"Normal Pressure" Hydrocephalus

LE JOURNAL CANADIEN DES SCIENCES NEUROLOG1QUES

Neurofibrillary Tangles in the Dementia of "Normal Pressure" Hydrocephalus

MELVYN J. BALL

SUMMARY: Routine neuropathologi- neuropathologically confirmed INTRODUCTION cal examination could not explain the Alzheimer's disease. The density of the Since the report of Adams et al dramatic improvement exhibited by one one who recovered was within normal (1965) of dramatic cures from shunt patient with "normal pressure" hydro limits. ing of hydrocephalic patients with cephalus after shunting. The improved The duration of dementia before dementia, ataxic gait, and inconti patient contrasted remarkably with the shunting, and the total duration of de unchanged condition of four others also mentia in these five patients rank in the nence, the syndrome of "low pres shunted successfully. The five brains same order as their degree of neurofibril sure" or "normal pressure" hydro were analysed by quantitative mor lary formation. Furthermore, a positive cephalus (NPH) has intrigued clin phometry to determine the degree of linear correlation exists between the ical neuroscientists. Several unsol neurofibrillary tangle formation in mes Tangle Indices and the total duration of ved mysteries of this disease remain ial temporal neurons. The density of dementia. The data suggest that early (Sugar, 1976). The biophysical tangle-bearing nerve cells in the four diagnosis may improve the chances of mechanism of its development, unimproved cases was markedly greater reversing the dementia of normal pres which should link the altered cere than in age-matched control brains from sure hydrocephalus before histological brospinal fluid (CSF) dynamics with nineteen normal subjects, and fell in the alterations prove too severe. the symptoms and signs, is not clear same range as that of eight dements with (Hakim and Adams, 1965; Gesch- wind, 1968; Ojemann et al, 1969). The clinical and laboratory criteria required to substantiate the diag RESUME: Un examen neuropatho- dements atteints de maladie d'alzheimer logique de routine ne pourrait expliquer confirmee neuropathologiquement. La nosis are imprecise (Messert and I'amelioration marquee apres "shunt densite de ces cellules anormales chez Wannamaker, 1974). Despite ing" exhibee par un patient avec hydro celui qui fut gueri etait dans les limites numerous radiographic studies cephalic normotensive, quandon la com de la normale. (pneumoencephalography, radio pare avec celle de quatre autres La duree de la demence avant iopera isotope scans, angiography, patients egalement operes avec succes tion et la duree totale de la demence echoencephalography, com dont la condition est demeuree in- chez ces cinq patients s'echelonnaient puterized axial tomography) and changee. Les cinq cerveaux furent an selon le meme ordre que leur degre alyses par morphometrie quantitative de formation neurofibrillaire. En plus, other laboratory investigations (e.g., pour determiner le degre de developpe- une correlation positive lineaire existe electroencephalography; saline in ment des formations neurofibrillaires entre I'indice des formations neuro trathecal infusion manometry), no dans les neurones mesio-temporaux. La fibrillaires et la duree totale de la de single examination or combination densite des cellules nerveuses porteuses mence. Ces donnees suggerent qu'un of clinical and laboratory abnor de formations neurofibrillaires dans les diagnostic precoce pourrait augmenter malities has been found that will quatre cas non ameliores etait re- les chances de reversion de la demence predict which patients with the syn marquablement plus grande que dans les dans I'hydrocephalic normotensive drome will respond favorably to cerveaux controles de dix-neuf sujets avant que les alterations histologiques normaux du meme age, et etait du ne soient trop severes. shunting procedures (Bannister, meme ordre de grandeur que celle de 8 1972; Wolinsky et al, 1973; Adapon et al, 1974; Stein and Langfitt, 1974). To date, very few patients with NPH, whether shunted or not, have been studied at autopsy. The brains of two patients treated unsuccess fully by ventriculoatrial shunt (Lorenzo et al, 1974) were said to From the Departments of Clinical Neurological Sciences and Pathology, University of Western On have multiple lacunes, consistent tario, London, Canada. with their histories of systemic Reprint requests to: Dr. M. J. Ball, Department of hypertension; and one of these also Pathology, University Hospital, 339 Windermere showed arteriolar sclerosis and foci Rd., London, Ontario, Canada N6A 5A5. of demyelination in the centrum

Vol. 3, No. 4 NOVEMBER 1976 - 227

Downloaded from https://www.cambridge.org/core. IP address: 170.106.40.219, on 27 Sep 2021 at 06:40:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0317167100025348 THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES

semiovale, characteristic of with an ataxic gait. His wife also sclerosis reflected the long-standing Binswanger's subcortical hyperten noted increasing loss of memory for systemic hypertension. The brain, sive encephalopathy. the previous five months and loss of weighing 1200 grams, showed mini Ernest et al (1974) reported nec interest in daily affairs. There was mal parasagittal thickening of the ropsy findings on two patients who no urinary incontinence. Blood leptomeninges, moderate patchy had had diabetes and systemic pressure on examination was atherosclerosis of the circle of Wil hypertension, both of whom were 240/120 mm. Hg. In addition to in lis, mild dilatation of the lateral ven considered to have clinical and appropriate moods including crying tricles, and no gross atrophy of the radiographic changes typical of without provocation and poor con cortical ribbon. Numerous tiny cys NPH. Their 66 year-old man had a centration, the man exhibited a posi tic infarcts were seen in the thalamus moderately good response to ven- tive glabellar tap, a sometimes bilaterally. Microscopically, moder triculoperitoneal shunt and expired shuffling gait, and a tremor of the ate concentric hypertrophy of the of hypertensive cardiac disease one hands, suggesting some Parkin walls of small arteries and some fib year later. Their 45 year-old woman sonian element to his problem. CSF rinoid necrosis of arterioles were ac died unimproved only a month after pressure at lumbar puncture was 110 companied by multiple old perivas ventriculojugular shunt, of pulmo mm. The cerebrospinal fluid was cular infarcts — "lacunes" (Fisher, nary embolism. Both brains con normal. Pneumoencephalography 1965; 1969), in the lentiform nuclei, tained many old cystic "lacunar" showed marked dilatation of the lat thalamus, subcortical white matter infarcts throughout cerebrum and eral ventricles and the third venticle, of frontal and temporal lobes, basis cerebellum, but especially in the with some air in widened cortical pontis, and cerebellar white matter. basal ganglia. The authors suggested sulci over the Sylvian regions al No senile plaques or neurofibrillary that the initial pathological process though none over the frontoparietal tangles were seen in sections of fron in some cases of NPH might be convexities. Radioiodine-labelled tal cortex. Rose's hi field of both hypertensive cerebrovascular dis human serum albumin (RHISA) scan Ammon's horns revealed a few ease with the "lacunar state" (mul of CSF by lumbar injection showed miniscule old infarcts with focal tiple small infarcts) in deep cerebral most of the CSF flow was abnormal, neuronal depletion and mild gliosis. grey matter. entering the ventricles in a retro With routine histological sampling, The brains of four other grade direction. A little radioisotope only an occasional pyramidal neuron "idiopathic" cases of NPH have did, however, migrate slowly over in the hippocampi demonstrated shown fibrous and obliterative thic the lateral aspects of the hemis Alzheimer's neurofibrillary degen kening of the leptomeninges, at the pheres, although almost none eration (Alzheimer, 1907). The incisura tentorii or over the cerebral reached the vertex. Pertechnitate neuropathology was essentially that convexities (Heinz et al, 1970; De- brain scan and EEG were not con of hypertensive cerebrovascular dis Land et al, 1972; Sypert et al, 1973). tributory. ease. The cause of the subarachnoid ob While in hospital, he developed a struction was unclear in three, and mild spastic quadriparesis, with The second patient, a 70 year-old secondary to old subarachnoid hypertonic, hyperreflexic limbs. The retail clerk (Patient No. 22), was first hemorrhage in the fourth. Two patient was discharged with the seen at age 64, with a seven-month further patients with CSF dynamics diagnosis of "normal pressure" history of increasing difficulty with indicative of NPH showed communicating hydrocephalus, recent memory. He also complained Alzheimer's disease at post-mortem probably in the developmental or of dizzy spells with transient blur examination (Sohn et al, 1973; Cob- transitional stage. Progressive clum ring of vision, slurred speech, and lentz et al, 1973). siness of his legs necessitated a re- falling to the floor, precipitated by In this report we document the admission three months later, at sudden movements of the head or by autopsy findings on five people which time urinary incontinence was quickly standing up. His failing shunted for NPH, four of whom also noted. Disorientation was more memory and brain-stem symptoms showed no response and one of marked than previously. In view of were initially suggestive of cere whom'improved dramatically. With his deterioration, a right ventriculo brovascular insufficiency. B.P. was quantitation of the neurofibrillary atrial Pudenz shunt was introduced, 140/90. Neurological examination tangle formation in the mesial tem some eight months after the onset of showed only the defect in mentation, poral cortex of these brains, a strik symptoms. Despite a functioning but upon re-admission three months ing difference was found which may shunt, his postoperative mental and later this had worsened to the extent explain the lack of recovery in the neurological status remained un that his overall I.Q. was 64. Aortic four unchanged patients. changed. A steady downhill course arch angiogram showed no atheroma marked by fecal and urinary inconti in the extracranial portions of the Clinical and Pathological Data nence and debilitating dementia ter four cerebral arteries. Pneumoence- The first patient, a 78 year-old minated in bronchopneumonia 2V2 phalogram revealed moderate dilata retired personnel manager (Patient years after his illness began. tion of the ventricles, some air in the No. 21 in the series to be discussed), Autopsy Findings: Hypertensive Sylvian fissures, and no air over the was admitted to hospital at age 76 cardiomegaly and severe nephro- hemispheres. The RHISA scan by

228 - NOVEMBER 1976 Neurofibrillary Tangles in the Deinei tia of "Normal Pressure" Hydrocephalus Downloaded from https://www.cambridge.org/core. IP address: 170.106.40.219, on 27 Sep 2021 at 06:40:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0317167100025348 LE JOURNAL CANADIEN DES SCIENCES NEUROLOG1QUES

