The purpose of the study is to determine whether transfusion is associated with mild pulmonary changes.

Transfusion-related acute lung injury (TRALI) is the leading cause of blood product related death. The precise mechanisms of TRALI are uncertain. We hypothesize that TRALI is a full spectrum of lung injury. One theory holds that the infusion of cytokines associated with the transfusion of stored blood combined with the trauma and stress of surgery produce lung injury. We propose to examine lung function following transfusion, and that stored blood (and by implication, cytokines) injures the lung. TRALI is a rare condition and we do not expect any of the patients that we observe will develop TRALI. We are merely observing their pulmonary function closely to see if there are any mild changes.

Patients will be randomized into three groups based on whether or not they have pre-donated blood for the surgery. Once divided into these groups they are randomized to one of three groups with patients in the first group randomized to receive their stored (either autologous or allogeneic) "unwashed" blood first in the operating room using the standard cell salvage system. The second group will receive their stored (either autologous or allogeneic) "washed" blood first and the third group will receive blood from the cell salvage system first. Those who get the stored blood first will get the cell salvage blood as their second transfusion and those that get the cell salvage blood first will get stored "unwashed" blood as the second transfusion in the operating room. We will closely monitor pulmonary function and oxygenation, and try to observe whether there is early evidence of lung changes. If there is evidence of lung changes, then the blood is tested for cytokines and antibodies.

Patients who are undergoing substantial posterior spinal reconstruction are chosen as the study population because (1) the need for transfusion is highly predictable; (2) all have endotracheal tubes to allow for controlled ventilation and control of inspired gases; (3) the availability of stored non-leukoreduced autologous packed red blood cells (prbcs) will enable comparison of effects of autologous blood containing elevated concentrations of pro-inflammatory cytokines and lyso-phosphatidyl choline with that of autologous red cells without elevated concentrations of pro-inflammatory cytokines (salvaged cells); (4) all have arterial catheters allowing for frequent sampling of arterial blood for blood gas determinations; (5) all have central venous catheters to guide fluid therapy and ensure the absence of fluid overload, thus eliminating an important confounder in assessing alterations of lung function after transfusion. For those not having autologous blood, it will be possible to compare the effects of blood containing elevated concentrations of pro-inflammatory cytokines and lyso-phosphatidyl choline normally associated with blood storage with that of autologous red cells without elevated concentrations of pro-inflammatory cytokines (salvaged cells).

Inclusion Criteria:

Patients, male or female, any race or ethnicity, age 16-75 years, undergoing elective posterior spine surgery at University of California, San Francisco (UCSF) with expected sufficient blood loss to require intra-operative cell salvage and blood transfusion, who have donated blood for themselves preoperatively. -

Exclusion Criteria:

Patients are excluded who are under 16 or over 75 year of age. Also excluded are those that have pulmonary disease, abnormal pulmonary function or gas exchange by history or physical examination, and pre-operative measurement of oxyhemoglobin saturation (percutaneous); any operative procedure within one week of study; active infection; cardiac failure.

The study population is all adult spine surgery patients aged 16 to 75 who are scheduled for orthopedic spine surgery at UCSF who are expected to lose at least 1000 cc of blood during their surgery and who will receive both stored blood and cell-salvaged blood during their surgery.