Article Text

Download PDFPDF

Persistent outer retinal fluid following non-posturing surgery for idiopathic macular hole
  1. Rubina Rahman1,
  2. Lucy Oxley1,
  3. John Stephenson2
  1. 1Department of Ophthalmology, Calderdale Royal Hospital, Halifax, UK
  2. 2School of Human and Health Sciences, University of Huddersfield, Huddersfield, UK
  1. Correspondence to Rubina Rahman, Department of Ophthalmology, Calderdale Royal Hospital, Salterhebble Hill, Halifax, HX3 0NJ UK; Rubina.Rahman{at}


Aim To present the anatomical and visual outcomes of patients with hypo-reflective cystic defects in outer fovea (outer foveal defect; OFD) in macular holes repaired with non-posturing vitrectomy and short term gas tamponade. To identify the incidence and risk factors for developing OFD foveal defect.

Method A prospective consecutive case series of 58 patients undergoing macular hole surgery was undertaken. OFD detected on 2-week postoperative Optical coherence tomography (OCT) was measured. In these patients OCT was performed monthly until resolution of OCT.

Results 27 eyes (46.6%) had an outer defect at 2 weeks, the presence of which was significantly associated with macular holes with larger base diameters preoperatively (p=0.006). All defects closed spontaneously without further intervention, and the final vision was not affected by the presence of an OFD. Visual recovery was only slightly (and not significantly) delayed by the presence of an outer defect.

Conclusions This is the first study of outcomes of OFDs following macular hole surgery in patients who did not posture postoperatively. OFDs are common but do not adversely affect visual outcomes.

  • Vitreous
  • Treatment Surgery
  • Retina

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Macular hole repair with vitrectomy and peeling of internal limiting membrane is reported to achieve successful hole closure in 90%–98% of cases.1–3 Clinical hole closure may be different to anatomical hole closure, especially in the era of increasingly higher definition imaging of the retinal architecture. Optical coherence tomography (OCT) has evolved from time domain to spectral domain providing as little as 5 μm of axial resolution of retinal anatomy, thus improving the visualisation of hole architecture. This has led researchers to demonstrate five different retinal abnormalities after flat closed macular holes.4 These include outer foveal hypo-reflective defects, persistent foveal detachment, moderately reflective foveal lesions, epiretinal membranes and nerve fibre layer defects.

An outer foveal defect (OFD) following macular hole surgery is one in which the hole clinically appears closed. However, on OCT evaluation, there is a hypo-reflective cystic space in the outer fovea representing persistent outer retinal fluid (figure 1). The incidence is reported to range from 34% to 49% in previous reports5–8 in relation to posturing macular hole surgery. We aim to determine whether lack of posturing postmacular hole surgery influences the incidence and visual recovery as well as final vision in these patients.

Figure 1

Outer retinal fluid defect after sealing of sensory retinal closure 2 weeks after macular hole repair.


A prospective consecutive case series of 58 eyes from 58 patients undergoing surgery for idiopathic macular hole was performed. Cases were operated by one surgeon (RR) between May 2009 and November 2011. Only patients in whom the macular hole was closed postoperatively were included.

Preoperative data included age, gender and best-corrected visual acuity (BCVA). BCVA refers to pinhole vision, apart from the final vision which was recorded with new distance prescription. All patients underwent OCT evaluation of macular hole base diameter and minimal linear diameter (MLD). OCT scanning was performed using spectral domain SD-OCT (Optovue RTVue-100 with V.4.0 software, Freemont, California, USA). Twelve radial 6 mm OCT images (MM6/Radial slice 0.27 s) through the centre of the macular hole were obtained by skilled operators.

All patients underwent 23-gauge transconjunctival pars plana vitrectomy, internal limiting membrane peel with Brilliant Peel (Fluoron, Geuder Germany) and endotamponade with sulfur hexafluoride (20% SF6). Drainage of fluid through the macular hole was not attempted in any of the study patients. In all, 52 patients were phakic and underwent combined microincision coaxial phacoemulsification with intraocular lens implant at the time of vitrectomy. The remaining six patients were pseudophakic on presentation after surgery. Patients were not given any positioning instructions.

As SF6 had completely resolved in 2 weeks, visual acuity and OCT scans were taken at 2 weeks postoperatively, and the maximum horizontal diameter of any OFD was measured using callipers on OCT. Patients noted to have an OFD at 2 weeks underwent serial monthly OCT scanning measuring the maximum horizontal diameter of the OFD until complete resolution of the was observed. The preoperative and postoperative BCVA was measured using a standard Snellen acuity, and converted to LogMAR scores for the purposes of statistical analysis. Patients were followed up for a minimum of 3 months or until their best corrected vision was stable on two consecutive visits.