lumbar subarachnoid injection cate fibrous tissue (sometimes sug expiring of bronchopneumonia eight showed reflux of CSF into the ven gesting a "lumen within a lumen"), years after his illness began. tricles, but no activity over the con consistent with the changes in pial Autopsy Findings: Adenocar vexities. EEG was non- vessels described in watershed cinoma confined to the prostate was contributory. A right ventriculoatrial (marginal) zones following periods also present. The brain, weighing Pudenz shunt was installed some ten of systemic hypotension (Romanul only 960 grams, was covered on the months after the onset of his trouble. and Abramowicz, 1964). The patient left by a thin (3 mm.) fibrous chronic The intraoperative ventricular pres had experienced severe chest pain subdural hematoma, possibly the re sure, although not measured, was six months before death, and an old sult of a war injury sustained 55 said not to be significantly elevated. myocardial infarct was confirmed at years before death. It did not com Despite the shunt's functioning well necropsy. press or distort the subjacent cere post-operatively, there was no ap Sections of basal ganglia showed a brum in any way. The parasagittal preciable improvement. Two years mild degree of hyaline arteriolar leptomeninges were minimally fibro- later, with the shunt still working sclerosis, mainly in each putamen tic. The shunt tube was in place well, he exhibited inability to dial the and globus pallidus. Mild widening through the right occipital lobe. The telephone, difficulty writing, shaving of perivascular spaces with rarefac circle of Willis was not remarkable. and dressing himself, and he fre tion of the neuropil was accom The narrowed cortical gyri and quently wandered away from home. panied by a few foci of astrogliosis widened sulci seen from the exterior He was severely disoriented to time and hemosiderin-laden mac correlated with a moderate degree of and place and showed some motor rophages; and in the head of the right atrophy of the entire cortical ribbon inco-ordination. In the last four caudate nucleus a single, slit-shaped on coronal sections, which also re years of life, he deteriorated further, old infarct was found. vealed a moderately severe dilata with frequent falls when walking tion of the lateral ventricles affecting The neuropathological picture about the hospital ward, gross men to a lesser extent the third ventricle, was thus not specific, but some fea tal confusion, profound dementia, though not the aqueduct or the tures characteristic of both and incontinence. He died of bron fourth ventricle. A small old cystic Alzheimer's disease and hyperten chopneumonia and ischemic heart infarct was seen in the left parietal sive vasculopathy were noted. disease a little more than six years lobe. after the onset of his disease. Microscopic evidence of two pro Autopsy Findings: The heart, kid The third patient, a 78 year-old cesses was found. Large numbers of neys and other viscera showed no retired telegrapher (Patient No. 23), senile plaques and neurofibrillary definite evidence of systemic hyper was said to have been psychoneuro tangles in the hippocampi were ac tension. The 1200-gram brain had tic for about 40 years. At age 70 he companied by many neurons with only very mild, patchy atheroscle began to suffer from progressive granulovacuolar degeneration and rosis of the circle of Willis, mini memory loss, and three years later many Hirano bodies. Senile plaques mal fibrosis of the parasagittal lep- was grossly disoriented to time and were abundant in frontal, temporal tomeninges, no cortical atrophy, no place, exhibiting severe ataxia of and parietal cortex as well, and in dilatation of the ventricular system, gait and urinary incontinence. the mamillary bodies. Modest num and no other gross changes. Micro RHISA scan of lumbar CSF showed bers of tangles were seen in middle scopically there were large numbers retrograde flow into the ventricles frontal and cingulate gyri and in in of senile plaques in the cortex of all and no evidence of normal absorp ferior temporal gyri. The frontal cor cerebral lobes bilaterally, and in the tion near the superior sagittal sinus. tex exhibited considerable neuronal mamillary bodies. The senile pla Pneumoencephalography revealed fall-out and gliosis, especially in ques were especially numerous in considerable dilatation of the 4th deeper laminae. the hippocampi, where many ventricle, aqueduct, 3rd and lateral In addition, however, the basal neurons contained neurofibrillary ventricles, including the temporal ganglia showed moderately severe tangles, and where granulovacuolar horns. The basal cisterns were en onion-skin hypertrophy of small ar degeneration of Simchowicz (Tom- larged but no air could be seen in the teries and severe hyaline arteriolo- linson and Kitchener, 1972) and the sulci over the convexities. EEG sclerosis, especially in each putamen showed non-localizing high voltage, rod-like bodies of Hirano (Hirano et and globus pallidus, where multiple irregular slow waves. A ventriculo al, 1968) were also prominent fea old infarcts (often with atrial shunt introduced four years tures. Tangles were not seen other hemosiderophages) and widened after the onset of his disease failed to than in the pyramidal layer of the perivascular spaces indicated a modify the clinical picture. Due to lacunar state. Although vessels in hippocampi. A few small arteries temporary cessation of function five and arterioles in the leptomeninges weeks post-operatively, the shunt other cerebral regions were not se and superficial cortex of frontal, was revised to a ventriculoperitoneal verely affected, occasional micro parietal, temporal and occipital connection; but despite its patency scopic old infarcts were seen in both lobes demonstrated mild focal thereafter he remained totally de Amnion's horns and in multiple sites hyaline hypertrophy of the media mented, incontinent and bed-ridden, in the subcortical white matter of and intimal thickening due to deli frontal and parietal lobes; and severe

