Sloping Shoulders Sign: A Practical Radiological Sign for the Differentiation of Alzheimer’s Disease and Argyrophilic Grain Disease

Abstract

Background:

Although hippocampal atrophy is a well-known imaging biomarker of Alzheimer’s disease (AD), this finding is not useful to differentiate AD from argyrophilic grain disease (AGD) which is a common AD mimicker presenting with similar amnestic symptoms and medial temporal atrophy. Instead, we propose use of the “sloping shoulders sign”, defined as a distinct configuration of the bilateral hippocampal heads showing lateral and downward slopes on axial magnetic resonance imaging (MRI).

Objective:

We investigated the diagnostic utility of the “sloping shoulders sign” as a simple radiological discriminator of AD from AGD.

Methods:

Using axial and coronal three-dimensional MRI, our newly proposed “sloping shoulders sign”, other quantitative indices including the axial hippocampal head angle (AHHA), and well-known medial temporal atrophy (MTA) score were evaluated in pathologically-proven 24 AD and 11 AGD patients.

Results:

Detection rate of the “sloping shoulders sign” was significantly higher in all AD groups (83%; 20/24) and AD with Braak neurofibrillary tangle V/VI stage subgroup (88%; 15/17) than in AGD patients (18% – 2/11; p < 0.001 and p < 0.001, respectively). In contrast to the MTA score, this sign as well as AHHA demonstrated higher diagnostic performance and reproducibility, especially to differentiate all AD patients from AGD ones (accuracies of 71.4% , 82.9% and 82.9%; Cohen’s kappa of 0.70 and 0.81, and intraclass correlation coefficient of 0.96, respectively).

Conclusion:

The “sloping shoulders sign” is useful to differentiate advanced-stage AD from AGD. Its simplicity and reproducibility based on visual inspection using axial MRI make it suitable for routine clinical practice.

Keywords

Alzheimer’s disease argyrophilic grain disease hippocampus magnetic resonance imaging sloping shoulders sign

INTRODUCTION

Hippocampal atrophy is closely linked to Alzheimer’s disease (AD) both neuropathologically and clinically, and atrophy of the medial temporal structures including the hippocampus is known to be a sensitive but non-specific imaging finding caused by other neurodegenerative pathologies as well [1]. For example, argyrophilic grain disease (AGD) is a common AD mimicker because it presents with amnestic mild cognitive impairment and medial temporal atrophy which are indistinguishable from those of AD [2]. AGD is an under-recognized 4-repeat tauopathy of the elderly, characterized by the principal neuropathological hallmarks including argyrophilic grains, oligodendrocytic coiled bodies, and neuronal intracytoplasmic pretangles [3, 4]. AGD-related neuropathological changes sometimes coexist with other neurodegenerative pathologies, and its prevalence was reported to increase up to 25.9% in patients with AD [5].

With an increasingly aging population and corresponding increase in age-related dementia, AGD is assumed to be the second most common neurodegenerative pathology after AD in very old demented patients [6, 7]. Due to the clinical similarities and lack of specific diagnostic criteria, making the clinical differentiation between AD and AGD using routine clinical diagnostic examinations including magnetic resonance imaging (MRI) is often challenging [8].

Against this background, the utility of a radiological finding indicative of hippocampal deformation with ventricular enlargement was recently identified to differentiate limbic tauopathies including AGD from AD [9]. This simple quantitative method using oblique coronal MRI was reportedly more useful than medial temporal atrophy, and easily available in daily clinical practice. Nonetheless, considering the prevalent use of the axial plane in routine MRI protocols irrespective of dementia, a visually assessed axial-based radiological finding would be more appropriate in the clinical setting [10, 11]. Intriguingly, a configuration of lateral and downward slope of the bilateral hippocampal heads in pathologically-proven AD patients resembles that of “sloping shoulders” (Fig. 1). We have named this finding the “sloping shoulders sign”, because the lateral and downward slope of the bilateral hippocampal heads resembles the frontal view of sloping shoulders. However, differences in the hippocampal configuration or dilated inferior horn, which reflect hippocampal and temporal atrophy, have not yet been compared between AD and AGD using axial MRI. This prompted us to investigate the utility of axial MRI findings which can be easily accessed in daily clinical practice. In this study we compared the diagnostic utility of this “sloping shoulders sign” with other quantitative and semiquantitative indices for the differentiation between AD and AGD, based on the premise that such simple visual inspection using axial MRI would be suitable for routine clinical practice.

