BackgroundTo investigate the incidence and clinical characteristics of unexplained visual loss in patients with fovea-sparing rhegmatogenous retinal detachment (RRD) during or after silicone oil (SO) tamponade.MethodsThe medical charts of all patients with macula-on RRDs, who underwent pars-plana-vitrectomy (ppV) with SO tamponade were retrospectively assessed regarding unexplained visual loss (UVL) of >= 3 Snellen lines and alterations on optical coherence tomography (OCT) during or after SO tamponade. The clinical data analysed included visual acuity, surgical parameters, OCT images, duration of SO tamponade and the time point of visual decline. Cases with re-detachment or secondary causes of visual loss such as SO emulsification, epiretinal membranes or macular edema were excluded.ResultsOver a 15-year-period, 22 cases with macula-on RRD, which had primarily been treated with ppV and SO tamponade, met the inclusion criteria. In most eyes (n = 20; 91%), the RRD was caused by a giant retinal tear (GRT). In 11 of these 22 cases (50%), best-corrected visual acuity (BCVA) had dropped by at least 3 lines for no apparent reason. In these 11 cases, mean preoperative logMAR BCVA was 0.2 (SD 0.13; range 0-0.5), equal to Snellen's VA of 0.63, and mean postoperative logMAR BCVA 1.0 (SD 0.24; range 0.5-1.3), equal to Snellen's VA of 0.10. Visual decline occurred about 12 weeks postoperatively (SD 6.2; range 3-20 ) and comprised 8 lines (SD 2.3; range -11 to -4). SO was removed on average 139 (SD 50.0; range 88-271) days after the first ppV. In 9 cases visual decline occurred while the SO was in-situ. In 2 patients, BCVA decline was noted 2 weeks after SO removal. In all eyes, preoperative central foveal thickness (CFT) was 254 mu m (SD 24.2), which decreased to 224 mu m (SD 29.6) during SO tamponade and increased to 247 mu m (SD 29.2) after SO removal, irrespective of the presence of UVL. The mean follow-up time was 20 months (SD 30.6) after SO removal.ConclusionUVL after SO tamponade for macula-on RRD is more frequent than expected. The incidence in our case series was 50%. The mechanism of this phenomenon is still unknown. In general, vitreoretinal surgeons should thoroughly question the need for SO tamponade, inform their patients of possible UVL and remove SO as early as possible.