THERAPEUTIC  EFFECTS OF AN ACETYLCHOLINESTERASE INHIBITOR (DONEPEZIL) ON MEMORY IN WERNICKE-KORSAKOFF’S DISEASE

Hüseyin A. Şahin, M.D.
İ.Hakan Gürvit, M.D.
Başar Bilgiç, M.D.
Haşmet A. Hanağası, M.D.
Murat Emre, MD, Professor of Neurology.

İstanbul University, İstanbul Medical School, Department of Neurology,
Behavioral Neurology and Movement Disorders Unit
34390 İSTANBUL-TURKEY

Address correspondence and reprints:
Professor Murat Emre M.D.
İstanbul University, İstanbul Medical School,
Department of Neurology,
34390 Çapa/İstanbul TURKEY
Tel: 90 212 5338575   Fax: 90 212 5338575       
e-mail: muratemre@superonline.com

Key Word: Wernicke Korsakoff’s disease, cholinesterase-inhibitors, memory.

Abstract

Objective- Wernicke-Korsakoff’s Disease (WKD) is cognitively an amnestic state resulting from strategic lesions in the limbic system subserving episodic memory network, due to thiamine deficiency. Neurochemical deficits have been implicated in the pathophysiology of amnesia based on the pathologic observations that various brainstem and basal forebrain nuclei are also affected. Previous treatment attemps with serotoninergic, noradrenergic and  cholinergic drugs gave controversial results. The objective of this study was to assess the effects of a cholinesterase inhibitor, donepezil, on memory, attention and executive functions of patients with non-alcoholic WKD.

Methods- Seven patients who developed WKD following a hunger-strike were included in this single-blind, placebo-controlled, cross over study.  The patients were given donepezil during the first 30, and placebo during the following 30 days. Neuropsychological tests to evaluate verbal and visual memory, attention and  executive functions were performed on days 0, 31 and 61.

Results- All patients completed the both phases of the study. There were no statistically significant differences between the three evaluations, except for a difference between active treatment and placebo phase in recall of Rey-Osterrieth Complex Figure, which was in favour of the placebo phase. Three were no statistically different changes in favor of the active treatment

Conclusions-  Cholinergic treatment with cholinesterase-inhibitors does not seem to be effective in WKD. This may be because pathways mediating both channel and state dependent functions are impaired in this disease, and enhancement of state-dependent cholinergic transmission may not be sufficient.

INTRODUCTION

Wernicke-Korsakoff’s Disease (WKD) is cognitively a pure amnestic state that is related to thiamine depletion¹.  Corpus mamillare (CM), dorsomedial nucleus of thalamus (DM) and periaquaductal gray matter are the major neural structures that are involved in the disease process^1,2,3.  First two structures are considered to be responsible for amnesia, since they are the basic relay stations on the two parallel pathways connecting hippocampus and amygdala with other structures of episodic memory neural network^4,5,6,7,8. Lesions in the ascending neurotransmitter nuclei, i.e. noradrenergic locus ceruleus, serotonergic dorsal raphe nucleus and cholinergic basal nucleus of Meynert, are also noted^{9,10,11,12,13,14,15}. The rationale for pharmacological intervention in WKD, via manipulation of different neurotransmitter systems is largely based on the latter pathologic observations^11,12,13,16.  In the past there have been a number of controversial reports regarding the beneficial effects of monoaminergic treatments^10,12,16,17.  Despite the implication of nucleus basalis of Meynert in the pathophysiology, to our knowledge, there have been only two reported studies on the use of cholinergic agents in alcoholic WKD, and none in non-alcoholic WKD^18,19. In the present study we assessed the therapeutic effects of an acetylcholinesterase inhibitor(AchE-I), donepezil on  memory, attention and executive functions of patients with non-alcoholic WKD.

METHODS

Patients

Seven patients with the diagnosis of WKD were included in the study. All patients were under our care since the beginning of their illness in 1996. They were all suffering from the consequences of a 69-day-mass scale hunger strike. Forty of the survivors had a cerebellar syndrome (nystagmus, ataxia) compatible with the sequel of Wernicke’s Encephalopathy, seven of whom had an additional amnestic state.  The latter group was labelled as WKD.  This group was among the 18 hunger strikers, who were referred to our care for a 3 to 6 week-hospital stay immediately after the termination of the strike. Clinical characteristics of patients was shown on Table 1.  Despite the variable degree of improvement in cerebellar signs, amnesia was largely stable over the ensuing 3 years.