Fig. 1

Exhibition of the “sloping shoulders sign”, axial hippocampal head angle (AHHA), and axial inferior horn area (AIHA). a) A magnified axial 3DT1WI in a patient with pathologically-proven AD showing the positive “sloping shoulders sign”. b) A magnified axial 3DT1WI in a patient with pathologically-proven AGD showing the negative “sloping shoulders sign”. c, d) Measurement of the AHHA and AIHA in a patient with pathologically-proven AD. Axial 3DT1WI at the level where the bilateral hippocampal heads were most clearly depicted showed a characteristic configuration of lateral and downward slope of the bilateral hippocampal heads reminiscent of “sloping shoulders” (a; curved arrows). On the other hand, a unilateral downward slope deformation was judged as a negative finding (b; curved arrow). AHHA and AIHA were measured at the same level. The angle between two crossing lines parallel to the long axes of the bilateral hippocampal heads was measured as the AHHA (c). Two gray-filled areas indicate the AIHA (d; arrowheads). 3DT1WI, three-dimensional (3D) T1-weighted image; AD, Alzheimer’s disease; AGD, argyrophilic grain disease.

MATERIALS AND METHODS

Subjects

The study population was selected by searching the available medical records between January 2013 and April 2021 at Fukushimura Hospital, Aichi, Japan. From the database of Fukushimura Brain Bank, 35 patients were selected by applying the following inclusion criteria: 1) neuropathological diagnoses according to the published criteria of AD and AGD [12, 13], 2) dementia status according to the major neurocognitive disorders of the Diagnostic and Statistical Manual of Mental Disorders (DSM)-V criteria, and 3) acquisition of MRI examinations including three-dimensional (3D) T1-weighted image (T1WI). Exclusion criteria were the presence of other comorbid neurodegenerative pathologies such as progressive supranuclear palsy, corticobasal degeneration, or large destructive lesions (e.g., cerebrovascular diseases, trauma, infections or neoplasms) in the limbic lobes [14, 15]. AD patients with argyrophilic grains were also excluded. On the other hand, slight accumulation of Lewy bodies and TAR DNA-binding protein-43 kDa (TDP-43) pathologies, which did not fulfill the diagnostic criteria of dementia with Lewy bodies and frontotemporal lobar degeneration with TDP-43 (FTLD-TDP), were accepted as reflecting minimal senile neuropathologic changes [16 –18]. This study protocol was approved by the institutional review board of Fukushimura Hospital (Research ID: 425, 2021), and was conducted in strict compliance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. A postmortem consent was obtained from the bereaved of all patients.

Neuropathological analysis

In this study, all 8μm-thick sections from the 34 formalin-fixed one-sided cerebral hemispheres were stained conventionally with hematoxylin and eosin, Klüver-Barrera, and Gallyas-Braak method. Additionally, selected sections were immunostained with antibodies against amyloid-β (IBL, Maebashi, Japan), phosphorylated tau (AT8; 1 : 1000, Innogenetics, Ghent, Belgium), TDP-43 (Cosmo bio, Tokyo, Japan), and α-synuclein (Wako, Osaka, Japan). AD-related neuropathology was scored using the Braak and Braak staging system, the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) criteria and the Thal phase system for amyloid-β plaques [19 –21]. Neuropathological diagnosis of AD was made for individuals showing, at least, intermediate AD neuropathologic changes [13]. Neuropathological diagnosis of AGD was made and categorized into three (I-III) stages according to the proposed staging paradigm for AGD [12, 22].

MRI protocol

All patients underwent MRI examination on a 1.5-T imager (MAGNETOM Aera®; Siemens Healthcare, Erlangen, Germany) with a 20-channel Head/Neck coil. A 3DT1WI was obtained on magnetization-prepared rapid gradient-echo (MPRAGE) for 28 patients and on fast low angle shot (FLASH) for seven patients. Scan parameters of MPRAGE were as follows: repetition time (TR)/echo time (TE), 1700 ms/4.2 ms; inversion time, 800 ms; flip angle (FA), 19°; field of view (FOV), 230 mm; matrix, 256×256; and 1.25-mm-thick gapless sections. Scan parameters of FLASH were: TR/TE, 27 ms/6.9 ms; FA, 19°; FOV, 180 mm; matrix, 320×224; and 0.65-mm-thick gapless sections. Additionally, the multiplanar reconstruction technique was performed to make a 1.25-mm-thick axial plane along the orbitomeatal line and oblique coronal plane perpendicular to the hippocampus body. To avoid biases due to head position and inappropriate reconstruction, a well-experienced neuroradiologist (K.S.) checked the quality of all reconstructed images.

