1 A History of Coma: Evolution of Ideas

We need to go back to the last century, a productive period that led to notable advances in our understanding of coma. If we go back earlier to the classic monographs in medicine, little nuance is found, but coma had been recognized as a signa mali ominis. One of the first works describing coma in some detail using clinical signs was by Wepfer in Historiae apoplecticorum (1658 edition). His patient with a cerebral hemorrhage was “deprived of all sensation” and developed “laborious” irregular breathing with “body shaken albeit by a movement.” Thomas Willis, in his work De Anima Brutorum (1672), reported coma, carus, and lethargy, and even mentioned coma vigil. The physician and taxonomist Boissier de Sauvages de La Croix, in his work Nosologia methodica Sistens Morborum Classes, Genera et Species (1763), defined coma as typhomania (awake but not aware), lethargus (easily awoken and responding to questions), cataphora (responding but falling asleep when unstimulated), carus (hard to awaken), and apoplexia (deep sleep and limbs flaccid).

In the 19th and early 20th centuries, disorders of consciousness were mentioned in medical texts, but attention was directed to their relation with organ failure.⁶⁷ Likewise, coma as a clinical condition was only dealt with sporadically in neurology textbooks. Open up Romberg’s Lehrbuch der nervenkrankheiten des menschen (1840 edition), and coma is mentioned only as a sign preceding death. Open up Hammond’s Treatise on Diseases of the Nervous System (1872 edition), and you will find coma referred in the “apoplectic form of cerebral congestion” and other catastrophes. Open up Bing’s Lehrbuch der Nervenkrankheiten (1913 edition), and you will generally find coma associated with epileptic fits, apoplectic stroke, and traumatic head injury.

Separate book chapters on coma appeared in the early 1950s (e.g., Russell DeJong, The Neurologic Examination, 1950), and they emphasized the value of noxious stimuli to look for responsiveness, brainstem reflexes, compensatory eye reflexes with head turning, and significance of decerebrate responses. In the early 1960s, several works appeared that focused on general patient care with recommendations on head positioning, ventilation and tracheostomy, fluid management, and temperature control.⁹⁴

As technology advanced, better laboratory tests became available, but groundbreaking changes came with novel neuroimaging. Rapidly establishing the cause of coma accelerated with the widespread availability of computed tomography (CT) scanning in the late 1970s. This assisted neurologists, but mostly neurosurgeons, who could now eliminate emergency cerebral angiograms and stop performing blind exploratory burr holes for acute subdural hematomas.³⁷ A decade later magnetic resonance imaging (MRI) allowed even better visualization of the damaged and displaced brain structures. More recently, functional MRI showed brain activation when certain tasks were asked of a patient with impaired consciousness. None of this was apparent during neurologic examination, and these findings provocatively implied that these patients were “disconnected” from the external environment.

Given the large body of work, it is important to identify distinguishable strands. Explaining the background to all of this requires not only a discussion of laboratory experiments, but also an exploration of clinical concepts and pathological observations.⁵³ This chapter will review how the clinical features of coma, its clinical course, and its outcome became known and how it became part of neurologic knowledge and practice.

As with so many other gradually unraveling medical syndromes, it will be difficult to trace the very first complete description of brain herniation in the medical literature. A possible time line is shown in Table 1-1.

Two milestones can be archived—the observation that coma could be due to compression of the brain, and the notion that it could lead to brainstem injury. One of the earliest clinicopathological correlations comes from Rostan from the Salpêtrière Hospital, who noted in the beginning of the 19th century that in patients with cerebral softening, coma was a sign of later deterioration. At autopsy, he noted compression of the brain, and clinically, he found dilation of the pupils, which then became fixed, and a respiration that was comparable to a very deep sleep.⁷⁴ It became clear that when brain tissue shifted, the brainstem carried the brunt of the injury. In fact, pontomedullary hemorrhages were linked to acute space-occupying intracranial lesions. Later, Duret found these lesions in 30% of traumatic and spontaneous cerebral hemorrhages and they would carry his name.²¹

In a series of dog experiments, Duret described in considerable detail the clinical and pathological features of les phénomènes de choc after injection of water or gelatin into the cranium. These original studies became part of a medical doctorate thesis in 1878 (Fig. 1-1).²¹ The signs included sudden acceleration of blood pressure, respiratory arrest, slow pulse, and tétanisme. Duret hypothesized that a shock wave in the cerebrospinal fluid (CSF) was augmented in the aqueduct of Sylvius, causing hemorrhage under the fourth ventricle. (This explanation was based on Bernoulli’s law, which states that hydrostatic pressure is the greatest in the immediate post-stenotic area.)

The earliest pathology account of brain herniation is probably herniation of the cortex into the arachnoid villae, particularly under the surface of the temporal lobe as an attempt to absorb pressure. Other accounts are on brain herniation from missile wounds showing extrusion outside the bone defect. Cushing was one of the first clinicians to describe tonsillar herniation and to warn of the risk of lumbar puncture.

Not a few instances of disaster in consequence of lumbar puncture have been recorded in the literature, and six have come under my personal observation. Three of them were fatalities in the medical ward after puncture in cases of unsuspected cerebellar lesions. If the brain, after such an incident, is removed from the cranial chamber soon after death and particularly hard in situ, it will show the imprint of the foraminal ring about the protrusion that has been tightly jammed into the open ring.¹⁹

Cushing mentioned tonsillar herniation in his Mutter lecture of 1901. In this paper, he also cited Quincke, who had already called attention to the dangers of lumbar puncture. Cushing named it protrusion of the brainstem and actual hernia cerebelli. These accounts of the tentorial pressure cone found at autopsy were the first clinical-pathological correlations, and they highlighted the dangers associated with sudden changes in the CSF compartments.^53,65

Other papers more specifically described tonsillar herniation. Collier, in his article in Brain in 1904, noted that supratentorial tumors can press not only the tentorium downward but also the brainstem and cerebellum. The cerebellum was deeply indented by the edge of the foramen magnum, forming a so-called conical plug. Collier described his observations of cerebellar herniation as follows:

In many cases of intracranial tumor of long duration, it was found postmortem that the posterior inferior part of the cerebellum had been pushed down and backwards into the foramen magnum and the medulla itself being somewhat caudally displaced, two structures together forming a cone-shaped plug tightly filling up the foramen magnum.¹⁷

Alquier reported, in Revue Neurologique, two cases that he designated as heterotopie du cervelet dans le canal rachidien.⁶ Autopsy in two cases with brain tumors showed tissue displacement, “forced to migrate,” because of the pressure effect of the tumor. Alquier could not choose between two hypotheses; one supported a developmental anomaly and the other favored herniated cerebellar tissue compressing the brainstem. The concluding remarks of his mentor, Pierre Marie, are interesting but he dismissed it as a postmortem artifact.

