Post-vitrectomy Macular Edema: Shedding Light on Incidence and Risk Factors

Abstract

Post-surgical macular edema is a common cause of delayed visual recovery. Although its occurrence after cataract surgery is well known, the specifics on post-vitrectomy macular edema (PVME) are less readily available in the recent literature. This narrative review identifies a wide-ranging incidence due to heterogeneous surgical indications, study designs, and lack of diagnostic standards in the current literature. PVME requiring pharmacological treatment ranges from 2.0 to 27.3%. Higher rates of macular edema are encountered after vitrectomy for retinal detachment, retained lens fragments, and epiretinal membranes. The intraretinal cystic changes after epiretinal membrane removal, however, often represent remnants of tractional disease, rather than true macular edema. This review elaborates on several patient-related and surgical risk factors, mostly in retinal detachment surgery. Predisposing factors are redetachment, proliferative vitreoretinopathy, silicone oil tamponades, extensive retinopexy, and macular involvement in cases with a retinal detachment. Patients who undergo procedures for macular holes, vitreous floaters, and secondary lens implantation appear less susceptible to developing PVME. Furthermore, higher rates of PVME are seen after crystalline lens removal, both in combined phacoemulsification-vitrectomy and when cataract surgery is performed after the vitrectomy. Additional research should lead to improved risk assessment and a reduced burden of this clinical issue.

Key Summary Points

This review elucidates the overall incidence of post-vitrectomy macular edema (PVME), requiring treatment in 3.0–27.3% of patients after vitrectomy.

It highlights a higher incidence of PVME after retinal detachment, in which several risk factors are identified, such as redetachment, proliferative vitreoretinopathy, silicone oil tamponades, and macular involvement.

Increased intraretinal cystic changes are reported after epiretinal membrane removal but can be considered a more complex entity due to chronic tractional disease.

High incidence of PVME is also found after removal of retained lens fragments, while other common surgical indications, such as macular holes, vitreous floaters, and secondary lens implants, lead to lower incidences.

Combined and sequential phacoemulsification leads to an increased risk of developing PVME.

Introduction

Post-surgical macular edema is often responsible for poor postoperative visual acuity. It is well known as Irvine–Gass syndrome following modern cataract surgery and occurs in 0.1 to 2.4% of patients [1]. Data on cystoid macular edema after pars plana vitrectomy, however, is limited. Post-vitrectomy macular edema (PVME) typically occurs at 4–12 weeks after surgery [2–4]. Indications for vitrectomy procedures are diverse, implying a more heterogeneous pathogenesis (Fig. 1). Inflammation, causing breakdown of the inner and outer blood–retinal barriers, plays a major role [5]. In vitreomacular traction syndromes and epiretinal membranes (ERMs), post-surgical intraretinal cystoid changes represent a more heterogeneous entity. Other than inflammatory cystoid macular edema, the mechanical traction can disrupt retinal and vascular integrity, facilitating fluid buildup, and microcystoid macular edema may appear because of Müller cell dysfunction [6].

Fig. 1.

Optical coherence tomography images showing post-vitrectomy macular edema (PVME) within 4 months after surgery for different indications, including preoperative images (upper images). A PVME after retinal detachment repair, B, C PVME after removal of idiopathic epiretinal membrane with and without vitreomacular traction: note the preoperative tractional cystoid changes that contribute to the development of postoperative PVME, D macular hole showing postoperative closure with persistence of subretinal fluid and an increase in intraretinal cystoid spaces, E PVME after removal of vitreous floaters, F more subtle PVME after removal of dropped intraocular lens (IOL) and secondary IOL implantation (iris claw lens)

