Held 14-15 January 2011 in Miami, Florida, the 5th annual Human Amyloid Imaging conference illustrated how this form of PET imaging is becoming broadly embedded throughout human Alzheimer’s disease research. Consider these examples.

Agneta Nordberg of the Karolinska Institute in Stockholm, Sweden, showed that amyloid imaging, in this case with 11C PIB, forms part a multi-tracer study by which this group is following patients with sporadic AD, MCI, and also families with autosomal-dominant mutations in APP and presenilin. Reminiscent of DIAN, this study measures neuropsychology, protein biochemistry in the CSF, brain structure with MRI, and brain metabolism with FDG-PET. The new piece Nordberg introduced in Miami was an attempt to see how neuroinflammation plays into AD by adding on PET imaging with 11C-d-deprenyl. This tracer visualizes astrocytosis, said Nordberg.

Inflammation is of intense interest in AD research, but has remained largely elusive to brain imaging. Some studies have used the only available PET tracer for neuroinflammation—(R)-PK11195 for microglia—together with PIB (Wiley et al. 2009; Okello et al., 2009; Yokokura et al., 2011; see also Schuitemaker et al., 2010), but many AD researchers consider PK11195 unsatisfactory. As for 11C-d-deprenyl, little is known about it. Nordberg mentioned recent work on this tracer in amyotrophic lateral sclerosis, Creutzfeldt-Jakob disease, and epilepsy, all indicating that astrocytosis happens early in disease. Arthritic joints have been reported to take it up (Danfors et al., 1997).

Comparing PIB and d-deprenyl retention in the same volunteers, Nordberg’s studies are ongoing. To date, her team sees that d-deprenyl tends to go up early, though amyloid and this astrocytosis marker appear not to be regionally correlated. D-deprenyl retention was high in MCI patients who are PIB-positive; it was also high in mutation carriers who are cognitively normal and whose amyloid burden is low. It was relatively lower in people who had advanced to Alzheimer’s, suggesting that astrocytosis appears to happen early on at the MCI or even pre-symptomatic stage. “This study shows we can do multimodal imaging studies to look at more variables at once in the same patients. This marker could potentially be incorporated into drug testing,” Nordberg said. “We are very excited about this and are trying to recruit more patients participating in this study,” she told ARF. That said, d-deprenyl does not image neuroinflammation exhaustively; more markers are needed.

On that note, Nordberg mentioned that amyloid imaging of patients with the Arctic mutation reinforces a related point that is sometimes overlooked as the first amyloid tracers are close to receiving regulatory approval for clinical use (see ARF related news story). That is, scientists urgently need PET tracers to image oligomeric/protofibrillar forms of pathogenic proteins to fill in the picture of what is going on in the brain. Case in point: Nordberg showed PET images of a 60-year-old woman who has AD resulting from the Arctic APP mutation, and this woman is PIB negative. The Arctic mutation causes early onset AD that progresses rapidly but is otherwise clinically quite typical. Rather than the classic cored, fibrillar amyloid plaques, this form of AD turned out on postmortem pathology to have caused ring-shaped plaques without a distinct core (Basun et al., 2008). The mutation does cause abundant formation of smaller protofibrillar forms of Aβ. These cases are extremely rare; however, they show that clinical AD can arise from forms of Aβ that neither of the current fibrillar amyloid tracers pick up. Imaging these smaller forms of Aβ would add much to understanding the disease. “We need more PET ligands,” Nordberg said.

A poster on antidepressants and brain amyloid created a buzz at HAI as well. Yvette Sheline of Washington University, St. Louis, Missouri, showed that among 186 cognitively normal elderly research volunteers who had undergone a PIB-PET scan, those who had taken antidepressant medication in the past five years had less brain amyloid. The longer these volunteers had been taking antidepressants, the less amyloid they had. The volunteers were otherwise closely matched in age, ApoE and other factors that might affect their amyloid load. The antidepressants in question here are the selective serotonin reuptake inhibitors. The human studies were prompted by earlier results in mice. Using microdialysis, Sheline’s coauthor John Cirrito and colleagues had found that both citalopram and fluoxetine, as well as direct infusion of serotonin, slashed soluble brain Aβ levels by a quarter to a third. This suggests that serotonin signaling might affect amyloid deposition, Sheline said.

