A, In the color fundus image, green lines indicate optical coherence tomographic scan locations for corresponding optical coherence tomographic images; black arrowheads indicate lesion sites. B, Three optical coherence tomographic images through Ebola retinal lesions demonstrate the proximity of multifocal discontinuities of the ellipsoid zone with an overlying V-shaped increased reflectance of the outer nuclear layer (equal to the reflectance of the adjacent outer plexiform layer). Black arrowheads indicate lesion sites.
A, An optical coherence tomographic image through an Ebola retinal lesion; arrowhead indicates an area of perilesional dark without pressure that corresponds to a thinned hyporeflective ellipsoid zone and absent interdigitation zone. B, An optical coherence tomographic image shows the transition between the normal retina and the circumferential extension of an area of dark without pressure near the nose. The enlarged inset highlights the transitional zone. C, Multiple Ebola retinal lesions and associated areas of dark without pressure are visible in a right-eye ultra-widefield retinal image. The specific length and direction of the line labeled “A” precisely correlates to the length and direction of the cross-sectional optical coherence tomographic slice shown in image A, and the line labeled “B” correlates to the length and direction of image B.
eMethods 1. Examination Protocol
eMethods 2. Humphrey Peripheral Nasal Visual Field Protocol
eTable 1. Ebola Virus Disease Survivor Cohort characteristics
eTable 2. Ebola Retinal Lesion Grading Classification
eTable 3. Ebola Retinal Lesion OCT Grading Results
eTable 4. Ebola Retinal Lesion OCT Grading Cohen’s Kappa Statistic
eFigure 1. Grading Classification Reference Images
eFigure 2. Survivor No. 1. Left eye
eFigure 3. Survivor No. 2. Right eye
eFigure 4. Survivor No. 2. Left Eye
eFigure 5. Survivor No. 3. Right eye
eFigure 6. Survivor No. 4. Right eye
eFigure 7. Survivor No. 4. Left eye
eFigure 8. Survivor No. 5. Right eye
eFigure 9. Survivor No. 5. Left eye
eFigure 10. Survivor No. 6. Right eye
eFigure 11. Survivor No. 6. Left eye
eFigure 12. Survivor No. 7. Right eye
eFigure 13. Survivor No. 7. Left eye
eFigure 14. Survivor No. 8. Right eye
eFigure 15. Survivor No. 9. Right eye
eFigure 16. Survivor No. 9. Left eye
eFigure 17. Survivor No. 10. Right eye
eFigure 18. Survivor No. 10. Left eye
eFigure 19. Survivor No. 11. Right eye
eFigure 20. Survivor No. 11. Left eye
eFigure 21. Survivor No. 12. Right eye
eFigure 22. Survivor No. 13. Left eye
eFigure 23. Survivor No. 14. Right eye
eFigure 24. Amalgamated retinal lesion spatial distribution
Customize your JAMA Network experience by selecting one or more topics from the list below.
Steptoe PJ, Momorie F, Fornah AD, et al. Multimodal Imaging and Spatial Analysis of Ebola Retinal Lesions in 14 Survivors of Ebola Virus Disease. JAMA Ophthalmol. 2018;136(6):689–693. doi:10.1001/jamaophthalmol.2018.1248
Can multimodal imaging of Ebola retinal lesions inform our understanding of their pathogenesis?
In this case-series study of 14 survivors, optical coherence tomography demonstrated a V-shaped increased reflectivity of the outer nuclear layer overlying discontinuities of the ellipsoid zone and interdigitation zone in the smallest lesions. A collapse of the overlying retinal structures was detected in larger lesions, corresponding visual field defects respected the horizontal raphe, and perilesional areas of dark without pressure (ellipsoid zone hyporeflectivity) accompanied 89% of lesions.
These findings are consistent with a neuronal rather than vascular pathogenesis; the relevance of dark without pressure is undetermined.
Differentiation between Ebola retinal lesions and other retinal pathologies in West Africa is important, and the pathogenesis of Ebola retinal disease remains poorly understood.
To describe the appearance of Ebola virus disease (EVD) retinal lesions using multimodal imaging to enable inferences on potential pathogenesis.
Design, Setting, and Participants
This prospective case series study was carried out at 34 Military Hospital in Freetown, Sierra Leone. Ophthalmological images were analyzed from 14 consecutively identified survivors of EVD of Sierra Leonean origin who had identified Ebola retinal lesions.
