ARTICLE
Bullous Keratopathy as a Progressive Disease: Evidence
From Clinical and Laboratory Imaging Studies
Naoyuki Morishige, MD, PhD, and Koh-Hei Sonoda, MD, PhD
Abstract: Bullous keratopathy is categorized as a corneal endo-
thelial disease. However, pathological changes, including subepi-
thelial fibrosis and the accumulation of extracellular matrix, have
been detected in the corneal stroma of individuals with this
condition. In vivo confocal microscopy allows the visualization of
human corneal cellular structures and has provided information
regarding how eyes are affected by various diseases. However, the
determination of disease pathogenesis on the basis of in vivo
confocal microscopic observations is problematic. We evaluated
the structural alterations in the corneal stroma of eyes affected by
bullous keratopathy using second harmonic generation microscopy
and laser confocal immunofluorescence microscopy of whole-mount
preparations. Using these approaches, we detected the transdiffer-
entiation of keratocytes into fibroblasts and myofibroblasts at the
anterior and posterior stroma and the presence of subepithelial
fibrosis at the anterior stroma and disorganized collagen lamellae at
the posterior stroma of the bullous keratopathy cornea. These
changes were only detected in specimens from eyes with stromal
edema lasting at least 12 months. Similar time-dependent changes
were apparent by using in vivo confocal microscopy in the corneal
stroma of patients with bullous keratopathy after performing
a Descemet stripping automated endothelial keratoplasty surgery
and were associated with an unfavorable outcome with regard to
postoperative visual acuity. Our observations suggest that patholog-
ical changes in the corneal stroma of patients with bullous
keratopathy are progressive and affect postoperative visual acuity
after a Descemet stripping automated endothelial keratoplasty
surgery is performed.
Key Words: bullous keratopathy, second harmonic generation,
scarring, endothelial keratoplasty
(Cornea 2013;32(Suppl):S77–S83)
B ullous keratopathy is characterized by corneal endothelial
decompensation associated with irreversible stromal
edema. Conditions or events that give rise to this disease
include Fuchs corneal endothelial dystrophy, cytomegalovi-
ral corneal endotheliitis, exfoliation syndrome, birth injury,
and endothelial injury caused by intraocular surgeries such as
cataract or glaucoma surgeries, including laser iridotomy for
From the Department of Ophthalmology, Yamaguchi University Graduate
School of Medicine, Yamaguchi, Japan.
The authors have no funding or conflicts of interest to disclose.
Reprints: Naoyuki Morishige, Department of Ophthalmology, Yamaguchi
University Graduate School of Medicine, 1-1-1 Minami Kogushi, Ube,
Yamaguchi 755-8505, Japan (e-mail: morishig@yamaguchi-u.ac.jp).
Copyright © 2013 by Lippincott Williams & Wilkins
Cornea  Volume 32, Number 11, Supplement, November 2013
primary angle closure or acute primary angle closure glau-
coma.1 Although novel treatment strategies for bullous ker-
atopathy have been proposed,2,3 keratoplasty remains the
treatment option adopted for most patients. However, the
surgical strategy is undergoing a change, from penetrating
keratoplasty (PKP) to endothelial keratoplasty such as
Descemet stripping automated endothelial keratoplasty
(DSAEK) or Descemet membrane endothelial keratoplasty.4
In bullous keratopathy, the aim of these surgeries is to
restore corneal endothelial function, which is responsible
for the removal of water from the corneal stroma. The
replacement of the corneal stroma, Bowman layer, and cor-
neal epithelium, which occurs during PKP, is not thought to
be necessary.
Studies have indicated that the clinical outcome of
endothelial keratoplasty is more favorable than that of PKP.5,6
The induction of corneal irregular astigmatism is less pro-
nounced for endothelial keratoplasty than for PKP, with the
result that the visual acuity after endothelial keratoplasty is
better than that after performing a PKP.7 In addition, the
frequency of graft rejection after a DSAEK is performed is
lower than that after a PKP.8,9 Such clinical results underlie
the transition from PKP to endothelial keratoplasty, with the
assumption being that the cornea will become clear after the
amelioration of corneal stromal edema by endothelial kerato-
plasty. However, some patients with bullous keratopathy do
not achieve favorable postoperative visual acuity after sur-
gery, with interface opacity thought to be responsible for this
failure. We have focused on the clinical course of bullous
keratopathy, especially the duration of stromal edema, in an
attempt to identify the factors that might affect visual acuity
after an endothelial keratoplasty is performed. In this review,
we describe our findings that suggest that stromal edema
gives rise to pathological changes in the corneal stroma in
a time-dependent manner in individuals with bullous keratop-
athy. Our observations suggest that bullous keratopathy is
a progressive disease with a pathology related to the duration
of stromal edema.
DETECTION OF PATHOLOGICAL CHANGES IN
THE CORNEAL STROMA BY APPLICATION
OF SECOND HARMONIC GENERATION
IMAGING MICROSCOPY AND
WHOLE-MOUNT IMMUNOSTAINING
To detect and analyze the structural changes in the
cornea with bullous keratopathy, we applied second harmonic
generation (SHG) imaging microscopy10–12 and whole-mount
www.corneajrnl.com | S77
Morishige and Sonoda Cornea  Volume 32, Number 11, Supplement, November 2013

