Ophthalmic Surgery, Lasers and Imaging Retina

Clinical Science 

Outer Retinal Defects Represent a Normal Recovery Pathway Following Internal Limiting Membrane Peeling in Macular Hole Surgery

Matthew A. Powers, MD, MBA; Ryan A. Shields, MD; Andrew A. Moshfeghi, MD, MBA; Darius M. Moshfeghi, MD

Abstract

BACKGROUND AND OBJECTIVE:

To examine perioperative factors associated with the development of outer retinal defects (ORDs) following surgical repair of macular holes (MHs).

PATIENTS AND METHODS:

An institutional review board-approved, retrospective, interventional cohort study was conducted. Patients who underwent MH repair during a 5-year period were identified. Statistical analysis was conducted to detect significant perioperative associations to ORD development.

RESULTS:

One hundred twenty-four eyes were included, and 54% developed an ORD following surgery. These defects correlated with lower preoperative stage (P = .0057), preoperative phakia (P = .036), and lack of prior macular surgery (P = .0016). Patients in the ORD group had significantly better preoperative and postoperative visual acuity (P = .031 and P = .0004, respectively), but there was no difference in change in acuity from preoperatively to 3 months postoperatively when compared with control patients (P = 42). The majority (89%) of ORDs resolved by 24 months postoperatively.

CONCLUSION:

The development of ORDs appears to be correlated with several factors indicative of favorable overall eye health and less advanced pathology and may represent a normal state of recovery after MH repair with internal limiting membrane peeling.

[Ophthalmic Surg Lasers Imaging Retina. 2018;49:e1–e8.]

Abstract

BACKGROUND AND OBJECTIVE:

To examine perioperative factors associated with the development of outer retinal defects (ORDs) following surgical repair of macular holes (MHs).

PATIENTS AND METHODS:

An institutional review board-approved, retrospective, interventional cohort study was conducted. Patients who underwent MH repair during a 5-year period were identified. Statistical analysis was conducted to detect significant perioperative associations to ORD development.

RESULTS:

One hundred twenty-four eyes were included, and 54% developed an ORD following surgery. These defects correlated with lower preoperative stage (P = .0057), preoperative phakia (P = .036), and lack of prior macular surgery (P = .0016). Patients in the ORD group had significantly better preoperative and postoperative visual acuity (P = .031 and P = .0004, respectively), but there was no difference in change in acuity from preoperatively to 3 months postoperatively when compared with control patients (P = 42). The majority (89%) of ORDs resolved by 24 months postoperatively.

CONCLUSION:

The development of ORDs appears to be correlated with several factors indicative of favorable overall eye health and less advanced pathology and may represent a normal state of recovery after MH repair with internal limiting membrane peeling.

[Ophthalmic Surg Lasers Imaging Retina. 2018;49:e1–e8.]

Introduction

Pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peeling for full-thickness macular holes (FTMHs) is a highly successful procedure with modern closure rates of more than 90%.1,2 Although not ultimately visual significant, the presence of postoperative signal voids (Figure 1) on optical coherence tomography (OCT), also known as outer retinal foveal hypofluorescence, or outer retinal defects (ORDs), following macular hole closure has been frequently described.3–12

Spectral-domain optical coherence tomography demonstrating outer retinal defect following macular hole repair.

Figure.

Spectral-domain optical coherence tomography demonstrating outer retinal defect following macular hole repair.

Various mechanisms have been hypothesized, including persistent subfoveal fluid, inner retinal bridge formation, decreased motility of the outer retina, and disinsertion of photoreceptors from the retinal pigment epithelium (RPE).5,8,10 Recent studies have demonstrated that small holes (ie, stage II), vitreomacular traction, and ILM peeling are all associated with the development of ORDs.3,12 Therefore, to help further elucidate perioperative factors associated with postoperative ORDs, we present a retrospective analysis of all patients having undergone FTMH surgical repair at a single institution for the past 5 years.