How this works remains unclear, as this line of research is at an early stage. The mouse studies were acute, not long-term. The human data came from a so-called sample of convenience, people who were asked about their antidepressant use, not from a drug trial designed to test this hypothesis. Only one such trial has published a reduction of PIB retention to date (Rinne et al., 2010; see also Part 5 of this series). Yet given the well-known safety profile of SSRIs, researchers might want to consider randomized trials to see if these drugs indeed could double as amyloid-lowering agents. At this point, Sheline’s message to readers is: Don’t try this at home.

Consider one more example of amyloid imaging in AD research. Keith Johnson of Massachusetts General Hospital, Boston, used it to take a closer look at microbleeds. These small ruptures in a blood vessel have caused concern in AD clinical trials and have generated a literature in mouse models, but remain poorly understood. The tiny dollop of blood eventually turns into hemosiderin, which makes a big black dot on MRI. A population-based imaging study had pinned their prevalence at about 30 percent by age 75, and had suggested that microbleeds seen in the brain’s upper, i.e., lobar regions, are related to CAA and AD, whereas microbleeds occurring deep inside the brain have to do with cerebrovascular disease (Vernooij et al., 2008).

Johnson and colleagues looked at both microbleeds and CAA together with an MRI/PIB/FDG-PET study in 92 cognitively normal or very mildly impaired people in their seventies, who also underwent cognitive tests and genotyping. When two raters analyzed the images separately, it became clear that distinguishing them from vessels that appear cut across in the plane of view was not trivial. “You have to be careful not to overcall them,” Johnson said.

How common were these signs of vascular injury in the brains of this set of older people? Twenty-three of the 92 folks had microbleeds, about evenly split between lobar and deep. Of the 23, most people had one microbleed, a few had two, two people had four or five. What were they related to? To ApoE status and age, to cortical PIB retention in the case of lobar microbleeds, and to a diagnosis of hypertension in the case of deep microbleeds. In other words, older people with ApoE4 who had amyloid deposition or high blood pressure were most likely to have a microbleed. Was having a microbleed in any way related to cognitive performance or cerebral hypometabolism? Not in this sample, Johnson reported. “In normal people, the microbleeds are somewhat common. They are a feature of preclinical AD but appear not to be associated with significant clinical findings,” Johnson said.

On balance, these data sounded less alarming than some previous studies; however, Johnson pointed out that the MGH sample was smaller and healthier than the Rotterdam cohort, for example. The HAI audience followed this talk with an animated discussion. The underlying idea is that vascular amyloid increases the risk for bleeding. Hence, when a microbleed shows up on MRI, that implies the person might have CAA and be at higher risk for stroke. But how that is interpreted in clinical practice varies quite a bit. Some centers go so far as to infer a person probably has AD when a microbleed shows up on MRI. Other scientists said this was jumping to conclusions and should not be done. Other practitioners assume that when they see one microbleed, it represents the tip of the iceberg. Here, too, Johnson’s data would suggest that is wrong. Johnson noted that his group used a highly sensitive technique and scrutinized the brains quite exhaustively. “My hunch is that there aren’t automatically many more if you see one,” Johnson said. Charles DeCarli of the University of California, Davis, an expert on vascular dementia, concurred. “People sometimes overestimate the import of a microbleed,” DeCarli said. “In a community setting, about 70 percent of people in this age group have single silent infarcts. On the cognitive side, these silent infarcts do not do much. When you have a PIB-positive person with a microbleed, that is probably CAA. But you cannot simply reverse this implication.”—Gabrielle Strobel.

This is Part 4 of a seven-part series. See also Part 1, Part 2, Part 3, Part 5, Part 6, and Part 7. View a PDF of the entire series.


Comments on this content

No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. DIAN Dispatch From Hawaii: After Slow Start, Network Is Humming
  2. Committee Shoots Down Florbetapir, Raising Bar for Field at Large
  3. Miami: Updates on J-ADNI, 18F Tracers, Biopsies
  4. Miami: Women Rock at Human Amyloid Imaging Meeting
  5. Miami: Amyloid in the Aging Brain—What Does It Mean?
  6. Miami: Multimodal Imaging, New Way to Test Amyloid Hypothesis
  7. Miami: Is Human Amyloid Imaging Ready for Clinical Trials?
  8. Miami: HAI Amyloid Imaging Conference Abstracts

Paper Citations

  1. . .
  2. . .
  3. . .
  4. . .
  5. . .
  6. . .
  7. . .
  8. . .

Other Citations

  1. View a PDF of the entire series.

Further Reading