Comparison of the effect of air tamponade versus no tamponade after pars plana vitrectomy for idiopathic epiretinal membrane

This study aimed to compare the surgical outcomes of pars plana vitrectomy (PPV) with and without air tamponade in patients with idiopathic epiretinal membrane (iERM). We prospectively enrolled 145 patients with iERM who underwent a 25-gauge transconjunctival sutureless PPV. Patients were assigned to either the air tamponade (air) group (79 eyes) or balanced salt solution (BSS; no tamponade) group (66 eyes). The central macular thickness (CMT), peripapillary retinal nerve fiber layer (pRNFL) thickness, and best-corrected visual acuity (BCVA) were compared for two years. At baseline, there were no significant differences between the two groups. CMT and BCVA were not significantly different between the groups for 2 years. However, the air group had a significantly lower thickness in the superior temporal pRNFL sector at 1 month (p = 0.01) and in the inferior temporal and superior temporal pRNFL sectors at 3 months (p = 0.02 for both). There were no significant differences between both groups in all the pRNFL sectors from 6 months to 2 years. The outcomes of PPV with air tamponade and that with no tamponade appear to be equivalent. This shows that air tamponade may not be an imperative procedure for iERM surgery and has no additional benefit.


Results
A total of 145 eyes were included in the study, with 79 eyes in the air group and 66 eyes in the BSS group. At baseline, there was no significant difference in age, sex, BCVA, IOP, CMT, or global pRNFL thickness between the two groups (Table 1). In both groups, there was no case of postoperative hypotony, vitreous hemorrhag, or endophthalmitis. Additionally, there was no case of reoperation due to postoperative complications, such as retinal detachment and recurrence of ERM.

Visual outcomes and central macular thickness after epiretinal membrane removal with and without air tamponade. BCVA and IOP were not significantly different between the two groups for
Temporal changes of best-corrected visual acuity and central macular thickness after epiretinal membrane removal with and without air tamponade. The preoperative and postoperative BCVA and CMT significantly differed. BCVA was improved from 0.36 ± 0.26 logMAR to 0.15 ± 0.11 logMAR in the air group (p < 0.001) and from 0.36 ± 0.27 logMAR to 0.17 ± 0.14 logMAR in the BSS group (p < 0.001) ( Table 2). CMT was decreased from 399.23 ± 80.12 μm to 355.32 ± 51.38 μm in the air group (p < 0.001) and from 388.82 ± 69.65 μm to 361.89 ± 43.49 μm in the BSS group (p = 0.021) (Fig. 1).
pRNFL thickness after epiretinal membrane removal with and without air tamponade. There was no significant difference in the global pRNFL at 2 years after surgery (Table 4; Fig. 2). However, there was a significant difference in the TS sector at the 1-month follow-up. The air group (139.89 ± 26.20 μm) had a significantly lower thickness than the BSS group (153.13 ± 21.28 μm) (p = 0.013). At 3 months after surgery, the air Temporal changes in pRNFL thickness after epiretinal membrane removal with and without air tamponade. The pRNFL values in both the air and BSS groups significantly increased in all areas, except for the T sector of the air group at 1 month after surgery, and then gradually decreased at 1 year ( Table 5). The thickness of the T and TS sectors of both groups showed a significant decrease at 1 year after surgery compared with that at baseline. The thickness of the TI sector of the air group also showed a significant decrease at 1 year after surgery compared with that at baseline, whereas the decrease in the TI sector of the BSS group was not significant (p = 0.086) at 1 year after surgery. At the end, in the second year after surgery, there was a significant difference only in the T sector of both groups. The T sector of the BSS group decreased from 101.73 ± 21.43 μm preoperatively to 75.94 ± 7.58 μm (p = 0.001), and that in the air group decreased from 108.62 ± 23.82 μm preoperatively to 83.00 ± 13.92 μm (p < 0.001). In both groups, the nasal side (N, NI, and NS sector) returned to its preoperative value within 6 months and did not show a significant difference when compared with the preoperative values.
Baseline values affecting final global pRNFL thickness and final central macular thickness. Final global pRNFL thickness had a significant relationship with preoperative global pRNFL thickness, whereas other factors including age, sex, SE, IOP, AXL, and the presence or absence of air tamponade were not significantly related (Table 6). Final CMT showed a significant association with preoperative CMT, but lacked a significant relationship with other factors (Table 6).

