Incidence and Causes

In the US just over 53,000 people die of traumatic brain injuries (TBIs) each year. These occur primarily from firearm-related events, motor vehicle-related events, and fall-related events. Firearm-related events remain the most prevalent causes of TBI, with 75% of deaths from self-inflicted injuries and 40% of deaths from assaults coming from brain injury. Homicide-related deaths have increased over the past decade in people 20 to 24 years of age. Motorcyclists are also at risk for TBI. Deaths from brain injury in this population have actually doubled over the past decade despite adoption in some states of universal helmet laws. Although implementation of helmet laws can decrease motorcyclist mortality, the lack of a nationwide universal helmet law limits the effectiveness of this protective measure. However, in the overall population of people who would constitute donor candidates, the death rate has decreased significantly, attributable primarily to increased safety measures aimed at motor vehicle drivers.

Physiologic Response

Brain tissue ischemia, the root source of brain death, triggers a series of well-defined effects as the damage progresses from the cerebral cortex through the brainstem. These effects proceed in a consistent sequence. Ischemia of the cerebral cortex and upper brainstem (the midbrain) results in a predominance of parasympathetic activity. Clinically, this manifests as hypotension and bradycardia. Subsequently, brainstem ischemia at the level of the pons triggers elevation of norepinephrine and epinephrine to well above normal physiologic levels, while leaving some functional parasympathetic nuclei. This results in the “Cushing reflex,” characterized by significant hypertension due to an increase in systemic vascular resistance (SVR), and a concomitant bradycardia as the still functional parasympathetic reflex arc attempts to compensate. As the ischemia progresses through the pontine region to the medulla, the parasympathetic nuclei can no longer function, leaving unopposed sympathetic input throughout the body. This is the period referred to as the “autonomic storm.” Finally, complete brainstem ischemia results in a decrease in sympathetic output and complete cardiovascular collapse, similar to the vasodilatory shock that occurs after a high spinal cord injury. These changes have very specific effects throughout the body, which are best considered by organ system.

Inflammatory

There are two triggers of inflammatory response to brain death. The first is direct neural tissue damage, which results in inflammation of the central nervous system. The second is in response to ischemia-reperfusion injury that occurs during the period of supranormal SVR in response to pontine ischemia. Release of inflammatory cytokines has been demonstrated locally in response to brain injury. Both cytokine profiles and complement levels have been studied specifically in organ donation, with effect on delayed graft function ; however, future work is necessary to identify ways to blunt these responses and improve donor graft function ( Fig. 6.1 ).

The distribution and pathophysiologic correlation of the rostral–caudal progression of cerebral-spinal ischemia termed coning, which eventuates in herniation and brain death.

Courtesy Kenneth E. Wood, DO.

Diagnosis

Death by neurologic criteria, or brain death, is a clinical diagnosis based on the presence or absence of a set of responses to neurologic stimuli. In the half century since the Harvard Committee first reported on “irreversible coma” as a new criterion for death, numerous variations on the determination of brain death have been proposed. However, after the President’s Commission for the Study of Ethical Problems in Medicine equated cardiac death and brain death in 1981, the modern criteria for brain death determination were set. These were then expounded upon by the American Academy of Neurology in 1995, and more recently reviewed and revalidated by Wijdicks and colleagues in 2010. Ranging from electroencephalograms to neuropathologic examination, none of the ancillary tests have proven as consistently reliable as a clinical examination done by a physician experienced in performing these tests.

Several prerequisites have to be met before initiation of a brain death examination. First, there needs to be evidence of a catastrophic brain injury that is compatible with a possible diagnosis of brain death. This can be established either clinically (evidence of gross head trauma) or using basic neuroimaging (computed tomography [CT]). However, it is important to remember that CT findings are not themselves confirmatory for brain death and may be misleading. Complicating medical conditions that may interfere with clinical assessment have to be addressed and resolved ( Box 6.1 ). Note that severe facial and ocular trauma will interfere with many of the brain death tests, making it difficult to form a definitive clinical diagnosis if they are present. However, in addition to traumatic injuries, severe electrolyte disturbances (hyper- or hyponatremia, hyper- or hypoglycemia), severe acid–base disturbances (profound acidosis), and endocrine dysfunctions (profound cortisol depletion or hypothyroidism) must be identified and corrected. No drug intoxication or evidence of poisoning can be present. This requires performing a drug screen and waiting for clearance of any alcohol to below the legal limit for driving (0.08%). In addition, it is necessary to identify any medications given in the hospital that could result in central nervous system depression; these need to have cleared the patient’s system before proceeding with testing. An appropriate time has been suggested at five times a drug’s half-life, assuming normal hepatic and renal function. Reversal agents are appropriate where available (opiates and benzodiazepines). Finally, the patient must at least have a core body temperature >32°C before starting the brain death examination. Severe hypothermia, defined as core body temperature <32°C, affects papillary light response, with complete loss of brainstem reflexes at core temperatures <28°C. Although rewarming techniques are beyond the scope of this chapter, for most patients a warming blanket should be sufficient to obtain appropriate temperatures before starting the examination.

