Objective: Brain amyloid imaging using positron emission tomography (PET) is of increasing importance in the premortem evaluation of dementias, particularly in relation to Alzheimer disease (AD). The purpose of this study was to explore the premortem diagnostic utility of 11C-PiB PET in sporadic Creutzfeldt–Jakob disease (CJD).
Methods: Two patients, 72 and 59 years old, underwent evaluation for rapidly progressive cognitive decline, dying after illness durations of 5 and 7 months, respectively. As part of their comprehensive assessment, 18F-FDG PET and 11C-PiB PET studies were performed approximately 2–4 weeks prior to death, and the brain regional distributions compared with those from cohorts of healthy controls (HC) and AD patients.
Results: Routine investigations, including brain MRI scans, revealed changes typical of sporadic CJD, with the diagnosis confirmed at autopsy in both patients. The 18F-FDG PET showed global hypometabolism in one patient and thalamic and frontal hypometabolism with unexpected hypermetabolism in the dentate nuclei of the cerebellum in the other. Neither patient displayed cerebral cortical 11C-PiB PET retention above the levels observed in HC.
Conclusions: No grey-matter 11C-PiB retention was observed in two pathologically confirmed cases of typical sporadic CJD. We speculate that low PrP plaque density and small plaque size, as well as a relatively low affinity of the radioligand, explain the absence of 11C-PiB retention. More studies to validate this hypothesis are warranted.
Statistics from Altmetric.com
The spectrum of transmissible human prion disease, characterised by the presence within the CNS of conformationally altered, protease-resistant isomers (PrPres) of the normal cellular form of the prion protein, PrPc,1 2 encompasses Creutzfeldt–Jakob disease, the variant form (vCJD) associated with bovine spongiform encephalopathy, kuru and Gerstmann–Straussler–Scheinker syndrome (GSS).3 4 Despite considerable improvement in diagnostic accuracy through the use of CSF biomarker assays5 and brain MRI,6 the search for additional diagnostic techniques continues. Neuropathological examination of brain tissue remains the only definitive method for diagnostic confirmation.
11C-2-(4′-methylamino-phenyl)-6-hydroxy-benzothiazole (11C-PiB) has a high affinity for Aβ fibrils, allowing in vivo quantification of amyloid deposits in the brain. There is a high correlation between 11C-PiB retention and Aβ concentrations assessed at postmortem or brain biopsy.7 8 9 Several commonly used neuropathological dyes have been shown to stain prion plaques. Thioflavin T (ThT), its derivative, 2-(4′-methylamino-phenyl)-benzothiazole (BTA-1),10 11 and the Congo Red derivatives BSB10 12 and X0413 bind in vitro and in vivo to PrPres in mouse and human prion-diseased brains. To date, however, the use of positron emission tomography (PET) in the diagnosis of sporadic CJD has been mainly centred in assessing glucose metabolism with 18F-FDG.14 15 A previous case report of two siblings with the same prion protein gene (PRNP) mutation described differential binding of two PET amyloid imaging agents (11C-PiB and 18F-FDDNP).16 Acknowledging that both ThT and its derivative, BTA-1, bind to PrP plaques, we explored the diagnostic utility of 11C-PiB in two patients with sporadic CJD confirmed by pathological examination.
Patients and methods
PET imaging studies
PET studies were approved by the Austin Health Human Research Ethics Committee. Written informed consent was obtained from the appropriate next of kin prior to the scans. A 40–70 min emission acquisition was performed after injection of 370 MBq of 11C-PiB. Ten minutes after completion of the 11C-PiB study, patients were injected with 185 MBq of 18F-FDG. Forty minutes later, a 20 min emission scan was acquired. Standardised uptake values (SUV) were obtained from regions of interest across cortical, subcortical and cerebellar regions. Acknowledging that PrP plaques are frequently found in the cerebellar cortex,17 11C-PiB and 18F-FDG SUV ratios were determined by normalising the regional SUV to the pons (SUVRpons). Regional distributions of each radioligand were compared against well-characterised cohorts of HC and AD patients.
