Effect of storage‐aged red blood cell transfusions on endothelial function in healthy subjects

Support for these studies was provided by NHLBI (R01 HL095479-01 and an administrative supplement to JDR), NIH Grant UL1 RR025008, and the National Blood Foundation (to AR). ClinicalTrials.gov identifier: NCT00838331. URL: www.clinicaltrials.gov/ct2/show/NCT00838331. Prolonged storage of red blood cell (RBC) units is associated with biochemical changes that may cause adverse transfusion effects. These alterations include depletion of adenosine triphosphate and 2,3-diphosphoglycerate, hemolysis, formation of microparticles, and oxidation of lipids among other metabolic and cellular disruptions. These changes may produce depletion of bioavailable nitric oxide (NO) in the recipient, with subsequent endothelial dysfunction and inhibition of NO-mediated vasodilation. We have previously shown that hospitalized patients receiving storage-aged RBCs (saRBCs) have significantly decreased flowmediated dilation (FMD), indicating reduced endothelial NO activity compared to those receiving fresh RBCs (fRBCs). To examine whether these effects occur in healthy subjects, we examined the impact of saRBC transfusion on FMD in healthy volunteers, with the hypothesis that impact of saRBC will differ between subjects with and without endothelial dysfunction. In a crossover study, 16 healthy subjects (mean, age 31 years; 63% males; Table 1) donated 1 unit of whole blood and returned 1 to 7 days later to receive an autologous fRBC transfusion (<7 days of storage). Subjects later returned for the second phase of the study at least 5 days after their initial transfusion and donated an additional unit of blood. When this unit had been stored for 35 to 42 days, subjects were brought back and this saRBC unit was infused. With ultrasound, brachial artery FMD was measured before transfusion and at 1 and 24 hours after both the fRBC and the saRBC transfusions. FMD was calculated as (hyperemic diameter – baseline diameter)/baseline diameter 3 100. Statistical analysis of allometrically scaled FMD was performed using a linear mixed5effects modeling for repeated measures to account for baseline diameter. Enrolled subjects did not have known acute or chronic illnesses with normal blood pressure and lipid profile (Table 1). There were no significant differences in pretransfusion hemoglobin (Hb) between the fRBC and saRBC phases of the study (12.4 6 1.4 g/dL vs. 11.8 6 1.2 g/dL, respectively, p 5 0.1) nor after transfusion (13.6 6 1.3 g/dL vs. 13.4 6 1.2 g/dL; p 5 0.1). Resting brachial artery diameter was similar at all time points during each transfusion phase (Table 1). The pretransfusion (8.1 6 7.1% vs. 7.4 6 6.9%, respectively; p 5 0.9) and posttransfusion FMD measurements did not differ between the fRBC and saRBC phases (p 5 0.6, Fig. 1). However, after 24 hours of saRBC transfusion, FMD was higher (7.7 6 3.9% vs. 9.8 6 4.2%; p 5 0.019). Thus, in healthy subjects we found no decrease in FMD with autologous saRBC compared to fRBC transfusion, suggesting that NO bioavailability is not reduced in these transfusion recipients. While these findings differ from our previous results in hospitalized patients (Supplementary Table), they indicate that the recipient’s clinical status is likely the major determinant of whether saRBC transfusion can cause adverse effects. In support of this hypothesis, the mean FMD in our hospitalized patients was significantly lower than the current health population studied (5.1% vs. 7.8%, p< 0.001). The studies by Berra and colleagues in human transfusion recipients are in agreement, as they found no differences in the hyperemia index in nine healthy adults after fRBC versus saRBC transfusion. In contrast, and consistent with our previous work, they found that obese adults with endothelial dysfunction showed significantly increased pulmonary artery pressure after transfusion of saRBCs whereas transfusion of fRBC had no effects. Similarly, in a number of murine, rat,