All seven patients were male with a mean age of 31 years (range 27-39).   Four of them were mildly ataxic and required no assistance in walking, 2 had moderate ataxia and were in need of walkers, and last one had severe ataxia, being practically wheelchair-bound.

Five of them had psychotic episodes during the follow-up period. The episodes always had a depressive character accompanied with fantastic paranoid delusions. Two had single episodes, one in the 2^nd month, one in 2^nd year after the hunger strike The former was markedly apathetic and the latter was psychiatrically symptom-free   during the present study. Three had recurrent psychotic episodes and showed some residual paranoid ideation. The remaining two did not develop psychosis, but displayed marked personality changes. One was apathetic, the other showed disinhibition with inappropriate, childish behavior.

The anterograde amnesia was classified as moderate to severe in 6, and mild in one. After 3 years, they all remembered that they had a hunger strike, they had some recollection of the medical team, but temporal aspects of the retained information were virtually absent: “when and how long was the hunger strike?”. A recently released patient could not answer whether he had been free since a week or six months.   Retrograde component went several years back.  One patient identified himself as belonging to a political group which had already been dissolved to merge into another one, some two years ago.   Another patient was telling how he had been surprised to see his 8-year old son as a big child, everytime he came to visit him in prison, while all his memories about him could not go beyond his infancy, how hard he had tried.

Evaluations

A neuropsychological battery, listed in Table 2 was used to assess the efficacy of cholinergic treatment^{20,21,22,23,24,25}.  Detailed neuropsychological assessments of patients had already been performed with a comprehensive battery, acutely and 1-year post-onset.  The results had been presented elsewhere²⁶.  Briefly, their cognitive profile can be summarized as a stable, isolated amnestic state, where linguistic, visuo-spatial and executive functions were well preserved.  For the purpose of this study, a relatively brief neuropsychological battery with an emphasis on amnestic functions was selected.  The same test material was used in all three sessions. After detailed explanation of the study, to the patients themselves and to the caregivers, consents were obtained from the caregivers.

In a single-blind, two-phased, crossover design, a single morning dose of donepezil (5mg   for the first fifteen days, 10 mg thereafter)   was administered for 30 days, after which donepezil placebo was administered for the same duration. Neuropsychological testing was performed on days 0, 31 and 61.

The three set of test scores at baseline, end of active treatment and end of placebo phase were compared with each other using Friedman test. Statistically significant differences were further analyzed with Wilcoxon-paried test in order to assess if the significance was due to changes from baseline to the end of active treatment or to changes from baseline to the end of placebo treatment.

RESULTS

All seven patients completed both phases of the study as planned. They all tolerated the drug well and there were no significant adverse events. None of the patients, neither the caregivers noted any subjective improvement or worsening during the two phases of the treatment.  Statistical analysis of neuropsychological test scores across the three testing sessions using Friedman test revealed statistically significant differences only for Camden topographical recognition memory test and Rey-Osterrieth Complex Figure immediate and delayed recall (Table 3). Further analysis using Wilcoxon’s paired-t test revealed, however, that the only significant difference between active treatment and placebo phase was in Rey-Osterrieth Complex Figure. This difference, however, was in favour of the placebo phase (Table 4). There were no other statistically significant changes, particularly in favour of the active treatment, in any other score.

DISCUSSION

The central monoaminergic and basal forebrain cholinergic pathways are vulnerable to disruption in WKD due to the critical location of the classical lesions.   McEntee et al. found a decrease in 3-methoxy-4-hydroxy-phenylglycol (MHPG) levels, a metabolite of noradrenalin, in cerebrospinal fluid of  patients with WKS and noted a strong correlation between the decrease in CSF MHPG and the degree of memory impairment^9,10,11. This finding prompted the use of noradrenergic substances in therapeutic trials.  Mc Entee and Mair reported favourable results with clonidine, an alpha-2 adrenergic agonist in two separate studies.  Measures of anterograde memory function improved in both studies, but apparently the impact of this improvement on activities of daily living (ADL’s) was negligible^10,16.  Mc Entee and Mair also reported decreased CSF levels of 5-HIAA, the metabolite of serotonin¹¹.   Based on this observation, Martin et al used fluvoxamine, a serotonin reuptake inhibitor, in an attempt to treat the memory impairment.  They found improvement on psychometric measures including memory and attention, but there was no mention of ADL’s¹².  In two studies, Arendt et al., and Cullen et al., demonstrated reduced cell counts in the Nucleus basalis of Meynert (Ch4) in patients with WKD^14,15. There have been two reports on the therapeutic use of  cholinergic agents in the patients.  O’Donnell et al. reported on the use of cholinergic precursor choline chloride and Franchesci et al. on AChE-I physostigmine, both  without any beneficial effects^18,19.  

