Cases - "Altered Mental Status"
The format of this chapter is different from the preceding ones. Case reports and their analyses to will be used to create a more realistic problem-solving environment. This approach gives a fuller understanding of each case problem and its ramifications.
For practical purposes, metabolic encephalopathy is defined as a potentially reversible abnormality of brain function caused by processes of extracerebral origin. These processes usually involve some metabolic upset (electrolytes, serum osmolarity, renal function or hepatic dysfunction), some deficiency (metabolic substrates, thyroid hormone, vitamin B12, etc), some toxic exposure (drugs, alcohol, medicines, etc) or systemic toxic states (sepsis, for example). Therefore, in metabolic encephalopathy, there is diffuse cerebral dysfunction. Of course, the degenerative diseases of the brain that result in dementia also produce diffuse dysfunction of the brain (see Chap. 16). Although both the degenerative disorders and metabolic encephalopathy must be considered in the differential diagnosis of diffuse cerebral dysfunction, dementing conditions progress slowly, while metabolic encephalopathy often presents rapidly and usually with fluctuating levels of alertness (attention and concentration). The primary cerebral degenerative disorders do make patients more susceptible to metabolic upsets and dementia and encephalopathy may coexist (complicating diagnosis).
Patients with metabolic encephalopathy may present in several ways depending on the magnitude and temporal course of the abnormality, the individual's age and neuronal reserve (i.e., the capacity to compensate for dysfunction). Ironically, patients with many causes of encephalopathy can look very similar, and it is usually not possible to determine the cause based on symptoms alone. The evaluation of the patient with encephalopathy intersects with the evaluation of stupor and coma, since they have many of the same causes (see Chap. 24) and often coexist. In addition to presenting with depressed consciousness, metabolic encephalopathy can present with progressive loss of intellect (dementia), hyperexcitable states such as agitated dementia (delirium) or seizures (multifocal and generalized myoclonus, generalized tonic-clonic seizures). Rarely, encephalopathy can present with focal abnormalities of brain function, entering into the differential diagnosis of stroke and slowly expanding masses (such as tumor or subdural hematoma).
In general, the presence of metabolic encephalopathy can be detected by physical examination, however, the etiology is usually not clear based on the exam. A reliable medical-neurologic history often delineates the diagnostic possibilities. Appropriate laboratory tests should then define the etiologic factor or factors.
Table 25-1, modified from Plum and Posner, is a comprehensive listing of the major processes capable of producing metabolic derangement of brain function. Analyses of examples of the various presentations of metabolic brain dysfunction follow. The plan is to give you a general approach to the diagnosis and management of metabolic encephalopathy.
A 54-year-old single man was brought to the emergency room by ambulance. He had been found unresponsive in his apartment. He was last seen 24-hours earlier by his friends at a social gathering, where his behavior was apparently normal. No medical or neurologic history was known.
The following were the initial findings:
- Mucous membranes were dusky and respirations were shallow at 10 per minute; within minutes the rate fell further. Nasotracheal intubation was achieved and respiration was supported mechanically.
- Blood pressure was 110/70, pulse 88 and regular, rectal temperature 36.0 C.
- Blood glucose was drawn and an intravenous started with 5% dextrose and water; 50 gm glucose was given by vein.
- There was no evidence of head trauma (i.e., there was no soft-tissue swelling or depression of the cranium and no contusion or lacerations over the face or scalp). There was no blood behind the tympanic membranes.
- The ocular fundus had flat disks, and venous pulsations were present.
- The patient's neck was supple to flexion.
- An indwelling urinary catheter was placed.
Neurologic evaluation was carried out simultaneously and revealed the following:
- No response to noxious stimulation (irritation of the nares with a cotton wisp and supraorbital pressure).
- No spontaneous respiration off the respirator for one minute.
- Pupils 1 mm in diameter that constrict consensually to bright light.
- Eyes in neutral position and immobile; no oculocephalic or cold caloric oculomotor-vestibular response.
- No reflexes or motor responses of any type observed or elicited.
Two diagnoses should be entertained for this dramatic picture. Acute transection of the pons and medulla by hemorrhage or infarction and, less likely, trauma could explain all the findings. Severe metabolic depression of hemisphere and brain stem (reticular formation) function is a much more likely diagnosis. Even if these conditions were equally common, one should approach the problem primarily from a therapeutic viewpoint and consider metabolic disorder the working diagnosis since acute transection is unlikely to be treatable.
Table 24-2 is an outline for the basic emergency room evaluation of patients with depressed consciousness of unknown cause. The emergency measures and laboratory studies are standard and straightforward. The neurologic examination, as formulated by Plum and Posner, is brief but comprehensive enough to tell the examiner whether the depression of consciousness has been caused by a process involving the contents of the supratentorial, infratentorial, or both compartments. Repeated examinations determine the evolution of the process. Last, the examination, when supplemented by selected laboratory studies, helps determine the specific diagnosis.
Let us further analyze our patient and determine why, on the basis of the neurologic examination alone, we should consider his condition typical of metabolic coma. The specific diagnostic possibilities can then be discussed and the laboratory procedures for making the final diagnosis reviewed.
