Cholinergic Markers in Elderly Patients with Early Signs of Alzheimer Disease

CLINICAL INVESTIGATION

Cholinergic Markers in Elderly Patients With Early Signs of Alzheimer Disease

Kenneth L. Davis, MD Context A central tenet of Alzheimer disease (AD) is the loss of cortical cholinergic Richard C. Mohs, PhD function and cholinergic markers in postmortem brain specimens. Whether these profound deficits in cholinergic markers found in end-stage patients are also found in pa- Deborah Marin, MD tients with much earlier disease is not known. Dushyant P. Purohit, MD Objective To determine whether cholinergic deficits in AD precede, follow, or Daniel P. Perl, MD occur in synchrony with the earliest signs of cognitive deterioration. Melinda Lantz, MD Design, Setting, and Patients Postmortem study of nursing home residents with Clini- cal Dementia Rating (CDR) Scale scores of 0.0 to 2.0 and 4.0 to 5.0 who underwent au- Gregory Austin topsy between 1986 and 1997, comparing the activity of the cholinergic marker en- Vahram Haroutunian, PhD zymes in the cortices of 66 elderly subjects with no (CDR score = 0.0; n = 18), questionable (CDR score = 0.5; n = 11), mild (CDR score = 1.0; n = 22), or moderate (CDR score = 2.0; CENTRAL TENET OF ALZHEIMER n = 15) dementia vs subjects with severe dementia (CDR score = 4.0-5.0; n = 15). disease (AD), established 20 Main Outcome Measures Activity of the cholinergic marker enzymes choline acet- years ago and repeatedly rep- yltransferase and acetylcholinesterase in 9 neocortical brain regions. licated, is the loss of cortical Results The activity of choline acetyltransferase and acetylcholinesterase in 9 neo- cholinergicA markers, specifically, cho- cortical brain regions did not differ significantly in subjects with CDR scores of 0.0 to line acetyltransferase (ChAT) and ace- 2.0, but was significantly lower in subjects with severe dementia (CDR score = 4.0-5.0). tylcholinesterase (AChE) activity, in Choline acetyltransferase levels were significantly correlated with severity of neuropatho- postmortem tissue from AD patients.1-3 logical lesions of AD, as measured by density of neuritic plaques and neurofibrillary tangles. This abnormality has been shown to cor- Conclusions Although neocortical cholinergic deficits are characteristic of severely relate with neuropathological markers4 demented AD patients, in this study, cholinergic deficits were not apparent in indi- and with the severity of dementia.4,5 viduals with mild AD and were not present until relatively late in the course of the Therapies designed to reverse the cho- disease. These results suggest that patients with more severe disease should be a tar- linergic deficit, such as AChE inhibitors, get for cholinergic treatment. are in large measure based on the impor- JAMA. 1999;281:1401-1406 www.jama.com tance of the cholinergic deficit to cog- nition and the symptoms of AD.6-8 Center, in collaboration with the Jewish topsieswereperformedbetween1986and Most postmortem studies assessing Home and Hospital (both in New York, 1997 and who had been residents of the cholinergic markers in AD derive from NY), has initiated a prospective study of Jewish Home and Hospital in Manhattan patients with end-stage dementia. Those itsresidents.Individualswhoarerelatively and the Bronx, NY. Autopsies were per- few studies in which brain biopsy speci- highly functioning are assessed at entry formedafterreceivinginstitutionalreview mens were obtained and cholinergic intothehomeandannuallythereafterand board approval and consent for autopsy markers were assessed antemortem gen- followed up until their death, when at- from each subject’s legal next of kin. erally are restricted to patients with ei- tempts are made to have an autopsy per- Clinical Dementia Rating Scale. formed.Inthisstudy,weevaluatethecho- First, subjects were assigned a Clini- ther a very early onset of dementia or 11,12 relatively advanced dementia.3,9,10 Thus, linergic markers of these patients with no cal Dementia Rating (CDR) Scale whether these profound deficits in cho- cognitive impairment, questionable cog- score using a multistep approach based linergic markers found in end-stage pa- nitive impairment, and early AD to address the relationship between early AD tients apply to patients with much less Author Affiliations: Departments of Psychiatry (Drs severe disease is not clear. and cholinergic markers. Davis, Mohs, Marin, and Haroutunian and Mr Aus- tin) and Pathology (Drs Purohit and Perl), Mount Si- To study the earliest changes in AD, METHODS nai School of Medicine, and the Jewish Home and Hos- the Mount Sinai Alzheimer’s Disease pital (Dr Lantz), New York, NY. Subjects Corresponding Author and Reprints: Kenneth L. Davis, Eighty-one subjects were selected from MD, Department of Psychiatry, Mount Sinai School For editorial comment see p 1433. of Medicine, One Gustave L. Levy Place, New York, 278 consecutive patients on whom au- NY 10029-6574.

