Key words
macular pucker - optical coherence tomography - prognosis - surgery - epiretinal membrane
- OCT
Introduction
Epiretinal membranes (ERMs) are a common finding in patients of increasing age [1]. The advent of optical coherence tomography (OCT) has led to the more frequent detection
of ERMs with close to one-third of affected patients found to have ERMs in both eyes.
By the age of 74, ERMs occur in over 50% of cases [2].
In routine clinical practice, it is not only the morphological appearance of the ERMs
but also the patientʼs visual acuity and subjective suffering that are of critical
importance. Patients described having blurred vision, metamorphopsia, and aniseikonia.
In addition, in severe cases, there can be a loss of visual acuity due to leakage
from the retinal vessels. In such cases, there can sometimes be a discrepancy between
the OCT findings and the patientʼs vision [3]. Newer, OCT-based, classifications not only consider the ERMs, but also neuroretinal
changes. All of these changes play a critical role in terms of the postoperative visual
acuity outcome [4].
ERM treatment consists of pars plana vitrectomy (ppV) combined with peeling of the
ERMs and the internal limiting membrane (ILM).
This study aims to summarize new findings on the clinical picture, classification,
and treatment of ERMs based on a review of the literature.
Method
To compile this review, we performed a literature search of the PubMed metadatabase
(https://pubmed.ncbi.nlm.nih.gov). We limited our search to publications from 2000 to 2022 and used the following
search terms: “epiretinal membrane”, “macular pucker”, “vitreoretinal traction”, “preretinal
fibrosis”, “gliosis”, and “epiretinal gliosis”. These search terms were also used
in combination with the terms “OCT”, “optical coherence tomography”, “postoperative”,
and “outcome”. As the main focus of this article is on idiopathic ERMs, publications
on secondary ERMs were excluded from the literature search.
Epidemiology, Pathophysiology, and Classifications
Epidemiology, Pathophysiology, and Classifications
Idiopathic ERM is one of the most common retinal diseases. Depending on the literature
consulted, it affects approximately 2 to 34% of all patients. The incidence of ERM
increases with increasing patient age [5], [6]. The Beaver Dam Eye Study showed that an ERM was present in 34% of patients in the
63- to 102-year age group [2]. Several other studies also provide supportive evidence of an increasing prevalence
of idiopathic ERMs with increasing patient age. According to Quinn et al., the odds
ratio per year is 1.97 (confidence interval 1.34 – 2.98) for patients aged 60 – 69,
and 4.61 (confidence interval 3.08 – 6.90) for patients over 70 years of age [7].
There is little proof that either the patientʼs ethnicity [8], [9] or sex is of relevance. It is possible that women may be affected more often than
men [7]; however, the available data is inconsistent on this point.
Idiopathic ERM refers to the form of ERM that occurs only after vitreous body detachment,
and in the absence of other risk factors or previous surgery. With vitreous body detachment,
a distinction is made between complete detachment without demonstrable attachment
or traction on OCT, and incomplete detachment. Incomplete vitreous body detachment
is characterized by residual adhesion of the vitreous body to the retina [10].
As the incidence of posterior vitreous body detachment increases with increasing patient
age, this may explain why the occurrence of idiopathic ERMs also increases with patient
age [6].
The formation of idiopathic ERMs is attributed to pathological posterior vitreous
body detachment with subsequent proliferation of various cells [11], [12]. Müller glia, hyalocytes, and the retinal pigment epithelium (RPE), as well as the
cytokines and growth factors they secrete, are involved in the formation and proliferation
of myofibroblasts [1], [10]. ERM growth is promoted by extracellular proteins such as type I collagen. The tractive
components of the ERM are attributed to α-smooth muscle actin (SMA). Other factors such as vimentin and anti-CD45 are also
involved [12].
Evidence shows that surgical procedures can play a critical role in the formation
of preretinal membranes. For example, a secondary ERM following a cataract operation
has been demonstrated in 16% of cases. These retinal changes following cataract operations
appear to be caused by the surgically induced effects on the vitreous body [5], [6].
