10 The differential diagnosis of normal pressure hydrocephalus
Adult Hydrocephalus, ed. Daniele Rigamonti. Published by Cambridge University Press. © Cambridge University Press 2014.
Introduction
Hakim and colleagues, in their paper describing the syndrome of normal pressure hydrocephalus (NPH) in three patients, recapitulated the clinical presentation as follows: “a mild impairment of memory, slowness and paucity of thought and action, unsteadiness of gait and unwitting urination” (Hakim’s triad) [1]. They also highlighted two important features of the syndrome: “The symptomatology was unobtrusive, having no assignable date of onset, and evolved over a period of weeks and few months … The most striking feature of the condition was the rapid restitution of the patient’s health as the result of lowering of the cerebrospinal-fluid pressure through surgical shunting.”
They further stressed that because “the clinical picture corresponds essentially to that of a mild form of dementia,” it was obligatory for the treating physician to distinguish such a mild dementia “occurring without obvious cause in a middle-aged person … from the large number of dementing illnesses of the presenium and the senium.” Interestingly, pneumo-encephalography demonstrated, in all three cases described in the original paper by Hakim, an obstruction to the CSF, due to a blockage of the subarachnoid space (SAS) at the cervical-medullary junction in the first two cases and to a colloid cyst in the third [1].
Nowadays, thanks to much improved imaging modalities (CT and MRI) we can quickly recognize ventriculomegaly. However, even in the presence of clinical symptoms, in a significant percentage of cases, in the elderly, no obvious blockage can be detected. NPH in the absence of a definitive cause has been called “idiopathic” NPH (iNPH). The late Professor Marmarou led an effort to better identify this condition from a clinical standpoint. The consensus guidelines published in the USA defined iNPH as probable, possible, and unlikely depending on presence or absence of several elements in the history, neurological examination, imaging, and response to specific diagnostic procedures [2,3]. The guidelines underlined the fact that because iNPH frequently is accompanied by multiple and significant comorbidities, its diagnosis remains challenging. The lack of awareness about NPH on the part of the medical profession further complicates matters [4].
The recently updated Japanese guidelines draw attention to a specific MRI pattern of disproportionately enlarged subarachnoid space hydrocephalus (DESH), believed to be pathognomonic of iNPH [5,6].
This chapter will discuss why establishing the diagnosis of NPH remains a challenge fifty years after its classic description.
The neurological examination of the NPH patient
Hakim and colleagues described the following cognitive elements as characteristic of the syndrome: “lack of spontaneity and initiative, faulty concentration, distractibility, lack of interest, apathy and inertia … Inner psychic life seemed to be impoverished, and the patient was bereft of thoughts.” The findings on the rest of the neurological anamnesis and examination were insignificant: “headache was absent or negligible. Papilledema did not occur. Mild horizontal nystagmus was present in all 3 cases … Strength of the limbs was maintained; there was no frank cerebellar ataxia. The plantar reflexes were usually flexor or equivocal until a late stage … The unsteadiness of gait, which was difficult to characterize, was not clearly ascribable to cerebellar deficit. Aphasia, dyspraxia and agnosia were not obvious. The circumference of the cranium was normal” [1].
Younger patients more often have hydrocephalus because of a visible obstruction and in most cases clinical symptoms and signs of elevated intracranial pressure. However, some of these may present later in life with very large ventricles and an “NPH like” clinical picture – an entity that has been described with the acronym LOVA (longstanding overt ventriculomegaly in adults) [7–9].
Elderly patients with presumed iNPH span a continuum from very healthy and functional to very sick and disabled. We have encountered people with impressive ventriculomegaly on MRI and both high IQ and great physical fitness with complaints of subjective difficulties. Given the fact that our tests are based on a control population whose “normal level of function” is at a lower level than the baseline function of these particular individuals (see Cases 1 and 2 at the end of the chapter), we could not detect any deficit.