Neuropathological examination and immunohistochemical detection of PrPres
Brain sections were treated with formic acid prior to formalin fixation and processing. Immunostaining for Aβ was performed with the 1E8 (1:50) monoclonal antibody, and for PrP with the 3F4 (1:1000, Covance) and 12F10 (1:2000, Cayman Chemical) antiprion protein monoclonal antibodies.18
Determination of prion protein gene codon 129 status and PrPres molecular subtype
Determination of whether the patient was methionine (M) or valine (V) homozygous or MV heterozygous at codon 129 of the prion protein gene and characterisation of the PrPres molecular subtype by western blot profile were performed as previously described.18
Statistical analysis was performed through Z scores generated against 11C-PiB and 18F-FDG regional SUVRpons from cohorts of both HC and AD patients. Z scores >±2.5 were considered significantly different.
Case report 1
Patient 1, a 59-year-old female, initially presented reporting 3 months of dizziness, gait unsteadiness and hearing difficulties, superimposed on a history of longstanding tinnitus and vertigo. MRI brain and ENT specialist review prior to neurological evaluation had not disclosed any abnormalities. One month later, the patient presented with increasing hearing disturbance manifesting as difficulty discerning words in conversation. Emotional lability, poor short-term memory, word-finding difficulties and prominent ataxia of gait were also evident. Brain MRI, including T1, T2-weighted, diffusion-weighted (DWI) and fast fluid-attenuated inversion recovery (FLAIR) acquisition sequences, reported non-specific small-vessel ischaemic changes. CSF examination was unremarkable aside a raised total protein, and presence of 14-3-3 protein. EEG displayed generalised slowing but no triphasic waves or periodic discharges. Six months after initial neurological evaluation, the patient was unable to hold a conversation or reliably name immediate family members, requiring assistance with feeding, standing and ambulating. There was a coarse postural tremor in the upper limbs. Repeat brain MRI revealed restricted diffusion and T2 hyperintensity in the left and right caudate nuclei and dorsomedial and pulvinar thalamic regions, with small foci of restricted diffusion in the left posterior temporal, left frontal and both occipital cortices. The patient died from respiratory difficulties approximately 7 months after initial neurological presentation. No definite myoclonus was ever observed.
Codon 129 genotyping revealed MV heterozygosity, and the PrPres western blot profile was type 3 (MV3).
PET imaging studies
Visually, the 18F-FDG scan showed hypometabolism of the thalami and frontal cortices, with hypermetabolism in the dentate nuclei of the cerebellum (see supplementary fig 1). There was no cerebral cortical or subcortical 11C-PiB retention (table 1) with the findings indistinguishable from HC (fig 1).
Macroscopic examination of the brain was unremarkable. Microscopic examination showed neuronal loss, gliosis and a spongiform encephalopathy. No Aβ plaques were observed. There was a fine “synaptic” pattern of PrP immunoreactivity in the corresponding areas with focal areas of granular PrP deposition (fig 1). A neuropathological diagnosis of CJD was made.
Case report 2
Patient 2, a 72-year-old female, was investigated for 4 months of progressive neurological decline. On hospital admission, dysarthria with limited speech output and reduced motor activity with shuffling gait, but no focal neurological signs nor myoclonus were observed. Brain MRI revealed increased signal in the left head of caudate nucleus on both FLAIR and DWI sequences, not present on a previous MRI 2 months earlier. CSF examination was unremarkable aside from detection of 14-3-3 proteins, while EEG demonstrated a generally slowed background with intermittent periodic sharp wave activity. The patient’s neurological condition inexorably declined, developing myoclonus and fluctuating consciousness, dying in an akinetic-mute state approximately 1 month after entering hospital.
Codon 129 genotyping revealed methionine homozygosity, and the PrPres western blot profile was type 2 (MM2).
PET imaging studies
Visually, the 18F-FDG scan showed global hypometabolism, more marked in the left frontal and left temporal cortices (see supplementary fig 1). There was no cerebral cortical or subcortical 11C-PiB retention, (table 1) with the findings indistinguishable from HC (fig 1).