The earliest anatomical description of uncal herniation can be attributed to Adolf Meyer in 1920.⁵⁶ The article comprised of a pathology series with virtually no explanatory text. The author declared at the outset: “The falx and tentorium constitute an important protection against any sudden impacts of pressure by keeping apart heavy portions of the brain, but they also provide an opportunity for trouble in case of swelling or need of displacement.”

Meyer’s article was rich in photographs (Fig. 1-2), but it drew few practical conclusions. Importantly, however, he described hemianopia as a false localizing sign of uncal herniation, resulting from compression of the posterior cerebral artery.

Other evidence of displacement was discovered at autopsy. Following a brief case report by Groeneveld and Schaltenbrand³² (Fig. 1-3A), pathologist Kernohan and neurologist Woltman⁴² published a seminal pathology work in 1929 on ipsilateral hemiplegia accompanying cerebral mass lesions. It provided the first comprehensive pathological observation that a groove of the shifted crus cerebri occurred on the side contralateral to a brain tumor (Fig. 1-3B). Pathological proof was provided: “herniation and displacement may be evidenced by a groove sweeping over the uncinate gyrus on the side of the tumor.” They concluded: “Notching of the crus cerebri by the free margin of the tentorium could, we believe, explain the homolateral signs of the pyramidal tract noted in most of our cases” (Fig. 1-3C).

Ipsilateral pyramidal signs with uncal herniation had puzzled physicians, and more than a few unnecessary craniotomies had been performed on the wrong side. Since it was now possible to explain homolateral hemiplegia, it could possibly reduce clinical error. Over the years, this observation has been referred to as the Kernohan-Woltman syndrome or Kernohan’s notch. However, even now, the mechanism by which this V-shaped indentation or groove is produced, whether by displacement of the brainstem at a diencephalic level or pushed by the herniating uncus, remains unclear. Moreover, this “notching” is more common as a result of gradual tumor growth and compression.

The anatomical relations of the tentorial hiatus were studied in greater detail in the following years. In 1938, Sir Geoffrey Jefferson³⁶ summarized it as follows:

The temporal lobes lie on the tentorium, which slopes away laterally as a gently inclined plane, so that pressure from above will tend to make them slide away from the midline. However if one lobe is enlarged it cannot escape overhanging the free edge. For this reason, a tumor of the temporal lobe will be the surest way of bringing it more firmly into contact with the midbrain and squeezing its inner border over the sharp edge of the falx, into a situation in which it can herniate downward into the posterior fossa.³⁶

Sir Geoffrey Jefferson (familiar to neurosurgeons by the eponym Jefferson’s fracture, a burst fracture of the atlas vertebra) also coined the term tentorial pressure cone or temporal pressure cone. This term introduced the word “pressure” in the dynamics of the process, but Jefferson was not quite able to pinpoint the main variables. Pathological findings of these cases seen at autopsy included compression of cranial nerves traversing the subarachnoid space, mesencephalic hemorrhages, and kinking of the posterior cerebral arteries.

In the same year, Moore and Stern⁵⁸ described 14 consecutive patients with brain tumors or abscesses that were specifically examined for evidence of vascular lesions. They described calcarine infarction and hemorrhages in the midbrain and pons. The hemorrhages in the brainstem were considered terminal events and were explained by reduced outflow, thus causing arterial congestion predisposing to hemorrhages. Acute increased intracranial pressure (ICP) was emphasized as a mechanism. Because the hemorrhages in the pons had major similarities to hypertension-induced hemorrhages, a similar mechanism was suggested with increased blood pressure causing hemorrhage in a fragile artery, the so-called locus minoris resistentiae. These secondary midbrain and pontine hemorrhages were not commonly seen in patients with transtentorial herniation and were topographically different from those described by Duret, who had noted them to be predominantly surrounding the fourth ventricle. Van Gehuchten,⁸³ Wolman,⁹⁶ and Friede and Roessmann²⁹ interpreted these lesions as ischemic of arterial origin and a consequence of prolonged tentorial coning. Wolman found thrombosed arteries in the same region, strengthening this concept.⁹⁶

An important advance came in 1939, when Reid and Cone⁷¹ published their experimental study. They induced acute compressive brain lesions in anesthetized Macacus Rhesus monkeys by infusing Ringer’s solution through trephine holes in the skull, with simultaneous measurement of ICPs. At will, they could induce and reverse pupillary dilatation through manipulation of the ICP (Fig. 1-4A–C). The animals were sacrificed after they developed pupillary changes. The oculomotor nerves were found to be compressed by the extruded hippocampal gyrus in most cases (Fig. 1-4C). The article additionally contained a description of clinical cases published in the literature and personal observations to augment the experimental findings.

Each clinical and experimental case that we have included has presented such a lesion in the form of a herniated hippocampal gyrus pressing on the third nerve. In some of our cases, the nerve was flattened or stretched and in one instance discolored...The amount of pressure necessary to produce the herniation in the normal animal may give some idea of the pressures in cases in human beings...In some of the animals, it was almost as high as systolic blood pressure and this may aid in the explanation of the infarctions that occur in man.⁷¹

This study was unable to duplicate findings such as the Kernohan-Woltman notch phenomenon or midbrain hemorrhages seen in human beings in similar circumstances. Nevertheless, the role of sudden elevation of ICP in the genesis of brain shift and uncal herniation was herein established. By demonstrating the reversible nature of the alleged signs of temporal lobe herniation, Reid and Cone were able to suggest the possible benefit of early surgery. Reid and Cone also described animal experiments reproducing the compressed third nerve with bilateral uncal herniation. Although they considered damage to the nucleus of the third nerve, they found that the herniated hippocampal gyrus flattened the third nerve in all monkeys and was considered the main mechanism.

Further refinement came in 1941, when Schwarz and Rosner⁷⁶ described a clinical-pathological study of herniation of the gyrus of the hippocampus derived from 43 cases, and with one detailed case study demonstrating the unraveling and sequence of clinical signs. These clinical findings included nuchal rigidity with resistance to lateral movement of the head and thermoregulatory disturbances. Often, the temperature reached 40°C, indicating that hippocampal herniation disturbed the blood supply to the midbrain. Lateral displacement of the brainstem was noted, with flattening of the aqueduct of Sylvius causing obstructive hydrocephalus. Oddly, the clinical signs were then used to extrapolate a temporal course, with the implicit assumption that such a course existed. Six stages were described: (1) fluctuation in state of consciousness, (2) anisocoria with or without disturbance in the light reflex, (3) nuchal rigidity, (4) impairment of extraocular muscles, (5) cardiorespiratory and thermoregulatory changes, and (6) paradoxical pyramidal tract signs followed by decerebrate rigidity.