Patients and Methods

A retrospective chart review was conducted of all patients presenting to the vitreoretinal service at a single institution for the past 5 years. Institutional review board approval was obtained prior to the study and was performed in compliance with the Health Insurance Portability and Accountability Act. Using CPT, ICD-9, and ICD-10 codes, patients with a diagnosis of FTMH who underwent PPV were identified and included in the study. Patients records were reviewed to identify gender, date of birth, age at surgery, and date of surgery. Perioperative variables were also collected, including laterality, macular hole (MH) stage,13 preoperative best-corrected visual acuity (BCVA), postoperative BCVA up to the third postoperative month, lens status at time of surgery, prior surgical repair attempts, use of indocyanine green (ICG) for ILM staining, use and type of tamponading agent, intraoperative use of autologous serum, postoperative complications, presence and severity of ORDs following surgery within the first postoperative 6 weeks on OCT, and days between surgery and the first OCT of sufficient quality to identify this defect.

Initial severity of ORDs, when present, was stratified on a four-point scale based on height in comparison to adjacent retinal thickness: maximum height less than 25% of adjacent retina (1), between 25% and 50% (2), between 50% and 100% (3), and greater than 100% of surrounding retinal thickness (4). Patients were excluded from the study for incomplete or ambiguous data, poor quality or absent imaging during the first postoperative 6 weeks, postoperative failure of the MH to close, or any severe postoperative complication, including endophthalmitis, retinal vascular occlusion, or retinal detachment. Patients undergoing vitrectomy for additional indications concurrent to the MH repair were also excluded.

Patient data were compiled and analyzed for perioperative factors significantly associated with the presence of ORDs using statistical software (SAS JMP version 13; SAS, Cary, NC). Snellen visual acuity was converted logMAR for statistical analysis using nonparametric testing, given non-normality of the data. Univariate testing, including Student's t-test, Wilcoxon test, Kruskal-Wallis test, analysis of variance, Pearson chi-square, Fisher's exact test, and logistic regression, were conducted. Multivariate testing, including ordinal logistic regression and nominal logistic regression, were also conducted to detect significant associations. Statistical significance was defined as alpha less than .05.

Results

Included Eyes

A total of 205 eyes were identified that underwent attempted surgical repair of a FTMH. Twenty eyes were excluded due to failure of surgical closure. Fifty-one eyes were excluded due to absent or inadequate OCT imaging. Eight eyes were excluded due to concurrent surgical indications including presence of rhegmatogenous retinal detachment, tractional retinal detachment, or choroidal rupture. Two eyes were excluded for postoperative complications, including one case of endophthalmitis and a central retinal artery occlusion.

Baseline Characteristics

One hundred twenty-four eyes were included in the study. Seventy-seven (62.1%) patients were female. Mean (SD) age at the time of surgery was 66.43 (9.71) years old. Sixty-six (53.2%) of the eyes were left eyes. Median (interquartile range [IQR]) preoperative BCVA was 0.6 (0.48–1) logMAR (Snellen equivalent 20/79.6), and median (IQR) postoperative BCVA at 3 months was 0.4 (0.3–0.54) logMAR (Snellen equivalent 20/50.2). Median (IQR) change in visual acuity from preoperative assessment to 3 months postoperatively was 0.24 (0.1–0.4). Seventy-nine (63.7%) eyes were phakic at the time of surgery. Fourteen (11.3%) eyes had history of prior attempted MH surgery. One hundred seventeen (94.3%) eyes had ICG used intraoperatively. Forty-six (37.1%) eyes had perfluoropropane (C3F8) used as a tamponading agent, 74 (59.7%) had sulfur hexafluoride (SF6) used, three (2.4%) had silicone used, and 1 (0.8%) had sterile air used. Seventeen (13.7%) eyes had autologous serum used intraoperatively (Table 1).

Baseline Patient Characteristics

Table 1:

Baseline Patient Characteristics

Postoperative Outer Retinal Defects and Resolution

Fifty-seven eyes (45.9%) had no evidence of outer retinal OCT signal void during the first 6 weeks postoperatively, whereas 67 (54%) eyes had evidence of some type of signal void. Average (SD) time from surgery to detection of ORDs on OCT was 16.39 (13.64) days. Of eyes with an initial ORD with imaging available (n = 52), 24 (46%) had resolution of the defect by 6 months postoperatively. At postoperative month 12, 34 (74%) patients with ORD with available imaging (n = 46) experienced resolution, and by 24 months, 89% of eyes with available imaging (n = 45) experienced resolution.

