Connections in the brain’s default mode network (DMN) begin to falter years before the onset of clinical symptoms in both sporadic and familial Alzheimer’s disease (AD), according to two new papers. Researchers led by Beau Ances at Washington University in St. Louis, Missouri, focused on cognitively healthy older adults. In the August 19 JAMA Neurology, they report that poor connectivity in the DMN, a network called into action during periods of introspection, correlates with changes in cerebrospinal fluid (CSF) biomarkers that indicate preclinical AD. A similar phenomenon occurs in familial AD, as described in the August 20 Neurology. Researchers led by Reisa Sperling at Brigham and Women’s Hospital, Boston, Massachusetts, found evidence of weak connections in the DMN years before expected symptom onset in people who have inherited a dominant AD gene.

“The results suggest that a relatively simple magnetic resonance imaging (MRI) measure might be a reasonable biomarker for early AD,” wrote William Jagust at the University of California, Berkeley, in an accompanying JAMA Neurology editorial. This imaging measure, called resting-state functional connectivity MRI, reveals synchronized activity within brain networks. As connections weaken, network activity grows discordant. Despite its promise, researchers still have work to do to improve the technique before it can serve as a biomarker, Jagust noted.

Previous research has shown that connectivity within the DMN is selectively vulnerable in AD. Numerous studies have found that fibrillar amyloid builds up in the DMN early in the disease (see ARF related news story). Moreover, people with a positive brain amyloid scan demonstrate less synchronized activity in this network than do age-matched controls without brain amyloid, hinting that amyloid deposits could damage connections (see ARF related news story; Hedden et al., 2009; Sheline et al., 2010; Mormino et al., 2011). However, these studies did not relate brain connectivity to fluid biomarkers, which are among the most commonly used preclinical markers.

To fill this gap, the WashU group looked at data from 207 cognitively normal older adults with an average age of 70 who were participating in aging and memory studies at the Knight Alzheimer’s Disease Research Center. First author Liang Wang found that people who had low CSF Aβ42 had poorer connectivity between two regions of the DMN—the posterior cingulate cortex (PCC) and the medial temporal lobe (MTL)—compared with people with normal CSF Aβ. The change was modest but significant. Because low CSF Aβ correlates with the presence of brain amyloid deposits, this data agrees with previous imaging findings, and supports the idea that Aβ disrupts neurotransmission.

Unexpectedly, however, high levels of phosphorylated tau in the CSF were independently associated with weak connections between the PCC and the MTL, implying that tau also damages wiring. Conversely, neither age nor brain shrinkage harmed connectivity. The PCC and MTL support episodic memory, and stronger connections between these regions correlate with better performance on memory tasks, Ances noted (see also Sperling et al., 2010; Wang et al., 2010).

The DMN opposes the dorsal attention network (DAN), which turns on when people engage with the outside world by performing a specific task. Normally, the DAN turns off when the DMN activates, and vice versa. In participants with low CSF Aβ42 or high p-tau, however, this switching between networks becomes sloppy, the authors found. In these people, some nodes of the DAN did not completely turn off when the DMN was active. In previous work, the authors showed that distinctions between multiple networks blurred in people with dementia (see ARF related news story).

Researchers do not yet know what pathology underlies such changes in connectivity. Based on autopsy studies, the posterior cingulate becomes clogged with amyloid early in disease, while the medial temporal lobe, which includes the hippocampus and entorhinal cortex, initially has little amyloid but many tau tangles. Why do these connections already fail while the MTL is still relatively free of amyloid? Ances speculated that the combination of amyloid and tau pathology may be to blame. It is also possible that amyloid deposits in the PCC alone cause aberrant electrical activity that disrupts connections to the MTL, or that soluble Aβ oligomers that go undetected play a role in the MTL. Tau tracers, which are under development (see ARF related news story), may help to disentangle the contribution of different pathologic proteins, Ances told Alzforum.

In ongoing work, Ances is collecting longitudinal data from participants to help pin down where changes in resting-state functional connectivity fit on models of disease progression (see ARF related webinar; ARF related news story). In these models, amyloid accumulates first, with tau pathology developing much later. Current data suggests that functional connections break down somewhere between the amyloid and tau pathologies, Ances said.

