Imaging techniques promise to revolutionize the diagnosis of Alzheimer’s and other neurodegenerative diseases, but at this point, data on how live images correlate with pathological brain changes are still limited. In the December 13 Brain, researchers led by Agneta Nordberg of the Karolinska Institute in Stockholm, Sweden, report on the postmortem examination of the first-ever AD patient to be followed by positron emission tomography of Pittsburgh Compound B (PIB-PET). In common with the handful of prior such analyses, Nordberg and colleagues found that PIB retention accurately identified fibrillar amyloid deposits. In addition, the authors performed detailed studies not done before and turned up intriguing correlations between amyloid accumulation and synaptic receptor density, as well as a surprising lack of correlation with markers of inflammation. The data from this patient help point the way toward fruitful areas of investigation for future postmortem studies.

“Although there are several correlative studies in the literature, few are as detailed and comprehensive as this one,” wrote Bill Klunk of the University of Pittsburgh, Pennsylvania, in an e-mail to ARF (see full comment below). “It will continue to be of value to perform exquisitely detailed postmortem in vivo correlations such as this in many more brains.”

The first such study, in 2007, verified that in vivo PIB imaging matched actual fibrillar Aβ deposits in a patient with dementia with Lewy bodies (see ARF related news story on Bacskai et al., 2007). In 2008, these findings were confirmed in an AD patient (see ARF related news story on Ikonomovic et al., 2008).

The current paper analyzed the brain of a woman with AD who had two copies of the ApoE4 allele. She died at the age of 61. In February 2002, at the age of 56, she volunteered for the first PET-PIB scan ever performed (see ARF related news story on Klunk et al., 2004). She received another PIB scan two years later, and over the course of her disease also got an MRI and three PET scans using fluorodeoxyglucose (FDG), a marker for glucose use and therefore brain metabolism.

These longitudinal data are one of the things that sets this case study apart, Klunk noted, as they allow researchers to look at disease progression. Over the eight years she was studied, the woman’s score on the Mini-Mental State Examination declined from a near-normal score of 27 down to five. The FDG data showed that her brain’s glucose metabolism decreased in parallel with her cognitive powers. By contrast, PIB retention, already high at first examination, showed little change over two years during which her cognition declined steeply. The amount of amyloid deposition seen at autopsy three years later also looked similar to PIB estimates, Nordberg said, suggesting no further change in amyloid between the second PIB scan and death three years later. This pattern matches the data from numerous previous studies, in which PIB retention increases during mild cognitive impairment, then seems to plateau during AD (e.g., see five-year study by Kadir et al., 2010).

Autopsy results confirmed the patient’s diagnosis of pure AD. First authors Ahmadul Kadir and Amelia Marutle validated the accuracy of PIB imaging by staining slices from several brain regions with five different anti-Aβ antibodies and a tau antibody. They also examined PIB binding to brain homogenates from nine regions, and compared both autopsy methods to PIB imaging data. The results confirmed that in vivo PIB retention correlates quite well with amyloid deposits, but does not correlate closely with tau and neurofibrillary tangles, as previous studies have found (see Ikonomovic et al., 2008).

To look at synaptic changes, Kadir and colleagues performed binding assays on brain homogenates with markers specific for two different subunits of the nicotinic acetylcholine receptor. Previous studies found that nicotinic receptor density is lower in AD brains than in healthy people (see Paterson and Nordberg, 2000), and that the decline in nicotinic receptors correlates with cognitive decline (see Kadir et al., 2006). Intriguingly, Kadir and colleagues found that the density of the α4β2 nicotinic receptor subunit was lower in brain regions with high amyloid, suggesting an interaction. They found no correlation between amyloid and the density of the α7 subunit, however. The results “indicate that [specific] nicotinic receptors might be closely involved with amyloid processes,” Nordberg said. “That might open up new strategies for developing drugs.”

Gerhard Koenig at drug discovery company EnVivo Pharmaceuticals in Watertown, Massachusetts, pointed out that if α4β2 subunits disappear over the course of the disease, then drugs targeting those subunits are less likely to work late in the disease course. Four clinical trials with compounds targeting α4β2 for AD have turned up negative, Koenig added, which is consistent with the new findings. However, the results from this patient need to be repeated in many more brains, Koenig said, and in people with other ApoE genotypes, in order to get the full picture. Some α7 agonists are currently in clinical trials (e.g., this EnVivo Phase 2 trial).