Downloaded From: https://jamanetwork.com/ on 09/23/2021 CHOLINERGIC MARKERS IN EARLY AD

on cognitive and functional status dur- other neuropathological lesions of suf- Final Case Selection ing the last 6 months of life. These steps ficient magnitude to contribute to cog- After the completion of the neuropatho- involved (1) assignment of a CDR score nitive dysfunction were excluded from logical studies and assignment of con- based on a careful review of all infor- consideration. These neuropathologi- sensus CDR scores, a final consensus mation contained within each pa- cal lesions included but were not lim- conference was held (with the partici- tient’s chart; (2) a blinded review of the ited to Pick disease, diffuse Lewy body pation of K.L.D., R.C.M., D.M., D. P. Pu- same records by a second reviewer ex- disease, Parkinson disease, stroke, rohit, D. P. Perl, and V.H. and the an- perienced in neuropsychological as- multi-infarction dementia, and severe temortem and postmortem assessment sessment of living elderly patients and cerebrovascular disease. Subjects with team) to select cases for measurement the assignment of a second indepen- mild cerebrovascular disease judged by of cholinergic markers and inclusion in dent CDR score; and (3) a telephone in- our neuropathologists (D. P. Purohit this study. All patients with neuropatho- terview with at least 1 family member and D. P. Perl) to be insufficient in se- logical lesions other than those of AD and/or caregiver for each subject and verity to affect cognitive function were were excluded. Because the aim of this assignment of a third CDR score. All 3 not excluded from consideration for study was to identify the relationship of CDR scores and all pertinent chart in- further study. cortical cholinergic markers to early and formation were subsequently pre- The neuropathological assessment mild dementia, only subjects with CDR sented to a senior clinician (D.M.) and consisted of examining representative scores of 0.0 to 2.0 were selected for in- a consensus CDR score was derived. blocks from superior and midfrontal clusion. These were considered as 4 CDR Validation. The reliability of the gyrus, orbital cortex, basal ganglia with groups, having CDR scores of 0.0, 0.5, postmortem chart review procedure for basal forebrain, amygdala, hippocam- 1.0, and 2.0. An additional group of 15 CDR scoring was determined by di- pus (rostral and caudal levels with subjects with CDR scores of 4.0 and 5.0 rect observation and patient interview adjacent parahippocampal and inferior (referred to hereafter as CDR scores of and by chart review alone for 35 sub- temporal cortex), superior temporal gy- 5.0) were included to represent severe jects. An interclass correlation coeffi- rus, parietal cortex (angular gyrus), cal- dementia. These subjects were selected cient of 0.86 was obtained for the 2 in- carine cortex, hypothalamus with mam- to match those with CDR scores of 0.0 dependent assessments of CDR. A millary bodies, thalamus, midbrain, to 2.0 as closely as possible with re- subset of the subjects (n = 22) had been pons, medulla, cerebellar vermis, and lat- spect to age, sex, and time from death neuropsychologically assessed during eral cerebellar hemisphere. Sections from to autopsy. Subjects with a CDR score life and had participated in longitudi- paraffin-embedded blocks were stained of 4.0 (n = 5) did not differ signifi- nal studies of cognitive function with using hematoxylin-eosin, modified Biel- cantly from subjects whose CDR score instruments such as the Mini-Mental schowsky, modified thioflavine S, anti– was 5.0 (n = 10) with respect to any de- State Examination13 and the Alzheim- ␤-amyloid, and anti-␶. Every case was mographic, neuropathological, or neu- er’s Disease Assessment Scale.14 When evaluated for the extent of neuropatho- rochemical parameter measured. available, the neuropsychological as- logical lesions using the Consortium to sessment results were also considered Establish a Registry for Alzheimer’s Dis- in deriving the final consensus CDR ease (CERAD) neuropathological bat- Measurement score. The correlation between the con- tery.17 The results of the neuropatho- of Cholinergic Markers sensus CDR score assigned and the logical studies were used to exclude all At the time of autopsy, the brain was Mini-Mental State Examination score cases that did not meet criteria as neu- halved midsagittally. The right half was for those 22 subjects who had been as- ropathologically normal or subjects who fixed in 4% paraformaldehyde solution sessed antemortem was r = −0.48 had neuropathological lesions other than and was used for neuropathological (P= .03). If only those subjects who had those associated with AD. Estimates of evaluation as described herein. The left received a Mini-Mental State Examina- the densities of neurofibrillary tangles half was sectioned into 0.5- to 0.8-mm tion score within 1 year of death were (NFTs) derived from the CERAD bat- coronal slabs, flash frozen in liquid ni- considered (n = 14), then the correla- tery and counts of neuritic plaque (NP) trogen–cooled isopentane, and stored at tion between the consensus CDR and density15,16 in the regions of interest were −80°C. The neocortical regions dis- the last Mini-Mental State Examina- used for correlational analysis with the sected for ChAT and AChE analysis cor- tion score rose to r = −0.73 (P = .003). activity of cholinergic marker en- responded to the middle frontal gyrus zymes. It should be noted that the neu- (Brodmann area 8); inferior frontal gy- Neuropathological Assessment ropathological measures were ob- rus (Brodmann area 44); anterior cin- The procedures used for the neuro- tained from examination of the right half gulate gyrus (Brodmann area 32); su- pathological assessments have been de- of the brain whereas the neurochemi- perior, middle, and inferior temporal gyri scribed previously.15,16 Initially, all sub- cal measures were obtained from corti- (Brodmann areas 22, 21, and 20, respec- jects with non-AD neuropathology or cal specimens derived from the left hemi- tively); the entorhinal cortex (Brod- AD neuropathology complicated with sphere. mann area 36/28); the inferior parietal