At the other end of the continuum, we have also many patients with a very complicated medical history and a neurological exam as well (see Cases 3 and 4). This reflects the presence of many comorbidities including hypertension in almost half of the patients, diabetes mellitus in about a fifth, cerebrovascular disease in various degrees in more than half of the patients, and dementia ranging from 15% to 75%, while “pure” cases of iNPH occur in barely a fifth of patients [10–12]. As a consequence, besides gait, cognitive, and urinary symptoms, the careful examiner will find paratonia of lower extremities, increased need of sleep, abnormal Romberg test, and retropulsion, features that are found in more than half of patients, while focal neurological signs, polyneuropathy, and cerebellar signs occur in at least one-fifth of the patients.
The diagnostic work-up of the NPH patient
The original diagnosis of NPH relied upon the presence of mild dementia, gait, and urinary difficulties (Hakim’s triad) seen in association with ventriculomegaly on pneumo-encephalogram.
Nowadays the recognition of NPH is complicated by two factors. While the availability of MRI has helped the early recognition of ventriculomegaly, this often precedes the appearance of the full triad. At the other end of the spectrum, in the older population, comorbidities are much more pronounced and might mask the NPH symptomatology especially if it is slow in development [13–15].
Several authors believe that the presence of the full triad most likely represents an advanced stage of the disease and advocate reaching an earlier diagnosis in every case [16]. We were able to document less likelihood of recovery after surgery, as the duration of symptoms increased [10]. Recent data from Sweden confirmed that a delay in intervention after confirmation of the diagnosis, in fact, is accompanied by a measurable deterioration of function and a lesser degree of improvement than in patients who underwent surgery more rapidly [17].
Furthermore, a recent publication claimed a high correlation between presence and severity of all major symptoms in iNPH [18].
Gait dysfunction
The presence of gait difficulties, as the first and prominent symptom, is essential to make the diagnosis of NPH [10].
Gait features that are classically described in iNPH are the wide base, the magnetic or “glued feet,” shuffling and slow pace. Gait impairment in NPH, however, ranges from difficulties running, fear of using escalators, difficulties walking on a slope or uneven surfaces or stairs, difficulties getting up from a chair, needing to use an assistive device such as a cane or a walker, to inability to stand and walk. (Please refer to Chapter 6 on pathophysiology of gait for a more detailed discussion of the gait characteristics.)
The 10-meter gait test, the timed up and go (TUG) test, the Tinetti gait and balance test, and the 6 ft walk are widely used, but not always satisfactory and adequate, tools to objectively measure the deficits. Objectively there could be no obvious sign in the patient who cannot run. Conversely, we could assess the patient who takes short steps (less than 1 foot in length), walks at a reduced pace and on a wider base, who has difficulty turning on his/her long axis (more than 4–6 steps to make a 360° turn).
The challenge resides in the fact that gait and balance problems in the elderly can be caused by a plethora of other processes, and it is unrealistic to expect the general practitioner or a family member to recognize the gait of NPH that is quite characteristic to the eyes of the expert. So the lack of awareness of iNPH and the presence of comorbidities frequently confound the general practitioner or the neurologist, leading unfortunately to a harmful delay in diagnosis.
There are several conditions that may deceive the practitioner and prevent the recognition of the causative role of iNPH in the development of gait problems. Parkinsonism (Parkinson’s disease and secondary parkinsonism) in its early phase may present a diagnostic challenge because several features are similar: small, shuffling steps, unsteadiness on turning, and a looming posture. However, the narrow-based gait, the loss of the arm swing, and the presence of tremor, hypokinesia, and rigidity are not features of NPH and should not mislead the thoughtful examiner. Lumbar stenosis usually causes low back pain, but weakness and radicular sensory disturbances precipitated by activity might affect the gait and create a source of confusion. Cervical spondylotic myelopathy causing upper motor neuron weakness in the lower extremities, with hyperreflexia and spasticity with or without sensory changes, will affect a person’s gait and balance. MRI and CT show the cord compression due to osteophyte formation. Peripheral neuropathy of a metabolic, infectious or other nature, and arthritis of the hip and knees can be aggravating gait problems, too [13–15].