Macular edema is defined as swelling or thickening of the neurosensory retina within the macular region. No strict diagnostic criteria are currently defined for post-surgical macular edema. In clinical practice, optical coherence tomography (OCT) is the most convenient method for detection and follow-up of macular edema. OCT images characteristically show intraretinal cystoid hyporeflective areas, predominantly in the inner nuclear layer (INL) and the outer plexiform layers, associated with retinal thickening and possible loss of the foveal depression [3, 7]. Cystoid spaces in the outer retinal layers or subretinal fluid are less common [3, 7]. Fluorescein angiography is often cited as the gold standard for diagnosing postoperative macular edema and is a useful tool for excluding other causes, such as diabetic retinopathy and retinal vascular disease. Angiographic post-surgical macular edema typically shows perifoveal petaloid leakage and optic disc staining but is found in up to 70% of patients after phacoemulsification, mostly without clinical significance [1].

Post-surgical macular edema with visual implications often requires treatment, including steroid formulations, non-steroidal anti-inflammatory drugs, or intravitreal injections with anti-vascular endothelial growth factor receptor medication. Despite these treatments, up to 20% of affected patients develops chronic disease [2–4]. The scope of this narrative review is to provide an overview of the available data on the incidence and risk factors of PVME.

Methods

A comprehensive literature research review was conducted using the PubMed and Embase databases for relevant articles in English. Search terms were ‘macular edema’ or ‘CME’ combined with ‘vitrectomy’ or indications such as ‘retinal detachment’, ‘macular hole’, ‘epiretinal membrane’ or ‘pucker’, ‘secondary lens implant’, ‘vitreomacular traction’, ‘floater’ and ‘vitreous hemorrhage’. Articles were selected on relevance and availability of data on either incidence or risk factors to develop PVME. Additional references cited in the read articles were included. The review is based on articles published between January 2000 and April 2025. This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Results

The current literature reports a wide-ranging incidence of PVME, up to 46.8% [8]. Prospective study designs report an incidence of intraretinal cystoid changes in 10.0–27.6% of patients [2, 9, 10]. Publications considering post-surgical macular edema requiring treatment report rates of 2.0–27.3% of patients [2, 3, 11–19].

Diabetes is often not a significant risk factor for the development of post-surgical macular edema [7, 8, 18, 20]. In a diabetic population, distinguishing diabetic macular edema from post-surgical macular edema can be difficult. A higher incidence of PVME is seen in patients with diabetic retinopathy, but one-third of these cases show some characteristics of diabetic macular edema [21]. Proliferative disease is no additional significant risk factor compared to non-proliferative disease (29.2% vs 26.3%) [21]. In the general population, older age is a minor but often insignificant risk factor [5, 10, 12, 20, 22–24].

Retinal Detachment

An inflammatory response is seen in detached porcine retina, which could contribute to the development of macular edema next to surgically induced inflammation and apoptosis of detached retinal cells [25]. Most data on PVME is available after retinal detachment repair, with incidence ranging from 1.9% to 33.9% [3, 5, 7, 10, 12, 23, 24, 26, 27].

The most important protective factor for PVME development is single surgery success [3, 5]. Macular detachment is an independent risk factor [3, 10, 12, 22], although significance is not always confirmed [5, 23, 24]. An increased incidence of PVME in delayed surgery is only found in one study [5, 7, 12].

Most publications find no significant difference between the phakic and pseudophakic population [7, 10, 12, 23, 24]. However, sequential cataract surgery within 6 months after retinal detachment may predispose patients to PVME [5, 7].

Endolaser retinopexy is an independent risk factor for PVME development compared to transscleral cryotherapy retinopexy, although this effect can be attributed to the more extensive retinopexy for which endolaser is reserved [7, 23]. A predisposing effect of 360-degree peripheral endolaser is not always found [26–28]. Low-grade proliferative vitreoretinopathy (PVR grade B) implies inflammation and expectedly shows a strong association with macular edema [5, 12, 23].

Other than aphakia, however, several known risk factors of PVR development do not significantly predispose to post-surgical macular edema, such as high myopia, the presence of vitreous hemorrhage and the number or size of retinal tears [3, 5, 7, 10, 12, 23, 29].