Alone, stupor or coma (particularly the latter) provides very little to differentiate metabolic from structural involvement of the reticular formation and cerebral hemispheres.
Excessive sighing and yawning, Cheyne-Stokes respirations, hyperventilation, hypoventilation, and apnea are nonspecific presentations of bilateral brain stem involvement whether it be structural or metabolic. On the other hand, ataxic respirations and apneustic breaths are seen uncommonly, if at all, with metabolic disease; if present, they indicate structural disease of the pons or medulla (see chapter 1).
Our comatose patient, who has progressed from hypoventilation to apnea, must have bilateral involvement of the reticular formation of at least the mesencephalon or upper pons. Involvement of the diencephalic reticular formation alone would more likely be reflected as stupor, whereas involvement of the reticular formation of the lower pons and/or medulla in isolation would not cause depression of consciousness. The apnea reflects loss of function of the medullary reticular formation containing the respiratory center. Therefore, we have evidence for involvement of at least the midbrain or pons and the medulla. If the involvement of this extensive amount of brain stem tegmentum were destructive or compressive, we would expect to see evidence of, or involvement of cranial nerve systems that lie therein. Specifically in our examination we should look for abnormalities of the oculomotor, abducens, and vestibular systems. Metabolic depression first involves the reticular formation and consciousness. Cranial nerve and vital function (respiration and cardiovascular support) abnormalities tend to occur late. Let us proceed with the examination.
The pupils of our patient were small (1 mm) and reacted to light. There is very little constriction possible from a resting size this small; therefore, it is necessary to use a strong light and at times a magnifying glass (an otoscope lens is handy) to confirm reactivity. With the exception of a few rare conditions (e.g., the Argyll Robertson pupils of tertiary syphilis or advanced diabetes), 1 mm pupils react to light. The pupil in narcotic overdose is usually maximally constricted (less than 1 mm) and cannot constrict further. Occasionally the pupils are maximally constricted following transection of the pons because of loss of the descending sympathetic systems and presumably a disinhibition and possibly an active facilitation of the Edinger-Westphal complex in the midbrain.
As a rule the last major reflex system lost with metabolic encephalopathy is the light reflex. The parasympathetic system is quite resistant, whereas the sympathetic system, an integral part of the reticular activating system, is depressed early in parallel with depression of consciousness. Small pupils are the rule with metabolic involvements, with the exception of toxicity from drugs that excite the sympathetic system (e.g., amphetamine) or depress the parasympathetic system (e.g., atropine, meperidine, glutethimide).
A primary destructive or secondary compressive lesion transecting the diencephalon could cause the pupillary abnormality in our patient, but it is likely that he would be stuporous and not comatose and it is highly unlikely that respiration would be lost. This would have to be caused by a separate lesion involving the medulla and sparing the midbrain. Midbrain transection causes loss of both sympathetic and parasympathetic systems, thus leaving the pupils in midposition (approximately 4-7 mm) and unreactive to light. Rarely, intraventricular dissection from an intracerebral arterial hemorrhage causes a skip picture. The hemorrhagic mass compresses the diencephalon, depressing consciousness, while it is hypothesized that the intraventricular hemorrhagic pressure wave causes depression of medullary functions when it is reflected against the floor of the fourth ventricle after passing down the ventricular channel. The midbrain and pons might then be initially spared. Acute respiratory arrest from medullary dysfunction is the most likely cause of sudden death associated with intraventricular hemorrhage.
Transection of the pons and the medulla could cause the changes in respiration, consciousness, and pupillary function observed in our patient. It is reasonable to leave this unusual possibility for a diagnosis of exclusion and consider the diagnosis metabolic until proved otherwise. It is statistically much more likely and is usually amenable to therapy.
With metabolic encephalopathy the basic oculomotor-vestibular reflex system (see Chap. 6) first becomes disinhibited as cerebral hemispheric and reticular formation influences are lost. These disinhibited reflexes present as tonic eye deviation when caloric tests are performed or as movements of the eyes away from the direction of head movement (oculocephalic reflex). The saccadic or checking component of the oculocephalic response is lost as is the fast or checking component of nystagmus elicited by caloric irrigation is also lost (the fast component is dependent on cerebral cortical function). With severe depression of brainstem function, this relatively resistant brain stem reflex system may be suppressed and, eventually, there is no oculomotor-vestibular response. On occasion bilateral medial rectus function may be lost first, with preserved pupil function. In metabolic encephalopathy, total suppression of the oculomotor-vestibular system occurs at about the same time as respiratory suppression. However, there is some individual variation.
Our patient had no oculomotor response at all, which is compatible with destruction of the medulla or pons or severe metabolic suppression of brain stem function.
As a rule, the motor abnormalities of metabolic encephalopathy are symmetric. In progression from stupor through coma, diffuse paratonia (perseveration of motor tone) is seen first, followed by diffuse weakness, and ultimately flaccid quadriplegia as the condition worsens. Decorticate and decerebrate reflex posturings with noxious stimulation are commonly seen with progressive metabolic suppression of any type. An excellent model for observing these motor changes is the patient undergoing general anesthesia. For that matter, the anesthetized patient is an excellent model for observing all the changes so far mentioned.
