Original article
Development of vitreomacular interface abnormality in patients
with diabetic macular edema
Chun-Kai Chang
a, Cheng-Kuo Cheng
a,b,c,*, Chyi-Huey Bai
a, Chi-Hsien Peng
a,b, Chao-Chien Hu
a,b aDepartment of Ophthalmology, Shin-Kong Wu Ho-Su Memorial Hospital, Taipei, TaiwanbCollege of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan cCollege of Medicine, National Taiwan University, Taipei, Taiwan
a r t i c l e i n f o
Article history: Received 6 March 2012 Received in revised form 26 April 2012
Accepted 10 May 2012 Available online 11 July 2012 Keywords:
diabetic macular edema diabetic retinopathy epiretinal membrane optical coherence tomography vitreomacular interface
a b s t r a c t
Purpose: To report the prevalence of vitreomacular interface abnormality (VMIA) and the incidence of the development of VMIA in patients with diabetic macular edema (DME). Factors associated with the development of VMIA were also investigated.
Methods: This is a retrospective observational study. Patients with DME who were followed for at least 6 months were reviewed. Ophthalmoscopic examination, fundus photography,fluorescein angiography, and optical coherence tomography were used to detect DME and VMIA.
Results: A total of 244 eyes in 180 patients were found to have DME at the initial visit. Of these, 16 eyes (6.56%) were also diagnosed with VMIA. Ninety-six eyes in 76 patients with DME who did not receive an intravitreal operation or medications were included in the follow-up study. VMIA developed in 8 eyes (8.33%) over a mean follow-up duration of 22.63 months, which corresponded to a calculated annual incidence of 4.42%. The mean time from the initial visit to the detection of VMIA was 19 months. Older age was found to be a risk factor for the development of VMIA. The occurrence of VMIA was associated with worsening visual acuity and a thicker central retinal thickness on thefinal visit.
Conclusion: This study reveals the prevalence of VMIA on the initial visit and the incidence and risk factors associated with the occurrence of VMIA in patients with DME. This study also found that the occurrence of VMIA was significantly associated with poor visual outcomes and worsened macular edema at the end of the observation period.
CopyrightÓ 2012, The Ophthalmologic Society of Taiwan. Published by Elsevier Taiwan LLC. All rights reserved.
1. Introduction
Diabetic retinopathy is one of the leading causes of blindness worldwide, and diabetic macular edema (DME) is the most important cause of visual loss in patients with diabetic retinop-athy.1A large population-based study, the Wisconsin Epidemiologic Study of Diabetic Retinopathy, reported that the incidence of DME over the studied 10-year period was 20.1% in the young-onset group, 25.4% in the old-onset group (which was taking insulin), and 13.9% in the oldest-onset group (which was not taking insulin).2For years, macular laser photocoagulation has been the standard treatment for DME. The Early Treatment of Diabetic Retinopathy Study (ETDRS) reported that laser photocoagulation
significantly reduces the risk of moderate visual loss in patients with clinically significant macular edema.3 However, approxi-mately 12% of treated patients still lost their vision.3For diffuse DME, the nonresponse rate to macular laser treatment was re-ported to be 24.6% in a long-term study4; persistent macular edema was believed to be the main cause of this unresponsiveness.3 Recently, vitreomacular interface abnormality (VMIA) has been found to be associated with persistent DME that is unresponsive to macular laser therapy.5e7
Clinically, VMIA is characterized by a highly reflective band over the inner retinal surface at the macular area.7It is defined by optical coherence tomography (OCT) as having the features of epiretinal membrane (ERM) and/or anomalous vitreomacular adhesions.6e8 The latter is sometimes referred to as posterior vitreomacular separation,9posterior hyaloid traction,5or vitreomacular traction7 in the literature and is characterized by ophthalmoscopic exami-nation as having the features of a taut, thick posterior hyaloid.9OCT is generally regarded as superior to conventional fundus exami-nation for the evaluation of the vitreomacular interface in order to
* Corresponding author. Department of Ophthalmology, Shin Kong Wu Ho-Su Memorial Hospital, Number 95, Wen-Chang Road, Shih-Lin District, Taipei 11120, Taiwan.
E-mail address:[email protected](C.-K. Cheng).
