Enhanced Depth Imaging Optical Coherence Tomography of Optic Nerve Head Drusen in Children

To assess the utility of enhanced depth imaging optical coherence tomography (EDI-OCT), compared with other conventional imaging modalities, for detecting and characterizing optic nerve head drusen (ONHD) in children.

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Enhanced Depth Imaging Optical Coherence Tomography of Optic Nerve Head Drusen in Children Peng Yong Sim, MBChB (Hons), BMedSci (Hons), Hibba Soomro, MBBS, FRCOphth, Michael Karampelas, MD, FEBO, MRCOphth, Faye Barampouti, MD, FRCOphth Background: To assess the utility of enhanced depth imaging optical coherence tomography (EDI-OCT), compared with other conventional imaging modalities, for detecting and characterizing optic nerve head drusen (ONHD) in children. Methods: We report a retrospective cross-sectional case series of consecutive pediatric patients (age #16 years) with ONHD confirmed using B-scan ultrasonography. All eyes were evaluated using spectral-domain OCT of the optic nerve head in conventional (non-EDI) and EDI modes, fundus autofluorescence (FAF), and standard automated perimetry. Detection rates and the capacity to characterize ONHD were compared between EDI-OCT, non–EDI-OCT, and FAF. Results: Twenty-eight eyes of 15 patients (mean age 11 years; 60% female) were identified with definite ONHD that were confirmed by B-scan ultrasound. Among the technologies, EDI-OCT, non–EDI-OCT, FAF, and automated perimetry had findings consistent with ONHD in 24, 21, 18, and 4 eyes, respectively. EDI-OCT had a significantly better detection capability (86% of eyes) compared with FAF (P = 0.04) but not with non–EDI-OCT (P = 0.15). Similar to results previously reported in adult patients, EDI-OCT detected ONHD at different levels of depth; most were located anterior to the lamina cribrosa. ONHD detected by EDI-OCT appeared as hypo‐ reflective ovoid regions bordered by hyper-reflective material or as isolated hyper-reflective bands without a hypo-reflective core. The mean greatest diameter of ONHD seen on EDI-OCT was 449.7 (SD ±114.1) mm. Conclusions: EDI-OCT detects ONHD in most eyes identified as having drusen on B-scan ultrasonography. This technique has the potential to be an effective alternative first-line diagnostic and monitoring tool for ONHD, particularly for detecting buried drusen in children. Journal of Neuro-Ophthalmology 2020;40:498–503 doi: 10.1097/WNO.0000000000000845 © 2019 by North American Neuro-Ophthalmology Society Ophthalmology Department, Watford General Hospital, Watford, United Kingdom. The authors report no conflicts of interest. Address correspondence to Peng Yong Sim, MBChB (Hons), BMedSci (Hons), Ophthalmology Department, Watford General Hospital, Watford, WD18 0HB, United Kingdom; E-mail: [e-mail] 498 O ptic nerve head drusen (ONHD) are intra- and extracellular deposits that accumulate within the optic nerve head and often become calcified over time. They occur in 0.4% of children and often pose a diagnostic dilemma; they may simulate true disc edema (1,2). Diagnosis in the pediatric population may also be more difficult because of the buried nature of ONHD that is typical in childhood; these drusen subsequently become more superficial with age (3–5). This may contribute to the underdiagnosis of ONHD, as demonstrated by a relatively higher prevalence of 2.4% in an autopsy series (6). Accurate diagnosis is imperative to avoid the route of unnecessary invasive investigations for more serious etiologies of optic disc elevation such as papilledema (7). Although ONHD can be diagnosed based on clinical examination alone, ancillary testing is commonly performed to secure the diagnosis. Testing includes fundus photography, fluorescein angiography (FA), fundus autofluorescence (FAF), B-scan ultrasonography, computed tomography (CT) scan, and more recently, optical coherence tomography (OCT) (8). B-scan ultrasonography and FA are more reliable than CT and FAF but have low resolution and fall short in providing structural and topographical information (3,9). The sensitivity of B-scan ultrasonography may also be lower in children than in adults because this technique does not reliably detect buried, noncalcified ONHD (often the case in children) (10,11). OCT is a relatively new imaging modality for ONHD. Conventional spectral-domain (SD) OCT provides higher resolution images compared with B-scan ultrasonography (12,13). However, this advantage is limited with increasing depth; deeper ONHD have lower resolution, often resulting in poorer demarcation of their posterior borders (14,15). Newer OCT technologies, such as enhanced-depth imaging (EDI) and swept-source, circumvent this issue by improving the image quality of deeper structures; these have shown promise in detecting ONHD (16–18). Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution Given that much research to date has focused on adults, there is a scarcity of the literature reporting the utility of enhanced depth imaging optical coherence tomography (EDI-OCT) in the detection of ONHD in the pediatric population. In this study, we assessed the usefulness of EDI-OCT for diagnosing and characterizing ONHD in children compared with conventional imaging modalities. METHODS We conducted a retrospective, cross-sectional case series analysis of consecutive pediatric patients presenting to the pediatric eye clinic of Watford General Hospital (Watford, United Kingdom) with ONHD confirmed on B-scan ultrasonography. This study received institutional review board (IRB) approval from the West Hertfordshire Hospitals NHS Trust Research & Development Department. All procedures and data collection were conducted in accordance with the Declaration of Helsinki. Consecutive pediatric patients (age #16 years) with ONHD evaluated between May 2017 and May 2018 were identified using a treatment register. The main exclusion criteria were the coexistence of previous posterior segment intraocular surgery, history of ocular trauma, or systemic conditions that affected the optic nerve head structure or visual field. Children were initially referred by their community eye care providers for assessment of suspected papilledema due to blurred optic nerve head margins. The diagnosis of ONHD was established based on evaluation of the patient’s history and physical examination, as well as review of diagnostic testing, such as neuroimaging, if necessary. All patients underwent a comprehensive ophthalmic examination, including best-corrected visual acuity (BCVA), slit-lamp biomicroscopy, standard automated perimetry (Humphrey VF Analyzer, 24-3 SITA-Standard Strategy; Carl Zeiss Meditec, Inc, Dublin, CA), and dilated fundus examination. Ocular imaging including FAF, SD-OCT retinal nerve fiber layer (RNFL), and SDOCT of the optic nerve head in both conventional (nonEDI) and EDI modes was performed for all patients (Spectralis, Heidelberg Engineering GMbH, Dossenheim, Germany). For FAF, images were obtained using a confocal scanning laser ophthalmoscope (Heidelberg Retinal Angiograph 2 [HRA2]; Heidelberg Engineering) with a view mode of 30°. Autofluorescence was performed using blue light excitation (488 nm), and emission was detected using a barrier filter of 500 nm. A standard procedure for FAF was used, including aligning and focusing of the optic disc image in the infrared reflection mode at 820 nm. Illumination was changed to the FAF mode once a sharply focused image and a mean of 9 frames were obtained. Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 For SD-OCT, both modes (non-EDI and EDI) were set to image a 15 · 15° rectangle for serial horizontal and vertical scans centered on the optic disc. Each rectangle was scanned with 73 sections, and 9 frames were averaged for each section. Based on the consensus of previous studies (14,16,17,19,20), ONHD masses on SD-OCT were defined as 1) hypo-reflective (signal-poor) round or ovoid structures bordered by hyper-reflective margins or 2) isolated or clustered hyper-reflective bands without a core. All cases of ONHD in this series were subsequently confirmed through B-scan ultrasonography performed by an