Pathologic heterogeneity in clinically diagnosed corticobasal degeneration

Methods.

Results.

The demographic information, clinical features, and pathologic diagnoses of the 13 patients are summarized in table 1. The mean age at onset was 64 years (range 49 to 72), and mean duration of illness was 6 years (range 3 to 12). Twelve patients had the core findings of progressive asymmetric limb rigidity and apraxia; the remaining patient had asymmetric rigidity without apraxia. Other cortical and extrapyramidal signs were present in every patient. None of the patients who were administered levodopa experienced any significant improvement. Dysarthria, oculomotor dysfunction, frontal release signs, and pyramidal tract findings were frequent accompanying features. Thus, every patient had the constellation of clinical findings most suggestive of CBD.

Despite all cases having the same clinical diagnosis of CBD, other neuropathologic conditions were identified at postmortem examination in 6 of the 13 cases. The characteristic pathologic findings of CBD were found in 7, whereas the findings were those of AD in 2 and PSP, Pick’s disease, nonspecific degenerative changes, and Creutzfeldt-Jakob disease in the remaining cases.

Tissue was available for immunohistochemistry in 12 cases. All CBD cases had NFP+ ballooned neurons, tau+ astrocytic plaques, and tau+ coiled bodies in oligodendroglia. The PSP and Pick’s disease cases had rare NFP+ achromatic neurons and sparse tau+ glial pathology, which were not of sufficient severity or topography to warrant the diagnosis of CBD.15 One CBD case (Case 7) had marked tau+ pathology in neurons and glia and scattered NFP+ swollen neurons in the frontoparietal region, as well as sparse globose neurofibrillary tangles and Pick body-like inclusions in the dentate fascia of the hippocampus, and this pathology was therefore considered most consistent with CBD. The H-E and anti-ubiquitin sections of several brainstem and cortical regions did not reveal any Lewy bodies in any case.

The topographic distribution and severity of histologic degenerative changes are shown in table 2. In every case but two, the severity of degenerative changes (i.e., neuronal loss and gliosis with or without status spongiosis) was maximal in the parietal cortex, with the cortical asymmetry corresponding to the severity of limb findings (e.g., left > right cortical degeneration corresponding to right > left apraxia, rigidity, etc.). The frontal cortex was equally or slightly less severely involved than the parietal cortex all cases. In two cases, the anterior cingulate cortex was maximally involved. More variable changes were noted in the occipital and temporal cortices. Degeneration of the substantia nigra was present to some degree in 11 cases but negligible in 2. Basal ganglia degeneration was absent in five cases. This lack of basal ganglia and nigral degeneration is noteworthy because all patients had had extrapyramidal signs.

Discussion.

This series adds to the growing literature that shows several neurodegenerative disorders and prion disease can underlie the constellation of clinical findings thought to be specific for CBD. These disorders include AD (Patients 8 and 9 in this series and other reported cases20,26-28), Pick’s disease (Patient 11 in this series and other reported cases19,20,29), PSP (Patient 10 in this series and other reported cases20,30), nonspecific degenerative changes (Patient 12 in this series and the reported case of Brown et al.31), Creutzfeldt-Jakob disease (Patient 13 in this series and the reported case of Cannard et al.32), as well as CBD (Patients 1 to 7 in this series and other reported cases2-12).

The seemingly high clinical misdiagnosis rate in this study is not surprising. Previous experience with clinicopathologic correlations in neurodegenerative disorders has tended to show the following: The initial reports on a unique and rare disorder suggest there is a strong correlation between a certain constellation of clinical findings and that particular disorder, but as additional cases with the same clinical appearance undergo autopsy, an appreciation of the pathologic heterogeneity evolves. Heterogeneity becomes especially apparent when a large number of patients carrying the same clinical diagnosis are followed prospectively until death. The misdiagnosis rate of 10 to 35% has been noted in studies performed in this manner in patients with clinically diagnosed PD33 and AD (see review by Klatka et al.34). If future studies indicate that CBD, Pick’s disease, and PSP represent phenotypes within the spectrum of a similar pathophysiologic process (see below), the misdiagnosis rate may become comparable with that of PD and AD.

There were also overlapping pathologic findings in the PSP and Pick’s disease cases and in one CBD case. These findings are consistent with other reports in which pathologic alterations characteristic of CBD and other presumably distinct disorders coexist.15,22,35

Although Creutzfeldt-Jakob disease has been reported to mimic CBD, and we have evaluated three similar cases of Creutzfeldt-Jakob disease presenting with “rapidly progressive CBD” (unpublished data, 1996), the rapid progression with duration of symptoms less than 2 years has suggested prion disease. Patient 13, our patient with prion disease pathology, is noteworthy because he had classic progressive CBD features over a 4-year period.

The findings from a recent study suggest that CBD is markedly underdiagnosed.23 The findings in this analysis suggest that CBD is overdiagnosed. This discrepancy likely reflects several factors, one being the manner in which clinical diagnoses were made because diagnoses were determined in our series based on prospective clinical evaluations as opposed to ratings of clinical vignettes in the study of Litvan et al.23 Although these authors suggest that the diagnosis may have improved if the neurologists had examined the patients, the clinical features that are listed as the best predictors of underlying CBD in their report were also present in most of our non-CBD patients. Myoclonus was identified as one of the better clinical predictors,23 but despite two or more examinations in each of our autopsy-proven CBD cases, myoclonus was never noted in any. Therefore, despite the rigorous attempts to improve diagnostic accuracy of CBD by several groups of experienced clinicians, the diagnosis cannot be made with accuracy in life at the present time.

