Possible Two Different Mechanisms of Engraftment Failure in Cord Blood Transplantation: Graft Rejection and Poor Graft Function

Engraft failure remains a critical issue to be solved especially after cord blood transplantation (CBT). The factors such as limited doses of infused total nucleated (TNC) and CD34⁺ cells, HLA disparity, and anti-HLA antibodies are considered to induce engraftment failure. We studied the patterns and mechanisms of engraftment failure for patients who failed engraftment after CBT.

Patients/methods: Medical charts on 429 transplant recipients of single unit CB with hematological diseases as the first allogeneic HSCT at Toranomon Hospital between January 2002 and May 2011 were retrospectively reviewed. Patients who did not meet the criteria of engraftment were subjected to the following analysis. Engraftment was defined when the neutrophil counts exceeded 0.5 × 10⁹/L for 3 consecutive tests. Patients who died or had progressive disease before day 28 post-transplant were excluded.

Among 429 recipients, 67 were excluded due to early death before day 28 (n=52) and disease progression (n=15). In remaining 362 patients, 31 were diagnosed as engraftment failure. Median age was 62 years (range, 17–71). Underlying diseases were AML (n=16), ALL (n=5), MDS (n=2), NHL (n=6), and SAA (n=2). Twenty-three (74%) were in high risk disease status (MDS RAEB and beyond, or AL, NHL not in remission). Conditioning regimens mainly comprised of purine analogue-based reduced-toxicity regimens with fludarabine phosphate (125-180 mg/m²), melphalan (80-140 mg/m²) or busulfan (8-16 mg/kg) and 0–4 Gy of total body irradiation (TBI), and others. Graft-versus-host disease (GVHD) prophylaxis comprised of tacrolimus (TAC) (n=13) or cyclosporine alone (n=8), and TAC + mycophenolate mofetil (MMF) (n=10). Median number of total nucleated cells (TNC) and CD34⁺ cells were 2.45×10⁶ /kg (range, 1.94 – 4.20), and 0.76×10⁵ /kg (range, 0.21 – 1.35), respectively. Three had 1 antigen mismatch and 28 had 2 antigen mismatches in serological HLA typing, and 15 had allelic mismatches in more than 3 loci between host and graft. Twenty-two (71%) showed recipient-dominant chimerism (donor type <5%) at the diagnosis of engraftment failure, which were designated as graft rejection, and the other 9 (29%) showed donor-dominancy (donor type >90%), designated as poor graft function. Anti-HLA antibodies were present in 7 of the 13 patients tested (54%), including 2 (28%) who had a donor-specific antigen that targeted against UCB unit used. All these 7 patients who had anti-HLA antibody showed the graft rejection pattern. T-cell donor-recipient chimeric status was assessed in 14 patients. Nine who showed complete recipient dominant chimerism in T-cell fraction (donor type <5%) at early transplant phase (median day15, range, 8–17) developed graft rejection. In 9 patients who achieved complete donor chimerism, 4 complicated persistent infection (sepsis, invasive pulmonary aspergillosis), and the other 5 developed hemophagocytic syndrome (HPS). There were no statistically significant differences between graft rejection group versus poor graft function group in terms of TNC, CD34⁺ cell dose, HLA disparity, disease type (myeloid vs. lymphoid), disease status (standard vs. high), pretransplant conditioning (presence vs. absence of melphalan, or TBI), and GVHD prophylaxis (presence vs. absence of MMF) in this study.

These data demonstrated that the graft rejection pattern comprised 70% of engraftment failure. Presence of anti-HLA antibody showed close correlation with graft rejection, suggesting antigen-mediated graft rejection mechanism following CBT. Recipient dominant T-cell chimerism at early transplant phase (around day 15) strongly indicates impending graft rejection. All patients who showed poor graft function pattern accompanied by severe infection and/or HPS. This pattern is rarely seen in HSCT of other stem cell sources, suggesting unique characteristics of immune cells in CB graft as we reported.

No relevant conflicts of interest to declare.