Once these prerequisites have been met, the brain death examination can proceed ( Fig. 6.2 ). Three major findings need to be identified and documented to confirm brain death. These are: (1) coma or unresponsiveness, (2) absence of brainstem reflexes, and (3) apnea.

• 1.
  

  Coma or unresponsiveness. This component requires that there is no motor response or eye movement to noxious stimuli, typically described as nail bed pressure or supraorbital pressure. Often, this can be the most difficult part of the examination for practitioners, as a wide range of spontaneous or reflex movements have been described in the literature. These include everything from isolated jerks of the upper extremities to cremasteric and abdominal muscle reflexes to respiratory-like movements. More extreme reflexes have elicited more fanciful descriptions, including the “Lazarus sign,” where a combination of shoulder, neck, and extremity movements makes patients appear to rise from the bed. Clinical expertise is necessary to differentiate between these movements, and central and cerebral motor responses to pain. Ultimately, these movements do not invalidate the diagnosis of brain death, but it is important for the clinician to be aware of these responses so as to counsel family members appropriately.

  
  
• 2.
  

  Absence of brainstem reflexes

  
  
  
  
  • a.
    

    Pupillary reflex (cranial nerve [CN] II and III)

    
    
    
    
    • i.
      

      Pupils are round or oval, typically of midrange size (4 mm), though some can be as dilated as 9 mm.

      
      
    • ii.
      

      Pupils show no response to light.

    
    
    
    
  • b.
    

    Ocular movement (CN III, VI, and VIII)

    
    
    
    
    • i.
      

      Oculocephalic reflex. Otherwise known as the “doll’s eye” sign, in brain death the pupils will not show any movement as the head is turned rapidly to one side or the other. Caution: This test is not to be utilized in patients who have a suspicion of spine instability or fracture.

      
      
    • ii.
      

      Vestibulo-ocular reflex. Also known as the “caloric reflex test,” this test requires confirmation of a clear external auditory canal and elevation of the head to 30 degrees before starting. Each external auditory canal is irrigated (separately, with an interval of at least 5 minutes) with approximately 50 mL of ice water. In the presence of brain death, no eye movement will be seen during the 1-minute observation period, regardless of the ear irrigated.

    
    
    
    
  • c.
    

    Facial sensation and facial motor response (CN V and VII)

    
    
    
    
    • i.
      

      Absence of a corneal reflex (eyelid movement/“blink reflex”—CN V ₁ and VII). Although many texts describe performing this test with a cotton swab, some centers have moved toward stimulation with a puff of air from an empty 10-cc syringe. This decreases the risk of corneal damage from direct contact, while still providing a sufficient stimulus to potentially evoke a response.

      
      
    • ii.
      

      Absence of a jaw reflex (masseter reflex—CN V ₃ ). For this test, the mandible is tapped at a downward angle just below the lips. A positive test would result in the upward movement of the mandible in response. Usually, this reflex is very slight.

      
      
    • iii.
      

      Absence of facial movement to noxious stimuli (CN V ₃ and VII). Deep pressure on the supraorbital ridge or mandibular condyles at the temporomandibular joint should not result in any facial muscle movement (grimacing).

    
    
    
    
  • d.
    

    Pharyngeal and tracheal reflexes (CN IX and X)

    
    
    
    
    • i.
      

      Pharyngeal reflex (“gag reflex”—CN IX and X). Posterior pharyngeal stimulation with a tongue blade or hard suction catheter should not elicit a response. Note that this is separate from the tracheal reflex.

      
      
    • ii.
      

      Tracheal reflex (“cough reflex”—CN X). Tracheal stimulation, usually achieved by suctioning the endotracheal tube, should not elicit a response over multiple passes.

    
    

  
  
  
  
• 3.
  

  Apnea. The absence of a drive to breathe is the final test in the clinical evaluation of brain death. Normally, an elevation in CO ₂ above a critical level (in the US defined as >60 mmHg) will stimulate the respiratory center in the brainstem (medulla), which then signals the respiratory muscles to breathe. The apnea test is designed to provoke this response in an effort to establish whether the medulla (the lowest anatomic segment of the brainstem) is alive.

  
  
  
  
  • a.
    

    Prerequisites

    
    
    
    
    • i.
      

      Normotension (systolic blood pressure ≥90 mmHg)

      
      
    • ii.
      

      Normothermia (core temperature >36°C)

      
      
    • iii.
      

      Euvolemia

      
      
    • iv.
      

      Eucapnea ( P a co ₂ 35–45 mmHg)

      
      
    • v.
      

      Absence of hypoxia

      
      
    • vi.
      

      No prior history of CO ₂ retention (no history of chronic obstructive pulmonary disease or obstructive sleep apnea)

    
    
    
    
  • b.
    

    Preparation

    
    
    
    
    • i.
      

      Preoxygenate the patient with 100% O ₂ before the test; target is a P a o ₂ >200 mmHg.

      
      
    • ii.
      

      Reduce the ventilation frequency to 10 to 12 breaths/min to achieve eucapnea.