Downloaded From: https://jamanetwork.com/ on 09/23/2021 CHOLINERGIC MARKERS IN EARLY AD

lobule (Brodmann area 7); and the pri- women in the different CDR groups did were statistically significant (PϾ.10 for ␹2 mary visual cortex (Brodmann area 17). not differ significantly ( 4 = 0.8; P = .82). all). The relationship between CDR The dissections of these regions were Subjects were grouped purely on the scores and AChE activity was nearly based on cortical maps similar to those basis of the CDR scores described identical to the relationship between published by Damasio and Damasio18 herein, without regard to neuropatho- CDR scores and ChAT activity (data not and were similar to the procedures de- logical diagnosis of AD. However, the shown). The significant differences scribed previously.5 One aliquot (ap- distribution of subjects on the basis of among cortical regions were attribut- proximately 100 mg) from each brain re- neuropathological diagnoses was able to higher ChAT activity in Brod- gion of each subject was used for the roughly similar to their distribution mann areas 21 and 36, higher AChE ac- ChAT and AChE activity assays. For 1 along the cognitive dimension as de- tivity in Brodmann areas 20, 21, 22, and subject with a CDR score of 0.0, corti- termined by the CDR score. The num- 36 relative to all other areas (PϽ.04 for cal tissue was available from only 3 Brod- ber of subjects in each CDR category all), and lower ChAT and AChE activ- mann areas (20, 21, and 22). The pro- receiving neuropathological diag- ity in the primary visual cortex relative cedures for the ChAT and AChE activity noses of definite, probable, possible, and to other cortical regions (PϽ.003 for all). assays were identical to those already de- no AD is shown in TABLE 2. scribed5,19 and were modified from the Correlation of ChAT and AChE procedures described by Fonnum20 and Activity of ChAT and AChE Activity With Cognitive Status Johnson and Russell,21 respectively. The as a Function of Cognitive Status The activity of ChAT correlated signifi- activity of ChAT and AChE was ex- The activity of ChAT and AChE in the cantly with CDR scores (range, r = −0.46 pressed as a function of protein concen- 9 cortical regions studied are shown in [Brodmann area 32] to r = −0.65 [Brod- tration that was estimated by the method FIGURE 1 and FIGURE 2. Analysis of vari- mann area 22]; PϽ.001 for all after Bon- of Bradford.22 ance of ChAT activity revealed a signifi- ferroni correction) when the entire co- cant effect of CDR groups (F4,74 = 17.57; hort was considered. However, when Data Analyses PϽ.001), a significant effect of brain re- subjects with CDR scores of 5.0 were ex- Ͻ The 5 CDR categories were used as the gions (F8,592 = 25.28; P .001), and a sig- cluded from the analysis, the ChAT ac- independent variable for subsequent nificant CDR group-by-brain regions in- tivity did not correlate significantly with Ͻ analyses. The dependent variables con- teraction (F32,592 = 2.49; P .001). CDR scores in any of the cortical re- sisted of the activity of ChAT and AChE Analysis of the effect of CDR showed that gions studied (range, r = −0.07 [Brod- in each of the 9 cortical regions. Re- the CDR 5.0 group differed signifi- mann area 8] to r = −0.27 [Brodmann peated-measures analyses of variance cantly from all other groups (PϽ.001 for area 20]; PϾ.05 for all after Bonferroni were used to analyze the activity of ChAT all) in all brain regions (PϽ.001 for all). correction). FIGURE 3 shows the corre- and AChE across cortical regions. Tukey No other differences among CDR groups lation of ChAT activity in the superior tests were