Statistical analyses using SPSS software (V.18.0) were undertaken on a data set of 58 patients, of whom 27 developed an OFD and 31 did not. Patients were discharged after resolution of the outer retinal fluid and improvement/stabilisation of vision on two consecutive visits.

A logistic regression analysis was undertaken to assess the effect of macular hole base diameter, MLD and demographic factors on the risk of developing a foveal defect. Due to the limited numbers of patients with comorbidities/complications of any kind, these data were discretised to compare those with and without either age-related macular degeneration or retinal detachment, without distinguishing between these conditions.

A Kaplan–Meier non-parametric time-to-event analysis was performed on the data to compare times with recovery of vision in patients with and without an OFD, with comparisons facilitated both graphically and using the log-rank statistic.

An analysis of covariance (ANCOVA)-type regression analysis was also undertaken to investigate the effect of various factors and covariates on change in vision following surgery assessed using the linear LogMAR scale. In addition to the baseline vision level, variables considered included macular hole base diameter, MLD, age, comorbidities/complications (defined as in the logistic regression analysis outlined above) and the size of any OFD if present. The extent of the correlation between baseline vision and macular hole base diameter was also assessed.


The 58 patients comprised 37 women (63.8%) and 21 men (36.2%) spanning an age range of 47–86 years (mean 70.4 years).

Effect of macular hole base diameter and demographic factors on risk of developing defect

A total of 27 patients (46.6%) had developed an OFD of whom 17 were women and 10 were men. OFDs observed ranged in size up to 754 μm (mean 309 μm) after 2 weeks and 829 μm (mean 205 µm) after 6 weeks.

The means and SD of age, preoperative base diameters and MLD of macular holes are shown in table 1.

Table 1

Summary of key descriptive statistics

Overall, 38 patients (65.5%) had MLD values of 400 µm or below. Among patients with lower MLD values, 18 (47.4%) had an OFD. Among patients with higher MLD values, 9 (45.0%) had an OFD. The association between MLD values (dichotomised as above) and OFD formation was non-significant (p=0.864).

The logistic regression analysis found base diameter, MLD and age to be significantly associated with the probability of OFD (p=0.006 for base diameter; p=0.043 for MLD; p=0.032 for age), with increased odds of OFD development in patients with larger macular hole base diameters and with smaller MLD measurements, and in younger patients (table 2). Comorbidities/complications were not found to be significantly associated with the probability of defect formation (p=0.362). Postoperative complications included one case of postoperative inferior rhegmatogenous retinal detachment with proliferative vitreo-retinopathy and three cases of posterior capsular opacification.

Table 2

p Values, ORs and associated CIs for factors/covariates potentially influencing development of foveal defect

Effect of developing defect on time to recovery of vision

All patients recovered vision by 11 months, with mean time to recovery of 2.98 months. Mean time for vision recovery ranged from 1.5 to 10 months (mean 3.74 months) among patients who experienced OFD formation. Mean time for vision recovery ranged from 1.5 to 11 months (mean 2.03 months) among those patients who did not experience OFD formation.

Hence mean time to vision recovery was greater in patients who experienced OFD formation. This difference in vision recovery experience was of near-borderline significance (χ2(1)=3.72; p=0.054) using the log-rank statistic derived from the Kaplan–Meier analysis.

Figure 2 illustrates distinct experience of vision recovery in ‘with defect’ and ‘without defect’ groups, indicative of a substantive difference between the two groups, with the proportion of patients with recovered vision being as great or greater among patients with no OFD than in patients with an OFD at all times up to 11 months. Despite the generally more rapid improvements in vision seen in patients without OFD, vision in all patients with OFDs was recovered 1 month earlier (at 10 months) than in patients without OFDs, all of whom had recovered vision by 11 months.

Figure 2

Times to vision recovery in patients with and without foveal defects.

Effect of ophthalmic and demographic factors on extent of recovered vision

The final vision on the LogMAR scale ranged from 0.00 to 1.00, with a mean score of 0.30. Change in vision from preoperative readings on the LogMAR scale ranged from a deterioration of 0.40 to an improvement of 1.38, with a mean improvement of 0.47. Patients with and without OFDs showed almost identical mean improvements in vision.

