To investigate amyloid accumulation by carbon 11–labeled Pittsburgh Compound B (11C-PiB) in hereditary cerebral amyloid angiopathy and APP locus duplication.
Design, Setting, and Patients
Positron emission tomography with 11C-PiB and magnetic resonance imaging were performed for 2 patients, 49-year-old and 60-year-old siblings with APP locus duplication, with hereditary Alzheimer disease and cerebral amyloid angiopathy.
Main Outcome Measure
Change in 11C-PiB uptake.
Uptake of 11C-PiB was increased especially in the striatum (caudate nucleus to 225% and 280% of the control mean and putamen to 166% and 185% of the control mean) and in the posterior cingulate (to 168% and 198% of the control mean), and it was marginally increased in other cortical brain areas. The pattern of increased 11C-PiB uptake was different from that seen in sporadic Alzheimer disease.
Amyloid imaging with 11C-PiB positron emission tomography is a useful tool for detecting in vivo amyloid accumulation in patients with hereditary cerebral amyloid angiopathy. However, the pattern of 11C-PiB accumulation differs between patients with typical AD and patients with APP locus duplication.
Deposition of the β-amyloid peptide (Aβ) in neuritic plaques in the brain is a hallmark of Alzheimer disease (AD). The accumulation of Aβ in vessel walls leads to cerebral amyloid angiopathy (CAA), which is an important cause of cerebral hemorrhage and microbleeds.1 Autosomal dominant CAA related to Aβ has been traced to mutations in the amyloid precursor protein (APP) gene (Online Mendelian Inheritance in Man 104760) and duplication of the APP locus.2- 6 The most prominent feature in all cases has been severe CAA in the leptomeningeal vessels together with superficial and deep intraparenchymatous small arteries, capillaries, and venules. Determination of CAA in vivo is often unspecific and is limited to the detection of hemorrhages in computed tomography or microbleeds with T2*-weighted gradient-echo magnetic resonance imaging.
Positron emission tomography (PET) has been used for the in vivo detection and quantification of amyloid burden in the brain. Imaging with PET using ligands that bind selectively to amyloid has been observed to be a promising method for the early detection of AD. Clear increases in carbon 11–labeled Pittsburgh Compound B (11C-PiB) uptake, a derivative of thioflavine T, have been demonstrated in several cortical areas in patients with AD relative to healthy control subjects.7 We have previously described a Finnish family with autosomal dominant dementia and frequent CAA and intracerebral hemorrhages due to APP locus duplication6,8 in whom neuropathological examinations revealed AD-type changes with Aβ in neuritic plaques and vessel walls. The present study was aimed at investigating the possibilities of 11C-PiB PET imaging for detecting amyloid pathological findings in CAA in vivo and comparing the distribution of amyloid deposits in patients with CAA and APP locus duplication with that in patients with typical AD.
A 49-year-old man started to experience cognitive problems 4 to 5 years before PET imaging. Neuropsychological examinations had been performed annually since then. At the time of PET scanning, his Mini-Mental State Examination score was 22/30. The neuropsychological profile revealed mild disorientation in time and progressive impairment in verbal memory. Visuospatial functions and construction were mildly impaired, but the patient performed better than in the previous examination, a feature compatible with the vascular component of the disease.
A 60-year-old woman had memory problems starting 5 years before the 11C-PiB PET imaging. At the time of the PET scan, her Mini-Mental State Examination score was 17/30 and the neuropsychological examination revealed disorientation in time and marked impairment of verbal memory. She also had mild visuospatial problems and dyspraxia. Her cognitive profile has been constantly progressive. Both patients had substantially better visuoconstructive skills than patients with AD at this stage. Both patients harbor the 0.55-megabase duplication of APP.6
At the time of PET scanning, these patients with APP duplication also underwent a magnetic resonance imaging scan with a Philips Gyroscan Intera 1.5-T CV Nova Dual scanner (Philips Medical Systems, Best, the Netherlands). This included axial T2-weighted spin-echo images (repetition time, 4488 milliseconds; echo time, 100 milliseconds; slice thickness, 6 mm) with a matrix of 512 × 512 and 2 excitations, coronal fluid-attenuated inversion recovery images (repetition time, 11 000 milliseconds; inversion time, 2800 milliseconds; echo time, 140 milliseconds; slice thickness, 6 mm) with a 512 × 512 matrix and 2 excitations, and axial T1-weighted 3-dimensional images (repetition time, 25 milliseconds; echo time, 5 milliseconds; slice thickness, 1 mm) with a 512 × 512 matrix and 1 excitation. The magnetic resonance imaging scans were interpreted by an experienced neuroradiologist (R.P.).