experienced observer at a tertiary hospital (Moorfields Eye Hospital, London, United Kingdom). A standard protocol using axial, transverse, and longitudinal approaches was used. Scanning was considered positive for ONHD when a hyperechoic lesion within the optic nerve head was detected and remained after gain was reduced by low-gain settings. An average of 4 still images focused on the optic nerve were captured. Multimodal images were independently assessed by 2 experienced attending ophthalmologists (defined as having 7 years of experience or more after residency training) who had no role in the study design, conduct, data analysis, or patient care. Apart from knowledge of inclusion criteria (i.e., ONHD presence on B-scan ultrasonography), no other clinical information or context was provided to the masked image examiners. The ONHD detection rates of EDI-OCT vs non–EDI-OCT and FAF were compared with the pairwise McNemar’s; kappa analyses were used to test for interobserver agreement. Associations of mean ONHD diameters and clinical variables (age, BCVA, and RNFL thickness) were evaluated using Spearman rank correlations; Bonferroni corrections were performed where appropriate. A P value of ,0.05 was considered statistically significant. Data analyses were performed using IBM SPSS, version 23 (IBM Corp, Armonk, NY). RESULTS Fifteen patients (60% female) with definite ONHD, confirmed by B-scan ultrasound in one or both eyes, were included. The mean age was 11 years (range 7–16). The median best-corrected visual acuity (logMAR unit) was 20.1 (range 20.2 to 0.3). Of the 15 patients, ONHD was bilateral in 13 and unilateral in the remaining 2, yielding 28 eyes for analysis. Of the 28 eyes, EDI-OCT, non–EDI-OCT, and FAF revealed ONHD in 24, 21, and 18 eyes, respectively; RNFL thickness by OCT and visual field was abnormal in 3 and 4 eyes, respectively (Fig. 1). A combination of both hypo‐reflective round/ovoid masses and hyper‐reflective bands was observed (Fig. 2). Both horizontal and vertical scans confirmed that these structures seen were at the same location. ONHD detected by non–EDI-OCT had 499 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Multimodal imaging—FAF image (top row, left), B-scan ultrasound (top row, middle; white arrow), RNFL (top row, right) and EDI-OCT (bottom row; asterisk, and white arrowheads) images showing ONHD. EDI-OCT, enhanced depth imaging optical coherence tomography; FAF, fundus autofluorescence; ONHD, optic nerve head drusen; RNFL, retinal nerve fiber layer. mostly similar features compared with EDI-OCT except for the posterior surface of ONHD masses that were often invisible or less clear (Fig. 3). EDI-OCT detected ONHD in 24/28 eyes identified as having drusen on ultrasound B-scan. There was no significant difference in the detection of ONHD between EDI-OCT and non–EDI-OCT (24/28 eyes vs 21/28 eyes; P = 0.15). EDI-OCT had significantly better detection capacity compared with FAF (24/28 eyes vs 18/28 eyes; P = 0.01). Adjusting for inter‐eye correlation using the general estimating equation (GEE) approach, this remained statistically significant (P = 0.04). Using kappa statistics (Table 1), there was strong interobserver agreement for the detection of ONHD on EDI-OCT (k = 0.83; CI 0.78–0.96). Moderate interobserver agreement was observed for the detection of ONHD on non–EDI-OCT (k = 0.57; CI 0.46–0.72) and FAF (k = 0.46 CI 0.32–0.59). The mean greatest FIG. 2. EDI-OCT images showing a large ovoid region of hyporeflectivity (top row; asterisk, and white arrowheads) and hyperreflective foci (bottom row; white arrows). The green lines are the scanned lines of the EDI-OCT images. EDI-OCT, enhanced depth imaging optical coherence tomography. 