Image analysis

All MR images were independently interpreted by two raters who were a neuroradiologist and neurologist with 17 years (rater 1, K.S.) and 13 years (rater 2, Y.U.) of neuroradiologic experience with MRI of dementia. For the analyses, the two raters were blinded to all of the patients’ information. Using ITK-SNAP software (Version3.8.0, http://www.itksnap.org/), the “sloping shoulders sign” was visually evaluated on the axial reconstructed 3DT1WI. At the level where the bilateral hippocampal heads were most clearly depicted, the distinct configuration of the bilateral hippocampal heads showing lateral and downward slopes was reminiscent of “sloping shoulders” (Fig. 1). To avoid the asymmetric changes often detected in AGD and FTLD, only bilateral downward slope deformations were judged as a positive finding. For the quantitative assessment of this sign, an angle between two crossing lines parallel to the long axes of the hippocampal heads was measured using a “line and rule mode” of ITK-SNAP software at the same level (axial hippocampal head angle: AHHA) (Fig. 1). After that, the area of the bilateral inferior horns of the lateral ventricles was measured by a “paintbrush mode (adaptive brush tool with the following settings: size, 40; granularity, 10; and smoothness, 100)” (axial inferior horn area: AIHA) (Fig. 1). Additionally, the semiquantitative Scheltens’ medial temporal atrophy (MTA) score was calculated to determine the differences in hippocampal atrophy between AD and AGD on coronal 3DT1WI [23]. The five-point MTA score ranged from 0 = no atrophy to 4 = severe atrophy for the rated brain regions.

Statistical analysis

Statistical analyses were performed using IBM SPSS statistics 24 (IBM SPSS Inc, Chicago, IL, USA). The one-way analysis of variance for normally distributed data (Mini-Mental State Examination score: MMSE), Kruskal-Wallis test for non-normally distributed data (age at MRI, age at death, disease duration at MRI, AHHA, AIHA, and MTA score), and Fisher’s exact test for categorical data (gender, and ratio of the “sloping shoulders sign”, Lewy body, and TDP-43 pathology) were performed for comparisons among AD subgroups subdivided by Braak neurofibrillary tangle (NFT) stage (i.e., IV and V/VI stages) and AGD group. When a significant level was found in multiple comparisons, the unpaired t-test or Mann-Whitney U test was also performed. Resulting p values were corrected according to the Bonferroni method and considered as statistically significant if < 0.05. Additionally, the relationship between AHHA and AIHA, and AHHA and MTA score was assessed by Spearman’s rank correlation coefficient. To evaluate the interrater reproducibility, the intraclass correlation coefficient, Cohen’s kappa, and weighted Cohen’s kappa were calculated. A receiver operating characteristic curve analysis was performed to assess the diagnostic performance of AHHA and AIHA. Youden’s index was applied to determine the cut-off values.

RESULTS

Clinical features

Patients’ characteristics are summarized in Table 1. AD group consisted of 24 patients with 17 severe pathological changes corresponding to Braak NFT V/VI stage with frequent neuritic plaques according to the CERAD and 7 with moderate pathological changes corresponding to Braak NFT IV stage with moderate or frequent CERAD neuritic plaques [19]. AGD group consisted of 11 patients with advanced AGD (Saito stage III) [12]. There were no significant differences in age at MRI, age at death, disease duration at MRI or co-existence rate of TDP-43 and Lewy-body pathology between the AD and AGD patients. On the other hand, MMSE score of AD with Braak NFT stage V/VI subgroup was lower than that of AGD.

Table 1

Patients’ characteristics

	AD		AGD	p
	AD Braak NFT IV	AD Braak NFT V/VI
Male/Female	2/5	1/16 ^a	5/6	0.03^†
Age at MRI (y)	86.1±7.2	83.3±7.0	87.7±6.4	0.15^*
Age at death (y)	87.4±7.1	84.8±6.8	88.1±6.4	0.29^*
Duration at MRI (y)	7.0±5.2	11.2±4.6	7.7±5.5	0.07^*
MMSE	14.0±6.1	6.2±5.7^b	14.6±6.6	0.005^‡
Braak NFT	IV 7	V 2, VI 15	I 2, II 9	N.A.
CERAD 0/A/B/C	0/0/2/5	0/2/0/15	3/2/4/2	N.A.
Thal 0/1/2/3/4/5	0/0/0/0/6/1	0/0/0/1/4/12	2/4/2/1/1/1	N.A.
LbP	3/7 (43%)	10/17 (59%)	3/9 (33%)^\|\|	0.53^†
TDP-43P	2/7 (29%)	11/17 (65%)	3/6 (50%)^\|\|	0.29^†
AGD Saito I/II/III	0/0/0	0/0/0	0/0/11	N.A.