Displacement of the brainstem itself was emphasized by Scheinker in 1945, who noted vertical shift and buckling for the first time. He stated that herniation of the hippocampal gyrus cannot be responsible for “gravity and danger of the clinical syndrome.” In his view—although not corroborated by data—occlusion of the aqueduct of Sylvius and intraventricular fluid pressure would rise, pushing the brainstem deeper into the tentorial opening.⁷⁵

In 1953, disturbances of ocular motor function were described in considerable detail by Sunderland and Bradley.⁸⁰ In patients with acute epidural hemorrhages, they found that the pupil on the affected side dilated. This was explained by the susceptibility of pupilloconstrictor fibers to deformation and pressure on the upper surface of the third nerve where the pupilloconstrictor fibers are located. Fixed pupils occurred, followed by loss of function of extraocular muscles. The changes in the opposite pupil were then explained by downward displacement of the basilar artery, pulling the posterior cerebral artery into the upper surface of the third nerve lying just below it and thus impairing pupil responses. In their reconstruction of events, central origin of a third nerve palsy was dismissed as an explanation because the extraocular eye mechanism would suffer and the opposite pupil was not affected until the signs were well advanced on the ipsilateral side.

Johnson and Yates,⁴¹ in 1956, also were interested in the pressure changes at the tentorium. This paper made two major contributions. First, it suggested that, with uncal herniation, the third nerve may be angulated across the petroclinoid ligaments, producing varying degrees of injury. Second, they described bilateral posterior herniation of the temporal occipital lobes, expanding cerebral hemispheres, gently sliding medially across the flat middle section, causing a pear-shaped compression, and compressing the dorsal midbrain, causing upward gaze palsy. It was of interest to the authors that upward gaze palsy produced a false localizing sign. It would suggest tumor in the pineal region, but, in fact, it was frontal and bilateral.

In 1959, Howell described upper brainstem compression and claimed a separate syndrome (Fig. 1-5).³⁵ He wrote:

Though the syndrome was complex and variable, its main features were sufficiently constant for it to be accurately distinguished from other diseases causing coma, in the majority of cases.

He correlated changes in respiration from noisy and labored breathing to apneic pauses followed by slow and gasping breathing, with heavy congested edematous lungs. Loss of pupil reflex (pupils were small or contracted) was a constant observation. Decerebrate rigidity was also common.

Similarly, the observations by McNealy and Plum,⁵⁵ examining 52 patients with supratentorial mass lesions (Fig. 1-6), distinguished herniation from displacement of the uncus into the tentorium from another syndrome that they felt was more frequently observed and named it central syndrome or bilateral diencephalic impairment. The discussion stated that “throughout this study the predictable inexorable manner in which decaying brainstem function followed an orderly rostrocaudal pattern has been emphasized, for this orderly deterioration greatly assisted the accurate clinical diagnosis of coma after supratentorial disease.” This central syndrome was characterized by the development of Cheyne-Stokes breathing; small reactive pupils; hyperactive oculocephalic responses; and bilateral motor changes, with the development of paratonia (resistance to passive movement) or decorticate rigidity in several stages, from an early diencephalic to late diencephalic, midbrain upper pons, lower pontine, and upper medullary stage. Key clinical features that include respiration, pupillary responses, oculocephalic responses, and motor responses could help in recognizing this syndrome. They further emphasized that “orderly progression of signs was invariable unless intraventricular hemorrhage or ill advised lumbar puncture rapidly altered cephalocaudal pressure gradients to produce medulla ischemia failure.” Key to its progression—named rostrocaudal deterioration—was a loss of function of the diencephalon, followed consecutively by loss of function of midbrain, pons, and terminally, the medulla oblongata. In their view uncal herniation also may progress with a rostrocaudal course with third-nerve compression but no impaired consciousness (due to sparing of the diencephalon structures). Signs of midbrain or pons involvement were noted in a subsequent stage when shift progressed. The downward displacement was facilitated by three factors: (1) arterial compression followed by infarction and edema (anterior cerebral artery under the falx and posterior cerebral artery from uncal compression), (2) venous compression followed by increased ICP (due to compression of the great cerebral vein), and (3) obstructive hydrocephalus at the aqueduct, further increasing ICP. According to the authors, these factors create a vertical force that could explain the rostrocaudal direction.

The clinical signs of brain herniation provided a touchstone that would remain the conventional standard. Questions of validity were raised most prominently by Finney and Walker in their monograph on transtentorial herniation.²³ It was only in the 1980s that Fisher²⁶ and Ropper⁷² challenged this dogma and surmised that many of the arguments used were unsatisfactory. Ropper felt that the uncal herniation was more a passive process than an active one as the mesencephalon was moved by the expanding cerebral mass and the ipsilateral ambient cisterns widened.²⁶ The main wrangling of Fisher was “the claim is being made that the aperture cerebellar herniation is an incidental late byproduct, a harmless telltale of increased posterior fossa content on the diffuse elevated pressure and not a special instrument for adding to the damage by throttling the brain stem.”26 Fisher noted two findings. He found that extreme cerebellar herniation can be asymptomatic and described an autopsy case of a 3- to 4-cm cerebellar herniation with hemorrhagic necrotic tips, proving that it had been present for several days. Secondly, in a pathological study, Fisher found cerebellar hemorrhage with posterior fossa compression and respiratory failure, but no pressure cone, in 17 of 18 patients. Fisher said in a passage of some importance, “displacement of cerebellar tissue into the foramen magnum may even afford perhaps some relief from crowding rather than being harmful.”26 He also questioned whether urgent decompression could rescue these patients from respiratory arrest. “Our concern is how often we see respiratory arrest before loss of brainstem reflexes. Acute respiratory failure causing death in this condition may be rare.”²⁶ The concept of uncal herniation was also considered disputable. Autopsy cases were described with horizontal shift and secondary brainstem hemorrhages, but absent temporal lobe herniation. Clinicians did note that hyperosmolar solution could reverse the clinical course, but it was hard to imagine that tissue wedged in the tentorial opening would become dislodged. An alternative explanation was that the lateral shift could deform the midbrain and stretch the extra-axial third nerve. His provocative, but ultimately correct, conclusion was that lateral displacement of the brainstem is the prime mover, not transtentorial herniation. Ropper added to this concept in an influential paper in 1986, in which lateral displacement of the brain tissue and level of consciousness were closely correlated⁷² (Figs. 1-7A, B). This work revisited the major tenets of brain herniation, with some CT scans showing a laterally displaced and rotated brainstem opening up the ambient cistern, but without the presence of herniated hippocampal tissue. This correlation of horizontal shift on CT scan and level of consciousness could be clinically relevant. If the degree of horizontal shift did not correlate with the level of consciousness, neurosurgical evacuation may not improve consciousness.

These observations correctly de-emphasized or even refuted brain tissue herniation as a clinical phenomenon. It is a fact that the pathologist always turned more attention to the herniation of brain tissue through openings rather than to the damage done to the thalamus and upper brainstem from displacement or compression.

At the end of the 19th century, investigators interested in the role of increased ICP in coma became more prolific. Earlier surgical pioneers who explained cerebral pressure following trauma included Von Bergmann.⁸⁴ His experiments described animals that developed clonic spasms when pressure was suddenly applied to the brain. A slow heartbeat, deep and snoring respirations, vomiting, and incontinence were other remarkable symptoms. These symptoms were not seen when pressure was slowly increased. To produce death, ICP must equal the carotid pressure.⁸⁴ Immediately after the injury, the blood pressure rises and then falls. There is a paralysis of the respiratory center; however, if mechanical ventilation is applied, the pulse remains strong. Hill suggested that these early effects were due to diminished flow in the bulbar centers.³⁴ Hill also suggested that death occurs when ICP equals the system blood pressure in the carotid arteries.

