15

www.jphmedical.com

Michael W. Belin, MD

INTRODUCTION

Most clinicians are familiar with the term“tomography”asinComputerizedTomography(CT)scanning. While the distinction between a standardX-ray and a CT scan is clearly understood, there isa generalized misunderstanding of the differencesbetween standard corneal “topography” and corneal“tomography.” Tomographyisanimagingtechniquewhereaseriesof cross-sectional images are merged to allow for acomputergeneratedthree-dimensionalreconstruction.The advantages of anterior segment tomographycompared to corneal topography are similar to theadvantages of CT scanning compared to standardroentgenograph.Anteriorsegmenttomographyallowsnot only the visualization of the anterior cornealsurface,buttheposteriorcornealsurface,theanteriorchamber, thecorneal thickness,portionsof theangleof the eye, and the lens. The computer generatedthree-dimensional reconstruction allows for a muchexpanded rangeof applications. Measurements suchasanteriorchamberdepth,cornealthickness,andlensdensity are samples of measurements available withanterior segment tomography thatwerenot availablewith older devices that just measured the anteriorcornealsurface. In order to generate a three-dimensionalreconstructionaseriesofplanaropticalcross-sectionalimages are merged (similar to creating a three-dimensional retinal photograph by merging a stereopairofphotographs). Whiletheearliersystemsuseda seriesof vertical optical cross sections,more recentsystems have utilized a rotating Scheimpflug camerawhich increased the accuracy of the tomographydevices.1

The added information supplied by anteriorsegment tomography has expanded the medical andsurgicalapplicationsofthesedevices.Whilescreening

patientsforrefractivesurgery,operativeplanningandthediagnosisofpost-LASIKectasia remain themostfrequentuseofthesedevices;otherapplicationsincludeuses such as: surgical planning for toric intraocularlenses,surgicalplanningforintra-cornealsegments(e.g.INTACS)2, surgical planning for phakic intraocularlenses, postoperative management of DS(A)EKpatients3,andpre-operativeevaluationofthecataractpatient. While there ismore thanonecommerciallyavailable rotatingScheimpflugsystem, theOCULUSPentacam (OCULUS Optikgeraete GmbH, Wetzlar,Germany)representsthevastmajorityofthesystemscurrently in use, though the discussion that followwouldapplytoallScheimpflugdevices.4

REFRACTIVE SURGERY

Themostfrequentapplicationofanteriorsegmenttomographyisintheevaluationofthepotentialrefrac-tive surgery candidate and identifying patients withearly or sub-clinical keratoconus. It is important toidentify these patients, who may be asymptomatic,because refractive surgery in these patients may leadto a rapid deterioration in vision, corneal distortionandprogressivethinning.Inordertoidentifypatientswhomayhaveearlyor sub-clinicalchanges,weneedtolookatmorethanjustvisualacuity,centralcornealthickness andanterior corneal curvature.5 Thevalueofnewtechnology is that itprovides informationonbothanteriorandposteriorcornealsurfacesandallowsfor the generation of a full pachymetric distributionmap.Theexamplebelow(Figure 1)demonstratesthebenefitofthistechnologyandthelimitationofrelyingsolelyonoldertechnologythatwaslimitedtomeasur-ing just anterior corneal curvature. In this example,while the anterior curvature appears completely nor-mal, the tomography reveals a significant“ectasia”orislandofpositiveelevationontheposteriorsurfaceandamarked inferiordisplacementof the thinnestpoint

Applications of Anterior Segment Tomography in Corneal Surgery

Michael W. Belin, MD • Professor of Ophthalmology, University of Arizona Health Sciences • Tucson, AZ (USA)

PUBLISHED BY JAYPEE-HIGHLIGHTS IN THE HIGHLIGHTS OF OPHTHALMOLOGY JOURNAL, VOLUME 38, NO. 2, 2010

16

Highlights of Ophthalmology • Vol. 38, Nº 2

on the pachymetry map. Changes on the posteriorsurface and/or changes in pachymetric distributionor progression will typically predate any changes ofthe anterior corneal surface and theuseScheimpflugimagingreconstructionwill identify“atrisk”patientswhowouldotherwisebemissedbyconventionaltopo-graphicscreening. Inadditiontoscreeningforectaticdisease,refractiveplanninginvolvescomputationofthepredictedminimalresidualbed.Thetraditionalmethodofestimatingtheresidualbedinvolvestakingthepre-operativecornealthickness by ultrasonic pachymetry and subtractingboththeflapthicknessandtheablationdepth.Ithasalwaysbeenassumedthatthelargestpotentialsourceoferrorwas intheestimationofflapthickness. Theestimatedablationdepthisprobablythemeasurewiththeleastamountofvariability. For corneal thickness, most practitioners utilizea hand-held ultrasonic pachymeter to determinethickness. Ultrasound, however, has a number ofsignificant limitations,oneofwhich is that it fails to