Imaging with 11C-PiB PET was performed with a GE Advance PET scanner in 3-dimensional mode (GE Medical Systems, Milwaukee, Wisconsin). About 500 MBq of 11C-PiB was injected intravenously and a 90-minute dynamic emission scan was performed.
Automated region-of-interest analysis was performed as described earlier9,10 with minor supplementation. To facilitate comparability of the results,11C-PiB uptake quantification at a 60- to 90-minute interval was made using both the region-to-cerebellum ratio and the region-to-pons ratio. The results were compared with previously published results in patients with AD and healthy volunteers examined at Turku PET Centre with identical imaging protocols and analysis methods.9 The AD group comprised 17 patients with mild to moderate AD (9 women, 8 men; mean age, 72.0 years; age range, 55-85 years; mean Mini-Mental State Examination score, 23.6), and the control subjects were 11 healthy volunteers (8 women, 3 men; mean age, 65.1 years; age range, 51-74 years; mean Mini-Mental State Examination score, 28.2).
Brain tissue from 2 autopsies in the family was available (family members III:10 and III:14).8 β-Amyloid was demonstrated immunohistochemically using monoclonal mouse antihuman antibody (clone 6F/3D; DakoCytomation Denmark A/S, Glostrup, Denmark), which recognizes residues 8 through 17 of the Aβ protein. Meningeal and superficial perforating arterial amyloid accumulations and amyloid deposits in senile plaques were evaluated semiquantitatively.
The PET study revealed that the uptake of 11C-PiB was increased especially in the striatum (caudate nucleus to 225% and 280% of the control mean and putamen to 166% and 185% of the control mean) and in the posterior cingulate (to 168% and 198% of the control mean) (Table 1). In the other cortical brain areas, the uptake was on the margin of the mean control values. This pattern of increased 11C-PiB uptake is different from that seen in pure AD, where the largest increases are in the frontal and parietal cortices, followed by the posterior cingulate, striatum, and lateral temporal cortex. An example of an image with 11C-PiB at the posterior cingulate and striatal level in case 1 is shown in Figure 1.
Carbon 11–labeled Pittsburgh Compound B positron emission tomographic scan of patient 1. A, Transverse section at the level of the posterior cingulate. B, Transverse section at the level of the basal ganglia. Red indicates areas of high uptake of carbon 11–labeled Pittsburgh Compound B. Note the increased signal in the posterior cingulate and basal ganglia.
The T2-weighted and fluid-attenuated inversion recovery magnetic resonance images of both patients with APP duplication at the time of the PET scan showed multiple punctate high-signal white matter lesions (Figure 2). Hippocampal atrophy was present in both patients, being of grade 2/4 on both sides in the 60-year-old woman and of grade 1/4 on the left side and grade 2/4 on the right side, evaluated on a visual scale, in the 49-year-old man.11 No findings incompatible with a diagnosis of AD were found, but several microbleeds had been seen with T2*-weighted gradient-echo magnetic resonance imaging of case 1 one year before the PET scanning.
Magnetic resonance imaging scans. A, Transverse T2-weighted section at the upper level of the lateral ventricles. B, Coronal fluid-attenuated inversion recovery section at the level of the occipital lobes and cerebellum. C, Coronal reconstructed T1-weighted section of the hippocampus. Note the hyperintense ischemic-degenerative white matter lesions near the ependyma of the lateral ventricles on T2-weighted and fluid-attenuated inversion recovery sections and hippocampal atrophy on the coronal T1-weighted section.
Brain tissue from 2 autopsies in the family was available (Table 2). The most prominent accumulation of Aβ was seen in the parietal region in both cases, but there was also accumulation in the frontal and temporal areas of the brain.
