SOME things improve with age. Vintage wine, firewood, and one's memory of past triumphs come to mind. Unfortunately, stored blood is not among them. Blood preservative solutions were developed to eliminate reliance on vein-to-vein transfusion and to improve blood supply and logistics. 1No one expected that, at the end of several weeks of refrigerated storage, blood cells would perform exactly as they do when they exit the donor's vein. Small compromises in erythrocyte quality and quantity, commonly referred to as the storage lesion , are tolerated as the price of increased blood availability. However, clinicians do not expect this price to include harm to their patients. In this issue of Anesthesiology, Leal-Noval et al.  2add their caution to a lingering suspicion that something in stored allogeneic erythrocytes may increase morbidity in surgical patients. How legitimate is the concern that prolonged storage is the culprit?

Freshly collected erythrocytes lack a nucleus, mitochondria, the ability to divide, and the capacity to synthesize nucleotides and proteins. In the refrigerator, the cells age somewhat less gracefully than they do in vivo . The erythrocyte's metabolic machinery declines progressively during storage, and a number of biochemical and biomechanical changes occur that affect survival and function. 3Membrane loss by microvesiculation alters its lipid composition. With membrane loss comes some decrease in the flexibility that allows the cell to transit the microcirculation and exchange gases efficiently. Loss of membrane proteins and a shift to glycolysis make the erythrocyte more vulnerable to oxidative stress and lipid peroxidation. When stored in the presence of leukocytes, erythrocytes leak potassium, and increased hemolysis is observed. Histamines, lipids, cytokines, and a variety of other substances accumulate in the supernatant solution in a time-dependent manner. The organic plasticizer di(2-ethylhexyl)phthalate migrates from the blood bag into the erythrocyte lipid membrane. Additional manipulations, such as filtration and gamma irradiation, prior to transfusion may further modify the cells. The clinical importance of these aggregate changes is unknown.

A better understanding of erythrocyte biochemistry and the additives that influence it, as well as a bit of luck and serendipity, have resulted in preservatives and containers that have increased the storage interval progressively. In 1914, blood could be stored for no more than 6 days. Today, it is refrigerated for up to 6 weeks. Efforts to further extend erythrocyte shelf life continue as we try to conserve a dwindling margin of safety in blood supplies. 4Yet as newer preservative solutions are developed, the clinical effects of prolonging blood cell storage are investigated only obliquely. We lack good scientific assays of erythrocyte transfusion efficacy in vivo . One surrogate measure, 2,3-diphosphoglycerate, becomes undetectable by the end of storage, leading to reduced oxygen offloading ability for 12–48 h posttransfusion. The clinical impact of this loss has been difficult to demonstrate. Adenosine triphosphate, the accepted surrogate assay for erythrocyte viability, turns out to be a poor predictor of in vivo  circulation. The gold standard for viability remains survival of 75% of injected radiolabeled erythrocytes at 24 h, an arbitrary standard that tolerates up to a quarter of nonviable erythrocytes transfused at the end of storage. However, until compelling evidence that these changes affect transfusion outcomes emerges, research to extend storage will inevitably trump efforts to improve erythrocyte quality.

How do these findings relate to the report of Leal-Noval et al.  2? Numerous observational studies using multivariate analysis models have suggested that allogeneic erythrocyte transfusions have unexpected adverse effects, such as enhancing cancer recurrence and susceptibility to postoperative infections, an association that has come to be called transfusion-related immunomodulation . 5Controlled clinical trials and meta-analyses have been less convincing. 6–8Nevertheless, several candidate mechanisms for transfusion-related immunomodulation have been proposed, and the length of erythrocyte storage numbers among them. 9In addition, a small number of observational studies of different patient populations have demonstrated a statistical association between prolonged storage of allogeneic blood and (1) rate of infection in trauma patients, (2) mortality in the intensive care unit, and (3) postoperative pneumonia in open-heart surgery patients. 9–11Each of these studies suffers from limitations in size, design, or methodology. In a retrospective analysis of 416 open-heart surgery patients, Vamvakas and Carven 12reported that the mean length of storage of all erythrocyte units and the mean length of storage of the two oldest blood components were associated with postoperative pneumonia and infection. The findings of Leal-Noval et al. , 2although similar, are neither identical nor comparable. Their prospective study of 897 consecutive patients undergoing open-heart surgery shows an association between the oldest stored unit (> 28 days) and nosocomial pneumonia. No relationship was found with the mean length of component storage. Both studies used highly sophisticated, although different, statistical techniques to deal with the multiple confounding variables. Vamvakas and Carven 12included in their analysis, and accounted for, leukoreduced blood, leukoreplete blood, and combinations of autologous and allogeneic blood, as well as platelet and cryoprecipitate transfusions. Leal-Noval et al.  2used buffy coat–poor erythrocytes in a different preservative and excluded from analysis patients who received autologous blood, as well as children younger than 16 yr, patients with preoperative fever or infection, those with anemia (hemoglobin < 11 g/dl), and those who died within 48 h of surgery. While there are good reasons for each of these decisions, they do make interpretation and comparison problematic.

