Key words
foveal-sparing ILM peeling - ILM flap transposition - full-thickness macular hole
Schlüsselwörter
foveaaussparendes ILM-Peeling - ILM-Flap-Transposition - Makulaforamen
Introduction
Full-thickness macular holes, defined as full-thickness defects of the fovea, lead
to central scotoma and deterioration of vision, and can be classified as primary (idiopathic)
macular holes (iMHs), or secondary macular holes. iMHs are hypothesized to be induced
by antero-posterior traction of the vitreous at the fovea [1] and/or vitreoschisis [2], whereas secondary macular holes are caused by conditions such as blunt trauma,
lightning strike, high myopia, macular schisis, macular telangiectasia type 2, exudative
age-related macular degeneration undergoing anti-VEGF treatment, macroaneurysm, and
surgical trauma [3]. Macular holes have been found in 0.5% of the participants of the “Beaver Dam Eye
Study” attending a follow-up examination with optical coherence tomography (OCT) [4], and McCannel et al. have described the incidence of iMHs with 7.8 persons per 100 000
[5].
Foveal-sparing peeling of the internal limiting membrane (ILM), initially proposed
by Ho et al. [6] and Shimada et al. [7] for patients with myopic foveoschisis in 2012, has the potential to minimize surgically
induced trauma on retinal structures. With the assistance of intraoperative OCT, Itoh
et al. found foveal-sparing ILM peeling avoiding alterations of the foveal architecture
during surgery compared to patients that underwent ILM peeling including the foveal
area [8]. Apart from myopic foveoschisis, foveal-sparing ILM peeling has resulted in a high
macular hole closure rate and better postoperative visual acuity among patients with
full-thickness macular holes compared to patients that underwent standard ILM peeling
[9], [10].
Kelly and Wendel were the first to describe vitrectomy as a promising treatment option
for surgical repair of macular holes [11]. Additional ILM peeling with gas or air tamponade has improved outcomes further
[12], [13], [14]. Nevertheless, macular hole closure rates of large macular holes are still poor
with classical ILM peeling techniques and could be increased with inverted ILM flap
techniques [15], [16]. ILM flaps, representing membranes of biological tissue harvested from the retinal
surface, form a scaffold for Müller cells to migrate into the macular hole [17], [18]. Peeling of the ILM is hypothesized to induce activation of Müller cells. ILM tissue,
as used for ILM flaps, was reported to be rich in neurotrophic factors to stimulate
migration of
Müller cells for closure of macular holes and retina regeneration [17], [19]. A recently published case report of foveal-sparing ILM peeling in combination with
transposition of a temporal ILM flap over the macular hole showed feasibility of this
combination of two successful techniques and reported encouraging outcomes [20].
The aim of this study was to examine outcomes in a cohort of patients that underwent
vitrectomy with foveal-sparing ILM peeling and transposition of an ILM flap over the
macular hole with respect to macular hole closure, postsurgical visual acuity, occurrence
of subfoveal atrophic areas, and central hyperreflective changes in the area of macular
hole closure.
Methods
This retrospective study included consecutive patients scheduled for pars plana vitrectomy
with foveal-sparing ILM peeling combined with ILM flap transposition over the macular
hole for macular hole repair between October 2020 and December 2022 at the Department
of Ophthalmology at the Hanusch Hospital in Vienna, Austria. All patients were operated
by a single surgeon. Inclusion criteria for this study were (a) presence of a full-thickness
macular hole before surgery, and (b) use of the surgical technique of vitrectomy with
foveal-sparing ILM peeling combined with ILM flap transposition over the macular hole.
All research and measurements followed the tenets of the Declaration of Helsinki and
were approved by the local ethics committee of the city of Vienna (EK 22 – 164-VK).
Clinical trials registration: NCT 05 897 671.
Surgery was performed with 23 G or 25 G pars plana vitrectomy and combined foveal-sparing
ILM peeling with ILM flap transposition over the macular hole, as reported previously
[20]. For ILM visualization, chromovitrectomy with a trypan blue and brilliant blue G-based
dye (MembraneBlue-Dual, D. O. R. C., Zuidland, The Netherlands), or a trypan blue
and blulife-based dye (Twin, Alchimia S. R.L, Ponte San Nicolo, Italy) was performed.
Decision for dye was according to availability due to COVID 19 pandemic-related problems
with supplies of medical devices. Foveal-sparing ILM peeling, using end-gripping forceps,
was performed at the inferior, nasal, and superior rim of the full-thickness macular
hole, using multiple curvilinear parafoveal peels [21] and leaving residual ILM tissue at the foveal rim. Furthermore, an ILM flap was
prepared from the temporal ILM and positioned in an inverted fashion over the macular
hole ([Fig. 1]). After successful peeling, fluid-air exchange was performed and correct ILM flap
positioning over the macular hole was checked. All patients received gas tamponade
(sulfur hexafluoride 20%, Alchimia S. R.L, Ponte San Nicolo, Italy) and were motivated
for postoperative face-down positioning for 48 hours after surgery.
