FiiRST-2 Trial Bolsters Argument Against Coagulation Factor Use as Alternative to Frozen Plasma in Severe Trauma

Interim results of the FiiRST-2 trial reinforce recent evidence from the CRYOSTAT-2 and PROCOAG studies that fibrinogen concentrate (FC) and prothrombin complex concentrate (PCC) are not superior to frozen plasma in severe trauma. The findings were published in JAMA Network Open.

FiiRST-2 was halted at interim analysis for futility, as pre-emptive FC/PCC offered no clear advantage in 24-hour blood product use for patients requiring a massive hemorrhage protocol (MHP). Conditional power was less than 25%, meaning that an unfeasibly large number of patients would be needed to demonstrate the superiority of FC/PCC.

The parallel-control superiority trial enrolled patients at six level I trauma centers in Canada aged 16 years or older and required MHP initiation within one hour of admission. Patients were randomized to receive FC-PCC or frozen plasma. Patients in the intervention group received two consecutively numbered MHP packs, each with 4 grams FC and 2,000 international units (IU) of PCC; those in the control group received two packs, each with two units of frozen plasma. Both groups received four RBC units in each pack, along with four units of packed platelets in the second pack. The intent was to limit the RBC:plasma ratio to no more than 2:1.

Patients in the control group were permitted FC if fibrinogen dipped to below 1.5-2.0 g/L. However, no PCC was allowed for the control group. Clinicians were allowed to decide if and when to release packs, but they were required to administer all PCC or FC in the first pack before beginning the second pack. In addition, if a second pack was opened, all components were to be administered within 24 hours or termination of the MHP (whichever came first). The authors noted that the order of transfusion of items within a pack was at the discretion of the treatment team, though it is common practice in Canada to alternate RBC and plasma administration.

If additional packs were required, these were administered per facility MHP, and plasma was used for factor replacement. All sites administered 2 g of tranexamic acid within three hours of injury.

The primary outcome was the average number of blood products administered within 24 hours; secondary outcomes included the average number of blood products administered within 24 hours, excluding frozen plasma in the first two packs; individual blood component use within 24 hours; the use of rescue hemostatic agents (recombinant factor VIIa, PCC and FC) from the third pack and after; laboratory endpoints; 24-hour and 28-day all-cause mortality; and time to death. Safety endpoints included the incidence of thromboembolic events (TEE) up to 28 days, ventilator days and intensive care stay length.

A total of 217 patients were randomized — 107 to receive FC/PCC and 110 to frozen plasma; 66 (61.7%) patients received the MHP intervention pack, and 71 (64.5%) received the control pack, for a total of 137 patients included in the intent-to-treat analysis.

At the time the trial was halted, the average number of blood products given within 24 hours was statistically similar for the two groups — 28 versus 23.8 units in the intervention and control groups, respectively. In addition, there were no significant differences in the numbers of each blood product administered by 24 hours. There was also no significant difference in mortality at 24 hours between the two groups, nor were there any differences for time to death or risk distribution for all-cause mortality at 28 days.

While not significant, TEEs were more common among patients receiving FC/PCC (14) compared with frozen plasma (10).

It is noteworthy that the results of the CRYOSTAT-2 and PROCOAG trials were not yet available when the FiiRST-2 trial was started. The CRYOSTAT-2 trial compared the use of plasma plus empiric fibrinogen replacement with standard care (plasma with delayed fibrinogen replacement). No differences in mortality of thromboembolic complications were seen between the two groups. The PROCOAG trial examined the use of pre-emptive four-factor PCC versus placebo as part of a ratio-based replacement strategy. While no differences in mortality were seen, the risk of thromboembolic events was greater with PCC.

In a related commentary, Pierre Bouzat MD, PhD; and Tobias Gauss, MD, argue that “the FiiRST-2 trial does not provide sufficient evidence to recommend the empirical use of FC and PCC as a replacement for plasma in patients with trauma at risk of massive transfusion.”

They noted that the findings of FiiRST-2 and PROCOAG that TEE was more common (though not necessarily statistically so) with administration of FC/PCC suggest a potential safety concern.

Therefore, they argue that “selecting patients with trauma who may benefit from CFCs [coagulation factor concentrates] is a critical consideration in managing massive hemorrhage. It seems reasonable to target FC administration to patients with low fibrinogen levels and to reserve PCC for patients with impaired thrombin generation.”

Both Bouzat and Gauss are with the Centre Hospitalier Universitaire Grenoble Alpes in France.