Hans Förstl

(the young medical student on the right is FH Lewy while working on his dissertation in Alzheimer's laboratory)

Friedrich Lewy

on Lewy-Bodies, Parkinsonism and dementia.

(the first description of dementia with Lewy bodies)

F. H. Lewy (1885—1950) recognized characteristic globoid and elongated inclusions with a surrounding hab (Lewy bodies) in the nerve cells of the dorsal nucleus of the vagus and in Meynert‘s nucleus basalis of patients with shaking palsy. Excerpts of his influential publications from 1912 and 1913 are translated. In 1923 he published a monograph which contains neurological, psychiatrie and neuropathological data about 43 patients with Parkinsonism; 21 were demented, 10 had affective disturbances and 12 were considered as mentally normal. Alzheimer tangles and/or plaques were found in the cortices of 12 patients. Struetures corresponding to Lewy bodies were described in the brains of 13 patients, but only once in the neocortex. Lewy‘s original work may deserve interest in view of current research on the prevalence and nature of the Lewy body and the nosological confusion caused by its rediscovery.

Friedrich H. Lewy studied medicine in his hometown Berlin and in Zürich. From 1909 he worked in Alzheimer‘s neuropathology laboratory at Kraepelin‘s psychiatric university hospital in Munich. He began to study serial sections from patients with ‘paralysis agitans‘ (Pa; shaking palsy). The first article in which he described his seminal finding was written in 1910 and published as his contribution to Lewandowsky‘s Handbook of Neurology (1912). He reported similar findings in 1913. In the same year Lewy followed Alzheimer to Breslau where he became director of the psychiatrie hospital‘s laboratory. He joined the German army during the first world war. In 1919 he returned to his work on Pa. The manuscript for his monograph on ‘Muscle tone and Movement‘, a book with 672 pages and 569 figures, was completed within three years while he worked at the medical department II of the Charite in Berlin. In 1934 he emigrated to the USA where he held several visiting professorships. In 1947 he was appointed professor of neuroanatomy and neuropathology in Pennsylvania. He died as ‘Frederic H. Lewey‘ in 1950 (Anonymous, 1951; Winkelman, 1951).
Tretiakoff (1919) was the first to use the term corps de Lewy for eosinophilic bodies in the substantia nigra. ‘Lewy bodies‘ have been found in 7—10% of the normal (preclinical?) population over 60, in cases of striatonigral degeneration, olivopontocerebellar atrophy, progressive supranuclear palsy in an broad speetrum of other neurodegenerative disorders (Gibb, 1986) and in a percentage of neuropathologically verified Alzheimer‘s disease and other dementing conditions which have received increasing attention in recent years (see Discussion).

Lewy‘s first description of the ‘Lewy bodies‘, based on the examination of 25 patients with Pa and without dementia, was published in 1912. He had observed characteristic changes in the dorsal nucleus of the vagus nerve in a ‘large number‘ of patients who showed ‘tremor and tension of their vocal cords or larynx‘. He had not found such features in six cases of ‘severely senile or arteriosclerotic‘ patients (Lewy, 1912, pp. 924—925):

