The Geneva brain collection

Abstract

The University of Geneva brain collection was founded at the beginning of the 20^th century. Today, it consists of 10,154 formaldehyde- or buffered formaldehyde–fixed brains obtained from the autopsies of the Department of Psychiatry and, since 1971, from the Department of Geriatrics as well. More than 100,000 paraffin-embedded blocks and 200,000 histological slides have also been collected since 1901. From the time of its creation, this collection has served as an important resource for pathological studies and clinicopathological correlations, primarily in the field of dementing illnesses and brain aging research. These materials have permitted a number of original neuropathological observations, such as the classification of Pick’s disease by Constantinidis, or the description of dyshoric angiopathy and laminar sclerosis by Morel. The large number of cases, including some very rare conditions, provides a unique resource and an opportunity for worldwide collaborations.

Historical background

The first studies in the field of neuroscience began as early as the 19^th century in Geneva, before the installation of the Department of Psychiatry in its current location. The first important publications appeared in 1877 by Jean-Louis Prévost, who studied oculomotor functions and hemianopsy in hemiplegia.¹

The earliest neuropathologic observations on postmortem brains in Geneva date from the last decade of the 19^th century. Dr Joannès Martin, the head of the “Asile des Vernets,” as the hospital for the insane was then known, and his assistant, Dr Jean Pilz, prepared the first histological slides from brain autopsies. Paul-Louis Ladame, privat-docent at the University of Geneva in neuropathology, studied mainly general paresis,^{2, 3} brain hemorrhage,⁴ and Huntington’s disease.⁵ These were the first steps to the extended neuropathological and clinicopathological studies that were continued into the 20^th century until today.

The new site of the psychiatry hospital called “Asile de Bel-Air” opened on November 9^th 1900, and Professor Rodolphe Weber, the first director of the psychiatric hospital also inaugurated the “Laboratory of neuropathology for anatomohistological studies of the brain.” In these days, autopsies were performed by medical doctors and psychiatrists. The first certified neuropathologist was hired only in 1947.⁶ From these first decades of the 20^th century, works originated on the aphasias,^{7, 8} hemianesthesia,^{9, 10} and vascular and neoplastic encephalopathies.^11-15 Also from the beginning of the 20^th century dates the collection of histological slides (the oldest one from 1901, the case of an enormous meningioma was published in 1905;¹¹ Fig. 1), and brain tissues embedded in paraffin blocks—of which more than 100,000 have been collected until today.

Figure 1.

(a) The oldest histological slide of the collection dates from 1901: it is an enormous meningioma of the orbitofrontal region. (b) The case was published in the Nouvelle Iconographie de la Salpêtrière in 1905.

In the first years, autopsies were performed only occasionally, in what were considered medically “interesting” cases. Later (around 1920), on the initiative of Professor Charles Ladame, an autopsy was systematically requested for any patient who died in the hospital. Charles Ladame was first hired as an assistant by Rodolphe Weber. He performed extensive histological studies on brains of mentally affected patients and published the summary of his medical thesis in 1909 in the then-prominent French medical journal, L’Encéphale.¹⁶ He described the histological lesions seen in syphilitic encephalopathy (granular ependymitis and rod-shaped microglia),^{17, 18} in vascular dementia (arteriolar hyalinosis),^{19, 20} revisited briefly the structure of senile plaques (foyers de sclérose miliaire, military sclerosis foci), and neurofibrillary tangles (scléroses neuronales, neuronal sclerosis)¹⁶—two years after Alzheimer’s original description.

In 1925, Charles Ladame replaced Rodolphe Weber as head of the psychiatric hospital and hired Ferdinand Morel as assistant in 1927. With Morel, an important period of the research in the field of dementia began. Morel was primarily interested in histological lesions in cases with dementia—principally, in senile dementia or late-onset Alzheimer’s disease. In addition to a great amount of research activity, he established a laboratory for research and postmortem neuropathological diagnosis (named Encéphale). Morel said that to obtain new knowledge, the only valid method was to use the “anatomo-clinical” method, as had been done in earlier years in cases of Alzheimer’s or Pick’s diseases.²¹ Morel was the first to perform silver impregnations (Bielschowsky and del Rió Hortega techniques) on brain tissues in this laboratory. Owing to his observations and descriptions of lesions, his name became internationally known, and two pathologies bear his name today: the Morgagni-Morel syndrome (hyperostosis frontalis interna, obesity, and virilization-hirsutism—in which he also described the associated menstrual disturbances)²² and the laminar sclerosis of Morel (seen in alcoholic dementia as a gliosis in layer III of the frontal cortex).²³

