Detection of clinical and subclinical retinal abnormalities in neurosarcoidosis with optical coherence tomography
- First Online:
- Cite this article as:
- Eckstein, C., Saidha, S., Sotirchos, E.S. et al. J Neurol (2012) 259: 1390. doi:10.1007/s00415-011-6363-8
- 181 Views
The aim of this work was to determine if neurosarcoidosis (NS) patients exhibit quantitative and/or qualitative in vivo evidence of retinal abnormalities on optical coherence tomography (OCT). Retinal imaging was performed using spectral-domain Cirrus HD-OCT in 20 NS patients (40 eyes) and 24 age-matched healthy controls (48 eyes). Study participants also underwent magnetic resonance imaging of the brain and spine, cerebrospinal fluid (CSF) analysis, and detailed neurological and ophthalmological evaluation. Quantitative OCT abnormalities of average macular thickness (AMT), peri-papillary retinal nerve fiber layer (RNFL) thickness, or both, were detectable in 60% of NS patients. Of NS patients with ocular symptomatology, 75% demonstrated quantitative OCT abnormalities, while only 25% had detectable abnormalities on detailed ophthalmological assessment. Furthermore, 33% of NS patients without ocular symptoms had quantitative OCT changes, while only 8% had abnormal ophthalmologic examination. RNFL and macular thinning and swelling were significant in the NS cohort compared to healthy controls (variance ratio testing; RNFL: p = 0.02, AMT: p = 0.006). AMT also correlated inversely with disease duration (rs = −0.65, p = 0.002). Patient proportions with OCT abnormalities did not differ according to NS subtype (myelopathic, meningeal, or encephalitic NS), CSF findings, or immunotherapy exposure. No qualitative OCT abnormalities were detected. Retinal abnormalities occur in all NS subtypes, and may be clinical or subclinical. Our findings suggest OCT may enable greater detection of retinal abnormalities in NS than ophthalmological assessment alone, and have implications for the assessment of ocular involvement in NS, and sarcoidosis in general. Longitudinal NS studies utilizing OCT are warranted.
KeywordsOptical coherence tomographyNeurosarcoidosisOptic neuropathyUveitisRetina
Sarcoidosis is a chronic, multi-system granulomatous disorder with predilection for the lungs, eyes, and skin . Symptoms consistent with central nervous system (CNS) involvement (referred to as neurosarcoidosis; NS) are reported in approximately 5–15% of sarcoidosis patients . Only 50% of NS cases at autopsy are recognized to have CNS involvement ante-mortem, suggesting a significant proportion of NS patients harbor subclinical disease . However, due to the lack of detailed, systematic neurologic studies in sarcoidosis, the prevalence of symptomatic or asymptomatic NS is unclear, and the clinical relevance of subclinical NS remains to be firmly established. Nonetheless, it is evident that NS may be debilitating and may require more aggressive immunotherapy than conventional pulmonary sarcoidosis .
Ocular involvement frequently occurs and is detectable in approximately 30–60% of systemic sarcoidosis cases [5, 6]. The most common ocular manifestation of sarcoidosis is uveitis, reflecting up to 70% of cases with ocular involvement [5–7]. Anterior uveitis, posterior uveitis (affects choroid and retina; part of the CNS), intermediate uveitis and panuveitis may all occur in sarcoidosis, acutely or chronically, and may be subclinical. Posterior segment inflammation may occur in up to 30–50% of patients with ocular sarcoidosis , and although it is suggested that posterior segment uveitis and NS may be associated , detailed ocular assessments including advanced high-resolution ocular imaging are lacking in NS.
Optic disc pallor is commonly observed in sarcoidosis , reflecting axonal degeneration in the retinal nerve fiber layer (RNFL) . The constituent axons of the optic nerve, derived from the RNFL, undergo retrograde degeneration secondary to optic nerve pathology, resulting in RNFL degeneration [12, 13]. Cystoid macular edema (CME) associated with posterior segment uveitis [14, 15], causing blindness in up to 10% of sarcoidosis patients , can be visualized with optical coherence tomography (OCT) [17–19].
OCT, a noninvasive, reproducible and reliable imaging technique enables high-resolution quantitative retinal imaging [20–22]. OCT has gained considerable interest in neurological disorders with affinity to cause retinal changes, such as multiple sclerosis (MS). OCT allows quantitative assessment of the integrity of the RNFL [13, 22]. Furthermore, OCT provides quantitative measures of average macular thickness (AMT), reflecting retinal neuronal integrity . Since uveitis and optic nerve involvement may occur in NS, OCT may provide insight into potential retinal changes resulting from these processes (clinically and subclinically). The objective of this cross-sectional study was to determine if NS patients exhibit in vivo evidence of retinal abnormalities on OCT and to characterize and quantify such changes.
