To the Editor:—We read with interest the article about the influence of erythrocyte concentrate storage time on postoperative morbidity in cardiac surgery patients published by Leal-Noval et al. in a recent issue of the Journal, 1which was accompanied by an editorial view. 2The main finding of this study was an association between storage time of packed red blood cells (PRBC) and the acquisition of postoperative nosocomial pneumonia, but not between the former and prolonged mechanical ventilation time. We would like to comment regarding two points. First, we wish to review some mechanisms by which the relationship between the number of units and pneumonia/ventilation might be connected. Second, we want to point out that the problem may not simply be the age of the transfused blood, but also the absolute numbers of units transfused.The current authors suggested that allogeneic transfusion may cause some deleterious effects that have not been recognized and which may explain the association between the number of PRBC transfusions and prolonged mechanical ventilation, but not between the duration of storage and prolonged mechanical ventilation. 1Induction of a subclinical transfusion-related acute lung injury (TRALI) could be one of those frequently underdiagnosed deleterious effects. It has been proposed that TRALI is the result of two clinical events, the first being a predisposing clinical condition (e.g. , a postoperative systemic inflammatory response syndrome) and the second being the transfusion of biologically active molecules in stored PRBC and platelets that have neutrophil priming activity. The concentration of some of these molecules (e.g. , cytokines, neutral lipids, lysophosphatidylserine) in the stored blood component increases with the product age, whereas that of other molecules (e.g. , human lymphocyte antibody class I, class II, and granulocyte antibodies) does not. To this regard, Silliman et al. 3have found that TRALI was associated with the transfusion of older blood products, especially platelets, and that patients with cardiac disease who required cardiopulmonary bypass were at particular risk to develop TRALI reactions. However, although a significant proportion of patients (135, 16.9%) received platelet transfusion, Leal-Noval et al. 1have not studied the possible association between either the number or the age of platelet concentrates and prolonged mechanical ventilation time. Despite that, because the number of PRBC transfused was significantly higher in patients with mechanical ventilation time greater than 24 h, the development of a certain degree of TRALI, in a dose-dependent manner, may result in a worsening of the pulmonary dysfunction induced by cardiopulmonary bypass with a prolonged mechanical ventilation time and, perhaps, contributes to a higher incidence of postoperative nosocomial pneumonia. 4Regarding PRBC storage time, there are several storage-dependent alterations in erythrocytes that, together with transfusion-related immunomodulation, may contribute to the increased risk of nosocomial pneumonia in the transfused patients. The decrease in the ability of erythrocytes to deform and unload oxygen in the peripheral tissues, possibly favoring the degree of visceral ischemia, have been mentioned, 1,2but other alterations may account for when erythrocytes are stored together with white blood cells. Leukocyte enzymes, including neuraminidase, are definitive contributors to the desialylation of erythrocytes during storage and to the exposure of phosphatidylserine residues in the membrane outer leaflet. 5These changes make the erythrocytes more adherent to the inflamed endothelium and convert it into a procoagulant surface, thus contributing to the alteration of microvascular blood flow. In addition, these changes are also signals of senescence and enhance erythrocyte phagocytosis by resident macrophages. 5All methods in current use for erythrocyte storage result in the maintenance of 90% or more of the in vivo recovery for about 2–3 weeks; later an increasing proportion of erythrocytes become nonviable. When old PRBC are transfused, a substantial number of cells (some 1012erythrocytes in a four-unit transfusion) must be phagocytosed, which obviously uses part of the capacity of the phagocytic mononuclear system and may compete with infection defense with possible serious consequences, such as postoperative nosocomial pneumonia. 6We agree with Leal-Noval et al. 1with regard to the need for continued research of the possible deleterious effects of old PRBC transfusion. However, in previous work, 4this group showed an increase in the incidence of postoperative nosocomial pneumonia in those patients transfused with more than 4 units. In the present work, 1patients who developed postoperative nosocomial pneumonia received more PRBC units than those who did not (8.8 ± 4.7 vs. 5.4 ± 4.7, respectively; P < 0.05). Because the authors found that each transfused PRBC unit increased the risk of postoperative pneumonia by 11%, 1we believe that the implementation of a perioperative blood-saving program, including several alternatives to allogeneic erythrocyte transfusion (e.g. , preoperative treatment of anemia, aprotinin in pump-prime, perioperative autologous salvage), together with a restrictive transfusion strategy, 7–9would probably be a better option to reduce postoperative morbidity in cardiac surgery patients.