The document traces the evidence around the use of hydroxyethyl starch (HES) in critical care from early reviews showing potential benefits to large trials published in 2012-2013 that associated HES with increased mortality and renal failure risk in sepsis patients, leading major clinical guidelines to recommend avoiding HES in high-risk patients like those with severe sepsis. While early evidence suggested HES may be useful for volume resuscitation, later trials involving thousands more patients established the risks of HES outweigh any potential benefits in critically ill populations. Questions remain around the safety and efficacy of low-dose HES for low-acuity non-sepsis

1. CHEST and Beyond
Hydroxyethyl starches in critical care
David Gattas
Feb 2013

2. Hydroxyethyl Starch (HES) is..
• A synthetic colloid
• Modified natural polysaccharide
– Amylopectin derived from waxy maize or potato
– hydroxyethyl substitution slows hydrolysis by amylase
– molecular weight and molar substitution influence
elimination
• in a saline or plasma-adapted carrier solution

3. HES in critical care
This topic has it all
• Science
– transformation of the evidence
– scale of trial design in critical care
• History
– shock and resuscitation
– evidence-based medicine
• Drama and Politics
– scientific misconduct
– clinical guideline development

4. HES in critical care
pre-2010
• Clinical paradigm
– “modern rapidly degradable HES”
• Actual usage
– highly variable
– based on geography, availability, local habits
• Best evidence
– Cochrane reviews
– other reviews

5. 2009: “Modern Rapidly Degradable HES”
Boldt Anesth Analg 2009;108:1574-82.

6. SAFE TRIPS Finfer Crit Care 2010;14:R185
• Every day in ICU, about 1 in 3 patients
receive an episode of fluid resuscitation
• About half of this is colloid
• About half of this colloid is HES
2007: Actual use of HES

7. SAFE TRIPS Finfer Crit Care 2010;14:R185
6%HES130 is the most frequently used
colloid in ICU globally
2007: Actual use of HES

8. 2007: Colloids
2010: Renal Safety of HES
Perel Cochrane Database Syst Rev 2007:CD000567.
Dart Cochrane Database Syst Rev 2010:CD007594.
“There is no evidence from RCTs that
resuscitation with colloids reduces the risk of
death, compared to resuscitation with
crystalloids, in patients with trauma, burns or
following surgery”
“Large studies with adequate follow-up are required to
evaluate the renal safety of HES products … There is
inadequate clinical data to address the claim that safety
differences exist between different HES products.

9. Schortgen 2001
• N=129
• Sepsis
• 6%HES
200/0.6
Schortgen Lancet 2001;357:911-6

10. Brunkhorst 2008 (VISEP)
• N=537
• Sepsis
• 10%HES
200/0.5
• Stopped
early
Brunkhorst N Engl J Med 2008;358:125-39

11. Published HES Reviews 1960-2010
• 223 HES reviews
– 165 made a recommendation
• 124 favourable to HES, 41 unfavourable
• Associated with favourable recommendation
– reviews with lower methodological quality
– reviews with no meta-analysis
– reviews with potential COI among authors
Hartog Intensive Care Medicine 2012;38:1258-71

12. Published HES Reviews 1960-2010
Hartog Intensive Care Medicine 2012;38:1258-71

13. Starch Use in Australia
• 6% HES 130/0.4 is the first starch approved by
the TGA late 2006.
• Marketed from 2008
• December 2008
– over >40 hospitals in Australia used HES
– > 200,000 units distributed
• > 30% of synthetic colloid market

14. Myburgh N Engl J Med. 2013 Feb 21;368(8):775. doi: 10.1056/NEJMc1215977

15. Anesth Analg 2009; 109(6): 1752-62

16. “Shortly after…”
Shafer Anesth Analg 2011; 112(3): 498-500.

17. Disclosure: no IRB approval
1st retraction
28 Oct 2010 http://www.aaeditor.org/NoticeofRetraction.pdf

18. Editors-in-Chief Statement #1
Acta Anaesth Scand Anaesthesia
Anästh und Intensivmedizin Anästh Intensziv Notfall Schmerztherapie
Anesth Analg Anesthesiology
BJA Can J Anesthesia
Der Anästh Eur J Anaesth
Intensive Care Medicine
• Regarding IRB approval:
– If the IRB investigation demonstrates no approval
 we will retract
• Regarding scientific fraud:
– If the separate hospital investigation demonstrates
any wrongdoing
 we will retract
4 Feb 2011 http://www.aaeditor.org/EICJointStatement.pdf

19. Editors-in-Chief Statement #2
66 new retractions (total 88)
prev + Crit Care Med, J Cran-Max-Facial Surg, Med Sci Mon, Minerva Anestesiol
• No IRB approval for 88 reports
4 Mar 2011 http://www.aaeditor.org/EIC.Joint.Statement.on.Retractions.pdf
9 Aug 2012 press release http://www.klilu.de/
Hospital Inquiry
• False data in at least 10 of 91 studies

20. Gattas Anesth Analg 2012;114(1):159-69
• 6% HES 130/0.4 RCTs only
• Acutely ill adults
• Date of search 24 Dec 2010
• 36 studies including 11 retracted
• 2149 participants including 541 retracted

21. Mortality
Gattas Anesth Analg 2012;114(1):159-69

22. Aim
To evaluate the safety and efficacy of 6% hydroxyethyl
starch (130/0.4) in 0.9% sodium chloride solution as
compared to 0.9% sodium chloride alone for fluid
resuscitation in adult patients treated in the Intensive
Care Unit.