Cushing is already acknowledged for his astute observation of brain herniation, and any historical account on the role of ICP must begin with Cushing’s contributions to its physiology. Cushing performed his animal experiments in Kocher’s laboratories (Figs. 1-8A,B). (A general surgeon, Theodor Kocher in Berne, Switzerland, received the Nobel Prize in 1909 for his work on goiters.) Cushing worked through the winter of 1900 on the following question given to him by Kocher: “To decide if incompression of the brain, the small veins and capillary vessels are dilated by stasis or compressed.” Increase in blood pressure with brain compression became known as “Cushing’s law” or “Cushing’s response” (Fig. 1-9), but this phenomenon had been observed earlier by other researchers. The work was first published in German in 1902¹⁸ and was notably followed by an acidic comment by Bernhard Naunyn, who claimed to have made the same observations 20 years earlier and felt undercredited. However, Cushing’s original experiments in dogs were highly important, and he introduced the measurement of intra-arterial blood pressure. ICP was increased by infusing saline in a rubber cannula, and Cushing documented this with increase of ICP; blood pressure would increase above the pressure that was applied to the vasomotoric center in the medulla.

With rapidly increased ICP, there was vagal activity with a decrease in pulse, sometimes asystole and shallow breathing. With further increased ICP, the regulatory function of the vasomotor center and medulla would become “paralyzed” and would not respond to hypoxemia of the medulla oblongata, and this would then result in hypotension. Therefore, these findings were best understood as the increase in systolic blood pressure due to increase of ICP as a result of increased activity of the vasomotor center in the medulla, in turn resulting from a decreased cerebral perfusion and ischemia. He could also document that after the vagal nerve and the spinal cord were cut, the blood pressure and pulse did not change with rising ICP. When only the vagal nerves were cut, the blood pressure would simply follow ICP.

Cushing’s work came to be treated with respect and even awe, despite attacks and futile attempts to discredit the findings. However, it became clear in subsequent papers that these changes in systemic blood pressure, pulse rate, and respiration were not demonstrated until the ICP reached or surpassed the level of diastolic pressure. Kocher and Cushing made substantial advances in the field by proposing the following stages of medullary compression: (1) First stage: accommodation—kompensation. In this stage, the CSF is displaced out of the cranial vault, followed by encroachment upon the cerebral venous bed with little change in the systemic circulation. (2) Second stage: stage of early manifest symptoms—anfangsstadium des manifesten hirndruckes. In this stage, blood from the capillaries has been expelled, and “anoxemia” of the vital bulbar centers results in rise of the systemic blood pressure. The pulse rate is retarded, but the pulse has a full quality. The respiratory rate is also reduced. (3) Third stage: stage of advanced manifest symptoms—hohe stadium des manifesten hirndruckes. In this stage, the respirations are more snoring and rhythmic and may be of the Cheyne-Stokes type. Papilledema is seen, and pupils become irregular. (4) Fourth stage: stage of medullary collapse—lähmungs stadium. In this stage, the “vital centers are exhausted.” The blood pressure is decreasing, and the patient is in shock, with all reflexes abolished, pupils dilated, and with irregular respirations with apneic episodes. These postulates, based on experimental findings, were accepted, and clinicians had no difficulty in putting them into practice and instructing nurses about them. In his work Hinnerschütterung, Kocher clearly recognized the lifesaving effects of decompression or trepanation in patients with increased ICP.

There were other important contributions, particularly on the consequences of increased ICP.⁸¹ Taylor and Page felt that arterial hypertension due to increased ICP was a combination of ischemia and mechanical compression. Others opted more for axial distortion of the brainstem to explain this vasopressor response.⁸² Langfitt, among others, documented that traumatic brain injury can cause a marked increase in ICP even without mass effect from a contusional lesion. Two major periods were observed: first, a brief rise in ICP due to the impact; second, a rise associated with cerebrovascular dilation from reduced cerebrovascular tone, resulting in an increase of cerebral blood volume.^46,88 Despite these landmark findings, the physiological changes in brain tissue shift with increased ICP remain incompletely understood.

One would like to believe that small pieces of a puzzle were gradually discovered, which led to the emergence of an easily recognizable clinical picture. However, clinical patterns are far from coherent. In this section, the basic clinical signs are further discussed.

Extensor rigidity with head retraction had been noted by clinicians in comatose patients and most notably in patients with cerebral hemorrhage extending into the ventricles. Sir Charles Sherrington, a Nobel laureate in medicine and physiology, described decerebrate rigidity in his animal experiments.¹³ Transection experiments in cats demonstrated the existence of a conceptional transverse plane at the level of the corpora mammillaria, red nucleus, and between anterior and posterior colliculi (line A in Fig. 1-10). Transection would produce decerebrate rigidity. Rigidity disappeared when a transection occurred caudal to that plane at the level of the vestibular nuclei (line B in Fig. 1-10). These findings were confirmed by others.⁶⁹

One of the first descriptions of decerebrate rigidity in man was by Walshe in 1923⁸⁷ (Fig. 1-11). He described that

the patient lay motionless and unconscious on her back with the head in a median position. There was no trace of head retraction. The arms lay across the chest, semiflexed at the elbows, with the forearm slightly pronated and the wrists and digits flexed. The legs lay extended and adducted with the feet plantar-flexed. There was spasticity of moderate degree in all 4 limbs, definitely more pronounced in the arms than in the legs.

In addition, the article described an abnormal tonic neck reflex of Magnus and De Kleijn (normal in neonates, known as the tonic neck reflex; Fig. 1-12).

To translate these experiments to clinical observations is difficult, but Walshe suggested that next to a midbrain lesion, a lesion of the forebrain or a ventricular hemorrhage could interfere with the activity of the midbrain centers⁸⁷ Walshe noted: “yet the lesion does not end here and in almost all of them, there is clear evidence of a progressive and ultimate fatal interference with the function of the vital medullary centers.” Decerebrate posturing became mostly recognized after traumatic brain injury and was recognized as a poor prognosticating sign.²⁰ In Fulton’s studies, decorticate responses with grasp reflexes were found in animals after removal of both motor and premotor cortices.³¹

Its localizing value in humans is less understood, and it can be observed in midbrain lesions or lesions involving injury to both hemispheres without evidence of brainstem injury or displacement. Moreover, worsening to flaccidity or change to withdrawal or decorticate responses is not clearly correlated with outcome. Both decorticate (pathological flexion) and decerebrate (pathological extension) responses indicate a high likelihood of a severe structural injury. There are many patients with a decorticate response on one side and a decelerate response on the other or even alternating in the acute stage. Often the “worse” response correlates with the hemispheric lesion. We can only conclude that it is more likely that the responses are different manifestations of a similar lesion rather than being precisely localizable.

