 7-14). PES occurred in 11/18 (61%) Gp 1 pts with robust haplo myeloid bridge vs 16/17 (94%) Gp 2-4 pts with rapid haplo-rejection. Overall, 3/27 (11%) PES pts required ICU care although marked improvement was associated with corticosteroids.

Conclusions: DCBT-haploCD34+ is feasible & neutrophil engraftment is enhanced. However, given the haplo-graft can be rapidly rejected, and sustained engraftment is not guaranteed, the quality & dose of the CB unit(s) in this platform is critical. Preliminary analysis suggests the haplo-graft dose & haplo-DSA may be relevant in haplo engraftment whereas there is a high degree of dominant CB unit-haplo HLA-mismatch and its role is unclear. Finally, PES may be more common with prompt haplo-rejection, may be clinically severe, and mandates early diagnosis and pulse corticosteroids. Analysis of serial cytokine samples to further investigate PES biology is ongoing.