ACETYLCHOLINESTERASE INHIBITORS FOR THE TREATMENT OF WERNICKE–KORSAKOFF SYNDROME–THREE FURTHER CASES SHOW RESPONSE TO DONEPEZIL

MURRAY COCHRANE, ASHLEY COCHRANE, PRAMOD JAUHAR* and ELIZABETH ASHTON

Parkhead Hospital, 81 Salamanca Street, Glasgow, G31 5ES, UK

* Author to whom correspondence should be addressed at: Parkhead Hospital, 81 Salamanca Street, Glasgow, G31 5ES, UK. Tel: +44(0)141 211 8359; Fax: +44(0)141 211 8431. E-mail: Pramod.Jauhar{at}glacomen.scot.nhs.uk

(Received 23 April 2004; first review notified 5 June 2004; accepted in revised form 1 November 2004)


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Three patients diagnosed with Wernicke–Korsakoff syndrome were treated with the acetylcholinesterase inhibitor, donepezil, for periods of 6 to 8 months. Cognitive testing [Alzheimer's disease assessment scale–cognitive subscale (ADAS–Cog), Mini-mental state examination (MMSE), Clock drawing test and six item 2 min recall] and carer questionnaires [Informant Questionnaire (IQ Code), Neuropsychiatric inventory scale (NPI)] were performed at baseline, mid- and endpoint of the treatment period and post-discontinuation. Progressive partial improvement occurred in cognitive measurements through the treatment period, some of which was sustained after discontinuing donepezil. Carer questionnaires also indicated improvement. Confounding factors necessitate caution when attributing improvements to the medication, but these cases suggest that this option merits further investigation.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Treatments for established Wernicke–Korsakoff syndrome remain elusive (Smith and Hillman, 1999Go). Only one quarter of individuals diagnosed with Wernicke–Korsakoff syndrome recover completely, the remainder continuing to suffer varying degrees of memory deficit. As with the dementias the impact on the individual, carers and health and social service resources due to this syndrome makes the discovery of a viable, cost-effective treatment highly desirable.

It is well established that acetylcholinerase inhibitors offer a partial benefit in Alzheimer's dementia and other dementias characterized by memory deficit (Birks et al., 2000). Theoretically, acetylcholinesterase inhibitors might similarly prove to be at least partially useful in Wernicke–Korsakoff syndrome. Kopelman (1995)Go and Lishman (1990)Go have explored the possible role of the cholinergic system in the memory deficit seen in Wernicke–Korsakoff syndrome. Lishman (1990)Go suggested that, in many ways, biochemical manipulation of neurotransmitter systems would seem ‘a better bet’ in Wernicke–Korsakoff syndrome than in Alzheimer's disease ‘in which the pathological process is progressive’.

Kopelman and Corn (1988)Go investigated the effect of cholinergic blockade on a spectrum of memory functions and found a pattern of deficit that corresponded more closely to that seen in Wernicke–Korsakoff syndrome than that in Alzheimer's disease. The common factor underlying the development of Wernicke–Korsakoff syndrome is almost certainly due to depletion of thiamine. Thiamine pyrophosphate (TPP), the active form of thiamine, is involved in three enzymatic reactions essential for the metabolism of glucose and the production of neurotransmitters, including that of acetylcholine. The inhibition of TPP-dependent pyruvate decarboxylase may cause an impairment in the synthesis of acetylcholine (Gibson et al., 1975Go). Cell counting has shown that Wernicke–Korsakoff syndrome could be associated with a reduction of neurones in the basal forebrain, including the nucleus basalis of Meynert, the major source of cortical acetylcholine (Arendt et al., 1983Go).

Our literature search revealed three case reports of alcohol-related Wernicke–Korsakoff syndrome treated with acetylcholinesterase inhibitors (Angunawela et al., 2001Go; Iga et al., 2001Go; Casadevall et al., 2002Go). Although these reports suggested promising results, the degree of success in each case is hard to interpret because of a variety of factors including short treatment periods, lack of reporting of sequential cognitive testing, and confounding variables such as the known spontaneous partial recovery in the first few months after diagnosis with Wernicke–Korsakoff syndrome and ongoing treatment with thiamine.

