Wernicke's Encephalopathy (presenting with optic neuropathy)
In this patient, the combination of encephalopathy, ataxia, and ophthalmoparesis, together with suspicion for nutritional deficiency (thin body habitus, recent significant weight loss) and evidence for autonomic dysfunction (resting tachycardia), suggested the possibility of a Wernicke's syndrome. The patient's optic neuropathy, however, occasioned additional diagnostic studies that were performed while thiamine was being administered. An MRI scan of the head demonstrated bilateral and symmetric hyperintensities on T2-weighted sequences in the medial thalami and along the border of the third ventricle, but signal abnormalities could not be convincingly demonstrated in the optic nerves. In screens for demyelinating disease, neoplasms, and infectious-immune disorders, we did not find an alternative cause of this patient's optic neuropathy.
On hospital Day 2, the patient's sister was contacted, who provided additional information. In contrast to the history obtained on admission, the patient in fact had a long history of alcohol dependence, with heavy alcohol consumption for the past 20 years. She had been functioning independently and was still working part-time until the previous year. She had recently been drinking one-half to one gallon of wine per day, not getting out of her home, and not taking care of herself. Her sister, who talked to the patient daily by telephone, observed that the patient was often confused and incoherent during the past week. When her sister visited the patient last week, she found medications all over the house, and generally the house was in a very poorly kept condition. The patient had hired assistance recently to manage her finances. Her sister also felt that the patient's personality had changed prior to her present illness, becoming more irritable and impulsive following her orbital fracture in 1992.
The morning after thiamine (100 mg daily) was initiated, the patient's visual acuity improved to 20/40 OU, her right afferent pupillary defect had resolved, and she displayed only mild bilateral abducens palsies. Within two days of starting thiamine, her visual acuity had normalized on the right. Her ocular movements were full, but she displayed horizontal nystagmus with leftward and rightward gaze, in the direction of the gaze. Within four days of starting thiamine, the patient had only trace nystagmus with horizontal gaze, while her MMSE score had improved to 27/30. A follow-up MRI scan demonstrated significant resolution of the T2-weighted signal abnormalities present in the medial thalami on the initial imaging study. This dramatic response to thiamine administration confirmed the clinical suspicion of Wernicke's encephalopathy, and also provided an explanation for her optic neuropathy.
At the time of discharge, the patient's MMSE score was 28/30 (-2 on recall), though she still showed poor insight into her condition, and difficulty with following through with treatment plans. A consultant psychiatrist felt that at the time of discharge, the patient did not have the capacity to make informed decisions regarding her care. Her sister obtained temporary guardianship, and the patient was discharged to a care facility near her sister's home.
A complete review of Wernicke's encephalopathy and the neurological manifestations of thiamine deficiency is beyond the scope of this discussion, which will concentrate mainly on the relatively uncommon feature of optic neuropathy and visual loss. Many general reviews of Wernicke's encephalopathy exist, including a comprehensive textbook on the subject by Victor, Adams and Collins (1), from which much of this discussion is referenced.
Briefly, Wernicke's encephalopathy may be defined as a neurologic disorder of abrupt or gradual onset, characterized by nystagmus, ocular motor palsies, unsteadiness of stance and gait, and a confusional-apathetic state (1). These symptoms may occur in isolation, but more often they occur in various combinations. However, only a minority of patients will present with the complete, classic "Wernicke's triad" of nystagmus, ataxia and confusion. In the large series by Victor et al., 66% of patients with Wernicke's syndrome had confusion as their presenting symptom, while 51% had gait imbalance and 40% had ocular symptoms. Less than half of all patients had a constellation of symptoms upon presentation. Of these, less than 30% had all three symptoms of the triad. Thus, lack of one of these symptoms should not lower one's clinical suspicion for Wernicke's encephalopathy.
Korsakoff psychosis, or Korsakoff syndrome, refers to an abnormality of mentation in which memory and learning are affected out of proportion to other cognitive functions, in an otherwise alert and responsive patient (1). Most patients with Wernicke's encephalopathy who survive and are treated will be left with a Korsakoff amnesic state, while most Korsakoff patients have initially had evidence of Wernicke's encephalopathy. Thus, both entities are commonly regarded as two facets of the same disease process, resulting from thiamine deficiency.