Melvyn J. Ball NOVEMBER 1976- 229

arteriolosclerosis in the basis pontis sures, suggesting "low-pressure" accompanied by extensive as- was accompanied by similar old mic communicating hydrocephalus. The trocytosis. Senile plaques and/or roinfarctions. RH1SA scan by lumbar injection neurofibrillary tangles were also Thus, gross and histological fea showed reflux of CSF into the lateral noted in mamillary bodies, posterior tures of Alzheimer's disease were ventricles but no activity over the hypothalamus, superior and middle present in a brain which also man convexities. A brain biopsy was per frontal gyri, middle and inferior ifested hypertensive cerebrovascu formed through a right frontal burr temporal gyri, and parietal and oc lar pathology. The severe arterial hole, and a pressure transducer was cipital cortex. A neurofibrillary nephrosclerosis at autopsy was con installed in the right lateral ventricle. tangle was also seen in each locus sistent with the latter, although only Histological examination of the ceruleus. The malformation of the a mildly elevated blood pressure had biopsy showed fibrous thickening of left temporal pole proved to be a been recorded clinically. the arachnoid; and many senile pla capillary telangiectasis involving The fourth patient, a 57 year-old ques and neurofibrillary tangles in cortex and subcortical white matter. housewife (Patient No. 24), began to the cortex, "consistent with A mild subacute leptomeningitis, have spells of weakness, light Alzheimer's disease". Although a ventriculitis and ependymitis, cor headedness and blurring of vision a 48-hour telemetric recording of the relating with post-mortem culture of few times per month, at age 49. In intraventricular pressure did not Staphylococcal organisms from the the next 3 years her husband noted show sustained Lundgren waves, meninges, was attributable to the some alteration of her personality repeated short spikes of raised pres shunt catheter. There was no evi and a shortening memory span. She sure (up to 25 mm. Hg.) were noted, dence of cerebral arteriolar disease had been a good bridge player, but especially during sleep. With these or of multiple old infarctions. increasing forgetfulness rendered changes and the air encephlographic In summary, the biopsy diagnosis her unable to play. At age 52 an and RHISA studies indicating some of Alzheimer's disease seemed ap episode of sudden severe pain in the features of NPH, a ven- propriate (despite the normal brain left eye and temple was accom triculoperitoneal shunt (Hockheim weight); but it also appeared possi panied by diplopia. An ophthal type) was installed. Despite excel ble that a degree of leptomeningeal mologist consulted for the persistent lent flow through the shunt system fibrosis, perhaps secondary to a headache noted left lateral rectus post-operatively, the patient showed minor episode of bleeding years ear palsy, mild left ptosis and slight no neurological improvement, be lier from a small vascular malforma miosis of the left pupil. Blood pres came incontinent of urine and feces, tion, had disturbed CSF flow, result sure was 130/75. Lumbar puncture and died of bronchopneumonia one ing in at least some clinical features four days later revealed an opening month later, some eight years after of "normal pressure" hydro pressure of 120 mm. and normal the onset of her dementia. cephalus. CSF. Bilateral carotid angiography Autopsy Findings: The brain was The fifth patient, the only one who showed no berry aneurysms, but a not atrophic, weighing 1275 grams. responded to shunting (Patient No. "small loop of vessels" was noted in A mild fibrous thickening of the lep- 20), was a 63 year-old foreman who the region of the anterior com tomeninges affected not only the despite therapy for systemic hyper municating artery, in an oblique parasagittal regions but extended out tension had blood pressures of up to view of the left carotid injection. laterally over the convexities to the 240/120 and electrocardiographic Pertechnitate brain scan was nor temporal lobes. The meninges over evidence of left ventricular hyper mal. Diabetes mellitus was diag the Sylvian fissures were also fibro- trophy for at least the last six years nosed with glucose tolerance testing, tic, but the basal meninges and out of life. About two years before his and she was discharged on Tol let foramina of the 4th ventricle were demise, a transient right hemiplegia butamide. Her impairment of mem normal. The vessels of the circle of and aphasia had been noted. While ory worsened gradually, so that by Willis showed neither atheroma or hospitalized for abdominal pain and age 57 she exhibited disorientation berry aneurysms. The lateral ventri depression nine months before to time and place, difficulty with cles were mildly dilated; the remain death, he exhibited early signs of simple commands, dyscalculia and der of the ventricular system was dementia, difficulty with gait, and dysgraphia, and a dressing apraxia. grossly normal. A small vascular some urinary and occasionally fecal Gait was slow and slightly wide- malformation was discovered oc incontinence. Pertechnitate brain based. The initial clinical impression cupying the most anterior 1.5 cm. of scan was normal; EEG showed was Alzheimer's disease. However, the left temporal lobe. poorly localized excessive slow- pneumoencephalography revealed Microscopically, large numbers of wave activity. Re-admission for not only moderate dilatation of the senile plaques and many Hirano congestive heart failure three lateral ventricles and the 3rd ventri bodies were present in the cortex of months later revealed poor orienta cle with considerable air in widened both hippocampi, where many tion for dates, poor knowledge of frontal sulci (indicative of cortical neurons exhibited neurofibrillary recent events, inability to recall atrophy) but also very little air pass tangles of Alzheimer and statements a few minutes later, ing over the vertex despite a con Simchowicz's granulovacuolar de moderate dyscalculia, and a wide- siderable amount in the Sylvian fis generation. Focal neuronal loss was based apractic gait. There was a

230-NOVEMBER 1976 Neurofibrillary Tangles in the Dementia of "Normal Pressure" Hydrocephalus Downloaded from https://www.cambridge.org/core. IP address: 170.106.40.219, on 27 Sep 2021 at 06:40:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0317167100025348 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

