Retinal Architecture and Melanopsin-Mediated Pupillary Response Characteristics: A Putative Pathophysiologic Signature for the Retino-Hypothalamic Tract in Multiple Sclerosis

Key Points

Question  What is the association between the melanopsin-mediated sustained pupillary constriction response with retinal architecture in multiple sclerosis?

Findings  In a case-control study of 24 patients with multiple sclerosis, ganglion cell layer and inner plexiform layer thinning determined by optical coherence tomography corresponded to a significant attenuation of the melanopsin-mediated sustained pupillary constriction response.

Meaning  Multiple sclerosis damage to retinal architecture indicates attenuation of the melanopsin-mediated sustained pupillary constriction response, with potential relevance for retinohypothalamic modulation of homeostatic mechanisms in health and in illness.

Importance  A neurophysiologic signature of the melanopsin-mediated persistent constriction phase of the pupillary light reflex may represent a surrogate biomarker for the integrity of the retinohypothalamic tract, with potential utility for investigating alterations in homeostatic mechanisms associated with brain disorders and implications for identifying new treatments.

Objective  To characterize abnormalities of retinal architecture in patients with multiple sclerosis (MS) and corresponding alterations in the melanopsin-mediated sustained pupillary constriction response.

Design, Setting, and Participants  The case-control study was an experimental assessment of various stimulus-induced pupillary response characteristics and was conducted at a university clinical center for MS from September 6, 2012, to February 2015. Twenty-four patients with MS (48 eyes) and 15 individuals serving as controls (30 eyes) participated. The melanopsin-mediated, sustained pupillary constriction phase response following cessation of a blue light stimulus was compared with the photoreceptor-mediated pupillary constriction phase response following cessation of a red light stimulus. Optical coherence tomography was used to characterize the association between pupillary response characteristics and alterations in retinal architecture, specifically, the thickness of the retinal ganglion cell layer and inner plexiform layer (GCL + IPL).

Main Outcomes and Measures  Association of pupillary response characteristics with alterations in retinal architecture.

Results  Of 24 patients with MS included in the analysis, 17 were women (71%); mean (SD) age was 47 (11) years. Compared with eyes from individuals with MS who had normal optical coherence tomography–derived measures of retinal GCL + IPL thickness, eyes of patients who had GCL + IPL thickness reductions to less than the first percentile exhibited a correspondingly significant attenuation of the melanopsin-mediated sustained pupillary response (mean [SD] pupillary diameter ratios at a point in time, 0.18 [0.1] vs 0.33 [0.09]; P < .001, generalized estimating equation models accounting for age and within-patient intereye correlations).

Conclusions and Relevance  In this case-control study, attenuation of the melanopsin-mediated sustained pupillary constriction response was significantly associated with thinning of the GCL + IPL sector of the retina in the eyes of patients with MS, particularly those with a history of acute optic neuritis. Melanopsin-containing ganglion cells in the retina represent, at least in part, the composition of the retinohypothalamic tract. As such, our findings may signify the ability to elucidate a putative surrogate neurophysiologic signature that correlates with a constellation of homeostatic mechanisms in both health and illness.

Progress over the past decade to model disease pathogenesis within the visual system has led to the proposed specific aim that the “eye can serve as a window” from which we can elucidate pathobiological underpinnings of neurodegenerative disease that are correspondingly amenable to the identification and monitoring of preventive, protective, performance-enhancing, and restorative properties of putative neurotherapeutic agents.^1-10 Understanding how retinal damage in multiple sclerosis (MS) may result in alterations in the melanopsin-mediated sustained pupillary constriction response may contribute to the identification of pathophysiologic signatures linking a simple and reproducible reflex with objective changes noted in homeostatic networks (eg, sleep-wake cycle transitions, regulation of mood and energy states, thermoregulation, cognition, eating and satiety, sexual and mating behavior, and neuroendocrine reflex arcs).^11-23 Furthermore, evidence is mounting that the hypothalamus participates in the coordination of immune regulatory and surveillance networks.²⁴

Light is well known to exert regulatory influences over several central nervous system–mediated states of homeostasis.^25,26 A decrease in light transmission from the retina to the suprachiasmatic nucleus (SCN) may result from retinal or optic nerve injury in MS. Such reductions in transmission of light may, in turn, produce dysfunction across a range of hypothalamic regulatory switch-points.^25-28