Data are shown as absolute numbers or the mean±standard deviation. AD, Alzheimer’s disease; AGD, argyrophilic grain disease; CERAD, Consoritum to Establish a Registry for Alzheimer Disease; LbP, Lewy body pathology; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging; N.A., not applicable; NFT, neurofibrillary tangle, TDP-43P, TAR DNA-binding protein 43 kDa pathology; y = years. ^*Kruskal-Wallis test; ^†Fisher’s exact test; ^‡One-way analysis of variance; ^||Lewy body and TDP-43 pathologies were not fully evaluated in some AGD patients; ^ap = 0.066 versus AGD by Fisher’s exact test with Bonferroni correction; ^bp = 0.006 versus AGD by unpaired t test with Bonferroni correction.

MRI features

Representative axial 3DT1WI of pathologically-proven AD and AGD patients are presented in Fig. 2. The “sloping shoulders sign” was observed in 20 AD patients, 15 with Braak NFT V/VI stage, and 5 with IV stage. The detection rate was higher in all AD (83% – 20/24) and AD with Braak NFT V/VI stage subgroup (88% – 15/17) than in the AGD patients (18% – 2/11; p < 0.001 and p < 0.001, respectively). All AD and AD with Braak NFT stage V/VI subgroup showed lower AHHA and larger AIHA than did AGD (91.1±32.1, 86.1±31.7, 644.9±447.3, and 738.4±479.7 versus 135.0±31.0 and 269.7±107.3; p = 0.008, 0.004, 0.016, and 0.006, respectively). Despite the relatively high MTA score, no statistical difference was observed between AD with Braak NFT V/VI stage subgroup and AGD patients (3.8±0.5 versus 3.3±0.6; p = 0.071). The weighted Cohen’s kappa of MTA score, Cohen’s kappa of “sloping shoulders sign”, and intraclass correlation coefficients of AHHA and AIHA for interrater reliability of two raters were 0.70, 0.81, 0.96, and 0.98, respectively. Significant correlations were found between AHHA and AIHA in all AD group and AD with Braak NFT V/VI stage subgroup (r = –0.563; p = 0.004 and r = –0.554; p = 0.021, respectively). Similarly, significant correlations were found between AHHA and MTA score in all AD group and AD with Braak NFT IV stage subgroup (r = –0.586; p = 0.003 and r = –0.756; p = 0.049, respectively).

Fig. 2

Representative axial 3DT1WI of pathologically-proven AD and AGD patients. a) An 89-year-old pathologically proven AD (Braak NFT stage VI stage) patient. b) A 96-year-old pathologically proven AD (Braak NFT stage IV stage) patient. c) A 92-year-old pathologically proven AGD patient. d) A 97-year-old pathologically proven AGD patient. In contrast to the downward slope configurations in pathologically-proven AD with the positive “sloping shoulders sign” (a, b; curved arrows), more horizontally shaped configurations similar to “square shoulders” were observed in pathologically-proven AGD with the negative “sloping shoulders sign” (c, d; arrows). 3DT1WI, three-dimensional (3D) T1-weighted image; AD, Alzheimer’s disease; AGD, argyrophilic grain disease; NFT, neurofibrillary tangle.

Accuracies of the “sloping shoulders sign” to differentiate AGD from all AD group, AD with Braak NFT V/VI stage subgroup, or AD with Braak NFT IV stage subgroup (82.9%, 85.7%, and 77.8%, respectively) were similar to those of AHHA (82.9%, 89.3%, and 77.8%, respectively). In contrast, the accuracies of AIHA to differentiate AGD from these AD subgroups were somewhat lower (71.4%, 82.1%, and 72.2%, respectively). Other measured values and diagnostic indices such as area under curve (AUC) and cut-off values are shown in Tables 2 and 3. ROC curves of AHHA, AIHA, and MTA scores to discriminate AD from AGD patients were exhibited in Fig. 3. Between all AD and AGD patients, both AHHA and AIHA showed higher AUCs than that of MTA score (p = 0.04 and 0.03, respectively). Furthermore, higher AUCs of AHHA and AIHA were also detected between AD with Braak NFT IV stage subgroup and AGD (p = 0.02 and 0.008, respectively). Details and analysis results including age- and gender-matched healthy controls are available in Supplementary Tables 1-3 and Supplementary Figure 1.

Table 2

Comparison of sloping shoulders sign and other quantitative indices

	AD			AGD	p
	All AD patients	AD Braak NFT IV	AD Braak NFT V/VI
SSS	20/24/ (83%)^a	5/7 (71%)	15/17 (88%)^b	2/11 (18%)	<0.001^*
AHHA	91.1±32.1^c	103.0±32.4	86.1±31.7^d	135.0±31.0	0.003^†
AIHA	644.9±447.3^e	417.8±261.6	738.4±479.7^f	269.7±107.3	0.005^†
MTA score	3.6±0.7	3.1±0.9	3.8±0.5^g	3.3±0.6	0.029^†