Visual Outcomes

In the group of eyes that did not experience postoperative ORDs, median (IQR) preoperative BCVA was 0.7 (0.54–1) logMAR. In comparison, eyes that went on to develop ORDs had a median (IQR) BCVA of 0.6 (0.4–0.88) logMAR (P = .031, Wilcoxon test). At postoperative month 3, median (IQR) BCVA of eyes that did not develop ORDs was 0.48 (0.3–0.6) logMAR, whereas median (IQR) BCVA of eyes that did was 0.4 (0.18–0.48) logMAR (P = .0004, Wilcoxon test). However, median (IQR) change in BCVA in the non-ORD eyes was 0.24 (0.67 to 0.49) logMAR and in the ORD eyes was 0.3 (0.15 to 0.4) logMAR (P = .42, Wilcoxon test), which remained non-significant even when controlling for preoperative phakic status (P = .64, nominal logistic).

Preoperative and Intraoperative Associations

Univariate testing was conducted to detect significant associations between preoperative and intraoperative factors and development of ORDs. Significant associations were detected between development of ORDs and lower-stage holes (P = .0057, Pearson's chi-square), preoperative phakia (P = .036, Fisher's exact test), lack of prior MH surgery (P = .0016, Fisher's exact test), and use of SF6 as tamponade agent over C3F8 (P = .0001, nominal logistic). No significant difference was detected between patient gender, age at time of surgery, surgical eye laterality, use of ICG, or use of autologous serum (Table 2).

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Nominal/Ordinal Variables)

Table 2A:

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Nominal/Ordinal Variables)

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Continuous Variables)

Table 2B:

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Continuous Variables)

Univariate analysis was also conducted to detect significant perioperative associations with ORDs stratified by severity, and detected significant associations between increasing severity of postoperative ORDs and decreasing preoperative hole stage (P = .0008, Pearson's chi-square), better preoperative visual acuity (P = .012, ordinal logistic), better postoperative visual acuity (P < .0001, ordinal logistic), lack of prior MH surgery (P = .024, Pearson's Chi-square), and type of tamponade agent used (P = .0033, Pearson's Chi-square). No significant difference was detected between patient gender, age at time of surgery, surgical eye laterality, phakic status, use of ICG, or use of autologous serum (Table 3). There was a trend toward better preoperative and postoperative BCVA in eyes with worsening ORDs (P = .075 and .0011 for pre- and postoperative BCVA, respectively, Kruskal-Wallis).

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

Table 3:

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

Time to first usable OCT differed significantly between the groups. Mean (SD) time to first usable OCT, and thus diagnosis in the ORD group, was 25.67 (19.1) days compared with 16.39 (13.64) days in non-ORD eyes (P = .006, t-test). This difference was due primarily to presence or absence of an obscuring gas bubble. Eyes in which SF6, a more rapidly resolving agent, was used had a mean (SD) time to first usable OCT of 13.23 (10) days, whereas time to first usable OCT in eyes in which C3F8 was used was 34.15 (17.82) days. This difference was statistically significant (P < .0001, t-test).

Multivariate testing via nominal logistic regression was also conducted. When viewing ORDs as a dichotomous variable (ie, not stratified by severity), no variables met statistical significance, but there was a trend toward an association with better preoperative BCVA (P = .09), lack of prior MH surgery (P = .06), preoperative phakia (P = .1), and use of SF6 gas over C3F8 (P = .1) (Table 4). When stratifying for severity of ORDs, ordinal logistic regression revealed significant associations between larger defects and better preoperative BCVA (P = .01), preoperative phakia (P = .01), and lack of prior MH surgery (P = .04) (Table 5).

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs

Table 4:

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

Table 5:

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

Median (IQR) time to resolution of ORDs was 5 months (range: 3 months to 12 months). Survival analysis was conducted to compare time to resolution between perioperative variables, but no significant differences were detected.

Discussion

Modern MH surgery is in general quite successful, with high rates of postoperative hole closure. The incidence of ORDs is also high, occurring in about half of cases according to prior reports,3,4,8,10,12,14,15 a proportion confirmed here. Previously, ORDs have been regarded as a pathologic change, yet we contend that they are likely part of the normal recovery process following MH repair with ILM peeling. This is based on several findings of the current and prior studies.