In the patient with early stages of metabolic encephalopathy, there may be some motor signs that are relatively specific for metabolic encephalopathy. Tremulousness, asterixis (usually symmetric, irregularly episodic loss of maintained motor tone of the limbs, less commonly of the axial muscles), multifocal myoclonus (diffuse, asynchronous twitching of portions of muscles often strong enough to cause movement at the joints, of larger magnitude than fasciculations), and generalized myoclonic jerks that may or may not lead to major motor seizures should all alert the examiner to the presence of metabolic disorder.
Our patient was flaccid, areflexic, and unresponsive to noxious stimulation, which is compatible with one of two scenarios: the motor shock state produced by acute transection of the lower brain stem or severe metabolic encephalopathy with suppression of the brain stem. This flaccid, areflexic condition is not caused by a supratentorial lesion with secondary compression and rostrocaudal deterioration of brain stem function since the pupils were small and reacted to light.
Summary of neurologic findings in case #1
The neurologic examination showed a patient who was flaccid, areflexic, and unresponsive to noxious stimulation, who was apnic off the respirator, and who had no oculomotor-vestibular response. His pupils were small and reactive to bright light. Although his pons and medulla may have been transected by a destructive lesion, the statistically likely and therapeutically oriented working diagnosis is metabolic encephalopathy. The former diagnosis can wait until we have disproved metabolic disease. Even if the picture were more abnormal - with the blood pressure needing exogenous pressor support and the pupils midposition and unreactive - one would still assume severe metabolic depression until proved otherwise. This latter situation, if it was secondary to a known destructive process, would indicate brain death. Metabolic conditions can totally depress brain function and still be compatible with full recovery if appropriate physiologic support is provided. Therefore, it is critical to rule out toxic and metabolic conditions in any patient meeting criteria for brain death.
The evaluation of the patient with coma and likely metabolic encephalopathy involves the consideration of a wide array of possibilities while, at the same time, treating acute issues that can complicate or worsen the coma. What toxic and metabolic categories must be considered for our patient? Table 25-1 gives essentially all the possibilities that might cause brain depression of this degree. Reviewing quickly the major categories should permit us to narrow the possibilities for our patient and carry out the laboratory tests necessary to define a specific cause. Obviously, the sooner the cause is identified, the sooner corrective actions can be instituted. Some nonspecific measures may be initiated as a matter of routine. For example, respiratory and cardiovascular support is initiated even before the patient reaches the emergency department (by first responders). Also, a quick survey of the place in which the patient was found often turns up clues (medicines, pill bottles, signs of trauma, etc). In the ED, after drawing blood for testing, thiamine followed by glucose is often given as a nonspecific measure in consideration of hypoglycemia. Naloxone (Narcan) is often given to reverse possible narcotic overdose. At that point, the detective work begins (including discussions with witnesses, acquaintances or family members). We will consider the thought process using this case as an example.
Basic support systems
These include oxygen, substrate, and cofactors.
Cardiorespiratory insufficiency caused by acute cardiac arrhythmia with spontaneous reversal might have occurred. Progressive subsequent deterioration with no good evidence for midbrain involvement (i.e., pupils remain reactive and small) to implicate rostrocaudal deterioration and good evidence (normal disks and venous pulsations) against progressive postanoxic cerebral swelling essentially rule out this possibility. Blood gases drawn on the respirator with room air revealed normal pH, Po2, and Pco2. An electrocardiogram showed only mild, nonspecific abnormalities and a regular rhythm.
Hypoglycemia seems an unlikely possibility; there is no history of insulin-dependent diabetes (overdose of insulin is a common cause). It must be remembered that it is quite possible to become severely hypoglycemic during prolonged alcohol drinking (days) with no intake of carbohydrates. Glucose and finally glycogen are depleted, gluconeogenesis from protein-amino acid sources is blocked by ethanol, and the blood glucose progressively drops and may reach low enough levels to cause encephalopathy. An insulin-secreting tumor is also a rare cause of severe hypoglycemia.
Hyperglycemic, nonketotic coma is usually preceded by days of progressive deterioration but must be considered a possibility as well as late-stage, hyperglycemic, ketotic coma.
A blood glucose determination was carried out on blood drawn prior to the infusion of glucose. A dip-stick test done within minutes showed a level of approximately 200 mgm%. The laboratory value, available one hour later, was 180 mg%. The normal arterial blood pH ruled out significant ketosis along with other major causes of metabolic acidosis severe enough to cause depression of consciousness (lactic acidosis, uremia, and exogenous acid poisoning).
We have been unable to determine the alcohol intake history of our patient. Wernicke's encephalopathy, a condition caused by thiamine deficiency and usually seen in chronic malnourished alcoholics, could cause this level of depression but seems unlikely because no obvious illness was noted by friends at a cocktail party 24 hours before.
Nevertheless, a multivitamin preparation was added to the glucose infusion because this history was inadequate, and because glucose infusion can precipitate Wernicke's encephalopathy in alcoholics or other malnourished individuals with borderline thiamine stores. The glucose initiates Krebs cycle activity, which places high demand on residual thiamine stores and deprives the brain of thiamine-dependent metabolic systems.