Contents lists available atSciVerse ScienceDirect
Taiwan Journal of Ophthalmology
j o u r n a l h o m e p a g e : w w w . e - t jo . c o m
2211-5056/$e see front matter Copyright Ó 2012, The Ophthalmologic Society of Taiwan. Published by Elsevier Taiwan LLC. All rights reserved. http://dx.doi.org/10.1016/j.tjo.2012.05.001
diagnose various macular diseases.6,10,11In clinical evaluations of patients with persistent DME, Ghazi et al found that OCT is 1.9 times more sensitive for the detection of VMIA than a combination of biomicroscopic fundus examination, stereoscopic fundus photography, andfluorescein angiography (FA).6
Recently, vascular endothelial growth factor (VEGF) was shown to be associated with the pathogenesis of diabetic retinopathy and DME,12e14and many clinical studies have also revealed the bene-ficial effects of intravitreal injection of anti-VEGF antibodies such as bevacizumab and ranibizumab for both reducing macular edema and improving visual acuity in patients with DME.15e20In a large-scale, randomized, controlled trial, intravitreal anti-VEGF injec-tion was found to be superior to macular grid laser therapy in terms of 1- and 2-year clinical outcome measurements.19,20However, in a recent study, Wu et al revealed that even with anti-VEGF treat-ment, DME and the features of VMIA may be associated with poor clinical results.21 On the other hand, growing evidence suggests that vitrectomy with the induction of posterior hyaloid separation and the release of the tractional membrane might provide bene fi-cial effects for the treatment of persistent DME in association with VMIA.22Therefore, determining the features of VMIA may provide significant clinical implications for the clinical management of patients with DME. However, clinical studies exploring the inci-dence and the risk factors of VMIA formation in DME patients are still very scarce in the literature.
In this study, we retrospectively investigated the incidence of VMIA formation in a group of patients with clinically evident DME who were regularly followed at our clinic for at least 6 months. The prevalence of VMIA, as indicated by the OCTfindings in all patients with DME on the initial visit, was also recorded. Since July 2005, the retinal clinic of our institute began to perform standardized macular OCT examinations on all patients on their initial visit, in addition to ophthalmoscopic examination and/or supplementary fundus photography and FA. Patients with diabetic macular edema should have OCT examinations performed at every scheduled visit. The occurrence of VMIA in patients with DME who regularly visit our clinic could, thus, be recorded. We excluded patients who had undergone intravitreal operations, such as vitrectomy or intra-vitreal injection of anti-VEGF or corticosteroids, prior to or within 6 months of the observation. To better understand the conditions associated with the formation of VMIA in DME patients, several systemic and ocular factors, including age, sex, the presence of hypertension and hyperlipidemia, mean level of glycosylated hemoglobin (HbA1c), baseline refractive status, best-corrected visual acuity (BCVA), central retinal thickness (CRT), and history of cataract surgery, macular laser treatment, and pan-retinal photocoagulation (PRP) were further analyzed to determine any correlations. We hope that this report may help ophthalmologists better understand the prevalence, incidence, and risk factors for the formation of VMIA in patients with DME.
2. Patients and methods
We conducted a retrospective study of patients with DME who were treated at Shin Kong Wu Ho-Su Memorial Hospital (Taipei, Taiwan) from March 2006 through January 2010. Approval from the institutional review board and ethics committee was obtained. The clinical records of all consecutive patients with DME were reviewed. The diagnosis of DME was made by the presence of exudative changes and the thickening of the macular area on the ophthalmoscopic examination and evidence of late macular leakage on FA. Increased CRT, cystic changes, and subretinalfluid might also have been found on the OCT examination. Because different combinations of these features were often found in the same patients, we did not classify our patients according to the FA
or OCT patterns that were noted in this study. Eyes with evidence of VMIA on the initial visit were recorded. Eligible patients were enrolled in this study according to the inclusion and exclusion criteria. The inclusion criteria for our study included an initial diagnosis of DME without VMIA by ophthalmoscopy, FA, or OCT with regular follow-up for at least 6 months after thefirst diag-nosis of DME. Patients were excluded if they had any of the following conditions: (1) regular follow-up<6 months; (2) VMIA or vitreous hemorrhage on the initial visit; (3) vitrectomy prior to or within 6 months of the follow-up period; (4) intravitreal anti-VEGF or corticosteroid injection prior to or within 6 months of the follow-up period; (5) vitreous hemorrhage that precluded a detailed macular examination within 6 months of the follow-up period; and (6) concomitant ocular diseases such as posterior uveitis, advanced age-related macular degeneration, or retinal vascular occlusions. If a vitrectomy or intravitreal injection of anti-VEGF or corticosteroids was administered to our patients during the course of this study, data collection was terminated and the follow-up period was considered until the last visit before surgery. On the other hand, macular focal, grid laser, PRP, and cataract operations before or during the follow-up period were recorded as items for subsequent analysis.