      
      
    • iii.
      

      Measure arterial P o ₂ , P co ₂ , and pH after these preparatory steps before starting the test.

    
    
    
    
  • c.
    

    Testing

    
    
    
    
    • i.
      

      Disconnect the patient from the ventilator.

      
      
    • ii.
      

      Continue to deliver 100% F i o ₂ at the level of the carina through a suction catheter or straight nasal cannula placed through the endotracheal tube.

      
      
    • iii.
      

      Observe closely for respiratory movements (abdominal, chest, neck) that could produce adequate tidal volumes.

      
      
    • iv.
      

      Continue the test as long as the patient remains stable. If, at the completion of 8 to 10 minutes, the patient remains stable, another 1 to 2 minutes can be taken before drawing an arterial blood gas. If the patient becomes hypotensive (systolic blood pressure <90 mmHg), hypoxic (O ₂ sat <85%), or develops cardiac arrhythmias, at any time before a full 8 to 10 minutes, an arterial blood gas needs to be drawn immediately and the ventilator needs to be reconnected.

    
    
    
    
  • d.
    

    Result interpretation

    
    
    
    
    • i.
      

      If respiratory movements are observed, the test is negative and the patient is not brain-dead.

      
      
    • ii.
      

      If respiratory movements are not observed, and the P a co ₂ is >60 mmHg or >20 mmHg above baseline normal P a co ₂ , the test is positive and the patient is clinically brain-dead.

      
      
    • iii.
      

      If respiratory movements are not observed, but the test was halted early for hemodynamic instability and the P a co ₂ parameters were not met, the test is indeterminate and additional testing should be considered.

      
      
    • 
      

      This test cannot be performed on every patient; approximately 10% of patients will be hemodynamically unstable at the time testing could occur, or before the conclusion of testing, requiring a premature stop. In these circumstances, other tests may be utilized.

    
    

  
  

General approach to the diagnosis of brain death.

Additional testing is not required by the American Association of Neurology guidelines, as brain death is a clinical examination. However, in patients for whom a complete examination cannot be performed, ancillary testing can be useful. Certain hospital or state guidelines on the declaration of brain death may also require an additional test, and therefore the practitioner is encouraged to review hospital and state-specific requirements ( Box 6.2 ).

Diagnosis

Deceased circulatory death is categorized using the modified Maastricht classification. Although the possibility for acute retrieval from an uncontrolled DCD does exist, for the purposes of this chapter only controlled DCD will be discussed.

Controlled DCD takes place when a planned withdrawal of cardiorespiratory support has been determined to be the best course for a particular patient. Naturally, this necessitates ongoing discussions with the family about the patient’s wishes and plans of care. After families express the desire to proceed with withdrawal, a representative from the organ procurement organization can present the option of DCD. Determining who will be a good candidate, or in other words, which patients will die within an acceptable warm ischemia time, is difficult ( Box 6.3 ). Various studies have demonstrated that increased ventilatory and circulatory support correlate with death in under 60 minutes. Decision support tools have been suggested to help with this process, although the opinion of an intensivist has proven similarly effective in the most recent large trial. With 20% to 25% of patients not correctly identified, it is important to counsel families who request withdrawal of care that this is an option, while at the same time preparing for ongoing patient care through the dying process if they do not meet criteria.

Once DCD has been authorized, every attempt is made to keep the patient medically stable until treatment is withdrawn. A short period of time is necessary to allow for organ allocation, but this process should not be extended over several days out of respect for the patient and family. Once arrangements have been made, the patient is transported to the operating room, at which point the family is allowed to pay their respects. The medical team responsible for the patient in the intensive care setting then proceeds with withdrawal of care, and after a period of 5 minutes of continuous asystole (monitoring with an electrocardiograph and arterial line), the patient is declared dead. Stiegler and colleagues have shown recently in an animal model that there is no return of brainstem function after 5 minutes of asystole, even if cardiopulmonary resuscitation is started at 5 minutes. At this point, the transplant team can proceed with organ recovery ( Fig. 6.3 ).

General approach to the diagnosis of brain death.

OPO, organ procurement organization; DCD, donation after circulatory death; OR, operating room.

Clearly, this process has significant ethical implications and has raised questions around the world regarding the nature of death and organ donation. Until the ethicists decide otherwise, the most important role the medical team can play in a DCD is to enforce the “dead donor rule,” wherein patients can only become donors after they are dead, and recovery of organs cannot cause a donor’s death. There is some thought that premortem interventions, such as obtaining blood samples and maintaining life-sustaining therapy while organ allocation processes take place, are acceptable, as the overall goal is to respect the patient’s final wish for organ donation. However, procedures that can cause serious harm, such as systemic heparinization, or cause pain, such as femoral cannulation, are not permitted until the patient has been declared clinically dead. It is also important to note that the transplantation team can have absolutely no involvement in patient management until after death has been declared, to avoid a conflict of interest. Critical care physicians should be familiar with their hospital’s policy on DCD, as they are often the individuals declaring death in these situations.