mild hypertonic hyperreflexia in the ities as far as the temporal lobes. mechanism for dementia was never right limbs. Lumbar puncture, which The circle of Willis showed only apparent; a dilated ventricular sys provided normal CSF, had an open mild patchy atherosclerosis. As in tem per se is unlikely, since shunting ing pressure of 120 mm. The RH1SA the four previous cases, the basal frequently returns the size of en CSF study showed some early and leptomeninges and outflow foramina larged ventricles to normal without sustained reflux into the ventricles; from the fourth ventricle were nor relieving the mental incapacitation but also good activity over the con mal. The ventricular system was not of patients with NPH. Because more vexities in the later films, with a enlarged, and there was no cortical detailed quantitative assessment of normal decline in count rate. This atrophy. The shunt tube was in place the (post-mortem) severity of mixed picture suggested an "early through the right occipital lobe. A neurofibrillary tangle formation in phase" of NPH. Air study showed small old infarct was seen in the right mesial temporal neurons has re symmetrical dilatation of the lateral putamen; and another slightly larger cently shown a strong correlation ventricles, particularly their frontal in the left basis pontis, explaining with the dementia of Alzheimer's horns; the third and fourth ventricle the right hemiparesis two years be disease (Ball, in press), morphomet and the basal cisterns were normal. fore death. ric analysis was applied to these five Air passing over the surface of the Microscopically, there were no brains as well. hemispheres reached their midpor- degenerative changes in the cortex After fixation by suspension in tions, but not the vertex. Sulci were except for a rare senile plaque in formalin for two weeks, the brain not enlarged. EEG showed more dif frontal, temporal and occipital lobes, stem and cerebellum were removed fuse theta activity, with deteriora and a very rare neurofibrillary tangle and on each side the entire mesial tion since the previous tracing three in the hippocampi. An occasional temporal lobe cut sequentially in the months before. miniscule focus of neuronal loss and coronal plane was blocked for serial The diagnosis was "adult com reactive gliosis was noted in 6- M thick paraffin sections. The municating hydrocephalus without Ammon's horns, and in the left oc area to be screened was outlined on consistently elevated intracranial cipital cortex. The basal ganglia the middle section of each tissue pressure"; and a ventri- were normal. A few small arteries in block stained with Congo red- culoperitoneal Pudenz shunt was the thalamus bilaterally, occluded by gallocyanin, using ink on the cover- therefore introduced, approximately old embolic material (with choles slip. The cortical regions examined 6 months after his dementia had terol clefts), were situated next to included Ammon's horn, the pro- begun. Within 3 days of surgery he old microscopic infarcts containing a subiculum, subiculum and pre- was mentally brighter, knew the date few residual hemosiderin-filled his subiculum of the hippocampal for and place, did serial 7's compe tiocytes. mation, and the parahippocampal tently, and had an improved gait, The obliteration of portions of the gyrus laterally to the collateral sul although noctural urinary inconti subarachnoid space frontally by the cus. Details of the method are dis nence was still present. Despite mild Fibrosis of leptomeninges was cussed in a previous paper (Ball, in post-operative cardiac and renal patchy, and no old blood pigment or press). No tangles were ever seen in problems he was discharged home 3 inflammatory cells were observed to neurons of the dentate gyrus. weeks after shunting, no longer in account for the reaction. The entire area within the inked continent of urine, able to walk well The neuropathology was therefore border was scanned in step-wise without support, and with consider non-specific, and no definite diag fashion using a Wild M501 micro ably improved mental status. Recur nosis of either Alzheimer's disease scope with a semi-automated (scan rent cardiac and renal failure forced or of frank hypertensive cerebral ning) mechanical stage, at a 200X a final readmission soon thereafter, vasculopathy could be substantiated magnification, with a square ocular though his neurological improve morphologically. (Weibel) graticule. A total of 77,784 ment persisted, and he died in pul microscopic fields were examined in monary edema some 9 months after Morphometric Methods these five cases. With polarized light the onset of his neurological syn Neither the gross or the micro from crossed Nicol prisms, and a drome. scopic changes found on routine lambda interference filter, Al Autopsy Findings: Hypertensive neuropathological examination pro zheimer tangles exhibit a bright con- cardiomegaly (450 grams) and vided a satisfactory explanation for gophilic birefringence and an shrunken kidneys (90 grams each) the striking difference in response to anomalous dichroism, appearing with severe arterial and arteriolar shunting between the fifth patient brilliant yellow-green against a faint nephrosclerosis confirmed the pres and the other four, although the sec blue-orange background (Stokes and ence of long-standing systemic ond, third, and fourth patients did Trickey, 1973). It has already been hypertension. The 1335-gram brain have considerable numbers of hip- shown that the nucleolar size, and showed a mild leptomeningeal fib pocampal plaques and tangles. The therefore probably the metabolic rosis which affected not only the significance of leptomeningeal fib rate, of a tangle-bearing nerve cell is parasagittal regions but extended rosis, somewhat more widepsread in significantly less than in adjacent medially into the interhemispheric the last two patients, was also un uninvolved neurons (Dayan and fissure and laterally over the convex clear. Furthermore, the precise Ball, 1973). The number of neurons

Melvvn J. Ball NOVEMBER 1976- 231

TABLE I Degree of Neurofibrillary Degeneration in Temporal Cortex

RAW TANGLE INDEX ADJUSTED TANGLE INDEX Both Both Fixed Brain Case Sex Age Left Right (Total) Left Right (Average) Weight (grams) 1 F 47 25 11 36 4.65 1.72 3.19 1000 2 M 52 22 63 85 2.84 6.31 4.58 1270 3 M 53 1 4 5 0.10 0.43 0.27 1430 4 F 56 3 4 7 0.44 0.59 0.52 1130 5 F 59 1 9 10 0.13 1.29 0.71 1300 6 F 60 15 9 24 2.28 1.24 1.76 1170 7 M 63 63 61 124 6.46 6.37 6.42 1330 8 M 67 2 1 3 0.33 0.11 0.22 1350 9 F 68 22 25 47 2.29 2.53 2.41 1250 10 M 69 68 98 166 7.75 12.65 10.20 1250 11 M 70 39 26 65 5.26 3.18 4.22 1190 12 M 76 81 26 107 9.79 3.27 6.53 1000 13 M 76 21 16 37 2.67 1.92 2.30 1380 14 M 76 4 4 8 0.47 0.39 0.43 1310 15 M 77 44 29 71 6.89 3.82 5.36 1530 16 F 81 35 52 87 7.58 12.78 10.18 1020 17 M 82 29 15 44 4.75 2.17 3.46 1230 18 F 83 69 79 148 10.24 8.95 9.60 1250 19 M 89 49 75 124 8.28 12.16 10.22 1040

