On May 16, 2005, the FDA published a notice to revise its "previous advice regarding the safety of albumin administration in critically ill patients." The notice formally acknowledged that the landmark SAFE (Saline vs. Albumin Fluid Evaluation) study has "resolved the prior safety concerns raised by the Cochrane Injuries Group" in a widely criticized 1998 meta-analysis that appeared in BMJ.
Access this notice.
The agency adopted the unanimous recommendation of its Blood Products Advisory Committee (BPAC) in mid-March, after hearing presentations by the lead SAFE investigator and others who pointed out (1) inherent flaws in the Cochrane study and (2) mounting evidence of important therapeutic advantages of albumin over crystalloids in specific clinical settings.
Some specific points from the FDA/BPAC meeting are worth noting. We hope that review and discussion of this information will precipitate rethinking of current policies about albumin access and use:
- While overall mortality did not differ across all ICU patients infused with albumin or saline in the SAFE study, a subset analysis revealed that the relative risk of 28-day mortality for the 1,218-patient subgroup of severe sepsis patients was 0.87 favoring albumin (87 albumin group deaths for every 100 saline group deaths). Per 100 patients treated, this equates to 4.6 fewer deaths for subjects enrolled with severe sepsis who received albumin. This strong indication approached, but did not attain, statistical significance (p = 0.09). FFF is actively encouraging an extension of the trial in severe sepsis patients to establish whether this albumin survival advantage is statistically significant.
- When SAFE study investigators excluded enrolled trauma patients from their analysis, the use of albumin was associated with a 4% lower risk of mortality than the use of saline (RR = 0.96; p = 0.04). In 492 brain-injured trauma patients, who constituted just 7% of the total study population, there were 21 excess deaths among those given albumin. (The death rate for non-brain-injured subjects was exactly 6.2% in both treatment arms.) As recently pointed out, "this mortality difference could easily have occurred by chance."1 The SAFE investigators have suggested that a new trial may be organized to resolve this question.
- The Cochrane reviewers evaluated three small and very poorly designed investigational burn studies, two of which specified vastly excessive albumin dosing.2,3 Possibly attributable to fluid volume overload, mortality in the albumin groups in both of those studies exceeded mortality in the crystalloid groups. Neither of these highly experimental protocols represent the standard of care in resuscitative therapy for severe burn patients, and they should have been excluded in the Cochrane analysis.
Because FDA's advisory committee was not briefed on this fundamental flaw, and the SAFE study specifically excluded burn patients, the committee reluctantly concluded that "the relative safety of albumin for use in patients with burns cannot be determined at this time." We believe that burn specialists are in the best position to interpret the available medical literature, and consider their own experience, in determining the best resuscitative strategy for their patients.
FFF Enterprises strongly endorses recommendations by FDA, the SAFE investigators and others to complete clinical evaluations of albumin resuscitation in severe sepsis and other defined critical care populations where the SAFE trial presented evidence of important outcomes.