Each patient underwent BCVA measurement using ETDRS charts at a distance of 4 m and ophthalmic slit-lamp biomicroscopy on each visit. Central retinal characteristics were analyzed using OCT (Stratus III; Carl Zeiss, Dublin, CA, USA) that utilized six diagonal 6-mm radial line scans through a dilated pupil and the macular thickness map program (versions 4.0). A macular thickness map algorithm centered on the fovea was also performed. Patients were regularly followed at 1w3-month intervals.
The OCT characteristics of VMIA were defined as having the features of ERM or anomalous vitreomacular adhesions.6,8The OCT configurations of ERM were defined as partially separated or globally adherent membranes above the macular area according to the classification reported by Wilkins et al.8 Partially separated ERM may be detected as thin, but distinctive, highly reflective bands just above the inner surface of the retina with some focally attached points. Globally adherent membranes may be detected when various combinations of the following configurations are noted: macular pseudohole, difference in the optical reflectivity between the membrane and retina, or a visible tuft or edge of the membrane. Anomalous vitreomacular adhesion was defined as an evident hyperreflective band that adhered to the surface of the retina at specific sites, was elevated elsewhere off of the surface, and continuous with the posterior vitreous surface, in agreement with the report by Ghazi et al.6 Ophthalmoscopic indications of VMIA were defined as either a fine or glistening membrane over-lying the macula or a thickened whitish tissue that induced wrin-kling of the retinal surface and obscured the underlying vasculature on ophthalmoscopic examination and fundus photography. Torturous vessels with varying degrees of leakage might be present on FA. Eyes were diagnosed with VMIA when an OCT configuration of ERM or anomalous vitreomacular adhesion was detected in at least two of the six diagonal scans on the OCT examination together with characteristicfindings on the ophthalmoscopic examination, fundus photography, or FA.
Systemic conditions, such as a history of hypertension, hyper-lipidemia, macular grid, focal laser, and PRP and the level of HbA1c were recorded on follow-up examinations. Diabetes mellitus was diagnosed from a self-reported history of medications and/or a fasting plasma glucose level7.0 mmol/L or a 2-hour postloading glucose level11.1 mmol/L. Hypertension was defined as systolic blood pressure140 mmHg, diastolic blood pressure 90 mmHg, or the concurrent use of antihypertensive medication. Hyperlipid-emia was defined as total cholesterol 200 mg/dL, low-density
lipoprotein cholesterol100 mg/dL, triglycerides 150 mg/dL, or the concurrent use of antihyperlipidemic agents.
Visual acuity was recorded in Snellen units and converted to logarithm of minimal angle of resolution (LogMAR) units for statistical analysis. The values are expressed as the mean standard deviation. The data collected at the initial visit were defined as baseline data; data from the last visit were defined as the final data. We also recorded the number of eyes that had received cataract surgery, macular grid or focal laser treatment, and PRP before or during follow-up.
We compared the mean age, mean HbA1c during follow-up, baseline refractive status, baseline BCVA,final BCVA, baseline CRT, final CRT, and the follow-up duration using the independent samples t test for patients who developed VMIA (VMIA [þ] group) and patients who did not (VMIA [-] group). The paired t test was used to compare baseline BCVA and CRT values with thefinal BCVA and CRT values, respectively, between the two groups. We also compared sex and the proportion of patients with a history of hypertension, hyperlipidemia, cataract surgery, macular laser, and PRP between the two groups using the Fisher exact test. Stepwise multiple logistic regression was also used to determine associations between the binary dependent variable (presence of VMIA) and the continuous and categorical independent variables, such as age, hypertension, hyperlipidemia, mean HbA1c, baseline refraction, cataract surgery, macular laser, PRP, baseline BCVA, and baseline CRT. A p value <0.05 was considered statistically significant. Statistical analysis was performed using SPSS 11.0 software (SPSS Inc., Chicago, IL, USA).
3. Results
3.1. Prevalence of VMIA on the initial visit
A total of 244 eyes in 180 consecutive patients with DME received complete standardized examinations on the initial visit at our retinal clinic. Of these, 16 eyes (6.56%) were found to have a preexisting VMIA that was found during the initial visit. These eyes, together with another 132 eyes, were excluded from the study according to the exclusion criteria. The remaining 96 eyes in 76 patients were included in the study when calculating the incidence of developing a subsequent VMIA.