The ANCOVA model found baseline vision to be significantly associated with the final recovered vision (p=0.001), controlling for other variables. Other demographic and ophthalmic parameters were not found to be significantly associated with the final recovered vision. No excessive correlations between predictor variables were observed.

Unstandardised and standardised coefficients of all parameters considered, p values and associated CIs are given in table 3. In this analysis, macular hole base diameter, minimum linear diameter and OFD size are measured in millimetres.

Table 3

Regression coefficients, p values and associated 95% CIs for analysis of extent of improved vision in patients


Previously it has been shown that excellent success rates can be achieved with non-posturing macular hole surgery using shorter acting gas.3 Rapid absorption of the gas bubble allows quicker visual rehabilitation. However, whether or not the combination of shorter acting gas and non-posturing surgery increases the risk of developing outer defects after macular hole surgery remains to be determined. This study of 58 patients with non-posturing surgery for idiopathic macular hole treated with vitrectomy, internal limiting membrane peel and SF6 gas found the incidence of OFD on OCT at 2 weeks to be 46.6%.

The incidence is comparable with other studies in which patients postured for 5–10 days post macular hole surgery.4–6 Kawano et al5 report an incidence of 49% at 1 month following combined phaco-vitrectomy and SF6 gas. The incidence in the study of Kang et al6 was 45.8%. The main risk factor for developing an OFD in our study was the preoperative base diameter of macular hole, with larger holes more likely to have an outer defect at 2 weeks (p=0.006). Controlling for other factors, MLD (p=0.043) and age (p=0.032) were also found to be significantly associated with OFD development, with increased odds of OFD development in patients with smaller MLD measurements, and in younger patients. Two earlier studies have shown that outer defects are more likely to occur in patients who have MLD preoperatively.5 ,6 It was postulated by Kang et al6 that this is because of the way in which glial cells commence proliferation in the healing process of the macular hole. No other theories have been put forward. The foveal defects did all gradually shrink and resolve over time, in agreement with previous studies.5 ,6 Our study is the first to measure the outer defects using SD-OCT; however, the size of defect measured after 6 weeks was not found to be associated with final vision (table 3) controlling for other variables. It may also be shown that defect size after 6 weeks did not correlate with preoperative base diameter.

The final vision was no different in patients who had a defect compared with those without. Final vision ranged from 0.0 to 1.0 with a mean of 0.3 LogMAR. The mean improvement of 0.47 was almost identical in both groups. Kawano et al and Kang et al also report no effect of a defect on the final visual outcome, achieving mean visions of 0.13 and 0.426, respectively.5 ,6 Kang et al6 actually found that their visual outcomes were better in the group who had a foveal defect, but the authors acknowledge that the difference in preoperative size of macular hole may explain this. The only factor found to influence final vision in our study was the preoperative level of vision.

Visual recovery was taken as the time to reach final visual acuity, and in all cases this occurred by 11 months. The presence of a defect did not significantly delay visual recovery in this sample (p=0.054), although there was only one patient in the without-defect group who took longer than 5 months for vision to stabilise (11 months). Without this patient (who had age-related macular degeneration), the difference would have been statistically significant. Including this patient in the statistical analysis, the mean time for vision to stabilise was 2.03 months in the without-defect group, and 3.74 months in the with-defect group, and it can be seen in the figure that the trend was for a more rapid recovery in patients without a defect. Previous studies have not commented on the time taken for visual recovery, so there is no corroborating evidence.

However, it follows that a defect in the photoreceptor layer which is gradually shrinking to reach resolution is likely to delay the recovery of vision.

In summary, OFDs are fairly common after macular hole surgery and do not adversely affect the final visual outcome. We found that they were more likely in patients with larger preoperative macular holes in contrast to other studies. The key difference between our study and previous ones is the lack of posturing. Non-posturing surgery has similar outcomes to posturing surgery, and there is no difference in the incidence of outer defects or in the final vision when patients do not posture. The presence of an OFD did not influence the final vision; only preoperative vision affected this. Finally, the presence of a defect has a borderline but not significant effect on the time taken for visual recovery, although larger studies would help to determine this effect.



  • Contributors RR: Conception and design of study, acquisition and analysis of data, revising the article critically and approval of the version to be published. LO: Collection and analysis of data, drafting and revising the article and approval of the published version. JS: Statistical analysis of the data, revising the article and final approval of the published version.

  • Competing interests None.

  • Ethics approval Local research and development (R&D) department.

  • Provenance and peer review Not commissioned; externally peer reviewed.