Cognitive deficit
iNPH is responsible only for a small proportion of cognitive difficulties experienced by the elderly; however, because of its reversible nature it must be considered in the differential diagnosis. Because iNPH patients are advanced in age (usually in their seventies or even older), the likelihood of finding Alzheimer’s disease (AD) or other forms of dementia in these patients is high.
The mini-mental state examination (MMSE) is useful in broadly assessing the level of dementia (or its absence); a low score in this test is not typical of NPH and should immediately raise the suspicious of a cortical dementia. More sensitive cognitive evaluation of iNPH patients requires specific tests for the assessment of subcortical frontal lobe deficits (executive dysfunction, impaired attention, visuospatial dysfunction, and psychomotor slowing) such as the Rey Auditory Verbal Learning Test, Stroop test, Grooved Pegboard, Trail Making A and B Test, and digit span test.
Aphasia, apraxia, or agnosia should immediately raise suspicion of conditions with cortical pathology (Alzheimer’s disease, frontotemporal dementia, and vascular dementia). Depression might mimic dementia and needs to be ruled out. The presence of resting tremor and lead-pipe rigidity with visual hallucinations should raise the suspicion of dementia with Lewy bodies (DLB) [13–15]. (Please refer to Chapter 7 on pathophysiology of the cognitive deficits for a more detailed discussion.)
AD is the most common form of dementia; it usually tends to follow a longer course, with prominent cortical dysfunction like apraxia and language disturbance and prominent memory disorders. Enlarged cortical sulci are seen on MRI. Early AD may resemble depression or pure memory disorders (Korsakoff amnestic syndrome). In AD positron emission tomography (PET) scanning may reveal hypometabolism and hypoperfusion in the temporal and parietal lobes. Cerebrospinal fluid (CSF) levels of Aβ42, tau, and phospho-tau are valuable biomarkers of AD. (Please refer to Chapter 15 on biomarkers for a more detailed discussion.)
Vascular dementia is very common in the elderly; it is often characterized by a long history of hypertension, focal neurological symptoms or signs, stepwise progression and fairly sudden onset of dementia. Pseudobulbar palsy (dysarthria, dysphagia, and emotional instability), focal motor and/or sensory deficit, ataxia, hyperreflexia, and Babinski sign are usually present. MRI shows either multiple cortical infarcts or several smaller infarcts of the deep white matter, basal ganglia, and thalamus (lacunar state). The major challenge to the clinician is subcortical vascular dementia, which can present with symptoms and signs that are more or less identical with those seen in iNPH [13,14].
The presence of a cerebrovascular history and stigmata on MRI should not prevent the recognition of the coexistent ventriculomegaly and reduce the responsibility to investigate if some symptoms might perchance be due to NPH. Recently Tisell et al. showed in a randomized controlled study that patients presented with pronounced white matter changes on MRI and symptoms and signs compatible with iNPH improved significantly after shunting [19].
The European multicenter study on iNPH confirmed that after treatment patients with vascular comorbidity improved to a similar degree as those without, suggesting that patients with vascular comorbidity should not be excluded from being considered for shunt surgery. However, by the same token, it should not be forgotten that severe vascular diseases as determined by a high Kiefer comorbidity score has a very negative prognosis [20–22].
Frontotemporal dementia (FTD) is a heterogeneous group of disorders that produce frontal and temporal lobe degeneration and affect behavior and language. FTD, the third most common cause of dementia after AD and vascular dementia, affects younger people (before age 60) and presents with behavioral problems or aphasia. MRI shows frontal and temporal lobe atrophy, and PET might show corresponding hypometabolism. Usually the most affected side correlates with the most prominent clinical feature: right-sided in the behavioral variant and left-sided in the language variant. We have occasionally treated patients who were diagnosed with FTD, but also had hydrocephalus: their dementia did not improve, but their gait benefited dramatically from the surgery (Case 3).
