326 AJVR, Vol 63, No. 3, March 2002

Llamas (Lama glama) and alpacas (Lama pacos) arepopular pet, show, and production animals in the

United States. Corneal disease is relatively frequentlyidentified in llamas. In a retrospective study,1 it wasdetermined that ocular disease happens significantlymore frequently in llamas than in horses or cattle and

that the cornea is the most commonly affected ocularstructure in llamas with ocular disease. The authors ofthat report theorized that this prevalence may havebeen attributable to the large size and anatomic posi-tioning of the globe, predisposing the eyes of llamas totrauma.

In that same retrospective study,1 cataracts werethe most common lens abnormality recorded in llamas,affecting 20 of 194 (10.3%) animals examined to deter-mine the cause of an ocular problem. Cataracts alsohave been reported in alpacas. In a survey2 of 29alpacas, 3 had cataracts. A guanaco (Lama guanacoe)reportedly had bilateral corneal edema and unilateralcataract and lens coloboma.3 Veterinary ophthalmolo-gists are hesitant to perform cataract surgery oncamelids because of a propensity for the formation ofsevere corneal edema after surgery.4,5 The only reporta ofcataract removal in a camelid describes a llama criathat underwent extracapsular cataract extraction anddeveloped severe corneal edema in both eyes and anulcer and perforation of 1 eye.

The aqueous humor of llamas and alpacas is simi-lar to that of other species (eg, dogs, horses) that cansuccessfully undergo cataract surgery.6 Because proteinconcentrations, pH, osmolality, and ionic contents ofthe aqueous humor of camelids do not differ signifi-cantly from those of other species, it was hypothesizedthat camelids may have a fragile corneal endotheliumand that damage to it may be responsible for postoper-ative corneal edema.6

Corneal endothelium is a single layer of hexago-nally shaped cells that forms a barrier between thecorneal stroma and aqueous humor. A minimum criti-cal endothelial cell density (ECD) is necessary tomaintain corneal function in humans.7 Cornealendothelial cells and corneal clarity may be compro-mised by disease processes or by surgical manipulationof the cornea or anterior chamber of the eyes.

Values for ECD have been reported in manyspecies.8-14,b Ocular pathologic and age-related changesthat affect ECD have been reported in humans15-17 anddogs.18-20 To the authors’ knowledge, there have notbeen any reports on ECD in camelids. The objectives ofthe study reported here were to determine centralcorneal ECD, corneal thickness (CT) in 5 areas, andhorizontal and vertical corneal diameters (CD) in nor-mal eyes of llamas and alpacas. We also intended todetermine whether there was an interaction ofresponse variables (age, sex, and side [left eye vs righteye]) on ECD, CT, or CD.

Materials and MethodsAnimals—Thirty-six llamas and 20 alpacas were used in

Received Aug 31, 2001.Accepted Oct 26, 2001.From the Department of Small Animal Clinical Sciences, College of

Veterinary Medicine, University of Florida, Gainesville, FL 32610-0126 (Andrew); and the Department of Clinical Sciences, Collegeof Veterinary Medicine, The Ohio State University, Columbus, OH43210 (Willis, Anderson). Dr. Willis’ present address is RowleyMemorial Animal Hospital, 171 Union St, Springfield, MA 01105.

Presented in part at the 31st Annual Meeting of the AmericanCollege of Veterinary Ophthalmologists, Montreal, QC, Canada,Oct 12, 2000.

Address correspondence to Dr. Andrew.