First, the relatively high incidence of ORDs does not seem to have an impact on postoperative visual outcomes, and although these defects do appear to delay visual recovery, there is no difference in final visual outcome.3,10,14 Similarly, we found no difference in change from baseline visual acuity between groups. Prior studies also indicated that the vast majority of ORDs resolve over time, with resolution after 7 months in 73%12 and resolution after 12 months in 93%.8 Our study demonstrated a resolution rate of 74% by 12 months and 89% at 24 months. Although the vast majority of ORDs resolved, it is unclear why the resolution rate was slightly lower than previous reports. The authors speculate this was related to incomplete OCT records at 12 months postoperatively.

Second, in the present study, there were consistent associations between ORDs and metrics associated with healthier eyes, as evidenced by significant associations to better pre- and postoperative visual acuity, lower-stage holes, and lack of prior macular hole or cataract surgery. Tranos et al. also found a correlation between smaller diameter holes and the incidence of ORDs.12 Reports have also indicated, which we have confirmed here, that there is ultimately no difference in visual acuity change between normal and ORD eyes.8,15 All of this information put together indicates that these defects hold minimal negative prognostic value.

The mechanism by which ORDs develop remains uncertain. Some have hypothesized that because the inner retina tends to close first, the persistence of outer retinal discontinuity mimics the appearance of subfoveal fluid.10 Other theories include postoperative RPE dysfunction, persistent tissue defects from a prior operculum, abnormal interdigitation of the photoreceptors and RPE, and decreased mobility of the outer retina during hole closure.8 Ehlers et al. demonstrated that an intraoperative increase in ellipsoid zone-RPE height and subretinal hyporeflectivity width was negatively correlated with the development of ORDs postoperatively.3 The authors hypothesized that these measurements were indicative of anterior retinal traction, which tends to loosen the retina from the RPE, increasing retinal mobility and allowing for more expedited hole closure. Peeling of the ILM may induce this effect and facilitate closure, and may ultimately be the reason for the relative success of this technique over vitrectomy alone.1,2 The small number of patients who did not undergo ILM peeling in the current study precluded adequate analyses of an association with ORDs.

The correlation with better overall eye health suggests that postoperative ORDs may be part of the normal recovery process following surgery. This may be related to the change in mobility of the retina over the RPE hypothesized above. The non-pathologic nature of ORDs would also explain why they have not been found to impact final visual acuity. Less healthy eyes with later-stage holes, worse vision, and previous surgery may not possess the proper ability to produce an ORD during hole closure, although this does not appear to ultimately affect closure rate in the literature.

Of note, we found a strong association between the use of SF6 gas and development of postoperative ORDs. Additionally, we found a strong association between the earliest available OCT and type of tamponade agent used. Namely, OCT imaging was available earlier in patients with SF6 gas used. This is likely due to the relatively rapid reabsorption time of SF6 as compared with C3F8. Patients in whom air or silicone oil was used also had earlier availability of OCT, although the small number of patients precluded adequate statistical analysis. Thus it may be the case that when C3F8 gas was used, ORDs resolved or lessened in severity before the gas bubble had resolved enough to utilize OCT. When postoperative OCT timing was included in multivariate analyses, the statistical significance of tamponade agent disappeared. Several comparisons have demonstrated relative equivalency between SF6 and C3F8 in MH repair, and although C3F8 may result in more frequent intraocular pressure spikes and hastened cataract formation,16 they have more or less been found to be equally effective and safe.17,18 It is certainly possible that SF6 has an effect on retinal or RPE cells, which modulate ORD presence or closure, but this has not been previously documented in the literature.

Conversely, ICG has indeed been shown to have toxic effects on both retinal and RPE cells.19 Unfortunately, the widespread use of ICG in macular hole surgery makes it difficult to draw statistical conclusions, even with large sample sizes. In the present study, nearly 95% of cases involved ICG use. And although there was a small increase in the likelihood of ORD development with ICG use on univariate analysis, this effect was not statistically significant and was not seen in the multivariate model. Moreover, prior research has demonstrated that focal macular electroretinogram (ERG) responses do not differ between ICG-assisted, brilliant blue G-assisted, and triamcinolone-assisted macular hole surgery, arguing against differential toxicity from these agents.20 In the present study, no patients underwent postoperative focal macular ERG, and thus we cannot draw any conclusions about retinal function between groups beyond the BCVA.

