Mean standard deviations for common glaucoma treatments.

In this case, macular hole was not observed preoperatively, as the preoperative OCT shows (except for a small subfoveal serous elevation), or intraoperatively. Although the exact timing of hole formation could not be established in this case, neither intraoperative tugging nor postoperative premacular fibrosis are likely to represent the mechanism for hole formation. Rather, pre-existing macular oedema, which can develop in a diabetes patient during the early postvitrectomy period, may have resulted in macular hole formation. The possibility of spontaneous closure cannot be excluded, as has been noted in some previous reports (Kokame & McCauley 2002; Lo & Hubbard 2006). However, it is unlikely to have occurred in our case because the macular hole persisted at 1 week prior to IVTA. Consistent with the theory that macular oedema may be an inciting factor that modifies foveal anatomy, leading to hole formation, the fact that the hole closed 2 weeks following IVTA in our case strongly supports the suggestion that IVTA plays a role in macular hole closure. A recent report on IVTA proposed that the treatment encouraged macular hole formation by causing thinning of the inner retina (LecleireCollet et al. 2007). However, Halkiadakis et al. (2003) demonstrated the closure of a full-thickness macular hole associated with uveitic macular oedema after peribulbar triamcinolone injection. In conclusion, inflammation and macular oedema might be causative factors in macular hole formation in patients with diabetic retinopathy. The present case indicates that the treatment of macular oedema with IVTA may lead to macular hole closure. References