Animal studies have demonstrated that the retinomesencephalic and retinohypothalamic pathways are almost exclusively driven by overlapping subtypes of melanopsin-expressing intrinsic photosensitive retinal ganglion cells (ipRGCs) (Figure 1).²⁷ Carefully constructed genetic knockout studies that specifically target the ipRGCs have determined that circadian rhythm dysfunction is associated with a decrease in the melanopsin-mediated sustained pupillary constriction phase of the light reflex.^11,26-29 In human studies, advanced glaucoma (which, like MS, leads to both axonal and retinal ganglion cell neuronal degeneration) has been associated with a decrease in retinohypothalamic suppression of melatonin by light, with a corresponding reduction in sleep quality in conjunction with daytime sleepiness and fatigue.^29,30 Several studies have described the complexity of sensitively and specifically differentiating melanopsin from rod- and cone-mediated pupillary light reflexes.^31-35

To initiate a series of stepwise investigations aimed at elucidating the physiologic consequences of damage to retinal architecture in MS, we used a novel wavelength-differentiated pupillary light reflex protocol (Figure 2).^12,14,36 Our purpose then was to first test the hypothesis that the melanopsin-mediated sustained pupillary constriction response would be attenuated or even abolished in the context of damage to retinal architecture objectively measured by using high-speed, high-definition spectral domain optical coherence tomography (OCT).

The study was conducted from September 6, 2012, to February 2015. We performed infrared-based pupillometry on 39 study participants, including 15 individuals serving as disease-free controls (30 eyes) and 24 patients with MS (48 eyes) (Table 1). Patient eyes were classified further according to history of acute optic neuritis (AON). This categorization was independent of the Kurtzke Expanded Disability Status Scale³⁷ or disease duration. All patients with MS fulfilled the revised McDonald diagnostic criteria.³⁸ Participants were recruited from the MS Neuro-Ophthalmology Research Laboratory in the Clinical Center for Multiple Sclerosis at the University of Texas Southwestern Medical Center at Dallas. Individuals were excluded if they had other neurologic or ophthalmologic conditions (including high myopia, defined as a correction of ≤−5 diopters of spherical measurement). All participants had previously undergone extensive neuro-ophthalmologic investigations and were well characterized (Table 1). The study protocol was approved by the University of Texas Southwestern Medical School Institutional Review Board in accordance with the Declaration of Helsinki.³⁹ The participants provided both written and oral consent and received a stipend.

Pupillometry was performed (DP-2000 device; Neuroptics Inc) using stimulus protocols as previously described.^12-14,36 Testing was conducted under uniform dark conditions following 10 minutes of dark adaptation. Participants were instructed to abstain from potential pupil-modulating substances (eg, alcohol, caffeine, and anticholinergic medications) for at least 24 hours before testing. Furthermore, each participant was interviewed before the inception of the investigations to confirm that these specific instructions were followed.

Investigations exploring the relative afferent pupillary defect typically aim to characterize the interocular differences in the transmission properties along the 2 afferent visual pathways. However, our goal was to better understand the association of damage to the retina as evidenced by ganglion cell layer and inner plexiform layer (GCL + IPL) thinning due to MS with attention of the melanopsin-mediated pupillary light reflexes (most particularly the persistent constriction phase of the response), similar to the rod- and cone-mediated responses, and the correlation of such derangements with discrete pathologic changes in retinal architecture (eg, thinning of layers that are germane for the high-fidelity transmission characteristics in response to light activation).¹⁴

Pupillometry was performed under conditions of varying stimulus intensities lasting 1 second with both red (622 nm) and blue (463 nm) stimuli administered alternatively in sequence to each eye. Park and colleagues³⁶ used high-intensity isoluminant red and blue stimuli with luminance values of approximately 2.6 candelas/m². We used a maximum stimulus intensity of 2.6 log lux to produce a maximum melanopsin-mediated sustained pupillary response (ie, defined as the most protracted duration of the constriction phase of the reflex) within the stimulus constraints of the instrument.