Grade C PVR is often excluded from publications but corresponds to higher rates of PVME (18.0–54.4%) [7, 15]. Additionally, retinectomy is identified as a risk factor but could be confounded by the presence of PVR [3, 7]. Similarly, PVME is seen twice as much after silicone oil tamponade compared to gas tamponade [3, 7]. Risk factors in silicone oil tamponade are the presence of a posterior staphyloma, macular detachment and longer duration of oil tamponade [20, 22, 30]. Correspondingly, the highest incidence can be found in a cohort with the longest mean duration of silicone oil tamponade [20]. Mechanisms contributing to macular edema in these cases could be decreased molecular transport in the remaining fluid volume, leading to permanence of inflammatory substances between the silicone oil and the retina. Gas or air tamponade, surgically induced posterior vitreous detachment, use of heavy carbon liquid, staining with triamcinolone, and preexisting glaucoma do not show correlation with PVME in patients who receive a vitrectomy because of a retinal detachment [3, 5, 7, 10, 12, 23].

Vitreomacular Disorders

Incidence of PVME in idiopathic ERMs ranges from 12.8% to 26.5% [14, 19, 31–35]. This relatively high incidence could be explained by the contribution of additional tractional forces causing cystoid intraretinal changes, foveoschisis, or vascular disruption prior to surgery; intraretinal cystoid spaces after ERM removal are found in 27.6% and in up to 40.0% if associated with prior vitreomacular traction (Fig. 2) [9, 36].

Fig. 2.

Optical coherence tomography images showing different modalities of intraretinal fluid after epiretinal membrane removal, including preoperative images (upper images). A, B Residual intraretinal cystoid spaces caused by both mechanical traction and microcystoid macular edema linked to Müller cell dysfunction in epiretinal membranes. The microcystoid macular edema is represented by distinctive oval hyporeflective cystoid spaces within the inner nuclear layer, sparing the fovea, C, D Post-vitrectomy macular edema associated with increased inflammation

Signs of significant retinal tractional disruption, such as increased ectopia (stage 4 ERMs) or preoperative cystoid spaces, are associated with a higher risk of developing or retaining intraretinal cystoid spaces after surgery [9, 14, 32, 36].

A predisposing biomarker for PVME is the cotton-ball sign, seen on OCT as a diffuse, round hyperreflective area between the ellipsoid zone and the interdigitation zone at the foveal center [34]. Preoperative separation of the ERM and internal limiting membrane (ILM) seems to play a protective role [34]. If present, the area of associated vitreomacular adherence or traction does not present a higher risk [36]. Other non-significant factors are phakic or pseudophakic status, surgically induced posterior vitreous detachment, and air or gas tamponade [9, 32, 36].

Peeling of the ILM is often performed to prevent ERM regrowth and has been proposed as treatment for refractory macular edema, but no significant impact is seen on the incidence of PVME in ERMs nor in retinal detachment repair [5, 7, 31].

The incidence of PVME in full-thickness macular holes appears generally lower (0.0–5.2%) [17, 33, 37], although one study surprisingly found similar PVME rates in ERMs (26.5%) and macular holes (27.3%) [14]. As this indication is less well published, no clear risk factors can be identified in the current literature. Importantly, it is suggested that macular edema can lead to a reopening of the macular hole [37].

Other Vitreoretinal Indications

Vitrectomy for floater removal boasts one of the lowest incidences of PVME (5.5%) [13]. Complicated cataract surgery with retained lens fragments requiring pars plana vitrectomy is often associated with an increased inflammatory response, explaining higher PVME rates in larger series (27.2–28.0%) [4, 38]. One publication fails to show a significant benefit of same-day over delayed surgery [38], although other data suggests lower rates of PVME after same-day vitrectomy due to limitation of lens-induced inflammation [39].