Retinal diseases such as retinal detachment, retinal foramina, retinal vein occlusion,
traumatic injury, or diabetic retinopathy can provoke the formation of a secondary
ERM [13]. Diabetes mellitus with accompanying hyperglycemia also appears to have an effect
on the vitreous body, which in turn favors the formation of an ERM [8], [14].
ERMs can be viewed clinically using fundoscopy, or on imaging using OCT, en-face OCT,
fundus photography, and autofluorescence. Fibrocellular membranes manifest clinically
through a cellophane-type retinal reflex and the formation of retinal folds, which
can change the anatomy of the retina [5]. On OCT, the ERM typically appears as a hyperreflective line above the retina. A
distinction can be made here between global attachment or focal partial attachment
of the membrane to the retina [15]. Advanced stages are characterized by fibrotic remodeling of the ERM. This preretinal
fibrosis may lead to macular edema, metamorphopsia, or loss of visual acuity due to
strain and deformation of the retinal layers caused by traction [9], [11], [16]. This deformation of the retina can be visualized using autofluorescence, a technique
which is particularly
suitable for visualizing deformation of retinal vessels. Fundus photography can also
be helpful in detecting deformations [11]. En-face OCT enables direct visualization of the changed morphology, as well as
the extent of the ERM [17]. On the microstructural level, ERM impairs neuronal transduction performed by Müller
cells inside the retina [16], [18].
A distinction should be made between classic idiopathic ERMs and epiretinal proliferations.
In contrast to an ERM where a hyperreflective line above the macula is visible on
OCT, with epiretinal proliferations, an isoreflective line is present above the macula
and above the ILM. In terms of histopathology, studies have shown that unlike ERMs,
epiretinal proliferations have no or only very few demonstrable tractive characteristics
[19].
Classifications
The clinical ERM grading proposed by Gass is the current established nomenclature.
The grades are classified as follows. Grade 0: cellophane maculopathy, stage 1: cellophane
maculopathy with puckering, stage 2: preretinal macular fibrosis [11]. The Xiao et al. review describes an overall prevalence of idiopathic ERMs of 9.2%
[95% confidence interval (CI): 4.7 – 13.8%]. Cellophane maculopathy has a prevalence
of 7.2% (95% CI: 3.3 – 10.8%), and preretinal macular fibrosis a prevalence of 2.0%
(95% CI: 1.3 – 2.8%) [9].
In addition to the grading proposed by Gass, idiopathic ERMs may also be distinguished
based on their appearance on OCT. A newer classification proposed by Govetto et al.,
based on OCT imaging, divides ERM into 4 stages (see [Fig. 1]):
Fig. 1 Images for Govetto staging of the epiretinal glia. a Epiretinal gliosis stage 1: presence of a foveal pit, retinal layers well defined.
b Epiretinal gliosis stage 2: cotton wool spot as a secondary finding marked with a
star; absence of a foveal pit, retinal layers well defined. c Epiretinal gliosis stage 3: absence of a foveal pit, retinal layers well defined,
presence of ectopic inner foveal layers (EIFLs). d Epiretinal gliosis stage 4: intraretinal cavity or macular edema marked with an arrow;
absence of a foveal pit, retinal layers disrupted, presence of EIFLs.
-
Stage 1: Presence of a foveal pit with well-defined retinal layers.
-
Stage 2: Absence of a foveal pit with well-defined retinal layers.
-
Stage 3: Absence of a foveal pit with well-defined retinal layers and presence of
ectopic inner foveal layers (EIFLs).
-
Stage 4: Absence of a foveal pit, disrupted retinal layers, presence of EIFLs [20].
Surgical Indications
Determining when to operate on an ERM does not depend solely on the OCT findings.
Primary consideration should be given to the patientʼs subjective level of suffering.