Main Outcomes and Measures
Multimodal imaging findings including ultra-widefield scanning laser ophthalmoscopy, fundus autofluorescence, swept-source optical coherence tomography (OCT), Humphrey visual field analysis, and spatial analysis.
The 14 study participants had a mean (SD) age of 37.1 (8.8) years; 6 (43%) were women. A total of 141 Ebola retinal lesions were observed in 22 of 27 eyes (81%) of these 14 survivors on ultra-widefield imaging. Of these, 41 lesions (29.1%) were accessible to OCT imaging. Retinal lesions were predominantly nonpigmented with a pale-gray appearance. Peripapillary lesions exhibited variable curvatures in keeping with the retinal nerve fiber layer projections. All lesions respected the horizontal raphe and spared the fovea. The OCT imaging demonstrated a V-shaped hyperreflectivity of the outer nuclear layer overlying discontinuities of the ellipsoid zone and interdigitation zone in the smaller lesions. Larger lesions caused a collapse of the retinal layers and loss of retinal thickness. Lesion shapes were variable, but sharp angulations were characteristic. Perilesional areas of dark without pressure (thinned ellipsoid zone hyporeflectivity) accompanied 125 of the 141 lesions (88.7%) to varying extents.
Conclusions and Relevance
We demonstrate OCT evidence of localized pathological changes at the level of the photoreceptors in small lesions among survivors of EVD with retinal lesions. The relevance of associated areas of dark without pressure remains undetermined.
We previously conducted a case-control study that identified retinal characteristics specific to survivors of Ebola virus disease (EVD) in Sierra Leone using ultra-widefield (UWF) retinal imaging. Of all retinal lesions characterized in the previous study, only 1 lesion of characteristic morphological appearance was exclusively identified in Ebola survivors by 2 masked graders.1 On this basis, this lesion was deemed most likely to be associated with Ebola virus infection. Identical lesions have been identified in other cohort studies of survivors of EVD.2,3 This expanded analysis provides optical coherence tomography (OCT) interpretation and functional visual field (VF) analysis to provide further insights into the pathophysiology of Ebola retinal sequelae.
A total of 160 survivors of EVD attended the ophthalmology clinic at 34 Military Hospital in Freetown, Sierra Leone, between January 2016 and April 2017. Of these, 14 survivors of EVD (8.8%) met the eligibility criterion derived from the findings of our previous study, which was having at least 1 Ebola retinal lesion identified on UWF retinal imaging.1 All 14 patients were recalled to the clinic and attended examinations that included OCT of accessible Ebola retinal lesions. Thirteen of these 14 survivors (93%) were recruited to this study, and 1 (7%) was excluded because of increased lens opacity that prevented fundus imaging. One further eligible patient was identified in March 2017 and directly enrolled into the study.
Informed consent was obtained from all participants. The study was approved by the Office of Sierra Leone Ethics and Scientific Review Committee on January 31, 2017, and followed the tenets of the Declaration of Helsinki.4
The appearance of lesions on OCT was categorized, and lesion grading concordance was compared by 2 ophthalmologists (P.J.S. and N.A.V.B.) who were each masked to the grades of the other grader. The examination protocol is summarized in eMethods 1 and eFigure 1 in the Supplement.
The 14 participants had a mean (SD) age of 37.1 (8.8) years. Six (43%) were women. The participants had a total of 27 eyes (1 patient had no view of the fundus available for retinal examination in 1 eye as a result of previous ocular trauma that had led to corneal and lens opacity followed by complicated surgery before Ebola infection). We analyzed 141 Ebola retinal lesions in 22 of the 27 eyes (81%; including 16 eyes of 8 patients with bilateral involvement and 6 eyes of 6 patients with unilateral involvement). Images were obtained by OCT of 41 of the 141 lesions (29.1%) in 20 of the 22 eyes (74% of the 27 total eyes). Characteristics of all 14 patients are summarized in eTable 1 in the Supplement. Corresponding multimodal imaging and VF images are available in eFigures 2 through 23 in the Supplement.
Ebola retinal lesions varied in size and shape, but distinctive linear borders with sharp angulations were characteristic (eFigure 11 in the Supplement). Multimodal imaging features varied according to severity and extent of retinal structures involved. A lesion severity grading from 1 to 5 is outlined in eTable 2 in the Supplement; results are shown in eTable 3 of the Supplement. The Cohen κ statistic of intergrader agreement was 0.77 (eTable 4 in the Supplement).