Discussion
In this study, we compared the use of air tamponade and no tamponade after PPV surgery for iERM. We found that every parameter, BCVA, CMT, and pRNFL thickness, improved gradually throughout 2 years postoperatively in both groups. Moreover, there was no significant difference in each parameter assessed between the two groups. Previously, two studies compared the use of air and BSS tamponade after vitreoretinal surgery 18,19 . One of the studies compared the difference in visual acuity between the two groups after lamellar macular hole surgery 19 , and the other compared the visual acuity and foveal contour changes between the two groups after iERM surgery 18 . However, prior to this report, there was no study comparing the difference in visual acuity, macular thickness, and pRNFL thickness between these two groups after iERM surgery with a long-term follow-up (2 years). In particular, this is the first study to compare pRNFL between two groups after iERM surgery.
Postoperative visual acuity and CMT were not significantly different between the two groups, but pRNFL showed a temporary difference after the surgery. At 1 month after surgery, the air group showed lower thickness than the BSS group in the TS segment, and at 3 months, the air group showed lower thickness in the TI segment as well as in the TS segment. This difference could be attributed to the compression effect of air tamponade. However, it is noteworthy that the difference in the pRNFL thickness between the two groups was significant up to 3 months, even though air tamponade in the vitreous cavity generally lasts for less than 1 month. Although Table 3. Comparison of the macular thickness of the BSS and air groups over 2 years. Values are presented as n or mean ± standard deviation. BSS balanced salt solution; Air air tamponade. p values were derived by independent t-tests.  18 compared the visual acuity, foveal surface symmetry, area-matched thickness, and area-matched contour of the air and BSS groups after ERM surgery. In that study, visual acuity, area-matched thickness, and area-matched contour did not differ between the two groups, but the foveal surface symmetry, comparing the shape of the nasal side of the macula and the temporal side of the macula with the fovea, was significantly higher in the air group up to 3 months postoperatively. These results are also one of the reasons for the inference that the tamponade effect of air can last up to 3 months. However, they only analyzed changes for up to 3 months, whereas we analyzed results up to 2 years and found that no significant difference in the pRNFL between the two groups from 3 months to 2 years. When comparing the preoperative to the postoperative 1-month pRNFL values for each group, the BSS group had an increase in thickness in all areas, while the air group had an increase in thickness in all areas with the exception of the temporal sector. For both groups, after the 1-month measurements, the thickness slowly decreased and by the 2-year time point, there was no significant difference from the preoperative values in all sectors, with the exception of the pRNFL thickness of the temporal sector that was significantly decreased. These observations are in line with previous studies. In the study by Balducci et al. 20 , 30 patients underwent vitrectomy, www.nature.com/scientificreports/ and the pRNFL thickness was increased in all areas, except the temporal sector, at 1 month; these then gradually returned to preoperative values. At 6 months, only the thickness of the temporal area was significantly reduced in their study. Lee et al. 21 also reported a significant decrease in pRNFL thickness at the temporal sector in 31 patients with ERM at 3, 6, and 12 months postoperatively. In our study, the thickness of the nasal side (N, NI, and NS sector) was restored to the preoperative value within 6 months in both groups and did not show a significant change even 2 years after the surgery. Lee et al. 21 also reported no change in the nasal side at 1, 3, 6 and 12 months after surgery. These results may be due to the fact that the nasal side was hardly affected by ERM, and the effect of air tamponade also did not seem to have a significant effect on the nasal side of pRNFL. A transient increase in the thickness of the pRNFL, which occurs 1 month after such surgery, may originate from the inflammatory response caused by the vitrectomy surgery itself. Additionally, according to the previous reports, direct mechanical damage to the optic head by the procedure such as ILM peeling, or inducing PVD with ocutome, or retinal toxicity due to ICG dye could be the causes 17,20,22,23 . The thickness of the temporal area gradually decreases, which may be due to the resolution of the traction force generated by ERM after surgery 21 or due to damage to the inner retinal layer during ILM peeling 20 .
Visual acuity improved significantly in both groups postoperatively. Leitritz et al. 18 reported no difference in visual acuity between the two groups during the short (3 month) follow-up period after vitrectomy due to iERM. From a practitioner's point of view, tamponade in the eye may cause lower vision and slower vision recovery immediately after surgery compared to no tamponade. However, our results with air tamponade showed that the visual acuity was identical between the two groups at 1 month postoperatively. In this study, there was also no significant difference in visual acuity between the two groups at all time points during the 2 years of follow-up.
This study has some limitations. First, the lack of randomization makes it entirely possible that there are baseline differences in disease severity between the two groups. Second, we enrolled a relatively small number of patients. To overcome this limitation, we refined our study design: we performed phacoemulsification and intraocular lens insertion on all 132 phakic eyes to exclude the effect of medial opacity on visual acuity according to the time change; the ILM peeling on every patient was performed by a single surgeon using the same technique; we excluded all patients with outpatient clinic or intraoperative finding of no PVD to remove or minimize the effect of PVD induction procedure on the pRNFL. Finally, we did not perform a visual field test. If we performed visual field examination, we would be able to obtain more information on how the change in pRNFL affects the visual field. Despite these limitations, this study is meaningful because of the long-term follow-up of 2 years, and it is the first report to prospectively compare visual acuity, macular thickness, and pRNFL between an air tamponade group and a no tamponade group in iERM surgery.
In conclusion, our study showed that there was no difference in postoperative visual acuity, macular thickness, and pRNFL thickness when air tamponade versus non-tamponade was performed during iERM surgery. This demonstrates that air tamponade might not be an imperative procedure for iERM surgery and has no additional benefit.