In some cases, an ERM may be present for a prolonged period of time without causing
any significant symptoms in the patient or morphological changes on OCT [21]. For this reason, the decision on when to operate should always be made in close
consultation with the patient, considering their subjective suffering; not all patients
need to be operated on immediately after diagnosis. In early-stage ERM with no subjective
impairment or loss of visual acuity, it is possible to monitor the course of the disease
before performing surgery [3].
If the ERM is accompanied by a loss of visual acuity, this is usually due to a combination
of tractions in the region of the inner and outer retina. An ERM may also have a negative
impact on the patientʼs visual acuity due to changed light refraction and the occurrence
of intraretinal edema [9]. According to data from a British case series, the average preoperative loss of
visual acuity was between 0.2 and 0.32; this improved, on average, to 0.5 after surgery
[22].
In patients with poor initial visual acuity, a considerable improvement in postoperative
visual acuity has been described. In patients with initially moderate to good visual
acuity, a form of “ceiling effect” is presumed to prevent further improvement of visual
acuity following surgery [3].
Considering the question of when to operate, it has been shown that an early operation
can lead to a better postoperative visual acuity outcome. Yusuf et al. were able to
demonstrate a bigger improvement in postoperative visual acuity following surgery
at an early stage compared to a “wait and see” approach [23].
Besides loss of visual acuity, patients often complain of aniseikonia or metamorphopsia;
if the OCT findings are consistent, this may be considered an indication for surgery.
OCT-supported analysis has shown that macropsia in the form of aniseikonia occurs
in the presence of changes to the photoreceptor distribution, while metamorphopsia
is attributed to changes in the inner retinal layers. Interestingly, the OCT findings
often do not reflect the patientʼs subjectively perceived loss of visual acuity.
While metamorphopsia can often be alleviated through surgery, this tends not to be
the case for aniseikonia [16]. A prospective study by Bouwens et al. investigating postoperative metamorphopsia
showed that out of 63 patients included in the study, 82% experienced an improvement
in metamorphopsia and 48% an improvement in visual acuity. Remarkably, according to
the study data, improvement in visual acuity did not correlate to the reported severity
of the preoperative metamorphopsia [24]. Moreover, analyses of the British population have shown that surgery was only considered
to be indicated due to tractive effects on the central retina in approximately 10%
of cases [14].
Anatomical Characteristics and Factors That May Be Predictive of Postoperative Outcome
Anatomical Characteristics and Factors That May Be Predictive of Postoperative Outcome
Vitreomacular interface
In a retrospective case series with a follow-up observation period of 24 months, Byon
et al. reported on 62 eyes with a visual acuity of ≥ 0.5. They subdivided ERM cases
into those with global attachment (GA) and those with partial attachment (PA), as
well as those with and without vitreomacular traction (VMT) ([Fig. 2 a– c] ). Overall, changes in ERM configuration were observed in 24 eyes (39%); these included
some changes that regressed spontaneously. During the 24-month period, 11 out of 33
eyes (33%) showed a progression from the GA type to the PA type. Four ERMs (6%) regressed
spontaneously from the PA type (n = 3) and the VMT type (n = 1); in these cases, an
improvement in the patientʼs vision was observed. Out of the 62 eyes, 4 eyes with
an intact ellipsoid zone (EZ) and PA-type configuration developed a weakened or disrupted
zone with simultaneous loss of visual acuity of more than 2 lines [15].
Fig. 2 Representative examples of ERM-associated OCT findings. a Vitreomacular traction: marked with an arrowhead, intraretinal cavities: marked with
stars, neurosensory detachment: marked with an arrow. b Partial adhesion: marked with arrows, cotton wool spot: marked with a star. c Global adhesion; cotton wool spot: marked with a star. d Disruption to the integrity of the interdigitation zone (IZ) and ellipsoid zone (EZ),
discontinuity of IZ and EZ marked with an arrow.