Density of corneal endothelial cells, cornealthickness, and corneal diameters in normal eyesof llamas and alpacas

Stacy E. Andrew, DVM; A. Michelle Willis, DVM; David E. Anderson, DVM, MS

Objective—To determine density of corneal endothe-lial cells, corneal thickness, and corneal diameters innormal eyes of llamas and alpacas.Animals—36 llamas and 20 alpacas.Procedure—Both eyes were examined in eachcamelid. Noncontact specular microscopy was usedto determine density of corneal endothelial cells.Corneal thickness was measured, using ultrasono-graphic pachymetry. Vertical and horizontal cornealdiameters were measured, using Jameson calipers.Results—Values did not differ significantly betweenthe right and left eyes from the same camelid. Therewas no significant effect of sex on density of cornealendothelial cells or corneal thickness in eitherspecies. Mean density of endothelial cells was 2,669cells/mm2 in llamas and 2,275 cells/mm2 in alpacas.Density of endothelial cells decreased with age in lla-mas. Polymegathism was observed frequently in bothspecies. Mean corneal thickness was 608 µm for lla-mas and 595 µm for alpacas. Corneal thickness anddensity of endothelial cells were negatively correlatedin llamas. Older (> 36 months old) llamas had signifi-cantly larger horizontal and vertical corneal diametersthan younger llamas, and older alpacas had a signifi-cantly larger vertical corneal diameter than youngeralpacas.Conclusions and Clinical Relevance—Density ofcorneal endothelial cells is only slightly lower incamelids than other domestic species. Density ofendothelial cells decreases with age in llamas. Age orsex does not significantly affect corneal thickness innormal eyes of llamas and alpacas. Specularmicroscopy is useful for determining density ofcorneal endothelial cells in normal eyes of camelids.(Am J Vet Res 2002;63:326–329)

01-08-0242R.qxd 2/8/2002 9:49 AM Page 326

the study. Llamas ranged from 9 to 200 months old (median,17 months). Eleven were sexually intact males (range, 3 to47 months old; median, 10 months), 10 were castrated males(range, 11 to 20 months old; median, 15 months), and 15were females (range, 9 to 200 months old; median, 80months). Alpacas ranged from 22 to 204 months old (medi-an, 48 months). Ten were sexually intact males (range, 22 to141 months old; median, 33 months), and 10 were females(range, 27 to 204 months old; median, 48 months).

All animals were housed at a university research farm.Each camelid was manually restrained, and the anterior seg-ment of both eyes was examined with diffuse illumination,using a transilluminator. Results of these examinations werenormal. Pupil dilation was accomplished by use of a solutionof 1% tropicamide.c Examination of the posterior segment ofeach eye was performed by use of a 20-diopter indirect lens,and results were considered normal. None of the camelidshad a history of ocular problems or clinical signs consistentwith ocular disease.

Corneal diameter—Topical anesthetic (0.5% propara-caine)d was applied to each eye, and Jameson calipers thenwere used to measure the diameter of each cornea. Diameterfrom limbus to limbus was measured in horizontal and verti-cal directions.

Density of corneal endothelial cells—Animals weresedated by IV administration of xylazine hydrochloride.e

Llamas were administered xylazine at a dose of 0.3 mg/kg,and alpacas were administered the drug at a dose of 0.6mg/kg. Animals were positioned in sternal recumbency, andtheir heads were maintained in an elevated position andplaced on a headrest. The ECD was determined for each eye,using a noncontact specular microscopef in automatic ormanual mode. The instrument used differential focusing onthe corneal epithelial and endothelial cell surfaces to imagethe endothelial cells in selected areas. A required workingdistance of 25 mm was used. A photographic area of the cen-tral cornea measuring 0.2 X 0.5 mm was analyzed. Fifteenrepresentative endothelial cells were selected for analysis.Values for minimum, maximum, and mean ± SD cell sizeswere calculated.

Corneal thickness—Ultrasonographic pachymetry wasused to determine CT at 5 locations (central, dorsal, ventral,medial, temporal) in both eyes of each animal. Corneal mea-surements at peripheral locations (ie, dorsal, ventral, medial,and temporal) were obtained approximately 5 mm from thelimbus in the clear portion of the cornea. Corneal thicknesswas determined, using a 20-MHz ultrasonographicpachymeterg that obtains consecutive A-scan measurementsof CT during recording and internally eliminates measure-ments > 5 µm from the mean before displaying a CT value.The transducer was apposed gently to the cornea in each ofthe 5 measurement locations without indenting the cornea.21

A value of 1,640 m/s was used for velocity of sound throughthe cornea.

