Article under review: Josephs KA, Ahlskog JE, Parisi JE, et al. Rapidly progressive neurodegenerative dementias. Arch Neurol 2009;66(2):201-7.
Background: Most of the common neurodegenerative disorders that present with dementia are insidious in onset and progress slowly over a period of years, with a total disease duration of more than 5 years.
Purpose: To characterize patients with neurodegenerative dementias that progress rapidly to death.
Methods: Using the Mayo Clinic records linkage system, the authors identified 22 patients evaluated between 2000 and 2007 with a disease duration from onset to death of less than 4 years and an autopsy-based histological diagnosis of a neurodegenerative disease.
Results: The age of onset ranged from 33 to 85 years. The interval from disease onset to initial neurologic evaluation averaged about 8 months.
The most commonly identified etiology was Creutzfeldt-Jakob disease, representing more than one third (n=8, 36%) of all cases, and all of those with Creutzfeldt-Jakob disease died within 1 year of onset. Myoclonus was a specific though somewhat insensitive marker of Creutzfeldt-Jakob disease, being present (recorded) in 6 of the 8 patients with Creutzfeldt-Jakob disease and in none of the 14 patients without Creutzfeldt-Jakob disease. EEG was inconsistently helpful in patients ultimately shown to have Creutzfeldt-Jakob disease: atypical triphasic waves, sharp waves, or epileptiform discharges were noted in 4 cases, excess slow waves in 3 of the remaining cases, and no abnormality in 1 case. However, all 3 patients with Lewy body dementia had atypical triphasic waves, which were intermittent in 2 and semiperiodic in the other. All of the patients with Creutzfeldt-Jakob disease were correctly diagnosed prior to death, but 2 patients with other conditions were also diagnosed as Creutzfeldt-Jakob disease clinically (1 with Lewy body dementia and 1 with progressive supranuclear palsy).
MRI was most useful in the Creutzfeldt-Jakob disease cases. Fluid-attenuated inversion recovery and diffusion-weighted imaging sequences were abnormal in all of the 6 patients in whom this was performed, with abnormalities in the cortical gyri, striatum, and thalamus.
Spinal fluid studies were generally not very helpful in discriminating among this group of patients with rapidly progressive dementias. Cell counts and glucose levels were uniformly normal. Spinal protein levels were elevated in 4 of 8 patients with Creutzfeldt-Jakob disease (range and mean or median was not given), and 3 of 7 patients with other rapidly progressive dementias who had spinal taps. Neuron-specific enolase levels in CSF were apparently elevated in 6 of 8 patients with Creutzfeldt-Jakob disease, and in 1 of 6 patients with other rapidly progressive dementias in whom this study was performed. If mildly abnormal values are included as "abnormal," all of the patients with Creutzfeldt-Jakob disease and 4 of the 7 other patients had abnormal results. Levels of 14-3-3 protein were abnormal in 3 of 6 patients with Creutzfeldt-Jakob disease and none of the 5 other patients in whom this was tested. Thus, with this small sample of patients with rapidly progressive dementia, neuron-specific enolase levels in CSF were fairly sensitive but nonspecific for Creutzfeldt-Jakob disease, and 14-3-3 protein in CSF was insensitive but seemingly specific for Creutzfeldt-Jakob disease.
Other etiologies were identified in 14 cases (64%), but only 4 of these were correctly diagnosed during life. Other identified etiologies included frontotemporal dementia with motor neuron degeneration in approximately one fourth (n=5, 23%), tauopathies (progressive supranuclear palsy or corticobasal degeneration; n=4, 18%), dementia with Lewy bodies (n=2, 9%), and Alzheimer disease (n=2, 9%). In almost all of these cases, mixed pathological findings were present, indicating likely contributions from more than 1 etiology, and the authors speculated that the mixed etiologies may have in fact contributed to the rapid course in these cases. The cases with etiologies other than Creutzfeldt-Jakob disease all involved patients who lived longer than 1 year.
MRI was less useful in the non-Creutzfeldt-Jakob disease cases. Varying degrees of cerebrocortical atrophy were apparent in all etiologic groups but were generally nonspecific, although severe frontal atrophy was present in 2 (of 5) patients with frontotemporal dementia with motor neuron features, and asymmetric left-greater-than-right frontoparietal atrophy was present in 1 patient (of 2) with corticobasal degeneration.