20 M 63 20 18 38 2.87 2.68 2.77 1335 21 M 78 183 130 313 19.94 17.81 18.88 1270 22 M 70 416 438 854 77.32 77.94 77.63 1170 23 M 78 455 470 925 112.35 98.74 105.55 960 24 F 57 1021 785 1806 171.17 120.44 145.80 1275

the nuclei and nucleoli of which normal and mentally sound accord shunting, was 2.77 (Table 1); this were well visualized, and in which a ing to detailed clinical information. density of tangle formation falls al neurofibrillary tangle was present in Their ages did not differ significantly most exactly on the regression line the perikaryon was thus recorded for from those of the NPH patients for normals of similar age (Figure 1). each slide. The total for all slides (t = 0.13, p>0.1). No significant The Adjusted Tangle Indices of the from both temporal lobes gave each neuropathological changes were four patients not responding, Pa case a "Raw Tangle Index" (Table found in these brains at autopsy tients No. 21, 22, 23 and 24, were I). (Ball, in press). 18.88, 77.63, 105.55 and 145.80 re The square area of grey matter spectively (Table 1); and they plot screened was determined on photo RESULTS on the graph at levels between four graphic enlargements of a matched When the Adjusted Tangle Index and seventy times greater than con hematoxylin-eosin-Luxol fast blue- of the nineteen control cases was trols of similar age (Figure 1). stained slide of each middle section plotted graphically against the age of The density of tangle-formation in (cortex and white matter cannot be each patient, a significant linear re the five patients with NPH was then well differentiated on Congo red- gression line could be shown (Figure compared with (a) the duration of gallocyanin stains), employing a di 1). The hatched area in Figure 1 their dementia before shunting; and gitizing device linked to a Hewlett- represents ± two Standard Devia (b) the total length of their dementia Packard calculator. As the true tions from this best regression line. from onset to death (Table 2). The thickness of each section was known The correlation coefficient rum is rank correlation coefficient rs be (5.85 J-t ), the number of tangle- 0.55, statistically significant at the tween the Adjusted Index and either bearing neurons per cubic millimeter 0.05 level. In mentally normal con parameter is -|-1.0. By the null of cortex could be calculated for trols, the degree of neurofibrillary hypothesis, both of these are thus each mesial temporal lobe. The av tangle formation in mesial temporal statistically significant correlations erage of both sides provided an "Ad cortex thus correlates well with age, at the 0.05 level. In other words, the justed Tangle Index" for each case rising very gradually with increasing probability that the increasing tangle (Table 1). age. indices would rank in the same order Nineteen control brains were also The Adjusted Tangle Index of Pa as the increasing periods of dementia examined, from patients aged 47 to tient No. 20, the normal pressure (either pre-shunt or total) merely by 89 years (mean 68.6), neurologically hydrocephalic who did improve after 'chance is less than I in 20.

232- NOVEMBER 1976 Neurofibrillary Tangles in the Dementia of "Normal Pressure" Hydrocephalus Downloaded from https://www.cambridge.org/core. IP address: 170.106.40.219, on 27 Sep 2021 at 06:40:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0317167100025348 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUES

TABLE 2 ganic dementia of the Alzheimer dis ease type — both presenile and Comparison of Tangle Density and Duration of Dementia senile (McMenemey, 1971) have in Duration of dicated that the number of neurons Dementia Total Duration bearing a neurofibrillary tangle of Before Shunting of Dementia Patient Age Adjusted Tangle Index Alzheimer, expressed per cubic mil (years) (tangles/cu. mm. grey) (months) (months) limetre of mesial temporal cortex, is =*20* 63 2.77 6 9 appreciably greater (from six to forty *21 78 18.88 8 30 times larger) than in serially sec *22 70 77.63 10 76 tioned hippocampi of age-matched 23 78 105.55 48 96 normal brains (Figure 3; Ball, in press). Since four of the five patients 24 57 145.80 95 96 with NPH in the present report who ("responded to shunting) failed to respond to technically suc If the degree of tangle formation months (range, 6 to 120 months); cessful shunting had a similar strik were a reflection of the length of the those with moderate improvement, ing increase in neurofibrillary de dementing process, then the tangle 21 months (1 to 39); those with com generation in their mesial temporal index at autopsy might be expected plete improvement, only 14 months lobes, whereas the one man who to correlate quantitatively with the (3 to 36). No pathogenetic improved dramatically after shunt total length of history. While mechanism was proposed to account had a normal, minimal degree of Tomlinson's laboratory (Blessed et for this correlation. No relationship tangle-formation (Figure 1), it is al, 1968) has shown a positive corre existed between either the corpus reasonable to speculate whether the lation between the severity of his callosum angle or the cerebral man magnitude of neurofibrillary change tological changes and the degree of tle thickness on pneumoencephalog could explain the difference in re dementia measured clinically, that raphy and either the surgical out sponse to therapy. Earnest et al report did not mention the total du come or the duration of illness. Pre (1974) also remarked that their 66 ration of dementia. A negative corre vious data (Ball, in press) from the year-old man who did respond to lation was shown between the de brains of eight patients showing or- shunting showed only a "few fibril- mentia test scores and the actual or the expected period of survival after 150-1 24 diagnosis; but there was no refer o ence to any direct correlation be tween length of dementing history and degree of histological change. In the present study such a positive correlation can be well demon • CONTROLS O NPH 23 strated by the best linear regression x • 1 2 SO. o line between Adjusted Index and W 100- total duration of dementia in months (Figure 2), which has a correlation UJ _l 22 coefficient un of 0.96 (p < 0.02). The z O degree of neurofibrillary degenera Q tion in the mesial temporal cortex UJ I- thus exhibits a remarkably close re V) lationship with the length of the de ~3 50- menting process.

DISCUSSION In a recent study of 25 patients 21 with NPH (Jacobs et al, 1976), the o post-shunt condition was classified • IT as "complete improvement", j=i5-«=ss°: - 0 T :5?7 "T -r»i4— i —r- ~l— "moderate improvement" or "no 45 50 55 60 65 70 75 80 65 90 95 improvement". The authors claim AGE (years) the improved patients' clinical signs Figure I—Relationship between Adjusted Tangle Index (number of tangle-bearing (dementia, motor dysfunction, in neurons per cubic millimeter of mesial temporal cortex) and patients' age in years. continence) had been present for The best linear regression line for the nineteen control brains (closed circles) has a less time than those who did poorly significant correlation coefficient (r = 0.55, p <0.05). Patients No. 20 to 24 (open (p<0.01). Those showing no change circles) had "Normal Pressure" Hydrocephalus: No. 20 responded to shunting; had been ill for an average of 40 Nos. 21, 22, 23 and 24 showed no improvement.

