Neurology -- Correspondence for Honig et al., 64 (3) 494-500

Correspondence published:

Atherosclerosis and AD: Analysis of data from the US National Alzheimer’s Coordinating Center: Alex E. Roher, Chera Esh, Tyler Kokjohn, Lucia Sue, and Thomas Beach (28 March 2005)

Reply to Roher et al: Lawrence S. Honig, Walter Kukull, Richard Mayeux (28 March 2005)

Atherosclerosis and AD: Analysis of data from the US National Alzheimer’s Coordinating Center 28 March 2005

Alex E. Roher,
Sun Health Research Institute
10515 Santa Fe Drive, Sun City, AZ 85351,
Chera Esh, Tyler Kokjohn, Lucia Sue, and Thomas Beach
Send Correspondence to journal:
Re: Atherosclerosis and AD: Analysis of data from the US National Alzheimer’s Coordinating Center

alex.roher{at}sunhealth.org Alex E. Roher, et al.

We welcome the report by Honig et al. [1] regarding a statistical association between intracranial atherosclerotic vascular disease (AVD) and Alzheimer’s disease (AD). The authors and readers would be interested to know that this work confirms our three previously-published studies.
Our first study examined 32 AD cases and compared them to 22 non- demented elderly controls. [2] The circle of Willis was removed and atherosclerotic occlusion of the component arteries was assessed comprehensively using computerized image analysis. Arterial occlusion was significantly greater in the AD group than in the control group and there were significant correlations between mean arterial occlusion and both plaque and tangle density.
In our second study, we tested the association of intracranial AVD with AD in a larger series of cases. [3] For this study, we compared circle of Willis AVD in 215 AD and control subjects, using a semi- quantitative method. Again, subjects with AD had significantly greater AVD. Although we, like Honig et al, did not find an increased prevalence of cerebral infarcts in AD, we did find that AVD score was significantly greater in subjects with infarcts than subjects without infarcts. As in the report of Honig et al, subjects with the apolipoprotein E - E4 allele did not have increased AVD measures. In the second study, we found that when E4 cases were excluded from the study, the associations of AVD with plaque and tangle density were even stronger than with these cases included.
In our third study, we evaluated occlusion in the major leptomeningeal branches of the circle of Willis [4], in 10 AD cases and 10 control cases selected from the 54 that were used for the first study. AVD was again significantly greater in the subjects with AD.
Unlike Honig et al, we found that the association between AVD and tangle density was statistically significant. This difference may be due to greater variability in the AVD score and tangle density estimates in the former study, where scores were derived from multiple centers, as compared to our studies, where measurements were performed consistently by only three individuals.
The relationship between AVD and AD has been the subject of controversy since Alzheimer’s time [5]. These findings demonstrate, for the first time, a significant statistical association between direct and specific measures of intracranial atherosclerosis and Alzheimer’s disease.
References
1. Honig LS, Kukull W, and Mayeux R, Atherosclerosis and AD: analysis of data from the US National Alzheimer's Coordinating Center. Neurology 2005;64:494-500.
2. Roher AE, Esh C, Kokjohn TA et al. Circle of Willis Atherosclerosis Is a Risk Factor for Sporadic Alzheimer's Disease, Arterioscler Thromb Vasc Biol. 2003;23:2055- 2062.
3. Roher AE, Esh C, Rahman A, Kokjohn TA, Beach TG. Atherosclerosis of cerebral arteries in Alzheimer disease, Stroke 2004:35;2623-2627.
4. Kalback W, Esh C, Castano EM et al. Atherosclerosis, vascular amyloidosis and brain hypoperfusion in the pathogenesis of sporadic Alzheimer's disease. Neurol Res 2004;26:525-539.
5. Beach TG. The history of Alzheimer's disease: three debates. J Hist Med Allied Sci 1987;327-349.
The authors have no conflicts of interest to report. Sun Health Research Institute is a non-profit organization.

Reply to Roher et al 28 March 2005

Lawrence S. Honig,
Columbia University College of Physicians and Surgeons
630 W 168th St (P&S unit 16), New York, NY 10032,
Walter Kukull, Richard Mayeux
Send Correspondence to journal:
Re: Reply to Roher et al

lh456{at}columbia.edu Lawrence S. Honig, et al.

We thank Roher et al for their comments on the relationship between atherosclerosis and Alzheimer’s pathological findings, and discussion of their three studies of 54, 215, and 20 selected autopsy cases.
There is consistency between the results from these smaller series of autopsy cases at the Sun Institute, and the results using the larger number of autopsies from the database formed from the combined national Alzheimer’s Disease (AD)Centers. While we did not indicate it in our paper, we also found a significant relationship between cerebral atherosclerosis and both clinical stroke history and neuropathological infarcts. Like Roher et al, we found that atherosclerosis, but not stroke, was associated with senile neuritic plaques, characteristic neuropathologic findings of AD.
We found a weak association of atherosclerosis with neurofibrillary tangles by Braak stage only in nondemented individuals, and only in the unadjusted data. When adjusted for age, atherosclerosis related to neuritic plaques, but not to tangles by Braak stage, whether in demented or non-demented individuals. Roher et al comment that they found a relationship between atherosclerosis and tangles, and speculate that such a relationship was not present in our data due to increased variance due to intercenter variability. Since our measure of tangles was Braak staging, this seems unlikely. However, other differences between the studies include that their individuals were nearly a decade older (average age ~88 versus ~80), and that our analysis included an adjustment for age at death.
We agree that the relationship between cerebrovascular disease and AD has been the subject of controversy. It remains an important biological question, with clear potential therapeutic implications. However, we reiterate that currently we can not be certain of causality or its direction, with respect to our demonstrated association of atherosclerosis and Alzheimer’s disease. Further research is needed.