In a retrospective case control study of ERM patients with a visual acuity ≥ 0.5,
the clinical picture progressed during the 31-month follow-up observation period in
15 out of 112 patients (13%). These patients showed a loss of visual acuity ≥ 2 lines.
In patients who experienced progression, a change in ERM configuration from GA type
to PA type occurred more frequently than in the control group (with no loss of visual
acuity) [25]. These results imply that ERM probably begins as the GA type, then progresses to
the more unstable PA type.
The vitreoretinal adhesion appears to influence the progression of ERMs. In a study
by Byon et al., it was shown that progression and loss of visual acuity occurred in
4 out of 10 eyes (40%) that had vitreoretinal adhesion on initial presentation, compared
with only 2 out of 52 eyes that had posterior vitreous detachment (PVD) (3.8%). The
authors postulated that this may have been due to an increase in proinflammatory factors
in the eyes that had vitreomacular or vitreopapillary adhesions [15].
Central foveal thickness (CFT) and related measurements
Central foveal thickness (CFT) has been investigated in numerous studies, leading
to differing results. A systematic review of 10 studies found no correlation between
preoperative CFT and postoperative visual acuity [18]. Likewise, a multiple regression analysis by Kim et al. did not find any correlation
in their retrospective case series [26]. In contrast, a systematic review from 2017 demonstrated a correlation between an
increased initial CFT and poorer postoperative visual acuity [27].
Overall, in most of the studies analyzed, there was no statistically significant correlation
between best-corrected visual acuity (BCVA) at 6 months [21], [28], [29], 12 months [30], [31], [32], [33], or 24 months [34], [35] after surgery.
Foveal contour and morphology
There are scarcely any indications of a correlation between preoperative foveal contour
and postoperative visual acuity. Scheerlinck et al. [18] were not able to determine any such correlation in their systematic review. Furthermore,
Ozdek et al. were unable to determine a correlation at 24 months after surgery. In
a large-scale, retrospective, multicentric study, they analyzed 634 cases and found
a statistically insignificant trend towards better visual acuity in patients with
a “foveal herniation” compared to patients with diffuse or pseudohole foveal morphologies
[35].
Zeyer et al. discovered in a retrospective study that eyes with a preoperative convex
foveal contour showed an average improvement in visual acuity of 2.4 lines at 12 months
after surgery compared to 0.6 lines in eyes with a flat or concave contour [36].
In a retrospective study, Kinoshita et al. subdivided the macular morphology into
the categories of diffuse edema, cystoid macular edema, pseudo-lamellar hole, and
vitreomacular traction ([Fig. 1 d]; [Fig. 2 a]). With the exception of the pseudo-lamellar holes, each of these groups showed a
significant postoperative improvement in visual acuity. However, the length of observation
periods differed (20.5 ± 14.6 months), and the study does not appear to have made
a distinction between lamellar holes and pseudoholes [13]. In their prospective study, Inoue et al. did not observe any correlation between
preoperative pseudoholes and postoperative visual acuity at 12 months after surgery
[32]. Another retrospective study did not show any correlation between BCVA and preoperative
intraretinal cystoid fluid accumulation. The absence of such a correlation may be
due to
the fact that cystoid fluid accumulation often does not completely regress after surgery
[37].
Govetto Staging and Ectopic Inner Foveal Layers (EIFLs)
An OCT-based classification of ERM and macular morphology, proposed recently by Govetto
et al., is being increasingly used in the literature ([Fig. 1]). Govetto stages 1 to 4 are associated with a poorer initial BCVA as the stages
become more advanced. Govetto et al. also described EIFLs and presented the hypothesis
that ERM-induced centripetal traction either displaces the inner retinal layers towards
the fovea or induces proliferation due to Müller cell damage and the secondary stimulation
of repair pathways. Interestingly, OCT angiography confirms that Govetto stage 2 and
3 ERMs show almost a complete loss of the foveal avascular zone due to the vascularity
of the EIFL [20].