fluorescence microscopic analysis of F-actin and a–smooth
muscle actin (aSMA).13 SHG imaging microscopy allows the
observation of 3-dimensional structures of collagen fibers and
lamellae in the cornea without the need for sectioning of the
tissue.11 It also allows the observation of a wider region of the
specimen than does a conventional microscopic analysis of
sectioned tissue (230 vs. 6–10 mm with a 40· objective lens).
A series of images obtained along the z-axis of a specimen
can be reconstructed into 3-dimensional images that can also
be analyzed as projection images. These properties of SHG
imaging microscopy render it more capable of detecting
abnormalities in the 3-dimensional structure of collagen fibers
and lamellae compared with light microscopy of sectioned
samples.14 In addition to SHG imaging microscopy, we
applied fluorescence microscopy to corneal tissue blocks
(2 · 2 mm). Again, sectioning is not required, allowing
the observation of keratocytes in the corneal stroma without
any artifactual damage inflicted by the sectioning process.
Staining with phalloidin and immunostaining with antibodies
specific for aSMA allow the detection of transdifferentiation
of keratocytes into fibroblasts and myofibroblasts.
CHANGES IN THE STRUCTURE OF COLLAGEN
LAMELLAE AND KERATOCYTE PHENOTYPE IN
THE BULLOUS KERATOPATHY CORNEA
Frontal sectional SHG images of the anterior stroma of
the normal human cornea revealed the presence of short and
densely packed collagen fibers (Fig. 1A). Fibers oriented in
the same direction formed lamellae that were feather shaped
and randomly oriented.11,12,16 Similar images of the posterior
stroma of the normal cornea revealed collagen fibers that were
longer than at the anterior stroma (Fig. 1B). Collagen lamellae
also appeared wider, indicating that the angle they form rel-
ative to the Bowman layer or Descemet membrane is flatter
than that formed by lamellae at the anterior stroma.11,16
Observation of the anterior region of the bullous keratopathy
cornea revealed the presence of an abnormal collagen

FIGURE 1. SHG microscopic images of the normal human cornea and the bullous keratopathy cornea. A and B, Frontal sectional
SHG images of the anterior and posterior stroma, respectively, of a normal cornea. C and D, Frontal sectional SHG images of
a lesion above the Bowman layer and of the posterior stroma, respectively, in a bullous keratopathy cornea. Note the loss of
alignment of collagen lamellae at the posterior stroma. E and F, SHG projection images reconstructed from stacks of frontal
sectional images of the anterior and posterior stroma, respectively, of a bullous keratopathy cornea. Scale bar, 50 mm. Modified
and reprinted from Morishige et al14,15 with permission from the Association for Research in Vision and Ophthalmology.

S78 | www.corneajrnl.com  2013 Lippincott Williams & Wilkins

Cornea  Volume 32, Number 11, Supplement, November 2013 Bullous Keratopathy as a Progressive Disease