Diseases of organs other than the brain
Significant renal and pulmonary failure has been ruled out by pH and blood gas determinations. Hepatic failure with secondary shunting of large-bowel-produced toxins into the systemic circulation must be considered. In a child, Reye's syndrome (parainfectious hepatic insufficiency and secondary brain failure) might appear this rapidly. Also, fulminant hepatic necrosis (toxic or infectious) could result in acute cerebral depression. Hepatic necrosis is unlikely to occur without significant jaundice, and 24 hours from alertness to this deep coma would seem unlikely. The condition of a patient with chronic cirrhotic liver disease and portal hypertension with significant portal-systemic shunting could rapidly deteriorate under certain stresses. A common scenario in cirrhosis is acute deterioration due to hemorrhage into the bowel (e.g., from varices or acute gastritis), causing a massive increase in metabolic toxins (including ammonia) due to the action of large-bowel bacteria on blood products (particularly protein).
No stigmata of chronic liver disease were seen, there was no jaundice, and the liver was not palpably enlarged or small to percussion. No stool blood was detected, a nasogastric tube aspirate showed only mucus, and the hematocrit was reported to be normal.
A clue to the presence of chronic portacaval shunting and encephalopathy in less severely depressed patients is respiratory alkalosis secondary to chronic mild hyperventilation. Elevation of the level of blood ammonia or cerebrospinal fluid glutamine (a reflection of elevated amounts of blood ammonia), are usually present to confirm the diagnosis.
Other than abnormalities of glucose metabolism, endocrine disease is unlikely to present as acute coma. Our patient's hypothermia (36 degrees C) was not low enough to consider deep coma caused by hypothyroidism. Also, a long history of progressive dementing deterioration would be likely. Mild hypothermia is nonspecific for deep-coma states. With lesser depression of brain stem function, this degree of hypothermia is probably most commonly seen with sedative intoxication, particularly the barbiturates.
There are other, rare considerations. The remote effects of cancer cause a chronic, occasionally subacute depression of brain function. Porphyria is unlikely to present primarily as depression of consciousness unless, as with other acute neuromuscular disorders (e.g., Guillain-Barre syndrome, occasionally myasthenia gravis), respiratory failure supervenes to cause hypoxia. These categories are unlikely in our patient, who did not lose independent respirations until after arrival in the emergency room and after respiratory support was established.
Abnormalities of ionic and acid-base environment of the brain
Significant abnormalities of the acid-base environment have been ruled out by pH determination. Relative hyponatremia secondary to excess water accumulation from excess intake of water or inappropriate secretion of antidiuretic hormone is the only ionic abnormality likely to cause this rapid and severe neurologic deterioration. Potassium abnormalities will effect cardiac and muscular function before they would affect the brain. Phosphorus deficiency would affect muscle function (strength) before it would impact the brain. Elevated serum calcium can suppress the brain (as can elevated magnesium levels), however, this is usually slowly evolving in the patient with serious endocrine or neoplastic disease. However, all of these were essentially ruled out since electrolyte values were available within an hour from admission and were all normal.
Infectious disease and subarachnoid hemorrhage
Meningitis or encephalitis could cause depression of brain stem function this severe. One might expect fever and possibly some nuchal rigidity; however, by the time these processes cause deep coma, both the fever and nuchal rigidity may be lost. Encephalitis would be the less likely of the two possibilities because severe cerebral swelling would be expected to have compressed the midbrain and caused pupil enlargement and loss of the light reflex. The presence of normal venous pulsations and flat optic disks, reflecting normal intracranial pressure, also rules strongly against this possibility. Coma due to bacterial meningitis would also be unlikely for these reasons, although rarely bacterial toxin could suppress brain function without causing major cerebral swelling or acute communicating hydrocephalus.
Subarachnoid hemorrhage may cause depression of consciousness to the level of stupor, but unless it dissects into the cerebral hemispheres and causes secondary compression of the brain stem, deep coma is unlikely in the acute stage. Evidence for increased intracranial pressure would be expected. Involvement of the midbrain (midposition and fixed pupils) would have occurred as part of the rostrocaudal process with the exception of the rare intraventricular hemorrhage phenomenon described earlier.
To rule out the distant possibility of bacterial meningitis, and with the security of a normal CT scan, a lumbar puncture was performed. The opening pressure was 140 mm cerebrospinal fluid (CSF). The fluid was clear and colorless. Three red cells were present in the first collection tube, none in the third; no white blood cells were present. No organisms were seen on gram stain (individuals with alcohol in their blood or who have for other reasons a poor white blood cell response to bacterial invasion may have a delayed CSF pleocytosis, but bacteria should be demonstrable on gram stain). The level of CSF glucose was 100 mg% and the protein level was 45 mg% (both normal). Rarely, LP can be done without cranial imaging if the fundi are normal and particularly if there are venous pulsations detected. This is usually only in the case of high suspicion of meningitis and no prospect of obtaining a cranial image in a timely fashion.
Disordered temperature regulation
This is unsupported by findings.