500 Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 3. A representative case of clinically definitive optic nerve head drusen (ONHD). Enhanced depth imaging optical coherence tomography (top row) visualized the entire ONHD mass (white arrowheads), whereas non–EDI-OCT (bottom row) showed mostly its anterior (superficial) aspect. OCT, optical coherence tomography. diameter of ONHD detected on EDI-OCT was 449.7 (±114.1) mm. Of the 28 eyes, FAF detected ONHD in 18 (64%) in regions that colocalized with the hyperreflective ONHD mass seen using EDI-OCT. Of interest, 2 patients had no apparent focal autofluorescence signaling from their fellow eye, but when examined with EDI-OCT, they had detectable ONHD, consistent in appearance with the larger drusen seen in the fellow eyes. Using Spearman rank-correlation coefficients, there were no significant correlations between the mean greatest diameter of ONHD measured on EDI-OCT with age (rs = 0.06, P = 0.7), best-corrected visual acuities (rs = 20.24, P = 0.3), and the mean global RNFL thicknesses (rs = 20.06, P = 0.8). There was also no significant correlation between the mean greatest ONHD diameter with RNFL thickness in the 4 different quadrants: superior (rs = 20.02, P = 0.9), nasal (rs = 0.04, P = 0.8), inferior (rs = 0.5, P = 0.06), and temporal (rs = 0.45, P = 0.09). DISCUSSION In this study, we demonstrated that EDI-OCT detected ONHD in a majority (86%) of children’s eyes identified as having drusen on B-scan ultrasonography. This is indirectly corroborated by previous studies that have primarily focused on adult populations, showing equivalent, if not better, diagnostic detection of EDI-OCT for ONHD than B-scan ultrasonography (16,17). Nonetheless, in contrast to adult studies, it is important to note that EDI-OCT failed to detect all cases of ONHD that were preidentified by B-scan ultrasonography in our study. A likely explanation for this is the higher incidence of buried ONHD in children (3–5). This is supported by a previous study that has reported the mean age at which drusen become superficial and visible to be 12 years (4), which is slightly higher than that of our cohort. Diagnosing buried ONHD can be more difficult in children, as it often resembles other causes of optic disc elevation such as papilledema, which subject patients to unnecessary invasive TABLE 1. Kappa coefficient calculation using 2-rater data samples Rater 2 ONHD present ONHD absent Row marginals 24 0 24 1 3 4 25 3 28 Rater 1 ONHD present ONHD absent Column marginals Kappa statistics (k) = 0.83 (CI 0.78–0.96). CI, confidence interval; ONHD, optic nerve head drusen. Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 501 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution tests. With B-scan ultrasonography, buried and calcified ONHD can be reliably detected as hyperechoic, highly reflective round structures with posterior acoustic shadowing (21). For this reason, this noninvasive technique is still widely regarded as the “gold standard” test for the detection of buried ONHD (21). Despite B-scan ultrasonography being superior in this front, it has 2 main challenges: 1) detection of ONHD is dependent on the degree of its calcification, making it unreliable for uncalcified ONHD and 2) its inability to provide high resolution images of ONHD structures and important structural information such as RNFL thickness or retinal ganglion cell population (22). These limitations are now circumvented with the advent of EDI-OCT, which enables deeper penetration and visualization than conventional OCT (22). Correspondingly, using EDI-OCT, we were able to visualize and measure ONHD in more detail as well as perform quantitative assessment of RNFL in children. Previous studies have indicated ONHD to be bilateral in 75%–95% cases (21,23,24). This is in agreement with our results, highlighting bilaterality as a feature of ONHD that can be discerned from early childhood years. The distribution and morphological characteristics of ONHD observed in our study are also similar to those of the adult population in which ONHD have been predominantly detected on the nasal side of the optic disc (17) and described as signal-poor core areas surrounded by hyperreflective margins (14,16,17,20,25). The prevalence of visual field defects in patients with ONHD has been described with considerable variability in the published literature, ranging from 11% to 51% (4,26). This is largely due to the