Assessment of Corneal Thickness and Keratocyte
Density in a Rabbit Model of Laser In Situ
Keratomileusis Using Scanning Laser Confocal
Microscopy
MICHAEL D. TWA AND MICHAEL J. GIESE
●
PURPOSE:
To determine the repeatability of corneal
thickness and keratocyte density using in vivo confocal
scanning laser microscopy in a rabbit model of laser in
situ keratomileusis.
●
DESIGN:
Prospective, experimental animal study.
●
METHODS:
En face tomographic images of corneal tissue
were captured from 5 New Zealand white rabbits. Central
corneal thickness was compared with conventional ultra-
sonic pachymetry. Keratocyte density was measured as a
function of stromal depth at baseline and 6 weeks after a
130-m lamellar incision in the following regions: first
countable stromal image (30 to 39 m), anterior stroma
(40 to 75 m), incision zone (76 to 150 m), mid stroma
(151 to 250 m), and deep stroma (251 to 400 m).
●
RESULTS:
The mean residual difference between ultra-
sonic and confocal corneal thickness measurements was
2.1 m (95% confidence interval [CI], ؊7.0 to 11.2
m; P ؍ .61). Before the lamellar incision, keratocyte
density was highest in the first countable frame of the
anterior stroma, 53 800 cells/mm
3
(95% CI, 35 000 to
72 000 cells/mm
3
) and was least in deep stroma, 27 100
cells/mm
3
(95% CI, 22 400 to 32 000 cells/mm
3
). Six
weeks after stromal lamellar incision, keratocyte density
was unchanged in the first countable frame of the
anterior stroma, 43 700 cells/mm
3
(95% CI, 31 800 to
55 500 cells/mm
3
; P ؍ .29). There were no changes in
cell density in deeper stromal regions.
●
CONCLUSIONS:
There was excellent agreement be-
tween ultrasonic and confocal microscopy measurements
of corneal thickness. In vivo repeatability of keratocyte
density estimation using scanning laser confocal micros-
copy is comparable with the results of previous reports
using tandem-scanning confocal microscopy. Keratocyte
density was more varied, but not significantly different,
in the anterior-most corneal stroma 6 weeks after a
lamellar incision. (Am J Ophthalmol 2011;152:
941–953. © 2011 by Elsevier Inc. All rights reserved.)
C
ORNEAL THICKNESS IS ONE OF THE PRIMARY
parameters that determine the structural integrity
of the cornea. It is reduced in keratoconus and
other ectatic degenerations and has been shown to de-
crease with prostaglandin analog use for the treatment of
glaucoma.
1
Keratocytes, the resident fibroblastic cells of
the cornea, are responsible for maintenance and integrity
of corneal tissue, and their physiology, morphologic fea-
tures, and numbers are altered by clinical treatments such
as corneal refractive surgery,
2
ultraviolet corneal cross-
linking,
3,4
mitomycin C use,
5
and other common interven-
tions. In this study, we evaluated the use of in vivo
scanning laser confocal microscopy to evaluate the accu-
racy and repeatability of in vivo measurements of corneal
thickness and keratocyte density to determine the suitabil-
ity of these measurements for laboratory and clinical
studies of these structural parameters.
There are several clinical instrument designs available
for in vivo confocal microscopy that provide en face
optical tomographic sections of the cornea and other
anterior segment tissues based on backscattered illumina-
tion principles that do not rely on exogenous contrast
agents. McLaren and associates compared corneal thick-
ness and cell density measurements using different in vivo
confocal microscope designs and showed that these param-
eters differ based on the design principles of the confocal
microscope.
6
Likewise, others have shown that measure-
ments of cell density and corneal thickness differ from
other standard measurement techniques, for example,
specular microscopy or ultrasonic pachymetry.
7,8
Recent
reviews of this clinical imaging technology are available.
9
Nipkow disc-based tandem scanning instruments were
one of the first designs capable of providing in vivo
microscopic images with micron-level resolution.
10
The
optical inefficiency of these instruments requires relatively
high illumination levels and longer image acquisition
times, which has limited their clinical use. A scanning-slit
microscope design was introduced in the 1980s (Nidek,
Inc, Freemont, California, USA), and although light levels
remain somewhat high, this design is optically more
efficient and provides improved image contrast and re-
duced acquisition times.
11
As with tandem scanning con-
focal microscopes, the scanning-slit confocal objective lens
Accepted for publication May 25, 2011.
From the College of Optometry, University of Houston, Houston,
Texas (M.D.T.); and the Nova Southeastern College of Optometry, Fort
Lauderdale, Florida (M.J.G.).
Inquiries to Michael D. Twa, College of Optometry, University of
Houston, 505 J. Davis Armistead Building, Houston, TX 77204-2020;
e-mail: mdtwa@uh.edu
©
2011 BY
E
LSEVIER
I
NC
.A
LL RIGHTS RESERVED
.
0002-9394/$36.00
941
doi:10.1016/j.ajo.2011.05.023