Thiamine (vitamin B1), when combined with phosphoric acid, constitutes thiamine pyrophosphate (TPP). TPP, the physiologically active form of thiamine, has a wide range of activity in intermediary carbohydrate metabolism. It acts as a coenzyme for pyruvate dehydrogenase (PDH) and alpha-ketoglutarate dehydrogenase (alpha-KGDH), which are responsible for the decarboxylation of pyruvate and alpha-ketoglutarate, respectively. In addition, TPP functions as a coenzyme for transketolase, an enzyme of the NADPH-producing hexose monophosphate shunt. Because of its central role in these processes subserving carbohydrate metabolism, the requirement for thiamine is greatest when carbohydrate is the principal source of energy, such as alcoholic patients or patients receiving large quantities of parenteral glucose. The brain, which relies exclusively on carbohydrates for immediate energy demands, is particularly vulnerable to thiamine deficiency.
Autopsy studies of patients with Wernicke's encephalopathy show symmetrical lesions in the paraventricular regions of the thalamus and hypothalamus, mammillary bodies, periaqueductal region of the midbrain and superior cerebellar vermis. The mammillary bodies are most consistently affected (2). There are varying degrees of necrosis. Myelinated fibers appear to be more affected than neuronal cell bodies, though essentially all affected structures are gray matter. There is also astrocytic, microglial and endothelial proliferation. Discrete hemorrhages are found in only 20% of cases (1), contrary to traditional views that Wernicke's encephalopathy represented a state of hemorrhagic necrosis.
The selective vulnerability of certain structures has no clear explanation (1), but does account for the specific clinical manifestations of Wernicke's encephalopathy, such as gait ataxia. Ocular palsies are attributable to lesions of the 6th and 3rd nerve nuclei, while nystagmus appears to derive from lesions in the vestibular nuclei. The lack of significant destruction of nerve cells in these lesions, at least initially, may contribute to the rapid improvement of symptoms with thiamine replacement (2). Hypothermia, found in 1% of a large series (1), likely results from lesions in the posterior nuclei of the hypothalamus. Other autonomic manifestations may also occur. While the global confusional state cannot be related to a discrete brain lesion (1), the chronic amnesic state of Korsakoff's syndrome appears to require lesions of the medial thalami. In autopsy cases of patients who died in the chronic stages of the disease, the only important difference in those patients with amnesia was involvement of the medial dorsal nuclei of the thalamus. These areas were not involved in those without amnesic symptoms (2). Mammillary bodies were affected in all chronic patients.
The diagnosis of Wernicke's encephalopathy is still a clinical diagnosis, without specific laboratory or ancillary markers. However, the selective vulnerability of the midline gray matter areas and their symmetric involvement allows neuroimaging to play an important role. In a study of 45 patients (15 with acute Wernicke's encephalopathy, 15 asymptomatic alcoholics, and 15 control subjects), CT and MRI of the brain were obtained within two days of admission and evaluated. On CT scans, 13% of patients with Wernicke's encephalopathy (and none of the asymptomatic alcoholics) showed low-density abnormalities in the medial thalamus. MR imaging revealed increased T2 signal of this area (as in our patient) in seven patients with Wernicke's encephalopathy and increased T2 signal in the periaqueductal area of the midbrain in six patients. These MRI findings were noted in one asymptomatic alcoholic. The sensitivity of MR imaging was found to be 53% with a specificity of 93% (3).
Other studies have shown higher sensitivity, but in a smaller group of patients (4). Areas of T2 signal abnormality may also show contrast enhancement acutely (5). Both enhancing and non-enhancing lesions may improve or resolve with thiamine replacement, as in our patient (3-6). Thus, T2 or FLAIR MRI can be helpful in the diagnosis, especially when only some of the typical clinical features are apparent, or when adequate history is unavailable, such as in a comatose patient.
MRI may serve another purpose in clarifying the cause of illness of a patient with a chronic amnesic state (7). Mammillary body atrophy is an irreversible marker of chronic Wernicke's encephalopathy and is best assessed on sagittal or coronal MRI (6). However, Antunez's study showed no significant difference in the MRI prevalence of mammillary body atrophy between asymptomatic alcoholics and patients with Wernicke's encephalopathy (3). Atrophy of the diencephalic or mesencephalic structures, characterized by dilatation of third ventricle or aqueduct, may also point to a contribution of thiamine deficiency in the patient with chronic amnesia (7).