Fig. 1 Method of foveal-sparing ILM peeling with ILM flap transposition over the macular
hole. In the temporal area, ILM is left for forming an ILM flap (1), while the residual
ILM around the macular hole is peeling with multiple curvilinear ILM peels (2).
In the early phase of use of foveal-sparing ILM peeling with ILM flap transposition
over the macular hole at our department, a broad area (reference length: one diameter
of the optic disc) of residual ILM was left at the foveal rim with the intention to
avoid any kind of surgically induced trauma at the ellipsoid zone (EZ) in this area.
Due to a high risk of failure observed among these patients, the surgical technique
was changed to leaving a narrow area of residual ILM at the foveal rim (reference
length: one-third of the diameter of the optic disc as a maximal reference length
for the distance of the edge of the peeled area to the foveal rim).
OCT imaging of the macula was performed with a stand-alone spectral-domain OCT (SD-OCT)
device (Spectralis OCT, Heidelberg Engineering, Heidelberg, Germany) before surgery,
and during follow-ups after surgery. Decision on successful macular hole closure in
clinical routine resulted from postsurgically performed OCTs, and for study purposes,
these decisions were based on added information on the type of macular hole closure
according to Kang et al. [22]. Evaluation of occurrence of discontinuities of the EZ, subfoveal atrophy and hyperreflective
changes, parameters with potential impact on outcomes according to the time point
of evaluation, was only driven from follow-ups in the time period of 3 months (± 1
month) after surgery, to ensure comparability of data. Distance-corrected visual acuity
(DCVA) was performed with decimal Snellen charts and only data assessed in the time
period of 3 months (± 1 month) after surgery was used for analysis of
outcomes. All patients received nonsteroidal and steroidal anti-inflammatory eye drops
during the first month after surgery.
Combined phacoemulsification with implantation of an intraocular lens and 23 G or
25 G pars plana vitrectomy with foveal-sparing ILM peeling and ILM flap transposition
over the macular hole was performed only in cases with coexisting vision affecting
cataract.
In case both eyes underwent vitrectomy with foveal-sparing ILM peeling combined with
ILM flap transposition over the macular hole, the eye first operated was chosen. Closure
of the macular hole was defined as a type 1 or type 2 closure, according to Kang et
al. [22].
Analysis of presurgical OCT images included measurement of the base and minimal macular
hole diameter by the OCT slide with the largest size of macular hole diameters. Minimal
macular hole diameter was measured in a parallel orientation to the base macular hole
diameter. Postsurgical OCTs were evaluated for (a) the presence of macular hole closure
and type of macular hole closure, (b) occurrence of discontinuities of the EZ, (c)
occurrence of subfoveal atrophy, and (d) occurrence of hyperreflective changes in
the area of macular hole closure.
Statistical analysis was performed in a descriptive fashion for mean values, standard
deviation, median, and interquartile range (IQR). All data were tested for normal
distribution using the Shapiro-Wilk test and the Kolmogorov-Smirnoff test. In case
of a normal distribution, mean and standard deviation were calculated, otherwise,
median, IQR, and range. A t-test was used for normally distributed data, otherwise,
the Mann-Whitney U test was used, and in case of dichotomic data, Fisherʼs exact test
was applied. Regression analysis was performed as a multiple regression analysis with
stepwise elimination of factors with p > 0.05. A p < 0.05 was regarded as indicating
significant differences between groups. Statistical analysis was performed using the
software tool BiAS (Epsilon Verlag, Darmstadt, Germany).
Results
In total, 42 eyes of 42 patients were included in this study. Demographic data and
baseline characteristics of macular holes are listed in [Table 1]. Foveal-sparing ILM peeling with ILM flap transposition over the macular hole was
first used in a case with coexisting large confluent drusen ([Fig. 2]), as there were concerns of accelerating the conversion of drusen to atrophy due
to surgical stress by a complete ILM peeling. Foveal-sparing ILM peeling with a broad
area of residual ILM at the rim of the optic disc and ILM flap transposition over
the macular hole worked well for this patient. Successful macular hole closure and
improvement of his vision from 0.1 before surgery to 0.6 (Snellen) 3 months after
surgery could be observed. Unfortunately, among the following three patients receiving
foveal-sparing ILM peeling with ILM flap transposition with a broad residual ILM area,
two patients failed to undergo postsurgical
macular hole closure and needed re-surgery. This high risk for failure was the reason
for adapting the technique to leaving only a narrow zone of residual ILM at the foveal
rim with one-third of the diameter of the optic disc as a maximal reference length
for the distance of the edge of the peeled area to the foveal rim.
