Watching the eyes of a person with ALS might help clinicians assess their brain function, according to a paper in the November 11 PLoS One. Researchers from the University of Ulm, Germany, used gaze-tracking technology to classify two stages of ocular deficit in people with the disease. First, conscious control over the line of sight weakens, suggesting problems with executive function that likely reflect pathology in the cortex. By the second stage, involuntary eye movements become erratic, indicating neurodegeneration in the brainstem. “Eye movement may serve as a window into the brain and its pathological alterations in ALS,” said senior author Elmar Pinkhardt. It could even serve as a biomarker for progression, the authors posit.

ALS causes paralysis, but the nerves controlling the eyes traditionally have been thought to be spared. However, scientists have observed that in people who survive for a long time with the disease, eye control is affected (reviewed in Sharma et al., 2011). Eye movements are affected in other diseases too, including Alzheimer’s (Molitor et al., 2015). Pinkhardt and co-author Albert Ludolph wondered if defects in eye movement might correlate with the spread of pathology across the brain. Specifically, they probed how eye-tracking lined up with a staging scheme for TDP-43 proteinopathy proposed in 2013. In this four-step staging, aggregates are thought to appear first in the motor cortex, brainstem, and spinal cord, and then spread to the prefrontal neocortex, precerebellar nuclei, and midbrain by stage 2. Next they creep into the postcentral neocortex and striatum, and finally affect the temporal lobe by stage 4 (see Nov 2013 conference news). However, those stages are only apparent upon autopsy. Eye-tracking, Pinkhardt and Ludolph reasoned, might give them an opportunity to observe the effects of TDP-43 proteinopathy in living patients.

Trailing gaze.

Compared to controls, people with ALS poorly tracked a moving target. [Courtesy of Gorges et al.]

First author Martin Gorges and colleagues recruited 68 people with ALS (none of them with frontotemporal dementia, which can co-occur with ALS), and 31 healthy controls for eye-tracking tests. Subjects faced a curved screen lined with tiny red and green lights. In diverse tests, the researchers asked participants to look at a newly lit dot, delay looking at the new light, or look away from it. The scientists could also project a smoothly traveling red dot onto the screen for subjects to follow (see image at right). In addition, they asked subjects to look back and forth between two lights as quickly as they could for 30 seconds. A camera tracked the movement of each eye, although, since they tended to move together, the authors averaged their movements as if the subjects were monocular. 

The subjects fell into three categories. Thirty people with ALS reacted just like healthy controls. Of the other 38, the authors characterized 25 as oculomotor stage 1, meaning they made more mistakes than controls when asked to delay glancing at a light or look away from it, and reversed gaze fewer times in the 30-second back-and-forth test. These characteristicjos indicate problems with executive control of eye motion, which is managed by brain regions affected in TDP-43 proteinopathy stage 1.

Twelve ALS participants had more severe executive defects, plus other problems. Their eyes moved slowly, and when tracking the moving dot, their gaze tended to lag initially, and then jerk, or saccade, to catch up. That happens when the brainstem or precerebellar networks malfunction. The authors attributed this to TDP-43 proteinopathy stage 2 and labeled it oculomotor stage 2 in ALS.

One subject manifested the brainstem symptoms of oculomotor stage 2 without the executive-function indicators of oculomotor stage 1. Pinkhardt speculated that a comorbidity other than ALS pathology affected eye movement in this person, noting that many conditions can influence eye tracking.

This study marks the first time that researchers have correlated eye movements with brain pathology staging, Pinkhardt said. In addition, Gorges found that that oculomotor stages correlated with worsening performance on the standard ALS functional rating scale, which rates abilities such as speech, swallowing, and walking. The authors wrote that this supports their assertion that eye-movement defects could serve as a marker of disease progression.

However, the link between eye-movement problems and TDP-43 pathological staging is somewhat speculative and inexact, noted co-author Johannes Brettschneider of Herford Hospital in Germany, who came up with the four TDP-43 stages. “The PlosOne paper is valuable because it collects thorough data on oculomotor changes in a large number of ALS cases, but it cannot prove that all these findings are caused by TDP-43 pathology, or that indeed the staging we published in 2013 is correct,” he wrote to Alzforum (see full comment below).

“I think it is a really solid study with a good number of participants,” agreed Malcolm Proudfoot of Oxford University in the United Kingdom, who did not participate in the project. However, he was not entirely sold on the ocular staging scheme. “For the time being, this is still being inferred as the Ulm group didn’t include any longitudinal data, unlike the pathological staging criteria which are based on brains from people who died of ALS ,” he said (see full comment below). The fact that stage 2 oculomotor defects almost always coexist with the stage 1 deficits suggests that stage 2 follows stage 1, he said, but it is not conclusive. In contrast, Proudfoot’s own analyses indicates that eye movements in people with ALS remain stable for up to two years, suggesting they do not progress from one stage to another (Proudfoot et al., 2015). Pinkhardt said longitudinal studies are now in progress.