A study of donepezil in the treatment of seven patients with non-alcoholic Wernicke–Korsakoff syndrome (Sahin et al., 2002Go) showed no benefit over placebo. The researchers acknowledged the small sample size and short (30 days) duration of treatment. The evidence for the response to donepezil in Alzheimer's disease was shown at 12 and 24 weeks (Birks et al., 2001Go). It may be that the shorter duration of treatment was insufficient to show a response in the treatment study of Sahin et al. (2002)Go. There might also be a slightly different pathology in non-alcoholic compared with alcoholic Wernicke–Korsakoff syndrome.

Three cases
We report three further cases of patients with a prior diagnosis of alcoholic Wernicke–Korsakoff syndrome treated with the acetylcholinesterase inhibitor, donepezil, for periods of 6 to 8 months. Cases A and C seem to represent a fairly pure form of Wernicke–Korsakoff syndrome, whilst case B represents a less clear-cut case complicated by small vessel disease. Donepezil was chosen as it is highly selective for acetylcholinesterase.

Protocol
Each patient had an established diagnosis of alcohol-related Wernicke–Korsakoff syndrome. Each was a routine reattender of the Addiction Psychiatry service. Following an initial meeting with the patient and carers, baseline tests were carried out. These consisted of physical tests (blood screen, ECG), baseline cognitive testing [Mini-mental state examination (MMSE), Alzheimer's disease assessment scale–cognitive subscale (ADAS–Cog), Clock drawing test, recall of six item name and address and recitation of months backward] and carer's questionnaires [Informant Questionnaire (IQ Code), Neuropsychiatric inventory scale (NPI)]. These tests were chosen because we apply the same protocol locally when assessing treatment of Alzheimer's disease with acetylcholinesterase inhibitors. Contemporary computerized tomography (CT) brain scans were arranged as required. No patient was excluded because of predefined exclusion factors (e.g. history of peptic ulceration or current alcohol misuse). Patients were then started on 5 mg of donepezil daily, which was increased to 10 mg daily after 1 month. Follow-up telephone contact was made on a monthly basis. Patient cognitive testing and, where possible, carer questionnaires were repeated at 3 and 6 months for cases A and B, and 4 and 8 months for case C. Treatment was then discontinued with one further follow-up assessment in cases A and B. All cognitive testing was administered by A.C. The NPI was administered by M.C.

ADAS–Cog is the Alzheimer disease assessment scale (cognitive subscale). The scale tests the components of memory, language and praxis and shows inter-rater reliability, test–retest reliability and construct validity (Rosen et al., 1984Go). It has been found to be a useful instrument for assessment of dysfunction in psychopharmacological studies on Alzheimer's dementia and due to symptom overlap, it may be applicable to other dementias.

The ‘Clock drawing test’ is a useful screen for cognitive impairment and is possibly useful as a marker for change in cognitive status (Shulman et al., 1986Go). It reflects frontal and temporo-parietal functioning and appears sensitive to change.

IQ Code is the Informant Questionnaire on cognitive decline in the elderly and shows internal consistency, reliability and validity (Jorm et al., 1989Go). The IQ Code also shows test–retest reliability over 1 year.

NPI is the Neuropsychiatric inventory and Caregiver distress scale. It is used to obtain information on the presence of psychopathology in patients with brain disorders, having specific application to patients with dementia. It shows content validity, concurrent validity, inter-rater reliability and test–retest reliability. It is sensitive to treatment effects, having ‘demonstrated the amelioration of behavioural symptoms in Alzheimer's dementia by cholinergic agents’. It is therefore a useful measure in assessing the efficacy of treatment (Cummings, 1997Go).