Subjects were recruited by unselected convenience sampling from the Johns Hopkins Neuroimmunology Clinic. Participants without CNS or extra-CNS histologic evidence of sarcoidosis were excluded. Subjects with ophthalmologic or neurologic disorders including glaucoma, diabetes, and hypertension, and/or a refractive error of greater than ±6 diopters, which may affect OCT measures, were excluded from the study. OCT scans acquired within 3 months of acute optic neuritis (AON) were excluded to minimize effects of optic disc edema on OCT measurements. Age-matched healthy controls (HCs) without known ocular or neurologic disease were recruited from volunteers within the Johns Hopkins medical system. Johns Hopkins University Institutional Review Board approval was obtained for study protocols. Participants provided written informed consent. The study was performed in accordance with Health Insurance Portability and Accountability Act guidelines.
NS was classified as definite or probable by the treating neurologist (CP), based on proposed diagnostic criteria . Magnetic resonance imaging (MRI) of the brain and spine were reviewed in a blinded fashion by a trained neurologist (SS), and the following characteristics were recorded: topography and burden of intra- or extra-axial lesions suspected to be associated with NS and lesion activity status based on contrast enhancement. Participants were further classified as having predominantly meningeal, myelopathic, or encephalitic subtypes of NS. Participants underwent lumbar puncture for the evaluation of cerebrospinal fluid (CSF) constituents including CSF glucose, protein, white cell count, and oligoclonal bands. Subjects with myelopathic NS were tested for serum anti-aquaporin-4 (AQP4) IgG antibodies characteristic of neuromyelitis optica. Participants were screened for the following ophthalmic symptoms: blurring, diplopia, photophobia, and excessive glare, and underwent detailed ophthalmological assessment including dilated fundus examination, fundus photography, intra-ocular pressure measurement, and slit-lamp examination. The clinical profile and evolution of sarcoidosis-related symptoms were evaluated to determine disease duration based upon first neurologic symptoms of NS, co-morbidities, and history of AON.
Optical coherence tomography (OCT)
Retinal imaging was performed using spectral-domain Cirrus HD-OCT (model 4000), with software version 5.0 (Carl Zeiss Meditec, Dublin, California) as described elsewhere [25, 26]. Briefly, peri-papillary data were obtained with the Optic Disc Cube 200 × 200 protocol, consisting of 200 horizontal scan lines (each composed of 200 A-scans) that form a 6 × 6 × 2 mm volume cube. Segmentation software determines the location of the inner limiting membrane and the outer boundary of the RNFL at each A-scan to create a 2D map of peri-papillary RNFL thickness. Macular data were obtained using the Macular Cube 512 × 128 protocol (128 horizontal scan lines each composed of 512 A-scans and one central vertical and horizontal scan composed of 1024 A-scans), forming a 6 × 6 × 2 mm volume cube. Different segmentation software identifies the inner limiting membrane and the inner boundary of the retinal pigment epithelium in this cube allowing determination of AMT. OCT scanning was performed by three trained technicians, who monitored scans to ensure fixation was reliable, as previously described . Scans with signal strength less than 7/10 were excluded from analyses.
The OCT software used an automated, computerized algorithm to compare measurements of average RNFL thickness and AMT against a normal distribution percentile scheme derived from the Cirrus HD-OCT normative database of 284 age-matched controls. This allowed designation of measures into the following categories: normal (5th–95th ‰), below normal (<5th ‰) or supranormal (>95th ‰).
Additionally, constituent B-scans of all acquired OCT images were reviewed for qualitative abnormalities, in a blinded fashion, by a trained ophthalmologist (SS).
Statistical analysis was completed on STATA version 11 (StataCorp, College Station, TX). OCT measures between the NS cohort and HCs were compared using Wilcoxon rank sum (Mann–Whitney) test and variance ratio testing. Spearman rank correlation was used to determine correlations between OCT measures and clinical characteristics. These analyses were then repeated using mixed-effects regression to account for inter-eye correlations. Fisher’s exact test was used in proportional analyses.