Stimulus intensity for the red and blue lights was matched for the photopic spectral sensitivity of cones, and pupillary diameter was captured at a sampling rate at 30 Hz. Retinal layer thicknesses were assessed by the use of high-speed, high-definition spectral domain OCT (Cirrus 4000 HD-OCT; Carl Zeiss Meditec). The instrument’s automated retinal segmentation analysis program was used to investigate the relationship between melanopsin-mediated sustained constriction of the pupillary light reflexes and corresponding changes in validated measures of retinal architecture (eg, mean thickness of the peripapillary retinal nerve fiber layer and mean thickness of the GCL + IPL).

The isolated melanopsin response was quantified by measuring the normalized pupil diameter 6 seconds after the red light stimulus and subtracting the normalized pupil diameter over a 1-second period that was centered at 6 seconds after a corresponding photopically matched-intensity blue light stimulus. The following paradigm is an adaptation from Park et al.³⁶

where RM_mel is the quantified activity of the melanopsin-driven retinomesencephalic tract, d_R6 is the pupil diameter during the red stimulus at 6 seconds, d_R0 is the pupil diameter immediately before the red stimulus, d_B6 is the pupil diameter during the blue stimulus at 6 seconds, and d_B0 is the pupil diameter immediately before the blue stimulus. As well as

where P_TR is the transient pupillary response and d_Bmin is the absolute minimal pupil diameter during the blue stimulus.

The means of the right and left pupillary responses were determined (an approach used to inform the detection of afferent pupillary defects without influence of any efferent response contribution), and the analysis was performed in a fashion that was blinded to AON status and patient-specific OCT data.

For the statistical analyses, we used generalized estimating equation (GEE) models to examine the capacity for the history of AON and GCL+IPL thickness above vs below the first percentile to determine mean melanopsin response at 2.6 log lux among patients with MS, accounting for age and adjusting for within-patient intereye correlations.

We grouped the eyes of patients with MS into defined cohorts contingent on specifically defined, OCT-measured thicknesses. Specifically, we dichotomized the eyes into objective designations as follows: (1) an affected eye from a patient with MS was defined by a retinal layer thickness threshold, where GCL + IPL thickness was less than the first percentile of that predicted on the basis of our normative database information; (2) alternatively, an unaffected eye from a patient with MS was defined in terms of a GCL + IPL thickness at the 5th to 95th percentile of that predicted based on the same normative database; and (3) we further dichotomized the data for individual eyes categorically grouped by a confirmed history of AON or no AON.

We also examined these categories and their capacity to predict mean melanopsin response among patients with intereye asymmetry based on GCL + IPL thickness abnormality or AON eye status. The correlation of OCT measures to mean melanopsin response was also determined using GEE models. All data were analyzed using Stata, version 13.0 (StataCorp).

Patient and control participant characteristics are described in Table 1. There was no significant difference in sex ratio between disease-free controls and patients with MS (67% vs 71%). All 24 patients with MS were characterized as having relapsing-remitting MS. The mean disease duration in the MS cohort with a history of AON could not be differentiated from the cohort with no history of AON (approximtely 10 years).

The melanopsin-mediated sustained pupillary blue light response with the absolute values for GCL + IPL thickness were compared with all disease-free control and all MS eyes, regardless of their AON history. When accounting for age and adjusting for within-patient intereye correlations, we confirmed that thinning of the GCL + IPL was associated with a significantly diminished melanopsin-mediated sustained pupillary constriction in response to a blue light stimulus (P < .001 for light-intensity stimuli of 2.6 log lux) (Figure 2 and Figure 3).

All 30 of the control participants’ eyes exhibited normal response characteristics to both red and blue light stimuli compared with the reduced blue light stimulus–induced constriction response in eyes of patients with MS who had either a history of AON and/or a reduced GCL + IPL thickness (Figure 3 and Video 1). We demonstrated that history of AON among the eyes of patients with MS was a significant indicator of diminished duration for the melanopsin-sustained pupillary constriction response to blue light (mean [SD] pupil diameter ratios at a point in time, 0.19 [0.14] for eyes of patients with MS who had a history of AON vs 0.30 [0.09] for eyes of patients with MS with no such history of AON; P < .001, GEE models) (Table 2 and Video 2).