Postseizure (postictal) depression
This is unlikely because the level of depression of brain function is too severe, and because it has progressively worsened in the absence of apparent seizure activity. However, unchecked major motor status epilepticus could result in this severity of brain dysfunction because of a mixture of associated metabolic dysfunctions. Hyperthermia, hypoxia, lactic acidosis, and prolonged excessive neuronal activity probably all contribute to brain depression and damage caused by generalized motor status. If our patient's seizures had arrested spontaneously some hours before he was discovered, these metabolic abnormalities might have been cleared. The lack of seizure history, although sedative (ethanol) withdrawal must be kept in mind, and the lack of evidence of expected brain edema make this possibility unlikely.
In a busy hospital emergency room, this is the most common diagnosis for patients admitted with coma of unknown etiology. The most common single subgroups are the sedative drugs (such as benzodiazepines and barbiturates) and aspirin.
For the various reasons given earlier, this major category would be strongly suspected by exclusion in our patient and a serum drug screen carried out. The general serum barbiturate level was determined and found to be in the high toxic range.
Management of the patient with metabolic coma should support vital systems -- respiratory and cardiovascular -- and, if possible and necessary, remove the metabolic abnormality or allow for its natural dissolution as with sedative excess. One of the major complications of prolonged coma is pneumonitis. It can usually be avoided by diligent pulmonary toilet. Another complication to be avoided is deep vein thrombosis and pulmonary embolism, and therefore pressure stocking should be used routinely and low dose heparin considered if not medically contraindicated.
The patient in this case made a full recovery, subsequently admitting to a reactive depression during which he had taken an overdose of pentobarbital, a relatively short-acting barbiturate. Hemodialysis is rarely necessary for treating overdoses of short- and even long-acting barbiturates. If careful physiologic support is continually available, the mortality from barbiturate or other sedative overdose should be less than 1%. With long-acting sedatives, the coma may be so prolonged (many days) that physiologic support is strained to the utmost, and then dialysis can shorten the period of critical care. The complications of hemodialysis appear minor enough now to warrant this aggressive approach.
A 60-year-old woman in an agitated, confused state was brought to the hospital by her husband. She had been ill with fever, chills, and muscle aches and pains for about a week and confined to bed, treated symptomatically with aspirin for "influenza". Her husband reported her to have been rational until early in the morning of admission, when she awakened and appeared agitated and frightened. She resisted her husband's attempts to take her to the hospital. There was no past history of significant medical illness. She had been taking no regular medications, although her husband claimed she occasionally took chlordiazepoxide (Librium) for insomnia. She smoked one pack of cigarettes a day and was considered a social drinker, consuming one or two drinks of wine or beer per day in the company of others.
Hospital findings and initial management:
- Blood pressure was 160/90; pulse was 110 per minute and regular; temperature was 38.5 C; respirations were 20 per minute and regular.
- General physical examination was resisted by the patient, who insisted on leaving, more than once walking toward the door wearing only her nightgown. She was disheveled, agitated, and frightened. As best could be judged there were no stigmata of chronic illness. Skin and mucous membrane color was normal. Nuchal mobility was irregularly limited in all ranges in keeping with diffusely present paratonia. No grimacing or Brudzinki's sign was elicited on neck flexion to suggest meningeal irritation.
- Neurologic examination revealed her to be confused about where she was, the time of day, and the day of the week. She was unable to remember the examiner's name and did not recognize him when he returned after stepping out of the room for several minutes. Essentially she was unable to imprint new information. Her attention span was shortened, reversals (e.g., saying days of week in reverse order or spelling WORLD in reverse) were impossible, and she was unable to interpret proverbs at all. She appeared to be episodically actively hallucinating, pointing to and identifying moving objects that were not present. Physical signs of diffuse cerebral dysfunction included the presence of diffuse paratonia (gegenhalten), disinhibited regressive feeding reflexes (snouting, sucking and rooting), bilaterally disinhibited palmomental responses, loss of checking on oculocephalic testing, and a bilateral forced-grasp response on palmar stimulation. The upper extremities were tremulous when outstretched, and asterixis was noted at the wrists when extended. The deep-tendon reflexes were symmetrically 3+. No other abnormalities were elicited.
Acute and agitated diffuse loss of cognitive function (delirium) is almost always a reflection of toxic, metabolic or infectious disorder. This presentation and also acute quiet or apathetic encephalopathy may be caused by most of the categories listed in Table 25-1 and is most likely to occur in the early stages of acute metabolic dysfunction. Frequently periods of agitation alternate with periods of apathy and depression of consciousness, ultimately culminating in stupor or coma if the dysfunction magnifies. Sedative (including alcohol, barbiturates, benzodiazepines, etc.) withdrawal states, hypoglycemia, hypoxia, hypermetabolic states including sepsis and hyperthyroidism, hypocalcemia, hypomagnesemia, and hepatic failure appear to be more commonly associated with a hyperexcitable state than other conditions. The excited stage of general anesthesia is a fair model of agitated dementia.