This single blind, crossover, placebo-controlled study failed to show any beneficial effects of donepezil treatment for 1 month on any of the cognitive parameters in patients with WKD, including verbal and visual memory. In general, similar almost identical performances were obtained across the three testing sessions within a two-month time period. Improvements on some measures were largely insignificant and clinically irrelevant.

It is now widely accepted that the  neural plasticity is essential for the registration and storage of new information^27,28. It is likely that hippocampus, amygdala and connected limbic structures are necessary for learning and recall of newly acquired information, whereas it seems that long term storage occurs in the neocortex where the specific information had been processed, via plastic changes^4,5,6. Cholinergic innervation of  the neocortex  is  thought to be largely responsible for the long-term  stability of  the memory traces. 

In our study AchE-I donepezil failed to induce any beneficial effects on the memory of patients with WKD. This negative result may be due to the fact that primary lesions which are responsible for amnesia in WKD are on two critical structures of the episodic memory network^2,3,4,5,6,7.  These are hippocampus and amygdala-related gateways or bottle-necks for access into the network^27,28. As suggested by Mesulam, cortico-cortical and cortico-subcortical monosynaptic reciprocal connections of a neural network subserve its channel-dependent functions, which define the type of information to be processed²⁷. On the other hand, state-dependent functions are subserved by diffuse projections from ascending neurochemical pathways to the network in question. These modulate the intensity and enable the stability of the specific information. Therefore, disruption of the former function is largely not amenable to pharmacological intervention, whereas the latter is. Deficits due to territorial strokes versus those in Parkinson’s disease are such examples, respectively²⁷.  Alzheimer’s disease, on the other hand, stands on the midway between these two examples.  Limbic degeneration in AD disrupts both channel (hippocampus and amygdala) and state-dependent (Nucleus basalis of Meynert) functions; cholinergic enhancement which has a modest yet significant place in AD treatment, may work through improvement in state-dependent functions. In WKS, however, where the core limbic damage is acute  and complete, no new information can gain access into the system in order to be processed and stabilised through cholinergic mediation²⁷.  

In conclusion, cholinergic treatment in WKD does not seem to improve memory function.   This may be due to the fact that, when channel-dependent functions fail because of the  disruption in the major relay stations of the memory network, enhancement of state-dependent modifying inputs cannot improve function²⁸.

Acknowledgement: We would like to thank to Dr. Sevda Özel for statistical analysis and to Pfizer inc., Turkey for supplying the study drugs.

REFERENCES

1.       Victor M, Adams RA, Collins GH. The Wernicke-Korsakoff syndrome and related  neurological disorders due to alcoholism and malnutrition. 2^nd ed. Philadelphia: F A Davies Company,1989.

2.        Mair WGP, Warrington EK, Weiskrantz L. Memory disorder in Korsakoff’s psychosis. A neuropathological and neuropsychological investigation of two cases. Brain 1979;102:749-783.

3.        Mayers AR, Meudell PR, Mann D, Pickering A.  Localization of lesions in Korsakoff’s syndrome: Neuropsychological and neuropathological data on two patients. Cortex 1988; 24:367-388.

4.       Markowitsch H, Pritzel M. Neuropathology of amnesia. Prog Neurobiol 1985;25:189-287.

5.       Tulving E, Kapur S, Markowitsch HJ, et al. Neuroanatomical correlates of retrieval in episodic memory: auditory sentence recognition. Proc. Natl. Acad. Sci. USA. 1994; 91:2012-2015.

6.       Squire LR, Zola SM. Structure and function of declarative and non-declarative memory systems. Proc. Natl. Acad. Sci. USA. 1996; 93:13515-13522.

7.       Fletcher PC, Frith CD, Rugg MD. The functional neuroanatomy of episodic memory. Trends Neurosci 1997; 20:213-218.

8.       Corkin S, Amaral DC, Gonzalez RC, Johnson KA, Hyman BT. H.M’s medial temporal lobe lesion: findings from magnetic resonance imaging.  J Neurosci 1997;17:3964-3979.