In 1938, Morel became head of the Department of Psychiatry. In 1943, he published—after Oppenheim, Bielchowsky, and Löwenberg at the beginning in the 20^th century—the third case of dyshoric angiopathy (drüsige Entartung or angiopathie topistique; Figs. 2a and 3),²⁴ one of the secondary histopathological signs of Alzheimer’s disease, seen most often in layer IVc of Brodmann area 17. It was Morel who first related this lesion to Alzheimer’s disease. Some years later, Morel, with Erwin Wildi, presented a series of 43 cases with dyshoric angiopathy.²⁵ They concluded that dyshoric angiopathy is seen only in the presence of senile plaques and is accompanied often by hyperproteinemia, and they postulated that dyshoric angiopathy is due to increased permeability of cortical capillaries.

Figure 2.

Morel’s first publication on dyshoric angiopathy (a) and Pantelakis’ article on congophilic angiopathy (b).

Figure 3.

Dyshoric angiopathy. (a) Thioflavin fluorescence, (b–e) modified Gallyas silver impregnation. Scale bar: (a) 200 μm, (b) 100 μm, (c–e) 50 μm̃.

Among Morel and Wildi’s collaborators, Stefanos Pantelakis published a study on 26 cases of another form of angiopathy, called congophilic angiopathy²⁶ (Fig. 2b). In contrast to dyshoric angiopathy, meningeal and perforant arteries are also affected. Today, these lesions together are referred to as amyloid angiopathy.

The microscopic research with Wildi on the aging brain continued with studies of granular atrophy,²⁷ an ischemic lesion corresponding to cortical scars (Figs. 4 and 5). Wildi also published on histopathologic changes in aged patients with schizophrenia.²⁸

Figure 4.

Severe granular atrophy of Morel in the parieto-temporo-occipital region.

Figure 5.

Histological picture of multiples cortical microinfarcts in the frontal cortex (black arrows). Inset: Neoformation of microvessels around the necrotic area (white arrow). Globus silver impregnation. Scale bar: (a) 500 μm, (b) 200 μm.

In 1959, Julian de Ajuriaguerra succeeded Morel as chair of Psychiatry. He is known as an important personality in the modernization and liberalization of psychiatry. Under his influence, Jean Constantinidis, engaged originally by Morel, pursued ongoing work in neurodegenerative disorders. One of Constantinidis’ important contributions was a new neuropathological classification of Pick’s disease into three types (A, B, and C), depending on the presence or absence of Pick’s bodies and ballooned neurons; this system was widely used until it was replaced by the recent classification of frontotemporal dementias, of which it was truly a forerunner²⁹ and is still frequently cited today (Fig. 6). Additionally, Constantinidis continued Morel’s clinicopathological studies of Alzheimer’s disease and, in 1964, introduced histochemical methods in his laboratory to examine the distribution of monoamines in the brain using histofluorescence.³⁰

Figure 6.

Asymmetrical brain atrophy of both temporal lobes in Pick’s disease (Loyez staining) (a). Constantinidis’ paper on a new classification of Pick’s disease (b).

Research on neuropeptides followed work investigating monoaminergic systems.^{31, 32} The production of antibodies against various neuropeptides allowed the Bel-Air team to perform detailed cartographies of substance P, Leu- and Met-enkephalins, somatostatin, vasoactive intestinal peptide, cholecystokinin,^33-37 delta sleep-inducing peptide (DSIP), corticotropin-like intermediate lobe peptide (CLIP), and orexin,^38-42,43 together with studies of the implications of these neuropeptides in human pathology.^{37, 44}