The incidence of PVME in secondary lens implantation ranges from 3.9% to 16.4% [11, 16, 18, 40, 41]. As expected, it is found more often after complicated cataract surgery as compared to intraocular lens (IOL) dislocation due to zonular dehiscence [18]. Myopia, pseudoexfoliation, and glaucoma show no significant associations [18]. When comparing anterior to retropupillary iris claw implantation, one meta-analysis comprising 347 patients favors retropupillary implantation in terms of PVME [42]. Proposed mechanisms are a more stable epithelial implantation compared to a stromal inclavation leading to less IOL movement and less induced inflammation [42]. When comparing Carlevale IOLs and other scleral fixation techniques to iris fixation, a recent review concludes on similar weighted mean PVME rates [43].

Surgical Technique

PVME after retinal detachment repair is more frequently encountered after combined procedures compared to a vitrectomy-only group [10]. However, consecutive phacoemulsification after retinal detachment repair also leads to a higher incidence of PVME [5]. In vitreomacular disorders, several publications fail to find a significantly higher incidence after combined procedures [2, 9, 14, 19, 31, 32, 34].

One recent study, however, found significantly higher rates of increased macular thickness after combined procedures for ERM removal (15.4% vs. 3.8%), suggesting that crystalline lens removal is the main culprit in the development of PVME [35].

Interestingly, macular subfield thickening after ERM removal is more prevalent after consecutive cataract surgery compared to combined procedures [19].

The use of smaller gauge (G) systems results in less early intraocular inflammation but has not proven to be beneficial regarding PVME [11, 13].

One publication fails to demonstrate an impact of operative time on the development of PVME [22]. No significant benefit is linked to the experience level of the surgeon [7, 33].

Discussion

Pars plana vitrectomy can provoke macular edema, as illustrated by the occurrence of after uncomplicated floater removal. This review sheds light on the incidence of PVME; however, it is important to notice that available studies differ in terms of heterogeneous surgical indications, diagnostic and inclusion criteria, follow-up period, and perioperative management (Table 1).

Table 1.