… These changes can be characterised as deposits whose origin appears to be related to forms which have been portrayed by Lafora. He—perhaps not quite rightly considered them as corpora amylacea. Thus they show some reactions similar to Corpora amylacea, but they are smaller, are shaped irregularly, and do not have their typical glass-like and laminated onion-bulb appearance. Other forms can also be found, but I cannot discuss their development in this context. Using Mann‘s stain we can see characteristic features of ball-, cord- or serpent-like forms, which are stained in shining red and lying in a blue mass of plasma. As can be demonstrated convineingly by serial slices, these different figures are only cuts or fragments of the same serpent-like, twisted deposit. Such sections show further that there are no neuronal nuclei in these structures, but glial nuclei which may have immigrated to prepare their dissolution. lt is also possible that some of these masses without a nucleus had earlier been ganglion cells with a nucleus, because apart from their often rather characteristic form, quite similar deposits can also be found in definite ganglion cells, which show a weIl-preserved nucleus. The circumference of such a mass of plasma exceeds that of a ganglion cell by up to three times and has globoid tumescences on the periphery which—when tangentially—appear like isolated balls. The plasma shows a peculiar honeycomb structure and is permeated by small vacuoles. Occasionally lumps can be found, which may correspond to remnants ofNissl bodies. Fibrils can never be demonstrated with absolute certainty. The histochemical reactions are quite typical. 1 cannot discuss here the histological details of simple senile dementia, but 1 would like to mention that 1 have found nuclei of satellite cells which were tranformed regressively into red globoids and that 1 have also found a small red globoid in a ganglion cell—but 1 have never observed any of these complicated structures. There is good evidence that similar changes can occur in axon-cylinders. 1 have already mentioned that such changes were—possibly by chance—present in every case which showed laryngeal tremor. Conversely, only one of those patients in whom this symptom could not be observed at the time of the clinical examination showed these changes in a limited number. They were not present in any of the others. The latter finding does not of course rule out the possible presence of a tremor in an earlier stage or that it may just have begun to develop. Even if—as 1 have mentioned—this coincidence may have occurred by chance, it should stimulate increased attention to the localisation of possible pathological processes.

Lewy mentioned similar findings in the basal nucleus of Meynert (Lewy, 1912, p. 926):

...In Meynert‘s magnocellular nucleus of the Substantia innominata 1 have always observed a cellular degeneration similar to the changes in the dorsal nucleus of the vagus nerve, preferentially of a honey-comb or granular structure. These findings exceeded senile changes by far. The lateral thalamic nucleus ... and particularly the aggregation of large cells, which is called paraventricular nucleus, participate in these changes. Our anatomical and especially physiological knowledge about these nuclei of the subthalamic area are still too inferior to understand more than the simple topographic relationship of the different groups of nuclei. With regard to the similar structure of the cells this could be one coherent aggregation of cells reaching from the Substantia nigra (Blocq and Marinesco found it destroyed by a tumour in a case of shaking palsy) to the septum pellucidum—a construction which would be similar to the reticular formation....

(Marinesco. Sur un cas de tremblement parkinsonien hmiplgique symptomatique d‘une tumeur du pedoncle cerebrale. Soc. de Biol., 1893.
Lafora and Glück. Beitrag zur Histopathologie der myoklon. Epilepsie. Zeitschr. f. d. ges. Neurol. u. Psych.6, 1911.)

When Lewy presented bis update to the Society of German Neuropsychiatrists in 1913 he had already investigated 60 cases of Pa. His prime suspects, responsible for Pa, were the basal ganglia. With regard to the ‘Lewy bodies‘, he had now shifted the emphasis from the dorsal nucleus of the vagus to Meynert‘s nucleus, which he discussed immediately after the Globus pallidus (Lewy, 1913, p. 53):

... The second major seat of the changes and therefore the other one which presents further indications to the disease process can be found in the so-called Meynert nucleus of the Ansa peduncularis or nucleus of the Substantia innominata. This nucleus, about whose connections and funetions no details are known, extends from the optic tract to the Septum pellucidum and contains large multipolar ganglion cells reminiscent ofmotor neurons. lt sometimes appears as if it would embrace the fornix and reach out to the wall of the ventricies. Detriment could be found within this area. But it would also be possible that these are peculiar changes of axoncylinders. The cells of this nucleus bear all signs of a senile change as described in cortex cells of senile dementia. They show the so-called Alzheimer changes of the fibrils besides extreme lipoidosis. In parallel, Mann‘s stain shows a globoid or oblong condensation of plasma, which yields some reactions similar to Corpora amylacea, but which can also be stained by Eosin. The plasma shows vacuoles, the core becomes detached and finally only these globoid forms, which are reminiscent of Corpora amylacea, remain lying in the tissue. A glial scar ends the process. Specific symptoms caused by the decay of this nucleus are not conceivable. The changes in the so-called sympathetic nucleus of the vagus nerve at the bottom of the fourth ventricle are very similar to the changes described above....

