Aug 312016

1269093_sJon-Emile S. Kenny [@heart_lung]

"We do not live in a privileged frame of reference."

-Carl Sagan

It is a common human experience to feel that one is the exception.  Indeed, the study of cosmology has demoted our - once privileged - place in the cosmos.  We have fallen from the center of the universe to inhabiting but one planet circling the sun, itself an average star in the Milky Way, comprising billions of others suns.  The vast Milky Way but one of billions of galaxies accelerating apart from one another ... but I digress.

Yet, in the ICU each consultant may hold his or her system in such regard - within a privileged frame of reference.  I have found this among some neuro-intensivists who seem to feel that - despite decades of negative trials - optimizing the exceptionally crude physiological variable of oxygen delivery is warranted, but only for the brain.

Certainly there is a simplistic rationale for this practice - the same explanation invoked 30 years ago when intensivists subscribed to supra-physiological oxygen delivery goals.  Hemoglobin determines oxygen content, oxygen content and cardiac output determine oxygen delivery [DO2], ischemic tissue needs oxygen, therefore giving hemoglobin (pRBCs) will mitigate tissue ischemia.


Yet the provision of pRBCs is not benign - transfused cells plug capillaries because they do not deform normally and have have abnormal oxygen-unloading characteristics.  They impair the immune system and increase the risk of circulatory overload.

So what underlies the practice of aggressively avoiding anemia in brain injury?  In a review, which should be read by all, Lelubre et al. summarize the evidence for and against transfusion practice in a large variety of brain-injured patients.

They list the innumerable retrospective studies which associate anemia with poor neurological outcome.  These analyses are consistent across a large swath of brain injured populations [e.g. traumatic brain injury, sub-arachnoid hemorrhage, ischemic stroke and intra-cerebral hemorrhage].  Yet they also highlight a comparably robust body of literature which associates the provision of red blood cell transfusion with no improvement in outcome or even worsened outcome in patients with brain injury.

The authors do highlight two recent prospective randomized trials.  One included 44 patients with SAH and targets systemic hemoglobin levels of 10 g/dL or 11.5 g/dL.  There were no statistically significant differences between the two groups while there was a trend towards fewer MRI-detected infarctions in the higher-Hb group (6/20 versus 9/22).

In the second prospective RCT, 200 TBI patients were studied with transfusion triggers of 7 mg/dL versus 10 mg/dL.  The two groups had relatively high Hb levels during the study [roughly 10 versus 11 mg/dL].  A dichotomized Glascow Coma Scale outcome showed no difference at 6 months and there were more thromboembolic events [22/101 versus 8/99] in the higher Hb target group.

In Practice

The authors leave us with a general approach to treating anemia in the brain-injured patient.  They rightly note that there is no strong evidence compelling the intensivist to transfuse a brain-injured patient with a hemoglobin < 7.0 g/dL - especially in the awake and conscious patient.

In those with a poor neurological examination, the authors invoke both systemic [e.g. central venous oxygen saturation and lactate levels] and cerebral [e.g. PbtO2 and SvjO2] markers of dysoxia as transfusion triggers.  Surprisingly, to support their argument for the cental venous oxygen saturation, they invoke the Rivers Trial of early, goal-directed therapy in severely-septic patients but make no mention of the TRISS Trial which convincingly erased the use of blood transfusion to augment oxygen delivery in the management of severe sepsis and septic shock.

The authors suggest "considering" blood transfusion in the neurologically-injured patient with the following dysoxia parameters: ScVO2 < 60% or an elevated lactate, PbtO2 < 15 mmHg or an SvJO2 < 55%.  Yet they acknowledge the data that blood transfusion tends to impact many of these parameters insignificantly.

What might be an interesting applied-physiology study?  Determine if the brain has reached its anaerobic threshold and then consider red blood cell transfusion!  How might that be done?  Consider the physiology described in this previous post; determine the carbon dioxide production [VCO2] and oxygen consumption [VO2] of the brain and the ratio of the two [i.e. the brain VCO2/VO2 ratio].  If the ratio is greater than 1.0, it is reasonable to postulate that the brain is below its critical oxygen delivery [cDO2].  At this point, it may be reasonable to augment DO2 with the provision of packed red blood cells.  Such analysis, however, would require venous blood gas analysis from the jugular vein.  It would also assume that an arterial blood gas obtained from the radial or femoral artery would reflect the oxygen content of the carotid artery.

Clearly what is needed is a large multi-center, randomized and controlled trial to give - at least some - real data upon which we can base our practice.  Until then the authors correctly emphasize the slim data set upon which transfusion in the brain-injured is based.

In summary - in the ICU - we must resist the Orwellian adage:

"All organs are equal, but some are more equal than others."


Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


Anemia & the Injured Brain