Delhi Journal of Ophthalmology

Challenges In Optical Coherence Tomography Interpretation For Diagnosing Glaucoma In Myopic Eyes

Kirti Singh
MD, DNB, FRCS, FAIMER
Dir Professor, Glaucoma Services,
Guru Nanak Eye Centre, Maulana Azad Medical College, New Delhi

Corresponding Author:

Kirti Singh
MD, DNB, FRCS, FAIMER
Dir Professor, Glaucoma Services,
Guru Nanak Eye Centre, Maulana Azad Medical College, New Delhi

Published Online: 20-DEC-2020

DOI:http://dx.doi.org/10.7869/djo.584

Abstract

Keywords :

Optical coherence tomography has witnessed a rapidly evolving technology since its debut by David Huang in 1991, to its present finesse. The non- invasive nature of this imaging tool coupled with speed, reliability and test re-test reproducibility in analyzing structural damage has made it the cornerstone imaging device for glaucoma. Intricate and precise segmentation algorithms delineating structure have further enhanced disease understanding and paradigm of glaucoma management .Increasing prevalence of myopia worldwide with 80- 90% young adults of Asia being affected, one fifth of them with high myopia, is generating a population who would be screened for glaucoma at some time of their life.1,2 Since the segmentation algorithms, so critical for glaucoma diagnosis, are tailored for normal retinal anatomy, variations from this normal in form of high refractive errors raises pertinent concerns regarding accuracy and reliability of OCT in these cases. This concern is underscored by evidence that myopes are 2-3 times at risk for developing glaucoma despite correcting for axial length and intraocular pressure. As this myopia cohort of young adults age, their proportion as glaucoma suspects would rise exponentially over coming decades. Understanding relevant optics and anatomical variations affecting interpretation of OCT in myopia, is thus essential to prevent mistaken diagnosis of red (false positive) or green (false negative) disease.

Optical coherence tomography uses optical interferometry to measure time delay of reflected light in near infra-red visible spectrum (810 - 1056 nm) and is therefore affected by refractive status and media opacities. Standard scanning protocols used for glaucoma diagnosis are - Circum-papillary Retinal nerve fibre layer quadrant scan (cp RNFL), Optic nerve head scan (ONH), Macular scan (macular ganglion cell complex, mGCC), Glaucoma progression analysis, with some new machines incorporating disc damage likelihood scale (DDLS) nomograms.

Segmentation software algorithms (layer seeking algorithms) to detect glaucoma have been tailored based on anatomy of retinal nerve fibre at disc and macula. Axons of ganglion cells (Retinal nerve fibres) arch in the macular area to converge on superior and inferior poles of disc. The circum-papillary Retinal nerve fibre layer scan (cpRNFL) protocol identifies  and plots RNFL thickness starting from temporal to nasal quadrant in each eye following the Temporal, Superior, Nasal, Inferior and Temporal (TSNIT) sequence. This is depicted as the double hump pattern in the TSNIT plot (Figure 1). Macular cube protocol scans a 6 × 6 mm2 area, measuring Retinal nerve fibre layer (axons) + Ganglion cell body + Inner plexiform layer IPL (dendrites), collectively called ganglion cell complex (mGCC). The central area, the foveola, which is devoid of ganglion cells is masked in the macula scan. (Figure 1). Probability levels for RNFL thickness are then displayed on four colour scale with red depicting thickness below thinnest 1 % percentile, yellow for measures within thinnest 1-5% of normative database, green for within 95% percentile and white for values beyond 95% percentile. This colour coding is arrived from population derived normative database involving adult patients over 18 year age and few races.


Figure 1: The Optic disc cp RNFL 1b. Macular cube (macular ganglion cell complex) scanning protocol for glaucoma diagnosis 


Myopic challenge in OCT interpretation
Axial myopia results in stretching of globe and thinning of retina. This thinning manifests as thinned retinal nerve fibre layer.3,4 Current machines do not incorporate normative data from eyes with high refractive error in their internal database, therefore scans of myopic eyes with such thinned out RNFL, are interpreted as pathological (red disease). Most machines have only data from patients with emmetropia or low myopia incorporated.

Red Disease
This is the sobriquet given to false positive errors in interpreting OCT test, with the non - glaucomatous patient being erroneously labelled as glaucoma. Causes of red disease in myopia can be due to: from poor centration of measurement circle on optic disc due to poor fixation by patient, lack of normative database for high refractive errors, altered distribution angle of RNFL with temporal crowding of nerve fibres, magnification factor in axial myopia, split and shifted RNFL peaks to disc size/ shape variations and lamina cribrosa biomechanics. The following text details these variations in myopic fundus in conjunction with screening protocol used, for ease of understanding their impact on interpreting scans.

1. Disc Size/ Shape Variations
The disc parameters studied Optic nerve head protocol are vertical cup disc ratio, rim area and mean RNFL thickness. Segmentation algorithms inbuilt in OCT software are designed for optic disc sizes and shapes within the Gaussian curve of emmetropic or low myopic populations with disc area between 1.3 - 2.5 mm2. Discs larger or smaller than this, make interpretation beyond scope of normative database, with erroneous labelling of pathological disease. For such discs, common in myopia, the measurement circle placement need to be checked and manually placed, prior to interpretation.

Disc margin on OCT is identified, as termination of Bruch’s membrane opening (BMO) and neuro-retinal rim is measured as the shortest perpendicular from that point to internal limiting membrane, called Minimum rim width (MRW).5 Tilted or dysplastic disc, often associated with myopia, would pose difficulties in identification of Bruch’s Membrane Opening BMO, the OCT landmark for disc margin identification, with subsequent erroneous rim measurement. (Figure 2) Myopic discs larger than 3.46 mm, would be erroneously labelled as damaged, since more distal dispersed peripapillary RNFL would get imaged. However there is a caveat to this, as evidence exists that circumpapillary RNFL thickness is positively correlated with optic disc size, with smaller discs having less and larger discs having more nerve fibres.6 This is in contravention to accepted anatomy of nerve fibres being constant in all sized discs.


Figure 2: OCT optic disc cube. The red line marks the internal limiting membrane and yellow the Retinal nerve fibre layer. The pink dot is the Bruch’s membrane opening (BMO) which identifies the optic disc

2. Circumpapillary RNFL Thinning
Retinal nerve fibres are both actually and artifactually thinner in myopia. Let us understand this. Axial myopia results in scleral stretching with ganglion cell axons spreading out to cover the wider globe area, therefore resulting in actual diffuse thinning of RNFL. This reduction in RNFL thickness with increasing axial length, is reported to be in the range of 2.2 -3.7 microns /mm. or 1.30 µm for every 1 diopter sphere increase in negative SE.7 Indian subcontinent data has verified thinned RNFL at 78.7 + 5.7 µ in high myopia, compared with 83.8 + 3.4 µ in moderate myopia and 91.3 + 2.99 in emmetropic eyes.8,9

3. Magnification Factor while scanning ONH
Increasing axial length, results in expanding circumference of cpRNFL measuring circle from 3.46 mm, due to magnification factor, especially for myopia greater than - 4 D.10 A myopic eye would result in increase of 3.46 mm scan circle, to 4.0 mm for a 28 mm axial length myopia.11 A longer myopic eye, generating an expanded measurement circle, measures RNFL at distance further from optic disc (Figure 3). Since nerve fibres fan out distal to optic disc, it would make nerve fibres appear more dispersed and artefactually less in number. To counteract this Littmann’s magnification factor correction needs to be applied for longer eyes. Some authors do not however ascribe to use of this magnification factor citing occurrence of larger discs in myopia, which in any case have diluted the effect of circle expansion.8 


Figure 3: Magnification of measuring circle in OCT commensurate with increasing axial length

4. Retinal nerve fibre layer distribution Pattern / Temporal Displacement Of RNFL Peaks
Volume-rendering MRI reveal globe shape in emmetropic eyes to be consistently spherical, with myopic eyes having symmetrical or asymmetrical anteroposterior elongation and posterior protrusions. With increasing myopia this asymmetrical anteroposterior elongation makes superotemporal (ST) and inferotemporal (IT) RNFL bundles converge temporally due to retina being dragged temporal to disc. The condition called as “Temporal displacement or convergence, reduces distribution angle bounded by ST and IT bundles and displaces RNFL peaks temporally, resulting in erroneous interpretation of thickened temporal RNFL and thinned nasal RNFL.7 (Figure 4) This temporal crowding in myopic eyes, could either be a result of an anatomical variation or image artefact due to increased vertical curvature of retina.12 The RNFL distribution angle has been shown to be negatively associated with axial length, with a 3.4 decrease for every 1-mm increase in axial length, and positively associated with spherical error, with a 2.7 increase for every 1.0 D increase in spherical error.7 This negative association between area of abnormal RNFL measurement and RNFL distribution angle persists even after adjustment of axial length and spherical error. This finding implies, that even if normative database including high myopia is incorporated in machine software, RNFL thickness analysis would be erroneous in eyes with a small RNFL distribution angle.


Figure 4: Temporal displacement of RNFL peaks, resulting in thickening of RNFL in temporal quadrants, apparent thinning in superior & inferior quadrants in Right eye. Reproduced from Akman A, Bayer A, Nouri-Mahadavi K. In Optical Coherence tomography in glaucoma pg 201

5. RNFL split / segmentation 
Split or segmented RNFL are anatomical variants of the RNFL bundles entering poles of disc. This variation initially identified as an artifact by the earlier GDx machine, is often seen in high myopes with peripapillary atrophy (PPA). In this condition retinal nerve fibres converging on superior and inferior poles of disc do not enter as complete bundles, but instead split in two. The space between two bundles masquerades as a local RNFL defect (Figure 5).


Figure 5: Split nerve fibre bundle in left eye superior pole. The split is still within the 95% perecentile, it is falls short of manifesting as red diseases

6. Peri-papillary atrophy (PPA) 
The presence of peripapillary atrophy with alpha and beta zones, has long been considered soft sign of glaucomatous disc damage and often known as halo glaucomatosus. Three zones namely alpha, beta and gamma have been identified in PPA. Alpha zone a is due to hyper or hypopigmentation (usually former) of RPE at edge of PPA along with slight thinning of choroidal tissues. Beta zone ß, present between a zone and disc margin, is visible as a white grey scleral crescent with loss of RPE, marked atrophy of photoreceptor layer and choriocapillaries. It is characterized by visibility of large choroidal vessels and sclera. Refined imaging by OCT advances, has identified two subtype of ß PPA namely beta zone PPA and gamma zone PPA in some myopic eyes. Gamma zone ?, previously called as myopic temporal crescent, is white coloured area interspersed between beta zone and optic disc border. Gamma zone differs from beta zone in lacking a Bruch’s membrane. Since OCT algorithms identify disc margin as Bruch’s membrane termination /opening, the presence of gamma zone ? would spuriously enlarge disc area in image interpretation. (Figure 6)


Figure 6:  a & 6 b Peripheral pigmented part of peripapillary atrophy (infero-temporal to disc) demarcated in pink is a zone


Figure 6: 6c, 6d & 6e Peripapillary atrophy , whitish grey zone with RPE loss, retinal and choriocapillary atrophy demarcated in green is ß zone and that in blue could be ? zone ( which is essentially an OCT finding)

Peri-papillary atrophy is often present in myopic retina and results in discrimination errors to differentiate glaucomatous damage in myopic disc from a non-diseased myopic optic nerve head. Presence of PPA in addition, makes placement of OD circle with subsequent measurement of disc and RNFL parameters difficult. The scanning circle diameter for measuring optic disc in OCT machines is 3.46 -3.5 mm, which may fail to traverse extent of the PPA. Presence of large PPA extending beyond scanning circle diameter, would result in RNFL measured from beta zone outward, resulting in erroneously low measures (red disease). (Figure 7 and 8) New versions of Spectralis have options of 4.1 and 4.7 mm circles, to bypass the of area peripapillary atrophy.13
Larger ß zone has been identified as a significant marker of glaucomatous damage. A recent report released from DIGS (Diagnostic Innovations of glaucoma study) states that the odds of having glaucoma get doubled for each 0.2 mm2 larger  ß PPA area, after controlling for age, central corneal thickness and axial length.14 On the other hand larger ? zone is due  to scleral stretching with myopic changes and is more often associated with healthy myopic, non-glaucomatous eyes.15,16 Advances in measurement of ß and ? zone have resulted in development of SALSA algorithms (San - Diego Automated Segmentation Algorithms), which measures areas of ß PPA with intact Bruch’s membrane, as surrogate indicators of glaucomatous damage.


Figure 7: Examples of Peripapillary atrophy 7a. RNFL calculation circle incorporates the PPA and therefore depicts less RNFL measure.7b Tilted disc with inferior PPA, the RCNFL calculation circle falls short of PPA and therefore depicts less RNFL measure. In addition the oval disc margin is not completely covered by Disc circle


Figure 8: Large peripapillary atrophy with posterior staphyloma. The cp RNFL print out gives erroneous results due to inability of measurement circle to stretch beyond the large PPA area.

Green disease
Myopic eyes suspected of glaucoma sometimes would be erroneously labelled as normal. This is seen in situations of epiretinal membrane, vitreoretinal tractions, Weiss ring or Internal limiting membrane (ILM) traction over optic nerve head, both causing artefactual increase in RNFL measurement (green disease).17 (Figure 9) However, once total posterior vitreous detachment occurs, the membrane would obscure light reflectance and that could cause a localized red disease. In Peri-papillary retinoschisis conditions, a transient increase in RNFL thickness is noted, again a cause of green disease.


Figure 9: Internal limiting membrane traction, resulting in artefactual increase in RNFL peak, Green disease

Macular cube: Macular Ganglion cell complex (mGCC)
Ganglion cell death is the hallmark of glaucomatous damage and macula has the unique position of containing 50% of ocular retinal ganglion cells. Due to the above defined confounders while measuring RNFL and optic disc parameters in the myopic suspect eye, macular GCC measurement has been explored as a more robust option to discriminate glaucoma patients from non-glaucoma subjects. Macular ganglion cell complex (mGCC) protocol measures ganglion cell, their dendrites (inner plexiform layer) and their axons (retinal nerve fibre layer). Some machines exclude RNFL and measure ganglion cell with inner plexiform layer (GC IPL). (Figure 10) A 6 by 6 mm2 area centered on macula is scanned , of which central elliptical region of 1.2 x 1.0 mm, represents foveola (which lacks ganglion cells) being masked. (Figure 11)18 Being less affected by anatomical vagaries in myopic eyes, macular scan has been established to have higher diagnostic power than cpRNFL parameters in detecting glaucoma in these patients.19,20,21 However some studies have disputed this with both macular and optic disc protocols having equivalent diagnostic efficacy in myopic eyes.22 Literature is equivocal on macular thickness in myopic eyes, with some authors reporting thicker macula23 with others documenting no change.24 In both situations a myopic normative database for ganglion cell internal plexiform layer (GCIPL) would improve glaucoma detection rate in myopic eyes.25 In myopic eyes with macular thining/hole, posterior staphyloma and Foster Fuchs however, the macular cube scanning protocol should not be attempted. (Figure 11)
The commonest challenge faced in discriminating glaucoma in myopic eyes, remains lack of myopic normative data. Diagnostic Innovations in Glaucoma Study (DIGS 1995) and African Descent And Glaucoma Evaluation Study (ADAGES 2002) are longitudinal cohort studies designed to study participants with or without myopia and rectify this gap in imaging technology. Structural, functional and vascular testing by Spectral OCT, in Caucasian and African population, seeks to generate a reference database of healthy myopic and glaucomatous myopic eyes.26
Recent advances in OCT use extended depth imaging and adaptive optics have been used to study lamina cribrosa structure.27 In myopic eyes, documenting changes in peripapillary sclera and lamina cribrosa dimensions may help in diagnosing glaucoma early.


Figure 10: 10a Macular cube protocol measures etinal nerve fibre layer & ganglion cells and inner plexiform layer 10b. Ganglion cell complex OU print out 


Figure 11:  11a. Pathological myopia with large annular crescent and Foster Fuchs spot, 11b Macular ganglion cell analysis protocol of same case, does not give any meaningful interpretation 


Conclusion 
Optical coherence tomography has become the imaging device of choice in diagnosing glaucoma. Its use in high to moderate myopia is limited by various anatomical alterations in myopic eye affecting scan acquisition and interpretation.28 Knowledge of this challenging aspect of OCT is essential for the ophthalmologist keeping in mind the exponential increase in myopia globally. This would reduce the erroneous labelling of non-diseased myopic as glaucoma and also missing the pathology in myopic eyes.

References
  1. Pei-Chang W, Hsiu-Mei H, Hun-Ju Y, Po-Chiung F, Chueh-Tan C Epidemiology of Myopia. Asia-Pacific Journal of Ophthalmology 2016, 5 (6): 386-93
  2. Morgan IG, Ohno Matsui K and Saw SM: Myopia. Lancet 2012, 379: 1739 48
  3. Schuman JS. Effect of refractive errors on RNFL thickness measures. JAMA Ophthalmol. 2016, 134(9): 1040.
  4. Joel S. Schuman JS, Pedut-Kloizman T, Pakter H, Wang N, Guedes V, Huang L et al. Optical Coherence Tomography and Histologic Measurements of Nerve Fiber Layer Thickness in Normal and Glaucomatous Monkey Eyes. Invest Ophthalmol Vis Sci. 2007;48: 3645–54
  5. Akman A. Role of Optical coherence tomography in glaucoma. In Optical coherence tomography in glaucoma. Eds. Akman A, Bayer A, Nouri Mahdavi. Springer international Publishing AF-GM Cham Switzerland, 2018, pg 20
  6. Schweitzer KD, Ehmann D and Garcia R: Nerve fibre layer changes in highly myopic eyes by optical coherence tomography. Can J Ophthalmol 2009. 44: e13 e16,
  7. Leung, C KS, Yu M, Weinreb RN, Mak HK, Lai G, Ye C and Lam DSC Retinal Nerve Fiber Layer Imaging with Spectral-Domain Optical Coherence Tomography: Interpreting the RNFL Maps in Healthy Myopic Eyes. Invest Ophthalmol Vis Sci. 2012; 53:7194–7200
  8. Singh D, Mishra SK, Agarwal E , Sharma R , Bhartiya S, Dada T. Assessment of Retinal Nerve Fiber Layer Changes by Cirrus High-definition Optical Coherence Tomography in Myopia. J Curr Glaucoma Pract 2017;11 (2):52-57
  9. Sihota R, Sony P, Gupta V, Dada T and Singh R Diagnostic Capability of Optical Coherence Tomography in Evaluating the Degree of Glaucomatous Retinal Nerve Fiber Damage. Invest Ophthalmol Vis Sci. 2006; 47:2006–2010
  10. Bae SH, Kang SH, Feng CS, Park J, Jeong JH, Yi K. Influence of myopia on size of optic nerve head and retinal nerve fiber layer thickness measured by spectral domain optical coherence tomography. Korean J Ophthalmol. 2016; 30: 335–43.
  11. Savini G, Barboni P, Parisi V, Carbonelli M. The influence of axial length on retinal nerve fibre layer thickness and optic-disc size measurements by spectral-domain OCT. Br J Ophthalmol. 2012; 96:57–61
  12. Zha Y, Zhuang J, Lin D , Feng W , Zheng H and Cai J . Evaluation of myopia on retinal nerve fiber layer thickness measured by Spectralis optical coherence tomography. Exp & Therapeutic Med 2017, 14: 2716-2720
  13. Akman A. Role of Optical coherence tomography in glaucoma. In Optical coherence tomography in glaucoma. Eds. Akman A, Bayer A, Nouri Mahdavi. Springer international Publishing AF-GM Cham Switzerland, 2018, pg 206
  14. Manalastas PIC, Belghith A, Weinreb RN, Jonas J, Suh MH, Yarmohammadi A, Medeiros FA, Girkin CA, Liebmann, JM and Zangwill LM. Automated Beta Zone Parapapillary Area Measurement to Differentiate Between Healthy and Glaucoma Eyes. Am J Ophthalmol. 2018; 191: 140–148.
  15. Jonas JB, Jonas SB, Jonas RA, Holbach L, Dai Y, Sun X, Panda-Jonas S. Parapapillary atrophy: histological gamma zone and delta zone. PLoS One. 2012;7(10):e47237.
  16. Dai Y, Jonas JB, Huang H, Wang M, Sun X. Microstructure of parapapillary atrophy: beta zone and gamma zone. Invest Ophthalmol Vis Sci. 2013; 54:2013–8.
  17. Akman A. Role of Optical coherence tomography in glaucoma. In Optical coherence tomography in glaucoma. Eds. Akman A, Bayer A, Nouri Mahdavi. Springer international Publishing AF-GM Cham Switzerland, 2018, pg 140
  18. Dharwadkar S and Nayak BK. Optical coherence tomography in glaucoma -I. Journal of clinical ophthalmology and Research 2017, 5 (1): 51-63
  19. Wang WW, Wang HZ , Liu JR , Zhang XF , Li M, Huo YJ , and Yang XG Diagnostic ability of ganglion cell complex thickness to detect glaucoma in high myopia eyes by Fourier domain optical coherence tomography. Am J Ophthalmol 2018; 11(5): 791–796.
  20. Zhang Y, Wen W, Sun X. Comparison of Several Parameters in Two Optical Coherence Tomography Systems for Detecting Glaucomatous Defects in High Myopia. Invest Ophthalmol Vis Sci. 2016 Sep 1; 57(11):4910-4915.
  21. Shoji T, Nagaoka Y, Sato H, Chihara E. Impact of high myopia on the performance of SD-OCT parameters to detect glaucoma. Graefes Arch Clin Exp Ophthalmol. 2012; 250 (12):1843-9
  22. Akashi A, Kanamori A, Nakamura M, Fujihara M, Yamada Y, Negi A. The ability of macular parameters and circumpapillary retinal nerve fiber layer by three SD-OCT instruments to diagnose highly myopic glaucoma. Invest Ophthalmol Vis Sci.2013, 5; 54(9):6025-32.
  23. Choi SW, MD, Lee SJ. Thickness Changes in the Fovea and Peripapillary Retinal Nerve Fiber Layer Depend on the Degree of Myopia. Korean Journal of Ophthalmology 2006, 20(4):215-219
  24. Lim MCC, Hoh ST, Foster PJ, LimTH, Chew SJ, Seah SKL, Aung T Use of optical coherence tomography to assess variations in macular retinal thickness in myopia. Invest Ophthalmol Vis Sci 2005; 46(3):974-8.
  25. Seol BR, Kim DM, Park KH, Jeoung JW. Assessment of optical coherence tomography color probability codes in myopic glaucoma eyes after applying myopic normative database. Am J Ophthalmol. 2017; 183:147–55
  26. Diagnostic Innovations in Glaucoma Study (DIGS). Grantome grantome.com › grant › NIH › R01-EY027510-03
  27. Kostanyan T, Wollstein G and Schuman JS. New developments in optical coherence tomography, Curr Opinion Ophthalmol 2015, 26(2) Invest Ophthalmol Vis Sci 2005;46 (3):974-8
  28. Nouri-Mahdavi K, Weiss RE. Detection of Glaucoma Deterioration in The Macular Region with Optical Coherence Tomography: Challenges and Solutions. Am J Ophthalmol. 2020 Sep 17:S0002-9394(20)30520-1

Search PubMed for

CITE THIS ARTICLE

Singh KChallenges In Optical Coherence Tomography Interpretation For Diagnosing Glaucoma In Myopic Eyes.DJO 2020;31:9-17

CITE THIS URL

Singh KChallenges In Optical Coherence Tomography Interpretation For Diagnosing Glaucoma In Myopic Eyes.DJO [serial online] 2020[cited 2021 Jan 26];31:9-17. Available from: http://www.djo.org.in/articles/31/2/Challenges-In-Optical-Coherence-Tomography-Interpretation-For-Diagnosing-Glaucoma-In-Myopic-Eyes.html