Despite variable pathologic diagnoses in this series, the one consistent histopathologic finding was asymmetric parietal and frontal cortical degeneration. Many of the clinical findings in this syndrome are therefore explained by this topography of cortical damage. Cortical sensory loss and apraxia occur with parietal lesions and occasionally with frontal lesions.36 Frontal or parietal lesions, or both, can also be associated with the alien limb phenomenon37 and mirror movements.38 Other less common findings such as pyramidal tract signs, oculomotor impairment, gait apraxia, apraxia of speech, aphasia, alexia, and agraphia can result from frontoparietal with or without temporal cortical dysfunction.

The specific neuroanatomic substrates underlying the motor findings are less clear. In our experience, some patients have clinical findings suggesting pyramidal tract pathology (e.g., asymmetric hyperreflexia, Hoffmans’s or Babinski’s signs, flexion or extension contractures, spasticity with velocity-dependent increases in muscle tone, etc.); extrapyramidal findings (e.g., rigidity, dystonia, tremor) are also a consistent feature. Dysfunction of neurons in the motor cortex and their afferent/efferent projections are the likely substrates for pyramidal tract findings; marked degeneration in the motor cortex has been reported in CBD39 and PSP.40 A recent study employing electrical stimulation in patients undergoing evaluation for surgical treatment of intractable epilepsy demonstrated that a significant proportion of motor responses were elicited from postcentral gyrus stimulation, and conversely, many sensory responses were elicited from precentral gyrus stimulation.41 Hence, damage to motor neurons residing in the parietal region could explain the motor findings in some who have primarily parietal lobe degeneration.

Rigidity, dystonia, and tremor have been presumed to reflect basal ganglia or nigral degeneration, or both.17 The lack of marked and sustained improvement in extrapyramidal signs after levodopa administration is an almost universal finding in cases clinically diagnosed with CBD,42 which presumably reflects postsynaptic nigrostriatal dysfunction rather than substantia nigra degeneration.17 However, basal ganglia or nigral degeneration, or both, have been absent on rare occasions in previous reports; in fact, the basal ganglia were spared in the index cases of Rebeiz et al.1 The clinicopathologic correlations in two of our cases are especially noteworthy; extrapyramidal findings were striking (in Case 12) or moderate (in Case 13) despite negligible degeneration of the basal ganglia and substantia nigra. Therefore, detectable degeneration of these subcortical gray and pigmented structures is not necessary to produce rigidity, dystonia, and tremor.

There are reports of similar extrapyramidal findings despite relatively normal basal ganglia and substantia nigra in patients diagnosed pathologically with AD26,43,44 and Pick’s disease.19,29 Limb akinesia, cataleptic posture, levitation, and rigidity with and without coexisting apraxia have also been reported in association with parietal cortex lesions.45 Therefore, extrapyramidal symptomatology might reflect dysfunction of the somatosensory, motor, and/or premotor cortices and their afferent/efferent projections.

All our patients but one had a distinct syndrome with progressive asymmetric rigidity and apraxia dominating the clinical picture, with the exception having asymmetric rigidity only. This constellation of findings, which represents the core features of the “CBD syndrome,” reflects the topography of parietofrontal degeneration, and a spectrum of disorders can manifest clinically in this fashion. The CBD syndrome can therefore be viewed in the context of other focal/asymmetric cortical degeneration syndromes in which a similar spectrum of disorders can manifest clinically as a frontal lobe dementia syndrome or a progressive aphasia syndrome.16 It is interesting to note that the preponderance of disorders that can present in a focal/asymmetric fashion have tau+ pathology. Furthermore, although features suggesting cortical and extrapyramidal system dysfunction are characteristic of dementia with Lewy bodies (DLB),46 DLB has not been identified in any patient with CBD syndrome findings to our knowledge. Because none of the patients in this series ever experienced complex visual hallucinations—a feature considered suggestive of DLB in the clinical setting of dementia and parkinsonism—the absence of visual hallucinations may be a key clinical feature that differentiates disorders with prominent tau pathology from those within the spectrum of disorders with Lewy bodies.

Recent analyses indicate that many cases with familial frontotemporal dementia with parkinsonism linked to chromosome 17 have prominent tau+ pathology and various combinations of achromatic neurons typical of CBD, Pick bodies characteristic of Pick’s disease, or globose neurofibrillary tangles typical of PSP.47 Whether CBD, Pick’s disease, and PSP represent distinctly different pathophysiologic processes or a spectrum of the same pathophysiologic process is a matter of debate.48,49 The rapidly evolving molecular research findings in these disorders increasingly indicate more similarities than differences. The accurate differentiation of CBD from other neurodegenerative disorders will therefore likely require analysis of biological material rather than rely on any clinical classification system.

Addendum.

Since this paper was accepted for publication, an additional four patients with clinically diagnosed CBD have died, and detailed neuropathologic examinations have been completed. The histopathologic diagnoses were CBD in two, nonspecific degenerative changes in one, and a combination Alzheimer’s and Pick’s disease changes in the other. The topographic distribution of degenerative changes in these cases was the same as that described in this article.