What do these kinds of eye measurements offer clinicians? Pinkhardt and Proudfoot agreed that eye movements might serve as a biomarker, but Proudfoot suspected they would not make a difference at the diagnosis stage. Nearly half of Pinkhardt’s ALS subjects tracked lights as well as healthy controls did, and those people presumably were in the earliest stages of disease, when a diagnostic test would be most useful. Eye movements also falter in conditions that might be confused with ALS, such as progressive supranuclear palsy, meaning they might not help with differential diagnoses either.

However, Proudfoot thought eye movements could provide a secondary endpoint in ALS clinical trials. The reason is that as patients become paralyzed and lose the ability to speak or write, some rely on their eyes to communicate. Assistive technology can even follow a person’s gaze, allowing them to pick out letters on a screen to compose messages or have an artificial voice articulate on their behalf. Communication is so important to quality of life, Proudfoot noted, that retaining the ability to control eye movements properly might be a good indicator of a drug’s efficacy in late stages. Plus, he noted that developers of eye-tracking communicators may need to consider the gaze defects.

For now, most clinicians cannot track eye movements with any more sophistication than asking a patient to follow their finger, but that could change thanks to smartphones and tablets, predicted Sebastian Crutch of University College London, who did not participate in the study. “It is not unreasonable to anticipate a day when a brief, automated, quantitative assessment of oculomotor functions could become part of standard clinical neurology assessments,” he said (see full comment below).—Amber Dance

Comments

Make a Comment

Comments on this content

  1. The authors of this study analyzed changes in oculomotor function in ALS cases and—based on what we know about the underlying anatomical structures—tried to relate these changes to the distribution of TDP-43 pathology. This relation is, of course, somewhat speculative, but many studies in the field try to establish a connection between clinical phenotypes and findings in neuropathology. The relation between the oculomotor changes reported here and the distribution of TDP-43 as described in our Annals paper is rather loose and not very exact.

    The PlosOne paper is valuable because it collects thorough data on oculomotor changes in a large number of ALS cases, but it cannot prove that all these findings are caused by TDP-43 pathology, or that indeed the staging we published in 2013 is correct.

  2. Although relatively early days (as the authors note, they await confirmation in larger samples and with longitudinal follow-up), this study valuably attempts to go beyond merely detecting a clinical deficit or correlating a clinical deficit with a particular neuroimaging finding or other biomarker, and instead tries to stage the level of dysfunction currently being exhibited by an individual.

    Eye-tracking is not yet a routinely available clinical tool, but with the proliferation of such technologies in smartphones and tablets, it is not unreasonable to anticipate a day when a brief, automated, quantitative assessment of oculomotor functions could become part of standard clinical neurology assessments. 

    More generally, computational, mathematical and engineering techniques are now routinely used to evaluate complex magnetic resonance imaging (MRI) datasets describing the structure of the brain in people with dementia. By contrast, we have barely examined whether similar techniques and technology can be harnessed to address highly complex clinical and cognitive datasets, such as those elicited by quantitative assessment of oculomotor function. 

  3. I think this is a really solid study by the Ulm group with a good number of participants. There are lots of overlapping findings with our study, which is great because it indicates the results are reproducible. For example, we also found that anti-saccade error rates correlated with falling ALSFRSr (Proudfoot et al., 2015). Here, Gorges and colleagues used a more sensitive measure of cognitive disturbance (the ECAS, as opposed to the ACE), which was very sensible, allowing them to find correlations between cognition and eye-tracking measures that were much less robust in our data.

    A lot of the discussion in their paper focuses on the utility of eye-tracking measures as a staging tool. They hope this can provide some clinical translation of their previous pathological staging criteria. The trouble is that our longitudinal data showed stability of eye-tracking measures rather than progressive decline. Having said that, they looked at a different range of measures, and it’s possible that some more basic eye movements (smooth pursuit and max peak velocity) may get worse as the disease progresses. For the time being, this is still being inferred as the Ulm group didn’t include any longitudinal data, unlike the pathological staging criteria which are based on brains from people who died of ALS (by definition, “end-stage”).

    I think this work is really important considering the ongoing development of eye-tracking devices for communication, which may need to take account of oculomotor deficits as the disease progresses. I don’t think we have enough data yet on what proportion of people with advanced ALS can actually make good use of these devices—it’s a try-it-and-see approach on an individual basis, which is of course perfectly reasonable.