In his monograph about ‘Muscle Tone and Movement; a Systematic Investigation of the Clinical Symptoms, Physiology, Pathology and Pathogenesis of Shaking Palsy‘ (1923), Lewy expanded his description of the histological changes which he considered specific for Pa. He knew that other researchers had described ‘Lewy bodies‘ in the substantia nigra of patients with Pa (‘. . . Tretiakoff has found the changes in 9 cases and Souques in 3 cases of typical Pa. Lhermitte and Corni pointed out that they have found similar changes in the substantia nigra ofpatients with tabes dorsalis, syringomyelia, tumours and transverse lesions of the spinal cord;‘ p. 280), where he has only observed ‘glassy decomposition, nuclear degeneration, and eosinophilia‘ (p. 279). By that time Lewy had examined 88 cases including a large number of patients with psychiatric impairment. The following paragraph is translated from his chapter on ‘The relationships of the clinical picture to the postmortem findings‘ pp. 359—360):

Senile dementia was diagnosed in 28 cases, arteriosclerotic dementia in 2 cases. Alzheimer‘s fibrillary changes were found in the cases 3, 11, 12, 14, 41, 49, 59, 63, 64, 65, 82. Apart from case 59, who was clinically completely normal, a considerable disturbance of the intelligence and memory prevailed, and in cases 14 and 64 primarily a disturbance of affect. Cases 14, 49, 82 showed only fibrillary changes, in 11 cases there were fibrillary changes together with plaques. These were present in cases 3, 11, 12, 28, 41, 59, 60, 63, 64, 65, 80, 82, 92; 59 and 61 who did not have a mental defect. Particularly extensive lipoid-deposits were present in 41, 71, 74, 81, 90; the last two were clinically normal. All these statements refer exclusively to the cortex. lt is remarkable that the early, long-lasting and severe cases are neither among those with fibrillary changes, nor among those with plaques. Accordingly the characteristic fibrillary changes in the basal nucleus are absent in these cases and other pathology prevails. Fibrillary changes in Meynert‘s nucleus are usually, but not always accompanied by such changes in the cortex.  (...) Fibrillary changes and numerous plaques were only found in such cases, who showed severe defects of memory of memorizing (Merkfähigkeitsdefekte). But cases exclusively showing extreme fatty cell-degeneration (‘Zellverfettung‘) were also in this clinical group. The purely paranoid patients did not show characteristic morphological findings. ...

Lewy did not comment on any relationship between ‘Lewy bodies‘ and cognitive impairment. He documented the results on which his statements were based on five large fold-out tables. In the hope of finding some indications of the prevalence and significance of ‘Lewy bodies‘ in his cases, we have tried to abbreviate and organize the data on the 43 patients for whom neurologic, psychiatric and neuropathologic descriptions were available (Table 1). The alphabetical order was kept to permit comparability with the original. There is some disagreement between text and tables. Empty cells cannot be regarded as representing ‘normal results‘, but should be considered as ‘missing data‘.

A typical description of the ‘Lewy body‘ was not presented in a single case. All of the 43 patients (13m; 30f) suffered from Pa; 21 (7m, 14f) were reported as being demented. Ten (2m, 8f) showed affective disturbances which were accompanied by paranoid ideas in three patients, by hallucinations in three, and in two by mild cognitive impairment. Twelve patients were described as psychiatrically normal. Cortical tangles (Alzheimer fibrils) and/or plaques were described in 12 of 37 patients, nine of them demented, two with affective disturbance, one with mild cognitive impairment and one normal. Seven demented patients showed neither plaques nor tangles in the neocortex. Lewy suspected neurosyphilis in one of those cases (Tschuschke), ‘severe cell loss‘ in another (Waldow), and a ‘peculiar process‘ in a third (Poinke). The four remaining cases had ‘glassy degeneration‘ or fuchsin-stained granules in the brainstem (Köhler, Madaus, Ziegler); in one ofthem this finding is not presented in the table, but as a figure in the chapter on Meynert‘s nucleus (Lewy, 1923, p. 268). (Fig. 1).