used for between-group com- parisons. Pearson product-moment and Spearman rank-order correlations pro- Table 1. Demographic Characteristics of the Study Subjects cedures were used to calculate the cor- Sex, No. Clinical Dementia (Age, Mean, y) relation of cholinergic marker enzyme Rating Scale Sample Postmortem Interval, Age, Mean (SD) activity in different cortical regions with Score Size, No. Mean (SD), h [Range], y Male Female CDR scores, NP densities, and ratings of 0.0 18 8.29 (5.96) 83.8 (9.9) [64-99] 3 (82.3) 15 (84.1) NFT densities. For the correlational 0.5 11 5.59 (4.63) 85.8 (8.3) [69-94] 2 (77.5) 9 (87.7) analyses, Bonferroni correction was ap- 1.0 22 4.79 (3.97) 82.9 (8.2) [74-103] 6 (88.0) 16 (89.7) plied where indicated by multiplying the 2.0 15 6.09 (6.04) 89.1 (5.7) [74-97] 3 (83.7) 12 (90.4) P value calculated by the total number 5.0 15 5.84 (7.39) 85.4 (10.3) [62-103] 3 (69) 12 (89.0) of analyses performed in that series. All groups 81 6.17 (5.07) 86.7 (8.9) [62-103] 17 (81.6) 64 (88.1) RESULTS The demographic characteristics of the Table 2. Neuropathological Diagnoses of Subjects Within Each CDR Group* final study cohort are presented in No. of Subjects CDR Scale TABLE 1. Groups formed on the basis of Score Non-AD Definite AD Probable AD Possible AD Total CDR scores did not differ significantly 0.0 13 0 1 4 18 Ͼ with respect to age (F4,76 = 1.4; P .25). 0.5 4 1 0 6 11 Although there were significantly 1.0 7 11 1 3 22 (PϽ.006) more women (n = 52) than 2.0 0 11 3 1 15 men (n = 14) in the study cohort as a 5.0 0 15 0 0 15 whole, the proportion of men and *CDR indicates Clinical Dementia Rating; AD, Alzheimer disease.

Downloaded From: https://jamanetwork.com/ on 09/23/2021 CHOLINERGIC MARKERS IN EARLY AD

temporal gyrus (Brodmann area 20) with 22), the inferior parietal lobule (Brod- AChE and the NFT densities in each cor- CDR scores for the entire study cohort. mann area 7), and the primary visual cortical region (range, r = −0.26 to r = −0.63 Identical relationships were observed tex (Brodmann area 17). When the en- by Spearman rank-order correlation; when correlations between AChE activ- tire cohort was used, the density of PϽ.02 for all). FIGURE 5 shows the cor- ity and CDR scores were considered. plaques in each cortical region correlated relation of NFT densities in the superior significantly with the activity of ChAT temporalcortexwiththeactivityofChAT Correlations of ChAT and AChE andAChEinthatregion(range, r = −0.27 in Brodmann area 22. For NPs, the sig- Activity With Severity to r = −0.47 by Pearson product-moment nificant correlations between NPs and of Neuropathological Lesions correlation;PϽ.01 for all).FIGURE 4 pro- the activity of ChAT and AChE were due As part of the neuropathological assess- vides an example of 1 of these correla- in large part to the inclusion of subjects ment of the subjects, the density of NPs15 tions, in which the NP densities in the with CDR scores of 5.0. When the CDR and NFTs16 was determined in several superior temporal gyrus are plotted 5.0 group was excluded from the analy- cortical regions, including the middle againsttheactivityofChATinBrodmann ses, only the correlation between ChAT frontal gyrus (Brodmann area 8), the su- area 22. Similar correlations were ob- activity and NP density in the middle perior temporal gyrus (Brodmann area served between the activity of ChAT and frontal gyrus was significant (r = −0.34; P = .03 after Bonferroni correction). When a similar analysis was performed Figure 1. Activity of ChAT in 9 Cerebrocortical Regions as a Function of Dementia Severity for NFT densities (eliminating the CDR