Case A
Case A is a 56-year-old woman who lived with her husband and youngest son. She was first diagnosed with Wernicke–Korsakoff syndrome 10 years earlier having been dependent on alcohol for at least 2 years prior to that time. Her husband had acted as her full-time carer following her diagnosis with Wernicke–Korsakoff syndrome. For several years she had been maintained on the same combination of medication, namely 20 mg of fluoxetine and one tablet of vitamin B compound daily. She had been partially deaf for 15 years and used a hearing aid. A CT scan of the brain performed in 1999 (4 years prior to commencing this trial) revealed ‘minor cerebral atrophy’.

Baseline physical tests revealed no significant abnormality and she was started on donepezil (administered by her husband); she tolerated the medication without side-effects. An improvement could be seen on the ADAS–Cog at 3 months (–4.0 points), which was sustained to the 6 month stage (–3.4 points from baseline). Marginal improvement could also be seen on MMSE at 6 months (+2 points). The score on MMSE dropped by 1 point on discontinuation. On 2 min recall of a six item name and address she had a baseline score of 0 out of 6. There was a marked improvement at 3 months (4 out of 6) and 6 months (5 out of 6). Interestingly, the recall score 4 months after discontinuing treatment showed further improvement (6 out of 6). There was no change at 3 months but deterioration was detected in the clock drawing test at 6 months (0.2 points).

Case B
Case B was a 62-year-old man who lived alone, but had regular contact with his son who provided carer ratings. He had a medical history that included emphysema and two cerebrovascular accidents (CVA), the most recent episode being 4 years earlier. He had a long history of alcohol dependence but had minimized his alcohol misuse to intermittent one-day binges. He had experienced a static level of memory dysfunction over the past 2 years that had led to a diagnosis of Wernicke–Korsakoff syndrome. His longstanding medication included 300 mg of thiamine daily. A CT scan of the brain revealed mild cerebral atrophy appropriate to his age and low attenuation in the periventricular matter bilaterally that was attributed to small vessel disease and ischaemia. This was unchanged from a previous scan performed 4 years earlier after his last episode of CVA.

Baseline blood tests and ECG revealed no significant abnormality and donepezil was commenced. The patient noticed some diarrhoea following the commencement of the medication but this quickly subsided. He reported greater alertness and a heightened awareness of his surroundings. He recounted an increased motivation and re-engaged in some previous interests, which was confirmed by his son. There were 2 periods of several days of non-compliance with medication, when his son confirmed that he had misused alcohol.

Score on the ADAS–Cog showed improvement at 3 months (–0.7 points) with further improvement at 6 months (–2.4 points), which was sustained after discontinuing treatment. MMSE also improved at 3 months (+2 points), which was again sustained throughout treatment and after discontinuation.

Baseline 2 min recall was 6 out of 6 (remaining unchanged throughout the treatment period). This might suggest that the diagnosis of Wernicke–Korsakoff syndrome was inappropriate or that he had improved partially in the 2 years since the diagnosis was made, perhaps related to abstinence and oral thiamine. However, the improvements in ADAS–Cog and MMSE were in the domains of short-term recall and orientation for time, suggesting that memory had improved in the treatment period. He had a perfect score on the Clock drawing test at all stages of treatment. Carer ratings, available only at 3 months, showed marked improvement in cognitive and non-cognitive psychopathological symptoms.

Case C
Case A was a 50-year-old man who lived alone but had daily contact with his sister who provided the carer ratings. His long history of alcohol dependency included periodic delirium tremens and withdrawal seizures. He had received a diagnosis of Wernicke–Korsakoff syndrome following the development of memory difficulties of an apparent sudden onset ~1 year earlier. He had received previous diagnoses of gastritis and alcoholic liver disease and had restricted his use of alcohol since that time, due to abdominal discomfort. However, he continued to have cycles of daily drinking lasting up to 3 weeks followed by ~2 months of abstinence. At the time of baseline testing he had been abstinent for 6 weeks. He had been receiving 300 mg of oral thiamine for 9 months daily. A CT scan of the brain revealed cerebral atrophy.

Baseline physical tests revealed no significant abnormality and donepezil was commenced. He noted an improved concentration and a renewed interest in drawing, his previous hobby. At the time of his 4 month reassessment he had discontinued medication for 6 weeks at the start of a 2 week lapse to daily drinking, followed by prescription of disulfiram, which was continued throughout the rest of the trial period with sustained abstinence thereafter. Repeat testing was carried out despite awareness of these confounding factors.