The legend to this figure reveals Lewy‘s view of the ‘Lewy body‘ changes as a dynamic process. He did not refer to the corresponding findings by a single term, but used different descriptions (eg ‘fuchsinophilic‘ or ‘argentophilic granules‘, ‘glassy degeneration‘, etc) depending on the surmised stage and the staining method used. Therefore the frequency of the changes which might correspond to ‘Lewy bodies‘ is certainly underestimated in the table (eg Stief). ‘Glassy degeneration‘ was described in the brainstem of seven patients (one demented, three with affective disturbance, three normal). Fuchsin-stained granules were present in six patients, all of them demented. Predominant gait disturbance was found in 10 of 13 patients (77%) with ‘glassy degeneration‘ of fuchsin-stained granules compared to only 20 of 43 patients (47%) without such changes (p < 0.05; x2-test). Only one patient (Weber) seems to have shown fuchsin granules in the neocortex. She suffered from dementia and also had abundament plaques and tangles.

Lewy explained his views on the relationship of Pa, dementia and ageing in his summary of the clinical findings (Lewy, 1923, pp. 42, 43):

... The psyche is seldom spared in the course of illness. The mental symptoms may have different characteristics depending on the age of their onset, but all belong to the group of senile dementla. In the early cases (Pa beginning before the age of 50; transl. note) anomalies ofaffect prevail, partly ofa euphoric, partly ofa depressive nature; they are often associated with hallucinations and paranoid ideas derived from them. In the later years, the disturbance of memorizing (Merkfähigkeit) and of memory (Gedächtnisschatz) predominate and lead to the most severe stages of dementia in a small number of patients....
Lewy felt that a distinct line between normal ageing and Pa could not be drawn and that Pa should therefore be considered as a form of senility with an early onset and a primary target in the subcortical areas of involuntary motor control, whereas senile dementia had its primary target in the cortex ...
(1923, p. 43).

Obviously Lewy had selected his cases because they had Pa and not because they were demented. A dementia rate of almost 50% in a series of patients with motor disorders is strikingly high and seems to conflict with Lewy‘s statement in the last sentence translated above. The postmortem series drawn from his 88 cases may be biased towards the more severe forms ofiliness. The question today is whether some of his patients would be classified as neuropathologically confirmed Alzheimer‘s disease rather than as Parkinson‘s disease with accompanying cognitive impairment. The features of one patient (Poinke), which are described in greater detail in Lewy‘s original table, are suggestive of corticobasal degeneration (Gibb et al., 1989a). Another patient (Köhler) might receive a diagnosis of ‘encephalitis subcorticalis chronica progressiva‘ or Binswanger‘s disease (Binswanger, 1893). Some of the cases would probably now be diagnosed as mixed forms of Parkinson‘s and Alzheirner‘s disease, others as ‘Lewy body variant of Alzheirner‘s disease‘, who are showing concomitant plaques, tangles and Lewy bodies (Hansen et ei., 1990). The presence of severe Parkinson syrnptoms would rule out a diagnosis of ‘senile dementia of the Lewy body type‘, which was recently defined as a form of dementia presenting with confusional states, fluctuations, hallucinations and mild extrapyramidal features (Perry et ai., 1990). None of Lewy‘s cases is compatible with a diagnosis of ‘cortical Lewy body dementia‘ in the sense of dementia with Lewy bodies cortical, but without significant Alzheimer pathology (Gibb et ai., 1989b). Gait impairment, agitation and hallucinations early in the course of disease have been described as being characteristic for ‘diffuse Lewy body disease‘ (Crystal et al., 1990). Lewy‘s patients who showed conspicuous changes in the brainstern (‘g.d.‘ or ‘fuchsin‘ in Table 1) had a significantly higher rate of gait disturbance. Further clinical features rnay not have been reported in sufficient detail and the absence or presence of characteristic histological changes in the cortex do not appear to be documented systematically enough to permit any other conclusions from Lewy‘s material.