25 5.0 cohort), only the correlation between BA 8 BA 22 ChAT activity and NFT density in the BA 32 BA 36 BA 44 BA 7 superior temporal gyrus remained sta- 20 BA 20 BA 17 BA 21 tistically significant (r = −0.40, P = .03 after Bonferroni correction). 15 Activity of ChAT and AChE in Subjects With Neuropathological 10 Protein per h Protein Diagnosis of AD The activity of ChAT and AChE was sig- 5 nificantly reduced if subjects were

Acetylcholine, Mean (SEM), nmol/mg of grouped purely on the basis of neuropathological diagnosis. Analysis of vari- 0.0 0.5 1.0 2.0 5.0 Clinical Dementia Rating Scale Score ance for the activity of ChAT in the 9 cortical regions of subjects neuropatho- Relative to the group without dementia (Clinical Dementia Rating [CDR] Scale score = 0.0), the activity of choline acetyltransferase (ChAT) was significantly reduced (PϽ.001 for all) in the CDR 5.0 group only. BA indicates Brodmann area. Figure 3. Correlation of Choline Acetyltransferase Activity in the Inferior Figure 2. Activity of AChE in 9 Cerebrocortical Regions as a Function of Dementia Severity Temporal Gyrus (Brodmann Area 20) With Clinical Dementia Rating Scale Scores 425 for the Entire Cohort 400 BA 8 BA 22 BA 32 BA 36 375 BA 44 BA 7 30 350 BA 20 BA 17 325 BA 21 300 25 275 250 20 225 200 15 175 Protein per h Protein 150 10 125 100 75 5 50 Acetylcholine, Mean (SEM), nmol/mg of 25 0 Acetylcholine, nmol/mg of Protein per h Acetylcholine, nmol/mg of Protein 0.0 0.5 1.0 2.0 5.0 Clinical Dementia Rating Scale Score 0 0.5 1.0 2.0 4.0 5.0 Clinical Dementia Rating Scale Score Relative to the group without dementia (Clinical Dementia Rating [CDR] Scale score = 0.0), the activity of acetylcholinesterase (AChE) was significantly reduced (PϽ.001 for all) in the CDR 5.0 group only. BA indicates For cohort, r = 0.58 and PϽ.001. Regression line de- Brodmann area. picts correlation for entire cohort.

Downloaded From: https://jamanetwork.com/ on 09/23/2021 CHOLINERGIC MARKERS IN EARLY AD

logically characterized as not having AD Figure 4. Correlation of Choline Figure 5. Correlation of Choline vs definitely having AD revealed a sig- Acetyltransferase Activity in the Superior Acetyltransferase Activity in the Superior nificant effect of diagnosis (F1,59 = 11.35; Temporal Gyrus (Brodmann Area 22) Temporal Gyrus (Brodmann Area 22) P = .002). Choline acetyltransferase ac- With Neuritic Plaque Density With Neurofibrillary Tangle Density Score tivity was significantly reduced in Brod- for the Entire Cohort for the Entire Cohort

mann areas 8, 20, 21, 22, and 36 in the 30 30 group neuropathologically diagnosed as definitely having AD relative to neuro- 25 25 Ͻ pathologically healthy controls (P .03 20 20 for all). Identical results were obtained for AChE activity. However, 15 15

when the regional activity of ChAT and 10 10 AChE in the 13 neuropathologically healthy elderly subjects with CDR 5 5 scores of 0.0 were compared with the 0 0