An improvement could be seen on ADAS–Cog at 4 months (–2.0 points), which was improved upon at 8 months (–7.0 points from baseline). MMSE showed no improvement at 4 months but did improve (+2 points) at 8 months. His baseline 2 min recall was 1 out of 6, remaining so at 4 months but scoring 5 out of 6 at 8 months. He scored perfectly on the Clock drawing test throughout the trial period. Carer ratings, IQ Code and NPI indicated little change at 4 months and the carer was unavailable for the 8 month testing.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Donepezil appears to have had a beneficial effect on both cognitive and non-cognitive functioning in each of the above cases.

Case A is the most straightforward and convincing and her improvements on MMSE, ADAS-Cog and 2 min recall are interesting. There was a 10 year history of established Wernicke–Korsakoff syndrome, with a corroborated history of complete abstinence since diagnosis and a static degree of memory dysfunction persisting well outside the time-period of spontaneous recovery expected following initial diagnosis. Compliance with donepezil was assured by her husband and there were no confounding episodes of alcohol misuse. For this case and also for cases B and C, we believe that the effect of learning through repeat testing was minimal due to the small number of repeat tests, the length of time between tests and the pre-existing memory deficit. However, there is, as for each of the three patients, a possible placebo effect of the drug as well as a possible ‘Hawthorn effect’ due to greater attention from therapists and carers during the treatment period.

There were numerous confounding factors in cases B and C. Recruiting patients without such confounding factors is difficult in this area of study, reflecting that patients whose alcohol use results in alcohol-related brain damage often have multiple pathologies, tend to be non-compliant with treatment and may relapse to alcohol use during the study. Patients with cerebral damage due to alcohol tend to have multiple cognitive impairments. Although we tried to isolate the individuals with a memory deficit suggestive of Wernicke–Korsakoff syndrome, such ‘pure’ cases are rare. In case B the presence of cerebrovascular disease complicates the diagnosis of Wernicke–Korsakoff syndrome. Nonetheless, improvements in the domain of short-term recall were evident during the trial. The ongoing prescription of thiamine was beyond any period where further benefit to memory may be expected but may have contributed to some improvement in cognitive function from the time of initial diagnosis such that he scored highly on baseline testing of 2 min recall. ADAS–Cog and MMSE did reveal evidence of ongoing short-term memory deficit at baseline. Other confounding factors included intermittent alcohol misuse, and co-occurring non-compliance with donepezil.

In case C, periods of alcohol misuse followed by abstinence up to the point of commencement on disulfiram 3 months into the trial confound the results. It may be that the dramatic improvements in 2 min recall and ADAS–Cog scores at the end of the trial period reflect his longest period of abstinence (5 months) for several years. Disulfiram, not donepezil, may be the important factor in this case.

All three patients described an increased feeling of well-being, including reduced anxiety and ‘clearer thinking’, following commencement on donepezil. Improvements seemed to be largely sustained following discontinuation of donepezil. If this was truly due to medication leading to sustained improvement beyond the treatment period this would suggest that Lishman's prediction was correct and that Wernicke–Korsakoff syndrome is ‘a better bet’ than the ‘progressive dementias’ when using cognitive enhancers.


    CONCLUSION
 TOP
 ABSTRACT
 INTRODUCTION
 DISCUSSION
 CONCLUSION
 REFERENCES
 
It may be that only partial response to acetylcholinesterase inhibitors can be expected. Lishman (1990)Go noted that ‘patients with established Wernicke–Korsakoff syndrome are likely to be suffering from the effects of a combination of alcohol neurotoxicity and thiamine deficiency, both needing to be present in severe degree before the syndrome is encountered in its fully fledged form’. Targeting the sequelae of thiamine depletion (such as cholinergic deficit) alone may be inadequate for significant resolution of the syndrome (it is notable that cholinergic blockade does not account for the long retrograde amnesia seen in Wernicke–Korsakoff syndrome (Kopelman and Corn, 1988Go). It may be that combining other pharmacological treatments with an acetylcholinesterase inhibitor, would result in greater recovery. The structural changes, including cortical (especially frontal) atrophy, subcortical damage and ventricular enlargement, found in individuals with Wernicke–Korsakoff syndrome (Lishman, 1990Go), suggest that pharmacological agents will ultimately prove to be of limited value. It is clear that no conclusions can be drawn on the basis of the three case reports, other than that the matter merits further investigation and research.