Lewy‘s clinical criteria for a diagnosis of ‘Pa‘ or of ‘dementia‘, his terminology, his description of clinical signs and of psychopathological symptoms, which were only briefly summarized in the original tables, have to be considered critically— even though he was trained in Kraepelin‘s department, which had once set international standards. The terminological inconsistencies, and inconsistencies between text and table are puzzling. The gaps in the neuropathologic data and their unstandardized documentation are disturbing. The reasons are to be sought in Lewy‘s difficult working conditions, in the speed with which he worked and in a meticulous approach to the characteristic features of each individual patient. Lewy‘s work shows the influence of his teacher Alzheimer and it may have been this style that enabled hirn to recognize the ‘Lewy body‘; the white matter disease in demented patients with Alzheimer pathology (Markfraß, see Table 1); the relationship between tangles in Meynert‘s nucleus and the neocortex (1923, p. 360); and the finding that elderly paranoid patients did not show one set of characteristic morphological findings (1923, p. 360), another issue that is currently under discussion. A rnore standardized approach and a rnore quantitative evaluation ofhis abundant clinical and neuropathological results rnight have led hirn to discern a link between the ‘Lewy body‘ and cognitive dysfunction. He shared the fate of devoted neuroscientists of the nineteenth century, who had important data readily available but who were not ready for the use of simple statistical rnethods to exploit their material (Förstl, 1991). In the introduction to his monograph (1923, p. IV), Lewy mdicated that he was aware of shortcornings in bis method and suggested that different approaches should be employed in the future. However, in view of the present state of affairs concerning the ‘Lewy body‘ and its significance, one feels that such different approaches would not necessarily lead to greater clarity in his work.

(a complete version of this paper by Hans Förstl and Raymond Levy was published in the "Int J Geriat Psychiat" 1991)

REFERENCES (selected)
Anonymous (1951) Obituary: Frederic H. Lewey. Arch. Neurol. Psychiat. 66, 114—115.
Binswanger, 0. (1894) Die Abgrenzung der allgemeinen progressiven Paralyse, 1-111. Berl. Klm. Wochenschr. 49, 1103-1105, 1137—1139, 1180—1186. Crystal, H. A., Dickson, D. W., Lizardi, J. E., Davies, P. and Wolfsen, L. 1. (1990) Antemortem diagnosis of diffuse Lewy body disease. Neurology 40, 1523— 1528.
Lewy, F. H. (1912) Paralysis agitans. 1. Pathologische
Anatomie. In Handbuch der Neurologie, Vol. 3/11 (M. Lewandowsky, Ed.). Julius Springer, Berlin, pp. 920— 933.
Lewy, F. H. (1913) Zur pathologischen Anatomie der Paralysis agitans. Dtsch. Ztschr. Nervenheilkunde 50, 50—55.
Lewy, F. H. (1923) Die Lehre vom Tonus und der Bewegung. Zugleich systematische Untersuchungen zur Klinik, Physiologie, Pathologie und Pathogenese der Paralysis agitans. Julius Springer, Berlin.
Tretiakoff, C. (1919) Contribution a l‘etude de l‘anatomie du locus niger de Soemmering avec quelques deductions relatives ä la pathogenie des troubles du tonus musculaire et dc la maladie de Parkinson. These de Paris.
Winkelman, N. W. (1951) Doctor Frederic H. Lewey, January 28, 1885—October 5, 1950. J. Neuropath. Exp. Neurol. 10, 85.

Fig. 1. Patient Stief. No. 30. Cells and forms from Meynert‘s nucleus. 347—3 50: drop-like and globoid decomposition of the Gell plasm, with partial detachment of the nucleus. Other cells show the formation ofa single, large inclusion body in a degenerating cell. 351, 352: strong silver impregnation of such a body. 353: final stage; only the inclusion body is left. 354: the detached inclusion body lies in a hab and is covered with argentophilic granules. 355: inclusion body with rest of tangles; on the right a (fibre-forming?) glial nucleus. 356: detached inclusion body. Drawing: Bielschowsky. Immersion 1.3 Comp. Eyepiece 6