    A future consideration is of course what the diagnostic value of these measures are (given that many other conditions also affect eye movements) and what their prognostic value turns out to be. We don’t yet have the data to assess its diagnostic potential fully, but my hunch is that it won’t be super-useful. You can see in the Ulm paper that a big portion of people with ALS do just fine on the tests, and you can expect that they represent the patients with the lowest overall burden of disease—just the people in whom there might be some diagnostic doubt. The other group with diagnostic doubt might be those patients in whom you know something is definitely wrong, but can’t exactly classify the neurological problem. Given that eye-movement dysfunction is a very prominent feature of, say, PSP, again I’m not sure it will help. It might be that functional neuroimaging using an eye-tracking task could distinguish the anatomical (and hence pathological) cause of an individual’s oculomotor dysfunction to provide diagnostic support. For example, there was a recent great paper by the Munoz group that evidenced dysfunction in the DLPFC—an expected culprit (Witiuk et al., 2014). 

    But yes, relative to “follow my finger,” I’m sure these assessments are more sensitive, and with much less inter-user variability that permits repeated longitudinal study. I don’t think it is novel “bedside” information, but it is way more quantitative. My hope is that early signs of dysfunction will be predictive of subsequent deterioration in the oculomotor domain, whereas patients with normal eye movements at diagnosis will remain normal. As communication is so important, you could easily envisage this being a secondary end point in a clinical trial. We came to a pragmatic conclusion, based on our data, that these tools could also support cognitive assessments in very disabled ALS patients, but you would first need to establish reliable normative ranges.

    The main findings of the paper are also the most contentious, that of classification. I think the evidence that executive impairments always coexist with catch-up saccades is suggestive, but that longitudinal evidence would be definitive. Further investigation of smooth pursuit in ALS looks to be worthwhile.

    References:

    . Eye-tracking in amyotrophic lateral sclerosis: A longitudinal study of saccadic and cognitive tasks. Amyotroph Lateral Scler Frontotemporal Degener. 2015 Jan-Feb;17(1-2):101-11. Epub 2015 Aug 27 PubMed.

    . Cognitive deterioration and functional compensation in ALS measured with fMRI using an inhibitory task. J Neurosci. 2014 Oct 22;34(43):14260-71. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. The Four Stages of TDP-43 Proteinopathy

Paper Citations

  1. . Oculomotor dysfunction in amyotrophic lateral sclerosis: a comprehensive review. Arch Neurol. 2011 Jul;68(7):857-61. PubMed.
  2. . Eye movements in Alzheimer's disease. J Alzheimers Dis. 2015 Jan 1;44(1):1-12. PubMed.
  3. . Eye-tracking in amyotrophic lateral sclerosis: A longitudinal study of saccadic and cognitive tasks. Amyotroph Lateral Scler Frontotemporal Degener. 2015 Jan-Feb;17(1-2):101-11. Epub 2015 Aug 27 PubMed.

Further Reading

Papers

  1. . Cognitive deterioration and functional compensation in ALS measured with fMRI using an inhibitory task. J Neurosci. 2014 Oct 22;34(43):14260-71. PubMed.
  2. . Early saccades in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Frontotemporal Degener. 2013 May;14(4):294-301. PubMed.
  3. . Eye movements in amyotrophic lateral sclerosis and its mimics: a review with illustrative cases. J Neurol Neurosurg Psychiatry. 2011 Jan;82(1):110-6. Epub 2010 Nov 19 PubMed.
  4. . Diffusion tensor imaging analysis of sequential spreading of disease in amyotrophic lateral sclerosis confirms patterns of TDP-43 pathology. Brain. 2014 Jun;137(Pt 6):1733-40. Epub 2014 Apr 15 PubMed.
  5. . Oculo-Visual Dysfunction in Parkinson's Disease. J Parkinsons Dis. 2015;5(4):715-26. PubMed.
  6. . Patients with mild Alzheimer's disease produced shorter outgoing saccades when reading sentences. Psychiatry Res. 2015 Sep 30;229(1-2):470-8. Epub 2015 Jun 27 PubMed.
  7. . Abnormalities of fixation, saccade and pursuit in posterior cortical atrophy. Brain. 2015 Jul;138(Pt 7):1976-91. Epub 2015 Apr 19 PubMed.
  8. . Diagnosis of mild Alzheimer disease through the analysis of eye movements during reading. J Integr Neurosci. 2015 Mar;14(1):121-33. PubMed.
  9. . The disengagement of visual attention in Alzheimer's disease: a longitudinal eye-tracking study. Front Aging Neurosci. 2015;7:118. Epub 2015 Jun 23 PubMed.
  10. . Saccade deficits in amnestic mild cognitive impairment resemble mild Alzheimer's disease. Eur J Neurosci. 2014 Jun;39(11):2000-13. PubMed.

Primary Papers

  1. . Eye Movement Deficits Are Consistent with a Staging Model of pTDP-43 Pathology in Amyotrophic Lateral Sclerosis. PLoS One. 2015;10(11):e0142546. Epub 2015 Nov 11 PubMed.