Blog with interesting cases and/or problems related to anesthesia with discussion based on best evidence in the literature.

September 3, 2012

Organ Procurement in the Brain dead Patient

I recently was called to return to the hospital at 1 AM to care for a patient who had been admitted at 10 am the previous morning from the ED after a drug overdose.  It wasn't clear what drug was used, but the patient experienced a cardiac arrest in the ED. After successful resuscitation she was found to be brain dead.  Studies to verify her as a candidate for organ donation were completed and a decision was made to harvest the kidneys only as she had evidence of aspiration pneumonia, cardiac dysfunction post arrest, and liver enzymes were significantly elevated indicating ischemic liver damage.  She was a only 33 years old.

General OR anesthesiologists are not often called upon to manage patients in the ICU who are brain dead and, therefore, the pathophysiology may not be fresh in the mind.  This case serves as review of the issues that the anesthesiologist will need to consider.

In most cases, the patient will be well prepared as they will most likely have all of the proper monitors and be ventilated in the ICU.  Nevertheless, it is important that the anesthesiologist approach this case as they would a critically ill patient who will need intensive and tight management of the cardiovascular and pulmonary systems.

Because these patients have suffered an insult resulting in brain death, swelling is likely which leads to an initial  surge  of catecholamines.  The result can be severe hypertension and bradycardia. This initial autonomic storm  is followed by a state of shock.  The causes include lack of sympathetic tone (spinal shock), cardiogenic shock from catecholamine toxicity, vasodilatory shock from inflammatory mediators, and finally, vasopressin depletion along with diabetes inspidus can result in hypovolemic shock.   Arrythmias, conduction abnormalities and ischemic changes may all be present on the EKG secondary to catecholamine toxicity. The brain swelling causing this surge may also lead to neurogenic pulmonary edema.  Therefore, both the heart and lungs are often damaged early on after brain death or in the process leading up to brain death.  Pulmonary edema may also result from left ventricular dysfunction leading to increased pulmonary vascular congestion, but increased hydrostatic pressure in the arterioles and increased vascular permeability from inflammatory mediators are also responsible.
Optimization of the cardiopulmonary system requires careful attention in many cases.  Many authors recommend avoiding elevated FiO2 (less than than 50%) if possible especially if the lungs are to be donated.  If the lungs are being donated then, the FiO2 should be no higher than to maintain the PaO2 greater than 75 mmHg.  Furthermore, given the propensity to increased lung water in these patients, if donation of the lungs is to occur, fluid administration must be carefully titrated to hemodynamics and cardiac output.  The use of a pulmonary artery catheter may be indicated, or perhaps, goal directed therapy with the  vigileo monitor (Edwards Life Sciences) is indicated.  Goal directed therapy in these patients (Lung donors), is ideal and the Flo Trac system can provide information on stroke volume variation in mechanically ventilated patients indicating the liklihood that they will increase stroke volume with a fluid challenge.  Volutrauma (barotrauma) must be avoided in the lung donor.  Thus, recommendations are that tidal volumes be limited to 6 to 8 mL/Kg, and PEEP should be kept to no more than 10 cmH20.  Lungs to be donated must be free of infection, and able to maintain PaO2 of 375 mmHg with FiO2 of 1.0 with PEEP of 5 cmH20.