Data are shown as absolute numbers or the mean±standard deviation. AD, Alzheimer’s disease; AGD, argyrophilic grain disease; AHHA, axial hippocampal head angle; AIHA, axial inferior horn area; MTA, medial temporal atrophy; NFT, neurofibrillary tangle; SSS, sloping shoulders sign. ^*Fisher’s exact test; ^†Kruskal-Wallis test; ^ap < 0.001 versus AGD by Fisher’s exact test with Bonferroni correction; ^bp < 0.001 versus AGD by Fisher’s exact test with Bonferroni correction; ^cp = 0.008 versus AGD by Mann-Whitney U test with Bonferroni correction; ^dp = 0.004 versus AGD by Mann-Whitney U test with Bonferroni correction; ^ep = 0.016 versus AGD by Mann-Whitney U test with Bonferroni correction; ^fp = 0.006 versus AGD by Mann-Whitney U test with Bonferroni correction; ^gp = 0.071 versus AGD by Mann-Whitney U test with Bonferroni correction.

Table 3

AUC, cut-off values, and diagnostic index in discriminating between AD and AGD patients

		AUC	CV	SENS (%)	SPEC (%)	PPV (%)	NPV (%)	ACC (%)
SSS	All AD versus AGD	N.A.	N.A.	83.3	81.8	90.9	69.2	82.9
	AD B V/VI versus AGD	N.A.	N.A.	88.2	81.8	88.2	81.8	85.7
	AD B IV versus AGD	N.A.	N.A.	71.4	81.8	71.4	81.8	77.8
AHHA	All AD versus AGD	0.84	100	79.2	90.9	95.0	66.7	82.9
	AD B V/VI versus AGD	0.86	100	88.2	90.9	93.8	83.3	89.3
	AD B IV versus AGD	0.79	110	71.4	81.8	71.4	81.8	77.8
AIHA	All AD versus AGD	0.82	462	58.3	100	100	52.4	71.4
	AD B V/VI versus AGD	0.86	462	70.6	100	100	68.8	82.1
	AD B IV versus AGD	0.71	263	85.7	63.4	60.0	87.5	72.2
MTA score	All AD versus AGD	0.68	3.5	75.0	63.6	81.8	53.8	71.4
	AD B V/VI versus AGD	0.75	3.5	94.1	63.6	80.0	87.5	82.1
	AD B IV versus AGD	0.47	3.5	42.9	63.6	42.9	63.6	55.6

ACC, accuracy; AD, Alzheimer’s disease; AGD, argyrophilic grain disease; AHHA, axial hippocampal head angle; AIHA, axial inferior horn area; AUC, area under the curve; B, Braak neurofibrillary tangle stage; CV, cut-off value; MTA, medial temporal atrophy; NPV, negative predictive value; PPV, positive predictive value; SENS, sensitivity; SPEC, specificity; SSS, sloping shoulders sign.

Fig. 3

ROC curves using AHHA, AIHA, and MTA score for the differentiation between AD and AGD patients. (a) all AD versus AGD patients, (b) advanced-stage AD with Braak NFT stage V/VI versus AGD patients, and (c) moderate-stage AD with Braak NFT stage IV versus AGD patients.

DISCUSSION

In this study, we clarified the utility of the “sloping shoulders sign” for the differentiation of pathologically-proven AD, especially with severe pathological changes corresponding to Braak NFT V/VI stage, and AGD. This visually assessed radiological finding was highly reproducible and showed high diagnostic accuracy similar to that of quantitative analysis (i.e., AHHA). In contrast, no statistical difference between AD and AGD was observed in the traditional Scheltens’ MTA score. These results suggest that this simple qualitative finding on axial 3DT1WI could contribute to the imaging diagnosis of AD and AGD.

Considering the non-specificity of medial temporal atrophy, more specific biomarkers reflecting amyloid and tau deposition using positron emission tomography and cerebrospinal fluid examination necessary to make the correct antemortem diagnosis of AD [24]. However, these examinations are highly expensive and invasive, precluding their use in the first-line diagnosis in routine clinical practice. Similarly, despite the utility of visualizing the characteristic atrophic pattern in demented patients, it is not realistic to apply advanced analytic techniques such as voxel-based morphometry and volumetry for all patients suspected of having AD [25 –27]. Thus, less invasive and more easily applicable techniques using MRI are required for routine purposes.

For these reasons, simple visual inspection using axial MRI, routinely scanned in various MRI protocols, is considered as a notable advantage of the newly proposed “sloping shoulders sign”. Moreover, this radiological finding can be assessed with high reproducibility and diagnostic accuracy comparable to those of quantitative analysis (i.e., AHHA) for differentiation between AD and AGD. Given its simplicity and convenience, it can be recommended as one key finding to supplement other differential points including hippocampal deformation with ventricular enlargement in AD and asymmetric atrophy relatively localized to the anterior temporal and limbic lobes in AGD [9, 27].