The observation that pupils dilate and become fixed to light came first from experimental studies. In the early 1800s the German internists and surgeons Von Leyden, Naunyn, and Bergmann all noted in their ICP studies that pupils dilate with increasing ICP. The pupillary changes were considered a result of medulla oblongata ischemia because their appearance was so closely related to the appearance of hypertension and periodic breathing.

Changes in pupils have fascinated clinicians and their significance sweeps far beyond any other sign in coma. The earliest clues to this time-honored sign can be traced back to 1867, when Sir Jonathan Hutchinson observed unilateral pupillary dilatation in a patient, although he dismissed its significance. It was forgotten for two decades before Sir William Macewen⁵¹ recognized its import in a treatise on the pupil (Fig. 1-13): “the patient was generally insensible at the onset when both pupils were dilated and fixed. As the patient recovered consciousness, one pupil became normal while the other remained dilated and fixed; this being on the side of the lesion.”

The surgeon Bergmann was convinced the lesion was located in the cortex. In his classic work (Deutsche Chirurgie: Die Lehre von den Kopfverletzungen, 1880) he located oculomotor dysfunction in the frontal eye field. His clinical observations also described widening of the pupil ipsilateral to the lesion.

Collier argued that oculomotor palsies were “always of peripheral type.”⁶⁵ Plum and Posner⁶⁷ suggested that downward movement of the posterior cerebral artery compresses the third nerve. As mentioned before, the correlation between fixed dilatation of the pupil and herniation was clearly described by the classic studies by Reid and Cone,71 and by Jennett and Stern,⁴⁰ who replicated the experiment in cats. Jennett and Stern had postulated the following:

The rapidity with which cardiorespiratory pupil and electroencephalographic changes usually return to normal on releasing the pressure; although, the hernia clearly persists—calls into question the rationale for splitting the tentorium in patients with persisting symptoms after removal of mass lesions.

The authors also stated that

the tentorium was removed in certain animals, and the changes were observed different from those that were noted with the tentorium intact; namely the severe respiratory changes were seen prior to pupillary dilation suggesting perhaps that with the tent gone, there was more ready transmission of the distorting effect to the lower brain stem.⁴⁰

Out of conformity, Fisher-Brugge coined the term Das Klivus Kanten Syndrome (the edge of the clivus syndrome). This syndrome largely consisted of unilateral or bilateral dilated fixed pupils and decreased consciousness; it was no different clinically than uncal herniation, but no compression by the uncus was found. Abnormal consciousness was attributed to compression or ischemia of the mesencephalon. Fisher-Brugge was convinced that the uncus of the hippocampus played no role in the cause of the fixed pupil.²⁴ The third nerve could be damaged owing to its position—wedged in between the edge of the clivus and the tentorial ridge. Petechial hemorrhages in the nerve were put forward as additional proof. It could also explain enlargement of the pupil on the opposite side of the mass.

Ropper suggested acute angulation of the third nerve over the clivus due to displacement of the brainstem in an autopsy study.⁷³ Therefore, the mechanism of pupillary enlargement has not been definitively explained, and more than one mechanism may be operative. How the opposite pupil enlarges with transtentorial herniation remains an anatomical mystery, with bilateral central (at the nucleus level) third-nerve damage being a more likely mechanism.

The initial discovery that the oculovestibular reflex is impaired in coma was by Klingon, who demonstrated disorders of the conjugate movements of the eyes after stimulation with cold water.⁴³ Disconjugate ocular responses (abduction of the eye at the stimulated site with the opposite eye frozen) correlated in a comatose patient with demyelination in the tegmentum (Fig. 1-14). Nathanson and colleagues⁶⁰ described, in 1957, the possible usefulness of oculocephalic and caloric responses in comatose patients and included patients who had complete absence of the oculocephalic reflex and caloric stimulation when treated with barbiturates. The clinical-pathological correlation was illustrated by massive brainstem lesions from basilar artery occlusion or swollen glioblastoma with brainstem hematoma. In these patients, disconjugate ocular movements were found; however, the presence or absence of cornea reflexes and pupillary light reaction did not correlate with the findings on oculocephalic and caloric tests. Patients who had a tonic ocular deviation had return of consciousness, while the absence of caloric stimulation and oculocephalic reflexes correlated with no recovery. Thus, oculocephalic and caloric tests were considered indicators of depth of coma and, when absent, indicative of a poor prognosis.⁴³ Furthermore, Ethelberg and Vaernet²² demonstrated the abnormality of conjugate eye movements similar to internuclear ophthalmoplegia in three cases of supratentorial space-occupying lesions. These earlier studies pioneered the use of clinical tests to assess brainstem function.

The discovery of the morphology of the respiratory center in the medullar oblongata can be attributed to Legallois, in his rabbit experiments in 1812.⁴⁸ This was followed by a series of studies linking brain injury to abnormalities of the rhythm of breathing.⁶⁶ Abnormal breathing patterns had been recognized as indicative of a primary brain lesion, and the most commonly known were periodic breathing patterns. The best recognized is the Cheyne-Stokes respiration (CSR), characterized by repeated periods of hyperpnea that alternate with apnea, with cycles that are not random but stereotypical. Each cycle may last approximately one to three minutes. In 1818, Cheyne¹⁶ described a patient with a stroke with a peculiar breathing pattern (Fig. 1-15). “For several days his breathing was irregular. It would cease for 1/4 of a minute and then it would become perceptive; although very low; and then by degrees it became heaving and quick and then it would gradually seize again.” Stokes described a similar pattern, of which he stated as follows: “the symptom in question was observed by Dr Cheyne, although he did not connect it with a special lesion of the heart.”^52,64 Studies that connected CSR with autopsy-proven structural lesions of the brain were subsequently reported.

Another classic central breathing pattern is Biot breathing.⁹⁰ Biot noted that this breathing pattern is different from CSR (Fig. 1-16). He distinguished this breathing pattern from Cheyne-Stokes breathing and, because he noted it in a patient with tuberculous meningitis, named it rhythme méningitique. The breathing pattern is irregular and rapid, with rhythmical pauses lasting 10 to 30 seconds, but sometimes with alternating periods of apnea and tachypnea. This breathing pattern lacked the crescendo–decrescendo cycles attributed to Cheyne-Stokes breathing and was completely irregular, with varying periods of apnea.^10,11

Central neurogenic hyperventilation was first described by Plum and Swanson⁶⁸ in 1959. Central neurogenic hyperventilation results in alkalosis due to a very high respiratory rate (60 to 100 per minute). In this study, the lesions that correlated with central neurogenic hyperventilation were mostly in the pons. Nine patients were described who developed “severe hyperventilation during the course of acute central nervous system disease.” In all of these patients, medial pontine damage was responsible for hyperventilation. In their hypothesis, “central neurogenic hyperventilation in man results from the uninhibited stimulation of both inspiratory and expiratory centers in the medulla by a lateral pontile reticular formation and bilateral located descending neuro pathways.” In all patients, there was profound hypocapnia with respiratory alkalosis but no hypoxemia.⁶⁸ Since this original description, many reports linked central neurogenic hyperventilation due to brainstem lymphomas.