Activity in recent years

Degenerative lesions in the human brain during normal aging and Alzheimer’s disease were investigated and published in details from early 1990s (Figs. 7, 8, and 9).⁴⁵ Numerous articles appeared that compared normal aging and Alzheimer’s disease in elderly subjects and centenarians. The particularity that in demented centenarians extensive neurofibrillary tangle (NFT) formation in the CA1 field of hippocampus is sufficient to develop dementia was the subject of several studies, underlining in the oldest old different cortical areas involved in the dementing process.^46-48 These studies led to the observation that NFT is a stronger correlate of cognitive decline than senile plaques (SP).^{45, 48-50} Another major contribution was the report of several cases presenting with complete Bálint’s syndrome as the first sign of Alzheimer’s disease. ^{51, 52} These cases showed a very particular, and specific, distribution of Alzheimer-type lesions in the visual association cortex, and these studies contributed greatly to our understanding of degenerative conditions, which are now generally referred to as posterior cortical atrophy.

Figure 7.

A slight decrease in brain weight during normal brain aging, based on the autopsy materials of the Geriatric Hospital of Geneva (between the years 1973 and 2009).

Figure 8.

Percentage of neuropathologically confirmed Alzheimer’s disease cases in the Department of Psychiatry and Department of Geriatrics (HOGER). The difference is due to the more specific psychogeriatric profile of the patients from the Department of Psychiatry.

Figure 9.

Representative examples of normal brain aging and brain atrophy in neurodegenerative disorders: (a, d) normal brains (74-year-old and 104-year-old persons, respectively; (b) typical fronto-parieto-temporal atrophy in Alzheimer’s disease; and (c) frontotemporal atrophy with the characteristic “knife-edge” atrophy (arrow) of the superior temporal gyrus in Pick’s disease.

Closer collaboration with psychiatrists and geriatricians on clinicopathological correlations promoted the use of modern clinical methods (Mini-Mental State Examination,⁵³ Clinical Dementia Rating Scale),⁵⁴ allowing validated neuropsychological tests to be correlated with neuropathological data. These collaborations made possible more detailed clinicopathological studies such as the pathological background of apraxia and agnosia in Alzheimer’s disease.^55-58 Specifically, more detailed clinical evaluation at onset was associated with semiquantitative neuropathological evaluation and, thereafter, with the stereologic quantification of the total number of neuron or degenerative lesions (NFT, SP) in various cortical and subcortical areas. ^{45, 48-50, 55-78}

In recent years, the neuropathology group expanded from investigating the aging brain and Alzheimer’s disease (Figs. 10 and 11) and used immunohistochemistry to explore other neurodegenerative illnesses (e.g., Lewy body disorders; Fig. 12a–c, or atypical parkinsonian syndromes; Figs. 12d and 13). The study of dementia associated with idiopathic Parkinson’s disease (PD) has pointed to the crucial role of limbic areas in developing cognitive decline during PD.⁷⁹

Figure 10.

Illustrative examples of Alzheimer type histological lesions. Neurofibrillary tangles (arrows in a, c, and e) and senile plaques (arrowheads in b, d, f); with classical stains (a,b: hematoxylin-eosin, c,d: Globus silver impregnation); and immunohistochemical methods (e) anti-tau immunohistochemistry with antibody AT8, and (f) anti-amyloid β immunohistochemistry with antibody 4G8. Scale bar: 50 μm.

Figure 11.

Alzheimer-type histological lesions. (a) neurofibrillary tangles (arrow) and senile plaques (arrowhead); (b–d) capillary structure of the neocortex. Note the presence of congophilic angiopathy in Figure 11c (arrow). (a) Thioflavin fluorescence and (b–d) modified Gallyas silver impregnation. Scale bar: (a, c, and d) 400 μm; (b) 50 μm.

Figure 12.

Synucleinopathies. (a, b) brainstem-type Lewy bodies; (c) pale body (arrows) in the substantia nigra in idiopathic Parkinson’s disease; and (d) glial cytoplasmic inclusion in oligodendrocytes in multiple system atrophy. (a–c) hematoxylin-eosin; and (d) anti-α-synuclein immunohistochemistry. Scale bar: (a, b) 20 μm, (c) and, and (d) 40 μm.

Figure 13.

