Chronic traumatic encephalopathy can only be formally diagnosed by examining the brain after death, but researchers are inching closer to spotting the neuropathological hallmarks of disease during life. A study published April 10 in the New England Journal of Medicine described the results of PET scans in 26 former National Football League players with mild cognitive symptoms. Researchers led by Robert Stern at Boston University School of Medicine and Eric Reiman of the Banner Alzheimer’s Institute in Phoenix reported that compared with controls, the players had a higher average uptake of flortaucipir, a tau tracer, in three regions of the brain typically affected in people with CTE. However, the players’ flortaucipir uptake was modest compared with that observed in people with AD, and substantial overlap existed between cases and controls. Only one of the former players had evidence of Aβ accumulation.

  • Former NFL players with cognitive complaints had higher flortaucipir uptake in three brain regions than controls.
  • Tracer uptake did not correlate with cognitive or neuropsychological test scores.
  • Years of playing football did correlate with uptake.

“This is a welcome and carefully written paper,” commented Christopher Rowe of the University of Melbourne in Australia. “It is the first good evidence that tau imaging can detect a signal in chronic traumatic encephalopathy, but the authors make very clear that this is a long way off from being a useful diagnostic test for individuals suspected of having the disease.”

To investigate whether PET scans could detect tau pathology in people with a history of repeated head trauma, the researchers scanned 26 former NFL players, aged 40–69, who reported cognitive, mood, or other behavioral symptoms. The retired athletes had played in the NFL for at least two years, and had played tackle football for a minimum of 12 years. The investigators also scanned 31 controls without a history of repeated head injury. Compared with controls, former players had modestly elevated tau tracer standardized uptake value ratios (SUVRs) in the bilateral superior frontal, bilateral medial frontal, and left parietal cortex—all regions previously identified as tau pathology hot spots in people with CTE. Only one former player had evidence of Aβ accumulation, as measured separately using florbetapir-PET scans. 

Though former players had a higher average tau tracer uptake than controls, individual measurements overlapped extensively between the groups, indicating that this form of tau-PET imaging could not be used for individual diagnosis.

The participants took a series of cognitive and neuropsychological tests, on which many players scored in the impaired range. Notably, 81 percent were considered depressed, and 35 percent had deficits in delayed recall. However, the extent of tau tracer uptake in any of the three affected brain regions did not correlate with test scores. In a post hoc analysis, the researchers found that tracer uptake correlated with the former players’ total number of years playing tackle football.

Reiman told Alzforum he was disappointed that flortaucipir uptake could not distinguish between former players and controls at the individual level. He and other commentators offered several potential explanations for the weak tau signals. For one, it is possible that in response to repeated head injuries, tau accumulates in a highly localized pattern, for example around the deep sulci as they warp under impact. This spotty pattern would preclude a strong tau tracer signal, Reiman said. Bill Jagust of the University of California, Berkeley, added that patchy, inconsistent deposition patterns make tau difficult to quantify, especially when averaging across subjects.

Another possibility is that flortaucipir simply does not bind CTE-tau filaments as strongly as it binds AD-tau, Reiman and Jagust said. Tau filaments in AD and CTE both consist of three-repeat and four-repeat isoforms, but recent structural studies of the two types of filament visualized distinct core structures (Jul 2017 news; Mar 2019 news). Michel Goedert, who co-led both structural studies with Sjors Scheres at the MRC Laboratory of Molecular Biology in Cambridge, U.K., said that future work will identify the binding sites of flortaucipir to AD-tau and CTE-tau filaments, which could clarify this issue.

“It will be interesting to see how second-generation tau tracers, such as MK6240, perform in this population,” wrote Gil Rabinovici at the University of California, San Francisco. “Overall, this [study] is an encouraging first step, but I agree with the authors that tau PET is not yet ready for ‘prime time’ as a diagnostic agent in CTE,” he added.

What to make of the finding that flortaucipir-PET did not correlate with cognitive or neuropsychological test scores? Reiman said he was unsurprised by that, given the small number of participants and their mild symptoms. Rowe raised the possibility that some of these former players do not have CTE at all. “This is highly likely given the nonspecific nature of the mood, cognition, and behavioral features used for clinical diagnosis of CTE,” he wrote.

Reiman thinks larger studies, of people with a wide range of symptom severity, might detect a connection between tau pathology and clinical symptoms. To that end, he is collaborating with Stern and other investigators to recruit participants to the DIAGNOSE-CTE study (see clinicaltrials.gov). This longitudinal study will enroll 120 former NFL players, 60 former college football players, and 60 controls. It will track tau, Aβ, and neurodegeneration via various neuroimaging techniques, search for potential biomarkers in cerebrospinal fluid and blood, and monitor cognition and behavior.