structure above the Bowman layer (Fig. 1C), corresponding to
subepithelial fibrosis, although the feather-shaped structure of
collagen lamellae in the anterior stroma was well main-
tained.14 In the posterior stroma of the bullous keratopathy
cornea, the arrangement of collagen lamellae changed from
well aligned to irregular (Fig. 1D), and cracked lamellae were
observed.15 Reconstructed 3-dimensional projection images
of the anterior stroma of the bullous keratopathy cornea also
revealed an abnormal collagen-containing structure above the
Bowman layer and the well-maintained interwoven structure
of collagen lamellae (Fig. 1E).14 Such images also revealed
irregular collagen lamellae at the posterior stroma (Fig. 1F).15
Fluorescence microscopic analysis of whole-mount
preparations revealed 3 different phenotypes for cells at the
anterior or posterior stroma. In the cornea of normal
individuals or some patients with bullous keratopathy, cells
at the anterior stroma were star shaped and did not exhibit
prominent F-actin structures (Fig. 2A). At the posterior stroma
of the normal cornea, the shape of cells was similar to that of
keratocytes at the anterior stroma, but the cells appeared larger
(Fig. 2B). In some corneas with bullous keratopathy, cells at the
anterior and posterior stroma were observed as a spindle shape
with pronounced F-actin structures (Figs. 2C, D), indicating
that keratocytes had been transformed into fibroblasts. Immu-
nofluorescence analysis of aSMA expression further showed
that some fibroblasts at the anterior or posterior stroma of bul-
lous keratopathy corneas had transdifferentiated into myofibro-
blasts (Figs. 2E, F).
WHAT INDUCES COLLAGEN STRUCTURAL
CHANGES AND TRANSDIFFERENTIATION OF
KERATOCYTES IN THE BULLOUS
KERATOPATHY CORNEA?
Changes in the extracellular matrix of the bullous
keratopathy cornea have been described,17,18 but the factors

FIGURE 2. Whole-mount fluorescence microscopic analysis of the normal human cornea and bullous keratopathy cornea. A–D,
Whole-mount staining for F-actin at the anterior (A, C) and posterior (B, D) stroma of normal (A, B) or bullous keratopathy (C, D)
corneas. The star-shaped keratocytes detected in the normal cornea appeared larger at the posterior stroma. Spindle-shaped cells
with prominent F-actin structures (fibroblasts) were observed in both regions of the bullous keratopathy cornea. F-actin was
stained with fluorescein isothiocyanate–labeled phalloidin (green), and nuclei were stained with Syto 59 (red). E and F, Whole-
mount immunostaining for aSMA (green) at the anterior and posterior stroma of the cornea with bullous keratopathy. Spindle-
shaped cells positive for aSMA expression (myofibroblasts) were observed in both regions (arrows). The asterisk indicates the
Descemet membrane. F-actin was also stained with rhodamine-labeled phalloidin (red), and nuclei were stained with Syto 59
(blue). Scale bar, 50 mm. Modified and reprinted from Morishige et al13 with permission from Wolters Kluwer Health and from
Morishige et al14 with permission from the Association for Research in Vision and Ophthalmology.

 2013 Lippincott Williams & Wilkins www.corneajrnl.com | S79

Morishige and Sonoda
or events that give rise to such changes are unknown. We
focused on the duration of stromal edema as a potential con-
tributing factor to the changes in the structure of collagen
lamellae and the transdifferentiation of keratocytes into fibro-
blasts and myofibroblasts in the cornea of individuals with
bullous keratopathy. We retrospectively reviewed the onset
of clinical stromal edema on the basis of clinical signs, such
as direct observation of epithelial or stromal edema, formation
of Descemet membrane folds, and foreign body sensation or
acute onset of ocular pain suggestive of epithelial erosion, in
patients who underwent a PKP or a DSAEK. We determined
the duration of stromal edema as the time from the onset of
such signs to the time of surgery, and we then evaluated the
relationship between this parameter and changes in collagen
structure and keratocyte phenotype detected by SHG imaging
microscopy and fluorescence microscopic analysis.
Figure 3 shows the relationship between the duration of
stromal edema and the presence or absence of abnormal struc-
tures visualized by SHG imaging at the anterior stroma of the
bullous keratopathy cornea detected by SHG imaging micros-
copy. All tissue specimens were derived from the corneal
button obtained at the time of PKP. Subepithelial fibrosis
was only detected in specimens from individuals with a dura-
tion of stromal edema of at least 12 months.14 We similarly
examined the relationship between the duration of stromal
edema and the presence or absence of fibroblasts in the ante-
rior stroma of the bullous keratopathy cornea (Fig. 4). Again,
fibroblasts were not detected in specimens from patients with
a duration of stromal edema of ,12 months.14 We also exam-
ined the relationship between the duration of stromal edema
and the presence or absence of myofibroblasts in the anterior
stroma of the bullous keratopathy cornea. Similar to the other
pathological changes, the presence of aSMA-positive myofi-
broblasts was only detected in tissue samples from individuals
with a duration of stromal edema of at least 12 months (Fig. 5).13
We subsequently examined the relationship between the dura-
tion of stromal edema and pathological changes at both the
anterior and posterior stroma of patients with bullous keratop-
athy (Fig. 6). Both subepithelial fibrosis at the anterior stroma
and the presence of highly scattered collagen lamellae at the
posterior stroma as revealed by SHG microscopy and the pres-
FIGURE 3. Relationship between the duration of corneal
stromal edema and the presence or absence of subepithelial
fibrosis as detected in projection SHG images for patients with
bullous keratopathy. Modified and reprinted from Morishige
et al14 with permission from the Association for Research in
Vision and Ophthalmology.
S80 | www.corneajrnl.com
Cornea  Volume 32, Number 11, Supplement, November 2013
FIGURE 4. Relationship between the duration of corneal
stromal edema and the presence or absence of fibroblasts at
the anterior stroma as detected by whole-mount staining for
F-actin in patients with bullous keratopathy. Modified and
reprinted from Morishige et al14 with permission from the
Association for Research in Vision and Ophthalmology.
ence of aSMA-positive myofibroblasts at the anterior or pos-
terior stroma determined by immunofluorescence analysis were
only detected in tissue specimens with a duration of stromal
edema of at least 12 months.15
CLINICAL IMPACT OF PATHOLOGICAL
CHANGES IN THE BULLOUS
KERATOPATHY CORNEA
Our laboratory observations of collagen structural alter-
ations (detected by SHG imaging microscopy) and the trans-
differentiation of keratocytes into fibroblasts or myofibroblasts
(detected by whole-mount fluorescence microscopy) in the
corneal stroma of bullous keratopathy patients with a duration
of stromal edema of at least 12 months suggested that
irreversible pathological changes may occur in the corneal
stroma of such individuals. In addition, these changes might
affect the visual acuity after DSAEK, given that in contrast to
PKP, the corneal stroma is not replaced by this procedure. We
FIGURE 5. Relationship between the duration of corneal
stromal edema and the presence or absence of myofibroblasts
at the anterior stroma as detected by whole-mount im-
munostaining for aSMA in patients with bullous keratopathy.
Modified and reprinted from Morishige et al13 with permission
from Wolters Kluwer Health.
 2013 Lippincott Williams & Wilkins
Cornea  Volume 32, Number 11, Supplement, November 2013
FIGURE 6. Relationship between the
duration of corneal stromal edema
and pathological changes at the
anterior or posterior stroma of in-
dividuals with bullous keratopathy.
Collagen structural abnormalities as
revealed by SHG imaging microscopy
included subepithelial fibrosis at the
anterior stroma and irregularly
arranged collagen lamellae at the
posterior stroma. Myofibroblasts
were detected by immunofluores-
cence analysis with antibodies to
aSMA. Individuals positive or nega-
tive for the pathological changes are
represented by closed and open
symbols, respectively. Reprinted from
Morishige et al15 with permission
from the Association for Research in
Vision and Ophthalmology.
therefore reviewed the clinical charts of patients with DSAEK
to evaluate the relationship between the duration of pre-
operative stromal edema and postoperative visual acuity. We
excluded individuals with macular diseases, severe glaucoma
affecting the central visual field, irregular astigmatism such as
that post-PKP, or obvious corneal scarring. Of 56 consecutive
patients with DSAEK seen at the Yamaguchi University
Hospital, 28 individuals were enrolled in the analysis. Figure 7
shows the relationship between the duration of preoperative stro-
mal edema and postoperative visual acuity at 6 months after
DSAEK. We divided the subjects into 2 groups on the basis
of the duration of preoperative stromal edema. Those with a dura-
tion of ,12 months (group A) showed a more favorable post-
operative visual acuity than did those with a duration of at least
12 months (group B).19 Statistical analysis revealed that the pro-
portion of subjects with a postoperative visual acuity of at least
FIGURE 7. Relationship between the duration of preoperative
stromal edema and best-corrected visual acuity (BCVA) at 6
months after the DSAEK was performed for individuals with
bullous keratopathy. Modified and reprinted from Morishige
et al19 with permission from Elsevier.
 2013 Lippincott Williams & Wilkins
Bullous Keratopathy as a Progressive Disease
20/40 was significantly larger for group A than for group B.
A similar pattern was also apparent for visual acuity at 3
months after DSAEK and over the entire follow-up period.19
We also examined the cornea of post-DSAEK patients by
in vivo laser confocal microscopy. For individuals in group
A, the cornea, including the shape of keratocytes, appeared
normal (Figs. 8A–C). In contrast, we detected subepithelial
fibrosis-like changes at the level of the Bowman layer or the
presence of cells resembling fibroblasts or myofibroblasts below
the Bowman layer in all group B subjects (Figs. 8D–F).19 The
pathological changes observed in the cornea with bullous ker-
atopathy by our laboratory investigations were also apparent
after the DSAEK was performed in patients with bullous kerat-
opathy with a long duration of preoperative stromal edema and
were likely responsible for the less favorable outcome of sur-
gery in terms of visual acuity.
LABORATORY AND CLINICAL EVIDENCE FOR
BULLOUS KERATOPATHY AS A
PROGRESSIVE DISEASE
Our laboratory and clinical studies13–15,19 focusing on
the impact of the duration of stromal edema on corneal
pathology and post-DSAEK visual acuity suggest that bullous
keratopathy is a progressive disease. Given that bullous ker-
atopathy develops as a result of endothelial decompensation
after endothelial injury caused by various conditions or
events, it is not surprising that pathological changes are
apparent at the level of the corneal endothelium or posterior
stroma. However, consistent with earlier observations,17,18 we
have detected such changes, including subepithelial fibrosis
as revealed by SHG imaging microscopy and fibroblastic or
myofibroblastic transdifferentiation of keratocytes as revealed
by whole-mount fluorescence microscopy, at the anterior
stroma of the bullous keratopathy cornea. Although the epi-
thelial side of the cornea is unlikely to be directly damaged by
www.corneajrnl.com | S81
Morishige and Sonoda Cornea  Volume 32, Number 11, Supplement, November 2013

FIGURE 8. In vivo confocal microscopy of the cornea of patients with bullous keratopathy at 6 months after DSAEK. A–C,
Confocal images for patients with a duration of preoperative stromal edema of ,12 months (group A in Fig. 7). D–F, Confocal
images for patients with a duration of preoperative stromal edema of at least 12 months (group B). Images were obtained at the
level of the epithelial basal cell layers (A, D), Bowman layer (B, E), and the anterior stroma (C, F). Note the subepithelial fibrosis-like
lesion in (D) and the presence of fibroblasts or myofibroblasts in (E) and (F). Scale bar, 100 mm. Reprinted from Morishige et al19
with permission from Elsevier.

the underlying events or conditions that give rise to bullous
keratopathy, individuals with this disease often manifest cor-
neal epithelial erosion. Such erosion may be a key factor in
the development of pathological changes observed in the
anterior portion of the bullous keratopathy cornea, given that
it results in the exposure of the Bowman layer and anterior
stroma to the external environment including tear fluid. Tear
fluid contains various cytokines and growth factors, including
transforming growth factor-b,20 which may promote the
structural changes observed at the anterior stroma of the bul-
lous keratopathy cornea. Indeed, we have found that patients
with bullous keratopathy with a clinical history of epithelial
erosion manifest myofibroblastic transdifferentiation of kera-
tocytes at the anterior stroma earlier than those without such
a history.13 With the exception of Fuchs dystrophy, infectious
endotheliitis, and exfoliation syndrome, the causes of corneal
endothelial damage in individuals with bullous keratopathy
are not ongoing, with the result that pathological changes
detected at the posterior stroma may not be progressive. How-
ever, structural alterations in the anterior portion of the cornea
caused by epithelial erosion would be expected to be pro-
gressive. Management of the condition of the corneal epithe-
lium in individuals with bullous keratopathy may thus be
important for the outcome of subsequent endothelial
keratoplasty.
CONCLUSIONS
In conclusion, we suggest that bullous keratopathy is
a pathologically and clinically progressive disease. This
notion has important implications for the management of
bullous keratopathy patients before endothelial keratoplasty
and thus warrants further investigation.
ACKNOWLEDGMENTS
The authors thank Atsushi Takahara for permission to
operate the multiphoton microscope used for our SHG
imaging studies. The authors also thank Teruo Nishida,
Tai-ichiro Chikama, Naoyuki Yamada, Norimasa Nomi, and
Yukiko Morita for contributions to our investigations.

S82 | www.corneajrnl.com  2013 Lippincott Williams & Wilkins

Cornea  Volume 32, Number 11, Supplement, November 2013
REFERENCES
1. Shimazaki J, Amano S, Uno T, et al. National survey on bullous kerat-
opathy in Japan. Cornea. 2007;26:274–278.
2. Okumura N, Koizumi N, Ueno M, et al. The new therapeutic concept of
using a rho kinase inhibitor for the treatment of corneal endothelial
dysfunction. Cornea. 2011;30(suppl 1):S54–S59.
3. Okumura N, Koizumi N, Ueno M, et al. Enhancement of corneal endo-
thelium wound healing by Rho-associated kinase (ROCK) inhibitor eye
drops. Br J Ophthalmol. 2011;95:1006–1009.
4. Boimer C, Lee K, Sharpen L, et al. Evolving surgical techniques of and
indications for corneal transplantation in Ontario from 2000 to 2009. Can
J Ophthalmol. 2011;46:360–366.
5. Price MO, Price FW Jr. Endothelial keratoplasty—a review. Clin Exper-
iment Ophthalmol. 2010;38:128–140.
6. Terry MA. Endothelial keratoplasty: clinical outcomes in the two years
following deep lamellar endothelial keratoplasty (an American Oph-
thalmological Society thesis). Trans Am Ophthalmol Soc. 2007;105:
530–563.
7. Price MO, Giebel AW, Fairchild KM, et al. Descemet’s membrane endo-
thelial keratoplasty: prospective multicenter study of visual and refractive
outcomes and endothelial survival. Ophthalmology. 2009;116:2361–2368.
8. Suh LH, Yoo SH, Deobhakta A, et al. Complications of Descemet’s
stripping with automated endothelial keratoplasty: survey of 118 eyes
at One Institute. Ophthalmology. 2008;115:1517–1524.
9. Shih CY, Ritterband DC, Rubino S, et al. Visually significant and non-
significant complications arising from Descemet stripping automated
endothelial keratoplasty. Am J Ophthalmol. 2009;148:837–843.
10. Morishige N, Wahlert AJ, Kenney MC, et al. Second-harmonic imaging
microscopy of normal human and keratoconus cornea. Invest Ophthalmol
Vis Sci. 2007;48:1087–1094.
 2013 Lippincott Williams & Wilkins
Bullous Keratopathy as a Progressive Disease
11. Morishige N, Petroll WM, Nishida T, et al. Noninvasive corneal stromal
collagen imaging using two-photon-generated second-harmonic signals.
J Cataract Refract Surg. 2006;32:1784–1791.
12. Morishige N, Nishida T, Jester JV. Second harmonic generation for
visualizing 3-dimensional structure of corneal collagen lamellae. Cornea.
2009;28:S46–S53.
13. Morishige N, Nomi N, Morita Y, et al. Immunohistofluorescence analy-
sis of myofibroblast transdifferentiation in human corneas with bullous
keratopathy. Cornea. 2011;30:1129–1134.
14. Morishige N, Yamada N, Teranishi S, et al. Detection of subepithelial
fibrosis associated with corneal stromal edema by second harmonic gener-
ation imaging microscopy. Invest Ophthalmol Vis Sci. 2009;50:3145–3150.
15. Morishige N, Yamada N, Zhang X, et al. Abnormalities of stromal structure
in the bullous keratopathy cornea identified by second harmonic generation
imaging microscopy. Invest Ophthalmol Vis Sci. 2012;53:4998–5003.
16. Morishige N, Takagi Y, Chikama T, et al. Three-dimensional analysis of
collagen lamellae in the anterior stroma of the human cornea visualized
by second harmonic generation imaging microscopy. Invest Ophthalmol
Vis Sci. 2011;52:911–915.
17. Ljubimov AV, Burgeson RE, Butkowski RJ, et al. Extracellular matrix
alterations in human corneas with bullous keratopathy. Invest Ophthal-
mol Vis Sci. 1996;37:997–1007.
18. Eagle RC Jr, Laibson PR, Arentsen JJ. Epithelial abnormalities in chronic
corneal edema: a histopathological study. Trans Am Ophthalmol Soc.
1989;87:107–119; discussion 119–124.
19. Morishige N, Chikama T, Yamada N, et al. Effect of preoperative dura-
tion of stromal edema in bullous keratopathy on early visual acuity after
endothelial keratoplasty. J Cataract Refract Surg. 2012;38:303–308.
20. Gupta A, Monroy D, Ji Z, et al. Transforming growth factor beta-1 and
beta-2 in human tear fluid. Curr Eye Res. 1996;15:605–614.
www.corneajrnl.com | S83