The impact of donor type and ABO incompatibility on transfusion requirements after nonmyeloablative haematopoietic cell transplantation

Abstract

We retrospectively analyzed transfusion requirements within the first 100 d among allogeneic haematopoietic cell transplantation (HCT) recipients with haematological malignancies given either myeloablative (n = 1353) or nonmyeloablative conditioning (n = 503). We confirmed that myeloablative recipients required more platelet and red blood cell (RBC) transfusions than nonmyeloablative recipients (P < 0.0001 for both). Myeloablative patients given peripheral blood stem cells required less platelet transfusions (P < 0.0001) than those given marrow while RBC transfusion requirements did not differ significantly. Subsequent analyses were restricted to nonmyeloablative recipients. Platelet and RBC transfusions were less frequent among related compared to unrelated recipients (P < 0.0001 for both), with comparable median numbers of transfused units. Major/bidirectionally ABO-mismatched recipients required more RBC transfusions than ABO-matched recipients (P = 0.006). Rates of graft rejection/failure, grades II-IV acute and chronic graft-versus-host-disease (GVHD), 2-year relapse, 3-year survivals and non-relapse mortality were comparable among ABO-matched, minor-mismatched, and major/bidirectionally mismatched recipients (P = 0.93, 0.72, 0.57, 0.36, 0.17 and 0.79, respectively). Times to disappearance of anti-donor IgG and IgM isohemagglutinins among major/bidirectionally ABO-mismatched recipients were affected by magnitude of pre-HCT titres (P < 0.001 for both) but not GVHD (P = 0.71 and 0.78, respectively). In conclusion, nonmyeloablative recipients required fewer platelet and RBC transfusions and among them, both unrelated and major/bidirectionally ABO-mismatched recipients required more RBC transfusions. ABO incompatibility did not affect nonmyeloablative HCT outcomes.

Keywords: ABO-incompatibility; allogeneic HCT; nonmyeloablative conditioning; transfusion.

Fig. 1. Comparisons of platelet and RBC (red blood cells) transfusions between nonmyeloablative and myeloablative HCT patients who received at least one transfusion

Nonmyeloablative recipients required less days of transfusion (A) and units of both platelets (B) and RBC (C).

Fig. 2. Comparison of platelet and RBC transfusions among related and unrelated nonmyeloablative patients who received at least one transfusion

Recipients given related grafts required less days of RBC transfusions (A). The days of platelet and the units of transfused platelets (B) and RBC (C) were not significantly different between the two groups.

Fig. 3. Influence of ABO-incompatibility on Platelets and RBC transfusion requirements among nonmyeloablative HCT patients who received at least one transfusion

(A) Recipients given major/bidirectionally ABO-mismatched grafts required more days of RBC transfusion than recipients given ABO-matched or minor mismatched grafts (P =0.004 and 0.026, respectively); (B) No significant difference between platelet units required by ABO-matched and –mismatched HCT recipients; (C) major/bidirectionally ABO-mismatched HCT required significantly more RBC transfusion units than the other two groups. *Reference group for comparisons.

Fig. 4. Patients with nonmyeloablative conditioning regimens

Cumulative incidences (solid lines) with 95% confidence interval (dashed lines) of disappearances of anti-donor IgG and IgM isohemagglutinin titres among 98 recipients of major/bidirectionally ABO-mismatched grafts.

Fig. 5. Impact of pre-HCT titre levels on post-HCT titre disappearance among major ABO/Bidirectionally mismatched recipients given nonmyeloablative conditioning

Hazard ratios (●) and 95% CIs (—) for the hazard of post-HCT disappearance of IgG (left panel) and IgM (right panel) titres for increasing quartiles of pre-HCT IgG and IgM titre levels, respectively, as compared to the lowest quartile for each (p<0.001 for both).

Fig. 6. Impact of ABO-compatibility on survival among nonmyeloablative HCT recipients

No statistically significant differences were observed among the three cohorts (P = 0.17).