“[The JAMA paper] drives home the point that different biomarkers can work together to provide an overall picture of network health,” said Jasmeer Chhatwal at Massachusetts General Hospital. Chhatwal and Aaron Schultz are co-first authors of the Neurology paper, but were not involved in the WashU research.

To see if resting-state functional connectivity declines in familial AD, Chhatwal and Schultz examined participants in the Dominantly Inherited Alzheimer Network (DIAN). They compared connectivity in 83 young, cognitively normal AD mutation carriers and 37 non-carriers. Those with an AD gene had weaker connections in several regions of the DMN up to 10 years before the expected age of symptom onset for their family. Connectivity was weakest in those closest to their expected age of onset. Echoing findings from the preclinical sporadic population, this weakening was most evident in the PCC.

The Neurology paper represents the first look at resting-state functional connectivity in DIAN, said Ances, who was a co-author on this paper. The fact that DMN connections fade in both familial and sporadic preclinical disease “stresses that what we are learning from DIAN has applicability to sporadic AD, as well,” Ances told Alzforum. In future work, he plans to directly compare connectivity in people with sporadic and autosomal-dominant AD.

Researchers agree that there are technical hurdles to surmount before resting-state functional connectivity could become a biomarker for preclinical disease. For example, different research groups currently use a variety of methods to analyze resting-state data. To help standardize protocols, Chhatwal and Schultz are developing new analytic techniques that will be easier to apply across different clinics and trials, Chhatwal told Alzforum. Researchers also need to make the technique more reliable and sensitive, and prove that it predicts disease, Jagust noted in his editorial.

Chhatwal argues that the effort will pay off because resting-state functional connectivity MRI has advantages over positron emission tomography (PET) imaging with fluorodeoxyglucose or Aβ ligands. It is cheaper, involves no radiation, and is easy to add to the MRI scans that are already required to monitor safety in clinical trials, Chhatwal said. The upcoming Anti-Amyloid Treatment in Asymptomatic Alzheimer’s Disease (A4) trial led by Sperling will use functional connectivity as a secondary endpoint (see ARF related news story). Researchers believe this biomarker might make a good outcome measure because it changes rapidly, and therefore might reveal a drug effect, Chhatwal said.—Madolyn Bowman Rogers

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Cortical Hubs Found Capped With Amyloid
  2. BOLD New Look—Aβ Linked to Default Network Dysfunction
  3. Communication Breakdown: Multiple Networks Decline in AD Brains
  4. HAI—Sharper Curves: Revamping a Biomarker Staging Model
  5. Solanezumab Selected for Alzheimer’s A4 Prevention Trial

Webinar Citations

  1. Together at Last, Top Five Biomarkers Model Stages of AD

Paper Citations

  1. . Disruption of functional connectivity in clinically normal older adults harboring amyloid burden. J Neurosci. 2009 Oct 7;29(40):12686-94. PubMed.
  2. . Amyloid plaques disrupt resting state default mode network connectivity in cognitively normal elderly. Biol Psychiatry. 2010 Mar 15;67(6):584-7. PubMed.
  3. . Relationships between beta-amyloid and functional connectivity in different components of the default mode network in aging. Cereb Cortex. 2011 Oct;21(10):2399-407. PubMed.
  4. . Functional alterations in memory networks in early Alzheimer's disease. Neuromolecular Med. 2010 Mar;12(1):27-43. PubMed.
  5. . Intrinsic connectivity between the hippocampus and posteromedial cortex predicts memory performance in cognitively intact older individuals. Neuroimage. 2010 Jun;51(2):910-7. PubMed.

Other Citations

  1. ARF related news story

Further Reading

Primary Papers

  1. . Cerebrospinal Fluid Aβ42, Phosphorylated Tau181, and Resting-State Functional Connectivity. JAMA Neurol. 2013 Aug 19; PubMed.
  2. . Biomarkers and Brain Connectivity. JAMA Neurol. 2013 Aug 19; PubMed.
  3. . Impaired default network functional connectivity in autosomal dominant Alzheimer disease. Neurology. 2013 Aug 20;81(8):736-744. PubMed.