The OCT images of the smallest lesions demonstrated multiple vertical discontinuities of the ellipsoid zone (EZ) and interdigitation zone (IZ) with overlying V-shaped increased reflectance of the outer nuclear layer (Figure 1 and eFigure 21 in the Supplement). Lesions appeared light gray in color on fundus photography and were predominantly nonpigmented.
Peripapillary lesions demonstrated variable curvatures depending on the optic disc perimeter location and resembled the arcuate anatomical pathways of the retinal nerve fiber layer (ganglion cell axons) (eFigures 4, 10, 11, and 15 in the Supplement). Visual acuity and color vision were preserved in all cases in the absence of other pathology. Corresponding absolute VF defects that respected the anatomical horizontal raphe were observed on 24-2 Humphrey visual field analysis (eFigure 10 in the Supplement) and with a peripheral nasal 24-2 test protocol (eMethods 2 and eFigure 10 in the Supplement).
Well-defined areas of dark without pressure (DWP) that corresponded on OCT imaging to a thinned, hyporeflective EZ and absent IZ (Figure 2) were seen adjacent to 125 of 141 Ebola retinal lesions (88.7%) in this series. The extent of DWP was variable, ranging from a confined circumferential marginal zone (eFigures 2, 6, 8, 14, 19, and 20 in the Supplement) to larger defined areas (eFigures 5, 7, and 15 in the Supplement and Figure 2) and in some cases 360° panretinal involvement (eFigure 16 in the Supplement). The extent of DWP in some eyes appeared to be associated with the density of Ebola retinal lesions (Figure 2 and eFigure 15 in the Supplement). In all cases, DWP appeared to spare the macula. No associated vitreous inflammation or vitreous traction were visible on OCT imaging.
The 120-point screen, 60-4 threshold tests, and peripheral nasal 24-2 protocol (eFigure 10 in the Supplement) were unable to conclusively identify any definitive VF defect corresponding to areas of DWP. The 24-2 Humphrey visual field analysis of 1 survivor with right hemiparesis after acute infection demonstrated a right-sided homonymous hemianopia and left inferior quadrantanopia (eFigures 19 and 20 in the Supplement).
Aligned and amalgamated retinal images with corresponding Ebola retinal lesion loci and longitudinal axis are shown in eFigure 24 in the Supplement. No overlapping axes or crossing of the horizontal raphe was observed.
We present a multimodal imaging analysis of a series of 14 survivors of EVD with Ebola retinal lesions, as characterized in our previous case-control study.1 While OCT analysis of larger lesions involving all retinal layers provides little insight into pathogenesis, OCT observations of small lesions revealed multifocal fine discontinuities of the EZ and IZ with overlying increased reflectivity of the outer nuclear layer (Figure 1). These findings mirror the histological appearance of early herpes simplex virus retinopathy observed in the contralateral retina via a retrograde axonal transmission after unilateral anterior chamber viral inoculation.5,6 They have also been observed in the ipsilateral retina following unilateral anterior chamber viral inoculation,6 although in all cases in this study, we did not identify signs of previous anterior chamber uveitis that would suggest a direct anterior to posterior dissemination. Peripapillary curvilinear lesions resembling the arcuate path of the ganglion cell axons were shown on both imaging and VF analysis to respect the horizontal raphe. Their presence provides further evidence that EVD involves retinal ganglion cells and creates a lasting insult to their afferent photoreceptors.
Possible pathogenic mechanisms for the characteristic retinal lesions observed in survivors of EVD could include retrograde neuronal transmission. Vascular ocular dissemination and involvement of the optic nerve leptomeninges has been demonstrated in a rhesus monkey model with acute fulminant Ebola infection.7
Although nonspecific to Ebola retinal lesions, the frequency of circumferential marginal zones of DWP around Ebola retinal lesions strongly suggests an association. This is supported by the correlation between Ebola retinal lesion density and the extent of DWP in some eyes (eFigure 15 in the Supplement). Areas of DWP in this study correspond to a hyporeflective thinning of the second hyperreflective band and loss of the third hyperreflective band on OCT imaging, currently termed the ellipsoid zone and interdigitation zone, respectively.8 Although controversy continues over the precise anatomical correlates of these bands,9,10 recent cellular characterization using immunohistochemistry markers concurs that the second band is generated by the photoreceptor ellipsoids and is probably the result of the high number of mitochondria that they contain, while the third band corresponds to the cone phagosomes located in the top of the retinal pigment epithelium.11
Because of the lack of histological evidence, preinfection imaging, and retinal imaging during acute infection, an absolute temporal association with EVD and the Ebola retinal lesions and associated DWP has yet to be established. We have not compared the OCT findings presented in this study with a control group of patients with retinal lesions associated with other pathologies to confirm that these characteristics are unique to Ebola retinal lesions.