In a later retrospective study, Govetto et al. [4] determined that EIFLs persisted after surgery in 91% of cases. According to a multivariate
analysis, EIFL thickness, defined as the area between the outer nuclear layer and
the ILM (ONL–ILM), was associated with poorer preoperative visual acuity regardless
of the CFT. However, the observed postoperative thinning did not correlate to an improvement
in visual acuity. In the end, it was hypothesized that the presence of EIFLs is associated
with a significantly poorer prognosis.
Moreover, it has been shown that Govetto stage 4 ERMs have poorer visual acuity outcomes
than stage 3 ERMs, even though EIFLs are present in both cases [4]. This implies that the disruption of other retinal layers has additional prognostic
significance. Gonzalez-Saldivar et al. [38] confirmed the negative prognostic significance of EIFLs in Govetto stage 3 to 4
ERMs. In a retrospective analysis, they reported that the postoperative visual acuity
after 12 months was better than the initial visual acuity in all stages. However,
the improvement was only statistically significant in stage 2 ERMs. Moreover, the
final visual acuity in stage 4 eyes was significantly poorer than in stage 3 eyes.
Other Inner Retinal Layers
There is little consensus in the literature regarding the use of other internal retinal
layers as prognostic markers. The multicentric, retrospective, DREAM study established
that a severe disorganization of retinal inner layers (DRIL) was associated with a
significantly smaller improvement in visual acuity after 12 months compared to the
absence of DRIL or only slight DRIL [39]. However, a retrospective study by Fernandes et al. [33] did not find any correlation between the severity of the initial DRIL and the BCVA
after 12 months, although the severity of the DRIL decreased in more than 50% of patients
over the same time period.
In their systematic review, Miguel and Legris discovered that a thinner ganglion cell–inner
plexiform layer (GC–IPL) at the start of the study was associated with a greater improvement
in postoperative BCVA [27]. In their prospective study, Kim et al. discovered that among all of the parafoveal
layers, only the thickness of the GC–IPL and the inner nuclear layer (INL) were negatively
associated with a postoperative improvement in visual acuity; only the INL association
persisted in the multivariate analysis [40]. In another prospective study, Zou et al. investigated the relationship between
initial BCVA and the thickness of seven retinal layers in the foveal, parafoveal,
and perifoveal regions. Multiple linear regression analysis showed that the INL in
all regions was associated with the visual acuity at the start of the study. In patients
whose visual acuity had improved by more than 2 lines 6 months after the
operation, there was a statistically significant correlation between visual acuity
and the INL. In contrast, no correlation could be found for any of the other retinal
layers [41].
Integrity of the Ellipsoid Zone (EZ)
On OCT, the EZ appears as a hyperreflective band in the outer retina. It represents
the ellipsoid part of the inner photoreceptor segment [30].
In the literature, data from numerous retrospective and prospective studies indicates
that disruption or absence of the EZ at the start of the study may be predictive of
a poorer postoperative visual acuity at 6 to 12 months [26], [32] ([Fig. 2 d]).
It is worth noting that in 18 – 36% of patients, the EZ defect was found to have resolved
after 6 months and 12 months. Visual recovery was significantly better in this patient
subgroup. The authors also showed that no further improvement could be expected after
a follow-up observation period of 12 months [35].
Integrity of the Interdigitation Zone (IZ)
On OCT, the interdigitation zone (IZ) appears as another hyperreflective line between
the EZ and the RPE [30]. It represents the part of the outer segments of the cone that is engulfed by apical
RPE cell processes ([Fig. 2 d)]. In many studies, this is referred to using the older term: cone outer segment tips
line. Analogous to the EZ, disruption of the IZ is considered to be an indicator for
photoreceptor damage; studies describe this in both qualitative and quantitative terms.