Statistical analysis—Response variables in the studyreported here were ECD, CT, horizontal CD, and vertical CD.Univariate and multivariate ANOVA were used to evaluatethe effect of sex, side (right eye vs left eye), and age on thesevariables in llamas and alpacas.h Data were expressed as mean± SEM. Values of P < 0.05 were considered significant.

ResultsDensity of corneal endothelial cells of llamas—

Mean ± SEM central ECD for all eyes of llamas was2,669 ± 56 cells/mm2. Values did not differ significant-

ly between right and left eyes of the same llama. TheECD decreased significantly (P = 0.004) with increas-ing age in llamas, but sex did not significantly affectECD. Mean ECD in younger (≤ 36 months old) llamaswas 2,778 ± 82 cells/mm2, whereas mean ECD in older(> 36 months old) llamas was 2,516 ± 63 cells/mm2

(Table 1). Heterogeneity of cells (polymegathism andpleomorphism) was observed frequently (Fig 1).

Density of corneal endothelial cells of alpacas—Mean central ECD for all eyes of alpacas was 2,275 ±90 cells/mm2. Values for ECD did not differ significant-ly between right and left eyes of the same alpaca. Therewas not a significant (P = 0.066) effect of age on ECDin alpacas, and sex did not significantly affect ECD(Table 1). Heterogeneity of cells was observed fre-quently.

Corneal thickness of llamas—Mean CT for alleyes of llamas was 608 ± 5 µm. Corneas of younger lla-mas were significantly (P = 0.04) thinner (592 ± 10µm), compared with corneas for older llamas (630 ±

AJVR, Vol 63, No. 3, March 2002 327

Table 1—Comparison of corneal values between older andyounger camelids

ECD CT HCD VCDSpecies Age (mo) (cells/mm2) (µµm) (mm) (mm)

Llama � 36 2,778 � 82a 592 � 10a 28.7 � 0.2a 21.5 � 0.2a

� 36 2,516 � 63b 630 � 12b 32.1 � 0.3b 24.2 � 0.4b

Alpaca � 36 2,472 � 122 590 � 4 30.1 � 0.5 21.6 � 0.3a

� 36 2,144 � 120 598 � 6 30.3 � 0.2 22.7 � 0.2b

Values reported are mean � SEM.a,bFor each species, values within a column with different superscript let-

ters differ significantly (P � 0.05).ECD = Endothelial cell density. CT = Corneal thickness. HCD = Horizontal

corneal diameter. VCD = Vertical corneal diameter.

Figure 1—Specularmicroscopic imageobtained from the lefteye of a 68-month-oldfemale llama with acentral endothelialcell density of 2,600cells/mm2. The whitebox indicates a photo-graphic area measur-ing 0.42 X 0.50 mmthat was analyzed.Notice the poly-megathism and pleo-morphism of thecorneal endothelialcells.

01-08-0242R.qxd 2/8/2002 9:49 AM Page 327

12 µm; Table 1). Although it appeared that sex signifi-cantly affected CT of llamas (young males had thinnercorneas than old females), there was an overrepresen-tation of young males and too few young females foraccurate comparisons of age-sex interactions.

Corneal thickness of alpacas—Mean CT for alleyes of alpacas was 595 ± 9 µm. We did not detect sig-nificant interactions between sex and age when datawere combined. Examination of scatter plots of CTversus age revealed a pattern of reduced CT in all quad-rants in younger alpacas.

Corneal diameters of llamas—Mean horizontalCD for all eyes of llamas was 28.2 ± 0.4 mm, whereasmean vertical CD was 22.5 ± 0.3 mm. Older llamas (n= 15) had significantly (P < 0.001) larger horizontal(32.1 ± 0.3 mm) and vertical CD (24.2 ± 0.4 mm),compared with corresponding values for 21 youngerllamas (28.7 ± 0.2 and 21.5 ± 0.2 mm for horizontaland vertical CD, respectively; Table 1).

