Safety and efficacy of 6% hydroxyethyl starch in patients undergoing major surgery The randomised controlled PHOENICS trial

Abstract

BACKGROUND Hydroxyethyl starch (HES) is often used for maintaining vascular volume during major surgery. Long-term high-dose HES in septic patients promotes renal injury, whereas meta-analyses of current HES products in surgical patients do not show such effects. OBJECTIVE We studied if the peri-operative use of HES is noninferior to crystalloids in terms of acute kidney injury. Secondary outcome was the noninferiority of HES on worsening of renal injury and/or the incidence of a composite endpoint of major complications and mortality until postoperative day 90. DESIGN Randomised double-blind trial in patients undergoing elective abdominal surgery with expected blood loss at least 500 ml. SETTING Multicentre trial at 53 study sites in 10 European countries. PATIENTS One thousand nine hundred and eighty-five (HES 977, crystalloid-only 981) patients aged 40 to 85 years with ASA status II-III. INTERVENTION Either 6% HES 130/0.4 or a crystalloid solution. Dosing was guided by mean arterial pressure and/or routine haemodynamic variables. MAIN OUTCOME MEASURE Change from pre-operative to lowest cystatin C-based eGFR during the first 3 postoperative days. Key secondary outcome was a composite endpoint of mortality and major postoperative complications after 90 days. RESULTS Mean change in eGFR from baseline to minimum was -3.4 +/- 17.7 ml min-1 1.73 m-2 in HES patients and -1.0 +/- 17.1 ml min-11.73 m-2 in crystalloid-only patients (P < 0.001 for noninferiority). The key secondary endpoint occurred in 35% of patients in each group. There were no clinically relevant differences in any safety endpoint including 90-day renal function. Any cause mortality-difference until the end of 1-year follow-up was not significantly different (8.6% in HES and 10.1% in crystalloid patients). CONCLUSION Peri-operative use of HES was noninferior to crystalloids in short-term renal function or a composite of mortality and major complications at 90 days. PHOENICS provides robust evidence that peri-operative in-label use of HES is well tolerated.