3.2. Demographic data
Demographic data of these patients are shown in Table 1. Ninety-six eyes (76 patients) were included in the follow-up study. The mean age was 60.03 8.74 years, and there were 34 male and 42 female patients. The mean follow-up duration was 22.63 14.13 months. The number of patients with hypertension and hyperlip-idemia was 50 of 76 patients (65.79%) and 50 of 63 patients (78.9%), respectively, of those who had a recorded blood lipid profile. The number of eyes that had received a cataract operation, macular laser, or PRP prior to or during the follow-up period were 13 (14.8%), 19 (19.79%), or 52 (54.17%), respectively. The mean refrac-tive status was 0.46 2.44 diopters. The mean HbA1c level was 8.58 1.72%.
The mean baseline BCVA was 0.46 0.50 LogMAR units (Snellen equivalent: 6/17.3), and the mean final BCVA was 0.58 0.55 LogMAR units (Snellen equivalent: 6/22.8), but there was no statistically significant difference between the baseline and final BCVA values. The mean BCVA at baseline deteriorated to level of the final visit by þ0.11 0.56 LogMAR units. The baseline CRT was 246.88 75.66
m
m, and thefinal CRT was 256.13 73.58m
m. The mean CRT from the baseline to the final visit deterioratedbyþ9.25 81.65
m
m. There was also no statistically significant difference between the baseline andfinal CRT (p ¼ 0.270).3.3. Incidence and factors associated with VMIA formation
The basic characteristics of the eyes that developed and did not develop VMIA are shown inTable 2. Eight eyes (8.33%) developed VMIA during the follow-up period (Table 1), of which seven eyes were classified as ERM and one eye was classified as anomalous vitreomacular adhesion on OCT (Fig. 1). The calculated mean annual incidence of VMIA (rate of VMIA formation/follow-up duration) was 4.42% per year (Table 1). The mean duration from the diagnosis of DME to the detection of VMIA was 19.00 8.94 months.
We performed additional statistical analyses to determine the associations between various systemic and ocular factors and VMIA formation in both groups (Table 2). Univariate analysis revealed that only age was associated with the development of VMIA in our study (p ¼ 0.03). Stepwise logistic regression also revealed that older age was the only risk factor associated with VMIA formation (p ¼ 0.04; odds ratio ¼ 1.101; 95% confidence
Table 1 Demographic characteristics. Eyes (cases) 96 (76) Sex (male/female) 34/42 Age (y)a 60.03 8.74 Hypertension 50 (65.8%) Hyperlipidemia* 50 (79.37%) HbA1c (%)a 8.56 1.72
Baseline refraction (diopters)a 0.46 2.44
Baseline BCVA (LogMAR)a 0.46 0.50
Baseline CRT (mm)a 246.88 75.66
Macular grid laser 19 (19.79%)
PRP 52 (54.17%)
Prior cataract surgery Surgery 5 (5.21%)
Follow-up period (mo)a 22.63 14.13
Time from diagnosis of DME to ERM development (mo)a 19.00 8.94
Total incidence of VMIA 8 (8.33%)
Annual incidence of VMIA 4.42%
BCVA ¼ best-corrected visual acuity; CRT ¼ central retinal thickness; DME¼ diabetic macular edema; VMIA ¼ vitreomacular interface abnormality; PRP¼ pan-retinal photocoagulation.
*Total number of patients with a known lipid profile ¼ 63. aData are presented as the mean standard deviation.
Table 2
Basic characteristics of the eyes that developed and did not develop VMIA.
VMIA (-) VMIA (þ) p Eyes (cases) 88 (70) 8 (6) Male/Female 31/39 3/3 >0.99 Agea(y) 58.84 8.84 66.00 7.84 0.030 Hypertension (%) 50 (71.4%) 5 (83.3%) 0.531 Hyperlipidemia*(%) 50 (87.8%) 6 (100%) 0.363 HbA1ca(%) 8.63 1.49 7.87 1.29 0.167
Refractive statusa(diopters) 0.34 2.49 1.72 1.47 0.128 Baseline BCVA (LogMAR)a 0.47 0.52 0.41 0.25 0.734 Baseline CRTa(mm) 248.20 78.54 232.25 27.64 0.571 Macular grid laser (%) 16 (18.2%) 3 (37.5%) 0.405
PRP (%) 47 (53.4%) 5 (62.5%) 0.723
Cataract surgery (%) 11 (12.5%) 2 (25.0%) 0.295 Follow-up (mo)a 22.05 13.78 29.00 17.32 0.184 VMIA¼ vitreomacular interface abnormality; VMIA (þ) ¼ eyes with VMIA forma-tion; VMIA (-)¼ eyes without VMIA formation; BCVA ¼ best-corrected visual acuity; CRT¼ central retinal thickness; PRP ¼ pan-retinal photocoagulation.
Analyzed using the independent-samples t test: age, refractive status, HbA1c, BCVA, CRT; Fisher exact test: sex, hypertension, hyperlipidemia, macular laser, PRP and cataract surgery.