Original research papers included in the review

First author	Year	Design	Indication	Patients	Diagnostic criteria of CME	%	Treated
Aljohani	2021	Retrospective	Macular hole or epiretinal membrane	161	OCT (CMT increase and loss of foveal depression)	9.2	No
Almeida	2012	Prospective	Macular hole	50	OCT (not specified)	2.0	Yes
Al-Dwairi	2022	Retrospective	Secondary lens implant	171	Not specified	8.8	Yes
Azzolini	2014	Retrospective	Retinal detachment repair with silicone oil	92	OCT (IRC and CFT increase)	21.7	No
Bae	2012	Retrospective	Retinal detachment repair with silicone oil	52	OCT (IRC and CFT increase and loss of foveal depression)	19.6	No
Bajgai	2018	Retrospective	Dropped intraocular lens	61	Not specified	16.4	Yes
Banker	2015	Retrospective	Retinal detachment	587	OCT (IRC)	15.2	No
Bernabei	2023	Retrospective	Retinal detachment	164	OCT (IRC)	17.1	No
Bernal Morales	2023	Retrospective	Secondary lens implant	325	OCT (CRT > 300 μm)	11.7	Yes
Chatziralli	2021	Retrospective	Retinal detachment	86	OCT (not specified)	16.3	Yes
Coussa*	2019	Retrospective	Epiretinal membrane and vitreomacular traction	30	OCT (IRC)	40.0	No
Choi	2021	Retrospective	Secondary lens implant	322	Not specified	3.9	Yes
De Nie	2010	Retrospective	Vitreous floaters	110	OCT (IRC)	5.5	Yes
Frisina	2014	Retrospective	Epiretinal membrane	242	OCT (IRC)	12.8	No
Gebler	2022	Prospective	Retinal detachment	49	OCT (IRC)	16.3	No
Gershoni	2022	Retrospective	Various in diabetic retinopathy	104	OCT (IRC and CMT increase of 60 μm)	26.9	No
Guber	2019	Retrospective	Epiretinal membrane	98	OCT (IRC)	16.3	No
Hamoudi	2018	Prospective	Epiretinal membrane	39	OCT (10% CSMT increase)	12.8	Yes
Iuliano	2021	Retrospective	Macular hole or epiretinal membrane	71	OCT (new or increased IRC and 10% CFT increase)	27.3	Yes
Kim	2009	Prospective	Various	109	OCT (CFT > 272 μm)	46.8	No
Kiss	2007	Retrospective	Retinal detachment repair with silicone oil	39	OCT (not specified)	18.0	Yes
Leisser*	2018	Prospective	Epiretinal membrane	69	OCT (IRC)	27.6	No
Merad	2022	Retrospective	Retinal detachment	493	OCT (IRC)	33.9	No
Modi	2013	Retrospective	Retained lens fragments	569	Not specified	27.2	No
Park	2024	Retrospective	Epiretinal membrane	267	OCT (10% CMT increase)	13.1	No
Passemard	2010	Retrospective	Macular hole	125	Not specified	5.2	No
Peters	2022	Retrospective	Retinal detachment	192	Not specified	9.4	No
Pole	2021	Retrospective	Retinal detachment	99	OCT (IRC)	25.0	Yes
Savastano	2010	Prospective	Macular hole or epiretinal membrane	110	OCT (not specified)	10.0	Yes
Scott	2003	Retrospective	Retained lens fragments	343	Not specified	28.0	No
Silva	2021	Retrospective	Epiretinal membrane	108	OCT (IRC and 10% CSMT increase)	18.5	No
Starr	2021	Retrospective	Retinal detachment	1466	OCT (IRC)	9.6	No
Wang	2020	Retrospective	Retinal detachment	2248	Not specified	6.9	No
Yang	2018	Retrospective	Retinal detachment repair with silicone oil	58	OCT (not specified)	36.2	No
Zaleski	2022	Retrospective	Secondary lens implant	175	Not specified	9.1	Yes

CME cystoid macular edema, CMT central macular thickness, CRT central retinal thickness, CSMT central subfield mean thickness, IRC intraretinal cystoid spaces, OCT optical coherence tomography, CFT central foveal thickness

*Presence of intraretinal cystoid spaces was not defined as macular edema

It is reasonable to classify the surgical indications into a lower (less than 10%) and a higher PVME risk group in which additional triggers are present (over 10%, Table 2). Populations with a higher PVME chance include patients with retained lens fragments after complicated cataract surgery and patients with retinal detachment, which both are associated with a higher degree of inflammation induced by either lenticular particles or detached retina. Any pro-inflammatory context like PVR, as well as preexisting inflammatory conditions, should warrant more caution towards PVME. In retinal detachment repair, redetachment and macular involvement are major risk factors, but unfortunately often inevitable. Delaying retinal detachment repair and long silicone oil tamponades should be avoided, where possible. Higher rates of intraretinal cystoid spaces are reported after removal of idiopathic ERM, although these changes do not always imply an underlying inflammatory process causing cystoid macular edema. While overlap is possible, these changes often represent sequelae of chronic mechanical traction or microcystoid macular edema caused by Müller cell dysfunction (Fig. 2) [6]. The presence and degree of preoperative cystoid changes and the distinctive oval hyporeflective cystoid spaces within the INL sparing the fovea are important clues in recognizing these modalities which do not need further treatment [6].

Table 2.

Summary of reported incidence of post-vitrectomy macular edema in common surgical indications

Indication	Incidence (%)	Median incidence (%)
Retinal detachment	1.9–33.9	16.3
Epiretinal membrane	12.8–26.5	14.6
Macular hole	0.0–27.3	4.4
Vitreous floaters	5.5	5.5
Retained lens fragments	27.2–28.0	27.6
Secondary lens implantation	3.9–16.4	9.1

Patient groups with a lower PVME risk include those who undergo floater removal and macular hole surgery. Although macular holes are often associated with a certain degree of tractional force, this seems to play a less significant role. Lastly, secondary lens implantation could be considered a low-risk procedure outside the context of a complicated phacoemulsification.