Observational studies serve the purpose of posing questions. Even a cursory reading of the several referenced studies that involve numerous confounding variables—disease state, patient demographics, nature and number of the blood components, concurrent treatments, and definition of adverse events—suggests that additional observational studies will not answer this question. No clinical study is better than its controls, regardless of the statistical legerdemain that is used. While the concern expressed by Leal-Noval et al.  2that prolonged storage may be a risk factor appears reasonable, their assertion that a randomized, controlled trial would not be ethically defensible is premature. There is insufficient scientific evidence to change practice and discard erythrocytes stored for longer than 28 days (or 14 or 21 days, as suggested in the other studies), a change that would have a huge impact on how transfusion services are delivered. Simply put, in the case of erythrocyte storage, we need to know whether and how much age matters.

Evidence that blood transfusions work remains largely empiric. Large-scale clinical trials of erythrocyte safety and efficacy are not currently required prior to extending the storage interval, nor would such studies likely be useful or practical. However, once observational studies suggest that stored blood is harmful, carefully controlled, hypothesis-driven, prospective trials to confirm or refute the results become imperative. In this case, we need to determine whether some storage lesion contributes to posttransfusion pneumonia, infection in trauma patients, or mortality in the critically ill. If a clinically significant effect is confirmed in any of these settings, we need to determine why it occurs and how to prevent it.

This Editorial View accompanies the following article: Leal-Noval SR, Jara-López I, García-Garmendia JL, Marín-Niebla A, Herruzo-Avilés A, Camacho-Laraña P, Loscertales J: Influence of erythrocyte concentrate storage time on postsurgical morbidity in cardiac surgery patients. Anesthesiology 2003; 98:815–22.

Robertson OH: Transfusion with preserved red blood cells. BMJ 1818; 1: 691–5
Leal-Noval SR, Jara-López I, García-Garmendia JL, Marín-Niebla A, Herruzo-Avilés A, Camacho-Laraña P, Loscertales J: Influence of red blood cell concentrates (RBCs) storage time on postsurgical morbidity in cardiac surgery patients. A nesthesiology 2003; 98: 815–22
Beutler E: Back to the future in RBC preservation. Transfusion 2000; 40: 893–5
Hess JR, Greenwalt TG: Storage of red blood cells: New approaches. Transfus Med Rev 2002; 16: 283–95
Klein HG: Immunomodulatory aspects of transfusion: A once and future risk? A nesthesiology 1999; 91: 861–5
Busch OR, Hop WC, Hoynck van Papendrecht MA, Marquet RL, Jeekel J: Blood transfusions and prognosis in colorectal cancer. N Engl J Med 1993; 328: 1372–6
Heiss MM, Mempel W, Jauch KW, Delanoff C, Mayer G, Mempel M, Eissner HJ, Schildberg FW: Beneficial effect of autologous blood transfusion on infectious complications after colorectal cancer surgery. Lancet 1993; 342: 1328–33
Houbiers JG, Brand A, van der Watering LM, Hermans J, Verwey PJ, Bijnen AB, Pahlplatz P, Eeftinck Schattenkerk M, Wobbes T, de Vries JE: Randomised controlled trial comparing transfusion of leukocyte-depleted or buffy-coat-depleted blood in surgery for colorectal cancer. Lancet 1994; 344: 573–8
Vamvakas EC: Possible mechanisms of allogeneic blood transfusion-associated postoperative infection. Transfus Med Rev 2002; 16: 144–60
Purdy ER, Tweeddale MG, Merrick PM: Association of mortality with age of blood transfused in ICU patients. Can J Anaesth 1997; 44: 1256–61
Offner PJ, Moore EE, Biffl WL, Johnson JL, Silliman CC: Increased rate of infection associated with transfusion of old blood after severe injury. Arch Surg 2002; 137: 711–6
Vamvakas EC, Carven JH: Transfusion and postoperative pneumonia in coronary artery bypass graft surgery: Effect of the length of storage of transfused cells. Transfusion 1999; 39: 701–10