23. Design and oversight
• Investigator-initiated, multicenter, prospective,
blinded, parallel-group, randomised-controlled trial
• 32 adult medical-surgical ICUs in Australia and New
Zealand
• Endorsed by the ANZICS Clinical Trials Group
• N=7000
• Powered to detect ARR 3.5% from baseline
mortality of 26%
• Powered to detect ARR 1.5% from baseline renal
failure of 6%

24. PRIMARY: All-cause mortality at 90 days
Incidence of acute kidney injury (RIFLE)
Use of renal replacement therapy
New organ failures (resp, cardiovasc, coag, hepatic)
Duration of ventilation
Duration of renal replacement therapy
Cause-specific mortality within 90 days
ICU and hospital mortality rates
Service utilisation (ICU, hospital LOS, mechanical ventilation, RRT)
Quality of life and functional outcome assessments (6 months)
Cost-effectiveness analysis
Outcomes

25. Study was conducted at maximum daily dose of 50mL/kg/day
Study treatment

26. Final recruitment

31. Study Fluids
Mean [ SD] daily average
HES: 526 425 ml
vs
Control: 616 488 ml
(P<0.001)

32. Net Fluid Balance
Mean [ SD] daily
average
HES: 921 1069 ml
vs
Control: 982 1161 ml
(P=0.03)

33. CVP
Mean [ SD] daily average
HES: 11.3 4.8 ml
vs
Control: 10.4 4.4 ml
(P<0.001)

34. Primary outcome: death at day 90

35. Probability of death at day 90

36. Renal Outcomes

38. Serum creatinine

39. Urine output

40. Adverse reactions

41. HES is associated with:
• Effects which are
– visible at the
bedside
– superficially
„positive‟
– minor
• Effects which are
– invisible at the
bedside
– negative
– major

42. So what else is happening?

43. 6S: Perner 2012
• N=804
• Severe sepsis
• 6% HES 130/0.42 vs Ringer‟s acetate
Perner N Engl J Med 2012;367(2):124-34.

44. BaSES: Siegemund (unpublished)
• N=241, single centre
• Sepsis
• 6% HES130/0.4 vs saline
• Mortality RR 0.97 (0.65-1.45)
• RRT RR 1.83 (0.93-3.59)
Siegemund M. NCT00273728 BaSES,personal communication, 2012-09-12

45. CRYSTMAS: Guidet 2012
• N=196
• Severe sepsis
• Lower volume of 6% HES 130 to achieve
haemodynamic stability
• Mortality RR 1.20 (0.8-1.74)
• RRT RR 1.83 (0.93-3.59)
Guidet B Crit Care 2012;16(3):R94.

46. Evidence transformation 2011-2013
• BASELINE approx 2100 subjects
• ADDED
– At least 8400 new subjects (~ 10 new RCTs)
• SUBTRACTED
– 541 subjects (11 RCTs)

47. Gattas Intens Care Med 2013 DOI 10.1007/s00134-013-2840-0

48. Haase BMJ 2013;346:f839

49. Zarychanski JAMA 2013;309(7):678-88

50. Patients Randomised into 6% HES 130 Trials
Gattas Anesth Analg 2012;114(1):159-69
Gattas Intens Care Med 2013 DOI 10.1007/s00134-013-2840-0

51. Patients Randomised into 6% HES 130 Trials
Gattas Anesth Analg 2012;114(1):159-69
Gattas Intens Care Med 2013 DOI 10.1007/s00134-013-2840-0

52. Gattas
Myburgh
ICM
Haase
Perner
BMJ
Zarychanski
McIntyre
JAMA
Published online 14 February 15 February 20 February
Population Critically ill Sepsis Critically ill
Intervention 6%HES130/0.4 6%HES130/0.4 All HES
Included trials 35 9 38
Included subjects 10391 3456 10880

53. Mortality
Gattas Intens Care Med 2013 DOI 10.1007/s00134-013-2840-0

54. Mortality
Haase BMJ 2013;346:f839

55. Mortality
Zarychanski JAMA 2013;309(7):678-88
RR for death
1.07
(95% CI, 1.00-1.14)
AR for death
1.20%
(95% CI 0.26-2.66%).
I2, 0%

56. Need for RRT
Gattas Intens Care Med 2013 DOI 10.1007/s00134-013-2840-0

57. Need for RRT
Haase BMJ 2013;346:f839

58. Need for RRT
Zarychanski JAMA 2013;309(7):678-88

59. Further exploration of these data
• Haase Perner
– analysis of trials with low risk of bias
– analysis of trials according to period of follow up
• Zarychanski McIntyre
– analysis of trials with and without Boldt

60. High vs
low risk
of bias
Haase BMJ 2013;346:f839

61. Long-term vs short-term follow up
Haase BMJ 2013;346:f839 suppl.

62. Boldt
Zarychanski JAMA 2013;309(7):678-88
Excluding Boldt
RR for death
1.09
(95% CI, 1.02-1.17)
Boldt trials
RR for death
0.94
(95% CI 0.82-1.09).
Overall
RR for death
1.06
(1.00-1.13)

63. So now what?

64. “recommendations reflect the high value we place
on the suggestion of harm in the setting of
available alternatives”
Published pre-CHEST, pre-6S
Reinhart Intens Care Med 2012;38:368-83.

65. ESICM task force – early 2012
• 1B: avoid HES with MW>200 in severe sepsis
• 1C: avoid HES in patients at increased risk for
AKI
• 2C: HES 130/0.4 should only be used in clinical
trial context in these patient populations
Reinhart Intens Care Med 2012;38:368-83.

66. Includes CHEST and 6S
Dellinger Crit Care Med 2013;41:580-637

67. The End (?)
Dellinger Crit Care Med 2013;41:580-637

68. Discussion
• What is the safety and efficacy of
administering low volumes of HES to low
acuity (non-sepsis) patients?
– in ICU?
– in anaesthesia?
• How important is the volume sparing effect
of colloids/HES?