Release of lens fragments during surgery leads to higher rates of PVME. In this regard, combined surgery probably results in higher chances of PVME. However, incidence after sequential cataract surgery is evenly elevated. In these cases, the prior removal of the vitreous buffer could facilitate exposure of the macular region to proinflammatory substances and explain proportionally higher rates compared to phacoemulsification in non-vitrectomized eyes.

Retinopexy in retinal detachment repair can induce inflammation and breakdown of the blood-retinal barrier and the extent has been hypothesized to be associated with PVME. The impact of 360-degree peripheral endolaser in elective vitreoretinal surgery has yet to be elucidated.

Other surgical factors, such as sclerotomy size as well as patient characteristics (myopia, glaucoma, diabetic status), play rather insignificant roles in the development of PVME. Additional factors could be evaluated in future research, such as the impact of vital stains and the viscosity value of silicone oil. Moreover, only a few prognostic imaging biomarkers have been identified so far.

OCT is most widely used in diagnosing macular edema; the most prevalent criterion is the presence of intraretinal cystoid spaces, which are found in up to 40.0% of patients after vitrectomy for vitreomacular traction [36]. An incidence up to 46.8% is reached using central macular thickness or central foveal thickness as diagnostic criterion [8].

As both criteria are non-specific to inflammatory post-surgical macular edema and also occur in retinal traction, it is more useful to look at an increase of said parameters in these pathologies. However, a clear-cut and optimal definition of PVME still has to be discovered.

It is important to note that these imaging features not always correspond to visual impairment or the need of treatment and may lead to an overestimation of the clinical burden of PVME. As mentioned earlier, publications considering PVME needing pharmacological treatment report lower rates. In a clinical setting, the diagnosis of inflammatory PVME can be complicated by tractional components or diabetic macular edema.

Other limitations of the current study could also be identified. A wide variety of inclusion criteria has been applied in available studies, often excluding more complex cases or patients with predisposing diseases such as diabetic retinopathy or uveitis. A wide range is also seen in the follow-up period, from 6 weeks in prospective designs up to several years in retrospective designs. Missing data could also account for under- or overestimation in some of these publications. As PVME can occur several months after surgery, a follow-up period of at least 12 months should be pursued. Finally, the benefit of perioperative and post-surgical medication has been poorly elucidated. Post-surgical regimens with steroids and non-steroidal anti-inflammatory drugs are variable and often not disclosed in the available literature. A more potent steroid prophylaxis may be warranted in cases with the risk factors identified in this review.

Conclusions

This review aims to provide an overview of the incidence of PVME found in the current literature. Vitrectomy procedures for retinal detachment, retained lens fragments, idiopathic ERMs, and combined procedures are more likely to predispose to macular edema development, as is sequential surgery. Furthermore, breakdown of the blood–retinal barrier in inflammatory conditions such as PVR or by retinopexy and release of crystalline lens particles is associated with increased PVME incidence. Diagnostic criteria play a major role in the reported incidence; some OCT-based definitions of macular edema tend to overestimate clinically relevant cases. Better prediction and understanding of PVME should lead to a better preventive strategy and reduced patient burden.

Author Contributions

Viktor Verplaetse (concept/design, drafting, critical revisions), Elon H.C. van Dijk (critical revisions, approval), and Koorosh Faridpooya (critical revisions, approval) contributed to the conception, literature review, manuscript drafting and critical revision of this narrative review. All authors approved the final version of the manuscript.

Funding

No funding or sponsorship was received for the publication of this article.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of interest

Viktor Verplaetse, Elon H.C. van Dijk, and Koorosh Faridpooya have nothing to disclose.

Ethical approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.