Methods
Subjects. This study was a prospective cohort study. Enrolled patients underwent surgery at Hangil Eye Hospital between January 2015 and January 2018. The study included only patients with confirmed iERM found using the Spectralis domain optical coherence tomography system (SD-OCT, Spectralis Version 5.6.1.0, Heidelberg Engineering, Heidelberg, Germany). Patients with other diseases that might affect pRNFL thickness and macular thickness, such as glaucoma, diabetic retinopathy, retinal vein occlusion, and age-related macular degeneration were excluded. Additionally, patients that did not present with posterior vitreous detachment (PVD) either intraoperatively, or in the outpatient clinic were excluded from the study. . ERM was removed under 2% lidocaine sub-Tenon anesthesia with monitored anesthesia care. Cataract surgery was also performed with vitrectomy in all phakic patients. After core and peripheral vitrectomy were performed, epiretinal membrane removal and ILM peeling were carefully performed with retinal forceps. Triamcinolone agent (MaQaid, Wakamoto Pharmaceutical Co., Ltd., Tokyo, Japan) was used for membrane peeling in some cases. If either an atrophic hole or lattice degeneration was present, air endo-tamponade was performed. The air tamponade was a full fill. These patients were then recommended to maintain a prone position for 1-day post operation. The surgery was completed after confirmation of the absence of any remaining membrane and macular hole using the intraoperative OCT (RESCAN 700, Zeiss, Germany).
Ethics statement. This study was approved by the Institutional Review Board of Hangil Eye Hospital and was carried out in accordance with the tenets of the Declaration of Helsinki; informed consent was obtained from all the patients.
Statistical analysis. Statistical analyses were performed using a commercially available software package (SPSS Statistics version 23; IBM Corp., Armonk, NY, USA). Independent t-test and Pearson's chi-square test were performed to compare the characteristics of the two patient groups: air tamponade (air group) and balanced salt solution (no tamponade; BSS group). Paired t-tests were used to compare pRNFL thickness before and after surgery in the two groups. A multiple regression model was used to evaluate whether there were preoperative factors affecting final global pRNFL thickness and final CMT. Age, sex, SE, IOP, AXL, the presence or absence of air tamponade, preoperative global pRNFL and preoperative CMT values were corrected. p values < 0.05 were considered significant.