regional activity of ChAT and AChE in per h Acetylcholine, nmol/mg of Protein per h Acetylcholine, nmol/mg of Protein subjects with CDR scores of 0.5 and 1.0 0 10 20 30 40 50 0 135 and neuropathological diagnoses of Neuritic Plaque Density, mm2 Neurofibrillary Tangle Density Score definite and probable AD, no signifi- For cohort, r = 0.43 and PϽ.001. For cohort, r = 0.59 and PϽ.001. cant group differences were found (F1,23 = 0.5; P = .48 for ChAT and F1,23 = 0.52; P = .48 for AChE). The ac- toward lowered cholinergic markers at ChAT and AChE activity in which sta- tivity of ChAT and AChE correlated sig- this level of dementia. tistically significant deficits are appar- nificantly with each other in each cor- It is noteworthy that pathological ently not seen, in this sample, until a CDR tical region. The overall correlation markers did not robustly distinguish those score of greater than 2.0 is reached. These between ChAT and AChE activity patients who have higher ChAT or AChE data suggest that plaques are an earlier was + 0.67. Regional analysis showed activity in the groups with CDR scores of marker of AD than cholinergic mark- that the correlation between AChE and 0.0, 1.0, and 2.0 from those subjects who ers. In further support of this conclu- ChAT activity was highest in Brod- have lower ChAT activity. When the sub- sion, although ChAT or AChE activity mann area 21 of the temporal cortex jects in the groups with CDR scores of 0.0 and NP density correlated significantly (r = −0.81; PϽ.001) and lowest in Brod- to 2.0 were divided into those with mean with each other (Figure 4), these corre- mann area 36 of the temporal cortex cortical NP densities of less than 10/ lations did not reach statistical signifi- (r = −0.53; PϽ.001). mm2 vs those with at least 10/mm2 (Kha- cance without the inclusion of the CDR chaturian23 criteria for AD in subjects aged 5.0 group. Thus, among the distribu- COMMENT 66-75 years), the regional activity of ChAT tion of controls and patients with ques- There is little evidence in the current was slightly lower in the high-plaque tionable, definite but early, and moder- sample that deficits in ChAT or AChE group relative to the low-plaque group ate AD cognitive deficits, no relationship are early markers of AD. Only the CDR (mean ChAT activity, 12.7 vs 10.7 nmol with cholinergic markers and NP den- 5.0 group differed from the other CDR of acetylcholine per milligram of pro- sity was found in any region studied. groups in ChAT or AChE activity. Sub- tein per hour, respectively), but the dif- It is clear from these data that it is un- jects having a CDR score of 1.0, 50% of ferences did not approach statistical sig- likely that a cholinergic marker would whom met neuropathological criteria for nificance (F1,27 = 0.35; P = .56). These be an early indicator of AD, at least in pa- definite AD (Table 2), did not have cho- conclusions are supported further by the tients of this older age group. It is even linergic marker activity that was signifi- lack of significant correlation of the NP less likely that a cholinergic deficit could cantly lower than the control group. Even and NFT densities with the activity of be identified prior to patients becoming if only those patients with a CDR score ChAT and AChE in subjects with mild symptomatic. These observations have of 1.0 who meet neuropathological cri- and moderate dementia. Hence, the pres- important implications for the use of cho- teria for definite AD are considered, they ence of the neuropathological character- linomimetics and, particularly, AChE in- still did not have a significant reduction istics of AD is not a strong contributor to hibitors, in the treatment of AD. Most pa- in cholinergic markers (P = .81). Of even cholinergic function in subjects with mild tients included in study groups that test greater surprise is the failure of the group dementia. On the other hand, even mild the efficacy of cholinesterase inhibitors with a CDR score of 2.0 (patients with or questionable dementia was associ- would fall into the CDR score range of moderate AD dementia) to differ signifi- ated with significant increase in cortical 1.0 or 2.0 in the current study and would 15 16 cantly from the control group (F1,29 = 3.7; NP and NFT densities. This con- have a mean age in the early 70s. It is well P = .06), although there was a clear trend trasts rather substantially with the case for recognized that dementia with onset as

Downloaded From: https://jamanetwork.com/ on 09/23/2021 CHOLINERGIC MARKERS IN EARLY AD

early as the fifth and sixth decades of life Indeed, greater response to cholinomi- though the neocortical activity of ChAT differs from the more common kind of metics has been reported in patients with and AChE did not decrease in sub- AD found in the eighth or ninth de- more severe illness.27 However, such pa- jects with CDR scores of 0.5, 1.0, and cades of life on a number of important tients commonly have Mini-Mental State 2.0, some degeneration of the fore- dimensions, including heritability and Examination scores of less than 10 and brain cholinergic system still could have course.24,25 As a group, these younger pa- have not been extensively studied for the occurred. It is possible that a subset of tients have a modest response to cholin- efficacy of cholinomimetics. forebrain cholinergic neurons degen- esterase inhibition.8 Because they are con- Given the presence of NP and NFT erated but compensatory changes in the siderably younger than the current series pathological findings in mild demen- remaining neurons normalized neocor- of patients, they might have more ex- tia and the absence of a cholinergic defi- tical ChAT and AChE activity.28-30 De- tensive cholinergic lesions.3 Alterna- cit, the question remains as to the neu- tailed stereological studies of the fore- tively, such patients are reminiscent of rochemical nature of the cognitive brain cholinergic neurons will have to healthy young people who previously deficit. As noradrenergic and seroton- be performed to address this possibil- have been shown to respond to physo- ergic deficiencies are more commonly ity conclusively. stigmine with an increase in their abil- seen in younger patients and the cur- Although neocortical cholinergic defi- ity to learn new information.26 It is pos- rent study population constituted a cits are characteristic of severely de- sible that the efficacy of cholinomimetics group with a mean age beyond that at mented AD patients, in this study, cho- in patients with AD of only mild to mod- which noradrenergic and serotoner- linergic deficits were not detected in erate degree results from boosting cho- gic abnormalities are found, it is un- individuals with mild AD and were not linergic activity to levels higher than the likely that these neurotransmitters present until relatively late in the course patients’ premorbid baseline rather than could have been the substrate for the of the disease. These results suggest that from reversing a cholinergic deficit. On cognitive deficits in the population with patients with more severe disease should the other hand, these data certainly sug- mild dementia. However, it is possible be a target for cholinergic treatment. gest that more robust effects with cho- that corticotropin-releasing factor, so- linomimetics might be found in pa- matostatin, or a subgroup of glutama- Funding/Support: This work was supported by grants tients with more severe disease, ie, tergic neurons might be affected in these AG02219 and AG05138 from the National Institutes patients with CDR scores of at least 2.0. early AD groups.5 In addition, al- of Health, Bethesda, Md.

REFERENCES 1. Perry EK, Perry RH, Blessed G, Tomlinson BE. Nec- RA, Savoy RN. Cortical biopsy results in Alzheimer’s 21. Johnson CD, Russell RL. A rapid, simple radiomet- ropsy evidence of central cholinergic deficits in senile disease: correlation with cognitive deficits. Neurol- ric assay for acetylcholinesterase, suitable for multiple dementia. Lancet. 1977;1:189. ogy. 1987;37:1201-1204. determinations. Anal Biochem. 1975;64:229-232. 2. Davies P, Maloney AJF. Selective loss of central cho- 11. Hughes CP, Berg L, Danziger WL, Coben LA, Mar- 22. Bradford MM. A rapid and sensitive method for linergic neurons in Alzheimer’s disease. Lancet. 1976; tin RL. A new clinical scale for the staging of demen- the quantitation of microgram quantities of protein 2:1403. tia. Br J Psychol. 1982;140:566-572. utilizing the principle of protein-dye binding. Anal Bio- 3. Bowen DM, Benton JS, Spillane JA, Smith CC, Allen 12. Morris JC. The Clinical Dementia Rating (CDR): chem. 1976;72:248-254. SJ. Choline acetyltransferase activity and histopathol- current version and scoring rules. Arch Neurol. 1993; 23. Khachaturian ZS. Diagnosis of Alzheimer’s dis- ogy of frontal neocortex from biopsies of demented 43:2412-2414. ease. Arch Neurol. 1985;42:1097-1105. patients. J Neurol Sci. 1982;57:191-202. 13. Folstein M, Folstein S, McHugh P. “Mini-Mental 24. Li G, Silverman JM, Smith CJ, et al. Age at onset 4. Perry EK, Tomlinson BE, Blessed G, Bergmann K, State”: a practical method for grading the cognitive and familial risk in Alzheimer’s disease. Am J Psychia- Gibson PH, Perry RH. Correlation of cholinergic ab- state of patients for the clinician. J Psychiatr Res. 1975; try. 1995;152:424-430. normalities with senile plaques and mental test scores 12:189-198. 25. Silverman JM, Li G, Zaccario ML, et al. Patterns of in senile dementia. BMJ. 1978;2:1457-1459. 14. Rosen WG, Mohs RC, Davis KL. A new rating scale risk in first-degree relatives of patients with Alzheimer’s 5. Bierer LM, Haroutunian V, Gabriel S, et al. Neu- for Alzheimer’s disease. Am J Psychiatry. 1984;141: disease. Arch Gen Psychiatry. 1994;51:577-588. rochemical correlates of dementia severity in Alzhei- 1356-1364. 26. Davis KL, Mohs RC, Tinklenberg JR, Pfeffer- mer’s disease: relative importance of the cholinergic 15. Haroutunian V, Perl DP, Purohit DP, et al. Re- baum A, Hollister LE, Kopell BS. Physostigmine: im- deficits. J Neurochem. 1995;64:749-760. gional distribution of neuritic plaques in nonde- provement of long-term memory processes in nor- 6. Santucci AC, Haroutunian V, Tsuboyama GK, Kanof mented elderly and cases of very mild Alzheimer’s dis- mal humans. Science. 1978;201:272-274. PD, Davis KL. Therapeutics of Alzheimer’s disease for ease. Arch Neurol. 1998;55:1185-1191. 27. Farlow MR, Brashear H, Hui S, Schneider LS, Un- clinical and pre-clinical issues. In: Iqbal K, Wisniewski 16. Haroutunian V, Purohit DP, Perl DP, et al. Neuro- verzagt F, for the Tacrine Study Group. The effects HM, Winblad B, eds. Progress in Clinical and Biologi- fibrillary tangles in nondemented elderly and very mild of tacrine in patients with mild versus moderate stage cal Research. New York, NY: Alan R Liss Inc; 1989: Alzheimer’s disease. Arch Neurol. In press. Alzheimer’s disease. In: Iqbal K, Mortimer JA, Win- 1111-1120. 17. Mirra SS, Heyman A, McKeel D, et al. The Con- blad B, Wisnieski H, eds. Research Advances in Al- 7. Mohs RC, Davis BM, Johns CA, et al. Oral physo- sortium to Establish a Registry for Alzheimer’s Dis- zheimer’s Disease and Related Disorders. New York, stigmine treatment of patients with AD. AmJPsy- ease (CERAD), II: standardization of the neuropatho- NY: John Wiley & Sons Inc; 1995:284-292. chiatry. 1985;142:28-33. logic assessment of Alzheimer’s disease. Neurology. 28. Bartus RT, Pontecorvo MJ, Flicker C, Dean RL, Figueir- 8. Davis KL, Thal LT, Gamzu ER, et al. A double- 1991;41:479-486. edo JC. Behavioral recovery following bilateral lesions blind, placebo-controlled multicenter study of ta- 18. Damasio H, Damasio AR, eds. Lesion Analysis in of the nucleus basalis does not occur spontaneously.Phar- crine for Alzheimer’s disease. N Engl J Med. 1992; Neuropsychology. New York, NY: Oxford University macol Biochem Behav. 1986;24:1287-1292. 327:1253-1259. Press; 1989:184-221. 29. Wenk GL, Olton DS. Recovery of neocortical cho- 9. DeKosky ST, Harbaugh RE, Schmitt FA, et al. Cor- 19. Haroutunian V, Davidson M, Kanof PD, et al. Cor- line acetyltransferase activity following ibotenic acid tical biopsy in Alzheimer’s disease: diagnostic accu- tical cholinergic markers in schizophrenia. Schizophr injection into the nucleus basalis of Meynert in rats. racy and neurochemical, neuropathological, and cog- Res. 1994;12:137-144. Brain Res. 1984;293:184-186. nitive correlations: Intraventricular Bethanecol Study 20. Fonnum F. A rapid radiochemical method of de- 30. Cotman C, Schieff S. Compensatory synapse Group. Ann Neurol. 1992;32:625-632. termination of choline acetyltransferase. J Neuro- growth in aged animals after neuronal death. Mech 10. Martin EM, Wilson RS, Penn RS, Fox JH, Clasen chem. 1975;24:407-409. Ageing Dev. 1979;9:103-117.

Downloaded From: https://jamanetwork.com/ on 09/23/2021