The eyes of patients with MS were grouped by GCL + IPL thickness thresholds. This approach yielded a highly discriminative capability with respect to the relationship between structure and function (our principal hypothesis). Specifically, reductions in GCL + IPL thickness to less than the first percentile were associated with significant attenuation of the melanopsin-sustained pupillary constriction response in contrast to those with thickness between the 5th and 95th percentiles (0.33 [0.09] for eyes with GCL + IPL thickness at the 5th-95th percentiles vs 0.18 [0.1] for eyes with thickness <1st percentile; P < .001, GEE models) (Table 2).

In analysis of only patients with intereye asymmetry with regard to AON history and GCL + IPL thickness status (5th-95th percentile vs <1st percentile), the unaffected eyes (no AON history or GCL + IPL thickness at the 5th-95th percentile) revealed significant preservation of the melanopsin response in the fellow eyes (melanopsin response of eyes with no history of AON, 0.32 [0.07] vs 0.19 [0.12] for the affected eyes with a history of acute AON; P = .002, GEE models) (Table 2). In patients with MS exhibiting objective intereye asymmetry in GCL + IPL thickness, there was a significantly diminished melanopsin-mediated sustained constriction response (0.17 [0.11]) for those with GCL + IPL thickness less than the first percentile vs those with thickness between the 5th and 95th percentiles (0.36 [0.09]; P < .001, GEE models) (Table 2).

The objective of this investigation was to assess the relationship between validated measures of retinal architecture (eg, GCL + IPL) and the blue light–induced sustained pupillary constriction responses, specifically in people with MS who either had or did not have a history of AON. The blue light–induced response represents a potentially important neurophysiologic signature for the integrity of ipRGCs that actively express melanopsin and exhibit characteristic physiologic responses to light stimuli restricted to the blue spectrum (ie, between 400 and 500 nm).

Previous studies had demonstrated correlations between retinal layer thinning and declines in several measures, including patient-reported low-contrast letter acuity, Humphrey visual fields, timing and amplitude responses of variable contrast, pattern reversal, multifocal visual-evoked potential stimuli, and the optic nerve head potential response characteristics of the multifocal electroretinogram.^40-42 Ultimately, we confirmed our principal hypothesis that, in eyes of patients with MS who have alterations of retinal architecture, the magnitude of the melanopsin-mediated sustained constriction phase of the blue light–induced pupillary response was significantly reduced. This observation is consistent with attenuation in the function of the retinomesencephalic tract (the localization for the integration of the afferent and efferent limbs of the pupillary light reflex).

A small population of ipRGCs (estimated to represent approximately 2%-3% of all ipRGCs) projects into the pretectal olivary nucleus for the coordination of the melanopsin-mediated pupillary light response, while some of the ipRGCs instead project into the SCN (the mammalian pacemaker) of the hypothalamus (via the retinohypothalamic tract).^11,26 This shared origin of the retinomesencephalic and retinohypothalamic tracts raises the possibility that the structural alterations of the retina in patients with MS may correlate not only to the diminished function of the retinomesencephalic tract (as evidenced by a decrease in the melanopsin-mediated persistent constriction response) but also to diminished function of the retinohypothalamic tract and concomitant alterations in a diversity of key states in body homeostasis.^25,26,29,30

The successful development of a surrogate signature for retinohypothalamic tract function could help to elucidate the pathobiological bases of some of the most debilitating symptoms faced by patients with MS as well as for those with other neurodegenerative disorders. For example, regarding the disabling symptom of fatigue in MS, a greater understanding of how retinal damage may compromise the high-precision regulation of how light input into the SCN serves to modulate the elaboration of pineal melatonin across the circadian cycle could be germane to the development of innovative treatment strategies. For instance, a patient with MS exhibiting an attenuated or even abolished persistent constriction response to a global blue light stimulus (eg, stimulation with 400-500 nm of global-spectrum blue light) may exhibit reconstituted, persistent constriction responses when the eyes are stimulated with highly discrete wavelength intervals of blue light stimulation (eg, test stimuli every 10 nm from 400 to 500 nm), with the purpose of identifying stimulus characteristics where activation of ipRGCs can be enhanced with greater stimulus specificity while using the persistent constriction phase response of the pupillary light reflex as a confirmatory neurophysiologic signature for the integrity of these cells and as a potential biomarker for the retinohypothalamic tract (a supposition we are currently investigating).

