Dr. Gold and Colleagues Reply

To the Editor: Drs. Pantoni and Inzitari make several interesting points regarding the validity of clinicopathological correlations in vascular dementia. They stress the great difficulty in diagnosing subcortical vascular dementia and rightly suggest that the sensitivities and specificities of the criteria for general vascular dementia may be lower for subcortical diseases than for multi-infarct dementia, which was the main focus of our study. Although subcortical lesions are particularly common in aged brains, their clinical significance is still controversial. In this light, it would be meaningful to explore the performance of newly proposed specific criteria for subcortical vascular dementia (1).

It is important to note that the neuropathological criteria we chose to diagnosis patients with multi-infarct dementia were far from arbitrary. As indicated in the text, we applied a restrictive definition (the presence of both microscopic and macroscopic infarcts in at least three or more associative neocortical areas exclusive of the secondary visual cortex and the absence of significant other dementia-related pathology) that could lead to broad agreement that such cases were indeed instances of vascular dementia. We agree that the diagnosis of dementia itself must be clinical (all of our patients had clinical dementia); however, postmortem examination of the brain has a key role in identifying the underlying pathology associated with the clinical expression of dementia. In this respect, neuropathology must remain the gold standard for the validation of clinical criteria. Furthermore, examples of numerous clinicopathological studies of Alzheimer’s disease have demonstrated that it is possible to establish valid correlations between the progression of neuropathological changes and both the severity of cognitive decline and the specific pattern of affected cognitive domains (2, 3). Demonstrating such relationships is not yet possible in vascular dementia. In order to address this issue, there is indeed a strong need to develop consensual and validated neuropathological criteria that can also reflect the amount of vascular burden for each of the various subtypes of vascular dementia.