Lower brainstem lesions can produce ataxia of respiration characterized by irregular breathing, prolonged inspiratory gasps, and apneustic breathing. Many of these observations were seen in patients with acute bulbar poliomyelitis,⁶⁸ but other cases involved patients with pontine hemorrhage and infarction, in whom irregular respiratory rhythm and apneic failure were reported. Steegmann⁷⁹ described patients with irregular slow and labored gasping stertorous respiration. Deep inspiratory gasps were described with diaphragmatic excursions but without intercostal movement. With each gasp, the chest wall retracted. In other cases, respiration was reduced to two respirations per minute or shallow without a change in rate. He correlated these inspiratory gasps to apneustic respirations in experimental animals. The experimental studies referred to in his paper were by Marckwald, who located a regulatory center at the inferior colliculus and found that long, powerful inspiratory cramps were interrupted by intervals with short expiratory pauses. Apneusis could be produced by severing the vagus nerve, transecting at the pons just posterior to the inferior colliculus.⁵⁴

Another neurogenic breathing type is cluster breathing. Clusters of regular breathing (including tachypneic periods) are interrupted by regular or irregular pauses. Although, allegedly, it has been associated with brainstem lesions, it may be more likely a variant of Cheyne-Stokes breathing (but missing the crescendo–decrescendo pattern).⁶⁷ Cluster breathing has been recently described with bihemispheric lesions sparing the brainstem.²⁸

In summary, one can only conclude that the localizing value of certain breathing patterns is limited, but its presence or sudden appearance has practical significance. It tells the clinician that the patient is potentially deteriorating neurologically and that oxygenation may become compromised, requiring endotracheal intubation and mechanical ventilation.

In the second half of the 19th century, most notably Osler⁶² noted that coma can be due to intoxications, infections, and organ failure.⁴⁴ Plum and Posner⁶⁷ in their 1966 textbook categorized it as metabolic, “caused by diffuse failure of neuronal metabolism,” and distinguished primary and secondary metabolic encephalopathies. Metabolic encephalopathy has since become a medical umbrella term and includes all conditions not associated with a new mass or destructive lesion.

Much of the understanding of the so-called metabolic causes of coma came from animal experiments. Experimental studies have consistently produced evidence of “selective vulnerability of the brain” and excessive glutamate in brain in the pathogenesis, among other mechanisms.⁷⁷ Uremic coma and encephalopathy at the earlier stages of renal failure have been best described by Bright¹² (Fig. 1-17). Bright noted multiple cases with headache, lassitude, intermittent confusion, and myoclonus evolving into a more serious state heralded by seizures, stupor, and coma. Osler emphasized “mania, noisy and restless patients with a delusional insanity.” Osler noted that seizures were not obligatory before a lapse into stupor, and noted focal signs. After dialysis became commonplace, severe forms became less noticeable. The uremic neurotoxins have remained elusive.

Initially, very little research has been done in understanding encephalopathy associated with acute metabolic derangements. Most of the laboratory experiments included both rapid lowering of plasma glucose and examination of the effects of hyperglycemia. These studies included exploration of the pathogenesis of hyperglycemia and, most notably, the introduction of the existence of idiogenic osmoles by Arieff and Kleeman.⁷ In these studies, animals became acutely hyperglycemic, and blood sugar levels were then corrected with insulin, fluids, or peritoneal dialysis without insulin. During hyperglycemia, the brain water content fell and equalized osmolality of the brain and the CSF. After several hours, the brain water content returned to normal, with no solutes that could explain this phenomenon. The return of brain water to normal levels was attributed to the formation of “idiogenic osmoles.” This experiment suggested that the diabetic brain accumulates an extra solute that defends against brain dehydration. These molecules have included glutamine, taurine, and myoinositol.

Hepatic encephalopathy and decreased consciousness had been noted in von Frerichs’ work. The emergence of jaundice marked the development of delirium, convulsions, and coma. In 1860, von Frerichs emphasized a transitional phase of “gloomy, irritable temper and restlessness” but also “quiet, harmless wandering.” Convulsions were noted in one third of the patients.⁸⁵ These symptoms paralleled the appearance of jaundice, a clinical sign that could be without neurologic manifestations until “a train of symptoms betokening danger supervened.” A landmark paper by Adams and Foley⁵ further delineated clinical symptoms and pathological changes of the brain. This clinicopathological study introduced asterixis as a key finding (Fig. 1-18). Adams and Foley documented asterixis with electromyographic recordings, but when coma occurred, prognosis was poor, with most patients dying within two weeks. Progressive hepatic encephalopathy in fulminant hepatic failure has only been recently recognized as a clinical syndrome associated with treatable cerebral edema. Treatment requires measures to reduce ICP and emergent liver transplantation. Other “metabolic encephalopathies” associated with hypothyroidism, Addison’s disease, acute pancreatic disease, and sepsis have remained poorly understood due to the lack of a specific animal model.

Of all the diffuse encephalopathies, anoxic-ischemic injury to the brain after cardiac arrest became of interest to pathologists in the early 1950s.⁵⁹ With the introduction of cardiopulmonary resuscitation (CPR), cerebral damage became recognized by clinicians and first by Negovsky, who named it post resuscitation disease. Some of the early hypothesis included cerebral reoxygenation injury and postischemic vasospasm cascading to necrosis.⁶¹ Some of the attention was directed to the different vulnerability of gray and white matter, and it became clear that the cerebral and cerebellar cortices were more affected with ischemia than the basal ganglia, with the reverse happening with hypoxemia. Levine reported in his rat experiments that white matter was more resistant to anoxic-ischemic injury than gray matter.⁴⁹ These observations became even more important when the clinical landscape changed with the introduction of CPR standards set by the American Heart Association in 1966.² Clinicians now noted that despite treatment of shock and CPR, the brain could still have irreparable damage.⁸ In his animal experiments, a good deal more optimistic than observed in practice, Heymans correlated improvement of clinical function with the time needed for revival and found five to 10 minutes for the cornea and pupil reflex and 15 to 30 minutes for the vasomotor and respiration regulation as the threshold for irreversibility.³³

Attention to psychogenic unresponsiveness may have started with the major hysterical attacks resulting in prolonged unresponsiveness and have been best described and shown in Charcot and Richer’s Les démoniaques dans l’art (Fig. 1-19). Visual hallucinations of animals (rats, snakes) precede trembling motions or contraction of one arm or leg, followed by the whole body. A “seizure” may follow with breath-holding pallor followed by redness, neck engorgement and upward eye deviation. Foam can appear on the lips, usually early and not during the resolution phase. The muscles are completely “relaxed.” Often, this phase is followed by “acrobatics” in some patients (see Fig. 1-19). Extreme opisthotonus may result with a patient curved forward with the head and feet touching the bed or may occur with the patient lying on the side. Charcot also described bizarre movements (as if wrestling with an imaginary being) and savage cries. Respiration may be hissing and interrupted by hiccups. The mouth may be open with tongue protrusion. Charcot often described these contortions as “being possessed.” His treatment of young afflicted women (but also men) included hypnosis and most memorably the compresseur ovarien (an abdominal vise with a knob applying pressure to the ovary) and both could stop the spells. Psychogenic unresponsiveness found entry in psychiatric and medical texts but remained merely descriptive and less ostentatious as Charcot’s examples, often as an afterthought after exclusion of other causes, and certainly not as a first consideration. The condition is very rare, and systematic study in a series of patients has not been reported.