To gain insight into how much MS-related reduction in optic nerve input can affect circadian rhythm, future investigations would involve studying how well a standard melanopsin-mediated input to the hypothalamus is capable of suppressing the rise in melatonin that occurs late in the day and triggers sleepiness (Figure 2B). This blue light–mediated function would provide a means for evaluating the association between the melanopsin-mediated pupillary response and sleep disorders in patients with MS.

Neither visual perception nor OCT thickness measures would constitute measures as specific as the blue light–mediated pupillary reflex for the ultimate assessment of retinohypothalamic function in health and disease states. In fact, recent work has demonstrated a correspondence between circadian disruption and pathways that regulate proinflammatory networks—an observation particularly germane to MS.⁴³

In the present study, a significant decrease in the sustained melanopsin response (ie, pupillary constriction) was observed when a blue light stimulus was applied to eyes with a history of AON compared with the unaffected fellow eye in the same patient. Furthermore, the melanopsin-mediated sustained pupillary response of the eyes of patients with MS without a history of AON was directly correlated with thinning of the inner retinal layers.

Patients with MS commonly develop varying degrees of tissue disorganization involving the optic nerve, retina, and central visual pathways, leading to loss in visual acuity (particularly low-contrast acuity), visual field deficits, color desaturation, illusory movements of objects in 3-dimensional space (eg, the Pulfrich phenomenon), and afferent pupillary defects.^1,2,6 However, to our knowledge, no prior human studies have functionally isolated the ipRGCs in patients with MS.

The investigations described herein report a significant dysfunction of the melanopsin-mediated sustained constriction responses in patients with MS that significantly correlated with a reduction in GCL + IPL thickness. Alternatively, most patients with MS with preserved GCL + IPL thickness exhibited a preserved melanopsin response, although eyes of patients with MS that exhibited an abnormally reduced melanopsin-mediated pupillary light reflex demonstrated corresponding and significant thickness reductions in the GCL + IPL of the retina.

Notwithstanding our observations, some eyes demonstrated reduced melanopsin-mediated pupillary responses despite normal GCL + IPL thickness, an observation most likely signifying that melanopsin-mediated pupillary responses may be more specific for demonstrating the neurophysiologic integrity of the retinal-melanopsin network (ie, functional deficits may precede changes in retinal architecture analogous to changes in visual-evoked potential response metrics that often antedate the subsequent development of retinal layer thickness reductions). As such, functional deficits of the pupillary response, concomitant with near-normal or even normal retinal thickness, may represent a cardinal pathophysiologic transformation in the properties of axonal conduction in MS characterized by attenuated or even complete block of axonal conduction velocity, albeit in the absence of altered architecture within the anterior visual system.

The correspondence between neurophysiologic response characteristics with validated measures of retinal architecture may yield, for the first time, the practical bases for the identification of multiparametric signatures that are derivatives of the inextricable linkage between structure and function within “eloquently elegant” neurobiological systems.

The corroboration of our findings, in conjunction with the application of innovative methods (eg, adjusting the stimulus characteristics from global blue light to highly precise and discrete blue light stimulation at 10-nm intervals across the spectrum of blue light), may translate the identification of retinal islands of response integrity into what we believe to be an intervention not previously available for the purpose of mitigating mood disorders, circadian resynchronization, chronic fatigue (through melatonin suppression and circadian synchronization), thermodysregulation, and other homeostatic systems that would appear to be dependent on the integrity of at least a proportion of the ipRGCs.

Correlative analyses of retinal structure and function in this study may be limited given that the melanopsin-containing retinal ganglion cells represent only a small subset of the total population of cells that contribute to thickness of the GCL + IPL.²⁸ For instance, the diffuse reduction in thickness of the inner retinal layers of nervous tissue (retinal nerve fiber layer and GCL + IPL) may not accurately indicate retinohypothalamic function.

