Prestorage Leukoreduction of Red Cells in Elective Cardiac Surgery: Results of a Double Blind Randomized Controlled Trial.

Abstract

Between November 1999 and August 2002, consenting adult elective cardiac surgery patients at Oregon Health & Science University, Portland Veteran's Administration Medical Center, and St. Vincent's Hospital who were undergoing cardiopulmonary bypass (CPB) were randomized at admission to receive either prestorage leukoreduced red cells (PSL-RBCs) or standard red cells (S-RBCs) in a prospective double-blind fashion. Only data from those transfused were analyzed. Outcome measures included death at 60 days, 60 day infection rate, and length of hospital stay (LOS). Patients at all 3 institutions were operated on by the same group of cardiovascular surgeons. Given higher baseline infection rates for coronary artery bypass grafts (CABG) randomization was stratified by CABG vs valve replacement (VR). All RBCs were issued with blinding hoods. All platelet transfusion were prestorage leukoreduced. RBC transfusion rates were 30% for CABG, 38 % for VR, and 63% for CABG + VR. Infections were determined by infection control nurses using standardized Centers for Disease Control criteria from hospital surveillance and records and follow-up phone calls. Deaths were determined from hospital records and follow-up calls, and verified by National Death Index data. The PSL-RBC arm included 304 patients and the S-RBC arm 258 patients. The two groups were well-matched demographically and by cardiovascular risk factors. Intent-to-treat analysis showed a 60 day mortality of 9.7% in the S-RBC arm and of 4.9% in the PSL-RBC arm (p=0.029). Heart failure as the sentinel cause of death accounted for most of the difference (45.5% of deaths in the S-RBC group vs 13.3% in the PSL-LR group). Death rates were procedure specific: CABG alone > CABG + VR > VR alone. There was no significant difference between the S-RBC and PSL-RBC groups with regard to overall infection rate at 60 days. Most infections were superficial wound infections in the CABG patients; however groups did not differ in more serious infections such as bacteremia (p=0.369) or pneumonia (p=0.360). There was no significant difference between the groups with respect to LOS exclusive of in-hospital deaths. Our results essentially replicate in a North American context those of a previous European trial (Van de Watering et al Circulation 1998; 97:562) involving elective cardiac surgery patients undergoing CABG and/or VR surgery randomized to receive S-RBCs prepared by the European buffy coat method vs leukoreduced RBCs. Despite technical differences in RBC preparation, the excess deaths in both studies in the S-RBC group vs the leukoreduced group suggests that elective cardiac surgery patients undergoing CPB constitute an at-risk group both in the US and Europe which may benefit from use of PSL-RBC. The significance of transfusion-related immunomodulation (TRIM) in man has been the subject of intense controversy. Interestingly the cause of the increased mortality in the S-RBC group, both in this study and the European study, could not be explained by differences in infection rates. Given the preponderance of deaths in the CABG patients it is tempting to speculate this may reflect an interaction between residual passenger leukocytes and ischemia which is independent of the TH1/TH2 lymphocyte shift postulated to underlie TRIM.