Demographics and clinical characteristics
Sarcoidosis (n = 20)
Controls (n = 24)
Age, mean (SD) years
Race, n (%)
Disease duration, mean (SD) years
Subtype, n (% of cohort)
Visual complaints, n (% of cohort)
Treatments, n (% of cohort)
CSF WBC count, mean/mm3 (range)
CSF protein, mean mg/dl (range)
Quantitative abnormalities of average RNFL thickness, AMT or both (<5th or >95th percentile compared to the Cirrus HD-OCT normative database) were detected in 60% of NS patients without associated qualitative OCT abnormalities. Of these, 50% had visual symptoms. Nine NS patients (45%) had RNFL or AMT abnormalities alone (unilateral in eight, bilateral in one). Three (15%) demonstrated concomitant RNFL thickness and AMT abnormalities (unilateral in two, bilateral in one). Two patients demonstrated increases in RNFL and AMT (one of whom had the temporoparietal lesion with mass effect), while one patient demonstrated reduction in both RNFL thickness and AMT (the patient with prior AON history).
Mean RNFL thickness and AMT
RNFL, μm (SD)
AMT, μm (SD)
There was no difference in the proportion of patients with OCT (RNFL or AMT) abnormalities with normal or abnormal CSF findings (Fisher’s exact test: p = 0.60). There was no difference in the proportion of NS patients with OCT abnormalities, either RNFL/AMT thinning or swelling, and history of ophthalmic symptoms (Fisher’s exact test: p = 0.37). Of eight patients with ophthalmic symptoms, six (75%) had abnormal OCTs, while only two (25%) had detectable abnormalities on ophthalmological assessment. Four of 12 patients (33%) without ophthalmic symptoms had abnormal OCTs, and only one (8%) had detectable abnormalities on ophthalmological assessment. All patients with detectable abnormalities on ophthalmological assessment (n = 3) had abnormal OCT scans. The patient with bilateral disc edema (with the temporoparietal lesion exerting mass effect) had increased RNFL and AMT bilaterally. The patient with the asymptomatic superior disc hemorrhage with superior disc elevation, and the patient with symptomatic intermediate uveitis had increased AMT ipsilaterally. However, the majority of patients with abnormal OCT scans (n = 10) had normal ophthalmological examination (n = 7). Patient proportions with either RNFL or AMT abnormalities did not differ by NS subtype (Fisher’s exact test: p = 0.19). Finally, the proportion of patients with prior or current exposure to prednisone or immunotherapies (mycophenolate mofetil, azathioprine, or hydroxychloroquine) did not differ in those with and without OCT abnormalities.
Above analyses were repeated using mixed effects linear regression accounting for within-subject inter-eye correlations. Results were similar to those presented above and did not result in loss or gain of significance.
Although OCT has been previously evaluated in a variety of neurological disorders, especially MS, little is known regarding potential OCT abnormalities which may occur in NS (or sarcoidosis). This study objectively demonstrates that symptomatic and asymptomatic retinal abnormalities can be detected in vivo in NS with OCT. Our results suggest retinal abnormalities occur in all NS subtypes, and although such changes may be symptomatic in approximately 60% of cases, they may only be detectable in a minority of cases on ophthalmological assessment. Furthermore, 33% of NS patients without any ophthalmologic symptoms in this study had detectable abnormalities on OCT, while only 8% of such patients had abnormalities on ophthalmological examination. These findings highlight the utility of OCT, beyond ophthalmological examination, in the ophthalmic assessment of NS patients and may have implications for the assessment of ocular involvement in sarcoidosis in general. Our findings are consistent with studies suggesting OCT is more sensitive than routine ophthalmological assessment alone for detecting RNFL thinning and primary retinal processes in neurological disorders [25, 27]. In one study, OCT detection of structural abnormalities consistent with primary retinal pathology (normal RNFL with AMT <5th ‰) was functionally corroborated by appropriate multifocal electroretinography (mfERG) abnormalities in a subset of patients .
Since sarcoidosis may directly and indirectly (through optic nerve involvement) produce retinal changes, complex heterogeneous OCT findings in a study of this nature are not unexpected. Although challenging to interpret our OCT findings in the absence of similar studies in sarcoidosis and NS, based on the known pathophysiology and OCT findings associated with optic neuropathies, retinopathies and various neurologic disorders, we propose there are a limited number of explanations underlying our observations. Increased RNFL thickness, with or without increased AMT may be a manifestation of either clinical or subclinical AON [11, 13, 22, 28]. Although there was no radiological evidence of AON, dedicated optic nerve protocol MRI was not performed as part of this study, and thus AON may not have been radiologically detectable. Alternatively, similar OCT findings may result from raised intracranial pressure (ICP) [29, 30]. This likely represents the etiology of OCT findings in the patient with the temporoparietal lesion exerting mass effect. Other than in this case, raised ICP was not suspected to be present in any of the other participants.