Underlying progressive degenerative brain processes such as Alzheimer's disease can emerge acutely as a florid dementia, agitated or quiet, under the influence of only mild metabolic stress, which alone is not enough to cause cognitive or emotional defect in a normal brain. This is true also for the aged brain, which under normal stress is able to function adequately. The 80-year-old with a temperature of 39 C is much more likely to be confused than the young adult whose neuronal safety factor (volume of neuronal processes and numbers of active neurons) is greater. Elderly patients admitted to the hospital on medical or surgical services are frequently seen in consultation for defective cognitive function. CT scans and MRIs reveal atrophic changes that are compatible with the patients' age but have not been well correlated with studies of cognitive function. Mild metabolic dysfunctions are frequently found such as minimal renal dysfunction, anemia, low-grade fever, mild sedative intoxication (e.g., from tranquilizers and sleeping medications), mild hypoxia from chronic pulmonary disease, and congestive heart failure, usually in various combinations. Correction of the disorder or disorders frequently leads to an improvement in cognitive function.
We have several clues from the history and examination of our patient. She has had a febrile illness and continues to be febrile. A temperature of 38.5 C is inadequate alone to cause clinically obvious cerebral dysfunction. Central nervous system infection, encephalitis or meningitis, could produce this picture. The absence of Brudzinski's sign or behavioral evidence for pain on neck flexion does not support meningeal irritation but does not rule it out either. A lumbar puncture should be done when there is: (1) suspicion of infection or (2) suspicion of subarachnoid hemorrhage in a patient with no focal neurologic defect. Lumbar puncture in our patient revealed normal pressure, no cells, normal protein and glucose levels, and normal results from gram stain and India ink preparations. Although the absence of elevated white cell count in the CSF does not entirely rule out meningitis or encephalitis, it does make CNS infection extremely unlikely, particularly in the face of normal pressure and normal CSF protein and glucose levels. Routine studies of blood revealed a white blood cell count of 15,000 with a shift to the left -- i.e., predominance of neutrophils and their precursors. Further fever workup included urine evaluation and culture, blood cultures, and chest x-ray, all of which showed normal results. Routine blood chemistries including glucose, electrolytes, BUN, Ca2+, Mg2+, liver function tests, and a screen for sedative drugs were normal.
The patient's husband was questioned closely concerning her use of chlordiazepoxide. The possibility that she could be physiologically dependent was emphasized, with the present picture being one of sedative withdrawal delirium (delirium tremens). He had been ministering to her illness during the week and was sure she had not had any sleeping medication. On our instruction he returned home and on careful searching found three partly empty bottles of 25-mg chlordiazepoxide capsules secreted in various parts of the house. They were prescribed by three different physicians; a general practitioner that the husband did not know, the patient's gynecologist, and their family physician. Each prescription had been filled at a different drugstore. This circumstantially confirmed the suspicion of withdrawal delirium, a condition that appears to be rebound hyperexcitability in a nervous system that has compensated for chronic sedative depressant effects (typically drugs or alcohol).
Treatment consisted of sedative replacement; in this case, diazepam was given intravenously until it had a calming effect. Gradual withdrawal was then undertaken, decreasing the drug dose by 10% each day. An uneventful recovery occurred, and subsequently the patient described how she had gradually become dependent on chlordiazepoxide. It is of some interest and instructive that the delirium occurred approximately one week following withdrawal. Sedative withdrawal states, which include hyperirritability, seizures, hallucinations, and frank delirium tremens, usually appear within 96 hours but sometimes delirium may emerge as late as 14 days after terminating or diminishing intake.
A 65-year-old man was referred to the neurology clinic for evaluation of his progressive loss of cognitive ability. His wife had noted forgetfulness, carelessness in dress and daily hygiene, and an increasing lack of spontaneity for at least six months.
Examination revealed evidence of diffuse cerebral dysfunction with difficulties in all aspects of cognitive function, and there was a short-term (recent) memory deficit with preservation of remote memory and out of proportion to other cognitive defects. This is rather nonspecific because both degenerative and metabolic processes may affect the memory areas of the medial temporal lobe more severely than other neocortical regions. Multiple regressive reflexes were easily demonstrated including forced grasping, disinhibited oculocephalic responses, snouting and sucking responses to labial stimulation, and a disinhibited glabellar blinking response. Diffuse mild paratonia was also present. No asterixis or myoclonus was present.
Approximately 80-85% of persons with progressive dementia without focal signs are shown to have a degenerative process, most often Alzheimer's disease. Unfortunately, none of these degenerative conditions can be stopped (although there are some treatments that should be considered, see Chapter 16). The most important initial consideration is to identify the 15-20% who have potentially treatable disorders. There are some clues that can help identify these treatable conditions and with the aid of laboratory screening most can be ferreted out. If the patient is younger than 50, the chances of having a treatable disorder are greater. If the disorder is of recent onset and rapidly progressive and if there are frequent fluctuations in the level of function, chances are higher that a treatable disorder is present. If motor stigmata such as tremulousness, asterixis, and multifocal myoclonus are present, it is very likely that a metabolic problem is, at least in part, responsible for the loss of cognitive function. Focal sensory or motor signs are not characteristic of metabolic disorders, and if they are present with the signs of diffuse dysfunction, they should suggest the possibility of a mass lesion, which may be amenable to therapy.
Although our main purpose is to discuss metabolic disorders that cause brain dysfunction, it is hardly realistic or practical to isolate only metabolic causes of diffuse cerebral dysfunction (dementia) from the other treatable causes. Table 16-3 lists the major causes of dementia. An asterisk marks the treatable categories, which are numerous. Some diagnostic clues and approaches to therapy are also listed.