However, the pathophysiology of the “sloping shoulders sign” has not been sufficiently elucidated. It is known that deformation of the hippocampal heads from a long horizontal elliptical to long vertical elliptical configuration becomes apparent with progression of hippocampal atrophy, especially the CA1 subfield, in AD [28]. Similarly, the ventricles expand with AD progression [29]. Considering the correlation of AHHA (a quantified index of the “sloping shoulders sign”), AIHA (a quantified index of inferior horn enlargement), and MTA score exclusively in AD, the “sloping shoulders sign” is surmised to be a consequence of not only hippocampal atrophy but also ventricular enlargement subsequent to atrophy of the surrounding lobar anatomy associated with advanced-stage AD [30]. Indeed, the “sloping shoulders sign” was not detected in four AD patients, mainly with Braak NFT IV stage subgroup, who showed higher AHHA and lower AIHA.

It has also been reported that the severity of NFT pathology correlates with the degree of atrophy and ventricular volume change in AD [25, 31]. So, we divided AD into two subgroups according to the Braak NFT stage to further evaluate the diagnostic performance of the “sloping shoulders sign”. Although not statistically significant, AD with Braak NFT IV stage showed not only a relatively low positive rate of this sign but also high MMSE and short disease duration. Similarly, AHHA was relatively high in patients with AD with Braak NFT IV stage. It makes sense to assume a correlation between the severity of the clinical manifestations and the neuropathological changes, and configuration of the hippocampus (i.e., “sloping shoulders sign” and AHHA). Therefore, clinicians should consider the disease severity when applying the “sloping shoulders sign” and AHHA for the differentiation of AD and AGD. Unfortunately, due to the nature of our Brain Bank in a medical long-term care sanatorium, it was difficult to evaluate long-standing severely demented patients using more detailed neuropsychological assessment in this study. Despite the relatively higher proportion of female in AD with Braak NFT V/VI stage, statistical analyses to eliminate the effect of sex imbalance was not applicable because of the non-normal distribution. In addition to the several medical, environmental, and lifestyle risk factors, female sex is known as a major risk factor of late-onset AD. Considering the effect of demographical variables, especially female sex which could cause more severe neurofibrillary degeneration and brain atrophy in AD patients, additional investigation with adjustment for demographical variables will be necessary to clarify the effect of sex for the differentiation between AD and AGD [32]. Furthermore, due to the insufficient case numbers, the utility of the “sloping shoulders sign” was not validated in patients with other kinds of neurodegenerative disorders such as senile dementia of neurofibrillary tangle type, FTLD with TDP-43, or pure limbic-predominant age-related TDP-43 encephalopathy (LATE), all of which can also cause severe medial temporal lobe atrophy. Additionally, some TDP-43 neuropathologies were included in this study. Considering the importance of TDP-43 comorbidity in accelerating hippocampal atrophy, further investigations to better clarify the effect of TDP-43 pathology for the differentiation between AD and AGD will be required [33].

In conclusion, the “sloping shoulders sign” is a reproducible and useful clue to differentiate advanced-stage AD from AGD. The simplicity and convenience of its use based on visual inspection of routine axial MRI make it suitable for routine clinical practice.

Footnotes

ACKNOWLEDGMENTS

We thank Mr. John Gelblum for his English proofreading.

A part of this work was supported by JSPS KAKENHI Grant Number JP 16H06277 (CoBiA).

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Barkhof

, Polvikoski

, van Straaten

, Kalaria

, Sulkava

, Aronen

, Niinistö

, Rastas

, Oinas

, Scheltens

, Erkinjuntti

(2007) The significance of medial temporal lobeatrophy: A postmortem MRI study in the very old. Neurology 69, 1521–1527.

Tolnay

, Clavaguera

(2004) Argyrophilic grain disease: Alate-onset dementia with distinctive features among tauopathies. Neuropathology 24, 269–283.

Braak

, Braak

(1987) Argyrophilic grains: Characteristic pathology of cerebral cortex in cases of adult onset dementia without Alzheimer changes. Neurosci Lett 76, 124–127.

Rodriguez

, Grinberg

(2015) Argyrophilic grain disease: An underestimated tauopathy. Dement Neuropsychol 9, 2–8.

Fujino

, Wang

, Thomas

, Espinoza

, Davies

, Dickson

(2005) Increased frequency of argyrophilic grain disease in Alzheimer disease with 4R tau-specific immunohistochemistry. J Neuropathol Exp Neurol 64, 209–214.

Saito

, Nakahara

, Yamanouchi

, Murayama

(2002) Severe involvement of ambient gyrus in dementia with grains. J Neuropathol Exp Neurol 61, 789–796.

