Dehydrated human-amnioticmembrane allograft versus conjunctival autograft after pterygium excision

CONJUNCTIVAL autograft after pterygium excision
is fast becoming the treatment of choice to prevent recurrences.
1-7
Conjunctival autograft as an adjunct was initially
reported to have a recurrence rate of 5.3%. Since then,
further studies have reported recurrence rates of up to
39% attributed to technique of pterygium excision.
Although deemed safer than other adjunctive treatments,
its use is limited in patients with larger defects and in
patients where the superior conjunctiva has to be
preserved.
4, 5
Conjunctival autograft has also been noted
to be unsuccessful in suppressing postsurgical fibrosis.
6, 8, 9
Transplantation of preserved human-amniotic
membrane was recently introduced as an adjunct to pterygium surgery.
10-12
Studies have shown that the basement
membrane promotes epithelial growth and differentiation, reinforces adhesion of basal epithelial cells and
prevents epithelial apoptosis. The stromal matrix
suppresses transforming-growth-factor-b signaling,
proliferation and myofibroblast differentiation of normal
human corneal and limbal fibroblasts, and human
conjunctival and pterygium body fibroblast, thereby
inhibiting unwanted extracellular-matrix production and
scarring.
10, 13
Some studies have shown it to be as effective
as conjunctival autograft in preventing pterygium recurrence without the drawback of disrupting the superior
conjunctiva and its inability to suppress postsurgical
fibrosis.
5, 14-16
Processing the amniotic membrane for transplantation
is a tedious process that requires freezing it at –80
o
C.
15
The advent of dehydrated processed membranes has
significantly made its use much more convenient since it
is ready to use and can be stored at room temperature.
Dehydrated human-amniotic-membrane allografts
(AmbioDry, Okto Ophtho, IOP Inc. Costa Mesa, CA, USA)
are dehydrated, decellularized, sterilized human-amnioticmembrane-tissue grafts. Using the dehydrated membrane
entails hydrating it with saline solution intraoperatively.
Fournier and McLachlan reported the successful use of
AmbioDry in destructive lesions of the conjunctiva, acting
as a reconstructive graft in nonhealing lesions of the ocular
surface.
17
Battle reported the use of AmbioDry following
excision of recurrent pterygium.
18
Both conjunctival autograft and amniotic-membrane
transplantation are successful in preventing pterygium
recurrence in most people. Both procedures, however,
cause a significant amount of discomfort for the patient
due to the number of sutures needed. Thus, tissue
adhesives such as the biologic and biodegradable fibrinbased adhesives have gained popularity lately. They also
induce minimal inflammation.
19
A recent comparative
study reported a 100% success rate for both fibrin glue
and nylon sutures in conjunctival autografts after pterygium excision.
20
It also found that using fibrin glue
resulted in significantly shorter surgery time and lesser
postoperative discomfort.
Our study compared the transplantation of dehydrated
human-amniotic-membrane allograft with conjunctival
autograft after pterygium excision using fibrin glue.

METHODOLOGY
Forty-two consecutive patients with primary pterygia
undergoing pterygium excision at the University of the
Philippines-Philippine General Hospital from November
2004 to March 2005 were prospectively enrolled. General
data, demographic factors, medical and ocular history
were obtained. All patients underwent a complete ophthalmologic examination including visual-acuity testing,
refraction, intraocular-pressure measurement, slitlamp
examination, and anterior-segment photography. Patients
with a history of ocular pathology (other than error of
refraction), ocular surgery or trauma, severe ocularsurface disease, chronic conjunctivitis, glaucoma, family
history of glaucoma, ocular hypertension, narrow
occludable angles, glaucomatous or physiologic disc
cupping, or known hypersensitivity to the components of
commercial fibrin glue (Beriplast P, CSL Behring, King
of Prussia, PA, USA) were excluded. Informed consent
was obtained from all patients. The institutional review
board approved the protocol and informed consent form.

Pterygium Grading
Pterygia were measured and graded according to the
grading scheme proposed by Tan et al. in 1997. The pterygia
were classified into grades 1, 2, or 3 based on slitlamp
biomicroscopy evaluation. Grade 1 (atrophic) included
pterygia in which episcleral vessels under the body of
pterygium are not obscured and clearly distinguished.
Grade 3 (fleshy) included pterygia in which episcleral
vessels underlying the body of the pterygium are totally
obscured. Grade 2 (intermediate) included all other
pterygia that did not fall under the other two grades.
21

Treatment Assignment
Using a table of random numbers, patients were
randomized to receive either pterygium excision with
amniotic-membrane transplantation or pterygium excision
with conjunctival autograft. One investigator examined all
patients pre- and postoperatively. All surgeries were
performed by two surgeons (KTL and RLBS).

