Starches Last Stand; Professor Chappell’s Keynote Lecture at Notfall Medizin 2018

Things are getting personal in the rearguard action to save the European Medicines Agency Marketing Authorisations for hydroxyethyl starch solutions. Starlingists will be familiar with my own view that ‘therapies’ based on the biophysical oncotic pressure of intravenous fluids are unlikely to make more than a short term difference to the distribution of fluid within the body compartments. Professor Chappell and his colleagues in Bavaria have long been committed to the colloid cause, especially starches, and have contributed to the available literature. They declare that their lectures are sponsored by Baxter Deutschland GmbH, Fresenius Kabi Deutschland GmbH, B Braun Melsungen AG and Serumwerk Bernburg AG. CSL Behring have provided Educational Grants to support research. The Twittersphere has shared Prof Chappell’s Keynote Lecture and here I offer my own thoughts on his defence of starches.

Prof Chappell’s lecture begins at about 44 minutes into the YouTube Video. We soon see a curious slide showing sodium and water passing freely between capillary lumen and interstitium; The Professor explains that water and electrolytes “are not held back at the vascular barrier, they can freely flow…” and as normal plasma volume is about 20% of the extracellular fluid volume, the volume effect of crystalloids “according to physiology” is 20%’. He backs up this claim by citing his own experiment published in 2012 helpfully titled The intravascular volume effect of Ringer’s lactate is below 20%: Let us look at the experiment. Subjects were ten anaesthetised women undergoing pre-hysterectomy normovolaemic haemodilution. After post-GA line insertions and stabilisation, around one litre of blood was withdrawn over about 40 minutes while three times the volume of isotonic salt solution (ISS) was infused. Why 3:1? It was a “compromise between the physiological expectations suggesting the five-fold amount and recent suggestions from literature 1.4-fold”. They report that heart rate, systolic and mean arterial blood pressures were unchanged by exchange transfusion. Central venous pressure was not reported. A slight drop in diastolic arterial pressure could be explained by reduced blood viscosity. So the volume equivalence of infused ISS to exsanguinated volume for circulatory stability is confirmed to be less than 3:1. Why then were these haemodynamically-stable and non-shocked patients next infused with 20% human albumin solution? Because the volume of distribution of indocyanine green (presumed to represent plasma volume) was reduced. After replacing removed albumin, the researchers calculated that the theoretical volume equivalence of infused ISS  to exsanguinated volume to maintain the ICG volume was greater than 5:1.

At about 50 minutes the Professor raises our eyebrows when he tells us that iso-oncotic colloids cannot pass an intact vascular barrier and so their intravascular volume effect is 100%. Applies to starch and albumin; “if it’s iso-oncotic it stays in your vessels”. To back this up he cites an earlier Bavarian experiment with ten patients undergoing normovolaemic haemodilution with 1.15:1 6% hydroxyethyl starch.  Indocyanine green and fluoroscein-labelled red blood cells are used as dilution indicators. (1) They report the phenomenon of colloid-induced anaemia in the following way;  “The hematocrit decreased disproportionally in relation to the residual intravascular volume. Consequently, estimating the volume effect from the changes in hematocrit led to an overestimation (about +30%).” (2)

“You should only give [colloids] if you have an isolated loss from your intravascular compartment” or “to restore cardiac preload in the case of acute shock”. Professor Chappell offers no thoughts on the use of blood transfusion to treat major haemorrhage. I am old enough to recall when modified fluid gelatin was said to be the colloid of choice in acute haemorrhage because its shorter plasma half time buys time to start a cross matched blood transfusion as the colloid is cleared.

Professor Chappell’s third experimental evidence is that starch resuscitation is quicker, and of smaller volume, than crystalloid resuscitation in a haemorrhagic hypotension porcine model. Worth noting that none of the animals became shocked and all survived, though Professor Chappell prefers to emphasise the time to return to pre-bleed mean arterial pressure. (3)

Professor Chappell then makes an impassioned plea for colloid boluses to increase stroke volume during peri-operative goal-directed therapy. The unacknowledged elephant in this room is that any colloid-induced increase in stroke volume is more than offset by colloid-induced anaemia so that there is no net increase in oxygen delivery.

At 1h 14mins Professor Chappell begins his tirade against the European Medicines Agency, and his dismay that even his own German agency favours suspending marketing authorisation for starches is evident. He then argues the best of a very weak case. By his own reasoning the appropriate dose of an iso-osmotic colloid solution is small, shall we say 1 litre? The alternative is said to be a larger volume of ISS, shall we say 3 litres? So do we really need the colloid?? I leave you to answer the question. I say RIP biophysical oncotic pressure therapy.

Readers wanting an Open Access Educational resource giving a credible account of the context-dependent half-times of infused crystalloids are strongly recommended to read Robert Hahn’s work, in spite of the reservation I note at ref 2! Let Hahn have the last word – “The question of how much colloids contribute to interstitial oedema remains unknown for lack of clinical data.”(5)

  1. Jacob, M; Rehm, M; Orth, V; Lötsch, M; Brechtelsbauer, H; Weninger, E; Finsterer, U. Exact measurement of the volume effect of 6% hydoxyethyl starch 130/0.4 (Voluven) during acute preoperative normovolemic hemodilution. 2003 .Anaesthesist 52(10) 896-904.
  2. (Just to add to our confusion, the substantial body of work published by Professor Robert Hahn over many years uses haematocrit changes to reflect plasma volume changes)
  3. Roger, C, Muller, L, Deras, P, Louart, G, Nouvellon, E, Molinari, N, Goret, L, Gris, JC, Ripart, J, de La Coussaye, JE, Lefrant, JY. Does the type of fluid affect rapidity of shock reversal in an anaesthetized-piglet model of near-fatal controlled haemorrhage? A randomized study. Br J Anaesth 2014 Jun. 112:1015-23. DOI 10.1093/bja/aet375

Leave a Reply

Your email address will not be published. Required fields are marked *