Melvyn J. Ball NOVEMBER 1976-233

150 lary tangles" in the hippocampus. Furthermore, if as Figure 2 sug gests, the intensity of tangle- formation might be a function of duration of the dementia, then the somewhat shorter period before shunting perhaps would explain the Q 100- Z excellent response of our fifth case y (Patient No. 20) — who was oper ated upon within six months of the z commencement of his dementia. The irreversibility of dementia in the

CO others might reflect their much more severe histopathological state. Re < sults of shunting in Jarpe's series (1970) hinted at the same phenome non: patients with "a defined arach noiditis" rarely failed to benefit, while those with "presenile demen tia" rarely improved. It seems unlikely that multiple 36 48 60 small infarcts would adequately ex DURATION OF DEMENTIA ( months) plain the lack of response in our four Figure 2—Relationship between Adjusted Tangle Index and duration of dementia in patients. While numerous old infarc NPH (in months). A striking linear correlation exists (r = 0.96. p < 0.02). tions were noted in cases No. 21, 22 and 23, the aggregate volume of in- farcted tissue in two of these brains (No. 21 and 22) almost certainly did not total 50 cu. cm., an amount usually exceeded in dementia of the so-called arteriosclerotic (Tomlin- 200- son et al, 1970) or "multi-infarct" variety (Hachinski et al, 1974). Al though multiple infarcts might have contributed to the clinical presenta M F O D CONTROLS tion in Case No. 23, the pathological changes typical of Alzheimer's dis 150- • • DEMENTS EH :2S.D. ease were also found. No infarcts at x all were seen in the brain of Case o No. 24, who also failed to respond. z Finally, numerous old infarcts not inconsistent with a mild lacunar 100- state (Fisher, 1965; 1969) were seen in Patient No. 20, in spite of which o I- he did improve dramatically. co Hypertension would also not read 2 ily account for the lack of response 50- in the four cases. Earnest et al (1974), drawing attention to the as sociation of lacunes and NPH, have speculated that many infarcts could reduce tissue bulk or tensile spsssc 3S«2: —r SSI —I strength, permitting ventricular dila 45 50 55 60 65 ° 75 95 80 90 tation in the face of the increased AGE ( years) CSF pulse pressure of hypertension. Figure 3—Comparison of Adjusted Tangle Index in nineteen control patients (open Only one of our four unimproved symbols), and in eight demented patients (closed symbols) whose brains demon cases had significant systemic strated Alzheimer's disease. Males, circles; females, squares. (From Ball, in press). hypertension; and by contrast, the The density of tangles in Alzheimer's dementia is similar to that noted in the four improved patient also was severely unimproved cases of NPH in Figure 1. hypertensive.

234- NOVEMBER 1976 Neurofibrillary Tangles in the Dementia of "Normal Pressure" Hydrocephalus Downloaded from https://www.cambridge.org/core. IP address: 170.106.40.219, on 27 Sep 2021 at 06:40:30, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0317167100025348 LE JOURNAL CANADIEN DES SCIENCES NEUROLOGIQUKS

