New Transfusion Threshold Affirmed for Cardiac Surgery
NOVEMBER 13, 2017
C. David Mazer, MDA restrictive red blood cell (RBC) transfusion strategy was non-inferior compared with a more liberal approach for patients undergoing cardiac surgery, according to phase 3 findings from the TRICS-III trial presented at the 2017 AHA Annual Meeting and simultaneously published in the New England Journal of Medicine.
In the large international trial, death from any cause, myocardial infarction (MI), stroke, or new-onset renal failure with dialysis by hospital discharge or by day 28 (the primary composite end point of the study) occurred in 11.4% of those in the restrictive group compared with 12.5% of those in the liberal threshold arm (odds ratio [OR], 0.90; 95% CI, 0.76-1.07). For the noninferiority analysis, this demonstrated a P value of <.001. Additionally, the two groups were also similar for secondary end points.
"The results were quite clear in establishing the noninferiority of the restrictive strategy for transfusion in cardiac surgery, that is there was no significant differences between groups in the primary outcome or any of the components of the primary outcomes, and the secondary outcomes," said lead investigator C. David Mazer, MD, from the St. Michael's Hospital and the Keenan Research Centre for Biomedical Science. "It will inform upcoming revisions to guidelines for care of patients undergoing cardiac surgery. I think it will provide the basis on which patients who are undergoing cardiac surgery will be transfused in the future."
In the open-label trial, 5243 patients were randomized to a restrictive RBC transfusion threshold of <7.5 g/dL hemoglobin or a liberal threshold of <9.5 g/dL. In the restrictive group, the <7.5 g/dL limit was started at the time of anesthesia and remained through discharge. In the liberal group, the limit was <9.5 g/dL in the operating room and intensive care unit (ICU) and was <8.5 g/dL in the non-ICU hospital ward through discharge.
Baseline characteristics were similar across groups, with a mean age of 72 years (±10). The mean EuroSCORE was 7.8 (±1.9). Overall, 35.4% of patients were women and surgeries consisted of coronary-artery bypass grafting (CABG) alone (26.1%), CABG with another procedure (27.7%), and a non-CABG procedure (46.2%).
Most events in the composite end point were MIs (5.9% in each group). The mortality rate was 3% in the restrictive group compared with 3.6% for the liberal threshold group (OR, 0.85; 95% CI, 0.62-1.16). In the restrictive and liberal groups, respectively, stroke was experienced by 1.9% and 2.0% of patients and renal failure with dialysis was seen in 2.5% and 3.0% of patients.
Transfusion was required for 52.3% of those in the restrictive group versus 72.6% of those in the liberal group, representing a significant reduction in the number of RBC transfusions needed (OR, 0.41; 95% CI, 0.37-0.47). The median number of RBC transfusions was 2 in the restrictive group versus 3 in the liberal arm.
"Noninferiority was consistent throughout both the prespecified subgroup analyses and the sensitivity analyses," noted Mazer. "One other component that was clearly demonstrated with this trial was that patients who were assigned to the restrictive transfusion were less likely to be transfused with allogeneic blood and when they were transfused they received fewer units of allogeneic blood than the liberal group."
Findings from the study provides data to support an evolution in care that was already under way at many institutions, he noted. Moreover, the study was conducted at several types of institutions globally, providing broad applicability to most settings.
"This provides for the first time a very high level of evidence that is broadly generalizable given the number sites in the study, there were 73 sites in 19 countries in every continent in the world, except Antarctica. I think there is broad generalizability of the findings," Mazer concluded.
Mazer CD, Whitlock RP, Fergusson DA, et al. Restrictive or Liberal Red-Cell Transfusion for Cardiac Surgery. N Engl J Med. 2017;11:DOI: 10.1056/NEJMoa1711818.