Revisiting Old Practices: More Restricted Indication of Preoperative Autologous Blood Donation in Healthy Bone Marrow Donors According to Baseline Hemoglobin Levels

Background: Despite the cost-effectiveness of preoperative autologous blood donation (PAD) having been intensely debated over the last decade (due to a significant reduction in the risk of transfusion-transmitted diseases), no consensus has been reached as to the risk-benefit status of PAD for healthy bone marrow donors. There is concern regarding the occurrence of pre-procedure anemia caused by the autologous donation itself, consequently increasing the risk of transfusion, and the possibility of unnecessary hospital costs in the case of donors with collected and non-transfused red blood cell concentrates (RBC). Thus, this study aimed to evaluate the changes in hemoglobin levels related to PAD in bone marrow donors followed at our institution and to seek a hemoglobin cutoff with a predictive power for the actual need for this procedure.

Study Design and Methods: We conducted a retrospective study at the Hematology and Transfusion Medicine Center at the University of Campinas, Brazil, evaluating the data for all bone marrow donors registered at our institution between 2002 and 2016 who had donated at least one unit of autologous blood. Mean Hgb values were compared separately for donors who donated 1 or 2 units, at 3 time-points: before PAD collection, the morning before marrow harvest and soon after harvest. ROC curves were used to investigate possible Hgb cutoff points for prediction of transfusion requirement.

Results: Donors identified in the study period comprised 80 individuals: 76 collected only 1 unit and 4 individuals collected 2 autologous units. Mean baseline Hgb values did not differ significantly between the 2 groups [1 unit: 14.9(12.3-18.1) x 2 units: 13.57 (12.6-14.8), p=0.069]. After PAD collection, there was a significant drop in Hgb levels for the whole cohort of donors [14.82 (12-18.1) x 12.75 (8.9-17.4), p<0.001], more pronounced for the group that donated 2 units (Hgb levels at this time point, 1 unit: 12.8(8.9-17.4) x 2 units: 11.55(11.2-12.1), p=0.045). However, after marrow harvest, Hgb levels were similar for the 2 groups, and despite significant declines in Hgb levels, none of the donors in our cohort required allogeneic transfusion and only 61.2% required autologous transfusion. To further evaluate the use of Hgb levels as a predictor for the efficacy of collecting autologous blood, ROC curve analysis identified baseline Hgb< 14.35g/dL as the sensitive cutoff to predict the need for transfusion after marrow harvest. These levels had a sensitivity of 52% and specificity of 80.4%, p=0.001 and OR=4.38 (95% confidence interval: 1.65-12.59).

Conclusion: In the current era of Patient Blood Management, mechanisms that prevent allogeneic blood transfusions, such as PAD, must be explored and debated, since they reduce patient exposure to risks related to allogeneic blood. Our analysis demonstrates the possibility of using hemoglobin thresholds as cutoff points for indication of PAD, tending to a more cost-effective approach. Furthermore, despite significant declines in Hgb levels after PAD, none of the donors in our cohort required allogeneic transfusion, demonstrating the safety of this procedure. Thus, the indication of PAD remains an option for those donors who feel insecure despite higher baseline Hgb levels.