Progressive supranuclear palsy. (a) globoid neurofibrillary tangle, (b) thorn-shaped astrocytes, and (c) coiled bodies. (a, b, and c) anti-tau immunohistochemistry. Scale bar: 100 μm.

To elucidate the pathological background of vascular dementia and its relatedness to Alzheimer’s disease, a microvascular ischemic score was developed. This permitted the assessment of the type and localization of small histological ischemic lesions (lacunes, microvascular cortical scars, or demyelination; Figs. 14 and 15) that have a cognitive impact on vascular and mixed dementia cases.^80-83 Further, the possible role of IgG was investigated in the physiopathology of degenerative lesions in normal aging and Alzheimer’s disease. ^84-86

Figure 14.

Lacunes (arrows) and deep white matter demyelination (arrowhead) (Luxol-van Gieson stain).

Figure 15.

(a) Loss of myelin due to multiples cortical microinfarcts. (b) lacune in the substantia nigra. (a,b) Luxol-van Gieson stain. Scale bar: 400 μm.

Extending into neuropsychological disorders, a large study of schizophrenic brains showed a substantial decrease of cortical thickness in anterior cingulate cortex;⁸⁷ late-life and post-stroke major depression was also investigated.^88-92

Finally, the collection includes 57 frontotemporal dementia (FTD) cases, with sporadic forms, progressive supranuclear palsy, corticobasal degeneration, and a neuropathologically confirmed FTD-family with representatives from three generations.^93-97 Detailed studies of lesion distribution in some rare conditions such as asymmetrical AD (Fig. 16), ^{65, 98} fatal familial insomnia,⁹⁹ Down syndrome,^{72, 100} a huge asymptomatic bilateral frontal arachnoid cyst (Fig. 17),¹⁰¹ Lowe syndrome,¹⁰² adult Dandy-Walker variant,¹⁰³ dementia pugilistica, ^{69, 104} postencephalitic parkinsonism, ¹⁰⁵, Pick’s disease (Figs. 18a, b, and e), ¹⁰⁶ corticobasal degeneration (Figs. 18c and d) ^{107, 108} or progressive supranuclear palsy (Fig. 13)^{109, 110} were also reported.

Figure 16.

A rare case of asymmetrical Alzheimer’s disease.

Figure 17.

Frontal arachnoid cyst.

Figure 18.

(a) a few, and (b) numerous Pick bodies (arrows) in the granule cell layer of the dentate gyrus of hippocampus; (c,d) corticobasal bodies (arrows) in the neurons of the substantia nigra; and (e) severe temporal atrophy in Pick’s disease. (a,b) anti-tau immunohistochemistry; (c,d) hematoxylin-eosin; and (e) Loyez stain. Scale bar: (a–d) 40 μm.

The collection today

Today, the collection holds more than10,000 formalin- of buffered formalin–fixed brains (post-mortem interval ranges from 30 minutes to 72 hours; Fig. 19) with a large spectrum of clinical and neuropathological diagnoses (Table 1). The collection continues to grow with the new authorized autopsy cases from both the Departments of Psychiatry and Geriatrics at the University of Geneva. The main aim of these autopsies is to establish the neuropathological diagnosis, and for each autopsy, full medical documentation is available. For diagnostic procedures, the brains are cut into 1 cm–thick coronal slabs, and, in the presence of macroscopic extra- or intra-cerebral lesions, photographs are taken. We have access to medical records and provide macroscopic neuropathological documentation as well, although histologic documentation is not performed electronically on a regular basis. When the postmortem delay is less than 24 hours, or for specific studies, fresh brain tissues are taken at the time of the autopsy. In each case, a full neuropathological report is produced and communicated to the clinicians. Thereafter, brains can be used for teaching and research activities and are preserved in formalin (since 2009, in buffered formalin). Use of brain specimens for research and teaching is reviewed and controlled by the ethics commission of the University of Geneva School of Medicine. The collection is maintained entirely through resources from the Department of Psychiatry.

Figure 19.

A partial view of the more than 10,000 fixed brains of the Geneva collection.

Table 1. A non-exhaustive list of diagnoses in the Geneva brain collection.