The present study has a few limitations to note. First, it is a retrospective study, and as such, there is variation in dates and quality of clinical data. Second, patient follow-up was highly variable, which made extracting relevant data many months out from surgery more difficult. This led to a high percentage of patients being excluded from analysis. And as mentioned above, OCT imaging and timing was not standardized between groups, and thus prone to fluctuations in the severity and true resolution of ORDs between groups. This study is, to our knowledge, the largest such study conducted on ORDs following MH surgery, but recruitment of additional patients would nonetheless help strengthen real associations and eliminate confounders. Future prospective studies will be needed to better gauge the impact of above and other surgical factors in this phenomenon.

Ultimately, ORDs after surgery for FTMHs largely resolve and do not affect final visual acuity. The present study revealed associations between several perioperative factors associated with healthy eyes and the development of ORDs, including lower stage at presentation, lack of prior cataract or MH surgery, better preoperative visual acuity, and — perhaps spuriously — use of SF6 gas as a tamponade agent. These associations suggest that ORDs are more likely to occur in healthier eyes without prior surgical history and may represent a normal state of recovery during MH closure.

References

  1. Lois N, Burr J, Norrie J, et al. Internal limiting membrane peeling versus no peeling for idiopathic full-thickness macular hole: A pragmatic randomized controlled trial. Invest Ophthalmol Vis Sci. 2011;52(3):1586–1592. doi:10.1167/iovs.10-6287 [CrossRef]
  2. Spiteri Cornish K, Lois N, Scott NW, et al. Vitrectomy with internal limiting membrane peeling versus no peeling for idiopathic full-thickness macular hole. Ophthalmology. 2014;121(3):649–655. doi:10.1016/j.ophtha.2013.10.020 [CrossRef]
  3. Ehlers JP, Itoh Y, Xu LT, Kaiser PK, Singh RP, Srivastava SK. Factors associated with persistent subfoveal fluid and complete macular hole closure in the PIONEER study. Invest Ophthalmol Vis Sci. 2014;56(2):1141–1146. doi:10.1167/iovs.14-15765 [CrossRef]
  4. Kawano H, Uemura A, Sakamoto T. Incidence of outer foveal defect after macular hole surgery. Am J Ophthalmol. 2011;151(2):318–322. doi:10.1016/j.ajo.2010.08.034 [CrossRef]
  5. Ehlers JP, Xu D, Kaiser PK, Singh RP, Srivastava SK. Intrasurgical dynamics of macular hole surgery: An assessment of surgery-induced ultrastructural alterations with intraoperative optical coherence tomography. Retina. 2014;34(2):213–221. doi:10.1097/IAE.0b013e318297daf3 [CrossRef]
  6. Bottoni F, De Angelis S, Luccarelli S, Cigada M, Staurenghi G. The dynamic healing process of idiopathic macular holes after surgical repair: a spectral-domain optical coherence tomography study. Invest Ophthalmol Vis Sci. 2011;52(7):4439–4446. doi:10.1167/iovs.10-6732 [CrossRef]
  7. Hasler PW, Prunte C. Early foveal recovery after macular hole surgery. Br J Ophthalmol. 2008;92(5):645–649. doi:10.1136/bjo.2007.130971 [CrossRef]