Table 1 Demographic data of all patients included into the study.
|
Demographic data
|
(n = 42)
|
|
Broad/narrow foveal-sparing zone
|
4/32
|
|
Foveal-sparing ILM peeling only partial or not successful
|
6
|
|
Median age
|
69 years (IQR: 63 – 77, range: 18 – 83)
|
|
Female/male
|
27/15
|
|
R/L eye
|
18/24
|
|
Lens status (phakic/pseudophakic/
phacovitrectomy)
|
18/15/9
|
|
Median minimal macular hole diameter
|
298 µm (IQR: 176 – 444, range: 56 – 912)
|
|
Median basal macular hole diameter
|
676 µm (IQR: 525 – 932, range: 200 – 1815)
|
Fig. 2 Foveal-sparing ILM peeling with ILM flap transposition over the macular hole was
first performed in a case with coexisting large confluent drusen, as there were concerns
of accelerating conversion of drusen to atrophy due to surgical stress by a complete
ILM peeling (upper panel: preoperative OCT, lower panel: OCT 3 months after surgery).
Foveal-sparing ILM peeling with ILM flap transposition over the macular hole with
a narrow residual ILM zone could be successfully performed in 32 eyes of 32 patients.
In the remaining 6 patients, foveal-sparing ILM peeling was unintentionally partly
successful or not successful (one patient had ILM peeling until the superior foveal
rim, one patient only foveal-sparing ILM peeling at the inferior foveal rim, a complete
ILM peeling was performed in four patients, and ILM flap transposition was possible
in five patients, while in one patient, the ILM flap was lost during surgery). Only
the subgroup with successful foveal-sparing ILM peeling with ILM flap transposition
over the macular hole with a narrow residual zone of ILM around the macular hole underwent
a detailed analysis of the outcomes.
Macular hole closure could be achieved in all patients after the first surgery, 31
of them showed type 1 closure (97%) and one type 2 closure (3%; [Fig. 3]). Postsurgical central atrophy of the EZ could be detected in one patient (3%),
a short discontinuity of the EZ in four patients (13%), and hyperreflective changes
in the area of macular hole closure in four patients (13%; [Fig. 4]). Cataract progression among phakic patients occurred in eight patients after surgery.
Fig. 3 Types of macular hole closure (upper left panel: preoperative OCT, DVCA = 0.2, lower
left panel: type 1 closure 3 months after surgery, DCVA = 0.5; upper right panel:
preoperative OCT, DCVA = 0.16, lower right panel: type 2 closure 2 months after surgery,
DCVA = 0.25).
Fig. 4 Postsurgical discontinuity of the EZ (lower left panel), and hyperreflective changes
(lower right panel).
Visual acuity in the subgroup of patients with foveal-sparing ILM peeling with ILM
flap transposition over the macular hole with a narrow residual ILM zone at the foveal
rim could be evaluated in 31 patients, while in 1 patient, visual acuity 3 months
after surgery was not available. DCVA after surgery improved in all patients, except
in one patient with postsurgical cataract progression due to the gas tamponade. Median
improvement of DCVA in the subgroup of patients with pseudophakic lens status 3 months
after surgery (n = 19) was 4 lines (IQR: 3 to 6), while phakic patients (n = 12) improved
3.5 lines (IQR: 2 to 5).
Multiple regression analysis (with stepwise elimination of factors with p > 0.05,
and the following predictors: preoperative DCVA, type of macular hole closure, lens
status 3 months after surgery, minimal macular hole diameter, basal macular hole diameter,
postsurgical central atrophy of the EZ, postsurgical EZ discontinuity, and central
hyperreflective changes in the area of macular hole closure) showed preoperative DCVA
to be a significant predictor for DCVA 3 months after surgery (p = 0.025; [Table 2]).
Table 2 Results of a multiple regression analysis examining potential predictors for DCVA
3 months after surgery. The table lists the examined potential predictors and the
p values before elimination of predictors with a p > 0.05. After elimination of predictors
with a p > 0.05 (beginning with the highest values), preoperative DCVA showed to be
a significant predictor for DCVA 3 months after surgery (p = 0.025).
|
Multiple regression analysis
|
(before elimination of predictors with p > 0.05)
|
|
EZ discontinuity
|
p = 0.949
|
|
Basal macular hole diameter
|
p = 0.761
|
|
Minimal macular hole diameter
|
p = 0.601
|
|
Central atrophy of the EZ
|
p = 0.577
|
|
Preoperative DCVA
|
p = 0.501
|
|
Type of macular hole closure
|
p = 0.379
|
|
Lens status
|
p = 0.152
|
|
Central hyperreflective area
|
p = 0.101
|
|
Multiple regression analysis
|
(after stepwise elimination of predictors with p > 0.05)
|
|
Preoperative DCVA
|
p = 0.025
|
Discussion
Foveal-sparing ILM peeling, with a narrow zone of residual ILM around the hole, combined
with ILM flap transposition over the macular hole achieved a high rate of macular
hole closure (type 1 closure: 97%, type 2 closure: 3%) after the first surgery and
an improvement of visual acuity in the majority of patients.