Our patient appeared to have had a recent onset of his dementia, which leads one to believe there is a fair chance of finding a treatable disorder. However, it is reasonable to equivocate about the time of onset of his dementia because family members may not notice a slowly dementing process until some obviously unacceptable disinhibited social behavior or blatant confusion occurs. The patient was 65 years old at the time of evaluation. He had taken early retirement at age 62 and had been relatively inactive with few hobbies or outside interests for the past three years. Business or work associates are more likely to notice dementia in a compatriot because it may affect productivity early (although a fairly routinized job may be accomplished effectively by a person well into the dementing process). The inactivity of retirement can easily veil deteriorating brain function, and therefore it is reasonable to suspect that deterioration of our patient's mental capacity may have been occurring for as long as three years. Even if this were true, a screening battery of tests should be conducted to identify any treatable disorder behind the decline. The specific tests should be guided by the history and examination, but a comprehensive battery should include at least the following:
|TEST||TO RULE OUT|
|Complete blood count (CBC)||General|
|Erythrocyte sedimentation rate (ESR)||Vasculitis (collagen disease), infection, occult tumor|
|BUN, creatinine||Renal insufficiency|
|Serum calcium||Hyper-or hypocalcemia|
|Liver function tests||Hepatic insufficiency; metastatic tumor|
|Thyroid function tests||Hyper- or hypothyroidism|
|Serum B12 level and/or Schilling test||B12 deficiency|
|Serum folate||Nutritional deficiency|
|Lumbar puncture||Increased number of cells (hemorrhage, infection), high level of protein (tumor, etc.), low level of sugar (infection, etc.), increased pressure (tumor, etc.), + VDRL (syphilis)|
|Chest x-ray||Lung tumor or other metastasized malignancy, diffuse pulmonary disease|
|CT scan or (better) MRI (if tolerated)||Mass, hydrocephalus, atrophy, infarctions, bilateral subdural hematoma|
|EEG||Metabolic encephalopathy; focal lesions; seizure activity; EEG of metabolic encephalopathy is liable to have diffuse slowing in the delta-theta frequencies, while degenerative disease may cause little (some theta slowing diffusely) or no abnormality|
Frequent runs of diffuse delta activity were seen on our patient's EEG, making a metabolic disorder highly suspect. A CT scan showed dilatation of the lateral, third and probably also the fourth ventricles; no cortical atrophy was evident. A lumbar puncture showed an opening pressure of 160 mm CSF, 10 lymphocytes, a protein of 130, and a glucose of 40 with a concomitant fasting blood glucose count of 100. Although an India ink preparation showed no cryptococci and the organisms did not grow on CSF culturing, cryptococcal antigen was detected in the spinal fluid. A course of treatment with amphotericin B was given, and within three weeks the spinal fluid protein and glucose levels had returned to normal and no cells were detected. Some improvement in the patient's dementia had occurred after two months, but a repeat CT scan showed a continued hydrocephalus. A ventriculoperitoneal shunt was placed and further improvement in mentation followed.
A 62-year-old man was brought to the emergency room on January 3 by ambulance, accompanied by his wife. He had been noted to be weak and uncommunicative that morning. Preliminary examination by emergency room personnel revealed a right hemiparesis and aphasia. The neurology consultant was called to evaluate the presumed stroke. The findings were confirmed. The patient was severely dysphasic; his communication was limited to a few perseverated and unintelligible jargon words. He appeared to have a right-sided hemianopsia to threat, responded poorly to noxious stimulation of the right extremities, and had a right-sided hemiparesis including the right lower facial muscles. The right deep-tendon reflexes were hyperactive and a Babinski response was present on the right.
General examination revealed an otherwise healthy-appearing individual. No evidence of arterial disease was present. A sweetish odor of the breath was present and this prompted the examiner to think of the possibility of metabolic disorder, specifically hyperglycemia with ketosis. The patient's wife volunteered that the previous January a similar event had occurred under the same circumstances, partying over the New Year. He had been found "blacked out", seen at another hospital, "sugared up," and sent home no worse for wear. He had not been nor was he being treated chronically for any illness. Within ten minutes after the patient was first seen, a dip-stick analysis of blood glucose was available. A blood specimen was sent to the laboratory for more accurate evaluation. The finger-stick value for glucose was less than 40mg%. Fifty grams of glucose were given intravenously, and within thirty minutes the patient was sitting up in bed conversing normally with only minimal residual evidence of right hemiweakness, which subsequently cleared. A laboratory glucose level was available within an hour and was 24 mg%.
It transpired that the patient was a once-a-year heavy drinker, and that this and the previous year he had avoided eating anything over the three days that he was continually drinking hard liquor. As noted, progressive depletion of glucose and glycogen stores would be expected and symptomatic hypoglycemia would ultimately occur because gluconeogenesis from amino acids is blocked by ethanol.