    ACKNOWLEDGEMENTS
 
We thank the staff at the Alcohol Day Hospital, Parkhead Hospital and, in particular, Charge Nurse Tracy Stafford for her help with baseline physical tests. The study was not funded by external sources, and the authors report no conflict of interest.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Angunawela, I. I. and Barker, A. (2001) Anticholinesterase drugs for alcoholic Korsakoff syndrome. International Journal of Geriatric Psychiatry 16, 338–339.[CrossRef][ISI][Medline]

Arendt, T., Bigl, V., Arendt, A. and Tennstedt, A. (1983) Loss of neurons in the nucleus basalis of Meynert in Azheimer's disease, paralysis agitans and Korsakoff's disease. Acta Neuropathologica 61, 101–108.[ISI][Medline]

Birks, J. S., Meltzer, D. and Beppu, H. (2001) Donepezil for mild and moderate Alzheimer's disease (Cochrane Review). The Cochrane Library 1, 1–14.

Casadevall, C. T., Pascual, M. L. F., Fernandez, T. T., Escalza, C. I., Navas, V. I., Fanlo, M. C. and Morales, A. F. (2002) Pharmacological treatment of Korsakoff psychosis: a review of the literature and experience in two cases. Revista de Neurologica 35, 341–345.

Cummings, J. L. (1997) The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology 48 (Suppl. 6), S10–16.[Abstract]

Gibson, G. E., Jope, R. and Blass, J. P. (1975) Decreased synthesis of acetylcholine accompanying impaired oxidation of pyruvic acid in rat brain minces. Biochemical Journal 148, 17–23.[ISI][Medline]

Iga, J.-I., Araki, M., Ishimoto, Y. and Ohmori, T. (2001) A case of Korsakoff's syndrome improved by high dose donepezil. Alcohol and Alcoholism 36, 553–555.[Abstract/Free Full Text]

Jorm, A., Scott, R. and Jacomb, P. A. (1989) Assessment of cognitive decline in dementia by informant questionnaire. International Journal of Geriatric Psychiatry, 4, 35–39.[CrossRef][ISI]

Kopelman, M. D. (1995) The Korsakoff Syndrome. British Journal of Psychiatry 166, 154–173.[Abstract]

Kopelman, M. D. and Corn, T. H. (1988) Cholinergic ‘blockade’ as a model for cholinergic depletion: a comparison of the memory deficits with those of Alzheimer-type dementia and the alcoholic Korsakoff syndrome. Brain 111, 1079–1110.[Abstract]

Lishman, W. A. (1990) Alcohol and the Brain. British Journal of Psychiatry 156, 635–644.[Abstract]

Rosen, W. G., Mohs, R. C. and Davis, K. L. (1984) A new rating scale for Alzheimer's disease. American Journal of Psychiatry 141, 1356–1364.[Abstract]

Sahin, H. A., Gurvit, I. H., Bilgic, B., Hanagasi, H. A. and Emre, M. (2002) Therapeutic effects of an acetylcholinesterase inhibitor (Donepezil) on memory in Wernicke–Korsakoff's disease. Clinical Neuropharmacology 25, 16–20.[CrossRef][ISI][Medline]

Shulman, K. I., Shedletsky, R. and Silver, I. L. (1986) The challenge of time: clock-drawing and cognitive function in the elderly. International Journal of Geriatric Psychiatry 1, 135–140.

Smith, I. and Hillman, A. (1999) Management of alcohol Korsakoff syndrome. Advances in Psychiatric Treatment 5, 271–278.