With the advent of spectral domain OCT along with low-contrast acuity measures, multifocal visual-evoked potential, multifocal electroretinogram, and the chromatic pupillary light reflex, we now have access to a toolbox with high-precision objective methods by which such characteristics can be detected and monitored noninvasively and longitudinally.^14-19 Given the significant correlation of the melanopsin-mediated, persistent-constriction pupillary light reflex response with GCL + IPL thickness, we can now at least hypothesize that patients with MS with significant changes in anterior visual system architecture may be at risk for retinohypothalamic dysfunction. If confirmed, this finding may advance our understanding of the origins of a constellation of disabling manifestations commonly affiliated with MS and how we might develop strategies to mitigate them.

Corresponding Author: Elliot M. Frohman, MD, PhD, Department of Neurology and Neurotherapeutics, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, Dallas, TX 75390 (elliot.frohman@utsouthwestern.edu).

Accepted for Publication: October 21, 2016.

Published Online: January 30, 2017. doi:10.1001/jamaneurol.2016.5131

Author Contributions: Drs Meltzer and Sguigna contributed equally to this study. Drs Kardon and E. M. Frohman had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Meltzer, Sguigna, Subei, Beh, B. S. Frohman, A. N. Frohman, Saidha, Calabresi, Rennaker, T. C. Frohman, Kardon, E. M. Frohman.

Acquisition, analysis, or interpretation of data: Meltzer, Sguigna, Subei, Beh, Kildebeck, D. Conger, A. Conger, Lucero, B. S. Frohman, A. N. Frohman, Saidha, Galetta, T. C. Frohman, Kardon, Balcer, E. M. Frohman.

Drafting of the manuscript: Meltzer, Sguigna, Subei, B. S. Frohman, A. N. Frohman, Saidha, T. C. Frohman, E. M. Frohman.

Critical revision of the manuscript for important intellectual content: Meltzer, Subei, Beh, Kildebeck, D. Conger, A. Conger, Lucero, B. S. Frohman, A. N. Frohman, Saidha, Galetta, Calabresi, Rennaker, T. C. Frohman, Kardon, Balcer, E. M. Frohman.

Statistical analysis: Meltzer, Sguigna, B. S. Frohman, A. N. Frohman, Saidha, T. C. Frohman, Balcer, E. M. Frohman.

Administrative, technical, or material support: Meltzer, Sguigna, Lucero, Calabresi, Rennaker, T. C. Frohman, E. M. Frohman.

Study supervision: Galetta.

Conflict of Interest Disclosures: Dr Subei has received speaker and consultant fees from Genzyme. Dr Galetta has received consulting fees from Biogen Idec. Dr Calabresi has provided paid consultation services to Novartis, EMD-Serono, Teva, and Biogen and has received grant support from EMD-Serono, Teva, Biogen, Genentech, Bayer, Abbott, and Vertex. Dr Rennaker is a part owner of Vulintus LLC. Ms T. C. Frohman has received speaker and consultant fees from Genzyme, Novartis, and Acorda. Dr Kardon has received funding from the Department of Defense, Veterans Affairs Rehabilitation Research and Development, and Novartis steering committee for Optical Coherence Tomography in Multiple Sclerosis (OCTiMS) multicenter study and is co-founder of MedFace, which uses facial features to assess light sensitivity and vision. Dr Balcer has received honoraria for consulting on development of visual outcomes for MS trials from Biogen and Genzyme and is on a clinical trial advisory board for Biogen. Dr E. M. Frohman has received speaking and consulting fees from Teva Neuroscience, Genzyme, Acorda, and Novartis. No other disclosures were reported.

Funding/Support: Funding for the study was received from National Multiple Sclerosis Society grants RG 3780a3/3 (Dr E. M. Frohman) and RG 4212-A-4 (Dr Balcer; subcontracted to Drs Calabresi and E. M. Frohman), National Eye Institute grants R01 EY 014993 and R01 EY 019473 (Dr Balcer; subcontracted to Drs Calabresi and E. M. Frohman), and Braxton Debbie Angela Dillon and Skip Donor Advisor Fund (Dr E. M. Frohman; subcontracted to Drs Calabresi and Balcer).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Jason Ooi created the artistic illustrations specifically formulated and produced for this article. Ron Cohen, MD (Acorda Therapeutics), provided critical and constructive review and revision of the manuscript. There was no financial compensation.