In several retrospective studies, there appears to be a correlation between the presence
and increasing length of preoperative IZ disruption and poorer postoperative visual
acuity at 6 months [29], 12 months [30], and 24 months [34]. It should be noted, however, that Fernandes et al. [33] found a correlation for both IZ and EZ in the univariate analysis, but only an IZ
correlation in the multivariate analysis. This indicates that patients with an intact
EZ may still have a poor outcome if they have a disrupted IZ. Itoh et al. [30] also found a correlation between postoperative visual acuity and IZ disruption,
but not EZ disruption. Shimozono et al. hypothesized that the IZ is more susceptible
than the EZ to damage induced by ERM traction [29]. There are several indications that IZ defects can resolve
postoperatively in a minority of patients [34].
Other Outer Retinal Layers and Length of Photoreceptor Outer Segment (PROS)
Other potential OCT markers in the outer retina may be located in the outer foveal
and parafoveal layers [photoreceptor outer segment (PROS), ONL] [18], [33]. Of these layers, a correlation with postoperative visual acuity has only been demonstrated
for the length of the PROS.
Shiono et al. [28] showed that it can be difficult to distinguish between EZ and IZ defects if there
are artefacts caused by intraretinal fluid and cataracts. For this reason, they proposed
to determine the PROS length as a quantitative evaluation of the photoreceptor layer
[28]. In a prospective study, the multiple regression analysis found a positive correlation
between PROS length and postoperative BCVA after 6 months. However, a similar correlation
could not be found for thickness of the outer fovea (from the external limiting membrane
to the RPE) or for thickness of the ONL [28]. Using multiple regression analysis, Kinoshita et al. [42] determined that there was a correlation between better visual acuity after 24 months
and a longer baseline PROS length. They also analyzed ONL thickness but did not find
any correlations for this parameter. In another
retrospective analysis, Hashimoto et al. [43] discovered that postoperative recovery of visual acuity had a positive correlation
with recovery of the PROS length.
Central Bouquet Anomaly (CBA)
The central bouquet is a subfoveal area around 100 µm in size consisting of a dense
accumulation of cones and Müller cells ([Fig. 1 b]; [Fig. 2 b], [c]). It is affected by tractive changes caused by ERM. Govetto et al. undertook a grading
of central bouquet anomalies (CBAs) and showed that the initial BCVA decreased with
increasing CBA grade. Moreover, they showed that the CBA tended to be associated with
a Govetto stage 2 ERM, and that it was negatively correlated to the presence of an
EIFL [44]. This has been confirmed by Ortoli et al. [21]. It has been hypothesized that the EIFL protects against CBA by reducing the traction
of the outer fovea [44].
CBA has been less well studied than IZ/EZ, but the limited results that are currently
available do not seem to indicate that it has a significant role for prognosis. This
is not surprising, considering the negative correlation with the EIFL. The retrospective
Brinkmann et al. [45] study established that 68% of eyes with CBA show postoperative improvement within
their classification subgroup.
Ortoli et al. described a postoperative resolution of CBA in 97.7% of eyes. However,
no correlation could be found between the presence or grade of preoperative CBA and
postoperative visual acuity after 6 months [21]. Other studies have shown a postoperative resolution of grade 1 CBA [4], [31].
Importance of Internal Limiting Membrane Peeling
Importance of Internal Limiting Membrane Peeling
The significance of the ILM in ERM pathogenesis remains to be fully elucidated. It
is suspected that the ILM, as a structure adjacent to the vitreous body, offers the
ERM cells a kind of guide structure that defines their final position and morphology
[23].
Based on this assumption, surgically removing the ILM may have a prophylactic benefit
as regards to ERM formation. This assumption has been upheld by several authors who
have described cases of previous rhegmatogenic retinal detachment in which an ERM
did not form postoperatively if the ILM had been surgically removed during treatment
of the retinal detachment [46], [47].
In a retrospective analysis, Forlini et al. investigated 159 patients who had been
treated with ppV due to retinal detachment [48]. In 78 eyes, ILM peeling was performed in addition to ppV. Whilst ERMs developed
postoperatively in 9% of patients in the cohort that underwent ILM peeling, this figure
rose to 31% of patients in the cohort that did not undergo ILM peeling (p = 0.001).