*Total number of patients with a known lipid profile ¼ 63. aData are presented as the mean standard deviation.
interval¼ 1.004e1.207). Other systemic and ocular factors, such as hypertension, hyperlipidemia, HbA1c, refractive status, and history of cataract operation, macular laser, and PRP, were not significantly associated with the formation of VMIA.
3.4. VMIA formation andfinal outcomes
Our study also determined that VMIA formation was signi fi-cantly associated with poor visual outcomes and thicker CRT at the final visit. Eyes that eventually developed VMIA demonstrated significant deterioration from the baseline to the final BCVA value (0.41 0.25 LogMAR units to 1.14 0.76 LogMAR units, p ¼ 0.034), while eyes without VMIA formation did not demonstrate this significant deterioration (0.47 0.52 LogMAR units to 0.52 0.51 LogMAR units, p¼ 0.30;Fig. 2A). The changes from the baseline to thefinal BCVA values were also significantly different (p < 0.001) between eyes with (deteriorated byþ0.73 0.79 LogMAR units) and without VMIA formation (deteriorated byþ0.06 0.50 Log-MAR units;Table 3). Similarly, the deterioration from the baseline to thefinal CRT level was statistically significant in eyes with VMIA formation (232.25 27.64
m
m to 355.63 81.66m
m, p¼ 0.012) but not in eyes without VMIA formation (248.20 78.54m
m to 247.08 66.20m
m, p¼ 0.883;Fig. 2B). Changes from the baseline to the final CRT level were also significantly different (p < 0.001) between eyes with (þ123.38 104.45m
m) and without VMIA formation (-1.13 71.38m
m;Table 3).4. Discussion
Our study investigated the incidence of the development of VMIA in patients with diabetic retinopathy over a relatively long period of time, which, to the best of our knowledge, has rarely been reported in the literature. In this study, eight (8.33%) of a total of 96 eyes developed VMIA over a mean follow-up period of 22.63 months, which corresponds to a mean annual incidence of 4.42% per year. The mean duration between of the development of VMIA and the first visit was 19.0 months. Of these eight eyes that developed VMIA, seven eyes (7.29%) were of the ERM configuration (Fig. 1A,B) and one (1.04%) demonstrated the features of anomalous vitreomacular adhesion (Fig. 1C,D). During the initial period of including patients in our study, 16 eyes were found to already have
been diagnosed with VMIA (16 of 244 eyes; 6.56%); therefore, 6.56% may represent the prevalence of VMIA in patients with DME who visit our retinal clinic.
We do not know if the calculated annual incidence of 4.42% per year for the occurrence of VMIA in our patients with DME is higher than that of the general population or not. As mentioned above, there are no reports on the incidence of VMIA in DME patients. A large survey on the general population, the Blue Mountains Eye Study, reported that the 5-year cumulative incidence of idiopathic ERM formation is 5.3% in thefirst eye and 13.5% in the second eye. ERM could be regarded as a specific type of VMIA that demon-strates the growth offibrous tissue. It seems that the incidence of VMIA (which predominantly presents as the ERM configuration) in the DME patients in our study is higher than that reported in the Blue Mountains Eye Study; however, because the study designs, methodologies, objectives, settings, and number of cases were all very different, it is difficult to directly compare various studies. Furthermore, our study used OCT in addition to ophthalmoscopic examination, fundus photographs, and FA to detect VMIA. OCT has been reported as more sensitive and superior for the detection of ERM and VMIA than traditional methods.6Studies that did not use OCT for the detection of VMIA may have underestimated its prev-alence. This is especially true for many large-scale epidemiological studies such as the Blue Mountains Eye Study that investigated the prevalence of ERM formation before the invention of OCT. As a result, it is still inconclusive whether or not the incidence of VMIA in patients with DME is higher than that of the general population. The prevalence of 6.56% for VMIA that was found during the initial visit of our 244 patients with DME is within the range of ERM reported by most epidemiological surveys of the general pop-ulation (2.2e18.5% according to the different methodologies used to study different racial and age groups in different studies).23e31A similar study performed by Thomas et al used OCT to detect VMIA, reporting a prevalence of only 10% for VMIA among 140 patients with DME who had been previously treated using macular laser therapy.9 They concluded that VMIA in DME might not be as prevalent as might be expected from the literature. Other studies, however, have found that the prevalence of ERM and VMIA seem to increase with the severity of DME.6,9,32,33 Ghazi et al specifically investigated cases of persistent and clinically significant DME that were refractory to macular grid laser treatment, reporting that the
Fig. 1. Optical coherence tomography (OCT) demonstrating changes in diabetic macular edema in representative cases. A, B) OCTfindings (A) before and (B) after the development of the epiretinal membrane (arrows). C, D) OCTfindings (C) before and (D) after the development of anomalous vitreomacular adhesion (arrows).
prevalences of ERM and VIMA are as high as 35.4% and 51.2%, respectively.33In our study, the mean initial CRT was only 246.88 um, which is much less than the mean CRT of the patients in the study performed by Ghazi et al (353.8 um). This might explain why our patients did not demonstrate a higher prevalence of VMIA on the initial visit.