From our observations, it may be ALZHEIMER, A. (1907). Ubereine eigenar- clinical and radiographic findings to im asked whether in some individuals tige Erkrankung der Hirnrinde. Allgemein provement following shunt surgery. Journal Zeitschrift fiir Psychiatrie. 64, 146-148. of the American Medical Association. 235. with clinical evidence of a NPH syn BALL, M. J. (1976). Neurofibrillary tangles 510-512. drome, an as yet undefined and the pathogenesis of dementia — a quan JARPE. S. (1970). Presenile dementia and pathogenetic process — ? "Alz- titative study. Neuropathology and Applied hydrocephalus. Acta Neurologica Scan- heimerization" — can lead, if se Neurobiology (in press). dinavica 46. Supplement 43. 89. vere enough and/or long-lasting BANNISTER, C. M. (1972). A report of eight KATZMAN. R. (1976). Cerebrospinal Fluid enough, to histological degeneration patients with low pressure hydrocephalus Physiology and Normal Pressure Hydro of the cerebral cortex. Preliminary treated by CSF diversion with disappoint cephalus. In: Terry. R. D. and Ger- ing results. Acta neurochirurgica, 27, 11-15. shon. S. (Ed.). Neurobiology of Aging. data from our laboratory suggest BLESSED, G., TOMLINSON, B. E. and Raven Press. New York. pp. 139-153. that cellular alterations identical to, ROTH, M.( 1968). The association between LORENZO, A. V., BRESNAN, M. J. and though somewhat less severe than quantitative measures of dementia and of BARLOW, C. F. (1974). Cerebrospinal those quantitated in this study in the senile change in the cerebral grey matter of fluid absorption deficit in normal pressure mesial temporal lobes, are also oc- elderly subjects. British Journal of hydrocephalus. Archives of Neurology. 30. Psychiatry, 114, 797-811. curing in numerous other portions of 387-393. COBLENTZ, J. M., MATTIS, S., ZINGES- McMENEMEY. W. H. (1971). Alzheimer's the cortical mantle. The resultant SER, L. H., KASOFF, S. S., WIS- disease. In: Blackwood, W., McMenemey, degree of generalized neuronal dam NIEWSKI, H. and KATZMAN, R. (1973). W. H.. Meyer. A.. Norman, R. M. and age, as reflected by the amount of Presenile dementia: clinical aspects and Russell. D. S. (Ed.). Greenfields's neurofibrillary change in the mesial evaluation of CSF dynamics. Archives of Neuropathology. Edward Arnold Ltd.. temporal cortex, might determine Neurology, 29, 299-308. London, pp. 529-546. whether alteration of CSF dynamics DAYAN, A. D. and BALL, M. J. (1973). MESSERT. B and WANNAMAKKR. B.13. Histometric observations on the (1974). Reappraisal of the adult occult hy by shunting could still reverse the metabolism of tangle-bearing neurons. drocephalus syndrome. Neurology (Min- dementing process. It is not known Journal of Neurological Sciences. 19, neap.). 24. 224-231. why the abnormal CSF flow is as 433-436. OJEMANN. R. G.. FISHER. C. M., sociated with neuronal degenera DELAND, F. H., JAMES, A. E., LADD, D. ADAMS, R. D., SWEET. W. H. and tion, although Katzman (1976) has J. and KONIGSMARK, B. W. (1972). NEW. P. F. J. (1969). Further experience recently drawn attention to this as Normal pressure hydrocephalus: a his with the syndrome of "normal" pressure tologic study. American Journal of Clinical hydrocephalus. Journal of Neurology. sociation in his speculation that the Pathology, 58, 58-63. Neurosurgery and Psychiatry. 31. 279-294. possible bulk removal of CSF by pial EARNEST, M. P., FAHN, S., KARP. J. H. ROMANUL. F. and ABRAMOWICZ. A. vasculature might be altered in the and ROWLAND, L. P. (1974). Normal (1964). Changes in brain and pial vessels in presence of meningeal fibrosis in pressure hydrocephalus and hypertensive arterial border zones. Archives of Neurol Alzheimer's disease. Whatever the cerebrovascular disease. Archives of ogy. II. 40-65. explanation, if this small series Neurology, 31, 262-266. SOHN, R. S.. SIEGEL. B. A.. GADO. M. proves representative, the best hope FISHER, C. M. (1965). Lacunes — small and TORACK. R. VI. (1973). Alzheimer's deep cerebral infarcts. Neurology (Min- disease with abnormal cerebrospinal fluid of success for treating the dementia neap.), 15, 774-784. flow. Neurology (Minneap.). 23. 1058-1065. of NPH would lie in earliest possible FISHER, C. M. (1969). The arterial lesions STEIN, S. C. and LANGFITT. T. W. (1974). diagnosis. underlying lacunes. Acta Neuropathologica Normal pressure hydrocephalus. Predicting (Berlin), 12, 1-15. the results of cerebrospinal fluid shunting. GESCHWIND, N. (1968). The mechanism of Journal of Neurosurgery. 41. 463-470. normal pressure hydrocephalus. Journal of STOKES, M. I. and TRICKEY. R. J. (1973). ACKNOWLEDGEMENTS Neurological Sciences, 7, 481-493. Screening for neurofibrillary tangles and This study was supported by the Canadian argyrophilic plaques with Congo red stain Geriatrics Research Society; and by a grant HACHINSKI, V. C, LASSEN, N. A. and MARSHALL, J. (1974). Multi-infarct De ing and polarized light. Journal of Clinical for equipment from the University Hospital Pathology, 26, 241. Research Trust Fund. The author thanks Drs. mentia. A cause of mental deterioration in the elderly. Lancet, 2, 207-210. C. G. Drake, H. J. M. Barnett, E. Davies, J. SUGAR, O. (1976). Editorial. Journal of the American Medical Association. 235. 534. Gilbert, B. Adilman and L. Beattie. The ex HAKIM, S. and ADAMS, R. D. (1965). The pert technical assistance of Mr. Patrick Lo is special clinical problem of symptomatic SYPERT. G. W.. LEFFMAN. H. and gratefully acknowledged. hydrocephalus with normal CSF pressure: OJEMANN. G. A. (1973). Occult normal Observations on CSF hydrodynamics. pressure hydrocephalus manifested by Journal of Neurological Sciences, 2. parkinsonism-dementia complex. Neurol 307-327. ogy (Minneap.), 23. 234-238. REFERENCES HEINZ, E. R., DAVIS, D. O. and KARP, H. TOMLINSON. B. E., BLESSED, G. and ADAMS, R. D., FISHER, C. M., HAKIM, R. (1970). Abnormal isotope cisternography ROTH, M. (1970). Observations on the S., OJEMANN, R. G. and SWEET, W. H. in symptomatic hydrocephalus. Radiology, brains of demented old people. Journal of (1965). Symptomatic occult hydrocephalus 95, 109-120. Neurological Sciences, II. 205-242. with "normal" cerebrospinal fluid pres HIRANO, A.. DEMBITZER, H. M. and TOMLINSON, B. E. and KITCHENER, D. sure. New England Journal of Medicine, KURLAND, L. T. (1968). The fine struc (1972). Granulovacuolar degeneration of 273, 117-126. ture of some intraganglionic alterations. hippocampal pyramidal cells. Journal of ADAPON, B. D., BRAUNSTEIN, P., LIN, Journal of Neuropathology and Experimen Pathology. 106, 165-185. J. P. and HOCHWALD, G. M. (1974). tal Neurology, 27, 167-182. WOLINSKY. J. S., BARNES, B. D. and Radiologic investigations of normal pres JACOBS, L., CONTI, D., KINKEL, W. R. MARGOLIS, M. T. (1973). Diagnostic sure hydrocephalus. Radiology Clinics of and MANNING, E. J. (1976). "Normal- tests in normal pressure hydrocephalus. North America, 12, 353-369. pressure" hydrocephalus: relationship of Neurology (Minneap.). 23. 706-713.

Melvyn J. Ball NOVEMBER 1976- 235