Corticobasal degeneration is a tauopathy related to FTD. It involves both cortex and basal ganglia, usually causing unilateral limb clumsiness/sensory loss/apraxia associated with extrapyramidal rigidity, bradykinesia, and postural tremor. In this situation as well, NPH might coexist and an attempt at dealing with it might be considered (Case 4). Progressive supranuclear palsy (PSP) usually is very distinctive in its presentation, with dementia, supranuclear ophthalmoplegia (especially down gaze), pseudobulbar palsy, and dystonia with or without extrapyramidal rigidity being the main features. In Huntington’s disease, psychiatric symptoms and chorea usually precede dementia, which is characterized by loss of memory and executive function.
Lewy body disease presents with fluctuating cognitive abilities and visual hallucinations associated with rigidity and bradykinesia typical of parkinsonism. Creutzfeldt–Jakob disease (CJD) is a prion disease causing rapidly progressive dementia that begins either as a mild global cognitive impairment or as a focal cortical disorder (aphasia, agnosia, or apraxia). Psychiatric symptoms and changes in personality can be significant. Myoclonus, extrapyramidal signs, and cerebellar signs are frequently present.
Chronic subdural hematomas (cSDH) usually present with confusion and dementia, and hemiparesis, several weeks or months after a mild trauma in patients with cerebral atrophy, history of alcoholism, or on hemodialysis or anticoagulation. Less common are visual field defects and aphasia. CT and MRI show the frequently bilateral, extra-axial collection of fluid of homogeneous or mixed density, associated with obliteration of the underlying sulci and mass effect on the ventricular system.
A brain tumor, such as an infiltrating glioma, invading the frontal, temporal lobes and the corpus callosum might produce dementia. Apathy, impaired concentration, and slowness are more prominent. In this case too, however, MRI is diagnostic. Radiotherapy and chemotherapy may be aggravating the symptoms.
Leptomeningeal spread of cancer may present with confusion and multiple cranial neuropathies. Cytological study of the CSF will provide the diagnosis if the images do not show meningeal enhancement in the areas where the deficits are caused.
Chronic traumatic encephalopathy following repeated head injuries leads to progressive impairment of concentration and memory, personality changes, rigidity, tremor, and bradykinesia. MRI shows cortical atrophy.
Infectious processes such as HIV, syphilis, and progressive multifocal leukoencephalopathy (PML) due to JC papova virus can produce minor neurocognitive disorders and progress to dementia. The history usually is helpful and appropriate laboratory tests will confirm the diagnosis.
Metabolic diseases such as alcoholism, hypothyroidism, and vitamin B12 deficiency should be considered in the differential diagnosis and the appropriate tests ordered if a suspicion occurs.
Organ failures may produce dementia and confound the NPH diagnosis: dialysis dementia, Wilson’s disease, and non-wilsonian hepatocerebral degeneration [13,14].
Urinary dysfunction
The bladder symptoms of increased frequency, urgency, and incontinence should be distinguished from other urological conditions. Urinary incontinence could be related to recurrent urinary tract infections and medications, benign prostatic hypertrophy (BPH) in men, and pelvic floor laxity and stress incontinence in women.
Bladder symptoms in NPH result from detrusor hyperactivity. Although several tests can objectively document the dysfunction in these patients, they are rarely used in a consistent fashion. The utilization of a validated questionnaire is a valid surrogate for the assessment and grading of the urinary dysfunction. (Please refer to Chapter 8 on pathophysiology of urologic dysfunction for a more detailed discussion.)
Imaging evaluation
Since Hakim and his colleagues diagnosed NPH on the basis of pneumo-encephalography, imaging has taken gigantic steps. Isotope cisternogram was for a while the gold standard for the diagnosis of NPH and demonstrated NPH secondary to a subarachnoid space blockage when it showed reflux into the ventricular system. But it was the advent of CT first, followed by MRI, that transformed the field. The causes of hydrocephalus, when present, are clearly revealed. Blockage at the foramen of Monro, at the aqueduct, at the outlet of the fourth ventricle, in the basal cisterns, and at the level of the convexity can now be promptly recognized (please refer to Chapter 11 on basic imaging for a more detailed discussion) and the pathophysiology of iNPH is now being actively investigated (please refer to Chapter 12 on advanced imaging for a more detailed discussion).
It is routine these days to use both CT and MRI in the evaluation of NPH patients, although we foresee that this practice might change in the future due to cost or safety concerns.
CT is frequently the initial test and is certainly useful in the diagnosis of hydrocephalus, secondary for example to a subarachnoid hemorrhage (SAH). A CT scan could immediately give information on the degree of ventriculomegaly, the presence of periventricular edema, and the tightness of the cortical SAS.
MRI provides a superior anatomical and physiological definition of the ventricular system and the SAS. There is evidence that MRI could help discriminate between NPH and AD by studying the degree of dilatation of the perihippocampal fissures (PHFs), which are much more dilated inAD [23].
Several sequences are used, some of which should be considered “core” images and others “advanced” images. Core sequences provide information that is essential to the management of the patient; advanced imaging sequences provide images that are critical in furthering the understanding of the mechanism of the disease. (Please refer to Chapters 11 and 12 on basic and advanced imaging for more details.)
Invasive evaluation of CSF physiology
This diagnostic test provides information about CSF dynamics and predicts outcome. It consists in either removal of CSF accompanied by pre- and post-functional evaluation, or an infusion (bolus or continuous) test.
Spinal tap test (STT)
STT consists of the removal of more than 30 ml of CSF from the patient. Opening pressure should be measured at the beginning. CSF should be sent for routine and special examinations. Routine examinations include cytology, microbiology, and chemistry. (For special testing please refer to Chapter 15.)
An unequivocal improvement in the functioning of the patient after the STT has a very strong positive predictive value (PPV) for a beneficial effect of shunting. Different authors have different criteria that rely on either percentage improvement or improvement beyond 1 standard deviation (SD) in either gait or cognitive testing.
Continuous spinal or ventricular drainage
The test requires the insertion of an indwelling spinal or ventricular catheter for a few days and an hourly drainage of 5–10 ml of CSF. The CSF is analyzed in the same fashion as with the STT. This test has the advantage of demonstrating an improvement in some patients who do not respond to the STT.
Infusion test
There are several infusion methods available: bolus, the continuous and the pressure controlled.
In spite of some data suggesting the contrary, several series reporting the results of intrathecal infusion technique with constant flow suggest that outflow resistance (Rout) is pathologically elevated in patients with NPH, while it is normal in patients with cerebral atrophy; the compliance (Cp) is normal (see Chapters 13 and 14 for more details).
Intracranial pressure (ICP) measurements
We have not found ICP monitoring helpful in our experience at Johns Hopkins. The problem might reside in the fact that measurements have been performed in patients with symptoms suggestive of a CSF circulation disorder and patients with “low” pressures have been presumed to be normal. Thus “normal values” have been extrapolated from a symptomatic population. If truly normal values were found to be lower than so far assumed, then NPH could be a disorder of moderately, chronically elevated ICP. Data from the department at the University of Copenhagen are suggestive of normal ICP around 0 mmHg in supine position and –10 mmHg or below in standing position.
According to the recently concluded European iNPH multicenter study, CSF tap test and Rout can be used for selecting patients for shunt surgery, given their high positive predictive value (PPV >90%), but they cannot be used for excluding patients for surgery, given the very low negative predictive value (NPV <20%) [24].
iNPH grading scales
Over the years several authors have attempted to grade the clinical severity of NPH patients and assess their outcome after treatment. These tools have helped us enormously to improve preoperative assessment of patients and assessment of outcome. (Please refer to Chapter 9 for more details on grading scales.)
Outcome
Considering that surgery produces long-term improvement in about 70% of cases and that improvement might affect all three elements of the triad, we lament the fact that in the medical community at large there is not enough interest in trying to recognize this reversible form of dementia, offering patients a chance of even partial recovery [10,25]. A reason for this lack of enthusiasm could reside in a negative attitude about aggressively managing what some neurologists believe to be a nonexistent condition. More sadly, the lack of interest could be the result of sheer ignorance as documented by Conn [4]. For whatever reason the patient goes undiagnosed, the delay leads to further progression of the deficits and a diminished recovery after a delayed surgery. Several authors have commented on the negative correlation between diagnostic delay and postsurgical recovery [10, 16]. The presence of coexisting AD (31.3%) might provide yet another reason for poor long-term outcomes after shunting [26]. (Please refer to Chapter 20 on outcome for a more detailed discussion of this topic.)
Final thoughts
A study from Japan demonstrated a prevalence of iNPH of 0.51% in people aged 61 years or older, while asymptomatic ventriculomegaly with iNPH features on MRI (AVIM) occurred in 1.01% of this elderly population [27]. A population study from Norway showed a prevalence of probable iNPH of 21.9/100 000 and an incidence of 5.5/100 000/year; the age-related prevalence climbed to 181.7/100 000 above age 70 [28]. The total incidence of surgery for hydrocephalus in Norway over a 5-year period was 1.09/100000/year (ranging between 0.84 and 1.47/100000/year) [29]. In Sweden the annual incidence of surgery for every type of hydrocephalus was 3.4/100000; surgery for iNPH was the most common (47%) [30). (Please refer to Chapter 5 on epidemiology for a more detailed discussion of this topic.)
These data suggest that only 18–31% of patients with iNPH were offered surgery in the Scandinavian countries, where the condition is better known than in the USA. This clearly suggests a failure of recognition of the disease.
The medical profession clearly does not think enough about NPH, and the reasons are unclear. We think that we should stop overwhelming the medical community with a complicated classification and an even more complicated explanation of the pathogenesis. We believe that focusing on “ventriculomegaly” would be much more beneficial in raising the awareness of the condition.
The patient with ventriculomegaly should be referred to a neurological or neurosurgical center, where an appropriate evaluation would place him/her into the appropriate category based on the degree of symptomatology:
(1) Asymptomatic patients or patients with minimal symptoms of NPH.
(2) Patients with clear symptoms of NPH, no obvious comorbidities and clear response to diagnostic tests.
(3) Patients with advanced symptoms of NPH and significant associated pathology.
Patients with minimal or no symptoms should be carefully monitored, evaluated with a detailed neurocognitive and motor examination their status, and encouraged to undergo a spinal tap at some point to determine if their perceived sense of well-being is masking an insidiously developing deficit.
Patients with clear symptoms and clear response to CSF challenge that can be unequivocally measured are easier to manage. There is, however, fairly powerful evidence from the European hydrocephalus study that the tap test and the infusion tests have a poor NPV [24]. It is then prudent in patients in whom the symptoms and the images are characteristic of NPH to discuss the possibility of surgery even in the absence of a positive response to a CSF challenge test.
Patients with advanced symptoms of NPH and associated pathology represent a major challenge. In these patients the task is to determine if by eliminating the contribution of NPH to the disability a significant improvement in the quality of life could be obtained. Given some recent evidence obtained by the group of Wikkelsø and our own personal experience, it seems wise to at least consider surgery in this group as well.
In all groups undergoing surgery we recommend paying close attention to the long-term function of the shunt and intervening at the first sign of malfunction.
Case 1 A 53-year-old executive underwent a work-up for thyroid problems that included an MRI. He walks for 40 minutes every day with his wife. His neurological exam is completely within normal limits. His timed up and go (TUG) test time is 6 seconds, his Tinetti gait and balance score is 28/28. He can stand on one leg for 1 minute and 46 seconds with no problem. MRI showed ventriculomegaly, crowding on the convexity, bowing of the corpus callosum, dilatation of the third ventricle with bulging outward of the wall and aqueductal pathology. (See Figure 10.1.)
Related posts:
25 – Chiari malformation and hydrocephalus in adults
26 – Pseudotumor cerebri syndrome
21 – Normal pressure hydrocephalus syndrome secondary to hemorrhage, infection, and malignancy
22 – Low-pressure syndromes and cerebrospinal fluid leaks
24 – Management of hydrocephalus with associated cerebrospinal fluid pathologies
19 – Endoscopic third ventriculostomy
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