Our study also investigated the possible risk factors associated with the development of VMIA in DME patients. In the literature, older age has been identified as the most consistent risk factor associated with the formation of idiopathic ERM in many large population-based studies.23e31Our statistical analysis also found that older age was a major risk factor associated with VMIA formation in our DME patients according to both univariate anal-ysis (p¼ 0.03;Table 2) and stepwise logistic regression (p¼ 0.04; odds ratio ¼ 1.101; 95% confidence interval: 1.004e1.207). This suggests that eyes with DME are associated with similar risk factor as those with idiopathic ERM.
In our study, although procedures such as cataract extraction surgery, macular laser, and PRP were associated with the devel-opment of VMIA, these differences did not reach a statistic level of significance (Table 2). In the literature, macular laser is not consistently associated with VMIA9,33; however, cataract surgery and thermal laser photocoagulation, such as PRP, are well-known risk factors associated with a high prevalence of VMIA for-mation.30,31,34e39It is possible that our study size was too small to have enough statistical power to detect categorized risk factors, such as the percentage of patients with a history of cataract surgery or PRP, for infrequent complications like VMIA. Other systemic and ocular factors, such as hypertension, hyperlipidemia, level of HbA1c, and baseline refractive status, were also investigated in this study and found to have no significant correlation with the devel-opment of VMIA (Table 2). These factors were also less consistently found to be associated with the formation of ERM in previous epidemiological reports.23e31
Our study also shows that the formation of VMIA in eyes with DME significantly affected the visual outcome and CRT at the final visit. Although the initial BCVA and CRT values were not signi fi-cantly different between the VMIA (þ) and VMIA (-) groups, the final BCVA and CRT values were significantly different. In eyes where no VMIA was noted, the BCVA and CRT values were not significantly different between the initial and final visits. However, in eyes that developed VMIA, the BCVA and CRT values both significantly deteriorated (both p < 0.001;Table 3andFig. 2). In fact, several patients with VMIA eventually received pars plana vitrec-tomy (PPV) or membrane peeling to treat the deterioration of visual function. It is unknown if the formation of VMIA was the cause of the deterioration of both BCVA and CRT or was a co-phenomenon in eyes that deteriorated over the course of the follow-up period.
In summary, our study reveals that our patients with DME demonstrated a cumulative incidence of 8.33% over a mean follow-up period of 22.63 months (calculated annual incidence: 4.42%), which was correlated with age. This result may contribute to better knowledge regarding the general clinical course of DME patients who are not receiving intravitreal injection of any medication. However, several limitations need to be noted. First, the retrospective design essentially precludes the strict control of all systemic and ocular variables. Second, we intentionally excluded all patients who had received vitreor-etinal surgery before or during the follow-up period, such as PPV or intravitreal injection of anti-VEGF or corticosteroids. Because the necessity of these procedures usually indicated active proliferative disease or severe macular edema, the exclusion of these cases might suggest that most patients in our study represent a population of patients with relatively less severe diabetic retinopathy. Therefore, caution should be taken when applying our result to patients with active diabetic retinopathy or severe DME. Also, the number of eyes enrolled in our study was probably insufficient for a statistical comparison of the incidence of infrequent complications such as VMIA or ERM formation. Further studies with prospective and large-scale designs are needed to verify these conclusions.
Fig. 2. A) Baseline best-corrected visual acuity (BCVA) andfinal BCVA in eyes with or without the formation of vitreomacular interface abnormalities (VMIA). B) Baseline central retinal thickness (CRT) andfinal CRT in eyes with or without VMIA formation. Both thefinal BCVA and CRT values were significantly worse than the initial BCVA and CRT values in the VMIA (þ) group. * p < 0.05 by the paired t test. VMIA (-) ¼ eyes without VMIA formation; VMIA (þ) ¼ eyes with VMIA formation.
Table 3
Comparison of eyes with and without VMIA formation in the non-IVB group.