  8. Christensen UC, Krøyer K, Sander B, Larsen M, la Cour M. Prognostic significance of delayed structural recovery after macular hole surgery. Ophthalmology. 2009;116(12):2430–2436. doi:10.1016/j.ophtha.2009.06.001 [CrossRef]
  9. Ko TH. Ultrahigh-resolution optical coherence tomography of surgically closed macular holes. Arch Ophthalmol. 2006;124(6):827–836. doi:10.1001/archopht.124.6.827 [CrossRef]
  10. Takahashi H, Kishi S. Tomographic features of early macular hole closure after vitreous surgery. Am J Ophthalmol. 2000;130(2):192–196. doi:10.1016/S0002-9394(00)00456-6 [CrossRef]
  11. Moshfeghi AA, Flynn HW Jr, Elner SG, Puliafito CA, Gass JDM. Persistent outer retinal defect after successful macular hole repair. Am J Ophthalmol. 2005;139(1):183–184. doi:10.1016/j.ajo.2004.06.082 [CrossRef]
  12. Tranos PG, Stavrakas P, Vakalis AN, Asteriadis S, Lokovitis E, Konstas AGP. Persistent subretinal fluid after successful full-thickness macular hole surgery: Prognostic factors, morphological features and implications on functional recovery. Adv Ther. 2015;32(7):705–714. doi:10.1007/s12325-015-0227-z [CrossRef]
  13. Johnson RN, Gass JD. Idiopathic macular holes. Observations, stages of formation, and implications for surgical intervention. Ophthalmology. 1988;95(7):917–924. doi:10.1016/S0161-6420(88)33075-7 [CrossRef]
  14. Herbert EN, Sheth HG, Wickremasinghe S, Luthert PJ, Bainbridge J, Gregor ZJ. Nature of subretinal fluid in patients undergoing vitrectomy for macular hole: A cytopathological and optical coherence tomography study. Clin Exp Ophthalmol. 2008;36(9):812–816. doi:10.1111/j.1442-9071.2008.01901.x [CrossRef]
  15. Sano M, Shimoda Y, Hashimoto H, Kishi S. Restored photoreceptor outer segment and visual recovery after macular hole closure. Am J Ophthalmol. 2009;147(2):313–318.e1. doi:10.1016/j.ajo.2008.08.002 [CrossRef]
  16. Modi A, Giridhar A, Gopalakrishnan M. Sulfurhexafluoride (SF6) versus perfluoropropane (C3F8) gas as tamponade in macular hole surgery. Retina. 2017;37(2):283–290. doi:10.1097/IAE.0000000000001124 [CrossRef]
  17. Casini G, Loiudice P, De Cillà S, Radice P, Nardi M. Sulfur hexafluoride (SF6) versus perfluoropropane (C3F8) tamponade and short term face-down position for macular hole repair: A randomized prospective study. Int J Retina Vitreous. 2016;2:10. doi:10.1186/s40942-016-0036-9 [CrossRef]
  18. Kim SS, Smiddy WE, Feuer WJ, Shi W. Outcomes of sulfur hexafluoride (SF6) versus perfluoropropane (C3F8) gas tamponade for macular hole surgery. Retina. 2008;28(10):1408–1415. doi:10.1097/IAE.0b013e3181885009 [CrossRef]
  19. Narayanan R, Kenney MC, Kamjoo S, et al. Toxicity of indocyanine green (ICG) in combination with light on retinal pigment epithelial cells and neurosensory retinal cells. Curr Eye Res. 2005;30(6):471–478. doi:10.1080/02713680590959312 [CrossRef]
  20. Machida S, Nishimura T, Ohzeki T, Murai KI, Kurosaka D. Comparisons of focal macular electroretinograms after indocyanine green-, brilliant blue G-, or triamcinolone acetonide-assisted macular hole surgery. Graefes Arch Clin Exp Ophthalmol. 2017;255(3):485–492. doi:10.1007/s00417-016-3478-8 [CrossRef]