Major concerns using classical ILM peeling for macular hole repair are the potential
negative effects of iatrogenic traction on retinal tissue at the rim of the macular
hole, especially enlargement of the macular hole during surgery with potential damage
of photoreceptors. In contrast to classical ILM peeling, foveal-sparing ILM peeling
was reported to be less traumatic with respect to intrasurgical alterations of the
retinal tissue among patients with myopic foveoschisis [8], a finding that could be hypothesized to be valid for surgical maneuvers during
repair of full-thickness macular holes, too. Two groups of authors reported outcomes
of foveal-sparing ILM peeling for repair of full-thickness macular holes, stating
a high macular hole closure rate and better postoperative visual acuity compared to
patients that underwent standard ILM peeling [9], [10].
The combination of foveal-sparing ILM peeling with ILM flap transposition over the
macular hole is a novel approach, offering a combination of the benefits of foveal-sparing
ILM peeling [9], [10], as mentioned above, with high rates of macular hole closure after ILM flap transposition.
ILM flap transposition for macular hole repair was introduced by Michalewska et al.
[15], [16] and represents a promising surgical method, with ILM flaps forming a scaffold for
Müller cell migration into the macular hole and ILM flaps being rich in constituents
for enhancing Müller cell migration into the macular hole [17], [18], [19]. Furthermore, in case of failure of macular hole closure after the first surgery,
ILM flaps offer potential for high chances of macular hole closure in a second surgery
with repositioning of the ILM flap over the macular hole and endotamponade.
Broad zones of residual ILM around the macular hole are associated with the risk of
failure with respect to macular hole closure. It can be hypothesized that tangential
traction of the residual ILM may have the potential to interfere with macular hole
closure. Therefore, broad zones of residual ILM should be avoided when performing
foveal-sparing ILM peeling. Nevertheless, foveal-sparing ILM peeling with ILM flap
transposition over the macular hole can be safely performed in case of leaving a narrow
zone of ILM around the macular hole.
Foveal-sparing ILM peeling with ILM flap transposition over the macular hole is associated
with a trend of less occurrence of postsurgical central atrophy of the EZ compared
to a historical group of patients that underwent a combination of classical ILM peeling
with ILM flap transposition over the macular hole (historical group: 13%, foveal-sparing
ILM peeling with ILM flap transposition and a narrow zone of residual ILM around the
macular hole: 3%, p = 0.2, Fisherʼs exact test) [23]. This trend maybe indicates less intraoperative surgically induced trauma among
patients that underwent foveal-sparing ILM peeling with ILM flap transposition over
the macular hole and underlines the need for a prospective randomized trial comparing
both surgical methods.
We did not perform a detailed analysis of improvement of visual acuity due to the
fact that it is a routine procedure at our department to perform phacovitrectomy only
in cases of coexistent vision affecting cataract. Nevertheless, gas tamponade, as
applied in all patients examined in this study, has the potential to induce cataract
progression in the postsurgical time period with potential vision-limiting effects.
Limitations of our study are the retrospective nature of the study and the relatively
high number of phakic patients with possible negative effects of cataract progression
on visual acuity data.
Concluding, the combination of foveal-sparing ILM peeling with a narrow residual ILM
zone combined with ILM flap techniques has been shown to be a safe and effective surgical
option for patients with full-thickness macular holes, resulting in a high macular
hole closure rate with a low rate of occurrence of postsurgical central atrophy of
the EZ, and improvement of visual acuity in the majority of patients.
Conclusion Box
Already known:
Foveal-sparing ILM peeling has resulted in a high macular hole closure rate and better
postoperative visual acuity among patients with full-thickness macular holes compared
to patients that underwent standard ILM peeling [9], [10].
Newly described:
The combination of foveal-sparing ILM peeling with a narrow residual ILM zone combined
with ILM flap techniques has been shown to be a safe and effective surgical option
for patients with full-thickness macular holes, resulting in a high macular hole closure
rate with a low rate of occurrence of postsurgical central atrophy of the EZ, and
improvement of visual acuity in the majority of patients.