VMIA (-) VMIA (þ) p
BCVA change þ0.06 0.50 þ0.73 0.79 <0.001 CRT change -1.13 71.38 þ123.38 104.45 <0.001 VMIA (þ) ¼ eyes with VMIA formation; VMIA (-) ¼ eyes without VMIA formation; BCVA¼ best-corrected visual acuity; CRT ¼ central retinal thickness.
Acknowledgments
This study was supported, in part, by a research grant (no. SKH-8302-101-DR-22) from Shin Kong Wu Ho-Su Memorial Hospital. The author thanks Chia-Yu Lin for the help collecting data. There are no proprietary or commercial interests related to this manuscript that need to be declared.
References
1. Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature 2001;414(6865):782e7.
2. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic Study of Diabetic Retinopathy, XV: the long-term incidence of macular edema. Ophthalmology 1995;102(1):7e16.
3. Photocoagulation for diabetic macular edema. Early Treatment Diabetic Reti-nopathy Study report number 1. Early Treatment Diabetic RetiReti-nopathy Study Research Group. Arch Ophthalmol 1985;103(12):1796e806.
4. Lee CM, Olk RJ. Modified grid laser photocoagulation for diffuse diabetic macular edema. Long-term visual results. Ophthalmology 1991;98(10):1594e602. 5. Kaiser PK, Riemann CD, Sears JE, Lewis H. Macular traction detachment and
diabetic macular edema associated with posterior hyaloidal traction. Am J Ophthalmol 2001;131(1):44e9.
6. Ghazi NG, Ciralsky JB, Shah SM, Campochiaro PA, Haller JA. Optical coherence tomographyfindings in persistent diabetic macular edema: the vitreomacular interface. Am J Ophthalmol 2007;144(5):747e54.
7. Kim NR, Kim YJ, Chin HS, Moon YS. Optical coherence tomographic patterns in diabetic macular oedema: prediction of visual outcome after focal laser photocoagulation. Br J Ophthalmol 2009;93(7):901e5.
8. Wilkins JR, Puliafito CA, Hee MR, Duker JS, Reichel E, Coker JG, et al. Charac-terization of epiretinal membranes using optical coherence tomography. Ophthalmology 1996;103(12):2142e51.
9. Thomas D, Bunce C, Moorman C, Laidlaw AH. Frequency and associations of a taut thickened posterior hyaloid, partial vitreomacular separation, and sub-retinalfluid in patients with diabetic macular edema. Retina (Philadelphia, PA) 2005;25(7):883e8.
10. Mirza RG, Johnson MW, Jampol LM. Optical coherence tomography use in evaluation of the vitreoretinal interface: a review. Surv Ophthalmol 2007;52(4):397e421.
11. Gallemore RP, Jumper JM, McCuen 2nd BW, Jaffe GJ, Postel EA, Toth CA. Diagnosis of vitreoretinal adhesions in macular disease with optical coherence tomography. Retina (Philadelphia, Pa) 2000;20(2):115e20.
12. Aiello LP. Angiogenic pathways in diabetic retinopathy. N Engl J Med 2005;353(8):839e41.
13. Aiello LP, Avery RL, Arrigg PG, Keyt BA, Jampel HD, Shah ST, et al. Vascular endothelial growth factor in ocularfluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med 1994;331(22):1480e7.
14. Lutty GA, McLeod DS, Merges C, Diggs A, Plouet J. Localization of vascular endothelial growth factor in human retina and choroid. Arch Ophthalmol 1996;114(8):971e7.
15. Nicholson BP, Schachat AP. A review of clinical trials of anti-VEGF agents for diabetic retinopathy. Graefe’s Archive for Clinical and Experimental Ophthal-mology 2010;248(7):915e30.
16. Michaelides M, Kaines A, Hamilton RD, Fraser-Bell S, Rajendram R, Quhill F, et al. A prospective randomized trial of intravitreal bevacizumab or laser therapy in the management of diabetic macular edema (BOLT study) 12-month data: report 2. Ophthalmology 2010;117(6):1078e86.
17. Michaelides M, Fraser-Bell S, Hamilton R, Kaines A, Egan C, Bunce C, et al. Macular perfusion determined by fundusfluorescein angiography at the 4-month time point in a prospective randomized trial of intravitreal bev-acizumab or laser therapy in the management of diabetic macular edema (Bolt Study). Retina (Philadelphia, PA) 2010;30(5):781e6.
18. Bandello F, Cunha-Vaz J, Chong NV, Lang GE, Massin P, Mitchell P, et al. New approaches for the treatment of diabetic macular oedema: recommendations by an expert panel. Eye (London, England) 2012;26(4):485e93.