Regardless of initial diagnostic uncertainty, we must emphasize that the best diagnostic measure in acute Wernicke's encephalopathy is responsiveness to a trial of thiamine. Treatment should be initiated emergently upon suspicion, and not delayed by investigation of alternative hypotheses or atypical features. Improvement in ocular symptoms of Wernicke's encephalopathy is universal and rapid, often within hours. Failure of the ocular palsies to respond should raise doubts about the diagnosis. Most patients with Wernicke's encephalopathy will show complete recovery of extraocular motility, although a horizontal nystagmus may persist. Ataxia responds somewhat later, and only 40% of patients recover completely (2). The global confusional state also recedes, but 80% of patients with Wernicke's encephalopathy are left with a residual memory disorder (2).
The symptoms and signs of nutritional optic neuropathy, or "nutritional amblyopia," are similar to most other optic neuropathies. No single feature is pathognomonic. However, certain characteristics appear frequently. In the vast majority of cases, nutritional amblyopia is painless. Patients initially notice a blur or fog at the point of fixation, following by a progressive decline in visual acuity, which may be rapid. Visual field defects are nearly always central or cecocentral, related to atrophy of the papillomacular nerve fiber layer (8). Although the extent of decline is variable, total blindness or visual loss less than 20/400 is unusual in nutritional optic neuropathies (9). Findings are nearly always bilateral, although often asymmetric; completely normal findings in one eye should cast doubt on the diagnosis.
Because of the bilateral involvement in nutritional optic neuropathies, a relative afferent pupillary defect (as was noted in this patient) is not a common finding. Pupils will typically be reactive to light. The retina is usually normal or slightly hyperemic. Retinal hemorrhages may be present. Optic atrophy is typically a late finding but can be variable, initially manifesting as temporal disc pallor (8).
In his original clinical and pathologic observations from three patients, Carl Wernicke (1881) noted impairment of visual acuity in one patient (1). In general, however, impaired vision in Wernicke's encephalopathy is unusual. In the large case series of 232 patients reported by Victor et al., only six patients (2.6%) had visual impairment on initial examination (1). All of these patients' impairment was limited to central or cecocentral scotomata, as was the case with our patient. As with nutritional optic neuropathies in general, these neuropathies showed minimal ophthalmoscopic changes.
There are only rare case reports of actual blindness in Wernicke's encephalopathy. Timmings et al. (10) describe a 47 year old female who was consuming several bottles of wine daily, and presented with three weeks of diplopia, followed by six days of complete visual loss. She was oriented, but exhibited slow thought and speech and had poor insight into her condition. In this patient, there was no light perception, and her pupils were unreactive to light. She had ophthalmoplegia, nystagmus, mild global weakness and severe ataxia. She was able to count fingers within four hours of intravenous thiamine administration. By the second day, her visual acuity improved to 6/36 (meters) OD and 6/24 OS and, by the third day, 6/12 OD and 6/9 OS. By Day 27, after having switched to oral thiamine supplements, visual acuity was 6/6 bilaterally. Her other neurological symptoms also improved within days.
Van Noort et al. (11) report a similar case of complete blindness in a 35 year old man with ulcerative colitis who was receiving parental feeding with large amounts of glucose. He suddenly developed severe optic neuropathy and oculomotor palsy. This patient initially noted painless visual deterioration, and within days his visual acuity fell to 1/300 bilaterally. Fundoscopic exam showed bilateral disc pallor and edema with flame-shaped hemorrhages on the retina. However, his initial neurological examination was unremarkable. Two days later, his visual acuity was zero and pupils were non-reactive, while the eyes became divergent, and he became apathetic and depressed. After treatment with thiamine injections, his visual acuity rapidly recovered, as well as his other neurological symptoms, while the disc edema and hemorrhages resolved slowly over the next few months.
Finally, Tesfaye et al. (12) describe a 21 year old female with hyperemesis gravidarum who rapidly developed visual loss over 10 days. She became virtually blind, unable to count fingers at one meter. Optic disc swelling and retinal hemorrhages were found in both eyes, but the retinal arteries and veins appeared normal. She also had conjugate gaze restriction, nystagmus, and mild confusion, but lacked other typical findings of Wernicke's encephalopathy, including cerebellar signs or ataxia. Her vision improved within a few hours of intravenous thiamine administration, and she was able to read 12 hours later.