Diagnostic	Number of cases	Mean age at death in years (range)	W/M
Normal brains	371	76.9	200/171
Alzheimer’s disease, total	1555	82.3 (48-101)	1081/471
Early onset Alzheimer’s disease	93	68.5 (48-86)	51/42
Parkinson’s disease*	175	81.3 (50-100)	76/99
Lewy body dementia (since 1998)	82	81.7 (58-100)	35/47
Progressive supranuclear palsy	8	84.1 (78-88)	2/6
Corticobasal degeneration	3	74.3(68-81)	0/3
Multiple system atrophy	1	66.0	0/M
Frontotemporal dementia	57	72.7(48-90)	38/19
with tau-positive inclusions (Pick’s disease)	18	73.7 (64-90)	14/4
with ubiquitin-positive positive inclusion**	26	71.3 (56-89)	15/11
without tau- or ubiquitin-positive inclusions	13	72.0 (48-85)	9/4
Amyotrophic lateral sclerosis	9	69.2 (62-80)	8/1
Huntington’s disease	16	61.0 (32-80)	5/11
Postencephalitic parkinsonism	7	49.4 (32-70)	5/2
Dementia pugilistica	1	58.0	0/1
Multiple sclerosis	16	66.2 (30-88)	12/4
Syphilitic encephalopathy	79	54.5 (26-84)	23/56
Chronic alcoholism with Korsakoff syndrome	14	63.4 (38-81)	9/5
Chronic alcoholism with Morel’s laminar sclerosis	4	54.8 (37-76)	1/3
Schizophrenia	116	68.3 (22-95)	69/47
without neuroleptic treatment	34	57.5 (22-79)	14/20
with neuroleptic treatment	82	71.5 (22-95)	55/27
Postmeningitic hydrocephalus	4	41.7 (21-71)	1/3
Down syndrome	2	60.0 (55-65)	1/1
Mental retardation, ~50% with brain malformations	70	37.1 (4-86)	31/39
Sturge-Weber syndrome	1	40.0	1/0
Lowe syndrome	1	23	0/1
Total	10154	3 - 104	5907/4247

^*including 2 cases with deep brain stimulation

^**TDP 43 and FUS immunohistochemistry as been initiated.

If necessary for the diagnosis, paraffin-embedded whole brain slices with classical staining methods (e.g., Nissl, hematoxylin-eosin, Bodian, Luxol Fast Blue-Van Gieson) are produced. More than 200,000 histological slides are available in the collection. Some staining methods like Schroeder or Loyez (Fig. 20), often employed in the past century, however, are no longer used.

Figure 20.

Schroeder (a) and Loyez (b) myelin-staining.

For diagnostic purposes, the Globus silver impregnation (Fig. 21) is used in each case, which allows for the visualization of vascular and most degenerative microscopic lesions such as NFT, SP, or Pick bodies. Antibodies anti-tau and anti-amyloid proteins have been in use since 1990 on frozen sections in cases with dementia. In the last decades, a rapid progression of the field of immunohistochemistry permitted a greater diagnostic accuracy. Today, many antibodies (against tau, amyloid, ubiquitin, α-synuclein, GFAP, α-B-crystallin, TDP43) are used routinely on paraffin-embedded slides for diagnostic purposes and also for quantitative analyses. For the routine diagnosis of dementing diseases, semi-quantitative measurements are used for defining the neuropathological severity. Stereologic analyses are now used as a gold standard for all clinicopathological studies. ^{59, 61, 63, 64, 68, 70, 77}

Figure 21.

Visualization of neurofibrillary tangles and neuropil threads with Globus silver impregnation in the (a) parietal, (b) temporal neocortex, and (c–e) the substantia nigra. Scale bar: (a) 300 μm; (b,c) 100 μm; and (d,e) 30 μm.

To date, this collection, remarkable for its rich historical background and its size, has generated a number of neuropathological discoveries and continues to be used, locally and in the context of international collaborations, to study the histopathologic and clinical characteristics of brain diseases. As for other historical and large-size collections, some of which are discussed in the present volume, the challenge for the future lies in their being adequately curated, documented, and preserved so that their value as unique resources continues to stimulate collaborative research in neuropsychiatric disorders.