Kaj Blennow of the University of Gothenburg in Sweden stressed the importance of such multimodal, longitudinal studies. “It will be interesting to learn whether regional cortical tau deposition also is associated with biomarker evidence of neurodegeneration in the same areas (evaluated by volumetric MRI), or by increases in fluid biomarkers reflecting neurodegeneration, such as plasma levels of tau or neurofilament light (NFL) protein, especially in longitudinal studies employing such measures.”—Jessica Shugart

Comments

  1. Developing in vivo biomarkers of CTE is absolutely critical for efforts to understand basic questions about disease prevalence, mechanisms, susceptibility and resilience factors, and ultimately for drug development. This paper by Stern and colleagues represents a significant advance by providing proof of concept that detecting tau pathology in CTE may be possible using PET imaging.

    However, the overall degree of flortaucipir binding was modest, solidly in the range of controls in some at-risk former NFL players, and even in the “high binders” their SUVR values were lower than we typically see in symptomatic patients with AD. It is not clear whether the overall modest binding is due to absent or low tau burden in the players, or to a lack of sensitivity of this tracer to detect the tau lesions in CTE. We know from the recent cryo-EM study (Falcon et al. 2019) and other work that tau lesions in CTE are microstructurally distinct from neurofibrillary tangles in AD, and this certainly may affect the affinity of a PET tracer to bind to the aggregates.

    Another concern in interpreting the data is the lack of specificity of flortaucipir, which we know binds with low affinity to targets other than tau. It will be interesting to see how second-generation tau tracers, such as MK6240, perform in this population. Overall, this is an encouraging first step, but I agree with the authors that tau PET is not yet ready for “prime time” as a diagnostic agent in CTE.

    References:

    . Novel tau filament fold in chronic traumatic encephalopathy encloses hydrophobic molecules. Nature. 2019 Apr;568(7752):420-423. Epub 2019 Mar 20 PubMed.

  2. This is a welcome and carefully written paper. It is the first good evidence that tau imaging can detect a signal in chronic traumatic encephalopathy, but the authors make very clear that this is a long way off from being a useful diagnostic test for individuals suspected of having the disease. The areas of tau tracer uptake in the group with clinically suspected CTE in the dorsal prefrontal and medial temporal cortex match well with the areas of highest tau density in stage 3 or greater stages of CTE as described at postmortem by McKee et al., 2013.

    The results are therefore encouraging and justify support for independent replication studies. Postmortem confirmation is needed to assess if the large overlap with controls and lack of correlation with cognition or other clinical features is due to absence of CTE in some of the symptomatic patients. This is highly likely given the nonspecific nature of the mood, cognition and behavioural features used for clinical diagnosis of CTE. Alternatively, a better tracer with higher affinity for tau and lower nonspecific background may be needed before tau PET can be used to assist the diagnosis of this condition.  

    References:

    . The spectrum of disease in chronic traumatic encephalopathy. Brain. 2013 Jan;136(Pt 1):43-64. PubMed.

  3. This is a very interesting and important study. For the first time, it is convincingly shown (at the group level) that former National Fotboll League players, with cognitive and behavioral symptoms, have increased tau PET ligand retention in several cortical brain areas, know to be affected by tau pathology in postmortem CTE case series.

    These data are of particular interest, since they support that the acute damage to brain cell structures (e.g., neuronal axons, glial cells, or capillaries) induced by repeated concussions initiates pathophysiological processes driving tau aggregation in the cerebral cortex. While no associations were found with neuropsychological test scores, flortaucipir SUVRs correlated with total years of participation in football, supporting a relation with repeated head impacts.

    It will be interesting to learn whether regional cortical tau deposition also is associated with biomarker evidence of neurodegeneration in the same areas (evaluated by volumetric MRI), or by increases in fluid biomarkers reflecting neurodegeneration, such as plasma levels of tau or neurofilament light (NFL) protein, especially in longitudinal studies employing such measures.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Tau Filaments from the Alzheimer’s Brain Revealed at Atomic Resolution
  2. Traumatic Tau: Filaments from CTE Share Distinct Structure

External Citations

  1. clinicaltrials.gov

Further Reading

Primary Papers

  1. . Tau Positron-Emission Tomography in Former National Football League Players. N Engl J Med. 2019 May 2;380(18):1716-1725. Epub 2019 Apr 10 PubMed.