We found that the uptake of 11C-PiB was increased especially in the caudate nucleus and posterior cingulate and slightly less clearly in the putamen of the patients with APP duplication. In the other cortical brain areas, the uptake was close to the mean control values. Previous studies had shown the typical retention pattern of 11C-PiB to be most prominent in the frontal cortex of patients with AD (1.9 times that in the control subjects) but to be high in other cortical regions such as the parietal, occipital, and temporal cortices and the posterior cingulate (up to 1.5-1.7 times).7,9 Thus, the topographical pattern of increased 11C-PiB uptake in patients with APP duplication is different from that seen in patients with sporadic AD. However, a similar pattern of prominent 11C-PiB retention in the striatum is detected in patients with presenilin 1 (PS1) gene mutations.12
Neuropathological examination of 2 members of the same family revealed Aβ accumulation in both plaques and vessel walls. These findings were consistent with a diagnosis of definite AD according to the Consortium to Establish a Registry for Alzheimer's Disease. The most prominent accumulation of Aβ was seen in the parietal region, but accumulation was also seen in the frontal and temporal areas of the brain, whereas lesions are usually less prominent in the thalamus, putamen, and caudate nucleus. The neuropathological findings were similar to those seen in other cases of APP duplication.13 The Aβ accumulation was surprisingly mild in the striatum, but 11C-PiB uptake was distinct in this region. In patients with PS1 mutations, both 11C-PiB uptake and Aβ-immunoreactive and 6-CN-PiB– or X-34–positive plaques were prominent in the striatum. However, neuritic amyloid pathological findings were not present.12 This type of immunoreactive amyloid could not be excluded in the present family with APP duplication. One limitation is that the postmortem samples were from individuals different from those who underwent 11C-PiB PET scans; the autopsy subjects were probably at a more advanced stage of the disease, so any comparison of their amyloid accumulations with in vivo 11C-PiB uptake should be undertaken with caution. Unfortunately, the posterior cingulate was not systematically examined at that time and no tissue was available for plaque and amyloid load determination. The most prominent feature was severe CAA in the leptomeningeal vessels, together with superficial and deep intraparenchymal small arteries, capillaries, and venules.8 This phenomenon has also been seen in other families with APP duplication.13 Cerebrovascular amyloid contains fibrillar amyloid and stains with PiB,14 so it may also contribute to the topographical pattern of 11C-PiB uptake in patients with APP duplication. Indeed, in vivo 11C-PiB and postmortem analysis of amyloid load in 1 patient with dementia and Lewy bodies revealed that the dominant source of the 11C-PiB signal in that case was due to cerebrovascular amyloid.14 Because cerebellar Aβ plaques have been detected post mortem in families with early-onset AD with PS1 mutations, we also calculated 11C-PiB uptake using the pons as a reference area, as has been done previously.13 The results were almost identical to those obtained with the cerebellar reference (Table 1). Accordingly, as suggested by postmortem examination of 2 cases (Table 2), there is not a significant amount of cerebellar Aβ plaques in our patients with APP duplication.
In conclusion, the pattern of 11C-PiB accumulation differs between patients with typical AD and patients with APP duplication. The different pattern of 11C-PiB accumulation is probably due to overproduction of Aβ as seen in families with PS1 mutations.13 However, patients with APP duplication have abundant cerebrovascular amyloid, which also binds 11C-PiB and thus could contribute to increased signal.
Correspondence: Anne M. Remes, MD, PhD, Department of Neurology, University of Oulu, Box 5000, FIN-90014 Oulu, Finland (firstname.lastname@example.org).
Accepted for Publication: September 16, 2007.
Author Contributions:Study concept and design: Remes, Laru, and Rinne. Acquisition of data: Tuominen, Kemppainen, Mononen, Någren, and Parkkola. Analysis and interpretation of data: Laru, Tuominen, Aalto, Mononen, Parkkola, and Rinne. Drafting of the manuscript: Remes, Laru, and Aalto. Critical revision of the manuscript for important intellectual content: Tuominen, Kemppainen, Mononen, Någren, Parkkola, and Rinne. Statistical analysis: Aalto. Obtained funding: Remes and Rinne. Administrative, technical, and material support: Laru, Tuominen, Aalto, Mononen, Någren, Parkkola, and Rinne. Study supervision: Remes, Parkkola, and Rinne.
Financial Disclosure: None reported.
Funding/Support: This work was supported by a grant from the Finnish Medical Foundation, the Päivikki and Sakari Sohlberg Foundation, project 205954 from the Academy of Finland, the Sigrid Juselius Foundation, and clinical grants from Turku University Hospital (EVO).
Remes AM, Laru L, Tuominen H, Aalto S, Kemppainen N, Mononen H, Någren K, Parkkola R, Rinne JO. Carbon 11–Labeled Pittsburgh Compound B Positron Emission Tomographic Amyloid Imaging in Patients With APP Locus Duplication. Arch Neurol. 2008;65(4):540-544. doi:10.1001/archneur.65.4.540