To assess differences in anatomical and functional outcomes when using C2F6, SF6 or C3F8 gas tamponade in macular hole surgery with pars plana vitrectomy, internal limiting membrane peeling and a non-strict postoperative posturing regimen.

Setting: Gloucestershire Hospitals NHS Trust, UK.


Clinical and surgical data were prospectively entered into an electronic medical record (Medisoft, Leeds, UK). Data of consecutive macular hole surgeries operated between 2005 and 2012 using 20G PPV from 3 consultant-lead teams were included in this study. A regimen of prone posturing for the first night and then no posturing except to avoid sleeping supine for 1 week was adopted by all surgeons. Selection of the gas tamponade was mainly based on surgeon’s discretion. Data collection included: demographic details, tamponade type, stage of macular hole, preoperative and postoperative visual acuity (VA), anatomical success rate by OCT and secondary cataract surgery rate.


299 eyes of 276 patients with a mean age of 69.3±8.2 years (mean±SD) and a 1:2.5 male to female ratio were included. The type of gas employed was C2F6 in 45.8% (137/299), SF6 in 18.1% (54/299) and C3F8 in 36.1% (108/299). No differences were observed in macular hole stage and type of gas employed at baseline. Mean preoperative BCVA was significantly worse in eyes treated with C3F8 compared to SF6 and C2F6 (0.94±0.37 LogMAR vs 0.82±0.28 vs 0.84±0.28, p:0.034). No significant differences in postoperative BCVA were observed between gases at any time point but mean BCVA change from baseline was significantly better with C3F8 than SF6 and C3F8 at 12 months (-0.51±0.38 LogMAR vs -0.43±0.30 vs -0.34±0.31, p:0.037). Closure rate assessed by OCT at 1 month was 95.5% for C3F8, 97.4% for SF6 and 92.3% for C2F6. However, closed holes with persistent foveal detachment on OCT 1 month after surgery were significantly less frequent in C3F8 eyes compared to SF6 or C2F6 (9.1% vs 37.8% vs 38.1%, p:0.056, p:0.036, respectively). At 3rd postoperative month, no significant OCT differences were observed between gases. Cataract surgery was performed at 11.61±6.1 months in C2F6 group, 9.62±5.5 in SF6 group and 11.54±7.19 in C3F8 group, with no significant differences between groups.


In our series, C3F8 was used in poorer vision eyes, provided better anatomical outcome at 1st month as assessed by OCT and showed significantly higher vision gain at 12 months than SF6 and C2F6.

Financial Interest: Robert Johnston is the medical director of ‘Medisoft’ the electronic medical record used in this study. No other author has any financial interest.