TRICS III: Restrictive RBC transfusion is noninferior to a liberal strategy

SAN DIEGO — In moderate to high-risk cardiac surgery cases, a strategy of restrictive red blood cell transfusion decreases transfusion of allogeneic red blood cells, and is noninferior to a liberal transfusion approach in terms of mortality and major morbidity, according to data from the TRICS III trial presented at the 38th Annual Cardiothoracic Surgery Symposium.

“As perioperative hematocrit decreases, the risk of a variety of outcomes – including respiratory failure, renal morbidity, cardiac dysfunction, stroke or mortality – all increase,” C. David Mazer, MD, of St. Michael’s Hospital, University of Toronto, said in his presentation. “The inflection point seems to be somewhere near a hematocrit of 22 to 26% On the other hand, red cell transfusion is also associated with increased mortality.”

For the TRICS III trial, Mazer and colleagues hypothesized that lower hemoglobin concentration for red cell transfusion (a restrictive strategy) would be non-inferior to a liberal strategy in terms of mortality and heart, brain and kidney function.

In the randomized study conducted at 74 sites in 19 countries, the researchers evaluated 5,243 patients aged 18 years and older who were slated to undergo cardiac surgery with cardiopulmonary bypass. Eligible participants also had a preoperative additive EuroScore I of 6 or greater. Participants were randomized to one of two groups: a restrictive regimen, with a threshold of less than 7.5 g/dL in the operating room, ICU and ward; or a liberal group with transfusion thresholds of 9.5 g/dL in the operating room and ICU, with a step down to 8.5 g/dL on the ward.

Also included in the randomization were 208 patients from the TRICS II trial. A total of 5,092 participants were assessed in a modified intent-to-treat analysis, and the final analysis included the per protocol group of 4,860 patients.

Preoperative hemoglobin levels were identical; a disparity between the groups emerged in the OR, and this disparity remained throughout the hospital stay.

“If you look at transfusion data, there was essentially a 59% reduction in the probability of receiving any red cells in the restrictive group,” Mazer said. “This was highly significant.”

The primary composite outcome (death, MI, stroke, dialysis or renal failure) occurred in 11.4% of patients in the restrictive group vs. 12.5% of the liberal group (OR: 0.90; 95% CI, 0.76-1.07).

“The estimated risk difference between the two was -1.1, which slightly favors the restrictive group, and our prespecified, noninferiority margin was 3%,” Mazer said. “It was highly significantly indicative of noninferiority of the restrictive group compared to the liberal group.”

Various secondary outcomes were also analyzed, including ICU and hospital length of stay, duration of mechanical ventilation, infection, bowel infarction, AKI, seizure, delirium and encephalopathy. The odds ratios or hazard ratios for these outcomes were not clinically or statistically significant between the two groups.

In several subgroup analyses, the researchers evaluated the impact of the following on the primary outcome: age, sex, diabetes, preoperative pulmonary disease, surgery, LD function, and preoperative hemoglobin.

Of these, a significant interaction was seen only with age.

“This went in the opposite direction of what most people might have thought,” Mazer said. “That is, patients who were greater than 75 years of age did significantly better with restrictive transfusion than liberal transfusion, and they did significantly better than the younger group with restrictive transfusion.”

At 6 months postoperatively, no significant difference was seen between the groups in the primary composite outcome, as well as in an expanded composite outcome - revascularization or readmission or emergency department visit.

Cumulative mortality occurrence was also comparable between the groups at 6 months.

“The mortality occurrence on the Kaplan-Meier curves are virtually superimposable after the 30-day period,” Mazer said. “This is in contradiction to the TITRe2 study, which only followed patents to 90 days. At 180 days, or 6 months, there was no significant difference between the two groups in terms of mortality.” — by Jen Byrne

Reference:

Mazer DC. TRICS III. Presented at: The Annual Cardiothoracic Surgery Symposium; Sept. 6-9, 2018; San Diego.

Disclosure: Mazer reports being an investigator in the BARTS study group, and has served on guidelines committees. The TRICS III study was peer-review funded.