External examination of the brain was unremarkable. Microscopic examination showed neuronal loss, gliosis and a spongiform encephalopathy. Occasional (+) Aβ diffuse plaques were observed in the frontal section. 12F10 immunoperoxidase studies demonstrated a “synaptic” pattern of PrP deposition, scattered small granular concentrations of PrP in and around vacuoles in the cortex (fig 1) but no plaque deposition. 3F4 immunohistochemistry additionally showed small plaque-like PrP deposits within the underlying white matter in frontal cortical sections and small granular plaques in the thalamus. There was also quite prominent granular PrP deposition in the midbrain neuronal areas. A neuropathological diagnosis of CJD was made.
To the best of our knowledge, this is the first report of 11C-PiB PET imaging in patients with confirmed sporadic CJD. Although previous studies have shown in vitro and in vivo binding of BTA-1 (the cognate parent compound of PiB) to PrP plaques, we found no significant brain retention of 11C-PiB in our patients. Our observations are in broad agreement with the findings of a study assessing 11C-PiB in siblings carrying a PRNP mutation.16
This lack of 11C-PiB binding might be attributed to PrP-plaque size and density within the brain. Due to the rapid progression and relatively short duration of symptomatic illness, sporadic CJD patients do not demonstrate appreciable PrP-plaque deposition at autopsy, where a positive correlation has been reported between illness duration and PrP-plaque density.19 Although some sporadic CJD molecular subtypes can manifest larger kuru-type plaques,20 the most common subtype (MM2, as in Patient 2) usually demonstrates only very fine “synaptic” and small perivacuolar deposits.19 20 Among the various molecular subtypes, MV3 commonly demonstrates PrP plaques, especially in the cerebellum.20 The lack of detectable 11C-PiB binding in our patients also concurs with a previous in vitro study of various forms of human prion disease, which only found BTA-1 binding to larger PrP plaques and not to smaller synaptic-type deposits.20 It appears possible that a certain threshold of plaque size and/or density may be required before appreciable 11C-PiB retention can be detected in vivo using PET. This is also highlighted by the presence of some Aβ diffuse plaques in Patient 2. While there is a positive correlation between Aβ burden as quantified by 11C-PiB PET and brain Aβ at post-mortem or biopsy,7 8 one patient with a high number of Aβ plaques presented with a 11C-PiB negative PET scan,9 highlighting the possibility that there might be certain conformational types of amyloid plaques that 11C-PiB does not detect in vivo.
The lack of 11C-PiB retention may also relate in part to reduced blood flow or to the relative affinities of 11C-PiB for the protein deposits at the concentrations achieved during PET studies. Despite almost identical PiB-negative scans, the respective 18F-FDG scans were quite distinct, suggesting blood flow was unlikely to constitute a significant determinant factor in 11C-PiB binding. In vitro studies demonstrating binding of BTA-1 to PrP plaques were performed using 1 μM BTA-1, while infected mice were injected with 10–30 mg/kg of BTA-1.10 These concentrations and doses are significantly greater than the low nanomolar range achieved in the brains after 0.00006–0.00007 mg/kg is injected into humans for a 11C-PiB PET study. Hence, it is likely that the relatively low affinity of 11C-PiB for aggregated PrP is translated in insufficient 11C-PiB concentrations to adequately label PrP plaques in vivo. This concurs with studies showing that despite their secondary β-sheet structure, 11C-PiB does not bind to other misfolded proteins at the concentrations achieved during PET studies.21 22
In conclusion, our observations suggest there might be a limited diagnostic role for 11C-PiB PET in the clinical evaluation of sporadic CJD, but it remains possible that other forms of human prion disease, such as GSS, vCJD and some sporadic CJD subtypes with higher burdens of larger PrP plaques may demonstrate significant binding of 11C-PiB. More studies to validate these hypotheses are warranted.
Funding PET studies were supported in part by funds from the Austin Hospital Medical Research Foundation, Neurosciences Victoria and the University of Melbourne. The Australian National Creutzfeldt–Jakob disease Registry is funded by the Commonwealth Department of Health and Ageing.
Competing interests None.
Ethics approval Ethics approval was provided by the Austin Health Human Research Ethics Committee.
Patient consent Obtained from the patients’ families.
▸ An additional figure is published online only at http://jnnp.bmj.com/content/vol80/issue9
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.