Similarly, in a multivariate analysis of different endotamponades, the risk was reduced
significantly by 75% in the peeling cohort compared to the cohort with no peeling.
Moreover, postoperative visual acuity was significantly better in the cohort with
peeling (0.32) compared to the cohort without peeling (0.16; p = 0.01).
This clear correlation becomes even more evident if we consider that the peeling cohort
in the Forlini et al. study contained a larger proportion of eyes with risk factors
for developing an ERM than did the cohort without peeling. Thus, 24 out of 78 patients
in the peeling group had suffered vitreous hemorrhage or proliferative vitreoretinopathy
in the eye that was included in the study. As these diseases can promote ERM formation,
they should be taken into account when considering whether or not to perform ppV with
ILM peeling [48].
In a previously published retrospective analysis of 135 eyes that had undergone vitrectomy
due to retinal detachment, Nam and Kim reported no cases of postoperative ERMs occurring
in the subgroup of patients who had undergone ILM peeling. In contrast, ERMs developed
in 21.5% of the group of patients who had not undergone ILM peeling. In this study,
all of the patients underwent ppV with gas injection [46].
In a study by Aras et al., 42 eyes that had been treated for retinal detachment with
silicone oil were observed over a period of around 25 weeks. The authors described
a 27% rate of postoperative ERMs in the patients who had not undergone peeling. The
outcome was markedly different in patients who underwent ILM peeling during vitrectomy;
none of the patients in this group developed an ERM [47].
Forlini et al. have postulated that the reduced occurrence of ERMs after ILM peeling
is due to the fact that RPE cells need to be anchored to a basal membrane in order
to proliferate and form an ERM. The ILM functions as a basal membrane for the Müller
cells in the retina. ILM peeling prevents detached RPE cells from accumulating and
proliferating epiretinally; this in turn prevents the formation of an ERM [48].
The studies presented here demonstrate that there is a clear benefit to removing the
ILM as part of treatment for a detached retina. However, it should be noted that peeling
can be technically challenging in eyes with retinal detachment, especially when the
detachment affects the macula. For this reason, the indication for prophylactic peeling
should also depend on the experience of the surgeon who is to perform it.
Risks of Surgery
In a case series of 1 131 eyes in which an ERM was treated by vitrectomy and peeling,
intraoperative complications occurred in 9.8% of cases. The most common adverse events
associated with this procedure were iatrogenic retinal tears (4.9%), iatrogenic retinal
trauma (1%), and intraoperative lens touch (1%). In procedures performed without a
cataract operation, complications occurred in 8.1% of cases. In 3% of cases, another
ERM operation was required 5.5 months after the primary operation. In 1% of cases,
retinal detachment occurred 3.2 months after the operation [22]. However, other studies have shown higher rates of postoperative retinal detachment.
For example, in a case series of 362 eyes, Guillaubey et al. reported retinal detachment
after 70 days following an ERM operation in 9 eyes (2.5%) [49]. The overall rate of endophthalmitis after vitrectomy is low, with an incidence
ranging from 0.14 to 0.84 reported in the
literature [50].
Summary
An ERM is one of the most common retinal diseases. It occurs in 35 to 50% of the population,
and its incidence increases with age.
As well as subjective symptoms such as metamorphopsia, aniseikonia, or loss of visual
acuity, changes can also be observed on OCT. On OCT, an ERM manifests as a hyperreflective
membrane on the surface of the retina. In severe cases, it can also be diagnosed clinically
using fundoscopy. A new OCT-based classification proposed by Govetto et al. can be
used to determine the prognosis for postoperative visual acuity [20].
Other retinal findings, such as an increase in central foveal thickness or the presence
of an EIFL accompanying the ERM, can also influence postoperative outcome.
Vitrectomy with peeling of the ERM represents the gold standard for surgical treatment.
Based on the current literature, ILM peeling can also be recommended.