9.       McEntee WJ, Mair R. Memory impairment  in Korsakoff’s Psyhosis: A correlation with brain noradrenergic activity. Science 1978;202:905-907.

10. McEntee WJ, Mair RG. Memory enhancement in Korsakoff’s psychosis by clonidin: further evidence for a noradrenergic deficit. Ann Neurol 1980;7:466-470.

11. McEntee WJ, Mair RG, Langlais PJ.  Neurochemical pathology in Korsakoff ‘s psychosis: implication for other cognitive disorders. Neurology 1984;34:648-652.

12. Martin PR, Adinoff B, Eckardt M, et al. Effective pharmacotherapy of alcoholic amnestic disorder with fluvoxamine. Arch Gen Psychiatry 1989; 46:617-621.

13. McEntee WJ, Mair RG. Korsakoff’s syndrome: a neurochemical perspective. Trends Neurosci 1990;13:340-344.

14. Arendt T, Bigl V, Arendt A, Tennstedt A.   Loss of neurons in the nucleus basalis of Meynert in Alzheimer’s disease, paralysis agitans and Korsakoff’s disease. Acta Neuropathol(Berl) 1983;61:101-108.

15. Cullen KM, Halliday GM, Caine D, Krill JJ. The nucleus basalis (Ch4) in the alcoholic Wernicke-Korsakoff syndrome: reduced cell number  in both amnestic and non-amnestic patients. J Neurol Neurosurg Psychiatry 1997;63:315-320.

16. McEntee WJ, Mair RG, Langlais PJ. Clonidin in Korsakoff’s disease: pathologic and therapeutic implications. Prog. Clin. Biol. Res. 1981;71:211-23.

17. O’Carrol RE, Moffoot A, Ebmeier KP, Murray C, Goodwin M. Korsakoff’s syndrome, cognition and clonidine. Psychological Medicine 1993;23:341-347.

18. Franchesci M, Tancredi O, Savio G, Smirne S. Vasopressin and physostigmine in the treatment of amnesia. Eur Neurol 1982;21:388-391.

19. O’Donnell V, Pitts W, Fann WE. Noradrenergic and cholinergic agents in Korsakoff’s syndrome. Clin Neuropharmacol 1986;9:65-70.

20. Delis D, Kramer J, Kaplan E, and Ober B: The California verbal learning test. The  Psychological Corporation. San Antonio, Texas, 1987.

21. Weintraub S. Neuropsychological assessment of mental state. In: Mesulam M, ed. Principles of behavioral  and cognitive neurology, 2^nd  ed. New York: Oxford University Press 2000:121-173.

22. Warrinton EK. The Camden memory tests. Psycholgy Press, an imprint of Erbaum (UK) Taylor&Francis Ltd. 1996.

23. Wechsler D: Wechsler adult intelligence scale-third edition. The psychological corporation,  San Antonio,Texas, 1998.

24. Reitan RM; Validity of the trail-making test as an indication of organic brain damage. Percept Mot Skills 1958;8:271-276.

25. Comalli PJ, Wapner S and Werner H. Interference effects of stroop color-word test in childhood, adulthood, and aging. J Genet Psychol 1962;100:47-53.

26. Gürvit H, Gökmen E,  et. al. Hunger strike-related Wernicke-Korsakoff’s Disease. J Neurol Scien 1997;150:s39.

27. Mesulam M. Behavioral Neuroanatomy. In: Mesulam M, ed. Principles of behavioral  and cognitive neurology, 2^nd  ed. New York: Oxford University Press 2000: 1-120.

28. Mesulam M. Neuroplasticity failure in Alzheimer’s disease: Bridging the gap between plaque and tangles. Neuron 1999,24:521-529.

Table 1            Clinical characteristics of patients       

Table 2             Neuropsychological Battery

Verbal Memory

California Verbal Learning Test²⁰

Visual Memory

            Rey-Ostherrieth Complex Figure²¹

            Camden Facial, Pictorial and Topographical Recognition Tests ²²

Attention

            Digit Span²³

Executive Functions

Trail Making Test A-B²⁴

Stroop Test²⁵

Controlled Oral Word Association Test (K-A-S)

Table 3             Results of the Neuropsychological Evaluation

A: Baseline (mean±SD)

B: Day 31 (mean±SD)

C: Day 61(mean±SD)

Table 4 Results of Wilcoxon paired-t test

A: Baseline (mean±SD)

B: Day 31 (mean±SD)

C: Day 61(mean±SD)