Brain death also results in a myriad of endocrine pathologies that might be expected when the hypothalamic pituitary axis is destroyed. In my patient an infusion of levothyroxine was infusing, along with levophed and insulin.  Thyroid hormone comes from the thyroid gland where it is secreted predominantly as T4 (80%).  However, T4 is not as biologically active as T3.    In addition,  peripheral de iodination of T4 to the biologically active form of T3 is diminished in this population perhaps due to elevated cytokine levels.  Lastly, peripheral conversion of T4 to a hormone referred to as reverse T3 occurs.  Reverse T3 is biologically inactive and gives rise to the so called sick euthyroid syndrome.  Therefore, attempts to replace thyroid hormone was felt to provide benefit in this population.  Unfortunately, IV formulations of T3 are not available in the US or are significantly more expensive than IV T4. Consequently, the administration of levothyroxine in brain dead patients is somewhat controversial. But, IV levothyroxine (T4) is part of a protocol set up by the United Network for Organ Sharing (UNOS).  While some published works have not found any improvement in hemodynamic stability with T4 infusions, others have shown benefit.  Salim et al.[1], studied 19 patients declared brain dead who demonstrated significant HD instability requiring high doses of vasopressors after optimizing fluid requirements.  In these patients, when 20 mcg of IV T4 was given as an initial bolus, then followed by an infusion of 20 mcg/hr, vasopressor requirements were decreased and in 53% of patients vasopressors were weaned off completely.  Unfortunately, this was not an RCT, (patients served as their own controls).  Furthermore, addition of IV T4 was part of a protocol including D50 with 20 units of insulin and methylprednisolone.  The insulin infusion was titrated to maintain blood glucose between 120 and 180 mg/dL. Several years later Salim et al. [2]published another study showing that IV levothyroxine improved the rate of organs that could be donated. Overall, organs per donor were 3.9 when T4 was used vs. 3.2 (P<.01) when it was not used.  However, the benefit was only statistically significant when the organ donated was the pancreas. T4 failed to reach statistical significance when the organ donated was the heart, lung, liver or kidney (i.e. P value greater than 0.05).  Rosendale et al. did a large retrospective analysis of over 10,000 transplants.  Those who received hormone replacement (HR) comprised of insulin, glucocorticoid, and thyroid hormone had improved organ donation after brain death that was statistically significant.  Unfortunately, the absolute improvement was small (the biggest improvement was for kidney donation with an absolute increase of 7%). It is not well understood why thyroid hormone might help in the HD unstable patient. It is believed that thryoid hormone can improve myocardial contractility via upregulation of adrenergic receptors as well as increased intracellular Ca2+ availability. Furthermore, it may cause cellular respiration to switch from anaerobic to aerobic thus improving cellular efficiency mitigating metabolic acidosis. 
In the end, studies to date that have shown an improvement in donation success utilized a combination of medications that included insulin as an infusion, thyroid hormone IV, and glucocorticoids replacement.
The inclusion of glucocorticoids in these studies has a physiologic basis.  ACTH is produced in the PG when stimulated by CRH (corticotrophin releasing hormone) from the hypothalamus. In the brain dead patient, the hypothalamus is often non functional.  Furthermore, studies have found that a majority of brain dead patients are not capable of responding to ACTH (cosyntropin) when provided [4], and in these patients, supplementation with solucortef resulted in HD stabilization. The following protocol is recommended by UNOS: 
  1. D50 1 amp with 20 units regular insulin IV followed by infusion
  2. Solumedrol 2 GM
  3. Levothyroxine 20 mcg then 20 mcg/hr titrating as needed up to 40 mcg/hr for low BP.

Studies have also shown that in the brain dead patient, vasopressin levels decrease. Therefore, in patients with significant hypotension, some authors recommend vasopressin as a first line agent [10].  Brain death results in diabetes insipidus due to a lack of ADH production, creating a second reason for choosing vasopressin as pressor of first choice.  Vasopressin can be given as a bolus, and then as an infusion of 0.5u/hr to 4 u/hr.  However, some protocols avoid going higher than 2.5 u/hr due to the risk of organ ischemia, particularly mesenteric ischemia.
As mentioned, diabetes insipidus is common in these patients (up to 65%). It is the second most common problem found in the brain dead patient after hypotension (81%). However, polyuria can result from other causes as well, including osmotic diuresis from mannitol and/or hyperglycemia, and diuretic pharmacotherapy (i.e. lasix).  As a result, careful monitoring of urine output in the OR is important, with most clinicians attemtping to avoid urine outputs that exceed 100 ml/hr.  Due to polyuria, electrolyte abnormalities are common.  Serum Na+ greater than 155 mmol/L is associated with poor liver graft function.  Therefore, this should be monitored carefully and water should be provided as necessary to treat hypernatremia.  K+, Mg2+, Ca2+, and phosphate are often depleted and may need to be replaced. Vasopressin is helpful in mitigating diabetes insipidus, but DDAVP may be required (dose is 0.3 mcg/kg IV). 
While serum hormones levels such as ADH, ACTH, TSH, etc drop due to central neurologic failure, hyperglycemia is not a result of failure of production of insulin; rather peripheral insulin resistance. 
Brain death can also lead to coagulopathy.  This is a result of necrotic brain tissue releasing thromboplastins which lead to increased fibrinolysis. Furthermore, DIC occurs in up to 28% of these patients.  Since thermoregulation is under the control of the hypothalamus, these patients are at risk for hypothermia.  As their body temperature drops, coagulation is further impaired. Therefore, the clinician should carefully follow coagulation parameters looking for evidence of DIC.
While recent studies have advocated for allowing for lower and lower Hgb levels prior to triggering transfusions in the regular OR arena, in the case of the brain dead patient, transfusions to maintain a Hgb of 10 g/dL is recommended.  The common reasons to avoid transfusions in patients who are to survive surgery don't apply in this case; maximizing oxygen delivery is the only goal. 