Figure 1. (4 map refractive display (Oculus Pentacam)). The map on the upper right is a standard curvature map and looks totally normal. The posterior elevation (bottom right) and the pachymetric map (upper left) show prominent abnormalities that identify this cornea as ectatic. This would have been missed if one only looked at anterior curvature and central pachymetry readings.

givea completepictureof theentire corneal surface.Theassumptionisthatthecenterofthecorneaisthethinnestzone,andwhilethisassumptionisreasonablytruein“normal”corneas,itisoftenwrongineyeswithkeratoconusor“sub-clinical”disease(exactlythecaseswearetryingtoeliminate)(Figure 1). Residualbedcomputation shouldutilize the thinnestpoint in thetreated portion of the cornea and not just a centralapical reading. For these reasons and others, I haveabandoned the routine use of ultrasonic pachymetryinfavoranautomatedopticaldevicethatallowsmetomeasurethecorneal thicknessover theentirecornealsurface. Insteadofusingasingleultrasoundreadingdeterminedatthe“apex”,Iutilizethethinnestcornealthicknessreadingonthepachymetricmap. We recently looked at 1,436 pre-operative eyesusing the Oculus Pentacam and recorded cornealthicknessattheapex,pupillarycenterandthethinnestreading(Table 1).

The average thickness readings at the apex(539.3µm),pupilcenter (538.8µm)andthe thinnestreading (536.3 µm) were similar. The differencesbetween the apex and both the pupillary center andthe thinnest region were also small with a relativelytightstandarddeviation(1.06+/-1/73,2.99+/-4.34respectively). The range, however, did show a fewsignificantoutliers.Atleastonepatienthada31µmdifferencebetweentheirapexandpupilcenterreadingandupto93µmcomparingthethinnestregiontotheapex. The use of the full pachymetric map and thethinnestpointreadingallowsfortheidentificationoftheseoutliers,thatotherwisewouldhavebeenmissed.6

In addition to screening preoperatively for earlyor sub-clinical keratoconus, it is essential to be ableto separate decreased vision after LASIK due toundercorrection,abnormalhealingresponseorsurfaceabnormalities from iatrogenic ectasia. While earlier

Apex Pupil Thinnest Apex-Pupil Apex-Thin Pupil-Thin

Mean 539.3 538.8 536.3 1.06 2.99 1.94

Median 539.0 539.0 537.9 1.0 2.0 1.0

Mode 542.0 542.0 539.0 0 1.0 1.0

S.D. 36.8 36.9 37.12 1.73 4.34 3.07

Range 411-664 410-664 409-664 0-31 0-93 0-61

TABLE 1 N= 1,436 eyes

17

www.jphmedical.com

Figure 2. (4 map refractive display (Oculus Pentacam)). This postoperative LASIK patient exhibits a normal anterior corneal surface (both curvature and elevation), but shows a prominent positive para-central island indicative of iatrogenic ectasia.

Figure 3. (Corneal thickness or Pachymetric map (Oculus Pentacam)). The pachymetry map shows the corneal thickness over the entire cornea allowing for a more refined determination of the proposed segment channel depth.

studies (using the Orbscan) found changes to theposteriorcornealsurfacecommonlyoccurredfollowingLASIK, more recent work has shown that routinepost-operativechangesdonotoccurontheposteriorcornealsurface.7Whilechangesontheanteriorsurfacemaybeseeninalloftheaboveconditions(includingpost-LASIK ectasia), only iatrogenic ectasia (or pre-existing keratoconus) show changes on the posteriorcorneal surface. Comparing the posterior cornealsurfacepre-operativelytothepostoperativemapisthemostsensitivemethodtoidentifyearlyectaticchange.Forward protrusion or displacement of the posteriorcorneal surface is found in eyes with post-LASIKectasia (Figure 2). Patients exhibiting such changesshouldbecautionedaboutanyfurtherablativesurgerythatmayfurtherweakenthecornea.

INTRACORNEAL RING SEGMENTS

Scheimpflug imaging has become invaluablein the preoperative, operative and post-operativemanagement. Elevation based tomography moreaccuratelycharacterizestrueconemorphology.Manysurgeonsdeterminethetypeofsegment,thenumberofsegmentsandtheplacementofthesegmentsbasedonconemorphology. Traditionally, themorphologyof the cone was categorized by descriptions based