Baseline Patient Characteristics

Gender N (%) Lens Status at Time of Surgery N (%)
  Male 47 (37.9%)   Phakic 79 (63.7%)
  Female 77 (62.1%)   Pseudophakic 45 (36.3%)

Eye N (%) Previous Macular Surgery N (%)
  Left 66 (53.2%)   Yes 14 (11.3%)
  Right 58 (46.8%)   No 110 (88.7%)

Age at Time of Surgery Mean (SD) ICG Used N (%)
  Years 66.43 (9.71)   Yes 117 (94.3%)
  No 7 (5.6%)

Stage N (%) Tamponade Agent N (%)
  2 18 (15.7%)   C3F8 46 (37.1%)
  3 40 (34.8%)   SF6 74 (59.7%)
  4 57 (49.6%)   Silicone Oil 3 (2.4%)
  Air 1 (0.8%)

BCVA (logMAR) Median (IQR) Autologous serum N (%)
  Preoperative 0.6 (0.48–1)   Yes 17 (13.7%)
  POM3 0.4 (0.3–0.54)   No 107 (86.3%)
  Difference 0.24 (0.1–0.4)

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Nominal/Ordinal Variables)

Odds Ratio (95% CI) P Value

Gender (Female) 1.39 (0.67–2.88) .36b

Eye (Left) 1.19 (0.59–2.42) .72b

Preoperative Stage .0057c
  2 versus 3 4.1 (1.02–16.4) .047c
  2 versus 4 7.39 (1.29–28.45) .0036c
3 versus 4 1.8 (0.8–4.09) .16c

Phakic at Time of Surgery 2.12 (1–4.46) .036a

Lack of Prior Macular Surgery 8.66 (1.85–40.6) .0016a

Use of ICG 1.61 (0.34–7.51) .07b

Tamponade Agent Used .0006d
  SF6 vs C3F8 4.76 (2.15–10.54) .0001d

Use of Autologous Serum 0.41 (0.14–1.19) .12b

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

X2 Value (DF)a P Valuea
Gender 1.69 (4) .8
Eye 2.45 (4) .65
Preoperative Stage 26.57 (8) .0008
Lens Status at Time of Surgery 7.9 (4) .09
Lack of Prior Macular Surgery 11.26 (4) .024
Use of ICG 2.25 (4) .69
Tamponade Agent Used 19.67 (12) .0033
Use of Autologous Serum 3.84 (4) .43

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs

Odds Ratio (95% CI) P Valuea

Gender (Female) 1.66 (0.59–4.66) .38

Age 1.03 (0.98–1.1) 017

Eye (Left) 1.04 (0.41–2.68) .83

Preoperative Stage .3
  2 versus 3 3.23 (0.72–14.57)
  2 versus 4 2.38 (0.51–11.19)
  3 versus 4 0.74 (0.23–2.31)

Preoperative BCVA 0.38 (0.11–1.23) .09

Phakia at Time of Surgery 2.04 (0.73–5.7) .10

Lack of Prior Macular Surgery 4.45 (0.79–24.9) .06

Use of ICG 1.13 (0.19–6.8) .80

Time to First Postoperative OCT 0.99 (0.96–1.03) .72

Tamponade Agent Used 0.27
SF6 vs C3F8 2.5 (0.8–8.19) .1

Multivariate Perioperative Risk Factors for Development of Postoperative ORDs (Stratified by Initial Size of ORD)

X2 (DF) P Valuea
Gender 0.84 (1) .36
Age 3.89 (1) .05
Eye 0.17 (1) .68
Preoperative Stage 3.42 (2) .13
Preoperative BCVA 5.55 (1) .01
Lens Status at Time of Surgery 6.07 (1) .01
Lack of Prior Macular Surgery 4.08 (1) .04
Use of ICG 0.43 (1) .63
Time to First Postoperative OCT 0.17 (1) .56
Tamponade Agent Used 2.33 (2) .17

Univariate Perioperative Risk Factors for Development of Postoperative ORDs (Continuous Variables)

No ORD ORD P Value

Age (Years) 65 (11.56) 67.64 (7.68) .15b
Mean (SD)

Preoperative BCVA 0.7 (0.54–1) 0.6 (0.4–0.88) .031c
Median (IQR)

Postoperative BCVA 0.48 (0.3–0.6) 0.4 (0.18–0.48) .0004c
Median (IQR)

Difference in BCVA 0.24 (0.67–0.49) 0.3 (0.15–0.4) .42c
Median (IQR)

Days to First Postoperative OCT 25.67 (19.1) 16.39 (13.64) .006a
Mean (SD)
Authors

From Byers Eye Institute at Stanford University School of Medicine, Palo Alto, CA (MP, RS, DM); and Gayle and Edward Roski Eye Institute at The University of Southern California, Los Angeles (AM).

Drs. Powers, Shields, and Darius Moshfeghi have received grant support from Research to Prevent Blindness through an unrestricted departmental grant. This sponsor has had no role in the study design, collection, analysis and interpretation of data, writing the report, or the decision to submit the report for publication. Dr. Andrew Moshfeghi reports no relevant financial disclosures.

Address correspondence to Matthew A. Powers, MD, MBA, Byers Eye Institute, Department of Ophthalmology, Stanford University School of Medicine, 2452 Watson Court, Palo Alto, CA 94303; email: powers.matthew.a@gmail.com.

Received: September 05, 2017
Accepted: December 04, 2017

10.3928/23258160-20180907-01

Sign up to receive

Journal E-contents