Focal signs, suggesting destructive lesions of the brain, are unusual neurologic manifestations of metabolic disease. Hypoglycemia, hepatic encephalopathy, and hypoxia are the most frequently diagnosed conditions producing this confusing picture. Because hypoglycemia is easily treated early, because it can cause permanent neurologic defects if it is prolonged, and because it can present as coma, stupor, delirium, dementia, generalized and rarely focal seizures, and focal ablative disease, it is reasonable to demand that a blood glucose determination be made early in the evaluation of patients with these symptoms. In patients with coma, stupor, or delirium of unknown etiology, it is reasonable to give 25-50 gm glucose intravenously after drawing blood for glucose evaluation. In patients having a stroke picture or seizures, early determination of the glucose level would appear adequate. If a finger-stick determination shows hypoglycemia, 50 gm glucose should be given.
Why do patients with metabolic disorders develop focal loss of function? Some have underlying structural disease such as old areas of infarction or contusion, borderline adequate focal blood supply, or neoplasms. It is presumed that neurons that are defective but functional within or surrounding these structural lesions are depressed early by metabolic stress.
There are some patients, however, in whom there is no evidence of an underlying focal lesion and, particularly with hypoglycemia and hepatic encephalopathy, the focal abnormalities may shift during the same or other episodes. If shifting does occur, it makes a metabolic problem seem more likely. No good explanation for this asymmetric metabolic suppression of brain function is available.
A 56-year-old man who was previously well was admitted to the hospital in a stupor with clonic twitching of his left extremities and face. He had been feeling ill for several days and his wife had noted him to be lethargic on the morning of admission. At about 8 a.m., when she went to check on him, she noted that his left side was intermittently twitching.
Examination in the emergency room showed respirations were
20 per minute and regular, temperature was 38.5 C, blood pressure 150/90,
and pulse was 90 and regular. He was stuporous with appropriate withdrawal
responses on the right, and he had continuous clonic twitchings of the left
extremities and face, which were synchronous. No other pertinent findings
were noted on the remainder of the neurologic examination. On general examination
the only feature of note was dryness of the mucous membranes, decreased skin
turgor, and myoedema (a bulging of subcutaneous and muscle tissue on percussion),
suggesting dehydration. The major laboratory finding was hyperglycemia (blood
glucose 1,100 mg%) in the absence of ketosis.
In the absence of a history of diabetes mellitus, which is typical for persons developing nonketotic hyperglycemic coma, it was elected to treat him with small doses of insulin and fluid replacement with careful attention to electrolyte levels. Some patients like this are found to be extremely sensitive to insulin and others fairly resistant. Therefore it is important to begin therapy with small doses of insulin to avoid a precipitous drop in the blood glucose level, which may result in rebound cerebral edema. This complication probably occurs because of a buildup of polyols (sorbitol, fructose, etc.) or other osmotically active substances intracellularly in the brain in the presence of prolonged hyperglycemia. When the amount of blood glucose is rapidly dropped, causing decreased serum osmolality, the polyols are only slowly metabolized and do not readily diffuse into the extracellular or intravascular space. This creates a relative intracellular hyperosmolality and water is imbibed by the cells, causing edema and dysfunction. The glucose level in our patient was gradually reduced over two days. When the level reached approximately 500 mg% at the end of the first day, the focal seizures ceased and the stupor cleared. Mild left-sided hemiparesis and confusion persisted for some hours after the seizures stopped. Twenty-five mg diazepam stopped the seizures only transiently when given intravenously in divided doses. Diphenylhydantoin was given intravenously in a dose of 750 mg over the first day and had no obvious effect on the seizures. However, no bilateral generalization of the seizures occurred. No maintenance anticonvulsant was given. It is typical for focal status epilepticus of metabolic origin to be poorly responsive to anticonvulsant medication. As rule, the best therapy is to clear the metabolic abnormality if possible.
Focal status epilepticus (epilepsia partialis continua) has been associated with multiple metabolic abnormalities. Nonketotic hyperglycemic hyperosmolar coma is one of the more common underlying disorders, but we have seen it with uremia, hepatic encephalopathy, hypoxia, hypoglycemia, hyponatremia, encephalitis, meningitis, and theophylline excess, and it has been reported associated with others. It appears to be a rather nonspecific and unusual manifestation of metabolic disorder. It has also been associated with structural lesions, particularly infarction, neoplasms, and other masses unaccompanied by systemic metabolic abnormality.
Irritative dysfunctions (such as focal seizures) appear in some cases to be associated with minor subclinical cerebral lesions such as old areas of encephalomalacia (ancient infarcts or hemorrhages) and neoplasms. Presumably damaged but viable neurons surrounding such lesions have a relatively low threshold for seizure activity, and this threshold is exceeded by the metabolic stressor. In other cases, even the best examination has shown no obvious structural defect to account for the focal seizures.
In summary, we have given case examples of the various cerebral manifestations of metabolic disorders. Depression of consciousness, progressive loss of intellect, hyperexcitable states such as delirium and seizures, and less often focal abnormalities of brain function may be the behavioral presentations of metabolic disorders either alone or in combination, and these patterns may coexist in a single patient or may progress from one to another state in a single patient.
- Plum, F., Posner, J.B.: Diagnosis of Stupor and Coma, ed. 3. Philadelphia, F.A. Davis Co., 1980.