heterogenous age population being studied, with visual field defects being more commonly observed in cases of superficial compared with buried ONHD (27). A similar situation also exists with RNFL thickness (a structural surrogate of ONHD) in which more prominent RNFL thinning is seen in eyes with superficial ONHD (28). Therefore, it is likely that the low prevalence of visual field defects and RNFL thinning (14% and 11% respectively) observed in our study reflects the higher proportion of buried ONHD in children. In line with previous studies reporting a gradual increase in size of drusen with age (3), we found the average ONHD size in children to be smaller than that reported for adults (449.7 mm vs 686.8 mm respectively) (17). Interestingly, an observational case series by Malmqvist et al (29) showed only minimal progression in size of drusen between early adulthood and old age. Based on their findings, they suggest a transition phase during adolescence in which there is more significant evolution in the size and location of drusen. This hypothesis would account for the difference in OHND size seen in our cohort compared with adults, although further studies are needed to verify this observation. There are several limitations, the most important of which being the retrospective and cross-sectional nature of our study. Owing to the aforementioned reasons, larger 502 longitudinal cohort studies are needed to investigate any dynamic evolution of OHND from childhood to early adulthood, as well as establish any meaningful temporal relationship between OHND size/location and RNFL and visual field defects. Second, our evaluation of interobserver agreement involved only a limited number of (2) examiners with extensive ophthalmology experience. This could be improved on by recruiting a larger number of ophthalmologists, from different institutions, with varying ophthalmic subspecialties and years of experience. In conclusion, EDI-OCT is able to detect ONHD in a majority of eyes identified as having drusen on B-scan ultrasonography in children. We also show that EDI-OCT can achieve a high interobserver agreement on the diagnosis of ONHD diagnosis as well as provide valuable quantitative information about ONHD topography, structure and size, and their relation to our structures such as RNFL thickness. With the rapid advancement of imaging technology, EDIOCT has the potential to not only be a new, less invasive, diagnostic gold standard for ONHD but also a monitoring tool to establish a structural–functional paradigm of ONHD progression in children. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: F. Barampouti and M. Karampelas; b. Acquisition of data: P. Y. Sim and H. Soomro; c. Analysis and interpretation of data: P. Y. Sim and H. Soomro. Category 2: a. Drafting the manuscript: P. Y. Sim and H. Soomro; b. Revising it for intellectual content: F. Barampouti and M. Karampelas. Category 3: a. Final approval of the completed manuscript: P. Y. Sim, H. Soomro, F. Barampouti, and M. Karampelas. ACKNOWLEDGMENTS The authors acknowledge the assistance of Gianfranco Ventura (ophthalmic photographer) with ophthalmic imaging and data collection. REFERENCES 1. Fong CY, Williams C, Pople IK, Jardine PE. Optic disc drusen masquerading as papilloedema. Arch Dis Child. 2010;95:629. 2. Hu K, Davis A, O’Sullivan E. Distinguishing optic disc drusen from papilloedema. BMJ. 2008;337:a2360. 3. Auw-Haedrich C, Staubach F, Witschel H. Optic disk drusen. Surv Ophthalmol. 2002;47:515–532. 4. Hoover D, Robb R, Peterson R. Optic disc drusen in children. J Pediatr Ophthalmol Strabismus. 1988;25:191–195. 5. Frisén L. Evolution of drusen of the optic nerve head over 23 years. Acta Ophthalmol. 2008;86:111–112. 6. Friedman AH, Gartner S, Modi SS. Drusen of the optic disc. A retrospective study in cadaver eyes. Br J Ophthalmol. 1975;59:413–421. 7. Leon M, Hutchinson AK, Lenhart PD, Lambert SR. The costeffectiveness of different strategies to evaluate optic disk drusen in children. J Am Assoc Pediatr Ophthalmol Strabismus. 2014;18:449–452. 