Esteban de Antonio

, López-Álvarez

, Rábano

, Agüera-Ortiz

, Sánchez-Soblechero

, Amaya

, Portela

, Cátedra

, Olazarán

(2020) Pathological correlations of neuropsychiatric symptoms in institutionalized people with dementia. J Alzheimers Dis 78, 1731–1741.

Jicha

, Nelson

(2019) Hippocampal sclerosis, argyrophilic grain disease, and primary age-related tauopathy. Continuum (Minneap Minn) 25, 208–233.

Sakurai

, Kaneda

, Inui

, Uchida

, Morimoto

, Nihashi

, Kato

, Ito

, Hashizume

(2021) Simple quantitative indices for the differentiation of advanced-stage Alzheimer’s disease and other limbic tauopathies. J Alzheimers Dis 81, 1093–1102.

10.

Filippi

, Agosta

, Barkhof

, Dubois

, Fox

, Frisoni

, Jack

, Johannsen

, Miller

, Nestor

, Scheltens

, Sorbi

, Teipel

, Thompson

, Wahlund

(2012) EFNS task force: The use of neuroimaging in the diagnosis of dementia. Eur J Neurol 19, e131–140, 1487-1501.

11.

Mehan

Jr. , González

, Buchbinder

, Chen

, Copen

, Gupta

, Hirsch

, Hunter

, Johnson

, Kelly

, Larvie

, Lev

, Pomerantz

, Rapalino

, Rincon

, Romero

, Schaefer

, Shah

(2014) Optimal brain MRI protocol for new neurological complaint. PLoS One 9, e110803.

12.

Saito

, Ruberu

, Sawabe

, Arai

, Tanaka

, Kakuta

, Yamanouchi

, Murayama

(2004) Staging of argyrophilic grains: An age-associated tauopathy. J Neuropathol Exp Neurol 63, 911–918.

13.

Montine

, Phelps

, Beach

, Bigio

, Cairns

, Dickson

, Duyckaerts

, Frosch

, Masliah

, Mirra

, Nelson

, Schneider

, Thal

, Trojanowski

, Vinters

, Hyman

(2012) National Institute on Aging-Alzheimer’s Association guidelines for the neuropathologic assessment of Alzheimer’s disease: A practical approach. Acta Neuropathol 123, 1–11.

14.

Litvan

, Hauw

, Bartko

, Lantos

, Daniel

, Horoupian

, McKee

, Dickson

, Bancher

, Tabaton

, Jellinger

, Anderson

(1996) Validity and reliability of the preliminary NINDS neuropathologic criteria for progressive supranuclear palsy and related disorders. J Neuropathol Exp Neurol 55, 97–105.

15.

Dickson

, Bergeron

, Chin

, Duyckaerts

, Horoupian

, Ikeda

, Jellinger

, Lantos

, Lippa

, Mirra

, Tabaton

, Vonsattel

, Wakabayashi

, Litvan

(2002) Office of Rare Diseases neuropathologic criteria for corticobasal degeneration. J Neuropathol Exp Neurol 61, 935–946.

16.

Mackenzie

, Neumann

, Baborie

, Sampathu

, Du Plessis

, Jaros

, Perry

, Trojanowski

, Mann

, Lee

(2011) A harmonized classification system for FTLD-TDP pathology. Acta Neuropathol 122, 111–113.

17.

Uchino

, Takao

, Hatsuta

, Sumikura

, Nakano

, Nogami

, Saito

, Arai

, Nishiyama

, Murayama

(2015) Incidence and extent of TDP-43 accumulation in aging human brain. Acta Neuropathol Commun 3, 35.

18.

McKeith

, Boeve

, Dickson

, Halliday

, Taylor

, Weintraub

, Aarsland

, Galvin

, Attems

, Ballard

, Bayston

, Beach

, Blanc

, Bohnen

, Bonanni

, Bras

, Brundin

, Burn

, Chen-Plotkin

, Duda

, El-Agnaf

, Feldman

, Ferman

, Ffytche

, Fujishiro

, Galasko

, Goldman

, Gomperts

, Graff-Radford

, Honig

, Iranzo

, Kantarci

, Kaufer

, Kukull

, Lee

VMY

, Leverenz

, Lewis

, Lippa

, Lunde

, Masellis

, Masliah

, McLean

, Mollenhauer

, Montine

, Moreno

, Mori

, Murray

, O’Brien

, Orimo

, Postuma

, Ramaswamy

, Ross

, Salmon

, Singleton

, Taylor

, Thomas

, Tiraboschi

, Toledo

, Trojanowski

, Tsuang

, Walker

, Yamada

, Kosaka

(2017) Diagnosis and management of dementia with Lewy bodies: Fourth consensus report of the DLB Consortium. Neurology 89, 88–100.

