Factors predicting outcome of vitrectomy for diabetic macular oedema: results of a prospective study.

AIM: To determine preoperative demographic, clinical, and optical coherence tomography (OCT) factors which might predict the visual and anatomical outcome at 1 year in patients undergoing vitrectomy and inner limiting membrane peel for diabetic macular oedema (DMO). METHODS: A prospective, interventional case series of 33 patients who completed 1 year follow up. Measurements were taken preoperatively and at 1 year. Outcome measures were logMAR visual acuity (VA) and OCT macular thickness. A priori explanatory variables included baseline presence of clinical and/or OCT signs suggesting macular traction, grade of diabetic maculopathy, posterior vitreous detachment, fluorescein leakage and ischaemia on angiography, presence of subretinal fluid, and peroperative indocyanine green (ICG) use. RESULTS: 33 patients completed 1 year follow up. On average VA deteriorated by 0.035 logMAR (p = 0.40). Macular thickness significantly improved by a mean of 139 microm (95% CI; 211 to 67, p<0.001). Patients with evidence of clinical and/or OCT macular traction significantly improved logMAR acuity (logMAR improvement = 0.08) compared with patients without traction (logMAR deterioration 0.11, p = 0.01). Presence of subretinal fluid significantly predicted worse postoperative result (p = 0.01) CONCLUSION: On average, patients showed a statistically significant improvement in central macular thickness following treatment but a marginal acuity worsening. Presence of subretinal fluid on OCT is hypothesised to be exudative rather than tractional in nature. The visual benefit of vitrectomy for DMO in this study was limited to patients who exhibit signs of macular traction either clinically and/or on OCT.