Many astute neurologists noted that patients with prolonged unconsciousness had similar clinical characteristics and a need for further classification emerged. How to best name these conditions did not seem easy. Earlier terms for prolonged unconsciousness were discarded because they failed to capture the essence of this state, namely a totally unconscious state with only autonomic function being retained. Kretschmer⁴⁵ introduced the term Das Apallische Syndrom in 1940 (Fig. 1-20), and the connotation is still used in a few European countries. Kretschmer coined the term to be similar to apraxia and agnosia. Apallia referred to a lesion of the pallium, the mantle of gray matter forming the cortex. Other, now abandoned, terms include coma vigile, la stupeur hypertonique post-comateuse, and vie végétative, and more recently wakeful unconscious state.

The first attempts to clinicopathologically define this syndrome (“neocortical death”) came from the Institute of Neurological Sciences in Glasgow, but this term was applied only to patients after cardiac arrest. Jennett and Plum³⁸ proposed the term persistent vegetative state and described clinical features that drew a distinction from other less severely disabling neurologic states (Fig. 1-21). In their 1972 Lancet communication, they wrote

the word vegetative itself is not obscure: vegetate is defined in the Oxford English Dictionary as “to live a merely physical life, devoid of intellectual activity or social intercourse,” and vegetative is used to describe “an organic body capable of growth and development but devoid of sensation and thought.”

Jennett and Plum coined the term to emphasize the “vegetative or noncognitive components of the nervous system.” Plum mentioned that “the term persistent autonomic state could have been employed almost equally well,” but the term was “less flexible” and “would have been less understood by the patient’s family.”⁷⁰ The term persistent vegetative state became transfixed in the medical vernacular.

Cairns and colleagues¹⁴ can be credited for introducing the term akinetic mutism. Patients are unaware and mostly immobile but track objects or fixate on their surroundings. In the original description of akinetic mutism, the clinical picture is described as follows:

The patient sleeps more than normally, but he is easily roused. In the fully developed state he makes no sound and lies inert, except that his eyes regard the observer steadily, or follow the movement of objects, and they may be diverted by sound. Despite his steady gaze, which seems to give promise of speech, the patient is quite mute, or he answers only in whispered monosyllables. Oft-repeated commands may be carried out in a feeble, slow, and incomplete manner; but usually there are no movements of voluntary character; no restless movements, struggling, or evidence of negativism.¹⁴

Cairns and colleagues noted fluctuations and episodes where the patient would respond by some speech and purposeful movements.¹⁴ The clinical entity is difficult to recognize and may disappear with the introduction of new less precise terms (Chapter 4).

Over the years rehabilitation physicians in recognized that patients in a persistent vegetative state were or became sometimes more awake, albeit minimally so. These patients at times were responsive and other times completely noncommunicative, and therefore an attempt was made to better define this condition tentatively named Minimally Conscious State and to set it apart from persistent vegetative state. This started a new and ongoing discussion on further classifying prolonged comatose states—and unfortunately further ambiguity (Chapter 4).

At the far end of the spectrum of comatose states is brain death. This clinical condition, characterized by coma with loss of all brainstem function including breathing and blood pressure regulation, became noted only in the early 1950s. Irreversible coma and brain death entered the medical lexicon. The initial descriptions of brain death came from neuropathological and electrophysiological observations. As in so many other conditions, neuropathologists were the first to find remarkable characteristics in comatose patients who were on the ventilator. The term respirator brain was coined quickly after the pathological features showed a dusky, congested, discolored appearance with liquefied portions and often crumbling cerebellar tonsils.⁴⁷ This finding was simply a consequence of a very high ICP resulting in pressure coning and global intracranial circulatory arrest. At least several days were needed for the brain to become discolored, showing vasocongestion and thrombosis in the venous system. However, outer portions of the cortex could remain intact due to retained extracerebral circulation.^47,86 Parallel to this finding was the observation by electroencephalographers of a new electrographic phenomenon called isoelectric or low-voltage electroencephalogram (EEG) in deeply comatose patients with loss of all brainstem reflexes (Fig. 1-22).²⁵

It took several years to clearly define the clinical accompaniment of both these laboratory findings, and the discovery was, therefore, in a somewhat reversed order. Although there were prior mentions of brain death (focused on absent flow by angiography, or on isoelectric EEG), two French groups proposed criteria. In 1959, Wertheimer, Jouvet, and Descotes⁸⁹ were among the first to propose criteria for this new clinical state. This manuscript largely focused, as many before, on the significance of an isoelectric EEG but also documented stopping the ventilator to stimulate the respiratory centers through increasing respiratory acidosis. Medulla oblongata function was further tested by carotid compression or ocular pressure (no change in pulse rate), and intravenous injection of atropine and amphetamine (no bradycardia). Neurosurgeon Wertheimer and neurologist Jouvet were set on defining brain death by electrophysiological criteria and the authors felt objective criteria were needed. They inserted fine bipolar electrodes into the medial thalamic structures, applied strong electrical currents, and found no motor response as proof of no brain function (Fig. 1-23).

Even more important and also in 1959, neurologists Mollaret and Goulon⁵⁷ published Le coma dépassé (best translated as “irreversible coma”), which described for the first time comprehensive neurologic findings in patients who lost brain function (Fig. 1-24). This type of coma (dépassé) was now classified as the deepest coma known so far (Fig. 1-25). Mollaret and Goulon reserved judgment on whether this condition meant death of the individual. Their description included immobility of eyeballs in a neutral position, fixed and dilated pupils, absent blinking with stimuli, absence of swallowing reflexes, drooping of the jaw, absence of motor response to any stimuli, muscle hypertonia, areflexia of tendon reflexes, and equivocal plantar reflexes. More clinical details included “medullary automatisms,” lack of spontaneous respiration after discontinuation of ventilation, immediate cardiovascular collapse as soon as the adrenalin infusion was stopped, and a disturbance of thermoregulation and hypothermia, depending on the environmental temperature (poikilothermia).⁵⁷ The paper also noted the absence of reactivity of the isoelectric EEG. The original observations of Mollaret and Goulon also included deterioration in these comatose patients with the development of oxygen desaturation, hypercapnia, and appearance of combined respiratory metabolic acidosis, polyuria, hypoglycemia, and glycosuria.⁵⁷ They also found that the heart rate often slowed to 40 beats per minute with no change in pressure of the eyeballs, or carotid sinus, or use of intravenous atropine.

