There are no advantages to liberal use of red blood cell (RBC) transfusion over restrictive use in moderate to high-risk patients undergoing cardiac surgery, the multicentre randomized noninferiority TRICS-III* trial has shown.

Liberal use refers to transfusion if haemoglobin level drops to <9.5 g/dL in the operating room or the intensive care unit and to <8.5 g/dL subsequently on the ward, restrictive use to transfusion if haemoglobin is <7.5 g/dL intraoperatively and postoperatively.

“Apparently, the restrictive approach was noninferior to the liberal approach,” said Dr Christopher Granger, director of the Cardiac Care Unit, Duke University Medical Center in Durham, North Carolina, US who is unaffiliated with the study. “Interestingly, the result was more favourable to the restrictive strategy – it was not statistically significant but was nonetheless reassuring. I think this will have impact on the guidelines.”

The primary composite outcome of hospital death, stroke, new renal failure requiring dialysis or new focal neurological deficit occurred in 11.4 percent of patients treated using the restrictive strategy vs 12.5 percent in those treated with the liberal strategy (p<0.001 for noninferiority). Mortality rates were no different between groups at 3 percent and 3.6 percent, respectively. [N Engl J Med 2017;doi:10.1056/NEJMoa1711818]

TRICS-III involved cardiac surgery patients from 19 countries with a preoperative European System for Cardiac Operative Risk Evaluation (EuroSCORE I) of ≥6, randomized to restrictive or liberal strategy and followed through day 28. Secondary outcomes included duration of hospital stay and mechanical ventilation; and individual perioperative adverse events including MI, stroke, delirium, infection, acute kidney injury, gut infarction, prolonged low cardiac output state, and platelets and plasma transfusion.

Of patients in the restrictive arm, 52.3 percent received allogeneic RBC transfusion vs 72.6 percent in the liberal arm (odds ratio [OR], 0.41; 95 percent CI, 0.37—0.47). There were no significant differences in the secondary outcomes between groups.

Optimal threshold for transfusion

“Current guidelines recommend transfusion for severe anaemia and recommend against it at haemoglobin level of >10 mg/dL,” said lead author Dr C David Mazer from the University of Toronto and St Michael’s Hospital in Toronto, Ontario, Canada. “In between, they don’t provide good guidance.”

In the setting of cardiac surgery, there is considerable uncertainty on the optimal threshold for transfusion. Observational studies suggest harm from the liberal transfusion approach, but there are very limited randomized controlled trials to confirm this. The TITRe trial, on the other hand, found no significant difference between the liberal and restrictive transfusion strategies in terms of morbidity or total costs postoperatively. [N Engl J Med 2015; 372:997-1008]

“In TRICS-III, we’ve shown that we can transfuse more sparingly and maintain patient safety and patient outcomes, while saving blood [which is a scarce resource] and its associated costs. There was remarkable consistency of results through various subgroups and sensitivity analyses, and I think that strengthens the message.”

Age does matter

Interestingly, in the subgroup analysis of the TRICS-III trial, patients 75 years and older did better with the restrictive strategy, suggesting that this group may be more susceptible to some of the harmful effects of transfusion (eg, infection and mortality).

“I find this very counterintuitive as most surgeons have a lesser threshold for transfusion in the elderly patients, those 75 and 80 years old. I was baffled by the results. This is extremely important and the long-term outcomes should be evaluated,” said discussant Dr Frank Sellke, Karl Karlson & Gloria Karlson professor and chief, Division of Cardiothoracic Surgery and director, Lifespan Cardiovascular Institute, Brown Medical School and Rhode Island Hospital in Providence, Rhode Island, US.

“Many people believe that the older you are, the higher your haemoglobin should be or the more liberally you should transfuse. The outcome that we saw was opposite to that,” Mazer said. “Restrictive transfusion is safe in these patients.”

Implication to practice

Two in three patients undergoing surgery receive blood transfusion. However, blood transfusion is not safe in some countries. “Blood is not only expensive, it is also toxic [in certain cases]. You don’t want to give more than what is necessary. With TRICS-III, there is now compelling evidence that less is better,” commented Dr Daniel Sessler from the Cleveland Clinic in Ohio, Canada, who is unaffiliated with the study. “So, with rare exceptions, we shouldn’t be transfusing above a haemoglobin of 7 [g/dL] and that applies to sicker and older patients, and even critical-care patients.”