The data on inflammation are more puzzling. Using an antibody to glial fibrillary acidic protein, which labels astrocytes, Kadir and colleagues found increased numbers of astrocytes populating areas with lots of amyloid. This fits with previous findings of increased astrocytosis in AD brains. However, when the authors used PET ligands specific for activated astrocytes and activated microglia, they saw no correlation between these cell types and areas of high amyloid deposition. This conflicts with some in vivo data (e.g., Edison et al., 2008). One possibility, the authors suggest, is that autopsy tissue simply does not bind these ligands well. Another explanation, Nordberg said, is that astrocytes are present but not activated. She also speculated that inflammation might be more pronounced earlier in AD, and therefore does not correlate well with amyloid at autopsy. Koenig suggests, on the other hand, that by late stages of the disease, the baseline inflammation of the brain might be so high that no further elevation can be seen in areas of high amyloid.

One limitation of PET-PIB is that it only measures fibrillar Aβ. Nordberg compared fibrillar, extracellular amyloid to snow that piles up outside a house: It can be quite deep and still cause no problems for those inside. Soluble, oligomeric Aβ, on the other hand, might be more like water that gets inside a house and causes extensive damage. “[Oligomers] would probably correlate much more with functional activity and cognitive impairment,” Nordberg said. She is working on methods to try to image oligomers in living brains. Nordberg also points to the value of studying things such as inflammation and synaptic function in living patients, as she believes this data will be crucial for the evolution of new drug therapies.—Madolyn Bowman Rogers


  1. This case report study of the first Pittsburgh Compound B (PIB) positron emission tomography patient with Alzheimer’s disease (AD) by Kadir et al. provides extensive descriptive information regarding the neuropathological changes found in the brain of a 61-year-old female patient who came to autopsy following an approximate eight-year duration of AD-like clinical symptoms. This thorough investigation provided pathological verification of the clinical diagnosis of AD as well as extensive correlative studies between PIB binding (presumably representing fibrillar Aβ burden) and several cognitive and neuropathological measures.

    Of note was a negative correlation between PIB binding and tissue homogenate assessment of nicotinic acetylcholine receptor binding. Specifically, the authors employed 3H-nicotine and 125I-α-bungarotoxin binding to assess the two major neuronal nicotinic acetylcholine receptor subtypes in the brain (α4β2 and α7) and nicotinic receptor subtypes, respectively. Interestingly, there was a negative correlation between 3H-nicotine binding and PIB retention, particularly in areas with the highest levels of PIB signal, whereas no correlation was found between PIB binding and 125I-α-bungarotoxin binding levels. The authors interpret these results as fibrillar amyloid having a negative drive on the expression of neuronal α4β2 receptors. This certainly may be the case, as there have been reports of decrements of 3H-nicotine binding in cortical tissue in postmortem AD brain, as well as more direct measurements of individual nicotinic receptor subunits via immunocytochemical and genomic-based approaches.

    A caveat to the present approach is that 3H-nicotine is a (well-established) binding assay that is selective, but not specific for α4β2 nicotinic receptors, just as 125I-α-bungarotoxin is selective, but not specific for α7 nicotinic receptors. Furthermore, these assays were performed in tissue homogenates that include admixtures of neuronal and non-neuronal cells, and cannot account for a decrease in 3H-nicotine binding as a function of neuronal loss, which is known to occur in the cortical regions assessed. One might argue that the observation of no changes in 125I-α-bungarotoxin binding in relation to PIB retention obviates the concern of neuronal loss, but this does not take into consideration the abundance of α7 nicotinic receptors on glial cells, particularly astrocytes, which may complicate any interpretation, as gliosis occurs in the cerebral cortex and neuronal loss takes place within the tissue that was used as input material for the assays. A method that our group has used to assess classes of relevant transcripts within vulnerable cell types, without the concern of neuronal cell loss, is single population gene expression profiling in combination with RNA amplification and custom-designed microarray analysis (Ginsberg et al., 2010). Specifically, we have performed expression profiling for acetylcholine receptors on vulnerable cholinergic basal forebrain (CBF) neurons within the nucleus basalis in postmortem AD and mild cognitive impairment (MCI).

    Although not within the cortex, CBF neurons displayed a statistically significant upregulation of α7 nicotinic receptor mRNA in subjects with mild to moderate AD compared with those with no cognitive impairment (NCI) and MCI (Counts et al., 2007). No differences were found for other nicotinic receptor mRNAs (including the α4 and β2 subunits) or muscarinic acetylcholine receptor subtypes across the cohort. Similar to the findings in the present study with PIB binding, expression levels of α7 nicotinic receptor mRNA was inversely associated with cognitive performance (Counts et al., 2007). Thus, expression of individual nicotinic acetylcholine receptor subunits is likely to be cell type-specific, and a correlation between expression levels and PIB binding may require additional studies at the single population level in tissue sections with genomic-based approaches or unbiased estimation techniques, in combination with immunocytochemical methods using subunit-specific antibodies (which is no easy feat in human postmortem brain material when it comes to acetylcholine receptors).