Corneal diameters of alpacas—Mean horizontalCD for all eyes of alpacas was 30.2 ± 0.2 mm, whereasmean vertical CD was 22.2 ± 0.2 mm. Values for bothCD did not differ significantly between the right andleft eyes of the same alpaca. Older alpacas (n = 10) hadsignificantly (P < 0.001) larger vertical CD (22.7 ± 0.2mm), compared with the value for 10 younger alpacas(21.6 ± 0.3 mm; Table 1).

DiscussionOcular disease is relatively common in camelids.

In a comprehensive retrospective study1 of ocular dis-ease in llamas, 6% of llamas examined at veterinaryteaching hospitals had an ocular problem, with thecornea being the most frequently affected ocular struc-ture (41%). Camelids have extremely large prominenteyes, which may predispose them to traumatic cornealulcers or lacerations.

The corneal endothelium is vital to maintainingcorneal clarity. During normal aging, cells are lost, andremaining cells spread to fill gaps. Corneal endotheli-um is a monolayer cell population that does not under-go mitosis.22 When there is damage to the endothelium,remaining cells enlarge to fill gaps.23 Human beings areborn with approximately 1 million corneal endothelialcells, and this number decreases by two thirdsthroughout life.24 A minimum critical ECD of 400 to700 cells/mm2 is required to sustain corneal viability.25

In the study reported here, mean ECD was 2,673cells/mm2 in llamas and 2,275 cells/mm2 in alpacas.These values are only slightly lower than those report-ed for normal eyes of dogs (2,816 cells/mm2),11 cats(2,668 and 2,794 cells/mm2),10,b and horses (3,155cells/mm2).12 Heterogeneity of endothelial cells was evi-dent in llamas and alpacas of various ages.Polymegathism (ie, variability in cell size) and pleo-morphism (ie, detection of cells that are not hexago-nal) may indicate stress to the endothelial layer or anabnormal cytoskeleton of the endothelial cells.26

Veterinary ophthalmologists are hesitant toremove cataracts or perform other intraocular surgeryon camelids because of a propensity for postoperative

complications.4 Cataract removal reportedly has limit-ed success because of severe postoperative cornealedema and uveitis.4,a The edema can result in cornealswelling and ulcers that may be progressive and blindthe animal.a Many cataracts in llamas are congenital,4

and it is presumed that cataract removal is most oftenperformed on young llamas with the highest possiblecell density. There are several theories regarding themechanism of this postoperative corneal edema. TheECD in llamas may be lower than expected, theendothelium may be extremely sensitive, or there maybe severe postoperative inflammation.4,5 Results of thestudy reported here did not support a low ECD valueas the sole cause of postoperative corneal decompensa-tion. Polymegathism and pleomorphism of endothelialcells may mean that evaluation of ECD alone is not suf-ficient for assessment of endothelial stability.27 Theslightly lower ECD values of camelids combined withan inherently unstable endothelial cell layer, concur-rent intraocular surgery, and inflammation may be suf-ficient to compromise the pumping function of theendothelium. The effect of phacofragmentation or var-ious irrigating solutions in the anterior chamber onECD in dogs has been reported28,29; however, to ourknowledge, this information has not been reported incamelids and could be another potential source ofcomplications.

Mean ECD decreased with increasing age in llamasbut not alpacas in the study reported here. Humanslose approximately 0.56% of their endothelial cellsannually.30 A correlation between decreasing ECD withincreasing age also has been reported in rabbits,9

cats,10,b dogs,19 and horses.12 We were unable to deter-mine the reason that the ECD in alpacas did notdecrease significantly with increasing age. However, alarger sample population that would include youngeranimals may have revealed that increasing age is animportant factor for decreases in ECD.

Corneal thickness is considered an indirect quan-titative measure of ECD and endothelial cell function.7

We determined that corneas of younger llamas werethinner than those of older llamas, and there also wasa pattern of reduced CT in younger alpacas. The CTthickness did not differ significantly among the 5 loca-tions measured. Central pachymetry is a representativemeasurement of CT that may be relevant in a clinicalsetting.

Horizontal and vertical CD were determined.Values did not differ significantly between species,although llamas are substantially larger than alpacas, asdetermined on the basis of body weight and overall size.

Use of a specular microscope to determine ECD is apractical way to evaluate corneal endothelium in sedat-ed llamas and alpacas. Additional studies need to be per-formed in camelids to determine the cause of severepostoperative corneal edema following intraocularsurgery. We believe that use of a specular microscope todetermine ECD may have great potential for use incamelids before and after intraocular or corneal surgery.

aIngram KA, Sigler RL. Cataract removal in a young llama (abstr), inProceedings. Annu Meet Am Assoc Zoo Vet 1983;95–97.

bIgarashi R, Maeda Y, Igarashi O, et al. Examination of the normal

328 AJVR, Vol 63, No. 3, March 2002

01-08-0242R.qxd 2/8/2002 9:49 AM Page 328

AJVR, Vol 63, No. 3, March 2002 329

feline corneal endothelium with a noncontact specular microscope(abstr), in Proceedings. 30th Annu Meet Am Coll Vet Ophthalmol1999;79.

cTropicamide, Bausch & Lomb Pharmaceutical, Tampa, Fla.dAlcaine, Alcon Laboratories Inc, Fort Worth, Tex.eRompun 30 mg/ml, Bayer Corp, Shawnee Mission, Kan.fTopcon SP-2000P, Topcon America, Paramus, NJ.gDGH500, DGH Technology Inc, Exton, Pa.hSystat 6.0 for Windows, SPSS Inc, Chicago, Ill.

References1. Gionfriddo JR, Gionfriddo JP, Krohne SG. Ocular diseases of lla-

mas: 194 cases (1980–1993). J Am Vet Med Assoc 1997;210:1784–1787.2. Gelatt KN, Otzen Martinic GB, Flaneig JL, et al. Results of

ophthalmic examinations of 29 alpacas. J Am Vet Med Assoc 1995;206:1204–1207.

3. Barrie KP, Jacobson E, Peiffer RL. Unilateral cataract withlens coloboma and bilateral corneal edema in a guanaco. J Am VetMed Assoc 1978;173:1251–1252.

4. Gionfriddo JR. Ophthalmology. Vet Clin North Am FoodAnim Pract 1994;10:371–382.

5. Gionfriddo JR, Friedman DS. Ophthalmology of SouthAmerican camelids: llamas, alpacas, guanacos and vicunas. In:Howard JL, ed. Current veterinary therapy 4: food animal practice.Philadelphia: WB Saunders Co, 1999;644–648.

6. Aubin ML, Gionfriddo JR, Mama KR, et al. Analysis ofaqueous humor obtained from normal eyes of llamas and alpacas. AmJ Vet Res 2001;62:1060–1062.

7. Faulkner WJ, Varley GA. Corneal diagnostic techniques. In:Krachmer JH, Mannis MJ, Holland EJ, eds. Cornea: fundamentals ofcornea and external disease. St Louis: CV Mosby Co, 1997;275–281.

8. Meyer LA, Ubels JL, Edelhauser HF. Corneal endothelial mor-phology in the rat. Effect of aging, diabetes, and topical aldose reduc-tase inhibitor treatment. Invest Ophthalmol Vis Sci 1988;29:940–948.

9. Doughty MJ. The cornea and corneal endothelium in theaged rabbit. Optom Vis Sci 1994;71:809–818.

10. Peiffer RL, DeVanzo RJ, Cohen KL. Specular microscopicobservations of clinically normal feline corneal endothelium. Am J Vet Res 1981;42:854–855.

11. Stapleton S, Peiffer RL. Specular microscopic observationsof the clinically normal canine corneal endothelium. Am J Vet Res1979;40:1803–1804.

12. Andrew SE, Ramsey DT, Hauptman JG, et al. Density ofcorneal endothelial cells and corneal thickness in eyes of euthana-tized horses. Am J Vet Res 2001;62:479–482.

13. Greiner JV, Kenyon KR. Corneal aging. In: Albert DM,Jakobiec FA, eds. Principles and practice of ophthalmology. Basic sci-ences. Philadelphia: WB Saunders Co, 1994;689–696.

14. Collin HB, Grabsch BE, Johnston AW. An accurate in vitroassessment of the corneal endothelium. Am J Optom Physiol Opt1982;59:5–12.

15. Rosenblum P, Stark WJ, Maumenee IH, et al. HereditaryFuchs’ dystrophy. Am J Ophthalmol 1980;90:455–462.

16. Gagnon MM, Boisjoly HM, Brunette I, et al. Cornealendothelial cell density in glaucoma. Cornea 1997;16:314–318.

17. Yee RW, Matsuda M, Schultz RO, et al. Changes in the nor-mal corneal endothelial cellular patterns as a function of age. CurrEye Res 1985;4:671–678.

18. Gwin RM, Polack FM, Warren JK, et al. Primary caninecorneal endothelial cell dystrophy: specular microscopic evaluation,diagnosis and therapy. J Am Anim Hosp Assoc 1982;18:471–479.

19. Gwin RM, Lerner I, Warren JK, et al. Decrease in caninecorneal endothelial cell density and increase in corneal thickness asfunctions of age. Invest Ophthalmol Vis Sci 1982;22:267–271.

20. Gwin RM. Canine corneal endothelium: its role in cornealclarity and response to age and noxious stimuli, in Proceedings. 30thGaines Vet Symp 1981;1–7.

21. Wheeler NC, Morantes CM, Kristensen RM, et al.Reliability coefficients of three corneal pachymeters. Am J Ophthalmol 1992;113:645–651.

22. Pepose JS, Ubels JL. The cornea. In: Hart WM, ed. Adler’sphysiology of the eye. 6th ed. St Louis: CV Mosby Co, 1992;29–70.

23. Van Horn DL, Sendele DD, Seideman S, et al. Regenerativecapacity of the corneal endothelium in rabbit and cat. InvestOphthalmol Vis Sci 1977;16:597–613.

24. Laule A, Cable MK, Hoffman CE, et al. Endothelial cell pop-ulation changes of human cornea during life. Arch Ophthalmol1978;96:2031–2035.

25. Bourne WM, Kaufman HE. Specular microscopy of humancorneal endothelium in vivo. Am J Ophthalmol 1976;81:319–323.

26. Schultz RO, Matsuda M, Yee RW, et al. Corneal endothelialcell changes in type I and type II diabetes mellitus. Am J Ophthalmol1984;98:401–410.

27. Yee RW, Matsuda M, Edelhauser HF. Wide-field endothelialcounting panels. Am J Opthalmol 1985;99:596–597.

28. Nasisse MP, Cook CS, Harling DE. Response of the caninecorneal endothelium to intraocular irrigation with saline solution,balanced salt solution, and balanced salt solution with glutathione.Am J Vet Res 1986;47:2261–2265.

29. Gwin RM, Warren JK, Samuelson DA, et al. Effects of pha-coemulsification and extracapsular lens removal on corneal thick-ness and endothelial cell density in the dog. Invest Ophthalmol Vis Sci1983;24:227–236.

30. Bourne WM, Nelson LR, Hodge DO. Central cornealendothelial cell changes over a ten-year period. Invest Ophthalmol VisSci 1997;38:779–782.

01-08-0242R.qxd 2/8/2002 9:49 AM Page 329