Initially termed "tobacco-alcohol" amblyopia, the optic neuropathy syndrome described above is likely due to nutritional deficiency and not to the toxic effects of alcohol or tobacco, as previously assumed. An amblyopia that is indistinguishable on clinical and pathologic grounds has been observed in many prisoners of war, in whom neither tobacco nor alcohol was a factor (1). Carroll supplied the most convincing evidence in 1944; 25 patients with "tobacco-alcohol amblyopia" were given their usual amounts of tobacco and alcohol but were also provided a nutritious diet or B vitamins; all the patients recovered vision completely or nearly completely (1).
The pathologic basis of this disorder has been studied in only two cases, neither of which showed the lesions of Wernicke-Korsakoff syndrome. The essential change was a degeneration of the optic nerves, chiasm, and tracts, largely confined to the papillomacular bundles (1). Normal electroretinograms associated with abnormal visual evoked potentials and non-reactive pupils in the patient described by Timmings et al. (10) is likewise consistent with an abnormality of the anterior visual pathway. The central scotoma appearing in thiamine deficiency may start as demyelination in the papillomacular bundles in the optic nerves, which spread centrifugally to involve more fibers (10). Axonal loss with secondary degeneration of retinal ganglion cells may eventually occur. The rapid and marked improvement of vision in these patients following administration of thiamine (often taking place within hours), however, indicates that a major component of the defect cannot be solely due to demyelination, axonal loss or petechial hemorrhage. More likely, the visual loss, as with many of the treatable symptoms of Wernicke's encephalopathy, is due to reversible metabolic dysfunction (10), such as the defect in glucose metabolism described above.
Ascribing optic neuropathy specifically to deficiency of thiamine is difficult. With the exception of vitamin B12 deficiency, no selective nutrient deficiency has been conclusively proven to cause amblyopia in humans (9). In some cases of nutritional optic neuropathy a deficiency of thiamine has been implicated, whereas in other cases, deficiencies of riboflavin, riboflavin or folic acid have been documented. However, these vitamins are all apt to be depleted in patients who are generally malnourished. The difficulty of controlling only one factor at a time to identify individual elements of significance continue to be problematic, and there are few good animal models of nutritional deficiency-related optic neuropathy (9).
Interestingly, in the case of the patient with ulcerative colitis described by Van Noort, et al., parenteral feeding with supplementary vitamins was begun long before onset of his visual deterioration. For this reason, Wernicke's encephalopathy was not considered as a cause of the optic neuropathy until several days after onset of symptoms, when a partial oculomotor palsy developed. Addition of thiamine alone alleviated the symptoms. In retrospect, it appeared that thiamine had not been added to the parenteral nutrient supplements (11). While Van Noort's example may establish that deficiency of thiamine alone was the causative factor, prisoner of war studies in the past have shown varying effects of thiamine replacement on visual loss (9), though this may have been due to the chronicity of symptoms. Meanwhile, when chronic thiamine deficiency was produced in human volunteers in the 1940's, their vision did not suffer (9). As only 2.6% of patients with Wernicke's encephalopathy in Victor's series developed loss of visual acuity, the authors suggest that an additional factor besides thiamine (or even any vitamin) deficiency is involved in the pathogenesis of nutritional optic neuropathy. This factor may be a form of genetic susceptibility, such as an enzyme or metabolic defect, symptomatic only in the setting of thiamine deficiency.
Wernicke's encephalopathy should be suspected in any individual with ocular motility symptoms, gait difficulties or alterations in mentation, regardless of the duration of symptoms or history of alcoholism (or lack thereof). Prompt empiric administration of thiamine is an important diagnostic test as well as a therapeutic intervention in these patients. Decreased visual acuity, especially when associated with clinical features typical for nutritional optic neuropathy, should also raise suspicion for a diagnosis of Wernicke's encephalopathy, and certainly should not dissuade one from this diagnosis.
It is with great regret that we note the death of Dr. Maurice Victor (June 21, 2001). His contributions to our knowledge of alcohol-related neurological disorders, to the field of clinical neurology, and to the education of students and residents, were extraordinary.
-- Dennis R. Mosier, M.D., Ph.D.
Email comments: