- Email: [email protected]
Diagnostic Criteria for Alzheimer’s Disease
Neurobiology of Aging, Vol. 18, No. S4, pp. S67–S70, 1997 Copyright © 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0197-4580/97 $17.00 1 .00
PII:S0197-4580(97)00072-9
Diagnostic Criteria for Alzheimer’s Disease J. L. PRICE1 Department of Anatomy and Neurobiology, and The Alzheimer’s Disease Research Center, Washington University School of Medicine, St. Louis, MO 63110 PRICE, J. L. Diagnostic criteria for Alzheimer’s disease. NEUROBIOL AGING 18(S4) S67–S70, 1997.—This paper summarizes changes that distinguish early Alzheimer’s disease (AD) from nondemented aging based on 49 well characterized cases (30 nondemented, 10 very mildly demented, and 9 severely demented). Tangles were found in all nondemented cases (aged 54 to 88) concentrated in limbic structures. The probability of high tangle density increases with age, even in the absence of plaques or dementia. Based on plaques, nondemented cases can be divided into three groups: 1) cases younger than 73 years of age with one-third of older cases had no plaques; 2) about one-half of cases over 74 years of age had a few diffuse plaques in restricted patches in the neocortex; 3) about one-quarter of cases over 74 years of age had many neuritic and diffuse plaques throughout the neocortex; these may represent “preclinical” AD. Very mildly demented cases had high concentrations of neuritic and diffuse plaques in the neocortex and tangles in limbic structures. The observations indicate that the minimal diagnostic criterion for AD is plaques throughout the neocortex together with neurofibrillary changes (tangles in limbic structures and neuritic plaques in cortex). Tangles are a necessary but not sufficient criterion. © 1997 Elsevier Science Inc. Tangles
Plaques
Aging
Pre-clinical Alzheimer’s disease
THIS paper primarily addresses the question of the pathoanatomical changes that distinguish late onset, supposedly sporadic, Alzheimer’s disease (AD) from “healthy,” nondemented aging. It represents the result of investigations by, and discussions with, several researchers from the Alzheimer’s Disease Research Center at Washington University School of Medicine in St. Louis, especially Drs. Leonard Berg, John Morris, Dan McKeel, Eugene Rubin, and Martha Storandt. In order to distinguish healthy aging from Alzheimer’s disease, it is necessary to develop clinical or psychometric methods that can recognize the earliest stage of dementia. This is essential for the identification of a group of cases in the earliest, very mild stage of Alzheimer’s disease. It is even more important for identification of a pure group of nondemented “control” cases that is not contaminated by inclusion of cases with very mild dementia. From an analysis of a pure set of nondemented cases, the pathoanatomical changes related to age alone can be defined. In turn, a comparison between age-related changes and those found in very mildly demented cases can be used to define the changes that are uniquely characteristic of AD. A secondary problem for the development of consistent diagnostic criteria for AD is the variation between different staining methods and different laboratories. This can produce large differences in numbers of senile plaques, neurofibrillary tangles, or other markers counted within the same structure (6,7). Although this problem may be partially solved by developing standardized staining methods and standardized criteria for recognition of tangles and plaques, there is a need for qualitative criteria that can be applied without strict quantitative limits. The observations presented here were based on a modified Bielschowsky stain that reliably demonstrates both diffuse and neuritic or cored plaques.
Clinical dementia rating
The density of tangles and plaques was determined from large sections with a computer aided microscope digitizer, as described previously (12). ACCURATE DIAGNOSIS OF VERY MILD AD
In our experience, the most accurate and sensitive method for premortem recognition of the very mild dementia is informant based clinical evaluation, structured to give a Clinical Dementia Rating (CDR) (2,9). In addition to the history and relatively simple tests of memory and the ability to do calculations, an essential part of these evaluations is an interview with an informant who has known the subject well (e.g., a spouse or adult child). Most importantly, the informant interview can provide information about changes in behavior or cognitive ability over time. In the hands of experienced and careful neurologists, this type of evaluation has proved to be very successful in detecting the very mildest stage of dementia (CDR 5 0.5) and in staging more severe stages (CDR 5 1 to 3). In cases where subjects have died and come to autopsy while still in the very mild CDR 5 0.5 stage, the brains have all had substantial numbers of plaques and tangles, sufficient to warrant a diagnosis of AD. Furthermore, this was also true for cases in which the clinical evaluation gave only a questionable indication of cognitive decline (CDR 5 0/0.5) (11). Because of the value of the informant interview, it has also been possible to obtain a reliable indication of dementia with a retrospective informant interview administered after the death of the subject (3). This has allowed brains to be collected and studied when premortem data was not available. Neuropsychological tests can also be useful in providing a less subjective basis for early diagnosis, but only if administered at more than one time point so the subject’s performance can be
1 Address correspondence to: Joseph L. Price, Department of Anatomy and Neurobiology, Campus Box 8108, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO 63110.
S67
S68
PRICE TABLE 1 AVERAGE TANGLE DENSITIES Group
n
Hippo. CA1
Entorhinal Ctx
Perirhinal Ctx
Area 20
Area 22
Non-demented (60’s) Non-demented (70’s) Non-demented (80’s) Very mild dementia Severe dementia
5 11 13 10 9
0 (0) 6.2 (0.1–12.6) 6.2 (0.2–30) 24 (1.7–61) 74 (8.7–141)
0.2 (0–0.9) 5.3 (0.1–21) 6.9 (0.5–20) 20 (8.1–41) 49 (6.8–120)
0.2 (0–0.4) 4.9 (0.1–22) 6.6 (0.5–17) 21 (3.3–42) 51 (8.7–117)
0 (0) 0.8 (0–7.8) 0.7 (0–3.9) 6.4 (0.4–18) 41 (8.7–117)
0 (0) 0.2 (0–2.4) 0.1 (0–0.1) 0.4 (0–0.1) 23 (2.2–84)
Average tangle density per mm2 (range in parentheses) in hippocampal field CA1 (Hippo. CA1), the entorhinal cortex (Ctx), perirhinal cortex, inferior temporal cortex (Area 20) and superior temporal cortex (Area 22) as a function of decade of age (in nondemented cases; CDR 5 0) or severity of dementia. The very mildly demented cases were rated CDR 5 0.5 or 0/0.5 (average age 85, range 78 –95); the severely demented cases were rated CDR 5 3 (average age 80, range 62–95).
monitored over time (15). Because of the substantial variation in performance between individuals, the range of “normal” values is too broad to allow the subtle changes characteristic of the very mildest stage of dementia to be detected by a single test. However, if a subject’s performance can be measured across time, so the test results at any one time can be compared with the previous performance of the same individual, and if several different tests are used, then neuropsychological tests can provide a reliable, objective, and highly sensitive indication of the earliest cognitive decline. For this to be useful in a general clinical setting, it would probably be necessary to administer routine tests at a relatively early age (e.g., 60 or 65), in order to have an index for comparison with subsequent tests at ages where the risk of dementia is greater. PATHOANATOMICAL CHANGES ASSOCIATED WITH AGING, WITHOUT DETECTABLE COGNITIVE DECLINE
Thirty subjects aged 54 to 89, in whom there was no indication of any dementing change and who did not have any other neurological disorder such as Parkinson’s disease, have been studied. The value of these cases is largely due to the quality of the clinical assessment that was available for them. They provide a pure nondemented group for definition of changes related to age alone, without contamination with very mildly demented cases. Observations on these cases have been described in several recent papers (2,7,10 –14). They will be briefly described here to provide a basis for comparison with very mildly demented cases. Tangles At least a few tangle-bearing neurons are found in all nondemented patients over the age of about 55– 60 (12). These are distributed in a very characteristic pattern in selectively vulnerable structures, including the entorhinal and perirhinal cortical areas and hippocampal field CA1. Within the entorhinal cortex, there is a characteristic laminar distribution, with the highest numbers in layers II and IV. There are very few tangles in the neocortex, although low densities of tangles may be found in areas situated near the hippocampus and other limbic structures, such as the inferior temporal cortex and the agranular insular areas. With increasing age (from 60 to 90), there is a progressive and highly significant increase in the density of tangles within the vulnerable structures, albeit with considerable individual variation [(13); Table 1]. In the youngest nondemented cases, the tangles are primarily found in the perirhinal and entorhinal cortex, and in slightly older cases they become concentrated in field CA1. There is also an age related increase in tangles in the inferior temporal cortex and other neocortical areas, although the numbers of tangles are very low in those areas. These increases are not seen in every
brain, however. Indeed, the most striking effect of age is an increase in the range and variance of tangle density in patients older than about 70 –75. Many older cases have very high numbers of tangles while others have very few. Plaques Unlike tangles, plaques are not found in all nondemented cases (12). There is an age effect in the prevalence of plaques (i.e., whether or not any plaques are present), but the number of plaques does not show any clear relation to age (13). In a series of 30 nondemented brains studied, the 9 cases aged less than 75 years contained no plaques, based on examination with either a sensitive modification of the Bielschowsky stain or with immunohistochemical stains for b-amyloid (Table 2). The 21 cases aged 75 years and older can be divided into three distinct groups based on plaque number, distribution, and type (Table 2). There is essentially no age difference between the groups. In the first group (7 of 21 cases, average age 81, including one case aged 88), no plaques could be detected in any region. The second group (9 of 21 cases, average age 81) had relatively small numbers of plaques, distributed in restricted patches in the cerebral cortex, which often occupy one bank of a sulcus or part of a gyrus. The plaques are all of the diffuse type, without thickened, dystrophic neurites or amyloid cores. In some cases, only a single patch of plaques can be detected, whereas in other cases several patches of plaques are found in many cortical areas. The average density of plaques across the cortex of these cases is very low (, 1 plaque/mm2), although it may be higher in the center of a patch of plaques. In the third group (5 of 21 cases, average age 80), more numerous plaques are found distributed extensively throughout the neocortex. Many of these plaques are neuritic or cored in type, although the majority of plaques are still diffuse. The average density of plaques in the cortex of each of these cases is much higher than in the first group (6 to 20 plaques/mm2), and the density may be still higher in specific regions (Table 2). As noted below, the severity of the pathological change in these cases approaches or even overlaps that seen in the very mildly demented cases; they may represent preclinical cases of AD. It is notable that there is very little difference in tangle density between the three groups of cases aged 75 and older, although the cases younger than 75 have fewer tangles. PATHOANATOMICAL CHANGES ASSOCIATED WITH VERY MILD OR QUESTIONABLE DEMENTIA
Ten cases have been studied that were assessed as either very mildly demented (CDR 5 0.5) or, when the status was less certain,
DIAGNOSTIC CRITERIA FOR ALZHEIMER’S DISEASE
S69 TABLE 2
PLAQUE DENSITIES
Group
n
Average age (range)
Non-demented, , 75 years old Non-demented, $ 75, no plaques Non-demented, $ 75, few plaques Non-demented, $ 75, many plaques Very mildly demented Severely demented
9 7 9 5 10 9
64 (54–73) 81 (75–88) 81 (75–89) 80 (75–88) 85 (78–95) 80 (62–95)
Plaques/mm2 Average (range)
none none 0.6 (0.1–1.4) 10.4 (4.8–20) 20 (9–38) 22 (11–42)
Plaque Distribution in cortex
patchy extensive extensive extensive
Plaque type
Tangle density, entorhinal cortex
all diffuse ca. 15% neuritic/cored ca. 15% neuritic/cored . 50% neuritic/cored
0.6 (0.1–3.5) 5.4 (0.1–15) 7.3 (0.5–20) 8.2 (0.7–21) 20 (8–41) 49 (7–120)
Plaque density distribution and type in groups of aging nondemented and demented cases. The nondemented cases were grouped by age around 75 years. Those cases 75 and older were grouped into those with no plaques, few plaques, and many plaques. Note that there is no overlap between the nondemented groups, although there is overlap between the nondemented cases with many plaques and the very mildly demented cases. For comparison, the tangle density in the entorhinal cortex (average and range) is given for each group of cases; there is complete overlap in tangle concentration between the three groups of nondemented cases aged 75 and older.
as questionably demented (CDR 5 0/0.5). All but one of these had their final assessment within a year before death; the other case was assessed with the retrospective informant interview. None had any neurological disorder other than AD that might have contributed to the cognitive decline (e.g., Parkinson’s disease). Together they provide a remarkable resource for analysis of the pathoanatomical changes associated with the earliest stage of AD. Observations on these cases have been reported in several papers (2,6,10 –13), and they will be briefly described here. All of the cases have sufficient pathological changes to meet the 1985 National Institute on Aging (NIA) criteria for AD (5,8,11). Tangles Large numbers of tangles are found in all of these cases, especially in the areas that have many tangles in the nondemented cases (e.g., hippocampal field CA1 and the entorhinal and perirhinal cortical areas; Table 1 and refs. (10,12)). These same areas also have substantial numbers of paired helical filament containing “neuropil threads.” Neocortical areas near the hippocampus and other limbic structures (e.g., the inferior temporal cortex) also have substantial numbers of tangles, although other areas (e.g., the superior temporal cortex) still have relatively few tangles. The relative pattern of distribution of tangles is therefore the same as in the nondemented cases, although the very mildly demented cases, as a group, have substantially greater density of tangles. On an individual case basis, however, there is considerable overlap between the density of tangles in nondemented cases over 70 and in very mildly demented cases. For example, in five of the nondemented cases aged between 70 and 89 the tangle density in the entorhinal cortex is greater 15/mm2, whereas in five of the very mildly demented cases it is less than 15/mm2. The correlation between tangle density and the earliest stage of dementia, therefore, is limited. There is apparently a better correlation between very mild dementia and cell loss. In a recent study of cell number in the entorhinal cortex in a subset of these same cases, no cell loss was found as a function of age in nondemented cases (4). In the very mildly demented and questionably demented cases, however, 32% of the neurons in the entorhinal cortex were lost, and in layer II the loss was as great as 60%. In severely demented cases of AD, the number and density of tangles is still greater. There are also many more tangles in cortical regions such as the area 22 in the superior temporal cortex (Table 1).
Plaques As compared to most of the nondemented cases, the most striking difference in all of the very mildly demented and questionably demented cases is the large number of plaques that are distributed widely throughout the cerebral cortex (6,10,14). The plaque density throughout the cortex averages 20/mm2 across this group of cases, and it is at least 9/mm2 in each brain (Table 2). All of the brains had still higher concentrations in localized cortical regions. Most of the plaques are diffuse in type, but there are also substantial numbers of neuritic or cored plaques. There are also plaques in the hippocampus and the entorhinal and perirhinal cortical areas, although fewer than in the neocortex. In severely demented cases of AD, the total number of plaques is very similar to that seen in the very mildly demented cases (Table 2). The type of plaque changes differs, however, so that the majority of plaques have thickened, dystrophic neurites or amyloid cores, with a commensurate decrease in the number of diffuse plaques. CONCLUSIONS FOR THE DIAGNOSIS OF AD
The observations summarized above suggest several features that characterize the earliest, very mild stage of AD, which presumably constitute the minimal criteria for diagnosis of AD. 1. There are substantial numbers of plaques widely distributed throughout many parts of the cerebral cortex (at least 10/mm2 in some areas). The majority of these plaques are diffuse in type, without thickened neurites or amyloid cores, but there are also substantial numbers of neuritic or cored plaques. The presence of such plaques may be the best indication of the presence of the disease process underlying AD. Plaque formation is not a ubiquitous consequence of aging, at least into the mid-to-late 80’s. Although some nondemented cases have substantial numbers of plaques, including some neuritic or cored plaques, they probably represent preclinical cases of AD (see below). 2. There is a high concentration of neurofibrillary tangles in the hippocampus and adjacent structures, especially hippocampal field CA1 and the entorhinal and perirhinal cortices. The average tangle density in the entorhinal cortex is at least 5/mm2 and is much higher in layers II and IV. Although these features are necessary characteristics of early AD, by themselves, they are not sufficient for diagnosis because tangle concentration tends to increase with age. Similar concentrations of tangles can
S70
PRICE
be found in these areas in many nondemented aged cases, including those with relatively few or no plaques. 3. There are indications of cytoskeletal change in the neocortex as well as the hippocampal formation. The most prominent such indication are the dystrophic neurites within neuritic plaques. In addition, there are also at least some tangles and neuropil threads in the neocortex. These are most prominent in cortical regions near the hippocampus and other limbic areas such as the inferior temporal cortex. With increasing severity of dementia in the later stages of AD, there is an increase in tangle density. This is especially true in the vulnerable areas in and around the hippocampus but is also seen in the neocortex. Although the total number of plaques does not increase substantially, the proportion of neuritic and cored plaques does increase with increasing dementia severity. A corollary of these observations is that the very mildly demented and questionably demented cases, at the threshold for clinical detection of AD, already have very substantial neuropathological changes. Therefore, it is clear that the disease process must begin before any cognitive change can be detected clinically or
neuropsychologically. The relatively few nondemented cases with substantial numbers of both diffuse and neuritic/cored plaques, as well as many tangles in and around the hippocampus, provide the best candidates for such preclinical cases. As a group, the numbers of both plaques and tangles in these cases are lower than those found in very mildly demented cases. The qualitative features or pattern of pathology are the same in both groups, and there is a quantitative overlap in some cases. It is very likely that these cases would have progressed to full symptomatic AD if the patient had lived long enough. If this is so, then it is possible that the patches of diffuse plaques that are found in other aged nondemented cases represent an even earlier stage in the disease process, which would have progressed to the preclinical and then clinically detectable stages. ACKNOWLEDGEMENTS
I thank Mr. Hieu Van Luu, Ms. Melissa Rundle, Mr. David Wavek, and Mr. Karan Randhava for their excellent technical assistance. This work was supported by Grant P01 AG03991 from the United States Public Health Service/National Institute on Aging.
REFERENCES 1. Berg, L. Clinical dementia rating (CDR). Psychopharmacol. Bull. 25:637– 639; 1988. 2. Berg, L.; McKeel, D. W., Jr.; Miller, J. P.; Baty, J. Neuropathologic indexes of Alzheimer’s disease in demented and nondemented people aged 80 years and older. Arch. Neurol. 50:349 –358; 1993. 3. Davis, P. B.; White, H.; Price, J. L.; McKeel, D.; Robins, L. N. Retrospective postmortem dementia assessment: Validation of a new clinical interview to assist neuropathological study. Arch. Neurol. 48:613– 617; 1991. 4. Gomez-Isla, T.; Price, J. L.; McKeel, D. W.; Morris, J. C.; Growdon, J. H.; Hyman, B. T. Profound loss of layer II entorhinal cortex neurons distinguishes very mild Alzheimer’s disease from nondemented aging. J. Neurosci. 16:4491– 4500; 1996. 5. Katchaturian, Z. S. Diagnosis of Alzheimer’s disease. Arch. Neurol. 42:1097–1105; 1985. 6. McKeel, D. W., Jr.; Ball, M. J.; Price, J. L.; Smith, D. S.; Miller, J. P.; Berg, L.; Morris, J. C. Interlaboratory histopathologic assessment of Alzheimer neuropathology: Different methodologies yield comparable diagnostic results. Alzheimer. Dis. Assoc. Disord. 7:136 –151; 1993. 7. Mirra, S. S.; Heyman, A.; Berg, L.; McKeel, D.; Sumi, S. M.; Crain, B.; Brownlee, L.; Vogel, F. S.; Hughes, J. P.; van Belle, G.; participating CERAD neuropathologists. The Consortium to Establish a Registry For Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathological assessment of Alzheimer’s disease and other dementias. Neurology 41:479 – 486; 1991. 8. Mirra, S. S.; Gearing, M.; McKeel, D. W., Jr.; Crain, B. J.; Hughes, J. P.; vanBell, G.; Heyman, A.; and participating neuropathologists. Interlaboratorycomparison of neuropathologic assessments in Alzhei-
9. 10.
11.
12.
13.
14.
15.
mer’s disease: A study of the Consortium to Establish a Registry for Alzheimer’s disease (CERAD). J. Neuropathol. Exp. Neurol. 53:303– 315; 1994. Morris, J. C. The clinical dementia rating (CDR): Current version and scoring rules. Neurology 43:2412–2414; 1993. Morris, J. C.; McKeel, D.; Storandt, M.; Rubin, E. H.; Price, J. L.; Grant, E. A.; Ball, M. J.; Berg, L. Very mild Alzheimer’s disease: Informant-based clinical, psychometric, and pathological distinction from normal aging. Neurology 41:469 – 478; 1991. Morris, J. C.; Storandt, M.; McKeel, D. W.; Rubin, E. H.; Price, J. L.; Grant, E. A.; Berg, L. Cerebral amyloid deposition and diffuse plaques in “normal” aging: Evidence for presynptomatic and very mild Alzheimer’s disease. Neurology 46:707–719; 1996. Price, J. L.; Davis, P. B.; Morris, J. C.; White, D. L. The distribution of tangles, plaques and related immunohistochemical markers in healthy aging and Alzheimer’s disease. Neurobiol. Aging 12:295–312; 1991. Price, J. L. The relationship between tangle and plaque formation during healthy aging and mild dementia. Neurobiol. Aging 14:661– 663; 1993. Price, J. L. Tangles and plaques in healthy aging and Alzheimer’s disease: Independence or interaction? Semin. Neurosci. 6:395– 402; 1994. Rubin, E. H.; Storandt, M.; Miller, J. P.; Kinscherf, D. A.; Grant, E. A.; Morris, J. C.; Berg, L. A prospective study of cognitive function and onset of dementia in cognitively healthy elders. Arch. Neurol., in press; 1997.
PII:S0197-4580(97)00072-9
Diagnostic Criteria for Alzheimer’s Disease J. L. PRICE1 Department of Anatomy and Neurobiology, and The Alzheimer’s Disease Research Center, Washington University School of Medicine, St. Louis, MO 63110 PRICE, J. L. Diagnostic criteria for Alzheimer’s disease. NEUROBIOL AGING 18(S4) S67–S70, 1997.—This paper summarizes changes that distinguish early Alzheimer’s disease (AD) from nondemented aging based on 49 well characterized cases (30 nondemented, 10 very mildly demented, and 9 severely demented). Tangles were found in all nondemented cases (aged 54 to 88) concentrated in limbic structures. The probability of high tangle density increases with age, even in the absence of plaques or dementia. Based on plaques, nondemented cases can be divided into three groups: 1) cases younger than 73 years of age with one-third of older cases had no plaques; 2) about one-half of cases over 74 years of age had a few diffuse plaques in restricted patches in the neocortex; 3) about one-quarter of cases over 74 years of age had many neuritic and diffuse plaques throughout the neocortex; these may represent “preclinical” AD. Very mildly demented cases had high concentrations of neuritic and diffuse plaques in the neocortex and tangles in limbic structures. The observations indicate that the minimal diagnostic criterion for AD is plaques throughout the neocortex together with neurofibrillary changes (tangles in limbic structures and neuritic plaques in cortex). Tangles are a necessary but not sufficient criterion. © 1997 Elsevier Science Inc. Tangles
Plaques
Aging
Pre-clinical Alzheimer’s disease
THIS paper primarily addresses the question of the pathoanatomical changes that distinguish late onset, supposedly sporadic, Alzheimer’s disease (AD) from “healthy,” nondemented aging. It represents the result of investigations by, and discussions with, several researchers from the Alzheimer’s Disease Research Center at Washington University School of Medicine in St. Louis, especially Drs. Leonard Berg, John Morris, Dan McKeel, Eugene Rubin, and Martha Storandt. In order to distinguish healthy aging from Alzheimer’s disease, it is necessary to develop clinical or psychometric methods that can recognize the earliest stage of dementia. This is essential for the identification of a group of cases in the earliest, very mild stage of Alzheimer’s disease. It is even more important for identification of a pure group of nondemented “control” cases that is not contaminated by inclusion of cases with very mild dementia. From an analysis of a pure set of nondemented cases, the pathoanatomical changes related to age alone can be defined. In turn, a comparison between age-related changes and those found in very mildly demented cases can be used to define the changes that are uniquely characteristic of AD. A secondary problem for the development of consistent diagnostic criteria for AD is the variation between different staining methods and different laboratories. This can produce large differences in numbers of senile plaques, neurofibrillary tangles, or other markers counted within the same structure (6,7). Although this problem may be partially solved by developing standardized staining methods and standardized criteria for recognition of tangles and plaques, there is a need for qualitative criteria that can be applied without strict quantitative limits. The observations presented here were based on a modified Bielschowsky stain that reliably demonstrates both diffuse and neuritic or cored plaques.
Clinical dementia rating
The density of tangles and plaques was determined from large sections with a computer aided microscope digitizer, as described previously (12). ACCURATE DIAGNOSIS OF VERY MILD AD
In our experience, the most accurate and sensitive method for premortem recognition of the very mild dementia is informant based clinical evaluation, structured to give a Clinical Dementia Rating (CDR) (2,9). In addition to the history and relatively simple tests of memory and the ability to do calculations, an essential part of these evaluations is an interview with an informant who has known the subject well (e.g., a spouse or adult child). Most importantly, the informant interview can provide information about changes in behavior or cognitive ability over time. In the hands of experienced and careful neurologists, this type of evaluation has proved to be very successful in detecting the very mildest stage of dementia (CDR 5 0.5) and in staging more severe stages (CDR 5 1 to 3). In cases where subjects have died and come to autopsy while still in the very mild CDR 5 0.5 stage, the brains have all had substantial numbers of plaques and tangles, sufficient to warrant a diagnosis of AD. Furthermore, this was also true for cases in which the clinical evaluation gave only a questionable indication of cognitive decline (CDR 5 0/0.5) (11). Because of the value of the informant interview, it has also been possible to obtain a reliable indication of dementia with a retrospective informant interview administered after the death of the subject (3). This has allowed brains to be collected and studied when premortem data was not available. Neuropsychological tests can also be useful in providing a less subjective basis for early diagnosis, but only if administered at more than one time point so the subject’s performance can be
1 Address correspondence to: Joseph L. Price, Department of Anatomy and Neurobiology, Campus Box 8108, Washington University School of Medicine, 660 S. Euclid Avenue, St. Louis, MO 63110.
S67
S68
PRICE TABLE 1 AVERAGE TANGLE DENSITIES Group
n
Hippo. CA1
Entorhinal Ctx
Perirhinal Ctx
Area 20
Area 22
Non-demented (60’s) Non-demented (70’s) Non-demented (80’s) Very mild dementia Severe dementia
5 11 13 10 9
0 (0) 6.2 (0.1–12.6) 6.2 (0.2–30) 24 (1.7–61) 74 (8.7–141)
0.2 (0–0.9) 5.3 (0.1–21) 6.9 (0.5–20) 20 (8.1–41) 49 (6.8–120)
0.2 (0–0.4) 4.9 (0.1–22) 6.6 (0.5–17) 21 (3.3–42) 51 (8.7–117)
0 (0) 0.8 (0–7.8) 0.7 (0–3.9) 6.4 (0.4–18) 41 (8.7–117)
0 (0) 0.2 (0–2.4) 0.1 (0–0.1) 0.4 (0–0.1) 23 (2.2–84)
Average tangle density per mm2 (range in parentheses) in hippocampal field CA1 (Hippo. CA1), the entorhinal cortex (Ctx), perirhinal cortex, inferior temporal cortex (Area 20) and superior temporal cortex (Area 22) as a function of decade of age (in nondemented cases; CDR 5 0) or severity of dementia. The very mildly demented cases were rated CDR 5 0.5 or 0/0.5 (average age 85, range 78 –95); the severely demented cases were rated CDR 5 3 (average age 80, range 62–95).
monitored over time (15). Because of the substantial variation in performance between individuals, the range of “normal” values is too broad to allow the subtle changes characteristic of the very mildest stage of dementia to be detected by a single test. However, if a subject’s performance can be measured across time, so the test results at any one time can be compared with the previous performance of the same individual, and if several different tests are used, then neuropsychological tests can provide a reliable, objective, and highly sensitive indication of the earliest cognitive decline. For this to be useful in a general clinical setting, it would probably be necessary to administer routine tests at a relatively early age (e.g., 60 or 65), in order to have an index for comparison with subsequent tests at ages where the risk of dementia is greater. PATHOANATOMICAL CHANGES ASSOCIATED WITH AGING, WITHOUT DETECTABLE COGNITIVE DECLINE
Thirty subjects aged 54 to 89, in whom there was no indication of any dementing change and who did not have any other neurological disorder such as Parkinson’s disease, have been studied. The value of these cases is largely due to the quality of the clinical assessment that was available for them. They provide a pure nondemented group for definition of changes related to age alone, without contamination with very mildly demented cases. Observations on these cases have been described in several recent papers (2,7,10 –14). They will be briefly described here to provide a basis for comparison with very mildly demented cases. Tangles At least a few tangle-bearing neurons are found in all nondemented patients over the age of about 55– 60 (12). These are distributed in a very characteristic pattern in selectively vulnerable structures, including the entorhinal and perirhinal cortical areas and hippocampal field CA1. Within the entorhinal cortex, there is a characteristic laminar distribution, with the highest numbers in layers II and IV. There are very few tangles in the neocortex, although low densities of tangles may be found in areas situated near the hippocampus and other limbic structures, such as the inferior temporal cortex and the agranular insular areas. With increasing age (from 60 to 90), there is a progressive and highly significant increase in the density of tangles within the vulnerable structures, albeit with considerable individual variation [(13); Table 1]. In the youngest nondemented cases, the tangles are primarily found in the perirhinal and entorhinal cortex, and in slightly older cases they become concentrated in field CA1. There is also an age related increase in tangles in the inferior temporal cortex and other neocortical areas, although the numbers of tangles are very low in those areas. These increases are not seen in every
brain, however. Indeed, the most striking effect of age is an increase in the range and variance of tangle density in patients older than about 70 –75. Many older cases have very high numbers of tangles while others have very few. Plaques Unlike tangles, plaques are not found in all nondemented cases (12). There is an age effect in the prevalence of plaques (i.e., whether or not any plaques are present), but the number of plaques does not show any clear relation to age (13). In a series of 30 nondemented brains studied, the 9 cases aged less than 75 years contained no plaques, based on examination with either a sensitive modification of the Bielschowsky stain or with immunohistochemical stains for b-amyloid (Table 2). The 21 cases aged 75 years and older can be divided into three distinct groups based on plaque number, distribution, and type (Table 2). There is essentially no age difference between the groups. In the first group (7 of 21 cases, average age 81, including one case aged 88), no plaques could be detected in any region. The second group (9 of 21 cases, average age 81) had relatively small numbers of plaques, distributed in restricted patches in the cerebral cortex, which often occupy one bank of a sulcus or part of a gyrus. The plaques are all of the diffuse type, without thickened, dystrophic neurites or amyloid cores. In some cases, only a single patch of plaques can be detected, whereas in other cases several patches of plaques are found in many cortical areas. The average density of plaques across the cortex of these cases is very low (, 1 plaque/mm2), although it may be higher in the center of a patch of plaques. In the third group (5 of 21 cases, average age 80), more numerous plaques are found distributed extensively throughout the neocortex. Many of these plaques are neuritic or cored in type, although the majority of plaques are still diffuse. The average density of plaques in the cortex of each of these cases is much higher than in the first group (6 to 20 plaques/mm2), and the density may be still higher in specific regions (Table 2). As noted below, the severity of the pathological change in these cases approaches or even overlaps that seen in the very mildly demented cases; they may represent preclinical cases of AD. It is notable that there is very little difference in tangle density between the three groups of cases aged 75 and older, although the cases younger than 75 have fewer tangles. PATHOANATOMICAL CHANGES ASSOCIATED WITH VERY MILD OR QUESTIONABLE DEMENTIA
Ten cases have been studied that were assessed as either very mildly demented (CDR 5 0.5) or, when the status was less certain,
DIAGNOSTIC CRITERIA FOR ALZHEIMER’S DISEASE
S69 TABLE 2
PLAQUE DENSITIES
Group
n
Average age (range)
Non-demented, , 75 years old Non-demented, $ 75, no plaques Non-demented, $ 75, few plaques Non-demented, $ 75, many plaques Very mildly demented Severely demented
9 7 9 5 10 9
64 (54–73) 81 (75–88) 81 (75–89) 80 (75–88) 85 (78–95) 80 (62–95)
Plaques/mm2 Average (range)
none none 0.6 (0.1–1.4) 10.4 (4.8–20) 20 (9–38) 22 (11–42)
Plaque Distribution in cortex
patchy extensive extensive extensive
Plaque type
Tangle density, entorhinal cortex
all diffuse ca. 15% neuritic/cored ca. 15% neuritic/cored . 50% neuritic/cored
0.6 (0.1–3.5) 5.4 (0.1–15) 7.3 (0.5–20) 8.2 (0.7–21) 20 (8–41) 49 (7–120)
Plaque density distribution and type in groups of aging nondemented and demented cases. The nondemented cases were grouped by age around 75 years. Those cases 75 and older were grouped into those with no plaques, few plaques, and many plaques. Note that there is no overlap between the nondemented groups, although there is overlap between the nondemented cases with many plaques and the very mildly demented cases. For comparison, the tangle density in the entorhinal cortex (average and range) is given for each group of cases; there is complete overlap in tangle concentration between the three groups of nondemented cases aged 75 and older.
as questionably demented (CDR 5 0/0.5). All but one of these had their final assessment within a year before death; the other case was assessed with the retrospective informant interview. None had any neurological disorder other than AD that might have contributed to the cognitive decline (e.g., Parkinson’s disease). Together they provide a remarkable resource for analysis of the pathoanatomical changes associated with the earliest stage of AD. Observations on these cases have been reported in several papers (2,6,10 –13), and they will be briefly described here. All of the cases have sufficient pathological changes to meet the 1985 National Institute on Aging (NIA) criteria for AD (5,8,11). Tangles Large numbers of tangles are found in all of these cases, especially in the areas that have many tangles in the nondemented cases (e.g., hippocampal field CA1 and the entorhinal and perirhinal cortical areas; Table 1 and refs. (10,12)). These same areas also have substantial numbers of paired helical filament containing “neuropil threads.” Neocortical areas near the hippocampus and other limbic structures (e.g., the inferior temporal cortex) also have substantial numbers of tangles, although other areas (e.g., the superior temporal cortex) still have relatively few tangles. The relative pattern of distribution of tangles is therefore the same as in the nondemented cases, although the very mildly demented cases, as a group, have substantially greater density of tangles. On an individual case basis, however, there is considerable overlap between the density of tangles in nondemented cases over 70 and in very mildly demented cases. For example, in five of the nondemented cases aged between 70 and 89 the tangle density in the entorhinal cortex is greater 15/mm2, whereas in five of the very mildly demented cases it is less than 15/mm2. The correlation between tangle density and the earliest stage of dementia, therefore, is limited. There is apparently a better correlation between very mild dementia and cell loss. In a recent study of cell number in the entorhinal cortex in a subset of these same cases, no cell loss was found as a function of age in nondemented cases (4). In the very mildly demented and questionably demented cases, however, 32% of the neurons in the entorhinal cortex were lost, and in layer II the loss was as great as 60%. In severely demented cases of AD, the number and density of tangles is still greater. There are also many more tangles in cortical regions such as the area 22 in the superior temporal cortex (Table 1).
Plaques As compared to most of the nondemented cases, the most striking difference in all of the very mildly demented and questionably demented cases is the large number of plaques that are distributed widely throughout the cerebral cortex (6,10,14). The plaque density throughout the cortex averages 20/mm2 across this group of cases, and it is at least 9/mm2 in each brain (Table 2). All of the brains had still higher concentrations in localized cortical regions. Most of the plaques are diffuse in type, but there are also substantial numbers of neuritic or cored plaques. There are also plaques in the hippocampus and the entorhinal and perirhinal cortical areas, although fewer than in the neocortex. In severely demented cases of AD, the total number of plaques is very similar to that seen in the very mildly demented cases (Table 2). The type of plaque changes differs, however, so that the majority of plaques have thickened, dystrophic neurites or amyloid cores, with a commensurate decrease in the number of diffuse plaques. CONCLUSIONS FOR THE DIAGNOSIS OF AD
The observations summarized above suggest several features that characterize the earliest, very mild stage of AD, which presumably constitute the minimal criteria for diagnosis of AD. 1. There are substantial numbers of plaques widely distributed throughout many parts of the cerebral cortex (at least 10/mm2 in some areas). The majority of these plaques are diffuse in type, without thickened neurites or amyloid cores, but there are also substantial numbers of neuritic or cored plaques. The presence of such plaques may be the best indication of the presence of the disease process underlying AD. Plaque formation is not a ubiquitous consequence of aging, at least into the mid-to-late 80’s. Although some nondemented cases have substantial numbers of plaques, including some neuritic or cored plaques, they probably represent preclinical cases of AD (see below). 2. There is a high concentration of neurofibrillary tangles in the hippocampus and adjacent structures, especially hippocampal field CA1 and the entorhinal and perirhinal cortices. The average tangle density in the entorhinal cortex is at least 5/mm2 and is much higher in layers II and IV. Although these features are necessary characteristics of early AD, by themselves, they are not sufficient for diagnosis because tangle concentration tends to increase with age. Similar concentrations of tangles can
S70
PRICE
be found in these areas in many nondemented aged cases, including those with relatively few or no plaques. 3. There are indications of cytoskeletal change in the neocortex as well as the hippocampal formation. The most prominent such indication are the dystrophic neurites within neuritic plaques. In addition, there are also at least some tangles and neuropil threads in the neocortex. These are most prominent in cortical regions near the hippocampus and other limbic areas such as the inferior temporal cortex. With increasing severity of dementia in the later stages of AD, there is an increase in tangle density. This is especially true in the vulnerable areas in and around the hippocampus but is also seen in the neocortex. Although the total number of plaques does not increase substantially, the proportion of neuritic and cored plaques does increase with increasing dementia severity. A corollary of these observations is that the very mildly demented and questionably demented cases, at the threshold for clinical detection of AD, already have very substantial neuropathological changes. Therefore, it is clear that the disease process must begin before any cognitive change can be detected clinically or
neuropsychologically. The relatively few nondemented cases with substantial numbers of both diffuse and neuritic/cored plaques, as well as many tangles in and around the hippocampus, provide the best candidates for such preclinical cases. As a group, the numbers of both plaques and tangles in these cases are lower than those found in very mildly demented cases. The qualitative features or pattern of pathology are the same in both groups, and there is a quantitative overlap in some cases. It is very likely that these cases would have progressed to full symptomatic AD if the patient had lived long enough. If this is so, then it is possible that the patches of diffuse plaques that are found in other aged nondemented cases represent an even earlier stage in the disease process, which would have progressed to the preclinical and then clinically detectable stages. ACKNOWLEDGEMENTS
I thank Mr. Hieu Van Luu, Ms. Melissa Rundle, Mr. David Wavek, and Mr. Karan Randhava for their excellent technical assistance. This work was supported by Grant P01 AG03991 from the United States Public Health Service/National Institute on Aging.
REFERENCES 1. Berg, L. Clinical dementia rating (CDR). Psychopharmacol. Bull. 25:637– 639; 1988. 2. Berg, L.; McKeel, D. W., Jr.; Miller, J. P.; Baty, J. Neuropathologic indexes of Alzheimer’s disease in demented and nondemented people aged 80 years and older. Arch. Neurol. 50:349 –358; 1993. 3. Davis, P. B.; White, H.; Price, J. L.; McKeel, D.; Robins, L. N. Retrospective postmortem dementia assessment: Validation of a new clinical interview to assist neuropathological study. Arch. Neurol. 48:613– 617; 1991. 4. Gomez-Isla, T.; Price, J. L.; McKeel, D. W.; Morris, J. C.; Growdon, J. H.; Hyman, B. T. Profound loss of layer II entorhinal cortex neurons distinguishes very mild Alzheimer’s disease from nondemented aging. J. Neurosci. 16:4491– 4500; 1996. 5. Katchaturian, Z. S. Diagnosis of Alzheimer’s disease. Arch. Neurol. 42:1097–1105; 1985. 6. McKeel, D. W., Jr.; Ball, M. J.; Price, J. L.; Smith, D. S.; Miller, J. P.; Berg, L.; Morris, J. C. Interlaboratory histopathologic assessment of Alzheimer neuropathology: Different methodologies yield comparable diagnostic results. Alzheimer. Dis. Assoc. Disord. 7:136 –151; 1993. 7. Mirra, S. S.; Heyman, A.; Berg, L.; McKeel, D.; Sumi, S. M.; Crain, B.; Brownlee, L.; Vogel, F. S.; Hughes, J. P.; van Belle, G.; participating CERAD neuropathologists. The Consortium to Establish a Registry For Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathological assessment of Alzheimer’s disease and other dementias. Neurology 41:479 – 486; 1991. 8. Mirra, S. S.; Gearing, M.; McKeel, D. W., Jr.; Crain, B. J.; Hughes, J. P.; vanBell, G.; Heyman, A.; and participating neuropathologists. Interlaboratorycomparison of neuropathologic assessments in Alzhei-
9. 10.
11.
12.
13.
14.
15.
mer’s disease: A study of the Consortium to Establish a Registry for Alzheimer’s disease (CERAD). J. Neuropathol. Exp. Neurol. 53:303– 315; 1994. Morris, J. C. The clinical dementia rating (CDR): Current version and scoring rules. Neurology 43:2412–2414; 1993. Morris, J. C.; McKeel, D.; Storandt, M.; Rubin, E. H.; Price, J. L.; Grant, E. A.; Ball, M. J.; Berg, L. Very mild Alzheimer’s disease: Informant-based clinical, psychometric, and pathological distinction from normal aging. Neurology 41:469 – 478; 1991. Morris, J. C.; Storandt, M.; McKeel, D. W.; Rubin, E. H.; Price, J. L.; Grant, E. A.; Berg, L. Cerebral amyloid deposition and diffuse plaques in “normal” aging: Evidence for presynptomatic and very mild Alzheimer’s disease. Neurology 46:707–719; 1996. Price, J. L.; Davis, P. B.; Morris, J. C.; White, D. L. The distribution of tangles, plaques and related immunohistochemical markers in healthy aging and Alzheimer’s disease. Neurobiol. Aging 12:295–312; 1991. Price, J. L. The relationship between tangle and plaque formation during healthy aging and mild dementia. Neurobiol. Aging 14:661– 663; 1993. Price, J. L. Tangles and plaques in healthy aging and Alzheimer’s disease: Independence or interaction? Semin. Neurosci. 6:395– 402; 1994. Rubin, E. H.; Storandt, M.; Miller, J. P.; Kinscherf, D. A.; Grant, E. A.; Morris, J. C.; Berg, L. A prospective study of cognitive function and onset of dementia in cognitively healthy elders. Arch. Neurol., in press; 1997.