Unstructured Eye Tracking as a Diagnostic and Prognostic Biomarker in Parkinsonian

New diagnostic and prognostic biomarkers of neurological disease are consistently sought after. In particular, identifying biomarkers which can diagnose a condition at an early or prodromal stage would identify a cohort of patients most likely to benefit from disease modifying treatments. Similarly, biomarkers which might serve as a surrogate of disease progression over time will be essential to monitor these patients' response to those treatments.

Parkinson's disease (PD) is one progressive neurodegenerative condition for which many disease-modifying trials are currently in development. PD remains an entirely clinical diagnosis and no diagnostic tests for the condition currently exist. It must be distinguished from clinically similar conditions with different pathological substrates, namely the atypical parkinsonian disorders such as progressive supranuclear palsy (PSP), corticobasal syndrome (CBS), multiple system atrophy (MSA) and dementia with Lewy bodies (DLB) as each substrate will require tailored treatments. Differentiating PD from atypical parkinsonism can be challenging, particularly early in the disease course. The diagnostic accuracy among those with autopsy proven PD early in the disease course is as low as 26% and although this accuracy rises with disease duration, the pre-mortem diagnosis remains incorrect up to 15-20% of the time.

Much effort in recent years has focused on the development of molecular diagnostic biomarkers.However, whether these biomarkers can explain the heterogeneity we see among people with PD (PwP) and other parkinsonian disorders, and whether they are useful markers of disease progression remains unclear. Either way, clinically relevant biomarkers which can differentiate parkinsonian disorders at an early stage, predict disease trajectory and track clinical progression over time will clearly be required. The clinical measures most commonly used to track disease progression in clinical trials are rating scales such as the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS). There are however concerns that these scales may not provide the precision necessary to detect small clinical changes, particularly in early PD. Before one can assess whether a clinical biomarker is useful longitudinally, their accuracy in a cross-sectional cohort must be assessed and, in particular, whether they are sensitive to the heterogeneity seen among the degenerative parkinsonian disorders. In addition, one must assess the effect, if any that symptomatic medications such as levodopa have on the performance of that biomarker. Quantitative assessment of eye movements may be a simple, reliable and sensitive clinical biomarker for PD and atypical parkinsonian disorders such as progressive supranuclear palsy (PSP), multiple system atrophy (MSA) and corticobasal syndrome (CBS) Indeed, eye movements have long been used at the bedside to localize pathology and to diagnose many movement disorders. Although clinically apparent oculomotor abnormalities are well documented in the advanced stages of these conditions, they may not be appreciable to the naked eye early in the disease course. The diagnosis of PSP, for example, relies heavily on the clinical assessment of eye movements - in particular the presence of limited vertical gaze. Although this is the classical eye movement abnormality seen in PSP, in the early stages of the disease, the only eye movement clue may be subtle slowing of vertical saccades or curvature of the arc of the saccadic eye movement. In a postmortem study of parkinsonian disorders, a vertical supranuclear gaze palsy, when present, has a specificity of 91% for pathologically definite PSP, increasing to 97% when present within 3 years after disease onset. However, its sensitivity in heralding a PSP diagnosis is only 31% when detected within the first 3 years of disease.

To date, the application of quantitative oculomotor analysis to movement disorders has been limited, largely as a result of technological restrictions. As such, any information gleaned from eye movements in clinical practice is largely restricted to what can be detected by the naked eye at bedside assessment. Since quantitative assessment of eye movements will be more sensitive to these oculomotor abnormalities than bedside assessment with the naked eye, detection of these abnormalities will be possible at an earlier stage. Advances in technology have allowed eye tracking to become a simple, non-invasive, and effective way to assess eye movements, pupillary function and blink rate with high resolution. Quantitative oculomotor parameters (saccades and pupillary changes) correlate with motor performance, cognitive function and postural abnormalities in PwP. Given that these are crucial milestones in progression of PD, the investigators hypothesize that early assessment of subclinical oculomotor abnormalities will have prognostic value in PD and related conditions. However, only a few small studies have investigated whether quantitative assessment of eye movements can be used to differentiate PD from similar conditions such as PSP.

Recently, pupillary dilatation has been shown to be a potential marker of goal-directed eye movement (specifically motor preparation when planning saccades), but these have not been explored in detail in PD. Given that PwP lose automaticity of movement in the early stages and can compensate with goal-directed attentional strategies, it is likely that early compensatory differences in these processes can be detected.