8. Chang MY, Pineles SL. Optic disk drusen in children. Surv Ophthalmol. 2016;61:745–758. Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 9. Chang MY, Velez FG, Demer JL, Bonelli L, Quiros PA, Arnold AC, Sadun AA, Pineles SL. Accuracy of diagnostic imaging modalities for classifying pediatric eyes as papilledema versus pseudopapilledema. Ophthalmology. 2017;124:1839–1848. 10. Kurz-Levin MM, Landau K. A comparison of imaging techniques for diagnosing drusen of the optic nerve head. Arch Ophthalmol. 1999;117:1045–1049. 11. Petrushkin H, Ali N, Restori M, Adams GGW. Development of optic disc drusen in familial pseudopapilloedema: a paediatric case series. Eye. 2011;25:1101–1102. 12. Johnson LN, Diehl ML, Hamm CW, Sommerville DN, Petroski GF. Differentiating optic disc edema from optic nerve head drusen on optical coherence tomography. Arch Ophthalmol. 2009;127:45–49. 13. Murthy RK, Storm L, Grover S, Brar VS, Chalam KV. In-vivo high resolution imaging of optic nerve head drusen using spectraldomain Optical Coherence Tomography. BMC Med Imaging. 2010;10:2008–2010. 14. Yi K, Mujat M, Sun W, Burnes D, Latina MA, Lin DT, Deschler DG, Rubin PA, Park BH, de Boer JF, Chen TC. Imaging of optic nerve head drusen: improvements with spectral domain optical coherence tomography. J Glaucoma. 2009;18:373–378. 15. Wester ST, Fantes FE, Lam BL, Anderson DR, McSoley JJ, Knighton RW. Characteristics of optic nerve head drusen on optical coherence tomography images. Ophthalmic Surg Lasers Imaging. 2010;41:83–90. 16. Merchant KY, Su D, Park SC, Qayum S, Banik R, Liebmann JM, Ritch R. Enhanced depth imaging optical coherence tomography of optic nerve head drusen. Ophthalmology. 2013;120:1409–1414. 17. Sato T, Mrejen S, Spaide RF. Multimodal imaging of optic disc drusen. Am J Ophthalmol. 2013;156:275–282.e1. 18. Silverman AL, Tatham AJ, Medeiros FA, Weinreb RN. Assessment of optic nerve head Drusen using enhanced depth imaging and swept source optical coherence tomography. J Neuroophthalmology. 2014;34:198–205. 19. Malmqvist L, Bursztyn L, Costello F, Digre K, Fraser JA, Fraser C, Katz B, Lawlor M, Petzold A, Sibony P, Warner J, Wegener M, Sim et al: J Neuro-Ophthalmol 2020; 40: 498-503 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Wong S, Hamann S. The optic disc drusen studies consortium recommendations for diagnosis of optic disc drusen using optical coherence tomography. J Neuroophthalmology. 2018;38:299–307. Slotnick S, Sherman J. Buried disc drusen have hypo-reflective appearance on SD-OCT. Optom Vis Sci. 2012;89:704–708. Hamann S, Malmqvist L, Costello F. Optic disc drusen: understanding an old problem from a new perspective. Acta Ophthalmol. 2018;1:673–684. Chen JJ, Costello F. The role of optical coherence tomography in neuro-ophthalmology. Ann Eye Sci. 2018;3:35. Rebolleda G, Kawasaki A, de Juan V, Oblanca N, MuñozNegrete FJ. Optical coherence tomography to differentiate papilledema from pseudopapilledema. Curr Neurol Neurosci Rep. 2017;17:74. Malmqvist L, Lindberg ASW, Dahl VA, Jørgensen TM, Hamann S. Quantitatively measured anatomic location and volume of optic disc drusen: an enhanced depth imaging optical coherence tomography study. Investig Ophthalmol Vis Sci. 2017;58:2491–2497. Ghassibi MP, Chien JL, Abumasmah RK, Liebmann JM, Ritch R, Park SC. Optic nerve head drusen prevalence and associated factors in clinically normal subjects measured using optical coherence tomography. Ophthalmology. 2017;124:320–325. Erkkilä H. Optic disc drusen in children. Acta Ophthalmol. 1977;129:3–44. Katz BJ, Pomeranz HD. Visual field defects and retinal nerve fiber layer defects in eyes with buried optic nerve drusen. Am J Ophthalmol. 2006;141:248–253. Malmqvist L, Wegener M, Sander BA, Hamann S. Peripapillary retinal nerve fiber layer thickness corresponds to drusen location and extent of visual field defects in superficial and buried optic disc drusen. J Neuroophthalmology. 2016;36:41– 45. Malmqvist L, Lund-Andersen H, Hamann S. Long-term evolution of superficial optic disc drusen. Acta Ophthalmol. 2017;95:352–356. 503 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.