All of the frontotemporal dementia cases with motor neuron degeneration showed TDP-43 immunoreactive lesions in hypoglossal nuclei or spinal cord anterior horn cells.
Other than myoclonus that was specific for Creutzfeldt-Jakob disease, clinical findings often felt helpful in distinguishing between different neurodegenerative dementing disorders proved insensitive or nonspecific (at least in isolation). Parkinsonism was common in all of the etiologic groups and was present overall in 15 of the 22 cases (68%). Visual hallucinations at dementia onset were reported in patients with Lewy body dementia, Creutzfeldt-Jakob disease, and Alzheimer disease. Rapid, severe fluctuations in cognitive function were reported in patients with Lewy body dementia and Alzheimer disease. Motor neuron disease was reported in 1 patient with Creutzfeldt-Jakob disease and in 2 of the 5 patients with frontotemporal dementia with motor neuron degeneration.
Mean age at onset varied by etiology, but there was considerable overlap in age of onset across different etiologic groups. As the etiologic groups were small for most categories, it almost gets to be an anecdotal comparison except for the largest groups. Patients with Creutzfeldt-Jakob disease had disease onset from 41 to 80 years of age, with a median age of 64.5 years. Patients with frontotemporal dementia with motor neuron degeneration had disease onset from 33 to 67 years of age, with a median age of 50 years. The remaining patients had disease onset from 54 to 85 years of age, with a median age of 74 years .
Conclusions and commentary: Consistent with clinical experience, Creutzfeldt-Jakob disease is the most common cause of rapidly progressive dementia overall and is almost exclusively the cause in cases that progress from onset to death in less than 1 year. However, other neurodegenerative conditions can also on occasion cause dementias that progress nearly as rapidly, including frontotemporal dementia with motor neuron degeneration, progressive supranuclear palsy, corticobasal degeneration, dementia with Lewy bodies, and Alzheimer disease. In the United Kingdom and France, variant Creutzfeldt-Jakob disease would need to be included among the conditions causing rapidly progressive dementia with a disease course of longer than 1 year.
Periodic complexes on EEG within the first year after onset in a patient with a rapidly progressive dementia are strongly suggestive of Creutzfeldt-Jakob disease, whereas atypical triphasic waves (semiperiodic or not) after the first year may suggest Lewy body dementia.
Other than the fluid-attenuation inversion recovery and diffusion-weighted imaging abnormalities in patients with Creutzfeldt-Jakob disease, MRI was not particularly helpful. Although severe symmetric frontal or frontotemporal atrophy in patients with rapidly progressive dementia may suggest frontotemporal dementia with motor neuron degeneration, this finding was not sensitive for this condition. Similarly, although asymmetric left-hemisphere-predominant frontoparietal atrophy in patients with rapidly progressive dementia may suggest corticobasal degeneration, this finding was also not sensitive (although the numbers here are very small and the findings, thus, anecdotal).
Because "rapidly progressive" was defined in this study on the basis of disease duration, rather than rate of change in cognitive function or functional performance, it is not straightforward to apply these results to a living patient who seems to be worsening rapidly. Furthermore, the distribution of etiologies for patients with "rapidly progressive dementia" will depend on the maximum disease duration when that is the basis for inclusion - cases with duration less than 1 year will be almost exclusively Creutzfeldt-Jakob disease (except for occasional cases with rapidly progressive dementias in which the patients die of intercurrent illness and rare cases of other etiologies), while cases with duration between 1 and 2 years will include mostly cases of frontotemporal dementia with motor neuron degeneration and cases of progressive supranuclear palsy.
Of note, as this was a retrospective study utilizing clinical information and test results obtained as part of clinical care, albeit at a sophisticated tertiary medical center, there was no systematic approach to ordering diagnostic tests. So, for example, only 5 had single-photon emission computed tomography, likely not all had EEG, etc. Indeed, for most of the tests discussed, the number who had the test for each etiologic group is not given. Therefore, no great weight should be given to the specific proportions of positive or negative test results for any specific etiology with this study design.