Fibrin Glue
Beriplast P is a fibrin sealant that imitates the final stage
of the coagulation process. Fibrinogen is converted into
fibrin on the tissue surface by the action of thrombin.
The fibrin formed is then cross-linked by factor XIIIa, thus
creating a firm, mechanically stable fibrin network.
Aprotinin from bovine lungs is then added to the fibrin
sealant to prevent rapid fibrinolysis. The potential transmission of contaminants is mitigated by screening of
plasma donors, virus removal and inactivation by the
manufacturing process, and pasteurization and clinical
virus-safety measures.
20

Surgical Technique
One drop each of 2.5% phenylephrine (Mydfrin, Alcon
Lab Inc, Fort Worth, TX, USA), 0.15% brimonidine
(Alphagan P, Allergan Inc., Irvine, CA, USA), and 0.5%
antazoline phosphate, 0.05% naphazoline HCl, 0.125%
Zn sulfate (Zincfrin-A, Alcon Lab Inc, Fort Worth, TX,
USA) were instilled thirty minutes prior to surgery. After
instillation of topical proparacaine HCl (Alcaine, Alcon
Lab Inc, Fort Worth, TX, USA), standard preparation and
draping of the involved eye were done. Prior to excision,
the widest dimensions of the pterygium were measured
using a caliper (head to limbus and width at limbal area).
The head of the pterygium was carefully dissected from
the cornea using a surgical blade (No.15). Lidocaine with
epinephrine solution (Xylocaine 2%,Astra-Zeneca,
Sweden) was injected subconjunctivally to balloon the
conjunctiva and delineate the fibrovascular tissue
underneath. The pterygium body was then dissected from
the underlying sclera using Wescott scissors. Subconjunctival fibrovascular tissue was dissected away from the
overlying conjunctival epithelium. The pterygium was
then excised at the base. Atrophic edges of the conjunctiva
were removed. Tenon’s capsule was removed to provide a
clean scleral bed for the placement of the graft. The
greatest vertical and horizontal diameters of the bare sclera
were then measured using a caliper.
The dehydrated human-amniotic-membrane allograft
was prepared according to the manufacturer’s instructions. The dry allograft was trimmed to be oversized by
1 mm at all sides and then hydrated while on the surgical
site with sterile saline solution orienting the graft
basement matrix side up. Several drops of sterile solution
were applied to the allograft at one- to two-minute
intervals for a period of 5 to 10 minutes. One drop of
the fibrinogen solution was applied to the bare sclera
coating the whole surface. Thrombin solution was then
applied over the fibrinogen solution. The amnion graft
was then slid onto the bare sclera making sure that the
graft was completely adherent or apposed to the recipient
conjunctiva. The hydrated amnion graft was noted to be
mobile even with the application of the fibrin glue. Four
anchor sutures (nylon 10-0 Alcon Lab Inc, Fort Worth,
TX, USA) were added to secure the amnion graft to the
underlying sclera. The suture knots were trimmed and
buried.
Free conjunctival grafts oversized by 1 mm at all sides
marked with gentian violet were harvested from the
superotemporal or superonasal conjunctiva of the same
eye for nasal and temporal pterygium, respectively.
Lidocaine with epinephrine solution was injected
subconjunctivally to balloon the area of the graft and
separate it from the underlying Tenon’s capsule. Using
atraumatic conjunctival forceps and Vannas scissors, the
conjunctiva was dissected from Tenon’s capsule taking
care to include as little Tenon’s tissue as possible in the
graft. The graft was then slid carefully on top of the
cornea and kept moist using sterile normal saline
solution. Fibrin glue was applied as mentioned previously.
The graft was then slid onto the bare sclera making sure
tha t it was completely adherent or apposed to the
recipient conjunctiva.
Both grafts were then smoothed, gently pressing them
onto the scleral bed for five minutes. Excess glue was
removed and the graft trimmed if necessary. The speculum
was removed and the patient was asked to blink to test
graft adherence and mobility.
Steroid-antibiotic ointment was then applied on all eyes
and a pressure patch applied for 24 hours.

Postoperative Regimen and Follow-up Examination
After removal of the patch, visual-acuity testing,
biomicroscopic examination, fluorescein staining, and
anterior-segment photography were done. The presence,
appearance, and integrity of the graft were noted. All cases
received identical postoperative regimen of topical steroidantibiotic drops and topical steroid-antibiotic ointment
(Tobradex, Alcon Lab Inc, Fort Worth, TX, USA) tapered
over two months.
All patients were followed up weekly for 2 weeks, then
at one month, two months, and three months. At each
postoperative visit, the same examinations were performed
by a single investigator.

Outcome Measures
This study compared surgery time, efficacy, early
recurrence, postoperative discomfort, and cosmesis in
both groups.
Surgery time was recorded from the start of graft
preparation to the end of surgery.
Graft success was defined as incorporation of the graft
into the surrounding conjunctiva with complete
epithelialization over it; graft failure was defined as
absence of the graft at 2 weeks or less postoperatively.
Pterygium recurrence was defined as the presence of
fibrovascular conjunctival reencroachment extending
beyond the limbus.
Subjective sensations such as pain, foreign body
sensation, tearing, and discomfort were determined and
measured using a 5-point scale adapted from Lim-Bon-
Siong et al:
22
NONE = no pain at all, VERY MILD =
presence of pain but easily tolerated, MILD = presence of
pain causing some discomfort, MODERATE = presence
of pain that interferes with usual activity or sleep, SEVERE
= presence of pain that completely interferes with usual
activity or sleep. The questionnaire was administered at
each follow-up visit.
Three separate masked observers were given photographs of the operated eye taken at 3 months postoperatively and were asked to grade the cosmetic appearance
using a 5-point scale: POOR = very congested graft with
numerous radial vessels, FAIR = congested graft with few
radial vessels, GOOD = slightly congested graft with
minimal radial vessels, VERY GOOD = graft not congested
with very minimal conjunctival vessels and EXCELLENT =
pristine, untouched appearance.
Postoperative complications such as graft ischemia and
necrosis, infection, graft dehiscence, persistent epithelial
defect, steroid-induced glaucoma, and granuloma
formation were likewise noted.

Statistical Analysis
All continuous numerical data were summarized using
descriptive statistics (percentages and frequency
distribution). Tests for homogeneity of sample included
comparing for significant differences in all numerical
continuous data. Data filtration and test for normality
of data revealed skewness values greater than 1. Owing
to the small sample size, nonparametric statistics were
employed. Mann Whitney U Test was used to compare
the significant change in mean ranks between the two
procedures while chi-square was used to detect
independence of the groups and the discrete categories.
Spearman’s rho correlation was used to investigate the
association between pterygium grade and cosmesis.
In order to compare the proportion of ratings for a
specific scale category, the Z-test for two proportions was
utilized.
Comparison of inter-rater agreement was facilitated
using kappa statistics and p-values compared using
Statistica version 1999 (Stat Soft Inc., Tulsa, OK, USA).
To compare for the change in pain, foreign body
sensation, tearing and discomfort severity at baseline and
3 months postsurgery, Wilcoxon signed rank test for
paired data was used for each treatment group while the
Mann Whitney U test was utilized to compare the mean
ranks of scores between the two groups at each period of
observation.
All statistical tests were carried out to reject the null
hypothesis assumed for each study objective at .05 level of
significance and were carried out using the Statistical
Package for the Social Sciences ver. 10 (SPSS, Inc.,
Chicago, IL, USA) and the Statistica version 1999.

RESULTS
All 42 patients met the inclusion criteria. They were
randomized to receive either dehydrated human-amnioticmembrane allograft (n = 21) or conjunctival autograft
(n=21). One patient in the dehydrated-human-amnioticmembrane-allograft group was lost to follow-up. A total
of 41 subjects were included in the final analysis. The
baseline clinical profile and preoperative evaluation are
summarized in Table 1. The youngest patient in this study
was 20 years while the oldest was 74 years (mean of 43 ±
13, median age of 43 years). There were 22 males (54%).
The left eye was most commonly involved (23 or 56%)
and a nasal location was most common. All patients had a
best-corrected visual acuity of ≥ 6/30 except for 1 patient
in the autograft group who had ≥1/60.
Tests of homogeneity revealed no significant difference
between the two groups at baseline in terms of age, sex
distribution, affected eye, location of pterygium, duration
of the pterygium based on history, use of topical
medications, baseline pterygium grading, horizontal and
vertical measurements of pterygium size and intraoperative defect size (p > .05 for all).
Mean surgery times (from application of graft to end
of surgery) were not significantly different between the
two groups (p = 0.16) (Table 2).
All patients in both groups had good graft survival from
day 1 to 3 months postoperatively (Table 2). One patient
in the conjunctival autograft group had graft dehiscence
at day 1 postoperatively, which had to be sutured, but had
subsequent good graft survival thereafter. One patient in
the conjunctival autograft group developed a granuloma
at the junction of the graft and recipient conjunctiva, and
underwent excision of granuloma.
No recurrences of pterygia were noted in both groups
at three months postoperatively (Table 2).
We compared the severity of pain, tearing, foreignbody sensation, and level of discomfort between the two
groups across the specified periods of observation.
Among all patients, only 1 in the conjunctival autograft
group complained of severe pain noted at day 1
postsurgery. Most of the subjects had mild to moderate
pain that gradually abated at 2 weeks and beyond. Pain
severity scores were not significantly different between
the two procedures across the periods of observation
(p = 0.07). The mean tearing-severity scores of patients
who received dehydrated human-amniotic-membrane
allograft were statistically lower compared with those
who received conjunctival autograft at day 1 postoperatively (p = 0.024). Foreign-body-sensation severity was
similar for both groups at baseline, but the rate of
decrease over the ensuing weeks was statistically significant (p < .001). The degree of severity change,
however, was not significantly different between the two
procedures (p = 0.82). The degree of change in
discomfort severity was also similar (p = 0.31).
Cosmetic-grading results showed a statistically higher
proportion of “GOOD” ratings by patients randomized
to the conjunctival autograft compared with those who
received dehydrated human-amniotic-membrane allograft
(Table 2). Although there was a slightly higher proportion
of “POOR” ratings in the amniotic-membrane group (12%
versus 3%), this was not statistically significant (p = .06)
(Table 2).

DISCUSSION
This study compared suture-anchored AmbioDry graft
with conjunctival autograft using fibrin glue to secure both
grafts. Mean surgery time was noted to be similar in both
groups. Both treatment groups also exhibited similar graft
survival rates. No recurrences occurred by the end of the
three-month observation period. No conclusion, however,
can be made from our data regarding long-term
recurrence rates. Among the postoperative symptoms,
only tearing scores showed a significant difference
between the 2 groups.
To the best of our knowledge, no clinical studies have
yet been published comparing dehydrated humanamniotic-membrane allograft with conjunctival autograft
after pterygium excision. Battle reported one case where
AmbioDry was used following excision of recurrent
pterygium.
18
We noted a significant learning curve in using AmbioDry.
In its dry state, the graft was difficult to handle and to orient.
Handling the translucent, hydrated graft also proved to be
a challenge especially in the presence of blood.
In this study, anchoring sutures had to be placed in the
AmbioDry group because of residual graft mobility after
fibrin-glue application. Preliminary results of a similar
study using a different preparation of fibrin glue (Tisseel,
Baxter, Vienna, Austria) showed complete adherence of
the graft to the bare sclera (personal communication with
Roy Chuck, MD, Wilmer Eye Institute, Baltimore, MD,
USA). Differences in the molecular component of the
fibrin-glue preparations may be the reason for the
disparity in results.
There were no serious complications related to both
procedures observed in this study. One patient developed
a granuloma 2 months postoperatively, which was
subsequently excised. One patient had a partial graft
dehiscence noted on the first postoperative day that had
to be sutured. All sutures were released 1 month
postsurgery.
Although cosmetic appearance of patients in the
conjunctival autograft group was significantly better,
AmbioDry remains to be a viable alternative for patients
with advanced or bipolar pterygium, scarred conjunctiva,
and glaucoma patients who may need filtering procedure
in the future.
The limitations of this study include the short followup period, the lack of an independent investigator to
evaluate recurrence and other complications, and the lack
of a more objective grading scale for cosmetic appearance.
In summary, dehydrated human-amniotic-membrane
allograft anchored with nylon sutures and attached with
fibrin glue (Beriplast P) is a safe and effective adjunct after
excision of primary pterygium. It is comparable to
conjunctival autograft in preventing early recurrence and
can be considered as a primary grafting method for primary
pterygium excision. However, conjunctival autograft may
be superior cosmetically. Further studies are needed to
evaluate the long-term efficacy of dehydrated amnioticmembrane allograft and to determine recurrence rates.