    In summary, this well-written and illustrated case report provides somewhat tantalizing information regarding the power of biomarker assays for fibrillar amyloid and postmortem neuropathological assessments. The research community will have to wait until a full cohort of subjects imaged with PIB comes to autopsy to get a greater understanding of the relationship(s) of amyloid levels with cognitive decline and neuropathology. However, the initial observation of a negative correlation between PIB retention and 3H-nicotine binding is certainly provocative and worth greater follow-up, even though there are many caveats to be considered before coming to a conclusion that fibrillar amyloid (or other neuropathological markers such as neurofibrillary tangles or neuropil threads) drives neurotransmitter-identified system (such as cholinergic and glutamatergic, to name just two) deficiencies that are increasingly becoming hallmarks of AD neuropathology.


    . Alpha7 nicotinic receptor up-regulation in cholinergic basal forebrain neurons in Alzheimer disease. Arch Neurol. 2007 Dec;64(12):1771-6. PubMed.

    . Microarray analysis of hippocampal CA1 neurons implicates early endosomal dysfunction during Alzheimer's disease progression. Biol Psychiatry. 2010 Nov 15;68(10):885-93. Epub 2010 Jul 23 PubMed.

    View all comments by Stephen D. Ginsberg
  2. This manuscript describes in detail the postmortem assessment of an AD patient studied twice with PIB-PET and three times with FDG-PET during life (last PIB 35 months before death). The subject happens to be the first patient ever studied with PIB-PET, and while this is of historical interest, the main points of the paper are correlation of postmortem measures with in vivo imaging. This sort of correlation remains of interest because, although there are several correlative studies in the literature, few are as detailed and comprehensive as this one. The findings of this study confirm what has been previously reported: 1) in vivo PIB retention is an accurate marker for the total insoluble (i.e., fibrillar) Aβ content of the brain, and 2) there is little progression of PIB retention over two years during the clinical phase of moderate AD—a time during which metabolism progressively decreases in parallel to worsening cognition. One unique aspect of this study is the longitudinal nature of the in vivo data. Another unique aspect is the demonstration of negative correlations between PIB and nicotine binding and positive correlations between PIB binding/retention and a marker of reactive astrocytosis. It will continue to be of value to perform exquisitely detailed postmortem-in vivo correlations such as this in many more brains studied with amyloid tracers in order to determine the limits of sensitivity of the in vivo tracers and to determine if exceptions to the tight in vivo-postmortem correlations can exist in brains with Aβ deposits that may not be primarily fibrillar.

    View all comments by William Klunk

Make a Comment

To make a comment you must login or register.


News Citations

  1. It Is Official: Autopsy Verifies Human PIB-Amyloid Connection
  2. Amyloid Imaging: Laying PIB Concerns to Rest
  3. Pittsburgh Compound-B Zooms into View

Paper Citations

  1. . Molecular imaging with Pittsburgh Compound B confirmed at autopsy: a case report. Arch Neurol. 2007 Mar;64(3):431-4. PubMed.
  2. . Post-mortem correlates of in vivo PiB-PET amyloid imaging in a typical case of Alzheimer's disease. Brain. 2008 Jun;131(Pt 6):1630-45. PubMed.
  3. . Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Ann Neurol. 2004 Mar;55(3):306-19. PubMed.
  4. . Dynamic changes in PET amyloid and FDG imaging at different stages of Alzheimer's disease. Neurobiol Aging. 2012 Jan;33(1):198.e1-14. PubMed.
  5. . Neuronal nicotinic receptors in the human brain. Prog Neurobiol. 2000 May;61(1):75-111. PubMed.
  6. . PET imaging of cortical 11C-nicotine binding correlates with the cognitive function of attention in Alzheimer's disease. Psychopharmacology (Berl). 2006 Nov;188(4):509-20. PubMed.
  7. . Microglia, amyloid, and cognition in Alzheimer's disease: An [11C](R)PK11195-PET and [11C]PIB-PET study. Neurobiol Dis. 2008 Dec;32(3):412-9. PubMed.

External Citations

  1. EnVivo Pharmaceuticals
  2. EnVivo Phase 2 trial

Further Reading

Primary Papers

  1. . Positron emission tomography imaging and clinical progression in relation to molecular pathology in the first Pittsburgh Compound B positron emission tomography patient with Alzheimer's disease. Brain. 2011 Jan;134(Pt 1):301-17. PubMed.