I opt to use inhalational anesthetics as tolerated for these procedures.  Of course, a patient that is brain dead will not experience pain.  However, other considerations make the use of volatile anesthetics an important part of the regimen.  Surgical stimulus to the brain dead patient will illicit a strong spinal response.  This may cause movement and/or a surge of catecholamines.  This catecholamine surge potentially results in damage to organs to be donated through direct toxicity (heart) or via severe hypertension.  Volatile anesthetics are potent inhibitors of spinal reflexes and adrenergic response.  In fact, Antognini and Berg showed that the brain has little to do with the inhibition of movement to surgical stimulus [5]. In their study,  MAC-BAR was determined for goats in the conventinal fashion (whole body group) and was compared to two other groups.  Study group A had the circulation to the cranium isolated by CPB and MAC-BAR was determine using isoflorane for the brain, and in group B, the torso was isolated with CPB and MAC-BAR for the spinal cord was determined.  The authors found that whole body MAC-BAR was 3.7%.  Interestingly, they had trouble determining MAC-BAR in the brain.  In several animals they could not achieve MAC-BAR in the brain even when using an enflurane vaporizor that allowed over delivery of isoflorane.  Nevertheless, they published a value of MAC-BAR of 5.6% for the brain and 2.2% was found for the spinal cord.  The take home message is that when anesthetizing patients for surgery, the importance of the volatile anesthetics is related to their blood levels in the spinal cord more than the brain in when attempting to dampen the adrenergic response to surgical stimuli. 
Another potential benefit with volatile anesthetics is their proven ability to reduce injury after an ischemic insult.  This has been demonstated in the kidneys, brain, liver, and heart. The rational is that since organs being procured for donation must suffer a degree of ischemia during the process and obviously the subsequent reperfusion, their function is likely to suffer in relation to the degree of ischemia suffered.  Ischemia-Reperfusion injury results from a cascasde of events related the production of inflammatory cytokines, aggregation of neutrophils at the site of injury, and eventually cell death if the insult is too large.
Another important cytokine that worsens the inflammatory soup generated from ischemia-reperfusion is tumor necrosis factor alpha (TNA-alpha). Fuentes et al. [7] were able to show that isoflorane could attenuate TNA-alpha levels in mice who were injected with LPS.  This was accomplished by inhibition of NF-kB (a proinflammatory transcription factor) induction of DNA transcription.  In addition, these authors were able to show that IL-10, an anti-inflammatory cytokine, was elevated in mice given isoflorane.  Others have shown that IL-10 decreases TNF-alpha levels and increases survival in several modesl of sepsis. Weber et al. [8]  were able to demonstrate the ability of isoflorane to inhibit the ability of TNF-alpha to cause increased NF-kB activity reducing the production of CAM (cell adhesion molecules).  Cell adhesion molecules are important in the inflammatory process resulting in increased "stickiness" of the endothelium allowing neutrophils and other pro inflammatory cells to be recruited to sites of injury such as can occur with ischemia-reperfusion. All of this leads to up regulation of the antigen-expression by cells and could be associated with an increased incidence of early acute graft rejection [9]. Several studies have demonstrated that organ grafts from brain dead donors are inferior to living donors over both the short term and long term. Of course the source of this difference is multifactorial, initial management of the organs in the brain dead donor is likely a crucial factor.
In summary, anesthesiologists at community hospitals are increasingly being requested to provide intraoperative mangement of brain dead patients.  While these patients do not require anesthesia in the conventional sense, careful management of these patients to ensure organ protection as well as maximize the potential for good graft function in the recipient are goals under our control.  Understanding the pathophysiology of brain death on donor organs aids in choosing intraoperative interventions that will optimize graft function.  Briefly, anesthesiologists must recognize potential for hypotension and choose appropriate therapy to maintain HD stability, understanding potential for lung damage and optimize lung function particularly if these are to be donated.  Furthermore, the threat for hypothermia and coagulopathy/DIC must be recognized and managed.  Lastly, the varied and complex endocrinopathies associated with the destruction of the Pit/Hypothalamic axis must be recognized and managed.  

1. Salim A et al. , Arch Surg. 2001; 136:137
2. Salim A et al. Clin Transpl. 2007; 21:405
3. Rosendale JD et al. Transplantation. 2003; 75:482
4. Nicolas-Robin A et al. Anesthesiology 2010;112:1204
5. Antognini JF, Berg K. Anesth Analg. 1995; 81(4):843.
6. Hashiguchi H et al. Anesth Analg. 2005; 100(6):1584
7. Fuentes JM et al.  Clin and Vac Imm. 2006; 13: 281
8. Weber NC et al.  Anesthesiology. 2008; 108: 199
9. Pratschke J et al.  Transplantation. 1999; 67:343
10. Hunt SA, et al.   Crit Care Med. 1996; 24(9): 1599

No comments: