·Clinical
Research·
Accelerated
versus standard corneal cross linking in the treatment of ectasia post
refractive surgery and penetrating keratoplasty: a medium term randomized trial
Hany A. Khairy, Moataz F. Elsawy, Khaled Said-Ahmed,
Marwa A. Zaki, Sameh S Mandour
Department of Ophthalmology,
Menoufia University Hospitals, Menoufia 35211, Egypt
Correspondence to: Hany A. Khairy. Department of
Ophthalmology, Menoufia University Hospitals, Menoufia 35211, Egypt.
khairyhany@hotmail.com
Received:
Abstract
AIM: To compare the clinical
outcomes of the standard corneal cross linking (CXL) and the accelerated CXL in
patients with progressive corneal ectasia post refractive surgery and
penetrating keratoplasty.
METHODS: Totally 120
eyes of 83 patients
scheduled to receive either standard CXL (3 mW/cm2 for a period of
30min) or accelerated CXL (18 mW/cm2 for a period of 5min). The main
outcomes for comparison were the change in: maximum-K reading (K-max), manifest
refractive spherical equivalent (SE), central corneal thickness (CCT), and the
best corrected distance visual acuity (CDVA).
RESULTS: One hundred and eleven eyes
completed the study. The main outcome measurement was the K-max reading. Both
group showed significant improvement in the value postoperatively at 6 and
12mo. The mean change in the standard group was 1.21±0.11 D and in the
accelerated group was 0.90±0.05 D at the end of 12mo postoperatively, with no
statistically significant difference between the 2 groups. Similarly, CDVA
improved significantly from their preoperative value in the standard group by
2.98±0.11 letters, and in the accelerated group by 2.20±0.06 letters, with no
statistically significant difference between the two groups. Both of the SE,
and CCT showed no statistically significant difference at the end of follow up
period in each group.
CONCLUSION: Both standard CXL and
accelerated CXL are safe and effective treatment in halting ectasia after
corneal refractive surgery. The accelerated CXL results are comparable to the
standard CXL with short time exposure of the cornea to ultraviolet irradiation,
leading to reduced operation time, reduced operative ocular discomfort, and
corneal haze.
KEYWORDS: corneal cross linking; accelerated;
refractive surgery; penetrating keratoplasty; corneal ectasia
DOI:10.18240/ijo.2019.11.08
Citation:
Khairy HA, Elsawy MF, Said-Ahmed K, Zaki MA, Mandour SS. Accelerated versus
standard corneal cross linking in the treatment of ectasia post refractive
surgery and penetrating keratoplasty: a medium term randomized trial. Int J
Ophthalmol
2019;12(11):1714-1719
INTRODUCTION
Corneal cross-linking (CXL) has
become a widely accepted treatment to halt the progression of corneal ectatic
diseases especially after the United States Food and Drug Administration
approval in 2016[1]. The standard Dresden CXL
protocol (with a continuous irradiation of 3 mW/cm2 for 30min) has
been shown to be effective in halting the progression of ectasia. However, the
procedure is time consuming lasting for almost an hour which results in long
operation time, patient discomfort, and postoperative ocular pain[2-6]. Prolonged exposure to ultraviolet
(UV) rays may increase the risk of complications such as sub epithelial haze,
corneal infiltrate, and infectious keratitis[7-11]. Hence, the use of radiation of higher intensity and
shorter duration (accelerated CXL) was introduced[12].
Variable accelerated CXL protocols were prescribed in the literature. The most
common ones involved UVA irradiation of 9 mW/cm2 for 10min, and 18
mW/cm2 for 5min[13].
Pooled data from several comparative
studies has been discussed in a recent Meta-analysis. It was found that the
reduction in the maximum keratometry value (K-max) as well as in the spherical
equivalent (SE) were significantly greater in standard than in accelerated
protocol. However, comparative outcomes of best corrected distant visual acuity
(BCVA), central corneal thickness (CCT) and endothelial cell density (ECD)
indicated no significant differences between the two procedures[6,14-27].
The aim of this study was to compare
the clinical outcomes of the standard CXL and the accelerated CXL in patients
with progressive corneal ectasia post refractive surgery or penetrating
keratoplasty.
SUBJECTS and methods
Ethical Approval This prospective, randomized study
recruited patients from Menoufia University Hospitals and its satellite clinics
(between January 2016 and June 2017). The study protocol was approved by the
Ethical Committee of Menoufia Medical School (adhering to the tenets of the
Declaration of Helsinki). Trial Registration Number: NCT03791684. The study
protocol was explained to the patients and all patients provided a written
informed consent.
The data used to support the
findings of this study are restricted by the Ophthalmic Ethics Committee at
Menoufia University Hospital, in order to protect patient privacy. Data are
available from Mr Hany A. Khairy (khairyhany@hotmail.com), for researchers who
meet the criteria for access to confidential data.
Inclusion and Exclusion
Criteria Eligible patients were adults aged
above 18y. They had been diagnosed clinically with progressive corneal ectasia
following refractive surgery procedure [laser in situ keratomileusis
(LASIK), photorefractive keratectomy (PRK), ReLex small incision lenticule
extraction (SMILE)], or penetrating keratoplasty with CCT above 400 µm and the
K-max>47.00 D (Table 1).
The diagnosis of recurrent ectasia
was based on clinical findings, such as poor BCVA, pachymetric corneal
thinning, Vogt striae, visible anterior bulge, and topographic irregular
astigmatism.
Evidence of progression included:
increase in keratometry readings (K-max) of greater than 1.00 D, increase in
mean spherical refractive equivalent by 0.50 D, increase in astigmatism by 1.00
D, or a decrease of patient’s visual acuity by 5 or more letters over a period
of 6mo. Enrolled patients had the ability to understand and the willingness to
follow study instructions and were likely able to complete all required visits.
The exclusion criteria involved:
history of herpetic ocular diseases (including herpes simplex virus and
varicella zoster virus) in the study eye; active or recurrent ocular disease in
either eye (e.g., uveitis, chronic moderate to severe blepharitis or
severe dry eye) or sight-threatening diseases (e.g., previous retinal or
optic nerve diseases) that would interfere with the interpretation of the study
data.
Data collected from patients
included age, previous ocular and medical history, medications, and family
history of ocular diseases. Patients had baseline assessment at their
preoperative visit, including: BCVA in letters, refraction with
auto-refractometer, intraocular pressure (IOP) measurement with applanation
tonometry, dilated fundus examination with +78 D volk lens, and corneal
tomography by Pentacam® (Oculus, Germany).
The randomization process used four
opaque envelopes in two containers. One contained for standard CXL, and for accelerated
CXL, and the other contained the name of two patients listed for CXL on that
day. The two patients were randomized to one of the procedures by asking an
independent person to choose one envelope from each container. For patients who
were listed for both eyes, envelopes had right and left instead.
Surgical Procedure Patients were draped, after topical
anesthesia (Benoxinate Hydrochloride 0.4% eye drops) application. After the
debridement of the corneal surface (
Postoperatively, Ofloxacin 0.3%
drops were instilled as a prophylaxis and an eye bandage contact lens was left
for 72h. Patients were asked to use dexamethasone 0.1% eye drops, and ofloxacin
0.3% eye drops four times daily for one week postoperatively and then reduce it
to twice a day for another week.
Outcomes The main outcomes for comparison
were the change in: K-max, manifest refractive SE, CCT, and the BCVA. Patients
were followed up clinically at 1d, 1wk, 1, 6, and 12mo postoperatively but the
topographic measurements were done only at the 6 and 12mo visits.
Statistical analysis was conducted
using the SPSS (Version 15, 2006) for Windows statistical package. Data were
presented as mean±SD. Results were analyzed using Mann-Whitney-Wilcoxon
test. P<0.05 was considered statistically significant.
Results
One hundred and twenty eyes of 83
patients were included in the study; 42 males and 36 females. Five patients (9
eyes) did not continue the follow up period and were excluded from the study.
One hundred and eleven eyes were enrolled; 54 eyes had the standard CXL, and 57
had the accelerated CXL. Most eyes with post refractive surgery ectasia had
LASIK, and the most common indication for penetrating keratoplasty was
keratoconus (Table 1). Thirty three patients had both eyes treated; all of them
had ectasia post refractive surgery. Twenty one patients had one eye treated
for post refractive surgery ectasia, and 24 patients (24 eyes) had treatment
for post penetrating keratoplasty ectasia. There was no statistically
significant difference between the 2 groups in the K-max measurement, BCVA,
manifest SE refraction, or CCT (Table 2).
Table 1 The distribution of surgery
type in each group
|
Patients distribution |
Type of refractive surgery |
Indication for penetrating keratoplasty |
||||
|
LASIK |
Femto-LASIK |
SMILE |
PRK |
Keratokonus |
Other |
|
|
Accelerated CXL |
36 |
5 |
2 |
2 |
9 |
3 |
|
Standard CXL |
34 |
5 |
1 |
2 |
10 |
2 |
CXL: Corneal cross linking; LASIK:
Laser in situ keratomileusis; SMILE: Small incision lenticule
extraction; PRK: Photorefractive keratectomy.
Table 2 Comparison of the outcome
parameters in each group and between the two groups
mean±SD
|
Parameters |
The standard CXL group |
The accelerated CXL group |
1P |
2P |
||||
|
Preoperatively |
6mo postoperatively |
12mo postoperatively |
Preoperatively |
6mo postoperatively |
12mo postoperatively |
|||
|
K-max (D) |
52.19±1.88 |
51.10± |
50.98± |
52.17±1.81 |
51.41± |
51.27± |
0.36 |
0.38 |
|
BCVA |
37.56±5.08 |
39.78± |
40.54± |
37.58±5.11 |
39.47± |
39.78± |
0.74 |
0.37 |
|
SE (D) |
-3.55±1.29 |
-3.18± |
-3.12± |
-3.55±1.32 |
-3.14± |
-3.15± |
0.97 |
0.97 |
|
CCT (µm) |
439.11±33.8 |
434.28± |
432.57± |
439.67±36.5 |
433.14± |
433± |
0.87 |
0.92 |
CXL: Corneal cross linking; K-max:
Maximum keratometry value; BCVA: Best corrected distant visual acuity; SE:
Spherical equivalent; CCT: Central corneal thickness. aP=0.001 vs
preoperatively; 1P: Comparing the parameters 6mo
postoperatively between the two group; 2P: Comparing the
parameters 12mo postoperatively between the two group.
The main outcome measurement was the
K-max reading. Both group showed significant improvement in its value
postoperatively at 6 and 12mo. The mean change in the standard group was
1.21±0.11 D and in the accelerated group was 0.90±0.05 D at the end of 12
months postoperatively. There was no statistically significant difference between
the 2 groups at all points of follow up (Table 2). There was no deterioration
of BCVA in either group. Both groups showed significant improvement in the BCVA
from their preoperative value. In the standard group they improved by 2.98±0.11
letters at 12mo postoperatively. In the accelerated group they improved by at
2.20±0.06 letters at 12mo postoperatively. There was no statistically
significant difference between the 2 groups at all points of follow up (Table
2). The mean SE of manifest refraction showed no statistical difference at the
end of follow up period in each group, with no statistical significant
difference between the 2 groups at all points of follow up. The CCT showed no
statistical difference at the end of follow up period in each group. There was
no statistically significant difference between the 2 groups at all points of
follow up.
Most of patients K-max improved,
80.5% (46/57) in the accelerated group, and 75.8% (41/54) in the standard
group), by up to 2.00 D. There was a recorded progression of ectasia following
treatment in 10.5% (6/57) of patients in the accelerated group, and in 9.2%
(5/54) of patients in the standard group. However, the progression never
exceeded 1.0 D in either group (Figure 1).
Figure 1 Distribution
of the changes of the maximum keratometry (K-max) value between the two groups.
The success rate was similar in
postrefractive surgery and post penetrating keratoplasty cases with progression
of ectasia of around 10% in each group. There were no recorded serious
complications as corneal infiltrate, stromal keratitis or corneal
decompensation, postoperatively. Corneal haze (mild) was reported in 5 eyes in
the standard group and 2 eyes in the accelerated group. All cases of corneal
haze resolved at the 6mo visit apart from one case post refractive ectasia
which resolved at the end of follow up period.
Discussion
The risk factors and the course of
development of corneal ectasia after corneal refractive procedure are still not
well understood. However, similar to keratoconus, successful management of ectasia after
refractive surgery particularly in cases with steep cornea includes early
detection before the symptoms become more pronounced, careful monitoring, and
prompt CXL treatment in progressive cases[28].
The efficacy of standard CXL in the
prevention of keratoconus progression has been established now according to
different clinical trials. However, studies about CXL for post refractive
surgery ectasia were rare until Hersh et al[27]
published his US phase 3 multicenter trial. The results of the trial showed
statistically significant difference in the K-max value between the treatment
group (decreased by 0.7 D) and the sham group (progression of 0.6 D) after 12mo
of follow up. They concluded that progressive ectasia has stabilized with some
degree of improvement in CXL treated patients, while the control group continue
to progress over the follow up period.
Wan et al[29]
in a recent Meta-analysis reviewed the safety and stability of CXL for the
treatment of post LASIK ectasia. In the 7 studies analyzed, the topographic
parameters of the treated patients did not improve as reported in studies
involving keratoconic eyes but it was successful in halting its progression.
This indicated that post refractive surgery corneas treatment with CXL was less
effective compared with keratoconic corneas, possibly due to of the post
refractive surgery corneas flap formation and tissue ablation[2,30-35].
The standard Dresden CXL protocol
used irradiation of 3 mW/cm2 for 30min. However, the procedure is
time consuming, with long exposure of the cornea to ultraviolet irradiation,
leading to postoperative ocular pain, and corneal haze. Several modifications
have been introduced to reduce these drawbacks, including: trans-epithelial
CXL, using radiation of high intensity and shorter duration (accelerated CXL)[2], the use of accelerated epithelium on CXL, and the use
of pulsed-light accelerated CXL[2].
There is no uniform protocol for the
Accelerated CXL, and the evidences of its efficacy are not well understood.
Accelerated CXL in the treatment of keratokonus has been compared to standard
CXL in different clinical trials. In a recent Meta-analysis by Liu et al[6], they reported that the K-max reduction (pooled mean
difference was 0.49) was significantly greater in the standard group. In
addition, spherical equivalent decreased significantly (mean difference was
0.62) for the standard when compared to the accelerated group. However,
corrected distant visual acuity, CCT, ECD indicated no significant difference
between the two CXL regimes.
K-max remained the most important
factor to evaluate the efficacy of treatment in post refractive surgery
ectasia. In this study, the mean change in the standard group was 1.21±0.11 D
and in the accelerated group was 0.90±0.05 D at the end of 12mo
postoperatively. There was no statistically significant difference between the
2 groups at all points of follow up. Hersh et al[27]
in their randomized controlled trial reported significant decrease in the mean
K-max value (0.7±2.1 D) at the end of 12mo postoperatively. In 76 treated eyes,
the K-max decreased by 2.00 D or more in 14 eyes (18%), and remained within 2 D
in 59 eyes (78%). Despite of the initial increase of mean K-max by 1.0 D at
month 1, it showed progressive decrease from 1.1 D between months 1 and 3, to
0.2 D between months 6 and 12.
Richoz et al[33] reported the long term effect of standard CXL in
corneal ectasia after LASIK (23 eyes) and PRK (3 eyes). At the end of 2y follow
up, the mean K-max after CXL (50.9±4.9 D) was significantly lower than the
pretreatment measures (52.8±5.0 D). Other studies have reported similar
outcomes but most of them were non-randomized with a small number of patients
(21-35 patients)[30-35].
Marino et al[34] studied the effect of accelerated CXL on 40 eyes of
24 patients with post LASIK ectasia. Patients were treated with irradiation of
9 mW/cm2. At the end of the 24mo follow up period, none of the
studied eyes showed signs of progression, with more than 70% of patients with
stable or better vision. Despite the long term follow up the stud was not
controlled or randomized. The topographic results of the study by Choi et al[35] showed that despite the use of high irradiance
exposure (30 mW/cm2 for 3min and 40s), the effect of corneal
flattening was less in the transepithelial accelerated CXL compared to the
standard CXL. A previous study that compared 4 different protocols, including
conventional (3 mW/cm2 for 30min) and modified (9 mW/cm2
for 10min, 18 mW/cm2 for 5min, and 30 mW/cm2 for 3min)
methods, showed that the flattening effect of the steep and flat keratometry
was reduced with higher irradiation and less time[25].
Aixnjueluo et al[36] in his study on
Japanese patients with progressive keratoconus, reported statistically
significant improvements in K-max and BCVA at the end one year post
trans-epithelial accelerated CXL.
There is an agreement between
investigators that the CXL treatment in post LASIK ectasia is successful in
improving or at least stabilizing visual acuity. The improvement of the SE was
statistically significant (decrease by 0.5 D) in the US multicenter trial[27] CXL treated group. In the crosslinking treatment
group, there was a significant improvement of 5.0 letters of visual acuity at
the end of follow up. In the control group, there was a loss of 0.3 letters.
Hafezi et al[7] reported that CDVA
increased by one more lines in 90% of the treated patients. Morino et al[34] in his study on accelerated CXL, reported that BCVA,
uncorrected distance visual acuity was stable or improved in all patients after
2y, and Richoz et al[33] in their study on
26 eyes who has the standard CXL for post LASIK and PRK ectasia, reported that
BCVA improved to a mean of 0.3 logMAR unit in 19 cases and remained stable in 7
patients.
The most common complications
recorded by this procedure are ocular pain caused by epithelium stripping and
long exposure to UV radiation, sub-epithelial haze, sterile infiltration,
corneal decompensation and infectious keratitis. In the US trial[27] out of the 76 treated eyes, 5 had persistent corneal
haze and one had corneal scar at the end of the 12mo period. There were no
recorded serious complications in our study as corneal infiltrate, stromal
keratitis or corneal decompensation, postoperatively. However, corneal haze was
reported in 5 eyes in the standard group and 2 eyes in the accelerated group.
All cases of corneal haze resolved at the end of follow up period.
To the best of our knowledge, there
is no published studies comparing the standard and the accelerated technique in
post refractive surgery or penetrating keratoplasty patient diagnosed with
corneal ectasia. One advantage of this study was the good sample size despite
the strict inclusion criteria. In contrast to the US trial by Hersh et al[27] who set the corneal thickness at 300 µm, we included
only patients with CCT above 400 µm, as the local ethical guidelines do not
allow the use of CXL in thickness less than 400 µm. In the US trial, they used
hypotonic riboflavin (0.1% riboflavin, no dextran), 1 drop every 10s for 2min
sessions, in cases with pachymetry less than 400 µm. It is not clear whether
this modification had an effect on the results of treatment or not. Although
they have not reported any case of corneal decompensation, there were no
evidences of its effect on the ECD. Limitation of this study was the lack of
studying the endothelial cell count by specular microscopy or the demarcation
line by OCT. However, it is well established that the use of CXL is very
unlikely to cause damage in corneas thicker than 400 µm.
The interpretation of this study
results showed that both standard CXL and accelerated CXL are effective
treatment in halting ectasia after corneal refractive surgery. The accelerated
CXL results are comparable to the standard CXL with short time exposure of the
cornea to UV irradiation and reduced operation time.
ACKNOWLEDGEMENTS
The paper is accepted for oral
presentation at the ESCRS Meeting (Vienna 2018).
Conflicts of Interest: Khairy HA, None; Elsawy MF, None;
Said-Ahmed K, None; Zaki MA, None; Mandour SS, None.
REFERENCES