19.

Mirra

, Heyman

, McKeel

, Sumi

, Crain

, Brownlee

, Vogel

, Hughes

, van Belle

, Berg

(1991) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology 41, 479–486.

20.

Braak

, Braak

(1995) Staging of Alzheimer’s disease-related neurofibrillary changes. Neurobiol Aging 16, 271–278; discussion 278-284.

21.

Thal

, Rüb

, Orantes

, Braak

(2002) Phases of A beta-deposition in the human brain and its relevance for the development of AD. Neurology 58, 1791–1800.

22.

Rodriguez

, Suemoto

, Molina

, Nascimento

, Leite

, de Lucena Ferretti-Rebustini

, Farfel

, Heinsen

, Nitrini

, Ueda

, Pasqualucci

, Jacob-Filho

, Yaffe

, Grinberg

(2016) Argyrophilic grain disease: Demographics, clinical, and neuropathological features from a large autopsy study. J Neuropathol Exp Neurol 75, 628–635.

23.

Scheltens

, Leys

, Barkhof

, Huglo

, Weinstein

, Vermersch

, Kuiper

, Steinling

, Wolters

, Valk

(1992) Atrophy of medialtemporal lobes on MRI in “probable” Alzheimer’s disease and normal ageing: Diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry 55, 967–972.

24.

Jack

Jr. , Bennett

, Blennow

, Carrillo

, Feldman

, Frisoni

, Hampel

, Jagust

, Johnson

, Knopman

, Petersen

, Scheltens

, Sperling

, Dubois

(2016) A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 87, 539–547.

25.

Whitwell

, Josephs

, Murray

, Kantarci

, Przybelski

, Weigand

, Vemuri

, Senjem

, Parisi

, Knopman

, Boeve

, Petersen

, Dickson

, Jack

Jr. (2008) MRI correlates of neurofibrillary tangle pathology at autopsy: A voxel-based morphometry study. Neurology 71, 743–749.

26.

Whitwell

, Dickson

, Murray

, Weigand

, Tosakulwong

, Senjem

, Knopman

, Boeve

, Parisi

, Petersen

, Jack

Jr. , Josephs

(2012) Neuroimaging correlates of pathologically defined subtypes of Alzheimer’s disease: A case-control study. Lancet Neurol 11, 868–877.

27.

Sakurai

, Tokumaru

, Ikeda

, Morimoto

, Inui

, Sumida

, Oba

, Nakagawa

, Matsukawa

, Hashizume

(2019) Characteristic asymmetric limbic and anterior temporal atrophy in demented patients with pathologically confirmed argyrophilic grain disease. Neuroradiology 61, 1239–1249.

28.

Hayashi

, Wada

, Uchida

, Kitagaki

(2009) Enlargement of the hippocampal angle: A new index of Alzheimer disease. Magn Reson Med Sci 8, 33–38.

29.

Nestor

, Rupsingh

, Borrie

, Smith

, Accomazzi

, Wells

, Fogarty

, Bartha

(2008) Ventricular enlargement as a possible measure of Alzheimer’s disease progression validated using the Alzheimer’s Disease Neuroimaging Initiative database. Brain 131, 2443–2454.

30.

Thompson

, Hayashi

, De Zubicaray

, Janke

, Rose

, Semple

, Hong

, Herman

, Gravano

, Doddrell

, Toga

(2004) Mapping hippocampal and ventricular change in Alzheimer disease. Neuroimage 22, 1754–1766.

31.

Silbert

, Quinn

, Moore

, Corbridge

, Ball

, Murdoch

, Sexton

, Kaye

(2003) Changes in premorbid brain volume predict Alzheimer’s disease pathology. Neurology 61, 487–492.

32.

Filon

, Intorcia

, Sue

, Vazquez Arreola

, Wilson

, Davis

, Sabbagh

, Belden

, Caselli

, Adler

, Woodruff

, Rapscak

, Ahern

, Burke

, Jacobson

, Shill

, Driver-Dunckley

, Chen

, Reiman

, Beach

, Serrano

(2016) Gender differences in Alzheimer disease: Brain atrophy,histopathology burden, and cognition. J Neuropathol ExpNeurol 75, 748–754.

33.

Josephs

, Dickson

, Tosakulwong

, Weigand

, Murray

, Petrucelli

, Liesinger

, Senjem

, Spychalla

, Knopman

, Parisi

, Petersen

, Jack

Jr. , Whitwell

(2017) Rates of hippocampal atrophy and presence of post-mortem TDP-43 in patients with Alzheimer’s disease: A longitudinal retrospective study. Lancet Neurol 16, 917–924.