Abstract

Introduction

Blood transfusion is associated with serious adverse events (SAE) such as poor surgical outcome, higher surgical mortality. The cost of transfusion is escalating due to increasing demand and limited blood supply. There is a worldwide growing interest in improving blood utilization methods and storage. Identification of reliable predictors of surgical bleeding is of upmost importance in order to reduce the number of unnecessary transfusions, save limited blood resources, and lower the related costs. Cell derived microparticles (MP) are small membrane vesicles of 0.1 µm in size, released during cell activation and apoptosis. They play an important role in hemostasis and thrombosis.

Methods

Data presented herein are from an NIH funded prospective randomized study to investigate role of cell derived MP in surgical complications in CABG. Of a total of 122 patients, 81 received transfusion during and/or after surgery (Tx group) whereas the remaining 41 did not (NoTx group). In addition to routine laboratory tests, special assays were performed pre-surgery including concentrations of MP from platelets (PMP), red cells (RMP), endothelia (EMP), and leukocytes (LMP). The two groups were compared with respect to pre-surgical variables in order to assess potential predictors of the need for surgery.

Results

There were no significant differences between the two groups with respect to most of the pre-surgical demographic, clinical, and lab variables, except that Tx group was associated with higher incidence of female gender, type O blood group, and heparin use. On the other hand, the NoTx group was associated with higher pre-surgical levels of RMP and PMP. Using stepwise logistic regression analyses, four factors were identified as the most significant predictors of the need for transfusion. As depicted by an the areas under the curve (AUC) of the ROC curve, the 95% Confidence Interval of the AUC, and their corresponding p-value for each factor, were as follows.

For the model combining all four variables, the AUC and its 95% CI were 0.86 [0.78 - 0.94]. In addition, analysis of post-surgery data revealed better outcomes for the NoTx group in terms of complications such as hours on mechanical ventilation (mean ± SD: 9.2 ± 7.0 vs 15.5 ± 17.1, p = 0.005), number of SAE other than death (0 vs 9 (11.1%), p =0.028), and number of in-hospital deaths (0 vs 5 (6.1%), p= 0.167).

FactorAUC[95% C.I]p-value
HGB 0.66 [0.55 - 0.78] 0.005 
aPTT 0.68 [0.57 - 0.79] 0.015 
RMP 0.69 [0.58 - 0.80] 0.005 
CD41+PMP 0.77 [0.67 - 0.86] <0.001 
FactorAUC[95% C.I]p-value
HGB 0.66 [0.55 - 0.78] 0.005 
aPTT 0.68 [0.57 - 0.79] 0.015 
RMP 0.69 [0.58 - 0.80] 0.005 
CD41+PMP 0.77 [0.67 - 0.86] <0.001 
Discussion

This is the first report to show that pre-surgical levels of circulating MP are highly significant predictors of need for transfusion in CABG. This finding is consistent with other evidence that MP play a major role in hemostasis. Implementation of pre-surgery MP assay will greatly imporove assessment of risk of bleeding in surgery, and requirements for transfusion. This could reduce the number of transfusions, with corresponding improvement in patient outcomes, as well as reduced costs and strains on limited blood supplies. Furthermore, it may be feasible to manipulate pre-surgical MP levels in patients to minimize bleeding and Tx requirements. This could be done by increasing rate of MP generation or retarding rate of clearance, or both. Alternatively, MP could be produced in vitro (autologous or not) for infusion during surgery to minimize blood loss. Such approach is in progress [Jy, et al Thromb Haemost, in press].

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.