There has been some historical discussion as to whether neurologists Mollaret and Goulon or neurosurgeon Wertheimer and neurologist Jouvet were the first physicians to describe brain death. Both papers—published months apart—have comparable value. A strong argument can be made that Mollaret and Goulon had a more detailed paper, a larger series of patients, significantly more detail on how the patient’s absence of brain function affected systemic instability, and further characterization of laboratory abnormalities. However, this and other earlier European papers did not generate much interest in the United Kingdom or the United States and very little academic debate ensued in the following decade.

A major development occurred in 1968 when the Harvard Medical School Ad Hoc Committee was formed to examine a definition of brain death and published a guideline.^3,92 This committee, spearheaded by anesthesiologist Beecher, commissioned neurologists Schwab and Adams to write the initial drafts. The committee worked diligently and in four months produced an important document that was published in JAMA (Fig. 1-26). The document included a definition of brain death, a legal commentary, and a supportive address by the Holy Father Pope Pius XII. No public opinion was voiced and little is known about the implementation at that time.

In the United Kingdom, eight years later, the Conference of Royal College of Physicians—with a major contribution by Pallis—published their document on “Diagnosis of brain death” in 1976.⁶³ Their clinical diagnosis was called brainstem death, and their position was that if the brainstem is dead, the brain is dead; and if the brain is dead, the person is dead. This also implied that confirmatory tests were not required to document absence of brain function.³ The document was important because it more clearly delineated confounding factors, a flexible period of observation, and a technique for apnea testing.

The centrality of the brainstem in death determination has been recognized by many and even Charcot (“le bulbe c’est ultimum moriens des centres nerveux”). Still there was a major focus of U.S. neurologists on cortical injury and the need to document a “flat EEG” when diagnosing brain death. The emphasis on cortical function may have been the reason that during this time, a confusing discussion appeared with new terminology. Brain death was now classified by some bioethicists as whole brain death (hemispheres and brainstem irreversibly damaged) but also higher brain death (hemispheres irreversibly damaged), brainstem death (brainstem irreversibly damaged), and even a super locked-in syndrome (brainstem irreversibly damaged but hemispheres functioning), dividing the death of the brain into segments. This discussion largely was helpful in distinguishing persistent vegetative state (higher brain death) from brain death (whole brain death). It became clear that the dividing line between these positions was the tentorium—brain versus brainstem.

In a series of papers published in the British Medical Journal, Pallis described the diagnosis of brainstem death and the pitfalls and preconditions.⁶³ His position was that these two different concepts could be reconciled by accepting the loss of all brainstem reflexes as the point of no return (Fig. 1-27). The mere fact that most patients with a catastrophic neurologic injury do not lose all brainstem reflexes points to its natural resilience. The UK position was that the brainstem was the defining part of the brain. However, the UK position also stressed that the patient had to pass through two filters, those of previous conditions and exclusions. Although prior criteria mentioned mimicking factors, this approach was novel in suggesting that no patients should be examined unless this issue was addressed. In addition, testing for apnea was described in more detail using preoxygenation and diffusion oxygenation. The Conference of Medical and Royal Colleges in the UK later changed “brain death” to “brainstem death,” again emphasizing the importance of the brainstem.

Another notable development was the National Institutes of Neurologic Disorders and Stroke (NINCDS) Multicenter Collaborative Study on Cerebral Death¹ (Fig. 1-28). This prospective study enrolled patients with “cerebral unresponsivity and apnea.” Of 189 patients, 187 died from cardiac arrest and two survived, but these were patients with drug intoxication. Their proposed criteria required one examination at least six hours after the onset of coma and apnea. Problems have been recognized and not all of the patients may have met the criteria of brain death by current standards. This study was followed by a report by the “Medical Consultants on the Diagnosis of Death to the President’s Commission on Ethical Issues in Medicine and Biomedical and Behavioral Research.”⁴

In 1995, the American Academy of Neurology published its evidence-based guidelines⁹⁰ (Fig. 1-29). This document specifically defined clinical testing of brainstem function, described conditions that could mimic brain death, listed observations that are compatible with brain death but suggested otherwise, and gave a clear description of the apnea test procedure. The validity of confirmatory laboratory tests was also critically reviewed. An update of the guideline, incorporating new literature and a checklist, followed in 2010. The document also concluded that in adults one full examination would be sufficient and that ancillary tests had little place in the diagnosis of brain death⁹³ (Chapter 5).

It was not until the 1960s, with the publication of several noteworthy works, that the evaluation and approach to the comatose patient were described. One of the first books dedicated to coma was by the neurologists Fazekas and Alman. Their book focused on the physiology of acute metabolic disturbances but lacked insight into the mechanisms of coma caused by mass effect (Fig. 1-30). As alluded to earlier, after O’Neilly and Plum published their observations of clinically deteriorating patients and outlined two brain herniation syndromes in 1962, a full textbook followed in 1966 (The Diagnosis of Stupor and Coma) (Fig. 1-31).⁶⁷ Subsequent editions (1972, 1982) expanded on outcome prediction and the use of CT scanning, followed by a fourth rewritten edition in 2007. After the publication of this now-classic work written with Posner, Plum became the primus inter pares of those studying comatose patients. The clinical material came from King County Hospital and the University of Washington Hospital in Seattle between 1953 and 1963. This book put the emphasis on patients with destructive lesions, supratentorial mass lesions, infratentorial mass lesions, and metabolic brain disorders. The development of clinical signs over time—the rostrocaudal patterns—characterized this book. The book broadly classified the metabolic causes and included meningitis, subarachnoid hemorrhage, and cerebral vasculitis as a metabolic encephalopathy. It cemented the distinction between “structural” and “toxic-metabolic” causes of coma and has remained a clinical guide for neurologists. The book was novel in producing a table of toxic and metabolic coma based on changes in respiration (hypoventilation or hyperventilation) and linking types of breathing abnormalities to certain locations in the brain and brainstem. The book also carefully details bedside methods of neurologic examination.

Soon thereafter, C. M. Fisher published an extraordinarily detailed paper, Neurological Examination of the Comatose Patient²⁷ (Fig. 1-32). Fisher emphasized that ocular signs are “the most important of the entire examination” in a comatose patient. After seeing 10 cases with clinical-pathological correlation, Fisher described ocular bobbing (an intermittent down and up conjugate movement of the eyes), which he correlated with pontine pathology. Other ocular signs were described for the first time, including “wrong-way eyes,” pontine miosis, ocular agitation, doll’s eyes, eye closure and blinking in coma, and reflex blepharospasm. Fisher pointed out that the eyelid tone or the length of time the eyes remain open after they are opened by the examiner are both findings that give an indication of the depth of the coma. New observations about the motor examination included bilateral decerebrate posturing resulting from acute lesions involving the supratentorial motor system. In a time before CT s