19. Elman MJ, Aiello LP, Beck RW, Bressler NM, Bressler SB, Edwards AR, et al. Randomized trial evaluating ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology 2010;117(6):1064e77.
20. Elman MJ, Bressler NM, Qin H, Beck RW, Ferris 3rd FL, Friedman SM, et al. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology 2011;118(4):609e14.
21. Wu PC, Lai CH, Chen CL, Kuo CN. Optical coherence tomographic patterns in diabetic macula edema can predict the effects of intravitreal bevacizumab injection as primary treatment. J Ocul Pharmacol Ther 2012;28(1): 59e64.
22. Laidlaw DA. Vitrectomy for diabetic macular oedema. Eye (London, England) 2008;22(10):1337e41.
23. McCarty DJ, Mukesh BN, Chikani V, Wang JJ, Mitchell P, Taylor HR, et al. Prevalence and associations of epiretinal membranes in the visual impairment project. American Journal of Ophthalmology 2005;140(2):288e94.
24. Miyazaki M, Nakamura H, Kubo M, Kiyohara Y, Iida M, Ishibashi T, et al. Prevalence and risk factors for epiretinal membranes in a Japanese population: the Hisayama study. Graefe’s Archive for Clinical and Experimental Ophthal-mology 2003;241(8):642e6.
25. Kawasaki R, Wang JJ, Mitchell P, Aung T, Saw SM, Wong TY. Racial difference in the prevalence of epiretinal membrane between Caucasians and Asians. The British Journal of Ophthalmology 2008;92(10):1320e4.
26. Kawasaki R, Wang JJ, Sato H, Mitchell P, Kato T, Kawata S, et al. Prevalence and associations of epiretinal membranes in an adult Japanese population: the Funagata study. Eye (London, England) 2009;23(5):1045e51.
27. You Q, Xu L, Jonas JB. Prevalence and associations of epiretinal membranes in adult Chinese: the Beijing eye study. Eye (London, England) 2008;22(7): 874e9.
28. Duan XR, Liang YB, Friedman DS, Sun LP, Wei WB, Wang JJ, et al. Prevalence and associations of epiretinal membranes in a rural Chinese adult population: the Handan Eye Study. Invest Ophthalmol Vis Sci 2009;50(5):2018e23. 29. Klein R, Klein BE, Wang Q, Moss SE. The epidemiology of epiretinal membranes.
Trans Am Ophthalmol Soc 1994;92:403e25. discussion 25e30.
30. Mitchell P, Smith W, Chey T, Wang JJ, Chang A. Prevalence and associations of epiretinal membranes: The Blue Mountains Eye Study, Australia. Ophthal-mology 1997;104(6):1033e40.
31. Fraser-Bell S, Ying-Lai M, Klein R, Varma R. Prevalence and associations of epiretinal membranes in latinos: the Los Angeles Latino Eye Study. Invest Ophthalmol Vis Sci 2004;45(6):1732e6.
32. Ophir A, Martinez MR, Mosqueda P, Trevino A. Vitreous traction and epiretinal membranes in diabetic macular oedema using spectral-domain optical coherence tomography. Eye (London, England) 2010;32(10):1545e53. 33. Kim BY, Smith SD, Kaiser PK. Optical coherence tomographic patterns of
dia-betic macular edema. American Journal of Ophthalmology 2006;142(3):405e12. 34. Mester U, Volker B, Kroll P, Berg P. Complications of prophylactic argon laser treatment of retinal breaks and degenerations in 2,000 eyes. Ophthalmic Surg 1988;19(7):482e4.
35. Kraushar MF, Morse PH. The relationship between retina surgery and pre-retinal macularfibrosis. Ophthalmic Surgery 1988;19(12):843e8.
36. Battaglia Parodi M, Iacono P, Di Crecchio L, Sanguinetti G, Ravalico G. Clinical and angiographic features in nasal branch retinal vein occlusion. Oph-thalmologica 2004;218(3):210e3.
37. Hayashi K, Hayashi H. Influence of phacoemulsification surgery on progression of idiopathic epiretinal membrane. Eye (London, England) 2009;23(4):774e9. 38. Fraser-Bell S, Guzowski M, Rochtchina E, Wang JJ, Mitchell P. Five-year
cumulative incidence and progression of epiretinal membranes: the Blue Mountains Eye Study. Ophthalmology 2003;110(1):34e40.
39. Jahn CE, Minich V, Moldaschel S, Stahl B, Jedelhauser P, Kremer G, et al. Epi-retinal membranes after extracapsular cataract surgery (1). J Cataract Refract Surg 2001;27(5):753e60.