In this study, we demonstrate pathological changes seen at the level of the photoreceptors on OCT in small lesions. We demonstrate associated areas of DWP that appear as a hyporeflective, thinned EZ in combination with an absence of the IZ on OCT imaging. The importance of this finding remains to be determined, and follow-up observations are ongoing. These findings suggest that survivors of EVD in future outbreaks would benefit from ophthalmologic evaluation, including via OCT analysis and visual field assessment.
Corresponding Author: Paul J. Steptoe, MBChB, St. Paul’s Eye Department, Royal Liverpool University Hospital, Prescot Road, Liverpool, Merseyside L7 8XP, United Kingdom (email@example.com).
Accepted for Publication: March 4, 2018.
Published Online: May 3, 2018. doi:10.1001/jamaophthalmol.2018.1248
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2018 Steptoe PJ et al. JAMA Ophthalmology.
Author Contributions: Drs Steptoe and Semple had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Beare and Semple are co–last authors.
Study concept and design: Steptoe, Scott, Vandy, Sahr, Beare, Semple.
Acquisition, analysis, or interpretation of data: Steptoe, Momorie, Komba, Emsley, Harding, Beare, Semple.
Drafting of the manuscript: Steptoe, Momorie, Komba, Emsley, Beare, Semple.
Critical revision of the manuscript for important intellectual content: Steptoe, Scott, Harding, Vandy, Sahr, Beare, Semple.
Statistical analysis: Steptoe, Scott.
Obtained funding: Steptoe, Scott, Semple.
Administrative, technical, or material support: Momorie, Komba, Emsley, Scott, Vandy, Sahr, Semple.
Study supervision: Harding, Vandy, Sahr, Beare, Semple.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Steptoe reports grants from Global Ophthalmology Awards Programme Grant supported by Bayer, the Dowager Countess Eleanor Peel Trust, and Enhancing Research in Epidemic Situations funded by Wellcome Trust; nonfinancial support from Optos, CBM Italia, and Topcon during the conduct of the study; and grants from National Institute for Health Research Health Protection Research Unit in Emerging and Zoonotic Infections and British Medical Association Humanitarian Fund outside the submitted work. Dr Sahr reports grants from Wellcome Trust Enhancing Research Activity in Epidemic Situations Programme award, Wellcome Trust, Bill and Melinda Gates Foundation, USA Department of Defense HIV/AIDS Prevention Program, Save the Children UK, US National Institute of Allergy and Infectious Diseases, the World Health Organization, US Centers for Disease Control and Prevention, and European Commission–Research Seventh Framework Programme, outside the submitted work. Dr Scott reports grants from the Wellcome Trust Enhancing Research Activity in Epidemic Situations Programme, National Institute for Health Research Health Protection Research Unit in Emerging and Zoonotic Infections during the conduct of the study. Dr Beare reports personal fees from Abbvie, Wellcome Trust, and Santen, outside the submitted work. Dr Semple reports grants from National Institute for Health Research Health Protection Research Unit in Emerging and Zoonotic Infections at the University of Liverpool and Wellcome Trust Enhancing Research Activity in Epidemic Situations Programme award during the conduct of the study and grants from Wellcome Trust and Bill and Melinda Gates Foundations outside the submitted work. No other disclosures were reported.
Role of the Funder/Sponsor: The funding bodies and organizations who provided equipment support for this research had 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: We thank Optos PLC for their generous donation of the Daytona Ophthalmoscope, which continues to improve patient care for the people of Sierra Leone; Topcon for the loan of their OCT device; Onlime Sierra Leone Ltd for supplying the clinic with internet access; CBM Italia for their 2WIN autorefractor donation; the administration at 34 Military Hospital, Freetown, Sierra Leone, for facilitating the study; and Ian Pearce, FRCOphth, MB ChB, BSc, St. Paul's Eye Unit, Royal Liverpool University Hospital, for institutional peer review. Dr Pearce was not compensated for his contribution to the study.
Create a personal account or sign in to: