PERSPECTIVEFrom Clinical Examination of the Optic Disc to ClinicalAssessment of the Optic Nerve Head: A Paradigm Change

BALWANTRAY C. CHAUHAN AND CLAUDE F. BURGOYNE

� PURPOSE: To review and interpret the anatomy of theoptic nerve head (ONH) detected with spectral-domainoptical coherence tomography (SD OCT) pertaining tothe clinical examination of the optic disc and to proposethat a paradigm change for clinical assessment of theONH is necessary.� DESIGN: Perspective.� METHODS: Presently, the clinician evaluates neuroreti-nal rim health according to the appearance of the opticdisc, the clinically visible surface of the ONH. Recentanatomic findings with SDOCThave challenged the basisand accuracy of current rim evaluation. We demonstratewhy incorporation of SD OCT imaging of the ONH intothe clinical examination of the disc is required.� RESULTS: Disc margin-based rim evaluation lacksa solid anatomic basis and results in variably inaccuratemeasurements for 2 reasons. First, the clinically visibledisc margin is an unreliable outer border of rim tissuebecause of clinically and photographically invisible exten-sions of Bruch’s membrane. Second, rim tissue orienta-tion is not considered in width measurements. Wepropose alternative anatomically and geometrically accu-rate SD OCT-based approaches for rim assessment thathave enhanced detection of glaucoma. We also argue fornew data acquisition and analysis strategies with SDOCT that account for the large interindividual variabilityin the angle between the fovea and ONH.� CONCLUSIONS: We propose a 4-point paradigm changefor clinical assessment of the ONH that is anchored to theeye-specific anatomy and geometry of the ONH and fovea.Our approach is designed to enhance the accuracy andconsistency of rim width, as well as of peripapillary andmacular intraretinal thickness measurements. (Am JOphthalmol 2013;-:-–-. � 2013 by Elsevier Inc. Allrights reserved.)

Accepted for publication Apr 11, 2013.From the Department of Ophthalmology and Visual Sciences (B.C.C.),

Dalhousie University, Halifax, Nova Scotia, Canada; and the Devers EyeInstitute (C.F.B.), Portland, Oregon.

Inquiries to Balwantray C. Chauhan, Department of Ophthalmologyand Visual Sciences, Dalhousie University, 1276 South Park Street, 2WVictoria, Halifax, Nova Scotia, Canada B3H 2Y9; e-mail: bal@dal.ca

0002-9394/$36.00http://dx.doi.org/10.1016/j.ajo.2013.04.016

� 2013 BY ELSEVIER INC.

SINCE VON HELMHOLTZ’S DESCRIPTION OF THE

direct ophthalmoscope in 1851, clinical examina-tion of the optic disc has been a cornerstone of

ophthalmic practice.1 The optic disc constitutes theclinically visible surface of the neural and connectivetissues of the optic nerve head (ONH). By currentconvention, clinical disc examination requires identifi-cation of the outer and inner borders of the neuroreti-nal rim, respectively, the optic disc margin, and theoptic disc cup. The amount of rim tissue then is esti-mated within the apparent plane of the disc margin aseither the ratio of the size of the cup to the size ofthe disc2 or the rim area.3 These concepts are appliedwhether the examination is performed with directophthalmoscopy, slit-lamp biomicroscopy, optic discphotography, or a growing number of quantitativeimaging methods.Advances in spectral-domain optical coherence tomog-

raphy (SD OCT) for the first time have permitted imagingof ONH anatomic features. Structures such as the ante-rior4–6 and posterior7,8 lamina cribrosa surfaces, Bruch’smembrane–retinal pigment epithelium complex and itstermination within the ONH,6,9,10 border tissue ofElschnig,5 and the scleral canal opening9 now can be visu-alized readily. Accurately colocalizing fundus photographsto SD OCT image data has allowed clinicians to identifystructures that correspond to common clinical landmarks,for example, the optic disc margin.9,10

In this article, we explain how recent SD OCT find-ings undermine the current concepts of the clinical discmargin and rim quantification from both anatomic10

and geometric11–13 perspectives. In light of thesefindings, we propose a 4-part paradigm for incorporatingnew insights provided by SD OCT imaging of ONHanatomic features into the clinical examination of theoptic disc to achieve a clinical assessment of theONH. The paradigm recognizes the variable relation-ship between the fovea, the path of the retinal nervefiber bundles, and the ONH14 among individuals. Itslogic links the evaluation of the neuroretinal rim andperipapillary and macular retinal nerve fiber layer(RNFL) to the specific anatomy and geometry of eachindividual ONH and its relative orientation with thefovea.

1ALL RIGHTS RESERVED.

ANATOMIC ASSUMPTIONSUNDERLYING THE CLINICALLY VISIBLE

OPTIC DISC MARGIN

THE OPTIC DISC MARGIN IS A CLINICAL LANDMARK THAT

traditionally is defined to be the inner edge of the sclerallip or crescent (Figure 1).15 Within this conceptual frame-work, the disc margin is assumed to be a single and consis-tent anatomic structure around the entire ONH and a trueouter border of the neuroretinal rim, and therefore thelandmark from which the width of the rim can bemeasured. Current examination methods require identifi-cation of the disc margin, whether the examination isperformed clinically, photographically, or with confocalscanning laser tomography, an imaging technique thatmaps the surface topography of the ONH.16

FIGURE 1. Diagrams showing current concepts of optic discmargin anatomy. (Top) Optic disc appearance of a right eye.(Bottom) Corresponding view of the optic nerve head. In thisexample, the temporal disc margin is defined by the inner edgeof the scleral lip or crescent (dotted lines), whereas the nasaldisc margin corresponds to the termination of retinal pigmentepithelium (RPE), which also coincides with the inner edge ofthe sclera. In both instances, the disc margin is depicted as thetrue outer border of the neuroretinal rim. RNFL[ retinal nervefiber layer. Modified from Hogan, Alvarado Wedell.15

ANATOMIC ERRORS IN THECURRENT EVALUATION OF THE

NEURORETINAL RIM

RECENT HISTOLOGIC FINDINGS IN MONKEY EYES9 AND SD

OCT findings in human eyes10 (in each case, colocalizedto optic disc stereophotographs with the disc margin tracedby a glaucomatologist) have revealed 2 new findings thatchallenge the anatomic assumptions that underlie opticdisc margin-based neuroretinal rim evaluation. First, theclinical disc margin rarely is a single anatomic entity, norare the structures that underlie it consistent in an indi-vidual eye. Hence, the structure corresponding to the discmargin at the 3-o’clock position may be different to thatat the 9-o’clock position. Similarly, the structure corre-sponding to it at the 3-o’clock position in 2 eyes also maybe different.10 Second, what the clinician perceives asthe disc margin in clinical examination or with photo-graphs frequently is not a true anatomic outer border ofthe neuroretinal rim because of regionally variable andinvisible extensions of Bruch’s membrane that have notbeen appreciated previously (Figure 2). In such cases, therecan be a serious overestimation of the remaining rimtissue.13

Separate from these new issues regarding the anatomicouter border of the neuroretinal rim, its inner border, whichconstitutes the cup margin, is evaluated clinically witha stereoscopic evaluation of the rim width within theperceived plane of the disc margin. However, the rationaleto support the existence of an anatomically defensiblejunction between the neuroretinal rim and cup is weak.Current imaging techniques similarly divide the rimand cup on the basis of an arbitrary depth below a fixedplane.

The consequence of these findings is that current param-eters, including cup-to-disc ratio and neuroretinal rim area,

2 AMERICAN JOURNAL OF

are unlikely to estimate accurately the amount of remain-ing neural tissue in the ONH.13

ANATOMIC RATIONALE FOR BRUCH’SMEMBRANE OPENING AS THE OUTERBORDER OF THE NEURORETINAL RIM

THE TERMINATION OF BRUCH’S MEMBRANE AT THE ONH

represents the opening through which retinal ganglioncell axons exit the eye to form the choroidal and scleralportions of the neural canal. As such, this anatomicopening, termed Bruch’s membrane opening (BMO), isa true outer border of the neural tissues because axonscannot pass through an intact Bruch’s membrane to exitthe eye. Whether BMO is clinically visible, it is an

--- 2013OPHTHALMOLOGY

FIGURE 2. Optic disc appearance with conventional optic disc photography and spectral-domain optical coherence tomography (SDOCT) showing the pertinent anatomic features in the left eye of a patient with glaucoma. (Top) Optic disc with infrared illumination(first and third columns) with SD OCT imaging and photography (second and fourth columns) precisely colocalized. (Bottom) Colo-calized B-scan images corresponding to the position (blue line) indicated in the fundus images. (Top right) Clinically visible optic discmargin (green dots) traced in the disc photograph projected to (Bottom right) the B-scan, and (Bottom right) Bruch’s membraneopening identified in the B-scan (red dots) projected to (Top right) the fundus photograph. In this case, there is a clinically invisibleoverhang of Bruch’s membrane in the superior temporal quadrant resulting in a significant overestimation of the neuroretinal rimwidth with disc margin based evaluation, whereas in the inferior nasal quadrant, the rim is actually wider than perceived clinically.

anatomically accurate landmark from which neuroretinalrim measurements can be made. It is also a structure thatis identified consistently with SD OCT.9,13,17

The stability of BMO under a variety of conditionsprovides another rationale for its usefulness as a landmark.BMO is unaltered in the face of large changes in intraocularpressure induced by glaucoma surgery.18 The 2-dimensionalplane that best fits BMO also is axially stable with surgicalreduction of intraocular pressure.18 There is no publishedevidence yet on the stability of BMO with glaucomaprogression; however, in experimental glaucoma inmonkeys, BMO position in 3-dimensional histomorphom-etry of the ONH seems to be unaltered despite changes inthe neural component of the ONH and in the positionsof the anterior and posterior scleral canal opening.19

GEOMETRIC ERRORS IN THECURRENT EVALUATION OF THE

NEURORETINAL RIM

NEURORETINAL RIM MEASUREMENT WITH CLINICAL,

photographic, or confocal scanning laser tomographictechniques is made along the 2-dimensional plane of theperceived optic disc margin. However, in a single eye, theorientation of rim tissue varies around the ONH. At one

VOL. -, NO. - CLINICAL ASSESSMENT OF TH

extreme, axons may exit the eye almost parallel to thevisual axis, whereas at the other extreme, they may exitthe eye almost perpendicular to it, typically in the temporalsector, which can have a shallow sloping rim.20 Hence,potentially significant errors in rim width can occur if themeasurement plane is fixed.13 For example, for the samenumber of axons, the conventional rim width will be largerwhen its orientation is more horizontal compared withwhen it is more perpendicular to the fixed plane ofmeasurement (Figure 3). Even if a rim measurement ismade from an anatomically accurate location, that is,BMO, its width along the approximately horizontal fixedplane of BMO will be influenced by the orientation ofthe overlying rim tissue (Figure 3).

GEOMETRIC RATIONALE FORFORMULATING A MINIMUM RIM

WIDTH MEASUREMENT

BECAUSE OF THE VARYING ORIENTATION OF THE NEURO-

retinal rim relative to BMO, Chen and Povazay and associ-ates and first proposed that the minimum distance fromBMO to the internal limiting membrane represents themost geometrically accurate measurement of neuroretinalrim width.11,12 We subsequently characterized the

3E OPTIC NERVE HEAD

FIGURE 3. Influence of neuroretinal rim orientation on rim width measurement. Portions of B-scans from the same eye withspectral-domain optical coherence tomography: (Top left) one in which the rim orientation is relatively flat and (Top right) in anotherwhere it is relatively steep. (Bottom) Measurement of rim width along the fixed Bruch’s membrane opening (BMO) plane, termedBMO-horizontal rim width (BMO-HRW), and along the axis resulting in the minimum rim width from BMO, termed BMO-minimum rim width (BMO-MRW). The actual amount of rim tissue represented by BMO-MRW is identical in these 2 sections;however, BMO-HRW is more than 60% wider in the section shown to the left compared with the one on the right.

difference between this rim measurement, termed BMO-minimum rim width (BMO-MRW), and conventionalones13 and its usefulness in the detection of progressiveONH change in experimental primate glaucoma.21 Mostrecently, we showed that BMO-MRW significantlyenhanced the ability to detect glaucomatous optic neurop-athy compared with current confocal scanning laser tomog-raphy and current SD OCT analyses.22

Automated algorithms for identifying (segmenting) BMOinB-scans of SDOCT images have been described and incor-porated into commercial software.17,23–25 However, to date,the primary rationale for this implementation has been forautomated detection of the optic disc margin and derivationof a rim width measurement along a single fixed plane,23,25

akin to current clinical methods. Hence, although SD OCTtechnology is providing unprecedented detail on theanatomy of the ONH, which heretofore has not beenpossible, its use has been limited to mimicking the mannerin which the optic disc is examined clinically, rather thanincorporating the anatomic insights that clinicians have notbeen able to appreciate before into clinical practice.

CLINICAL EXAMPLES

THEOPTICDISCPHOTOGRAPHAND24COLOCALIZED SDOCT

radial B-scans around the ONH of each of 30 glaucoma

4 AMERICAN JOURNAL OF

patients and10 control subjects illustrating (1) that the clin-ical disc margin is not a single anatomic location, (2) thatclinically invisible extensions of Bruch’smembrane internalto the disc margin are regionally present in most eyes,and (3) that the dependence of rim width measurementon rim tissue orientation are available online (http://ophthalmology.medicine.dal.ca/research/onh.html).

ANATOMIC VARIATION IN FOVEAPOSITION RELATIVE TO THE OPTIC

NERVE HEAD

IN CLINICAL FUNDUS IMAGES, THE FOVEA IS LOCATED

below the level of the center of the ONH in most individ-uals. A recent study on the angle between the fovea andBMO center relative to the horizontal axis defined by thefundus image, termed the fovea-BMO center axis, in 222patients with ocular hypertension or glaucoma showedthat although the mean angle of this axis was �7 degrees(the fovea being 7 degrees below), the range wasfrom �17 degrees to þ6 degrees, or 23 degrees (DemirelS, written communication, January 25, 2013; Figure 4).Although the positions of these 2 structures relative tothe fundus image vary considerably between subjects, theanatomic path of RNFL bundles is governed primarily bythese 2 structures as the bundles approach the ONH and

--- 2013OPHTHALMOLOGY

FIGURE 4. Interindividual variability in the axis connecting the fovea and Bruch’s membrane opening (BMO) center (fovea-BMOcenter axis) in 222 patients with ocular hypertension or glaucoma. (Left) Schematic demonstrating the range of the fovea-BMO centeraxis (solid line), from (Top left) D6 degrees above to (Bottom left) L17 degrees below the horizontal in the image frame (dashedline). (Right) Frequency histogramwith a mean value ofL7 degrees (center left). Dots[BMO; filled circle[ fovea; filled square[BMO center. Data courtesy of Dr Shaban Demirel, Devers Eye Institute, Portland, Oregon.

exit the eye.14 In fundus images, the positions of the foveaand ONH also may vary slightly within the same individualfrom day to day because of cyclotorsion;26 however, thepath of RNFL bundles remains constant relative to thefovea-BMO center axis.

RATIONALE FOR REGIONALIZATION OFTHE NEURORETINAL RIM AND

PERIPAPILLARY AND MACULAR NERVEFIBER LAYER RELATIVE TO THE FOVEA-

BRUCH’S MEMBRANE OPENINGCENTER AXIS

CURRENTLY, IMAGE ACQUISITION AND DATA ANALYSIS

algorithms report regional data according to the temporal,superior, nasal, and inferior sector positions that are estab-lished relative to the fixed horizontal and vertical axes ofthe image. Hence, for example, the neuroretinal rim widthor peripapillary RNFL thickness in a given sector is assumedto refer to precisely the same anatomic location amongdifferent persons. However, because the fovea-BMO centeraxis can vary by as much as 23 degrees, the anatomic

VOL. -, NO. - CLINICAL ASSESSMENT OF TH

difference between 2 eyes of an individual geometric sectorposition also may be as large as 23 degrees (Figure 5).Because sector positions refer to measurements from

different anatomic locations, artificially large interindividualdifferences in sectoral neuroretinal rim width and peripapil-lary and macular RNFL thickness likely occur. As a result,the limits of normal variation in these measurements, as re-flected in normative databases, are probably artificiallyincreased, decreasing the diagnostic precision of imagingdevices. Errors in mapping ocular structures to the visualfield also may be induced and contribute to the poorobserved correlation between measures of structure andfunction in glaucoma.27,28

RATIONALE FOR IMAGE ACQUISITIONRELATIVE TO THE FOVEA-BRUCH’SMEMBRANE OPENING CENTER AXIS

BECAUSE IT IS IMPORTANTTOREGIONALIZENEURORETINAL

rim width and peripapillary and macular RNFL thicknessaccording to the fovea-BMO center axis, it is reasonableto extend the same logic also to data acquisition. Currently,data acquisition by imaging devices occurs invariably

5E OPTIC NERVE HEAD

FIGURE 5. Impact of regionalization of neuroretinal rim width and peripapillary retinal nerve fiber layer thickness measurementsbased on the axis connecting the fovea and Bruch’s membrane opening (BMO) center (fovea-BMO center axis). (Top center) Eyewith fovea-BMO center axis ofD6 degrees. (Top right) Eye with fovea-BMO center axis ofL17 degrees. (Middle left) Sector region-alization according to positions that are fixed relative to the imaging frame of the imaging device, currently applied to all eyes in theanalyses of rim width and retinal nerve fiber layer thickness. (Middle center and Middle right) Current regionalization with fixedsector orientation leads to measurements from variably different anatomic locations. (Bottom center and Bottom right) Regionaliza-tion relative to fovea-BMO center axis. In this case, rotating the sector orientation according to the fovea-BMO center axis ensuresthat sectors contain measurements from the same anatomic locations. IN [ inferonasal; IT [ inferotemporal; N [ nasal; SN [superonasal; ST [ superotemporal; T [ temporal.

according to a fixed coordinate system for sectors preset inthe device.

It is logical, therefore, that the fovea-BMO center axis bedetermined first in each eye, followed by data acquisitionrelative to this axis, to ensure that measurements aredistributed according to the pertinent anatomy of the indi-vidual eye. Such an approach will allow the most robustsampling of the constituent anatomy and the most mean-ingful comparison between individuals. Furthermore,because of image registration software, follow-up imagesthen can be acquired at the same anatomic positions ineach eye.

6 AMERICAN JOURNAL OF

PROPOSED PARADIGM CHANGE FORCLINICAL ASSESSMENT OF THE OPTIC

NERVE HEAD

BELOW, WE PROPOSE A 4-POINT PARADIGM CHANGE THAT

incorporates the new anatomic insights provided by SDOCT imaging of the ONH into the clinical examination ofthe optic disc for clinical assessment of the ONH (Figure 6).1. Optical Coherence Tomography Imaging of the Optic

Nerve Head Should not Mimic the Clinical Examination ofthe Optic Disc.

--- 2013OPHTHALMOLOGY

FIGURE 6. Schematic of incorporating spectral-domain opticalcoherence tomography (SDOCT) imaging that includes quanti-tative analyses and an understanding of optic nerve head (ONH)anatomic features into a clinical examination of the optic discwith ophthalmoscopy, slit-lamp biomicroscopy, or photographyto yield a clinically informed assessment of the ONH.

Because of the fundamental differences between theclinical examination of the optic disc and SD OCTimaging of the ONH, we believe that constraining SDOCT analyses to mimic the disc examination is erroneous.Instead, we propose that clinical practice should incorpo-rate knowledge of ONH anatomy detected by SD OCTinto the clinical disc examination in the same mannerthat diagnostic imaging techniques in other areas of medi-cine have advanced their fields. For example, pulmonolo-gists are trained to review imaging studies (x-ray,computed tomography, and magnetic resonance images)to enhance their clinical chest examination (palpation,percussion, and auscultation). In this regard, the practiceof referring to the SD OCT-detected BMO as the opticdisc margin is inaccurate and potentially misleading. Incor-porating the anatomic insights provided by SD OCT intothe clinical examination will be informative and shouldenable clinicians to make more accurate diagnoses.Constraining SD OCT to conventional concepts of discanatomy and terminology will inhibit its clinical potential.

2. Neuroretinal Rim Measurement Should be Anatomicallyand Geometrically Accurate.

Quantifying neuroretinal rim width requires identifica-tion of its inner border relative to an anatomic landmarkthat is a true outer border of the rim. Currently, BMO isthe most consistent SD OCT-detected outer border ofthe neuroretinal rim and a stable landmark that is visiblereadily in all but a few exceptional B-scans.9,22 Theminimum distance from BMO to the internal limitingmembrane represents the geometrically correct width ofthe neuroretinal rim. Automated algorithms that detectBMO and the internal limiting membrane with SD OCTimages are now available.17,23–25 Those that quantifyBMO-MRW regionally and determine its probability ofbeing within age-specific norms could be developed andbe available for clinical use.

VOL. -, NO. - CLINICAL ASSESSMENT OF TH

3. Acquisition and Regionalization of Data Based on Fovea-Bruch’s Membrane Opening Center Axis.Acquisition and regionalization of ONH and retinal

imaging data should respect the anatomic relationshipbetween the fovea and the ONH. In doing so, regional esti-mates of the neuroretinal rim and the peripapillary andmacular RNFL layer thickness will be correct anatomicallyand consistently will be comparable among individuals.The practical effect of correct regionalization should bereduced interindividual variation, leading to potentiallymore accurate and sensitive identification of early damagebecause of statistically narrower limits of variability.Correct regionalization also may increase the correspon-dence between structural and visual field measures. Suchregionalization strategies will not impact global measure-ments, because the latter represent sums or averages aroundthe entire ONH or macula.4. Optical Coherence Tomography Imaging of ONH

Anatomic Features Should Enhance, Rather Than Replace,the Clinical Optic Disc Examination.One of the key goals of the clinical optic disc examina-

tion is to assess the amount and health of the neuroretinalrim. In the clinical examination of the optic disc, theamount of rim tissue should be estimated based on theassumption that the disc margin represents the outer borderof rim tissue. The examination should continue to charac-terize the color and health of the rim regionally, or by clockhour, around the optic disc and to check for the presence ofpallor, bowing, or excavation of the ONH surface; opticdisc hemorrhages; and peripapillary RNFL defects. Theclinical examination is essential even if SD OCT imagingis available because many of the qualitative aspects of theexamination, such as the color of the rim and the presenceof hemorrhages, cannot be detected readily with SD OCTor other imaging devices.However, the clinician should be cognizant of the limi-

tations of current methods of rim width measurement andbe aware that potentially serious errors in rim evaluationcan occur because of the reasons discussed above. For prag-matic reasons, it is unrealistic to expect or require that eachclinical optic disc examination by all physicians in all envi-ronments be accompanied by SD OCT imaging. When theclinical disc examination is unequivocally normal, SDOCT imaging is unlikely to be additionally informative(Figure 7). However, when there is suspicion of abnor-mality, SD OCT will provide an anatomically accurateassessment of the neuroretinal rim (Figure 8).Finally, we believe that SD OCT-detected anatomic

features of the ONH should be displayed relative to a clin-ical photograph or infrared image of the disc by sectors orclock-hour positions according to the fovea-BMO centeraxis (Figures 7 and 8). Hence, clinicians can incorporateknowledge of this anatomy into their clinical discexamination and formulate a clinical assessment of theONH at the time imaging is performed. In the future,segmental rim widths in a given image may be compared

7E OPTIC NERVE HEAD

FIGURE 7. Incorporation of spectral-domain optical coherence tomography (SDOCT) into a clinical examination of the optic disc ina normal left eye. In this case, where the clinical appearance (Top) seems unequivocally normal, SD OCT with analysis of Bruch’smembrane opening minimum rim width (BMO-MRW) by clock hour (Bottom) is unlikely to be additionally informative. There isa minor mismatch between the clinically visible disc margin (green dots) and BMO (red dots), indicating that disc margin orBMO-based rim values are similar. At most clock hour positions, BMO-MRW appears wide. Future algorithms that indicate whetherindividual BMO-MRW values lie within or outside age-specific normal values could be clinically valuable.

with normative databases in an anatomically andgeometrically accurate manner.

SUMMARY

FROM RECENT FINDINGS ON ONH ANATOMY DETECTED

with SD OCT, we have argued that the foundation ofcurrent clinical optic disc margin-based evaluation of theneuroretinal rim is inaccurate because it lacks a solidanatomic and geometric foundation. We propose a para-digm change from the current clinical examination of the

8 AMERICAN JOURNAL OF

optic disc to the clinical assessment of the ONH thatincludes SD OCT imaging (Figure 6). The immediateconsequence of the new quantitative measures proposedis that our ability to detect glaucoma is enhanced.22 Weanticipate that as data acquisition and analysis takes placeaccording to the specific fovea-BMO center axis of an indi-vidual eye, interindividual variation in several importantquantitative measures, including thickness of the peripapil-lary and macular RNFL and other segmented layers of theretina, will decrease and further will enhance detection ofglaucoma. Finally, as longitudinal data are acquired, ourability to detect earlier and clinically meaningful changesimilarly may also be improved.

--- 2013OPHTHALMOLOGY

FIGURE 8. Incorporation of spectral-domain optical coherence tomography (SD OCT) into a clinical examination of the optic disc(Top) in a right eye suspected of glaucoma. In this case, SD OCT with analysis of Bruch’s membrane opening minimum rim width(BMO-MRW) by clock hour (Bottom) provides valuable additional information. There is a considerable mismatch between the clin-ically visible disc margin (green dots) and BMO (red dots), indicating a clinically invisible extension of Bruch’s membrane, especiallyinferiorly and nasally. In these locations as well as the superior sector, which is the most suspicious, the neuroretinal rim is consid-erably thinner than the clinician would estimate from a disc margin-based evaluation. Future algorithms that indicate whether indi-vidual BMO-MRW values lie within or outside age-specific normal values could be clinically valuable.

ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTERESTand the following were reported. Dr Chauhan is a consultant for and receives lectures fees fromAllergan, receives equipment support and lecture fees fromCarl Zeiss Meditec, and receives unrestricted research and equipment support from Heidelberg Engineering. Dr Burgoyne is a consultant to and receiveslectures fees fromMerck, Sharpe, and Dohme; receives unrestricted research and equipment support fromHeidelberg Engineering; and is a nonpaid consul-tant to and receives equipment support from Reichert. Dr Chauhan is supported by Grant MOP11357 from the Canadian Institutes of Health Research,Ottawa, Ontario, Canada. Dr Burgoyne is supported by Grants RO1EY011610 and RO1EY021281 from the National Eye Institute, National Institutes ofHealth, Bethesda, Maryland. Involved in Design of study (B.C.C., C.F.B.); Conduct of study (B.C.C., C.F.B.); Analysis and interpretation of data (B.C.C.,C.F.B.); and Preparation, review, and approval of manuscript (B.C.C., C.F.B.). The authors thank Drs Marcelo Nicolela, Neil O’Leary, Alexandre Reis,Faisal AlMobarak, Brad Fortune, Nicholas Strouthidis, and Hongli Yang for their contribution and discussions leading to the various research projects andideas represented in this article; Drs. Shaban Demirel, Stuart Gardiner, and Brad Fortune of the Portland Progression Project for providing the data inFigure 4; Glen Sharpe and Donna Hutchison for data collection and processing; and Juan Reynaud for software support.

VOL. -, NO. - 9CLINICAL ASSESSMENT OF THE OPTIC NERVE HEAD

REFERENCES

1. von Helmholtz H. Beschreibung eines Augenspiegels zurUntersuchung der Netzhaut am lebenden Auge. Berlin: AForstner’sche Verlagsbuchhandlung, 1851.

2. Armaly MF. Lessons to be learned from a glaucoma survey.J Iowa State Med Soc 1960;50:501–507.

3. Balazsi AG, Drance SM, Schulzer M, Douglas GR. Neurore-tinal rim area in suspected glaucoma and early chronicopen-angle glaucoma. Correlation with parameters of visualfunction. Arch Ophthalmol 1984;102(7):1011–1014.

4. Agoumi Y, Sharpe GP, Hutchison DM, Nicolela MT,Artes PH, Chauhan BC. Laminar and prelaminar tissuedisplacement during intraocular pressure elevation in glau-coma patients and healthy controls. Ophthalmology 2011;118(1):52–59.

5. Strouthidis NG, Grimm J, Williams GA, Cull GA,Wilson DJ, Burgoyne CF. A comparison of optic nerve headmorphology viewed by spectral domain optical coherencetomography and by serial histology. Invest Ophthalmol VisSci 2010;51(3):1464–1474.

6. Srinivasan VJ, Adler DC, Chen Y, et al. Ultrahigh-speedoptical coherence tomography for three-dimensional and enface imaging of the retina and optic nerve head. InvestOphthalmol Vis Sci 2008;49(11):5103–5110.

7. Lee EJ, Kim TW, Weinreb RN, Park KH, Kim SH, Kim DM.Visualization of the lamina cribrosa using enhanced depthimaging spectral-domain optical coherence tomography.Am J Ophthalmol 2011;152(1):87–95e81.

8. Park HY, Jeon SH, Park CK. Enhanced depth imaging detectslamina cribrosa thickness differences in normal tension glau-coma and primary open-angle glaucoma.Ophthalmology 2012;119(1):10–20.

9. Strouthidis NG, Yang H, Reynaud JF, et al. Comparison ofclinical and spectral domain optical coherence tomographyoptic disc margin anatomy. Invest Ophthalmol Vis Sci 2009;50(10):4709–4718.

10. Reis AS, Sharpe GP, Yang H, Nicolela MT, Burgoyne CF,Chauhan BC. Optic disc margin anatomy in patients withglaucoma and normal controls with spectral domain opticalcoherence tomography.Ophthalmology 2012;119(4):738–747.

11. Chen TC. Spectral domain optical coherence tomography inglaucoma: qualitative and quantitative analysis of the opticnerve head and retinal nerve fiber layer (an AOS thesis).Trans Am Ophthalmol Soc 2009;107:254–281.

12. Povazay B, Hofer B, Hermann B, et al. Minimum distancemapping using three-dimensional optical coherence tomog-raphy for glaucoma diagnosis. J Biomed Opt 2007;12(4):041204.

13. Reis AS, O’Leary N, Yang H, et al. Influence of clinicallyinvisible, but optical coherence tomography detected, opticdisc margin anatomy on neuroretinal rim evaluation. InvestOphthalmol Vis Sci 2012;53(4):1852–1860.

14. Airaksinen PJ, Tuulonen A, Werner EB. Clinical evaluationof the optic disc and retinal nerve fiber layer. In: Ritch R,Shields MB, Krupin T, eds. The Glaucomas. St. Louis:Mosby-Year Book, 1996:617–657.

10 AMERICAN JOURNAL OF

15. Hogan MJ, Alvarado JA, Weddell JE. Histology of theHuman Eye. Philadelphia: Saunders; 1971:523–606.

16. Zinser G, Wijnaendts-van-Resandt RW, Dreher AW,Weinreb RW, Harbarth U, Burk ROW. Confocal laser tomo-graphic scanning of the eye. Proc SPIE 1989;1161:337–344.

17. Hu Z, Abramoff MD, Kwon YH, Lee K, Garvin MK. Auto-mated segmentation of neural canal opening and optic cupin 3D spectral optical coherence tomography volumes ofthe optic nerve head. Invest Ophthalmol Vis Sci 2010;51(11):5708–5717.

18. Reis AS, O’Leary N, Stanfield MJ, Shuba LM, Nicolela MT,Chauhan BC. Laminar displacement and prelaminar tissuethickness change after glaucoma surgery imaged with opticalcoherence tomography. Invest Ophthalmol Vis Sci 2012;53(9):5819–5826.

19. Bellezza AJ, Rintalan CJ, Thompson HW, Downs JC,Hart RT, Burgoyne CF. Deformation of the lamina cribrosaand anterior scleral canal wall in early experimental glau-coma. Invest Ophthalmol Vis Sci 2003;44(2):623–637.

20. Jonas JB,GusekGC,Guggenmoos-Holzmann I,NaumannGO.Correlations of the neuroretinal rim area with ocular andgeneral parameters in normal eyes. Ophthalmic Res 1988;20(5):298–303.

21. Strouthidis NG, Fortune B, Yang H, Sigal IA, Burgoyne CF.Longitudinal change detected by spectral domain opticalcoherence tomography in the optic nerve head and peripapil-lary retina in experimental glaucoma. Invest Ophthalmol VisSci 2011;52(3):1206–1219.

22. Chauhan BC, O’Leary N, Almobarak FA, et al. Enhanceddetection of open-angle glaucoma with an anatomicallyaccurate optical coherence tomography-derived neuroretinalrim parameter. Ophthalmology 2013;120(3):535–543.

23. Abramoff MD, Lee K, Niemeijer M, et al. Automatedsegmentation of the cup and rim from spectral domainOCT of the optic nerve head. Invest Ophthalmol Vis Sci2009;50(12):5778–5784.

24. Lee K, Niemeijer M, Garvin MK, Kwon YH, Sonka M,Abramoff MD. Segmentation of the optic disc in 3-D OCTscans of the optic nerve head. IEEE Trans Med Imaging

2010;29(1):159–168.25. Mwanza JC, Chang RT, Budenz DL, et al. Reproducibility of

peripapillary retinal nerve fiber layer thickness and opticnerve head parameters measured with cirrus HD-OCT inglaucomatous eyes. Invest Ophthalmol Vis Sci 2010;51(11):5724–5730.

26. Patel NB, Wheat JL, Rodriguez A, Tran V, Harwerth RS.Agreement between retinal nerve fiber layer measures fromSpectralis and Cirrus spectral domain OCT. Optom Vis Sci2012;89(5):E652–E666.

27. Caprioli J, Miller JM. Correlation of structure and function inglaucoma. Quantitative measurements of disc and field.Ophthalmology 1988;95(6):723–727.

28. Hood DC, Anderson SC, Wall M, Raza AS, Kardon RH. Atest of a linear model of glaucomatous structure-functionloss reveals sources of variability in retinal nerve fiber andvisual field measurements. Invest Ophthalmol Vis Sci 2009;50(9):4254–4266.

--- 2013OPHTHALMOLOGY

Biosketch

Balwantray C. Chauhan, PhD, is theMathers Professor of Ophthalmology and Visual Sciences at Dalhousie University. His

research centres on alterations in the visual field and optic nerve head in clinical glaucoma, and on the effects of intraocular

pressure and endothelin on the neuronal and glial components of the optic nerve in experimental glaucoma. He is Principal

Investigator of the Canadian Glaucoma Study, a multi-centre investigation on the risk factors for the progression of open-

angle glaucoma.

VOL. -, NO. - 10.e1CLINICAL ASSESSMENT OF THE OPTIC NERVE HEAD

Biosketch

Claude F. Burgoyne, MD, is the Van Buskirk Chair for Ophthalmic Research and Director of the Optic Nerve Head

Research Laboratory at Devers Eye Institute in Portland, Oregon. His laboratory studies the effects of aging and

experimental glaucoma on the neural and connective tissues of the monkey optic nerve head within post-mortem 3D

histomorphometric reconstructions. It is now translating its 3D visualization and quantification capabilities from the

monkey to the human optic nerve head with spectral domain optical coherence tomography.

10.e2 --- 2013AMERICAN JOURNAL OF OPHTHALMOLOGY