Hydroxyethyl starches kill people with severe sepsis; use crystalloid instead (RCT)

Colloids are believed to be more effective expanders of plasma volume than crystalloids, making them a tempting therapy to use for patients in septic shock and other causes of hemodynamic instability.

Problem is, some colloid solutions seem to kill people. The VISEP Trial using hydroxyethyl starch 200 kDa/0.5 demonstrated that this high molecular weight HES likely led to renal failure requiring dialysis, with a suggestion of increased mortality starting about three weeks after the initial infusion. Manufacturers of colloids, and their physician proponents, then switched to lower molecular weight colloid preparations (like a 130 kDa/0.4 substitution ratio), believing the larger molecules were causing the kidney damage and possible deaths, but that smaller ones would be okay (and somehow superior to crystalloids).

However, as the just-published Scandinavian Starch for Severe Sepsis/Septic Shock randomized trial in the New England Journal of Medicine shows, these lighter-weight colloid preparations are also potentially lethal.

What They Did

Investigators randomized 798 patients with severe sepsis in Denmark, Norway, and Iceland, and Finland at 26 ICUs to receive either HES 130/0.4 or Ringer's  acetate throughout their illness. The composite primary outcome was death or need for dialysis 90 days after randomization.

What They Found

At 90 days, more patients receiving HES had died, compared to patients in the group receiving Ringers acetate (201 or 51% versus 172 or 43%, P = 0.03). Virtually all the events in the composite endpoint were deaths -- with an additional 1 patient in each group on dialysis at day 90 contributing to the composite endpoint -- a strikingly damning result against hetastarch. More patients receiving HES/hydroxyethyl starch required dialysis overall, as well.

What It Means

It was already clear that higher molecular weight preparations of HES colloid were harmful for patients with septic shock. It appears that lower molecular weight compounds are harmful, as well, causing renal dysfunction and death. What's the mechanism? Experiments in pigs show that hydroxyethyl starch can be nephrotoxic.

True believers in using colloids will have an even more difficult time justifying their use in the wake of this trial, especially after the Surviving Sepsis Guidelines committee has indicated it will strongly recommend the use of crystalloids as first-line therapy for fluid resuscitation in severe sepsis in its upcoming revision of the guidelines scheduled for this year (its 2008 guidelines did not advise physicians which fluid preparation to use), and strongly recommend against the use of hetastarches greater than 200 kDa -- giving that admonishment its own "strong" 1B recommendation. I don't know if they had the benefit of this data when forming these tentative revisions, or if they had, whether they would have also included 130 kDa hetastarch on their blacklist.

Albumin is a colloid that seems to be safe for fluid resuscitation in severe sepsis -- at least, that's what the SAFE trial in NEJM 2004 suggested. The Surviving Sepsis folks intend to give the addition of albumin to crystalloid the blessing of a weak 2B recommendation/suggestion in the forthcoming revision.

Clinical Takeaway: Crystalloids, not colloids, and certainly not hydroxyethyl starches, are recommended as first-line therapy for fluid resuscitation in severe sepsis and septic shock. The proper role of albumin, an apparently safer colloid, in the management of severe sepsis is still debated, but it's a reasonable adjunct therapy.

Somebody stick a fork in the hetastarch potato: looks like it's done.

Perner A et al. Hydroxyethyl Starch 130/0.4 versus Ringer's Acetate in Severe Sepsis. NEJM June 27, 2012 ePub ahead of print.

Surviving Sepsis Guidelines to be updated: Preview from SCCM Meeting