RNFL thinning, with or without AMT reduction, likely represents sequelae of prior AON or subclinical optic neuropathy . Since only one patient in this study had a known history of AON, we suspect subclinical optic neuropathy accounted for the majority of these OCT changes, consistent with observations in MS . AMT changes in the absence of RNFL changes most likely represent primary retinal changes . Increased AMT detected in this study is the likely derivative of clinical or subclinical uveitis [31, 32], possibly causing either CME  or retinal vasculitis . Notably, previous histopathologic reports describe the presence of intraretinal and preretinal granulomas in sarcoidosis [35, 36], which have recently been shown to be detectable with OCT . Although qualitative assessment of OCT scans in this study did not reveal granulomas, this does not definitely exclude their presence and potential contribution to increased AMT. Improved OCT resolution in the future may enable greater detection of retinal granulomas or inflammatory retinal changes. AMT reductions may also be the sequelae of uveitis, perhaps causing chronic or recurrent CME or vasculitis [38, 39]. Retinal vasculitis may be undetectable clinically and cannot be definitively excluded in the absence of fluorescein angiography, which was not performed in this study. Notwithstanding these potential explanations for our observations, we acknowledge only a comprehensive longitudinal study may help to elucidate the precise etiology of OCT changes in NS, and sarcoidosis in general. This however does not mitigate the relevance of our findings, which suggest symptomatic and asymptomatic retinal abnormalities can be detected in NS with OCT, even in the absence of identifiable clinical abnormalities. This indicates the potential utility of OCT in clinical and epidemiological studies of sarcoidosis (with and without CNS involvement) and suggests longitudinal OCT studies may be warranted.
This study has a number of limitations warranting discussion. The sample size in this study was small, reflecting the relatively low frequency of NS. Studies of larger cohorts of NS patients in the future may be more informative. Beyond this, similar studies assessing ocular findings including OCT should also be performed in sarcoidosis patients without CNS involvement, since they may harbor subclinical ocular disease. mfERG was not performed in this study. Although this may have provided useful functional information, particularly in those with suspected primary retinal processes, there is known disconnect between OCT and mfERG, such that both tests are not abnormal in roughly 50% of retinal disorders . Furthermore, mfERG is invasive, time-consuming, expensive, and potentially painful. Another limitation of this study is the absence of opening pressure data at the time of lumbar puncture, enabling objective ICP estimation. The presence of raised ICP was the likely cause of abnormal clinical and OCT findings in at least one of the patients included. However, we do not expect this to be a confounding factor in other subjects, since the majority of participants in this study had either myelopathic or patchy meningeal NS without MRI brain features consistent with raised ICP. Nonetheless, omission of opening pressure is a limitation and should serve as a reminder to investigators to include this in future studies assessing OCT findings in NS.
In summary, we demonstrate that greater than half of NS patients may have retinal abnormalities detectable with OCT. Retinal changes in NS may be clinical or subclinical, and OCT may possibly enable greater detection of retinal changes in NS than ophthalmological assessment alone. The etiology of retinal changes in NS most likely relate to the combination of optic neuropathy and uveitis (although contribution from retinal granulomas may also be possible). Our findings shed light on the pathophysiological mechanisms underlying retinal changes in NS, and in a broader sense imply there may be utility to the use of OCT, in addition to ophthalmological assessment, for the identification of ocular involvement in sarcoidosis. Future larger NS studies and studies of sarcoidosis without CNS involvement utilizing OCT are warranted, and may provide further insight into this highly complex and heterogeneous disorder.
National Multiple Sclerosis Society TR 3760-A-3 and Project Restore-Bart McLean Fund for Neuroimmunology Research.
Conflicts of interest
Dr. Christopher Eckstein receives funding through a Sylvia Lawry Physician Fellowship from the National Multiple Sclerosis Society. Dr. Shiv Saidha has received consulting fees from MedicalLogix for the development of continuing medical education programs in neurology, and educational grant support from TEVA Neurosciences. Dr. Peter Calabresi has provided consultation services to Novartis, Teva, Biogen-IDEC, Vaccinex, Abbott, Genentech; and has received grant support from EMD-Serono, Teva, Biogen IDEC, Genentech, Bayer, Abbott, and Vertex. Stephanie Syc has no disclosures. Michaela Seigo has no disclosures. Aleksandra Stankiewicz has no disclosures. E’Tona Ford has no disclosures. Gita Byraiah is supported by the Accelerated Cure Project for Multiple Sclerosis. Elias Sotirchos has no disclosures. Srilakshmi Sharma received funding from The Hirschorn Foundation Carlos A. Pardo receives funding through the Project Restore-Bart McLean Fund for Neuroimmunology Research and the Guthy-Jackson Charitable Foundation.