solelyonanteriorcurvatureanalysis.Sagittalcurvature,however,isaverypoorindicatorofconelocationandmorphology. Curvaturemaps,whether generatedbyplacidosystemsorgeneratedbyelevationsystems,donotdepict shape and convey false information aboutthe size and location of the cone. Elevation mapsmoreaccurately locate theconeandgive thesurgeonamore reliable roadmap toplan their surgery. Thecomplete pachymetric map (Figure 3) is useful fordeterminingthedepthofthetunnel(eithermechanicalor femtosecond). A circle can be drawn on thepachymetric map corresponding to the location andsize of the proposed segment tunnel. The thinnestreadinginthetunnelareamaythenbeusedtocomputethedesireddepth (e.g.75%of the thinnest reading).Usingthefullpachymetricmapisamorereliablewayto compute the proposed tunnel depth as there maybesignificantvariationincornealthicknessindifferentregionsofthecornea.Thisiseventruerinabnormalcorneas. Postoperatively,theScheimpflugimagesthemselvesmay be evaluated to determine the placement of theringsegments(Figure 4).Thequalityoftheindividualimages is oftenof sufficient quality todetermine theshape of the insert and to measure the depth of theplacement.

18

Highlights of Ophthalmology • Vol. 38, Nº 2

DESCEMETS’ STRIPPING ENDOTHELIAL KERATOPLSTY - DSE(A)K

Descemets’ stripping endothelial keratoplastyis becoming the treatment of choice for cornealedema secondary to endothelial dysfunction. Oneof the advantages of this surgery compared to full-thickness penetrating keratoplasty is the preservationoftheanteriorcornealarchitecture.8Eventhoughtheanterior corneal surface is unchanged, patients oftenhave a hyperopic shift secondary to changes to theposteriorcornealsurface.Standardreflective(Placido)topography is incapable of imaging the posteriorsurface. Scheimpflug cross-sectional imaging canclearlydemonstratetheminuslenseffectofthedonorlenticule(Figure 5). Anterior segment tomography can also be usedto document and follow corneal deturgescence, fordocumenting areas where the lenticule may not beadherent and to see fluid accumulation between thedonor lenticule and the recipient posterior stroma.In short, Scheimpflug imaging allows the surgeon tofollowtherateofhealing,therefractiveeffectandthepositionandhealthofthegraft.

Figure 4. (Single Scheimpflug image (Oculus Pentacam)). This single cross-sectional image shows the hexagonal shape of the intra-corneal segment (INTACS) and the measurement of the depth of placement.

Figure 5. (Single Scheimpflug image (Oculus Pentacam)). This single cross-sectional image clearly demonstrates the minus lens effect of the donor lenticule in DSEK.

Figure 6. (Anterior chamber depth map (Oculus Pentacam)). The anterior chamber depth map shows the distance between the posterior cornea and either the iris or lens. This is a somewhat shallow, but otherwise normal distribution.

PHAKIC INTRAOCULAR LENSES

CurrentphakicIOLsfallintothreemaincategories(anterior chamber, iris supported, and posteriorchamber).Whiletheclearance(distance)betweentheIOLandtheendotheliumisimportantinallthree,itis of critical importance in the surgical planning for

19

www.jphmedical.com

anteriorandirissupportedlenses. Thisisevenmoreproblematic with hyperopic phakic lenses as theiroverallthicknessisgreater. Centralanteriorchamberdepth readings are often inadequate to adequatelyaccess the risk of endothelial touch or compromise,as(dependingonthelensstyleandeyeanatomy)theminimal lens-corneadistancemayoftenoccur in theperiphery. Anterior segment tomography allows forthegenerationofathree-dimensionalreconstructionoftheanteriorsegmentandthegenerationofananteriorchamber depth map. This map shows the anteriorchamberdepthnotonlyatthecenterbutalmostouttothelimbus(Figure 6). With knowledge of the anterior segment andinformationon the geometryof thephakic lens it ispossibletosimulatethelensintheeyeandtogeneratethe approximate clearances that would exist betweentheendotheliumanddifferentpartsoftheeye.Withthis capability, different lenses can be evaluated todeterminesuitability(Figure 7).

IOL COMPUTATIONS IN THE POST REFRACTIVE PATIENT

StandardIOLformulasassumeanormalrelationshipbetweentheanteriorandposteriorcornealsurface.In

fact,allofourmeasurements,suchaskeratometry,haveassumed the same relationship that was described inGullstrand’sreducedeye.TheaccuracyofmodernIOLformulas isa testament tohowwell thisassumptionsworkinnormalpatients.Refractivesurgery,however,haschangedtheplayingfield.Modernlaserrefractivesurgeryalterstheanteriorcornealsurfacewhilenormallyleavingtheposteriorsurfaceunchanged.This,however,dramaticallychangestherelativerelationshipbetweenthetwosurfaces.Theassumptionsthatworkedsowellin normal (not post refractive surgery) eyes, do notapplytoeyesthathavebeensurgicallyaltered.Itisnotuncommontosee“refractivesurprises”of3Dormorein post refractive surgery eyes that undergo cataractandimplantsurgery. Anteriorsegmenttomography,byindependentlymeasuringthetwosurfacesallowsforthecomputationofacornealpowermapthatdoesnotrelyonapresumedposteriorsurface,butononethatisdirectlymeasured.Thereareanumberofapproachesthat have been taken to IOL computation in therefractivesurgerypatient. Twoofthemorecommonones are; true or net corneal power (Figure 8), andJack Holladay’s effective keratometry reading (EKR)(Figure 9).9 Bothof thesehavebeenuseful in caseswhere pre-refractive surgery keratometry (historicalmethod)isnotavailable.

Figure 7. (Phakic IOL simulation (Oculus Pentacam)). Using the known geometry of the phakic lens and the tomographic information of the anterior segment the distances between the proposed phakic IOL and other ocular structures can be computed. Here (upper left) we can see the proposed phakic IOL in the eye.

Figure 8. (True net power (Oculus Pentacam)). The True net power map is a power computation that takes the posterior surface measurements into consideration as opposed to using a standard assumption as is done with keratometry.

20

Highlights of Ophthalmology • Vol. 38, Nº 2

TORIC IOL AND ASTIGMATIC SURGERY

Theamountanddirectionofastigmatismismostoftenapproximatedbythesubjectiverefraction.Thestandard refraction, as well as a wavescan analysismeasuresthetotalrefractionandnotitscomponents.Particularly in cataract patients, the lens opacityor change in lens density often contributes to thetotal astigmatism. Astigmatic surgery, whether by atoric IOL or incisional should correct the non-lensastigmatism.Anteriorsegmenttomographyiscapableof measuring the anterior and posterior cornea,anteriorchamberdepth,anteriorlenssurfaceandlensdensity or transmission. This information allows, atleastinpart,abreakdownofthetotalrefractionintoitscomponents.Often,especiallyineyeswithsignificantlenschange,thesubjectiveastigmatismandthecornealastigmatismcandiffergreatly. NoelAlpins,MDhasbeenaproponentofastigmaticvectoranalysisforyearsandhasrecentlyreleasedasoftwareprogram(Assort®)that can be incorporated directly into a number oftopographicdevicesincludingtheOculusPentacam.10

CONCLUSION

Rotating Scheimpflug cross-sectional analysisallows for a three-dimensional reconstruction of theanterior segment. Just as computerized tomography(CT scanning) represented a huge advancementover standard X-rays, anterior segment tomography

represents a major advance compared to standardreflectivetopography.

References

1. Ciolino JB, Khachikian SS, Belin MW “Comparison of Corneal Thickness Measurements by Ultrasound and Scheimpflug Photography in Eyes that have undergone LASIK” Amer J Ophthal, 2008; 45:75-80

2. Ertan A, Kamburoglu G. Analysis of centration of Intacs segments implanted with a femtosecond laser. J Refract Surg 2007; 33(3):484-487

3. Kwon RO, Price MO, Price FW, Ambrosio R, Belin MW. Pentacam Characterization of Corneas with Fuchs’Dystrophy Treated with Descemet Membrane Endothelial Keratoplasty (DMEK) J Refract Surg (in press)

4. Belin MW, Khachikian SS. (ed) Elevation Based Topography. Highlights of Ophthalmology, City of Knowledge, Panama, 2008

5. Belin MW, Khachikian SS: Keratoconus: I know it when I see it.. Amer J Ophthal. Am J Ophthalmol 2007; 1143:500 - 503

6. Khachikian SS, Belin MW, Ciolino JB. Intrasubject Pachymetric Asymmetry Analysis. J Refract Surg, 2008; 24:606-609

7. Ciolino JB, Khachikian SS, Belin MW. Long-Term Stability of the Posterior Cornea after Laser Insitu Keratomileusis, J Cataract Refract Sur. 2007 Aug; 33(8): 1366-70

8. Terry MA, Chen Es, et. al. Endothelial cell loss after Descemet’s stripping endothelial keratoplasty in a large prospective series. Ophthalmology. 2008; 115(3): 488 - 496

9. Holladay JT, Hill WE, Steinmueller A. Corneal power measurements using Scheimpflug imaging in eyes with prior corneal refractive surgery. J Refract Surg. 2009; 25(10): 862-868

10. Alpins NA. Refractive versus corneal changes after photorefractive

keratectomy for astigmatism. J Refract Surg 1998; 14(4):386-396

Figure 9. (Effective keratometry reading (Oculus Pentacam). The map (lower right) shows the distribution of power somewhat analogous to the true net power above. The graph (lower left) is a representation of how the power changes or progresses. A “normal” distribution suggests greater accuracy or less variability under different lighting conditions. The table (upper left) shows the effective or equivalent K readings at varying optical zones.